i. Three Annotated Models of Research Thesis Statements

APPENDIX

i. Three Annotated Models of Research Thesis Statements

Writing Research Thesis Statements

Annotated Models: Example of the Problem of Transportation in Panama PROBLEM: Drivers in developing urban areas such as Panama City experience a great deal of traffic congestion, delays and accidents. QUESTIONS: 1) Advancing existing research: Where in Panama City are the most pressing congestion problems? Transportation problems, including congestion, involve a spatial dimension, for which there now exist a large body of literature and set of methodological tools available to researchers. Drawing contributions from geographers, planners, engineers, computer scientists, and more, specialized geographic technologies are currently available to analyze interesting environmental, economic and social patterns through modeling (Horner 2003). Increasingly, the dimension of time has been recognized as a critical element for advancing research, leading to the development of multidimensional transportation location referencing systems (Koncz and Adams 2002) that can enable dynamic modeling of congestion and other problems. Still, the field is plagued by the difficulties of adequately representing the many elements of data needed for addressing pressing and complex questions in empirical settings. Goodchild (2004) notes that a major challenge for further progress is the development of solutions that permit the co-existence of multiple, conflicting representations and analysis at multiple scales.

The first sentence states the problem to open the introduction to existing research. The second sentence introduces the existing body of research literature.

The third sentence focuses on and summarizes a specific part of the existing research literature.

The next sentence shows the limits of existing research, indicating a possibility of where it can be advanced. The last sentence supports the need for the proposed research to advance or make progress within existing research literature.

2) Resolving a contradiction: How would installing traffic signals impact public safety? As urban population and related development expands in the Panama City area, traffic volume and congestion increases, along with accident rates. Transportation research indicates that the installation of traffic signals at intersections can effectively resolve congestion and

The first sentence expresses the problem. The second sentence presents one side of the relevant research.

improve flow of vehicles (Strickland and Berman 1995). However, other research shows that intersection crash rates frequently increase with signal installation (Kuciemba and Cirillo 1992). The literature reveals that signalization usually leads to a shift in crash types, with fewer angle and turning collisions and more rear-end collisions, with some indications that crashes at signalized intersections may be less severe (NCSA 1999). The apparent contradiction between improving traffic flow and reducing accident rates suggest that transportation problems need to be studied in situ in order to design effective solutions that optimize overall public safety.

The third sentence presents a conflicting side of the relevant research.

The fourth sentence further explains the contradiction within the research literature. The final sentence shows the need for the proposed research based on the contradictions in existing research.

3) Developing a new line of inquiry: What unique cultural driving behaviors contribute to either causing or avoiding accidents? A fertile and multidisciplinary body of research focuses on the role of driver behavior within traffic incidents (Edsall and Driscoll 2004). Such research has been dominated by lines of inquiry that expose high-risk behavior such as driver fatigue, driver distractions, and driver impairment (low visibility, alcohol, medications, etc.). However, almost all such studies have occurred within highly economically developed areas such as in North America, Europe, and Australia. A few possible exceptions include comparative studies looking at drivers who cross borders such as at the US-Mexico and Finland-Russia interfaces (Lajunen 2001). What current transportation research does exist on developing countries focuses on infrastructure (TRB 2004). As visitors to cities such as Santiago, Bangkok, or Ouagadougou would attest, an obvious question arises: what role do the particular driving cultures in developing countries play in urban traffic accidents? There remains a dearth of research addressing the potential relationship between accident rates and culturally specific driving behaviors in urban areas of developing countries.

The first sentence introduces the existing body of research as a whole. The second sentence describes and summarizes the various lines of inquiry within existing research. The third sentence reveals the limitations of these lines of inquiry. The following sentence carefully notes a few possible exceptions to the case that the proposed line of inquiry is completely new. The next sentence shows how these exceptions do not quite follow the proposed new line of inquiry. The following sentence expresses the proposed new line of inquiry as a question. The final sentence explicitly states the need for a new line of inquiry as proposed.

APPENDIX

ii. Elwood, Sarah. NSF CAREER Award: Transforming the Politics of Place: GIS, Knowledge Production, and Community-Based Organizations in Urban Governance.

Project Summary This project examines the urban spatial and political impacts of two related developments in US cities: the growing adoption and use of geographic information systems (GIS) by communitybased organizations, and the expanding responsibilities being assumed by these organizations in urban planning, revitalization, and service delivery. The primary research objective of the project is to carry out a long-term ethnographic study of the production of GIS-based data and maps that analyze neighborhood needs, resources, and conditions; and the application of this spatial knowledge within local political processes that shape urban space at neighborhood level. The educational objective of the project is to implement service learning components into two undergraduate geography courses in which students will provide technical assistance for the community-based GIS analysis being studied as part of the project’s research component. These objectives are to be accomplished through an annual cycle of integrated research and educational activities carried out in collaboration with two community development organizations in a distressed inner city neighborhood in Chicago. Directed by staff and residents from the case study organizations, the project will create a spatial data library for use in their neighborhood revitalization activities. Students in the geography courses will work in collaboration with the organizations to create GIS-based spatial analysis projects using these data. The central research questions of the project will be examined using information gathered through participant observation of student-community interaction in educational activities; ethnographic interviewing of participating organization staff and neighborhood residents, as well as local government officials with whom the organization works; and analysis of maps and reports produced for the community-based project. With respect to its intellectual merits, this project will advance theories in urban geography and societal studies of GIS, by fostering a greater understanding of the role that spatial analysis technologies play in the construction of space and place, and conceptualizing the changing nature of democratic practice within local level political processes through which urban spatial change is negotiated. The project will advance knowledge about how urban spaces are constituted and transformed in the contemporary US city, through its examination of the interdependency of spatial analysis technologies, GIS-based knowledge construction, and the negotiation of agency and authority between civil and state actors in urban governance. The educational component of the project advances knowledge about effective pedagogy in urban geography and GIS, including ways of structuring experiential and service learning activities so that students are best able to connect conceptual and applied geographical knowledge, and critically reflect upon educational and societal importance of their active learning efforts. The ideas to be developed in this project will have an important impact in ongoing debates about effective practices for improving quality of life in US inner city neighborhoods. Through detailed knowledge about the changing roles and relationships of civil and state institutions, and the impact of different forms of knowledge, information and technologies within urban political processes, the project findings will inform efforts to create effective strategies and mechanisms for fostering urban revitalization. Finally, the model for integrated service learning and participatory research advanced in this project has the potential to reconfigure our understanding of the roles that universities play in the production and reproduction of urban spaces.

a. RESULTS OF PRIOR SUPPORT – not applicable, no prior support b. PROJECT DESCRIPTION Research Project Statement and Key Research Questions Over the past decade, non-governmental, non-profit, and voluntary citizen organizations have take an increasing responsibility for planning, problem solving and service delivery in U.S. cities. Simultaneously, a growing number of these organizations have begun utilizing spatial analysis technologies such as geographic information systems (GIS) to inform their efforts. Both of these developments have captured the attention of geographers, as each involves important shift in the processes through which local level urban change is negotiated, planned, and implemented (c.f. Brenner and Theodore 2001; Craig, Harris, and Weiner 2002). Theorization of the interdependent nature of these two developments remains under-developed in either urban geography or critical GIS research. The proposed research seeks to better understand the mutually constitutive relationship between GIS-based knowledge production by communitybased organizations, and the changing role and power of these institutions within urban political processes shaping land use, neighborhood revitalization, and community development efforts. “Community-based organizations” refers to organizations whose activities are geographically specific within a locality – directed at fostering change within a defined area in a city – with no implication that the social community represented is singular, unchanging, or uncontested. This research will develop stronger conceptualizations of how GIS use by such organizations, and application of the knowledge produced alters the processes and socio-political relations through which urban spaces are constituted and transformed. The project will advance urban geography and social studies of spatial technologies by building a better understanding of changing relationships between state and civil society in contemporary urban governance practices shaping city spaces, and conceptualizing the significance and impacts of spatial technologies and knowledge in these changing relationships and practices. Key research questions include: How and through what means does GIS-based spatial knowledge production by community-based organizations alter urban political processes through which neighborhood level change in urban spaces is negotiated, planned and implemented, and what are the implications of these changes for the relationship between the state and institutions of civil society in urban governance? How do changing urban governance practices within planning, revitalization, and service delivery affect the priorities, strategies, and goals advanced through GIS-based spatial knowledge production by community-based organizations? What forms of democratic practice and citizen participation are fostered through these changing governance strategies? More specifically, within GIS-based knowledge production by community organizations, what is the relationship between local state agendas, priorities and strategies for revitalization, planning, and service delivery, and those of community organizations?

1

Career Development Plan Prior Research and Education Accomplishments and Implications for this Project The proposed research and educational activities build on the central themes that have guided my career thus far. My scholarly efforts center on understanding the constitution of space, place, identities and power relations through locally inscribed collective action and citizen organizations. In particular, I consider the urban political significance of the knowledge and discourses produced by marginalized social groups and institutions based in socially, politically, and economically disadvantaged places. My earliest work examined these themes in the context of local level environmental justice activism, and negotiations of identity and “community” by lesbians occupying inner-city neighborhoods (Elwood 1998, 2000). Most of my research, however, examines these themes through study of the complicated interface between GIS-based knowledge production by community-based organizations and the social, political, and spatial impacts of that knowledge within urban planning, problem solving, and revitalization. My work in this area began with analysis of the unique spatial information technology needs and applications of community organizations and efforts to conceptualize key local support structures for GIS use in this context (Craig and Elwood 1998, Elwood and Leitner 1998; Leitner, Elwood et al. 2000). I continued with extended study of the ways in which GIS use in community-based planning alters discourses of decision-making about neighborhood space, with unequal impacts upon the kinds of knowledge considered relevant in neighborhood revitalization, and upon the participation and authority of social groups within a neighborhood (Elwood 2000). Most recently, I have focused upon spatial technologies and knowledge production as inextricably related to the changing nature of urban governance and its shifting role for institutions of civil society (Elwood 2002a, 2002b, 2002c, Elwood and Leitner 2002). The proposed project is designed to extend this interest through greater understanding of the forms of spatial knowledge and democratic practice advanced through community-based GIS use in urban planning and problem solving. As well, the research and educational activities in this project continue my focus on qualitative methodologies and participatory research approaches (Elwood and Martin 2000), and on the significance of university-community collaboration in fostering critically informed scholarship and pedagogies (Elwood 1996 and 2000d). Pragmatically, my research activities to date have developed a strong background in creating and sustaining research partnerships with community-based organizations. To prepare for the proposed project, I have been building relationships with Chicago area organizations over the past 2 years, and seeking to better understand the local political context of citizen participation and urban revitalization in which these groups are embedded. As well, I have recently completed several interviews with Chicago community-based organizations concerning the local support infrastructure for spatial data and technology provision to citizen institutions. Goals Through this project and beyond, I seek to position myself as an active and innovative contributor in geography, through activities that closely interweave research, education, and service. In my scholarly activities, I am working to be a leader in urban geography and critical GIS research, through research that further intersections in thought and theory from these subareas of geographic research. The proposed project would further this goal through its potential

2

research impacts, which include building stronger theorization of 1) the role and impacts of geographic information technologies and spatial knowledge in shaping urban spaces through processes of planning, revitalization and service delivery; and 2) the changing engagement of grassroots groups and other civil institutions in these governance processes through which urban change is negotiated. These contributions are most significant within urban and political geography, critical GIS research, and social studies of technology. The project design will advance thinking in geography concerning participatory research methods and experiential learning pedagogies. My goals as an educator are two-fold. I seek to build creative and effective learning environments for undergraduate students with diverse backgrounds and learning styles, and build knowledge about effective practices and curricula in urban geography and GIS that foster students’ critical understanding of socio-spatial inequality and their own social and political roles as citizens. One of the ways I will pursue these goals is through design of course activities that engage students in linking experiential knowledge with conceptual knowledge – to enable critical reflection upon the social and political geographies of the places in which their experiential learning is occurring. Specifically, the proposed project will work toward these goals through two linked experiential/service learning courses in geography in which students will work collaboratively with two community development organizations. The partner organizations in this project work in Chicago’s Humboldt Park neighborhood, an African American and Latino community that is one of Chicago’s most distressed neighborhoods. This site and the two organizations are selected because of their theoretical significance relative to the research questions, but also because they provide an opportunity for the students to examine socio-spatial inequities and their causes and implications in U.S. inner-city neighborhoods. The service learning activities of these courses, described in more detail in sections below, further my goals in student education, but also generate key data supporting the project’s research component. The service goals of this project are situated within DePaul University’s mission of urban service, which includes creating an accessible and effective educational environment for first generation college students and lower income households, and taking an active role in promoting social justice and addressing critical urban problems (Meister 1998). The students involved in the project will represent DePaul’s diverse student body, in which nearly 40% are racial and ethnic minorities, nearly half are first generation college students, and many come from central city neighborhoods precisely like the one targeted in this project. In seeking to foster strong interactive relationships between the university, students, and the Humboldt Park community, this project will create opportunities for students to connect real-world experiences (their own and those of community residents) to theory and concepts learned in the classroom, and to critically reflect upon the social and political significance of these experiences and ideas. Finally, this project seeks to build capacity for spatial analysis within the collaborating organizations. This goal will be met through creation of a community development spatial data library for their ongoing use, involvement of organization staff members in GIS education throughout the project, and investment in technological infrastructure of the organizations to facilitate independent GIS use and further manipulation of projects produced by the students. Objectives and Significance of the Proposed Research / Education Activities

3

There are three related sets of objectives for this project. First, I seek to build stronger theorization of the urban spatial and political impacts of GIS technologies and GIS-based knowledge production by community-based organizations as they seek change in urban neighborhoods and involvement in political processes and decisions through which these places are transformed. The project seeks to develop a better understanding of how community-based organizations are able to affect change in urban spaces through their involvement in planning, revitalization, and service delivery; and how relationships between state and civil society may be altered through deployment of GIS-based knowledge in these activities. The primary significance of these objectives lies in their capacity to extend geographers’ understanding of the role of spatial analysis technologies in the construction of space and place, the changing nature of democratic practice being fostered in urban political processes through which urban spatial change is negotiated, and the importance of spatial knowledge in these negotiations. “Spatial knowledge” here refers to representations of the characteristics, conditions, and needs of neighborhood spaces and structures. Given recent theorization of local citizens and civil institutions as increasingly marginalized in the face of neoliberal urban governance imperatives, and of urban spaces as being produced and reproduced by processes situated outside of localities, the proposed project is particularly important because it tries to understand, within this context, the remaining (or reconfigured) significance of local citizens, community-based organizations, and locally-produced spatial knowledges in shaping urban space. Second, in undergraduate education, the principal objective of the project is to design and implement linked experiential/service learning activities in 2 geography courses - the capstone course in my department’s urban geography sequence, and an advanced GIS applications course. Students in these courses will work with the two collaborating community-based organizations on GIS-based spatial analysis projects for use in the organizations’ neighborhood revitalization activities. Groups of students in the urban geography course will work in consultation with community organizations to produce plans for the project goals and potential applications within organizational activities. The GIS students, using a spatial data library created and maintained for the project, will collaborate with the organizations’ staff members to create the maps and analysis needed for the project. Both courses will include working sessions attended by students and organization staff to ensure community definition and guidance of student activities. As well, each course will conclude with a roundtable discussion of results, implications, and potential applications of materials produced, attended by all participants. The activities of each course will further include reflective writing and discussion of the experiential activities, so that these can be linked to the students’ conceptual learning. These educational activities reconstitute undergraduate students as active and engaged knowledge producers – a key step in fostering their critical thinking skills, and underscore for them the real-world relevance and impact of geographic knowledge. Finally, these educational objectives develop a strategy for teaching GIS as a socially constructed technology – a goal that has been called for in GIS literature (c.f. Warren 1997) but remains unfulfilled by most undergraduate GIS curricula. These educational objectives extend existing efforts to consider the benefits of experiential and service learning in the context of undergraduate geography education. The final objective of the project is to develop a sustainable model for university-community collaboration in community-based GIS projects, and developing internal capacity for spatial analysis in community-based organizations. Such a model would make a tremendously

4

important contribution to the theory and practice of community-based GIS. All existing empirical research points to the extreme difficulty community-based organizations face in sustaining GIS use, with further observation of the challenges of sustaining universitycommunity collaborations in this arena. The project design and activities outlined in this proposal lay the groundwork to develop a well-tested and robust model of community-GIS use and university collaboration that enable community-based organizations to surmount these challenges. The project will accomplish these objectives through tangible skill building of organizational staff members, and through involvement of the organizations throughout the process of planning for, creating, and applying GIS-based spatial knowledge. Each year of the project, one staff member from each organization will enroll in DePaul’s introductory GIS course. This multi-year commitment to internal skill building is essential given the relatively high staff turnover in community development agencies, and should increase the likelihood of retention and sharing of expertise in the organizations. Close collaboration through all aspects the GIS process also plays an important role in sustaining community GIS capacity – ensuring not just software skills but broader understanding of process of GIS-based knowledge production. The long term process of data development and skill building supported under this project will additionally lay the groundwork for university-community GIS collaboration beyond the duration of this proposed project, through the ongoing collaboration of the Humboldt Park community and DePaul’s Egan Urban Center, a participatory urban research institute. The Relation of the Research to the Current State of Knowledge in the Field Research activities of this project are informed by and seek to advance existing knowledge in geography concerning the social, political, and spatial impacts of GIS use, and of the shifting role and significance of citizens and local civic institutions within urban political processes shaping urban space and place. Study of the societal implications of spatial information technologies is rooted in the work of a number of scholars who have studied the social construction of GIS the kinds of spatial analysis and representation of space, place, and people advanced through this technology (Aitken and Michel 1995, Harris et al. 1995, Lake 1993, Pickles 1995, Sheppard 1993, Yapa 1991). These examinations of the epistemological assumptions about space and place inherent in GIS technology and types of knowledge and decision-making practices likely to be fostered through its application share a common concern with the technology’s purported tendency to advance quantitative, rationalist forms of spatial data and logic. One of the crucial contributions of this early work is its identification of knowledge construction as a primary process through which the social, political, spatial impacts of the technology occur. This conceptualization underlies the research investigations of this project by suggesting the following questions to be explored: What kinds of spatial data and community knowledge do the community-based organizations deem important and legitimate for inclusion in GIS analysis? What representations of and knowledge about community space – particularly conditions, needs, and resources – are produced under community-directed GIS analysis? Critical GIS research has particularly focused on the use of technology by and its implications for socially, politically, and economically marginal communities, organizations, and social groups. Case studies from around the world suggest a rapid expansion in GIS use in community, grassroots, and non-governmental organizations, in spite of the financial, time, and expertise

5

limitations they face and profile a wide array of successful strategies for fostering communitybased GIS use. Responding to such case studies, Craig, Harris, and Weiner (2002) emphasize that the impacts of GIS are highly context dependent, shaped by and responding to the particular socio-geographic situations in which they are utilized (See also Barndt 1998, Carver 2001, Elmes 2001, Leitner et al. 2000). The research questions and goals guiding this project are informed by this emphasis on the context-dependency of GIS use and impacts, which imply a need to develop situated theoretical frameworks in which the impacts of GIS and GIS-based knowledge are conceptualized as inextricably connected to and affected by the context in which they are embedded. For the proposed research, such a framework implies examining precisely how knowledge and authority are constructed and negotiated in GIS-based knowledge production and its deployment in urban political processes that constitute urban spaces. What priorities, activities, and agendas of community-based organizations are informed transformed, or initiated using GIS-based knowledge, and how? How and in what interactions and negotiations is this knowledge deployed by community-based organizations as they participation in local politics of urban revitalization and community participation? Drawing upon these understandings of the contextual nature of the impacts of GIS, and the important role that spatial knowledge production and application play in fostering these impacts, the research component of the proposed project is further informed by earlier conceptualizations of the mechanisms through which they impacts occur. In previous research (Elwood 2002a), I argued that GIS use by community-based organizations altered the participation and power of citizens and community organizations by shifting the language, practices, and underlying logic and priorities of community decision-making and neighborhood revitalization. The design and methods of this past project produced data useful in explicating the impacts of GIS within a community’s own dialogue and decision-making, but with a much more limited capacity to conceptualize how and with what impacts GIS-based knowledge was deployed by these organizations in the broader urban political arenas in which they engage – one of the central issues explored in the research component of this project. In this focus, the project directly engages ideas being generated in geography and urban studies concerning the implications of changing urban governance practices for citizens and community organizations. For the past decade, scholars have studied the changing involvement of citizens and voluntary organizations in urban planning, revitalization and service delivery, situating this development as part of a move toward participatory planning (c.f. Healy 1997, McCann 2001), collaborative governance (Kearns and Paddison 2000, North 2001), or a neoliberal agenda of downsizing the state, with devolution of key local state responsibilities to community organizations and local citizens (Brenner and Theodore 2001, Jessop 2001). Concerns are raised about declining resources available to citizen organizations to carry out these responsibilities (Boyte 1989, Grieder 1992, Jacobs 1992, Putnam 1995), and about whether these new responsibilities are accompanied by a parallel expansion in community voice and power in urban decision-making (Fisher 1994, Handler 1996, Hasson and Ley 1994, Lake and Newman 2002). Others worry that the neoliberal policy regimes or purportedly “collaborative” governance strategies under which this expansion is occurring have the potential to pre-occupy citizens and voluntary organizations in the service of state priorities (Peck and Tickell 1994 and 2002, Taylor 2000), and to limit democratic debate in urban politics by situating these organizations within state agendas for neighborhood revitalization and other kinds of urban spatial change (Atkinson 1999, Gough

6

2001, Taylor 2000). Within these debates, other scholars argue that the expanding responsibilities and activities of community organizations in fact create new spaces of engagement that may be advantageous. These studies are premised upon an understanding that citizens and community organizations continue to be active agents in the negotiation of urban spatial change. They emphasize opportunities for community organizations to formulate of new alliances with the expanding range of actors/institutions involved in urban governance (Kearns and Paddison 2000), to create alternative strategies and priorities for action (Sandercock 1998), and to develop new capacities and knowledge that strengthen the basis of their engagement in urban political processes affecting their communities (North 2001, Taylor 2000, Maloney et al. 1994). Debates within this body of knowledge inform the proposed project in several ways. First, they outline important new developments within the urban political context in which communitybased GIS use is situated – particularly highlighting current shifts in the way that priorities, practices and agendas for urban spatial change are mutually constructed and negotiated by state and civil actors and institutions. As well, just as GIS research conceptualizes knowledge construction and application as critical to understanding the socio-spatial impacts of GIS, some scholars with the urban governance debates see construction of knowledge and its application to community-level activities and priorities as essential components in ensuring an active and influential role for these organizations in the new urban political context in which they operate. The research component of this project thus conceives of knowledge and practices developed within the changing structures of urban governance as emerging from interaction among state and civil actors and institutions. Under this conceptualization, key questions to be explored in this project include what sort of knowledge and authority do community-based organizations construct in their activities and interactions with the local state, and what are the relationships between this knowledge and that produced by state actors and institutions with whom these organizations interact? My early proposition would be that community-based organizations actively modify the priorities and strategies advanced in their GIS-based knowledge production with specific attention to particular situation and audience to be engaged. In sum, the research activities of the proposed project are motivated by the need for stronger theorization of the complex interdependencies between spatial analysis technologies, GIS-based knowledge construction, and the negotiation of agency and authority by civil and state actors within urban governance. Development of these ideas will foster better understanding of the processes through which urban spaces are constituted and transformed in US cities. The research component rests on an assertion of the centrality of knowledge production and application to the active engagement of community-based organizations in spatial decision making processes affecting their communities, and seeks to understand the mutually constitutive involvement of state and community actors in negotiating priorities and practices for urban spatial change. The ideas developed through the research activities of this project are of course shaped by the local context in which they are conducted. Chicago has a lengthy history of community organizing that has been somewhat more focused upon labor politics than explicitly “placebased” politics. Place-based politics refers to the local decision-making processes that explicitly constitute and transform urban spaces. In Chicago, local state institutions instead have attempted to channel citizen involvement in place-based politics through the city’s ward system and

7

aldermanic decision-making processes (Bennett 1997, Ferman 1998). Thus, I expect the processes of GIS-based knowledge construction and application studied in this project to involve negotiation of priorities and goals for neighborhood transformation between community-based organizations and the individuals and institutions who participate in place-based politics through the aldermanic system. Chicago has a history of limited public information access (Mayfield et al. 1999), and my preliminary investigations for this project show a relative absence of local government efforts to ensure public availability of information, or support digital data provision and GIS access for community-based organizations. In light of these conditions, I expect that GIS-based knowledge production by these organizations to have a groundbreaking impact upon the negotiation of place-based politics in Chicago, perhaps enabling clearer insight into these impacts than has been possible in case studies that have been conducted in more progressive local state contexts of information access and citizen participation. Relation of Education Activities to Knowledge of Effective Learning in Geography The educational activities of the project are informed by and contribute to current knowledge about effective learning in human geography and GIS in higher education. The close integration of research and education activities in this project, in which student learning activities will play an essential role in producing knowledge necessary for achieving the research goals, respond to Warf’s (1999) argument that geographic research and education can and should be practiced as mutually constitutive. The experiential and service learning activities undertaken in the project respond to a growing body of knowledge concerning their positive societal and educational implications. Scholars have shown experiential learning to be an effective strategy for helping students link abstract and applied knowledge (Kolb 1984), further illustrating how service learning can foster active and responsible citizenship (Kahne et al. 2000, Varlotta 1996), and critical examination of societal stratification along lines of race, class, gender, and ethnicity (Densmore 2000). A growing body of knowledge about the particular significance of experiential and service learning in geography informs the design of the educational activities in this project. Some geographers contend that geography education is rooted in the same constructivist theory of learning through experience, reflection, experimentation, and evaluation that inform service and experiential learning (Dorsey 2001). Constructivist theories of learning further understand knowledge as actively produced and revised by students, teachers and communities (Merrett 2001). This emphasis on the production and modification of knowledge through social interaction has a particularly important role to play in student learning in this project, because it reinforces one of the key learning goals of the course – understanding the production and impacts of GIS-based knowledge. This overlap in student learning goals and assumptions guiding the learning process will, I expect, foster stronger analysis and learning by the involved students. Geographers have further emphasized the possibilities of promoting social justice through the constructivist pedagogy of service and experiential learning (Flint 2002, Merrett 2001). To these ideas about the positive impacts of service and experiential learning for geography education, I would add that these approaches make the important contribution of helping students understand the relevance of geographic concepts for understanding and fostering change in the spaces and places they inhabit. This linkage of conceptual and applied knowledge is essential in geography education at university level where, in spite of strong

8

national progress toward ensuring geography education at lower levels, many undergraduate students have had little to no prior exposure to geography. The service learning activities incorporated in the educational component of this project are informed by Cone and Harris’ (1996) conceptual model for service learning. This model draws upon Kolb’s (1984) more widely known experiential learning cycle, but more explicitly defines instructors, students, and community members all as active agents in the process of knowledge production and learning. This conceptualization is essential to fulfilling the research objectives, educational objectives, and community capacity building objectives that are woven together in this project. Cone and Harris’ model involves the following steps: 1) definition of the service tasks by all participants in pragmatic and conceptual terms, 2) engagement of students in experiential tasks and in critical reflection upon the process and knowledge produced, and 3) promotion of “mediated learning” forums in which the instructor guides students in developing greater understanding of the meaning and significance of the knowledge created, within community and scholarly settings. While Cone and Harris conceive of community members as primarily present in the first step, I involve them throughout to strengthen the student and community learning that can occur, and to create forums for participant observation that will generate data for the research component of the project. The activities of students enrolled in Geography 333: City Problems and Planning and Geography 245: Community-Based GIS in Theory and Practice will be structured around this cycle of activities and learning. I will promote the courses to students as linked, because I expect the strongest learning to occur for those students involved in the entire process of planning and producing the community spatial analysis projects. Each course will include two to three student community working sessions, and a capstone roundtable discussion – forums through which much of the task definition, experience, reflection and mediated learning will occur. In Geography 333, participants in the working sessions will lay out the goals and potential applications of the community spatial analysis projects, and in Geography 245 these sessions will provide an opportunity for community members to direct the implementation of these projects and provide in-progress assessment and ideas for modification of the maps and analysis. In each case, these working sessions integrate task definition into experiential activities. Critical reflection and mediated learning from these activities will occur in follow-up discussions and in written field journals. Students will be asked to reflect upon questions that might include: What types of information and data are being prioritized by the community in these projects? What are the “targets” they envision for this project – places to change, problems to solve, people to inform, or political processes to influence? As you reflect on the maps and analysis the community wants to produce, what sort of “story” is the organization trying to communicate about spaces, places, and people in the community? What do these stories suggest to you about capacity of and strategies through which community-based organizations try to create change? Seminar discussions relating course readings to the ideas students produce in these reflections will enable further guidance of the mediated learning process. The roundtable discussion of results, implications, and potential applications of materials produced, attended by all participants, also encourage such reflection and mediated learning. My observation of students as they participate in the interactive forums built into these two courses, together with review of their reflective field journals, will provide a rich source of information through which to assess their learning.

9

The experiential learning design of Geo245 responds especially to needs within GIS education. Institutions such as the National Center for Geographic Information and Analysis and University Consortium for Geographic Information Science advocate active learning methods in GIS education as well as inclusion of societal implications of GIS within these curricula (Kemp 1997, Kemp and Wright 1997). In spite of these efforts, discussion of GIS pedagogy is noticeably absent the literature on experiential and service learning in geography. Embedding GIS education within experiential service learning, as this project seeks to do, has tremendous potential to advance understanding of how to integrate conceptual and “practical” learning in GIS courses – identified in the GIS education literature as a key challenge. As well, service learning in GIS has potential to encourage students to thoughtfully consider the implications of maps and analysis produced reflect upon the possibilities of employing GIS to highlight and challenge spatial inequity. Outline of the Plan of Work The proposed project is designed as an annual cycle of linked research and education activities. While there is some degree of overlap, the bulk of education activities are concentrated in DePaul’s Winter and Spring Quarters, and with research activities and analysis concentrated in the summer and in a 6-week break between Fall and Winter Quarters. Evaluation and revision of research and educational activities will occur throughout. Project activities will begin with creation of a spatial data library for use by the participant organizations. The data library will include demographic, housing, and economic data from national, state and local sources, neighborhood knowledge developed by the partner organizations, and digital photos taken by project assistants and community residents. These data will be used in the two experiential learning courses, as the students, community organizations and I carry out each annual cycle of spatial analysis project planning and implementation. Finally, project assistants will create and maintain a project website through which the digital data and maps will be accessible to the community organizations on an ongoing basis. My own participant observation of these various university-community interactions and analysis of materials produced in these activities, together with annual interviewing of participating community members, organizational staff, and local government officials will generate a rich data set informing the project’s research goals. Table 1. Schedule of Research and Education Activities Research Activities

Education Activities

• Meet

Summer 2003

Fall 2003

with community organizations to define data needs for data library • Preliminary interviewing with organization staff and local government officials to build understanding of placebased politics in Chicago. • Inventory/collect all relevant data held at DePaul to be included in data library • Gather local government and community data for data library • Initial spatial analysis project planning

10

•

Develop course plans and materials

•

Finalize new course approval for Geo 245, experiential learning approval for Geo333.

•

Student recruiting

with community organizations Finalize Human Subject approval for Geo333/Geo245 • Purchase and set up hardware & software at community sites • Finish data library development for first iteration of courses • Observation of student-community working sessions, roundtable discussions • Evaluation of goals and procedures • Observation of student-community working sessions, roundtable discussions •

Winter 2004

Spring 2004

Year 2 Summer 2004

Fall 2004

Winter 2005 Spring 2005 Year 3 Summer 2005 Fall 2005 Winter 2006 Spring 2006 Year 4 Summer 2006 Fall 2006

•

Geo 333 (creation of spatial analysis project plans and goals for each group) • Student-community working sessions and roundtable discussions • First 2 community participants enroll in DePaul’s Introduction to GIS course • Geo 245 (implementation of spatial analysis project plans with each group) • Student-community working sessions and roundtable discussions Research Activities Education Activities • Interviews with organization participants • Revision of course plans / activities based and local government officials on student and community evaluation • Compilation / analysis of in-progress • Produce community report field notes and student field journals • Prepare Human Subject renewal for st upcoming year’s activities • 1 projects/maps added to web site • Acquire/learn qualitative data analysis software program • Update / expand data library for upcoming year • Analysis of data gathered to date • Planning meetings with participating organizations to evaluate / revise course • Finish interviews from summer activities and spatial analysis project • Continue collection / analysis of any outcomes organizational documentation using spatial analysis project results • Begin writing journal articles of • Continue as in Winter 2004 preliminary results. • Continued writing and analysis • As in Spring 2004 Research Activities Education Activities • As in Summer 2004: interviewing, • Evaluation and revision, as in Summer document collection and analysis, 2004. update data library and web site • As in Fall 2004 • As in Fall 2004 • As in Winter 2004 • As in Winter 2004 • As in Spring 2005 • As in Spring 2005 Research Activities Education Activities • As in previous summers – interviewing, • As in Summer 2005 • Additionally, prepare for Year 4 offerings analysis, data maintenance, evaluation / of Geo 245 and Geo 333 revision of procedures • Research leave for intensive midpoint • (research leave) analysis, continuing activities begun in summer 2006: extended interviews, compilation and analysis of all data gathered to date, continued preparation / revision of journal articles

11

Winter 2007 Spring 2007 Year 5 Summer 2007 Fall 2007

Winter 2008 Spring 2008 Summer 2008

•

•

•

•

As in Winter 2006 As in Spring 2006 Research Activities • Interviewing • Transcription • Write book prospectus • Circulate book prospectus to potential publishers • Ongoing analysis and writing

•

As in Winter 2007 As in Spring 2007 • Interviews and document analysis, as before • Final evaluation of project • Begin preliminary writing for book •

As in Winter 2006 As in Spring 2006 Education Activities • Compile community report

•

Meet with participating organizations, geography department, and DePaul Egan Urban Center to prepare plans for sustaining data library and service learning courses • As in Winter 2007 • As in Spring 2007 • Develop project report with community participants • Final evaluation of educational outcomes

Specific Research Issues This section outlines elements of research design, theoretical significance of site selection, data sources, research methods, and other key implementation issues, to further clarify project goals, theoretical contributions and requested resources. 1. Research Design and Field Sites One of the key observations in existing research on GIS use and impacts for community-based organizations is the high level of local variability in these processes, even for organizations operating in the same place (c.f. Elwood and Leitner 1998). I propose that examining this interorganizational variability may provide key insights into how GIS-based knowledge production and application are negotiated through and shaped by intersecting organizational and local political contexts. Involving two organizations situated in the same local political environment and responding to similar neighborhood conditions and needs should further clarify the nature of and reasons for this variability. The West Humboldt Park Family and Community Development Council, and the Near Northwest Neighborhood Network are both active in Chicago’s Humboldt Park area, and have similar organizational missions of improving quality of life through housing improvement, expanded economic activity and job opportunities, crime reduction, and increased community cohesion. Nonetheless, each has its own history, opportunities, and limitations within politics of urban revitalization in Chicago, and may thus produce distinctly different kinds of GIS-based knowledge and rely upon different strategies for deploying this knowledge in place-based politics, with varying impacts upon the production and reproduction of space in Humboldt Park. The selection of Humboldt Park are a field site is motivated by ongoing focus in critical GIS research upon the use of GIS technologies by socially, politically, and economically disadvantaged communities, and the impacts for these places and people. One-third of community residents live below the federal poverty level, and approximately 25% of adults are unemployed. 44% of residents are Latino and 49% are African American. Half of the housing stock in the area is seriously dilapidated or vacant. A number of practical considerations also guide selection of Humboldt Park as the site for this project. DePaul’s Egan Urban Center has 12

worked extensively with the community for many years, developing a level of institutional and personal trust that has already significantly enhanced preparation for the proposed project. In this collaboration, the Egan Center has greatly enhanced technology access for neighborhood residents – through neighborhood computer centers and computer distribution programs, and to enhance digital technology capabilities of Humboldt Park community organizations through technical assistance and training programs. These activities further lay the groundwork for implementing the proposed project. 2. Data Sources and Issues of Validity Project activities are designed to yield a rich array of ethnographic data addressing key research questions, particularly oriented toward developing data that enables multiple perspectives on these questions. The spatial analysis project plans produced through Geo333, and the maps produced through Geog 245 will provide insight into the spatial data and community knowledge which the participating organizations deem important and legitimate for community development applications in their communities, and a source for analyzing the representations of and about community space that produced under the community-directed GIS projects. Further data addressing these issues will be drawn from my own field notes prepared from viewing video recordings of the student-community working sessions and the capstone roundtable discussion for each course. I intend to video-record these interactions for later analysis because when they are occurring I will be occupied in facilitating the gatherings, making simultaneous research observation difficult. Ethnographic interviews targeting two groups of interviewees will generate data about the way in which the GIS-based maps and spatial knowledge created in the project are informing and transforming organizational community development goals, strategies, and activities; as well as how and to what effect these representations and knowledge are being deployed by the organizations in these arenas of neighborhood planning, revitalization, and service delivery. I expect to interview participating organizational staff members annually, but also local government officials with whom these organizations interact in their activities. These individuals will most likely include elected and appointed members from city ward offices in the Humboldt Park area, and as well as staff members in Chicago’s Departments of Planning and Development, Community Development, Housing, Zoning, and Workforce Development – the city institutions with whom these organizations work most frequently. These interviews will provide multiple perspectives upon the application and impacts of GIS-based spatial knowledge in urban political processes in which the community organizations are engaged. To further address this issues, I will collect and analyze reports, strategic planning documents and proposals produced by the organizations in which they draw upon maps and analysis produced as part of the project. To ensure thorough and rigorous interpretive analysis of the large volume of ethnographic data gathered in the project, I will code and analyze data using a qualitative data analysis software. Given the close linkages between research and educational activities in this project – particularly the collection of key data informing the research through activities in which I and the participating students will be directly involved – it is important to clarify several issues of validity. Reliance upon data gathered in university-community collaboration to inform research questions about GIS use and impacts is evident throughout critical GIS research practice, so the

13

design of the project is situated within bounds of accepted research practice in this sub-area. The key variable affecting validity of project findings vis-à-vis the urban political and spatial impacts of GIS-based knowledge production and application is the role that the students and I will play in these processes. I have specifically designed the project to engage us in a technical facilitation role, and to engage the participating community members as directors of the focus, scope, and application of the spatial analysis being carried out. Additionally, my reliance on multiple sources of data and collection techniques enables triangulation of ideas within the analysis and interpretation of these data – a key strategy to ensure rigor and validity of research findings in ethnographic research (Burawoy 1991, Yin 1994). 3. Human Subjects and Dissemination of Results I am presently working with DePaul’s IRB office to prepare necessary documents for human subjects approval, including clearance for my own activities and those of the students in each class. In designing these courses, I will include readings and activities to ensure that students receive explicit training in human subjects protection and research ethics. As well, the small size of these courses (approximately 20 students) will ensure that I am further able to provide individual mentoring throughout these courses. In Geo245, I will particularly highlight ethical issues in GIS, to try to make more transparent the ethical issues inherent in data acquisition, analysis, and representation processes in which the students will be engaged. I will disseminate results of the project in academic forums, the Humboldt Park community, and broader Chicago community development forums. Scholarly dissemination will take several forms and target several research areas in geography. I expect to produce multiple manuscripts for dissemination in urban geography, GIS, and geographic education journals. In the 2 years following completion of the project, I plan to produce a book focusing on the research results of the project, targeting the book toward an audience in urban geography, urban politics, community development. Throughout, I will make presentations at academic conferences focusing upon the project’s research and educational initiatives. In terms of community dissemination, all data developed for the spatial analysis data library and materials produced in the spatial analysis projects will be available to the partner organizations, and will be disseminated in local community development and neighborhood organizing forums such as the Great Cities Conference – an annual gathering of scholars and practitioners to share results of research and action throughout Chicago. Data and results will also be shared with Humboldt Park’s many other community organizations in presentations, and through the project website. This process will be facilitated by the ongoing partnership between DePaul’s Egan Urban Center and the Humboldt Park community. I am particularly committed to this local dissemination of data and results, given the underdeveloped infrastructure of local support resources for community-based spatial analysis. Specific Education Issues Having previously described the educational goals, activities, and specific strategies to be advanced in this project, I focus here on implementation issues and feasibility. An important element in the design of the educational component is the inclusion of two courses, rather than simply focusing on a single GIS course. At a philosophical level, incorporating Geo333 into this project provides opportunity to engage students in thinking critically about the politicized

14

processes through which urban spaces are created and modified, and to examine the active agency of citizens and civil institutions in this process. At the practical level, the spatial analysis projects to be produced cannot be completed from start to finish in a 10-week undergraduate course. With an eye toward maintaining a feasible workload in light of my other job responsibilities, I have targeted Geo333 because it is an existing course within my department’s curriculum, making Geo245 the only new course to develop under the educational activities of this project. Enhancing both the feasibility of the educational plan, and the quality of learning that can occur in these courses, both will be included in DePaul’s liberal studies curriculum as “experiential learning” – a designation that reduces their enrollment cap to 20, from the usual limit of 40 students. For the purposes of the experiential activities related to this project, I expect to have students organize themselves into 4 smaller working groups. I expect the courses to enroll an interdisciplinary group of students. Target groups outside geography include: public policy, sociology and anthropology students in each program’s respective urban studies concentrations, undergraduates from DePaul’s computer science school, and participants in DePaul’s GIS Certificate Program. In Fall 2002, the Geography and International Studies Departments will launch an additional certificate program in which students will learn conceptual and technical skills in spatial analysis technologies, through applications that can be broadly understood as analyzing or promoting social justice. I expect the Spatial Justice Certificate Program to play a strong role in developing student interest in Geo245 and preparing students to perform well in the course. Implications of the Proposed Study’s Integrated Research and Education Activities The interconnection of the research and education components of this project is significant in several ways. By integrating sustained investigation of the urban political and spatial impacts of GIS-based knowledge construction and application with development of active pedagogy in GIS and urban geography, the project furthers my career development goals of making strong and creative contributions in urban geography, critical GIS, and geographic education. The 5-year duration of the project is essential to the interconnected goals of the project, ensuring development of an effective and sustainable service learning model for GIS collaboration, as well as development of more detailed and robust theorizations of the urban political and spatial impacts of GIS-based knowledge. The latter is particularly significant, given the short-term nature of many previous research efforts through which current knowledge about societal impacts of GIS have been developed. Interconnection of the research and education components shifts the role of participating students from one of passive knowledge consumption in their educational activities to one of active knowledge production – a shift that will enhance their learning, critical thinking skills, and conceptualization of their own responsibility and efficacy as citizens. Finally, the linking activities and interactive circulation of information and ideas through the research and educational portions of the project and among the diverse range of participants will enable rich, multi-faceted data development and analysis informing to project’s key research questions. In this manner, the integrated project will foster significant conceptual advancement in urban geography and GIS to better understand changing relationships between citizens and the state, changing practices of urban governance, and multi-faceted implications of these developments for the production and reproduction of urban space.

15

References Aitken, S. and Michel, S. 1995. Who Contrives The 'Real' In GIS? Geographic Information, Planning, And Critical Theory. Cartography And Geographic Information Systems 22:17-29. Atkinson, R. 1999. Discourses Of Partnership And Empowerment In Contemporary British Urban Regeneration. Urban Studies 36(1): 59-72. Barndt, M. 1998. Public Participation GIS: Barriers To Implementation. Cartography And Geographic Information Systems 25(2): 105-112. Bennett, L. 1997. Neighborhood Politics: Chicago And Sheffield. New York: Garland Press. Boyte, H. 1989. CommonWealth: A Return To Citizen Politics. New York: Free Press. Brenner, N. and Theodore, N. 2001. Cities And The Geographies Of ‘Actually Existing Neoliberalism’. Paper presented at ‘Neoliberalism and the City’. Chicago, September 2001. Burawoy, M. 1991. The Extended Case Method. In M. Burawoy, Ed., Ethnography Unbound: Power And Resistance In The Modern Metropolis, pp. 271-287. Berkeley, CA: University of California Press. Carver, S. 2001. Participation And Geographical Information: A Position Paper. Workshop on Access to Geographic Information and Participatory Approaches Using Geographic Information. Spoleto, Italy, December 2001, http://www.shef.ac.uk/~scgisa/spoleto/workshop.htm Cone, D. and Harris, S. 1996. Service-Learning Practice: Developing A Theoretical Framework. Michigan Journal of Community Service Learning 3: 31-43. Craig, W. and Elwood, S. 1998. How And Why Community Groups Use Maps And Geographic Information. Cartography and Geographic Information Systems 25(2): 95-104. Craig, W., Harris, T., Weiner, D. 2002. Community Participation And Geographic Information Systems. London: Taylor and Francis. Densmore, K. 2000. Service Learning And Multicultural Education: Suspect Or Transformative? In C. O’Grady, Ed., Integrating Service Learning And Multicultural Education In Colleges And Universities, pp. 45-58. Mahwah, NJ: Lawrence Erlbaum and Associates. Dorsey, B. 2001. Linking Theories Of Service-Learning And Undergraduate Geography Education. Journal of Geography 100(3): 124-132.

1

Elmes, G. 2001. Responses To Papers On Access And Public Participation Using Geographic Information. Workshop on Access to Geographic Information and Participatory Approaches Using Geographic Information, Spoleto, Italy, December 2001, http://www.shef.ac.uk/~scgisa/spoleto/workshop.htm Elwood, S. 2002a. GIS And Collaborative Urban Governance: Understanding Their Implications For Community Action and Power. Urban Geography 22(8): 737-759. Elwood, S. 2002b. GIS Use In Community Planning: A Multi-Dimensional Analysis of Empowerment. Environment and Planning A 34(5): 905-922. Elwood, S. 2002c. Whose Neighborhood Is It? Revitalization Programs, Community Organizations, and the Local State. Rights to the City Conference. Rome, May 2002. Elwood, S. 2002d. Histories And Practices Of Activism In Minneapolis And Chicago: Implications For University-Community Collaboration. Annual Meeting of the Association of American Geographers. Los Angeles, CA, March 2002. Elwood, S. 2000. Lesbian Living Space: Multiple Meanings Of Home. The Journal of Lesbian Studies 4(1): 11-27. Elwood S. 1998. Community-Based Environmental Research: Exploring Issues of Equity and Democracy. Annual Meeting of the Association of American Geographers. Boston, MA, March 1998. Elwood, S. 1996. Community-Based GIS Education: A Twin Cities Experiment. Second International Symposium on GIS in Higher Education. Columbia, MD, September 1996. Elwood, S. and Leitner, H. Forthcoming 2003. Community-Based Planning and GIS: Aligning Neighborhood Organizations with State Priorities? Journal of Urban Affairs. Elwood, S. and Leitner, H. 1998. GIS and Community-Based Planning: Exploring the Diversity of Neighborhood Perspectives and Needs. Cartography and Geographic Information Systems 25(2): 77-88. Elwood, S. and Martin, D. 2000. ‘Placing’ Interviews: Location as a Consideration in Doing Qualitative Interviews. The Professional Geographer 52(3): 649-657. Ferman, B. 1996. Challenging The Growth Machine: Neighborhood Politics In Pittsburgh And Chicago. Lawrence, KS: University of Kansas Press. Fisher, R. 1994. Let The People Decide: Neighborhood Organizing In America, 2nd ed. New York: Maxwell Macmillan. Flint, C. 2002. Introduction: For A Pervasive Political Geography In The Curriculum. Journal of Geography 101(2): 61-62.

2

Ghose, R., and Huxhold, W. 2001. The Role Of Local Contextual Factors In Building Public Participation GIS: The Milwaukee Experience. Cartography and Geographic Information Systems, 28(3), 195-208 Gough, J. 2001. Neoliberalism And Socialization In The Contemporary City: Opposites, Complements And Instabilities. Paper presented at ‘Neoliberalism and the City’, Chicago IL, September 2001. Grieder, W. 1992. Who Will Tell The People: The Betrayal Of American Democracy. New York: Touchstone. Handler, J. 1996. Down From Bureaucracy: The Ambiguity of Privatization And Empowerment. Princeton, NJ: Princeton University Press. Harris, T., Weiner, D., Warner, T., and Levin, R. 1995. Pursuing Social Goals Through Participatory GIS: Redressing South Africa's Historical Political Ecology. In J. Pickles, editor, Ground Truth: The Social Implications of Geographic Information Systems. New York: Guilford, 196-221. Hasson, S. and Ley, D. 1994. Neighbourhood Organizations And The Welfare State. Toronto: University of Toronto Press. Healy, P. 1997. Collaborative Planning: Shaping Places in Fragmented Societies. Vancouver: University of British Columbia Press. Jacobs, B. 1992. Fractured Cities: Capitalism, Community, And Empowerment In Britain And America. New York: Routledge. Jessop, B. 2001. Good Governance And The Urban Question: On Managing The Contradictions Of Neoliberalism. Paper presented at ‘Neoliberalism and the City’, Chicago IL, September 2001. Kahne, J., Westheimer, J., Rogers, B. (2000). Service Learning And Citizenship: Directions For Research. Michigan Journal of Community Service Learning 7: 42-51. Kearns, A. and Paddison, R., 2000, New Challenges For Urban Governance: Introduction To The Review Issue. Urban Studies 37(5/6): 845-850. Kemp, K. 1997. The NCGIA Core Curricula in GIS and Remote Sensing. Transactions in GIS 2(2):181-190. Kemp, K. and Wright, R. 1997. UCGIS Identifies GIScience Education Priorities. GeoInfo Systems 7(9): 16-20.

3

Kolb, D. 1984. Experiential Learning: Experience As A Source Of Learning And Development. Inglewood Cliffs, CA: Prentice-Hall. Lake, R., 1993, Planning And Applied Geography: Positivism, Ethics, And Geographic Information Systems. Progress in Human Geography 17(3): 404-413. Lake, R. and Newman, K. 2002. Differential Citizenship In The Shadow State. Paper presented at Rights to the City Conference. Rome, Italy, May 2002. Leitner, H., Elwood, S., Sheppard, E., McMaster, S., McMaster, R. 2000. Modes of GIS Provision and Their Appropriateness for Neighborhood Organizations: Examples from Minneapolis and St. Paul, Minnesota. The URISA Journal 12(4): 43-56. Maloney, W., Jordan, G., and McLaughlin, A. 1994. Interest Groups And Public Policy: The Insider/Outsider Model Revisited. Journal of Public Policy 14: 17-38. Mayfield, L., Hellwig, M., and Banks, B. 1999. The Chicago Response To Urban Problems: Building University-Community Collaborations. American Behavioral Scientist 42(5): 863-875. McCann, E. 2001. Collaborative Visioning Or Urban Planning As Therapy? The Politics Of Public-Private Policy Making. The Professional Geographer 53(2): 207-218. Meister, R. 1998. DePaul University: Catholic, Vincentian and Urban. In Rury, J. and Suchar, C., Eds, DePaul Unversity: Centennial Essays and Images, pp. 5-50. Dubuque, IA: Kendall/Hunt Publishing Co. Merrett, C. 2000. Teaching Social Justice: Reviving Geography’s Neglected Tradition. Journal of Geography 99(5): 207-218. North, P. 2001. Community Action And Partnerships For Urban Regeneration: New Sites Of Struggle? Paper presented at Annual Meeting of the Association of American Geographers. New York, February 2001. Peck, J., and Tickell, A. 1994. Too many partnerships…the future for regeneration partnerships. Local Economy 9: 251-265. Peck, J. and Tickell, A. 2001. Neoliberalizing Space: The Free Economy And The Penal State. Paper presented at ‘Neoliberalism and the City’. Chicago, September 2001. Pickles, J. 1995. Representations In An Electronic Age: Geography, GIS, And Democracy. In Ground Truth: The Social Implications of Geographic Information Systems, ed. J. Pickles, pp. 1-30. New York: Guilford. Putnam, R. 1995. Bowling Alone: America’s Declining Social Capital. Journal of Democracy 6: 65-78.

4

Sandercock, L. 1998. Towards Cosmopolis: Planning For Multicultural Cities. New York: John Wiley. Sheppard, E. 1993. Automated Geography: What Kind Of Geography For What Kind Of Society? The Professional Geographer 45(4): 457-460. Taylor, M. 2000. Communities In The Lead: Power, Organisational Capacity And Social Capital. Urban Studies 37(5/6): 1019-1036. Varlotta, L. 1996. Service-Learning: A Catalyst for Constructing Democratic Progressive Communities. Michigan Journal of Community Service Learning 3: 22-30. Warf, B. 1999. Constructing A Dialogue: Geographic Education And Geographic Research. The Professional Geographer 51(4): 586-591. Warren, S. 1995. Teaching GIS As A Socially-Constructed Technology. Cartography and Geographic Information Systems 22(1): 70-77. Yapa, L. 1991. Is GIS Appropriate Technology? International Journal of Geographical Information Systems 5(1): 41-58. Yin, R. 1994. Case Study Research, Design and Methods, 2nd ed. Newbury Park, CA: Sage Publications.

5

APPENDIX

iii. Kenney, Martin and Patton, Don. Geography and Regional Science: The Geography of Entrepreneurial Support Networks. (non-funded version with reviews)

This study examines the spatial relationships between successful startup firms and various constituents of a startup's institutional support network that contribute to its birth and growth. In particular, this project requests funding to extend an existing database to include data on all domestic firms that went public between June 1996 through 2000. The existing database includes geographic information on the firms themselves, and five members of their support network: lawyers, venture capitalists, investment bankers, advisors in terms of members of the board of directors, and the universities that trained the management team of these firms. In two earlier papers we demonstrated the feasibility of using this data from U.S. Securities and Exchange Commission filings to study the geography of entrepreneurial support networks in three industries: semiconductors, telecommunications equipment, and biotechnology (Patton and Kenney 2005; Kenney and Patton 2005). This project will extend our initial research in two directions. First, the geographical location of the venture capitalists and other board members will be determined, primarily through Internet search, of approximately 1,700 more firms that have gone public in addition to these three high technology industries studied. This permits comparison of the support network in venture capital-financed and non-venture capital-financed firms. The size and rich detail of the database will permit the exploration of a number of hypotheses suggested by theoretical work and qualitative research regarding the geography of the support services for entrepreneurial clusters across all industry groups. Second, Internet search will significantly enhance the existing data on the educational background of the managers and directors of these newly public firms. This project, by examining the contribution of the university to start-ups through its role as educator, can provide insight into the role of particular universities in local entrepreneurship. In addition to the research questions and hypotheses explored through this database as described in the project description, this database will be converted into a relational database from its current Excel format. Through a password-protected system, we will place this database online, so that bona fide researchers may have access to it.

Entrepreneurship is increasingly understood as a spatially and organizationally embedded activity in which entrepreneurs mobilize other actors and resources in their quest to establish a new organization (Aldrich 1999; Rocha 2004; Thornton 1999).1 In certain regions, organizations dedicated to assisting entrepreneurial startups have been established and a wide variety of authors have noticed the role that networks of these support organizations have played in supporting entrepreneurship (Kenney and von Burg 1999; Saxenian 1994). Most recent research on what we have termed “entrepreneurial support networks” has been qualitative or examines only a single constituent of the network, usually the venture capitalists. These studies have provided suggestive insights and hypotheses, but suffer from the problem of a relatively small N, while the quantitative studies do not capture the multiple constituents of the organizational network facilitating new firm formation. Outside the myriad studies of entrepreneurship in biotechnology (see, for example, Bagchi Sen 2004; Cooke 2004; Powell et al. 2002; Zucker et al. 1998), the geography of the networks of organizations supporting hightechnology entrepreneurship has received much less attention In two earlier papers we demonstrated the feasibility of using data from U.S. Securities and Exchange Commission filings to study the geography of entrepreneurial support networks in three industries: semiconductors, telecommunications equipment, and biotechnology (Patton and Kenney 2005; Kenney and Patton 2005). We showed that different industries, and actors, exhibited different network geographies. Our Economic Geography paper suggests that the support network for IT-based firms differs markedly from those in biotechnology. This project will extend our initial research in two directions. First, we will add all the other firms that went public from 1996 to 2000, i.e., approximately 1,700 more firms from a wide variety of industries (see Table One below). This permits comparison of the support network in venture capital-financed and nonventure capital-financed firms. Also, the database contains every Internet IPO from the inception of the industry allowing an evolutionary analysis of the geography of an entrepreneurial support network over time. Our larger database will provide much larger Ns through the inclusion of firms from the other IT sectors and other biomedical fields. The size and rich detail of the database permits the exploration of a number of hypotheses suggested by theoretical work and qualitative research regarding the geography of the support services for entrepreneurial clusters. The second extension over previous research is the addition of data that will permit mapping of the university as a provider of trained personnel and this facet of the university’s role as a constituent of the entrepreneurial support network. For approximately 70 percent of the focal firm’s management team, the current database includes their educational history. This allows an examination of the most fundamental contribution the university makes to entrepreneurship, namely the initial training of personnel. Because the names of the universities are available it is possible to trace the flow of individuals through space. An example of the questions that can be answered is the role of second-rank regional universities in the formation of entrepreneurial firms. It will be possible to measure the level of education and location of contributing 1

On social embeddedness, see Granovetter (1985). On the embeddedness of economic activity in a regional context, see Storper and Salais (1997). For a critique of the networks-and-embeddedness framework see Peck (2005).

1

universities for firms in different industries. This work will provide new insights into the labor market geography of these elite entrepreneurial managers. In the database and methodology section, there is a more complete description of the database, however the important point is that it is now largely complete. The funding is requested to: First, complete the database (remaining tasks discussed below); second, fund our research using the database and; third complete the conversion of the database into a mySQL database with a graphical user interface. With this completed, the database will be placed on-line (password-protected) to encourage further research on the geography of entrepreneurship and entrepreneurial support networks. Literature Review and Hypotheses to be Tested Marshall (1890) recognized many, but not all, industries exhibit clustering (see also, e.g., Jacobs 1969; Perroux 1950; Porter 1990; Storper and Walker 1989). The tendency of similar and related types of economic activity to concentrate geographically is a widely observed phenomenon over time and across countries. These concentrations are most frequently referred to as clusters or industrial districts, and the relationship between innovation, entrepreneurship, and geography of these clusters has attracted the attention of academics from a variety of disciplines. The importance of industrial clustering for innovation and firm growth has been widely noted beginning with Alfred Marshall (1890), extending through Michael Piore and Charles Sabel (1984) to contemporary geographers (Malecki 1980; Scott 1993, 2005; Storper 1995). However, others have noticed that some industries do not exhibit strong clustering -- in those cases it might be expected that the geography of the entrepreneurial support network would be different. Krugman (1991), in a restatement of Alfred Marshall's observations from 1890 and the research of economic geographers, observes that there are three distinct reasons for localization. First, clusters allow for a large market of workers with highly specialized skills. For many firms such skilled labor can only be found within a cluster. Second, a cluster supports a wide range of specialized local suppliers of inputs and services. Certain specialized inputs may be far more readily available in clusters. This study differs from others in that we examine the providers of services to the entrepreneurial process, or what might be termed the “production of new firms.” The third aspect of industrial clusters that has drawn the particular interest of academics and policy makers is the tendency for knowledge to spill over from firms and individuals within a cluster. A large number of empirical studies demonstrate that knowledge spillovers are geographically mediated, which is to say that proximity matters (Jaffe et al., 1993; Audretsch and Feldman, 1996; Zucker et al, 1998). However, these studies do not address the far more complicated issue of entrepreneurship, whereby an individual or a team establishes a new enterprise. The spatial sciences offer a promising route for understanding the process of entrepreneurship and the actors that have emerged to support entrepreneurship. Scott and Storper (2003) recast Krugman’s three reasons with greater attention to the role of ambiguous information and face-to-face contacts within a geographically limited region. Each of these three factors underlying geographic concentration, they argue, has the effect of creating positive externalities for both firms and workers (p.583).

2

Drawing upon Polanyi et al. (1971) and, more recently, Nonaka and Takeuchi (1995), a good deal of recent theorizing explicitly recognizes the role of knowledge, in particular tacit knowledge as distinguished from codifiable knowledge, as being a key determinant in the geography of innovation (Gertler 2003: 79). Codifiable knowledge is ubiquitous and travels well, while tacit knowledge is related to a specific setting and a particular place, or geography (Maskell and Malmberg 1999). The entrepreneurial act catalyzes knowledge, explicit and tacit, using them as a basis for praxis. Pinch et al. (2003) combine this insight with the knowledge-based view of the firm to distinguish between component knowledge, which refers to scientific and technical skills, and architectural knowledge, which refers to routines and organizational capabilities. In this view simple geographical proximity is not sufficient for knowledge transfer within clusters unless those firms within a cluster share a system for organizing component knowledge (Pinch et al. 2003: 382). This concept of a region differs from the learning regions perspective where spatial proximity is essential for the effective transmission and production of tacit knowledge (Gertler 2003). In more recent work, Storper and Venables (2004) argue that all of the reasons originally cited by Marshall as to why economic activity tends to cluster depend critically on the most fundamental aspect of physical proximity -- face-to-face contact. All of the various theories of agglomeration, beyond those based on simple scale economies, ultimately rely on the importance of human interaction and the face-to-face contacts that provide a richness of information sharing and a type of emergent knowledge not available through other media. Depending on the nature of the information exchanged, tacit, codified or transparent, and the market environment in which these exchanges take place, stable or fluid, the need for geographical proximity ranges from low to very high. These recent works point to different implications for clustering depending on the nature of the knowledge exchanged among economic actors. Our proposal is to enhance and expand an existing database of a set of such actors involved in the initial public offerings of new firms to examine some of these implications. This database will provide information on the members of a firm's entrepreneurial support network; venture capitalists, law firms, investment bankers, independent directors, as well as detailed information about the education background of the firm's board and management. Clusters, Innovation, and Support Networks The literature investigating clusters has found that both traded and untraded interdependency benefits are responsible for the success of these regional economic agglomerations (Storper 1995; Porter 1990). Michael Porter (1998), in conclusions not very different from those of economic geographers such as Walker (1985) and Storper and Walker (1989), identified three broad ways in which clusters affect competition. First, the externalities present in a cluster operate to increase the productivity of all member firms. Many of these are what Storper (1995) terms “untraded dependencies.” Second, the cluster accelerates the innovative capacity of its firms. Third, the concentration of specialized skills and knowledge within the cluster reduces the barriers to entry and facilitates new firm formation. Baptista and Swann (1998) found evidence to suggest that all of these factors are at work and that innovation, firm entry and growth are all stronger in clusters. In qualitative work directed at particular industrial clusters,

3

Saxenian (1994) and Kenney and von Burg (1999) have argued that these benefits are responsible for the success of innovative regions such as Silicon Valley and Route 128. These are what Markusen (1999) characterizes as Italianate industrial districts, though she recognizes the emphasis on firm growth for an eventual stock offering or merger in clusters such as Silicon Valley is quite different from the Italian districts. Perhaps the most notable feature of these types of clusters is the degree of innovation associated with them. As early as Malecki (1980), it was observed that there was regional variation in R&D and as a result there were significant differences between the ability of regions to innovate. Cooke et al. (1997) terms these regional systems of innovation (see also Cooke and Morgan 1990). In earlier work, Florida and Kenney (1988b) conceptualized these locations as hosting “social structures of innovation.” Feldman (1994), using data collected by the Small Business Administration, found that innovations in particular industries were highly concentrated in states such as California and Massachusetts for electronics and New Jersey and New York for medical instruments. Jaffe et al. (1993) found that patents will cite other patents originating in the same location more frequently than patents outside the location controlling for the existing geography of related research activity. Almeida and Kogut (1997) obtained the same results in studying patents in the semiconductor industry, indicating that patent citations are localized. These empirical studies demonstrated that knowledge spillovers are geographically mediated and implies that innovatory activity clusters. Audretsch and Feldman (1996) show that even after the geographical concentration of production is accounted for, innovations are found to cluster in industries where industry R&D, skilled labor, and university research are important inputs. Zucker et al. (1998) show a strong relationship between the location of biotechnology intellectual human capital in the form of star research scientists and startup biotechnology firms. The literature has clearly noted the importance of local business services (Porter 1998; Bennett, et al. 1999; Muller and Zenker 2001; Scott 2002). Not surprisingly, small firms are more likely to depend on their networks. A number of studies examining the changing geography of business or producer services conclude that they are increasingly concentrated in larger cities (Gillespie and Green 1987; Tickell 1999; Keeble and Nachum 2002). Bennett and Smith (2002) found that nearly all business service firms catering to small firms were located within a 50-kilometer radius. All services are, of course, important to the small firm, but some service providers such as financial supporters can also provide legitimacy beyond simply the funds to continue the innovative process. All firms are embedded in social and business networks, both local and extralocal (Fritsch 2001) and the character of this embeddedness can vary within industries (Hendry et al. 2000), and almost certainly does between industries. A number of industrial districts such as the film industry have networks of co-located agents with differing skills that cooperate in making a movie (Christopherson and Storper 1986; Scott 2002). Sydow and Staber (2002), drawing upon structuration theory, examine project networks in two different regions producing content for the German television industry. They find in the more successful region an infrastructure of public support contributed to its success.

4

The significance of the support networks in Silicon Valley is illustrated by the fact that the region has the largest concentrations of high technology firms in both biotechnology (Kenney 1986; Powell et al. 2002; Zucker et al. 1998) and a number of information technology sectors such as the Internet (Zook 2002; Kenney 2002), hard disk drives (McKendrick et al. 2000; McKendrick 2004), and semiconductors (Almeida and Kogut 1999). There is ample qualitative work that unambiguously argues that benefits from clustering exist for small startup firms, however there have been fewer studies attempting to map the support networks in these clusters and measure their impact quantitatively. In one of attempt to quantify the benefits of location, Deeds et al. (1997) found that for a biotechnology firm not located in one of the eight regions with the highest concentration of biotechnology firms, relocation to San Francisco would have on average allowed it raise $6.3 million more capital at its IPO. In the case of biotechnology, location appears to have a concrete wealth effect. Eisenhardt and Schoonhoven (1990) found that Silicon Valley semiconductor startups that had received venture capital funding grew significantly faster than those located in other regions. Because of the broad range of industries in our database we can test hypotheses regarding the impact and interaction of variables such as location in a cluster and the types of support networks for various industries. In reviewing the recent literature on location and innovation Maryann Feldman (2000, p. 389) concludes that, "The consensus is that knowledge spillovers are geographically bounded within a limited space over which interaction and communication is facilitated, search intensity is increased, and task coordination is enhanced." This was echoed in a study of the concentration of the British motor sport industry in a small region of England that did not find thick layers of institutions as described by Cooke and Morgan (1990), though they found a highly competitive, but also interactive environment that they characterized as a community of knowledge (Wenger and Lave 1993; Pinch and Henry 1999), thereby echoing Brown and Duguid (2000) who termed Silicon Valley a “network of practice.” This project maps the spatial aspects of the institutional infrastructure of the intermediaries involved in supporting entrepreneurial startups. Previous studies have highlighted the role and proximity of various institutions such as venture capitalists, specialized legal firms, accounting firms, members of scientific advisory boards etc. Most, but by no means all, of the quantitative studies have either used aggregate data such as regional R&D expenditures or venture capital investments and thus were not focused on the individual firms, or they have focused on single firms and the spatial dimensions of their relationships with financial agents such as venture capital or scientific advisory board members. These studies have been limited because they have not measured the multiple intermediaries assisting the new firm formation and growth process. The Constituents of an Entrepreneurial Support Network The natural assumption is that a startup would most benefit from having access to support network members in close proximity. The face-to-face contact would facilitate the transfer of difficult to communicate tacit knowledge, thereby reducing information

5

asymmetries. A support network has many members including law firms, venture capitalists, investment bankers, accounting firms, universities, research institutes, and various types of advisors. Our database includes geographic information on five members of the support network: lawyers, venture capitalists, investment bankers, advisors in terms of members of the board of directors, and the universities that trained the management team (the role of the university will be discussed in a separate section).2 On the basis of the literature and our previous research, we postulate two general hypotheses: H1: The proximity of support network members to the focal firms will be directly related to the concentration of similar firms in a location. H2: The constituents and location of the entrepreneurial support networks differ by industry. In the U.S., legal assistance is usually the first type of support an entrepreneur is likely to seek. As Suchman (2000) shows in the case of Silicon Valley firms, lawyers are likely to not only provide assistance in the technicalities of incorporation and intellectual property law if the entrepreneur is separating from another organization, but also provide introductions to venture capital firms and other business services. In this sense, the entrepreneurial firm’s outside legal advisor is likely to be a “counselor.” The legal profession is, quite naturally, local in practice even though most large legal firms have numerous branch offices. In keeping with the results of our previous research, we would hypothesize that: H3: A firm's legal counsel will be the most proximate of the four network actors across all industries. For firms in more remote locations such legal advice may not be available. Therefore, we would hypothesize: H4a: The startup would have an relatively inexperienced lawyer or H4b: The startup will seek counsel from a legal firm in a distant cluster of startups in the same industry. Access to capital is often necessary for most startups. Two financial intermediaries, the venture capitalists and investment bankers, are included in this study. The role of spatial and network proximity for financial intermediaries has attracted significant attention recently. Agnes (2002) in a study of the interest rate swaps industry found that "different financial services have differing informational contents, with implications for the local embeddedness of financial services firms." This is confirmed by the finding that formal institutional networks are actually embedded in informal relationships through which transactions and information flows (Clark and O’Connor 2

We have extracted the firm’s accountants, but we are unable to unambiguously establish their location. This information could be used to establish whether certain accounting firms have established strong regional or industrial practices.

6

1997; Pryke and Lee 1995; Thrift and Leyshon 1994). In other words, as Uzzi (1999) illustrates, formal relationships such as the lender-borrower relationship are embedded in a social context, and this social embeddedness, what Garud and Jain (1996) in their study of technological change refer to as “just-embedded,” actually reduces the cost of loans and reduces risk. Abolafia (1997) confirms this finding qualitatively by showing that the necessity of social and physical proximity differs by the nature of the financial product. So, for highly standardized products such as listed equities and government bonds, traders need not be proximate, whereas for other more idiosyncratic financial instruments proximity is of greater importance. There is an ample literature suggesting that venture capital investing, being an idiosyncratic financial practice with high uncertainty, is a locally embedded practice.3 This is true, because of the importance of monitoring and informal assistance functions that go beyond simply providing capital (Florida and Kenney 1988a; Gompers and Lerner 1999; Black and Gilson, 1998). Indeed, Greenwald and Stiglitz (1992) observed that the venture capital industry shares the same aspects of localism that characterized early financial market communities. Because venture capital firms operate in a tightly knit community and have detailed information on the projects they fund and the industries in which their entrepreneurs operate, there is a strong reliance upon trust and reputation in the relationship between venture capitalists and the firms they fund. The critical venture capitalists in a startup are what are termed the “lead” venture capitalists who are the board members and those most intimately involved in monitoring and assisting the firm (Gompers and Lerner 1999). For this reason they are the most important venture capitalists in a startup and the ones that one would expect to be local. Therefore, we hypothesize: H5: Regardless of its location, the focal firm will have at least one local VC board member H6: The proportion of VC directors serving focal firms will be correlated with the concentration of firms in its industry in the location. Investment bankers are central to the IPO process as the lead investment banker is responsible for organizing the syndicate that takes the firm public (Ritter 1998). They provide advice to the startup firm on the most auspicious time to go public among other services. In the U.S., New York City is the center of the investment banking industry, however the larger firms have branches and there are many smaller local investment banks. From our previous research, we found that in the case of the more dispersed biotechnology industry, many of the lead investment bankers were located in New York City despite the fact that there were very few biotechnology startups in New York City. In this case the financial cluster in New York City appeared to have a greater attractive force than the smaller clusters. For this reason, we hypothesize that:

3

Unfortunately, with only a few exceptions regional scientists have not actively studied the geography of the venture capital industry (see, Mason and Harrison 2002; Zook 2002).

7

H7: Firms within the Silicon Valley and Boston regions will be served by investment banks/bankers in those regions. Those outside these regions will tend to be served by New York. The intellectual attention to clusters has been intense, however there are industries such as retail that do not appear to be as clustered geographically. For whatever reason (s), the location of successful startups in these industries has received far less scholarly attention. Our database permits us to examine their support networks, also. Given the dispersed nature of these startups, we would hypothesize: H8: In more dispersed industries, the financial intermediaries, such as investment bankers and venture capitalists, are more likely to be located in the New York financial cluster. The final members of the entrepreneurial support network are the non-venture capital board members. This is an important, but polyglot group, that can include representatives from critical customers or suppliers, academics, etc. They are also the most widely dispersed group. H9: As the most polyglot group, the board members will be the most dispersed support network members. The previous hypotheses are generated from the existing literature and our previously reported research. However, during the data analysis we expect that new understandings and conceptualizations of support networks for entrepreneurship will emerge because of the richness of the data and the fact that we will be exploring successful startups in industries such as retail and manufacturing, which have received only minimal study, as the scholarly literature is focused on venture capital-financed, technology firms because of the accessibility of information through the VentureXpert and patent databases. Comparing the Geography of Entrepreneurial Support Networks Across Industries Studies of clusters have explicitly recognized horizontal clustering, i.e. between competing firms, and vertical clustering, i.e. separate segments of the value chain (for example a supplier and an assembler).4 Unfortunately, there have been few empirical studies of the location of business services that support entrepreneurship. The common assumption is that these services are highly clustered in close proximity to the startup (Sorenson and Stuart 2001; Zook 2002). This project explores this assumption in greater detail by examining four members of the support network: law firms, venture capitalists on the firms’ boards of directors, investment bankers, and the independent members of the firms’ boards of directors across all industry groups. Our previous work with just three high technology industries produced important results. In examining the support network of the semiconductor industry it was found that the geographical proximity between these actors and the firms they support varied significantly, with a firm's legal counsel being the most proximate, followed by 4

In economic geography, see such classic texts as Storper and Walker 1989 and Perroux 1988.

8

investment bankers, venture capitalists, and independent directors (Patton and Kenney, 2005). This was followed by a comparison of the semiconductor industry with biotechnology and telecommunications equipment (Kenney and Patton, 2005). The results obtained indicated that conclusions about networks drawn from single industries can be misleading. The economic geography of the biotechnology support network differed significantly from the networks in semiconductors and telecommunications equipment, in that biotechnology had a far more dispersed entrepreneurial support network structure than did the two electronics related industries. Clustering in biotechnology (and, perhaps, in other industries, such as medical devices) may be inhibited because the source of entrepreneurs in this industry is not as concentrated in existing firms, but rather might be more dependent upon universities. Also, it is possible to speculate that biotechnology firms may not be as dependent upon each other and on proximate suppliers, thus limiting the centripetal forces drawing them together. Thus, the normal cluster effects may not be sufficient to create the winner-takeall regional dynamics that scholars have predicted. If the centripetal forces are not as strong for the firms and the sources of knowledge are more dispersed, then it is not surprising that the constituents of the entrepreneurial support networks are not as concentrated in close proximity to the firms. Oddly enough, this suggests that the literature on biotechnology clustering overestimates the significance of the centrifugal forces in biotechnology, while explaining why every local jurisdiction believes it can build a biotechnology cluster, i.e., they will be able to encourage the development of biotechnology firms, but are unlikely to receive the benefits of cluster synergies. This database makes it possible to conduct cross-industry comparisons of entrepreneurial firms at the same moment in their life cycle. To the extent industries rely upon university research they, and their support networks, will tend to be more dispersed. If the sources of knowledge that are the seeds of new firms are dispersed, so will these new firms other things being equal. Start-ups in industries whose clusters are coincident with New York, Boston, and Silicon Valley may incorrectly imply a necessary proximity to the venture capitalists and investment bankers found in these regions. Separating out these other influences on the geography of new firm formation cannot be made without the cross-industry comparisons suggested here. Universities, Human Capital, and New Firm Formation Universities are recognized as an essential institution in many of the most celebrated innovative regions of the world (Etzkowitz 2002, 2004; Storper and Salais 1997). Indeed, the evolution and development of Silicon Valley has been profoundly influenced by the role of Stanford University (Kenney 2000; Lecuyer 2005; O’Mara 2004; Saxenian 1994).5 Stanford University, along with the University of California, Berkeley, are two of the crucial institutions within Silicon Valley that observers have termed an "ecosystem", a "social structure of innovation", or an "incubator region"

5

The influence of Stanford University through the role of Frederick Terman, department chair of the electrical engineering department and university provost, would be hard to exaggerate given his encouragement of William Hewlett, David Packard, and the Varian brothers to establish firms in the area (Lecuyer 2005).

9

(Bahrami and Evans 2000; Florida and Kenney 1990; Eisenhardt and Schoonhoven 1990). There has been a growing appreciation of the university's role in transferring the knowledge developed through public research to the larger society, particularly the private sector in the form of commercial ventures. Scholarly interest in this role has tended to focus on the university's direct role in promoting new firm formation on the basis of university inventions. In particular, this interest has centered on university spinoffs in high-technology, science-based fields, such as biotechnology and, more recently, nanotechnology. The university's direct role in promoting such high technology start-ups, though, is but one of several channels by which the university's knowledge is transmitted to commercial enterprises. These other transfer mechanisms include journal publications, participation in professional conferences, professorial consulting, and hiring graduates in addition to the entrepreneurship of professors. This project extends previous research by examining the contribution of the university to start-ups through its role as educator. As such our unit of analysis in this section is the educational histories of the individual managers and directors of the focal firm. University education is an important component of the careers of the skilled personnel that fuel the U.S. start-up economy. Indeed, the mobility of individuals from universities and established firms to new firms may be the primary conduit not only of knowledge spillovers in knowledge-based industries (Audretsch and Keilbach 2004), but also of the managerial expertise needed to render this knowledge commercially viable (for a discussion of this in biotechnology, see Kenney 1986; Higgins 2005). Regions able to attract and/or retain these commercially creative individuals are likely to experience out-size economic success. The database can provide insight into the role of particular universities in local entrepreneurship. For example, it is often believed that the graduates of local second-tier universities located in regions with high-levels of specific local industrial knowledge benefit from a “regional” effect. Our database allows this to be tested in terms of entrepreneurship, i.e., will local second-tier universities located in an entrepreneurial region be more entrepreneurial. This can be stated formally: H10: Firms within clusters will have a higher proportion of management receiving their degree from a local university than will firms outside of a cluster. An extension of this is: H10a: The geographic "reach" of universities is greater for prestigious universities than for less prestigious colleges. (e.g., Santa Clara University vs. Stanford). Database and Methodology Every firm wishing to go public must file a prospectus with the U.S. Securities and Exchange Commission (SEC) prior to its initial offering of stock (IPO). The database has been constructed from the SEC filings of all the domestic firms that made an initial public offering from June 1996 through 2000. The initial list of the IPOs over this time period contained almost 2,000 firms. To restrict our attention to start-up firms only, we

10

excluded from this list spin-offs, mutual funds, real estate investment trusts, unit offerings, reverse LBOs, foreign ADR listings, and so-called small business IPOs. The resulting population contains 1,843 firms as shown in Table One. The scheme for categorizing these firms by industry group includes the complete range of SIC codes. The particular industry groups reflect the makeup of the firms that went public over this time period. The IPO is a defining event in the history of any firm. In return for being able to raise capital from the public, the firm must conform to the reporting and transparency requirements imposed by the SEC under the Securities Act of 1933. One of the primary objectives of the Securities Act of 1933 is to require companies making a public offering of their securities to publicly disclose relevant business and financial information about their company so that potential investors can make an informed investment decision regarding the offering. To achieve this end the 1933 Act requires companies going public to file disclosure documents with the SEC, the most important of which are the general form S-1 registration statement and the 424B prospectus. These documents, in effect, provide a detailed snapshot of the firm at the time it goes public, and it is these documents which provide the basis of the data used in this study. The sheer magnitude of the task is worth a brief, but incomplete, description. A PERL script parsed over 80 gigabytes of data. The resulting data when cleaned included 20,000 managers and directors (each with their appropriate title) and including over 10,000 full educational histories that included approximately 30,000 separate title entries. There were nearly 4,000 individual lawyers, their law firms, and their complete address. The sheer magnitude of data points overwhelmed Excel (so the data is held in four different Excel spreadsheets), and must be moved to mySQL to be used effectively. The database contains variables that pertain to the 1,843 firms going public. It includes the firm name, state of incorporation, address, stock exchange, ticker, offering size, initial share price, shares outstanding, underwriter discount, and company auditor. There are two law firms in this database; one is the law firm representing the firm in the offering, and the other is the law firm representing the underwriters (this is important because it is an effective proxy for the location of the lead investment banker who is an individual, e.g., Frank Quattrone). The information here includes each lawyer's name, their law firm, and the address of the law firm. The variables related to the underwriters include: the name of all underwriters of the IPO, and the number of shares each underwriter has agreed to purchase for the offering. The database also includes information on each manager and independent director including their name and age, their current position, their term of office, whether they were a firm founder, and the prospectus biography of every director and executive officer that describes their business experience over the previous five years. The employment of each manager prior to their joining the firm is recorded, as is the year in which they first established a relationship with the firm. In addition, frequently the individual provides a description of their education in their prospectus biography. This information is most often provided by firms in high technology industries where the level of education attained by the management team and Board of Directors is an important indicator of competence. In cases where this information is missing, we propose to use an Internet search to improve the response rate. On the basis of our work so far we estimate that slightly more than half of all biographies provide information on the person's education.

11

Because the database is comprised of only the most successful startups, i.e., those making an IPO, this is an elite and censored group of startups. This objection is valid, unfortunately there is no similarly detailed database available for all startups. To provide some control, we gather the location of all venture capital-funded firms from the VentureXpert database and this will be compared to the firms in our database to discover any systematic geographical bias that might exist. To measure the distance between various actors Microsoft MapPoint software is used. It converts each address in our database into latitude and longitude coordinates so that the distance between any two points on the globe can be determined. This allows us to construct a matrix of distances for any two addresses in the database.

12

Table 1. The Population of Firms Going Public between June 1996 through 2000 Firm Industry Group Standard Industrial Classification (SIC) Count Code 48 Semiconductors 3674 69 Biotechnology 283X 61 Telecommunications Equipment 366X 35 General Instruments 382X 47 Medical Instruments 384X 33 Computers 357X 65 Computer Systems 7373 54 Computer Programming 7371 60 Computer Services 7374, 7376, 7377, 7378, 7379 192 Software 7372 76 Telephone and Telegraph 4812, 4813, 4822 38 Broadcasting and Services 4832, 4833, 4841, 4899 46 Machinery All 35XX (except 357X), 37XX 44 Manufactured Goods I All 2XXX (except 20XX, 21XX, 283X) 54 Manufactured Goods II All 3XXX (except 35XX, 36XX, 37XX, 382X, 384X) 28 Electronic Equipment All 36XX (except 366X, 3674) 53 Retail Trade I 52XX, 53XX, 56XX, 57XX, 591X, 594X, 598X, 599X 56 Retail Trade II 54XX, 55XX, 58XX, 592X, 593X, 596X 62 Wholesale Trade 50XX, 51XX 70 Finance I 601X, 602X, 603X, 6712 43 Finance II 606X, 608X, 609X, 61XX 25 Transportation and Services 40XX, 41XX, 42XX, 44XX, 45XX, 47XX 17 Electricity, Gas, and Sanitation 49XX 47 Services I All 7XXX (except 73XX) 46 Services II 4311, all 8XXX (except 80XX, 82XX, 873X), 9XXX 117 Business Services 732X, 734X, 738X 41 Health Services 80XX 45 Securities, Insurance, and Real Estate 62XX, 63XX, 64XX, 65XX 11 Holding and Investment All 67XX (except 6712, 6726, 6798) 46 Advertising, Employment, and Leasing 731X, 733X, 735X, 736X 73 Information Retrieval 7375 67 Education and Research 82XX, 873X 8 Agriculture 0XXX 22 Oil, Gas, and Mining 10XX, 12XX, 13XX, 14XX, 46XX 14 Construction 15XX, 16XX, 17XX 22 Food and Tobacco 20XX, 21XX 8 Unclassified n.a. 1843 TOTAL

13

Database Preparation Tasks Remaining There are two remaining tasks to complete the database: 1) For approximately 46 percent of the directors, their educational background is not available in their biography. Our previous work indicates that Internet searches can fill this gap. An Internet search of the biotechnology industry yielded education histories for 92 out of the 134 missing managers and directors (68.7%) not reporting on education. The success rate in finding missing education histories for semiconductors and telecommunications equipment was 83.6% and 53.7%. However, in the interim Zoominfo.com, an online biography website, has emerged and provides extensive information garnered through “web scraping.” This should increase our conversion rate and speed the process. 2) The addresses of directors is occasionally given in the prospectus, but this information is provided for only a small proportion of directors. Therefore the remainder of these addresses must be found through Internet searches. All of the independent directors in the database are divided into two mutually exclusive sets: the directors affiliated with a venture capital (VC) firm, and the remaining board members that were not venture capitalists. Internet searches have been shown to be quite effective. For high technology firms less than 5% of the VC directors could not be located. A precise address was also found for 95.6%, 86.9%, and 81.9% of the non-VC directors in semiconductors, telecommunications equipment, and biotechnology, respectively. Project Outcomes We plan three outcomes for the project: 1) We will publish articles on the research results in scholarly journals in the spatial sciences (e.g., Economic Geography, Regional Studies, etc.), small business studies (e.g., Small Business Economics, Journal of Business Venturing, etc.), and research management (e.g., Research Policy) journals. 2) Through a password-protected system, we will place the final database online, so that bona fide researchers can use it. Other researchers will be able to use the database with the only proviso being that they should properly cite the creators. 3) We will provide an online user’s guide to the database.

14

Abolafia, M. 1997. Making Markets: Opportunism and Restraint on Wall Street (Cambridge, MA: Harvard University Press). Agnes, P. 2002. "The 'End of Geography' in Financial Services: Local Embeddedness and Territorialization in the Interest Rate Swaps Industry." Economic Geography 76: 347-366. Aldrich, H. 1999. Organizations Evolving (London: Sage Publications). Almeida, P. and B. Kogut. 1999. “Localization of Knowledge and the Mobility of Engineers in Regional Networks.” Management Science 45, (7): 905-917. Almeida, P. and Kogut, B. 1997. "The Exploration of Technological Diversity and the Geographical Localization of Innovation." Small Business Economics 9: 21-31. Audretsch, D. B., and Maryann P. Feldman. 1996. "R&D Spillovers and the Geography of Innovation and Production." American Economic Review 86 (3): 630-640. Audretsch, D. B. and M. Keilbach. 2004. "Entrepreneurship Capital and Economic Performance.” Regional Studies 38, (8): 949 – 959. Bagchi-Sen, S. 2004. “Firm-specific characteristics of R&D collaborators and noncollaborators in US biotechnology clusters and elsewhere.” International Journal of Technology and Globalisation 1, (1): 92-117. Bahrami, H. and S. Evans. 2000. "Flexible Recycling and High-Technology Entrepreneurship." In Martin Kenney (ed.) Understanding Silicon Valley. (Stanford: Stanford University Press): 165-189. Baptista, R. and P. Swann. 1998. “Do Firms in Clusters Innovate More?” Research Policy 27: 525-540. Bennett, R. J. and C. Smith. 2002. “The influence of location and distance on the supply of business advice.” Environment and Planning A 34, (2): 251 - 270. Bennett, R. J., D. J. Graham, and W. Bratton. 1999. "The location and concentration of businesses in Britain: Business clusters, business services, market coverage and local economic development". Transactions of the Institute of British Geographers 24: 393420. Black, B. S. and R. J. Gilson. 1998. "Venture Capital and the Structure of Capital Markets." Journal of Financial Economics 47: 243-277. Brown, J. S. and P. Duguid. 2000. "Mysteries of the Region: Knowledge Dynamics in Silicon Valley." In C. Lee et. al. (eds.) The Silicon Valley Edge., (Stanford: Stanford University Press): 16-45.

1

Christopherson, S. and M. Storper. 1986. 'The city as studio: the world as back lot: the impact of vertical disintegration on the location of the motion-picture industry', Environment and Planning D: Society and Space 4 (3): 305-320. Clark, G. L. and K. O'Connor. 1997. "The Informational Content of Financial Products and the Spatial Structure of the Global Finance Industry." In K. R. Cox (Ed.) Spaces of Globalization Reasserting the Power of the Local (New York: Guilford press). Cooke, P. 2004. “The molecular biology revolution and the rise of bioscience megacentres in North America and Europe.” Environment and Planning C: Government and Policy 22 (2) 161 – 177. Cooke, P., M. G. Uranga, and G. Etxebarria. 1997. “Regional Innovation Systems: Institutional and Organizational Dimensions.” Research Policy (26): 475-491. Cooke, P. and K. Morgan. 1990. Learning through Networking (Cardiff: University of Wales Press). Deeds, D., D. DeCarolis and J. Coombs. 1997. “The Impact of Firm-specific Capabilities on the Amount of Capital Raised in an Initial Public Offering: Evidence from the Biotechnology Industry.” Journal of Business Venturing 12 (1): 31-46. Eisenhardt, K. M. and C. B. Schoonhoven. 1990. "Organizational Growth: Linking Founding Team, Strategy, Environment, and Growth among U.S. Semiconductor Ventures, 1978-1988." Administrative Science Quarterly 35: 504-529. Etzkowitz, H.A. 2002. MIT and the Rise of Entrepreneurial Science (London: Routledge). Etzkowitz, H.A. 2004. “The Evolution of the Entrepreneurial University.” International Journal of Technology and Globalisation 1 (1): 64-77. Feldman, M. P. 2000. "Location and Innovation: The New Economic Geography of Innovation, Spillovers, and Agglomeration." In Gordon L. Clark, Maryann P. Feldman, and Meric S. Gertler (eds.) The Oxford Handbook of Economic Geography (Oxford University Press, New York): 373-394. Feldman, M. P. 1994. The Geography of Innovation. Boston: Kluwer Academic Publishers. Florida, R. L. and M. Kenney. 1988a. "Venture Capital, High Technology and Regional Development." Regional Studies 22: 33-48. Florida, R. L. and M. Kenney. 1988b. “Venture Capital-Financed Innovation and Technological Change in the U.S.” Research Policy, 17, 119-37.

2

Florida, R L., and M. Kenney. 1990. The Breakthrough Illusion: Corporate America’s Failure to Move from Innovation to Mass Production. New York: Basic Books. Fritsch, M. 2001. “Co-operation in Regional Innovation Systems.” Regional Studies 35 (4): 297-307. Garud, R. and S. Jain. 1996. "Technology Embeddedness." In J. Baum and J. Dutton, (eds.) Advances in Strategic Management (13): 389-408 (Connecticut: JAI Press). Gertler, M. S. 2003. "Tacit Knowledge and the Economic Geography of Context, or The Undefinable Tacitness of Being (There)." Journal of Economic Geography 3: 75-99. Gompers, P. A. and J. Lerner. 1999. The Venture Capital Cycle (Cambridge Mass. : MIT Press). Granovetter, M. 1985. "Economic Action and Social Structure: The Problem of Embeddedness." American Journal of Sociology 91: 481-510. Greenwald, B., and J. E. Stiglitz. 1992. "Information, Finance and Markets: The Architecture of Allocative Mechanisms." Industrial and Corporate Change 1: 37-63. Gillespie, A. E. and A. E. Green. 1987. “The Changing Geography of Producer Services Employment in Britain.” Regional Studies 21: 397–411. Hendry, C., J. Brown and R Defillippi. 2000. “Regional Clustering of High Technologybased Firms: Opto-electronics in three countries.” Regional Studies 34 (2): 129-144. Higgins, M. C. 2005. Career Imprints: Creating Leaders Across an Industry San Francisco: Jossey-Bass. Jacobs, J. 1969. The Economy of Cities (New York: Random House). Jaffe, A. B., M. Trajtenberg, and R. Henderson. 1993. "Geographic Localization of Knowledge Spillovers as Evidenced by Patent Citations." Quarterly Journal of Economics, 108: 577-598. Keeble, D. and L. Nachum. 2002. “Why Do Business Service Firms Cluster? Small Consultancies, Clustering and Decentralization in London and Southern England.” Transactions of the Institute of British Geographers 27, (1): 67-90. Kenney, M. and U. v. Burg. 1999. "Technology and Path Dependence: The Divergence between Silicon Valley and Route 128." Industrial and Corporate Change 8 (1): 67-103. Kenney, M. and D. Patton. 2005. "Entrepreneurial Geographies: Support Networks in Three High-Technology Industries." Economic Geography 81 (2):201-228.

3

Kenney, M. 1986. Biotechnology: The University-Industry Complex. Yale University Press, New Haven, CT. Kenney, M. (ed.). 2000. Understanding Silicon Valley (Stanford: Stanford University Press). Kenney, M. 2002. "The Growth and Development of the Internet in the United States." In The Global Internet Economy, ed. Bruce Kogut, 69-108. Cambridge: MIT Press. Krugman, P. 1991. Geography and Trade. The MIT Press, Cambridge, MA. Lecuyer, C. 2005. Making Silicon Valley: Innovation and the Growth of High Tech, 1930-1970. The MIT Press, Cambridge, MA. Malecki, E. 1980. "Dimensions of R&D Location in the United States." Research Policy 9: 2-22. Markusen, A. 1999. "Sticky Places in Slippery Spaces: A Typology of Industrial Districts." In T. Barnes and M. Gertler (eds.) The New Industrial Geography: Regions, Regulation and Institutions (London: Routledge): 98-126. Marshall, Alfred. 1890. Principles of Economics. (London: Macmillan & Company). Maskell, P. and A. Malmberg. 1999. "Localized Learning and Industrial Competitiveness." Cambridge Journal of Economics 23: 167-185. Mason, C. M. and R. T. Harrison. 2002. “The Geography of Venture Capital Investments in the UK.” Transactions of the Institute of British Geographers 27 (4): 427-451. McKendrick, D. G. 2004. “Leveraging Locations: Hard Disk Drive Producers in International Competition.” In Locating Global Advantage M. Kenney (ed.) (Stanford: Stanford University Press). pp. 142-174. McKendrick, D. G., R. Doner, and S. Haggard. 2000. From Silicon Valley to Singapore: Location and Competitive Advantage in the Hard Disk Drive Industry (Stanford: Stanford University Press). Muller E. and Zenker, A. 2001. "Business Services as Actors of Knowledge Transformation: The Role of KIBS in Regional and National Innovation Systems". Research Policy 30: 1501-16. Nonaka. I. and H. Takeuchi. 1995. The Knowledge-Creating Company (Oxford: Oxford University Press).

4

O'Mara, M. P. 2004. Cities of Knowledge: Cold War Science and the Search for the Next Silicon Valley (Princeton: Princeton University Press). Patton, D. and M. Kenney. 2005. "The Spatial Configuration of the Entrepreneurial Support Network for the Semiconductor Industry." R&D Management 35 (1): 1-16. Peck, J. 2005. "Economic Sociologies in Space." Economic Geography 81, (2): 129-176. Perroux , F. 1950. "Economic Space: Theory and Applications." The Quarterly Journal of Economics 64 (1): 89-104. Perroux, F. 1988. "The Pole of Development’s New Place in a General Theory of Economic Activity". In Regional Economic Development, ed. B. Higgins and D. J. Savoie, 48-76. Boston: Unwin Hyman. Pinch, S., N. Henry, M. Jenkins, and S. Tallman. 2003. "From 'Industrial Districts' to 'Knowledge Clusters': A Model of Knowledge Dissemination and Competitive Advantage in Industrial Agglomerations." Journal of Economic Geography 3: 373-388. Pinch, S., and Henry, N. 1999. "Paul Krugman’s Geographical Economics, Industrial Clustering and British Motor Sport Industry". Regional Studies 33: 815-27. Piore, M. and C. Sabel. 1984. The Second Industrial Divide (New York: Basic Books). Polanyi, K. et al (eds.). 1971. Trade and Markets in Early Empires. Chicago, IL: Henry Regnery. Porter, M. 1998. “Clusters and the New Economics of Competition.” Harvard Business Review (November-December): 77-90. Porter, M. 1990. The Competitive Advantage of Nations (New York: Free Press). Powell, W. W., K. W. Koput, J. I. Bowie, and L. Smith-Doerr. 2002. "The Spatial Clustering of Science and Capital: Accounting for Biotech Firm-Venture Capital Relationships." Regional Studies 36 (3): 291-305. Pryke, M. and R. Lee. 1995. "Place Your Bets: Towards an Understanding of Globalization, Socio-Financial Engineering and Competition within a Financial Centre." Urban Studies 32: 329-344. Ritter, J. R. 1998. "Initial Public Offerings." Contemporary Finance Digest 2: 5-30. Rocha, H O. 2004. “Entrepreneurship and Development: The Role of Clusters.” Small Business Economics 23: 363–400. Saxenian, A. 1994. Regional Advantage. Harvard University Press, Cambridge, MA.

5

Scott, A. J. 2005. “Entrepreneurship, Innovation, and Industrial Development: Geography and the Creative field Revisited.” http://www.diw.de/deutsch/produkte/veranstaltungen/Regionalization_InnovationPolicy_Conf2004/papers/scott.pdf. Scott, A. J. 1993. Technopolis: High-Technology Industry and Regional Development in Southern California (Berkeley: University of California Press). Scott, A. J. 2002. "A New Map of Hollywood: The Production and Distribution of American Motion Pictures." Regional Studies 36: 957-76. Scott, A. J. and M. Storper. 2003. "Regions, Globalization, Development." Regional Studies 37 (6): 579-593. Sorenson, O. and T. E. Stuart. 2001. “Syndication Networks and the Spatial Distribution of Venture Capital Investments.” American Sociological Review 106 (6): 1546-1588. Storper, M. 1995. "The Resurgence of Regional Economies, Ten Years Later: The Region as a Nexus of Untraded Interdependencies," European Urban and Regional Studies 2 (3): 191-221. Storper, M. and R. Salais. 1997. Worlds of Production: Frameworks of Action in the Economy (Cambridge: Harvard University Press). Storper, M. and A. J. Venables. 2004. "Buzz: Face-to-Face Contact and the Urban Economy." Journal of Economic Geography 4: 351-370. Storper, M, and R. Walker. 1989. The Capitalist Imperative: Territory, Technology, and Industrial Growth. London: Basil Blackwell. Suchman, M. C. 2000. "Dealmakers and Counselors: Law Firms as Intermediaries in the Development of Silicon Valley". In Understanding Silicon Valley, ed. M. Kenney, pp.7197. Stanford, Calif.: Stanford University Press. Sydow, J. and U. Staber. 2002. “The Institutional Embeddedness of Project Networks: The Case of Content Production in German Television.” Regional Studies 36 (3): 215227. Thornton, P. H. 1999. “The Sociology of Entrepreneurship.” Annual Review of Sociology 25: 19-46. Thrift, N. and A. Leyshon. 1994. "A Phantom State? The Detraditionalization of Money, the International Financial System and International Financial Centres." Political Geography 13: 299-327.

6

Proposal Title: The Geography of Entrepreneurial Support Networks Context Statement NATIONAL SCIENCE FOUNDATION GEOGRAPHY AND REGIONAL SCIENCE PROGRAM CONTEXT STATEMENT -- SPRING 2006 ADVISORY PANEL MEETING The Geography and Regional Science (GRS) Advisory Panel met on April 20-21, 2006, to evaluate proposals that were submitted in response to the January 15, 2006, target date. The GRS Advisory Panel considered 131 individual proposals in this round. Not all of these proposals required separate action by the panel, however, as 20 were secondary components of collaborative research proposals, with principal investigators from two or more institutions proposing collaborative work together through separately submitted proposals. These secondary components did not require independent consideration. Of the remaining 111 proposals, 49 were jointly reviewed with one or more other NSF programs. All other proposals were evaluated solely by GRS. The GRS Advisory Panel evaluated proposals using the NSF criteria of intellectual merit and broader impacts as articulated in the Grant Proposal Guide (NSF 04 23). All regular research proposals were sent to at least six external reviewers and to at least two advisory panel members. Every proposal was evaluated using at least three written reviews; most had five to seven written reviews. All written reviews (from both external reviewers and panelists) were made available to panelists prior to the panel meeting. On the basis of written reviews and extensive discussion during the panel meeting, the GRS Advisory Panel made recommendations regarding GRS and NSF support for each proposal. For the 111 proposals, the panel judged 57 proposals to be "fundable," and it declined to recommend funding for 52 proposals. Of the "fundable" regular proposals, the panel recommended that 11 be given high priority for support, 6 be given high-medium priority, 14 be given medium priority, 10 be given medium-low priority, and 16 be given low priority. Depending on budgetary constraints and on the willingness of other programs to participate in joint funding of meritorious proposals, GRS program directors anticipate recommending funding for between 12 and 15 projects. Verbatim copies of all completed reviews and any relevant panel summaries are made available to principal investigators via the FastLane system. GRS panelists and program officers do not necessarily agree with or endorse all statements by reviewers. Some reviews may contain irrelevant comments, which are not used in making final funding decisions. Panel members are asked to reflect on the substance of the written reviews and to draw generalizations that extend beyond the summary ratings alone. The reviews and panel summary contain evaluative material and constructive suggestions that may be used by the principal investigator in the conduct of future research regardless of whether a proposal is awarded or declined.

Panel Summary #1 ========================== INTELLECTUAL MERIT: POSITIVE ASPECTS OF THE PROPOSAL AND PROPOSED RESEARCH: The principal strength of this study is that it potentially clarifies inter-firm, interorganizational linkages in agglomeration economies by focusing on the geography of entrepreneurial support networks. The proposal is scientifically tight and the deliverable (a database) is clear. The PI will construct a database of IPO firms that detail their relations with four supports actors. This database will become available for bona fide researchers. The budget is reasonable and the research plan is generally tight. SHORTCOMINGS AND WEAKNESSES OF THE PROPOSAL AND PROPOSED RESEARCH: The study may have over-emphasized localization economies at the expense of urbanization economies given its inter-industry focus. It is also unclear how analysis of the hypotheses will proceed. ========================== BROADER IMPACTS: POSITIVE ASPECTS OF THE PROPOSAL AND PROPOSED RESEARCH: The study should be of interest to business organizations and business scholars in addition to geographers because it teases out the variation in geographic effect on entrepreneurial support networks. SHORTCOMINGS AND WEAKNESSES OF THE PROPOSAL AND PROPOSED RESEARCH: None identified. ========================== SYNTHESIS COMMENTS: A study that seeks to construct a database of supporting networks underscoring successful entrepreneurial (IPO) firms will enhance economic geographers' ability to conduct potentially exciting work on inter-firm linkages in the future. It is unclear if the study will

result in theoretical verification since few details on how the hypotheses will be tested are provided. However the panel recommends medium-priority fundability given the relatively low sum of money requested for the study. ========================== PANEL RECOMMENDATION (CHECK ONE): [ ] Fund with High Priority [X] Fund with Medium Priority [ ] Fund with Low Priority [ ] Decline to Fund [ ] Defer Funding Recommendation

Panel Recommendation: Fund with Medium Priority

REVIEW#1: What is the intellectual merit of the proposed activity? The merit of this proposal is quite good. It addresses an important(and unanswered) question concerning the spatial relationship between successful entrepreneurial firms and their support networks. It is a question that many people have written about and this proposal seeks to address it at a more general level than many of the regional cases studies have done. It also seems (judging from the budget) to involved undergraduates in the research process, albeit mostly in the process of data collection. It is primarily focused on the collection of data (although a number of hypotheses are laid out it is not clear how the researcher will proceed in answering them). This is a minor point as one can easily imagine how (given the data collected) the research would proceed. Nevertheless it is primarily a database building proposal which the author plans to share with the research community. What are the broader impacts of the proposed activity? The outcome of this project will likely impact beyond geography and regional science, most likely in economics and business studies. Particularly the willingness to make the database available to other researchers (although I don't believe process of doing this was outlined to a great extent) will greatly increase the impact. I would imagine that the data would be of use to researchers far beyond those interest in spatial relationships, e.g., network analysis would have a fieldday with this kind of data.

Summary Statement The PI does a good job of laying out the issues of this topic and I agree it an important topic and think this is a good proposal overall. My concerns (and why I did not rate it excellent) has to do with the reliability of the spatial data for lawyers, VCs, IB and directors. Although I am not as fluent with the S1/424B forms as the PI, I think that information on the lawyers and VCs is solid but am concerned with the location of investment bankers and directors. At the heart is the issue of multi-establishment firms and which address is used in the filings. For example if the IB headquartered in NY but the lead investment banker is in a satellite office in Palo Alto what address is used? I'm not sure if the law firm for the IB is a good proxy (p. 11) or not. Additionally I am concerned about the validity of determining the location of a director (circa 1996-00) based on Internet searches today. I assume that they will be looking at career histories but how easy is that to find. Again the multi-locational issue comes up. Someone might be working for HP but where? Zoominfo.com sounds promising but there wasn't much information on it. It sounds like a constant process of judgement calls by an undergraduate who may (or not) be well prepared to do it. A lesser concern is also how easy it is to identify the correct director and match them to their educational history. I don't think these issues are debilitating but I would have like to have known more about the nuts and bolts of how this takes place. I think this approach (gathering firm/individual data on the web) has great promise and this project is also useful for its ability to act as a pilot project of sorts for this kind of approach. Also why stop with firms with IPOs in 2000?

REVIEW#2: What is the intellectual merit of the proposed activity? The PI requests support to 1) complete a database of companies went public from 1996 to 2000; 2) study the geographical dimensions of the entrepreneurial support networks of these start-up companies; and 3) make available the database to the public to support further research. The proposal provides adequate background on the entrepreneurial network research, and the lack of quantitative and geographical dimensions in the research as most works are qualitative in nature. The mapping of the five members of the supporting network is interesting and quite systematic. Studying the roles of the university members should cast new insights on this often ignored aspect. The hypotheses

developed throughout the proposal are quite reasonable, though some seem a bit simplistic. However, the major weakness of this proposal is the lack of analysis. Though hypotheses were described, no information was provided to explain how the data will be analyzed and how the hypotheses will be tested. The methodology section is just the method to complete the database. The measurement of distance among actors is quite simplistic and computing the distance matrices of large size is not trivial. The bulk of the work seems to be doing internet search for information not provided by the filings of IPO.

What are the broader impacts of the proposed activity? This proposal is based upon previous work on three industries: semiconductors, telecommunications and biotechnology, and these industries have different patterns in their network geography. The potential to extent the current work to other industries should have significance outcomes to describe the network geography more comprehensively across industries. The roles of university, which have not been studied much in the past, should be better revealed through this study from a support network perspective. The obvious benefit is making the database available to the public, though the database does not capture all the start-ups, but only those went public. Summary Statement The creation of the data is no doubt a very solid contribution of the proposed work. However, how the data will be used, i.e. analyze, in supporting the proposed work is not clear. Potential results should add to the literature.

REVIEW #3: What is the intellectual merit of the proposed activity? This propsal supports research, including that of the PI, demonstrating differences among industries in the network geographies of high technology industries. The proposed research would make a significant contribution to the literature by further illuminating the nature and origin of differences among industry clusters. Because the proposed study uses data from U.S. Securities and Exchange Commission filings, however, the insights will be limited to limited to those firms that have gone public. This is relatively small proportion of SMEs in industrial clusters. I think the PI could benefit from consulting work by Boschma and Lambooy, and Rutherford to place his research in the broader picture of entrepreneurship and industrial clusters.

What are the broader impacts of the proposed activity?

This research will significantly expand knowledge of differences among industrial clusters and the capacity for entrepreneurship wihin differently composed firm neworks Summary Statement

This proposed research project builds on previous research by the PI and so, can be confidently supported. As the PI suggests, the IPO is "a defining event". This analysis will focus on a slect but very critical set of firms and can be expected to produce valuable reserch and policy relevant results.

REVIEW#4: What is the intellectual merit of the proposed activity? The proposed research will make significant contribution to theory: the role of networks of producer services (e.g., venture capital, legal counsel) in advancing the business of high technology. Several papers have been published by Powell and his group in this area. However, continuous validation of sources of support in high tech business is necessary through empirical analysis. Martin Kenney has demonstrated through his research that he is more than capable of pulling together large databases and deciphering meaningful relationships among different actors in the high tech industry. He is also well cited in Economic Geography. What are the broader impacts of the proposed activity?

The broader impact will be felt in understanding what facilitates IPOs and their role/potential as new product developers. Some explanations will also be offered in understanding what is called the "Fuzzy Front End" - this is the stage (e.g., start-ups) where new concepts are developed which could be radical, incremental, or platform technologies (see Peter Koen's work).

REVIEW #5: What is the intellectual merit of the proposed activity? It contributes to the economic geography literature on agglomeration economies and nature of inter-firm spillovers. What are the broader impacts of the proposed activity? Its focus on IPO firms will provide insights on the support (institutional) networks of successful firms. This is a topic that will not only interest economic geographers but business and economics scholars as well. Summary Statement This research proposes to investigate the geography of four major institutional network actors that support successful startup firms. Successful startup firms are defined as firms that have gone public from 1996 to 2000 (i.e. IPO). The PI seeks funding to construct a relational database that will potentially be accessible to researchers. A study of supporting networks is a worthwhile endeavor given a relatively rich literature on clusters and agglomeration economies. Much of this literature has yet to elaborate on the geography of inter-organizational, inter-firm linkages and spillovers other than to theorize that they exist arising from the nature of knowledge flows and face-to-face interactions. A more detailed empirical investigation into these conceptual insights particularly from the vantage point of industrial variation should potentially add to this body of knowledge. The PI is well-qualified to undertake the research. Comments for improvement are provided below. The research draws solely from SEC data. By definition, SEC firms are geared towards the stock market and equity funding. In turn the New York area is a central pole for the two principal network actors involved in equity funding, namely lawyers and investment bankers. From this point of view, some of the hypotheses (e.g. H8) are not particularly insightful, and are rather obvious. Furthermore, the nature of data implies that other forms of financial actors (e.g banks) are excluded as support networks, yet local banks across U.S. cities play a role in some industries like the retail sector. Hence generalizations about the geography of entrepreneurial support networks need to be caveated here with IPO firms more likely to be driven by the geography of venture capitalists and investment banks (these networks in turn are less spatially dispersed than those of banks). Next, it is not clear what the 4 network actors in H3 are. In the first para of that page, the PI lists 5 including universities. Third, this is not a major point but strictly speaking, some of the hypotheses listed are not hypotheses. In scientific hypothesis testing, the statement should contain variables and relationships that can be evaluated, and in this case, statistically evaluated. I suggest using propositions.

Fourth, the literature review misses the more important concept of agglomeration economies for this type of research, that is, urbanization economies. The actors being examined are from different industries (legal, finance, and some 30+ industries in Table 1), and conceptualizing the interactions and linkages in terms of localization economies is less insightful here. The literature has also gone beyond tacit=local and codified=global for knowledge forms; hence it is no longer true that entrepreneurial firms may only act on local tacit knowledge (see Bathelt, H., Malmberg, A., Maskell, P. (2004) in Progress in Human Geography, 28: 32-56). Indeed a potential contribution of this research is that proximity is less important for some industries because local tacit knowledge plays a less important role

REVIEW#6: What is the intellectual merit of the proposed activity? The proposed activity will create a database for the study of the geography of knowledge networks associated with IPO supported start up firms. The researchers will test hypotheses concerning the geography of these networks and inter-industry differences in network patterns. This is creative work on the cutting edge of firm-based innovation studies. What are the broader impacts of the proposed activity? The research activity will generate insight into a topic of important economic significance, namely, the role that the organization and geography of knowledge networks play in supporting successful start-up firms. Summary Statement This is a very straight forward and well prepared proposal. The value generated from the modest funding requested is very high. The authors have found a way to bring together existing sources of data to provide a basis for systematic quantitative study for a topic that to date has been dominated by case study analysis. Everything about this proposal is solid: the authors have a strong recent track record of productive research, the research builds in stepwise fashion on earlier research, the authors are going to make the data available for other scholars to use, the review of the literature is appropriate, the authors involve students in the research, the budget is appropriate. I can't wait to see the data. This is terrific proposal. Fund it if you can. I apologize for not writing a longer review but I don't think there is much more to say. The strength of the proposal lies in its focus on a particular research task (assembly of the database). The only reason I could see for not giving this project the highest ranking is if you are looking to fund work that is 'out of the box' - breaking open new fields. Otherwise I would say that this is exactly the kind of focused project that pushes research forward. Thanks for the opportunity to comment.

APPENDIX

iii. Kenney, Martin and Patton, Don. Geography and Regional Science: The Geography of Entrepreneurial Support Networks. (revised, funded version, reviews)

The role of social actors supporting entrepreneurship has recently been recognized as critical, yet their geography is not well understood. This is true despite the fact that the literature has recognized the importance of entrepreneurship for the birth and growth of industrial clusters. Drawing upon the results of research in high-technology clusters, there has been a general assumption that supporting actors, such as lawyers, venture capitalists, and others, should be located in close proximity to the entrepreneurs because of the necessity for face-to-face interaction to transmit tacit knowledge. More recently, some spatial scientists have remarked that even in highly clustered industries characterized internally by “local buzz,” “global pipelines” to actors outside the cluster exist and are important. The proximity and role of these support network actors across industries and regions will be explored through testing a variety of propositions derived from the literature on clusters and entrepreneurship. Chi square analysis will be used primarily for statistically testable propositions based on nominal level data regarding the proximity of startup firms and members of their entrepreneurial support network, while a hazard model will be used in examining the spatial influences on the speed with which startups go public. The project draws upon a unique database that the PI has been building on the entrepreneurial support network of approximately 2,400 firms that went public between 1996 and 2005 in the United States. This database indicates the exact location of the startup and its lawyer, venture capitalists and other members of its board of directors, and its lead investment banker – all key contributors to the birth, growth, and maturation of an entrepreneurial firm. In addition to the scientific contribution, this project will make two other important contributions: First, for practitioners interested in regional economic development, the results will suggest industries within which successful entrepreneurial firms can be created despite a dearth of local support actors. This is particularly important for regions not as richly endowed with venture capital and other support network actors. Second, it will make available, via an accessible website, the database with its numerous variables of geographic interest to other scholars. This will permit other researchers to build and extend our work with strict comparability of data.

Joseph Schumpeter (1947) trenchantly observed that the entrepreneur1 usually did not have capital or even excellent managerial skills – the job of the entrepreneur was to attract these resources. Today, entrepreneurship is understood as an embedded activity in which entrepreneurs mobilize other actors and resources to support their quest to establish a new organization, i.e., they create a new network (Aldrich 1999; Rocha 2004; Thornton 1999).2 Geographers and spatially-oriented social scientists have shown that there is a proximityrelated effect upon the networks entrepreneurs mobilize (Acs and Armington 2006; Florida and Kenney 1988a, 1988b; Malecki 1990; Romanelli and Schoonhoven 2001; Sorenson and Audia 2000; Suarez-Villa 1989). On the other hand, there is also a recognition that entrepreneurs may also have extra-local supporters. Using a unique database containing all the firms newly listed on U.S. stock exchanges from 1996-2005, we propose to explore the location of four members of what we term a startup’s entrepreneurial support network (ESN): the listing firm’s law firm, venture capitalist(s) (if there is one on the board of directors), the lead investment banker, and other members of the board of directors.3 This research will explore when and under what conditions specific actors that support successful entrepreneurship are proximate versus distant. In certain regions, organizations dedicated to assisting entrepreneurial startups have been established and a wide variety of authors have remarked upon the role that these support organizations have played in supporting entrepreneurship (e.g., Kenney and von Burg 1999; Nijkamp 2003; Saxenian 1994). Noticeably, in many other regions some or all of these ESN constituents are missing. Most recent research on ESNs has been qualitative or quantitatively examined only a single constituent of the ESN, usually the venture capitalists (Sorenson and Stuart 2001; Zook 2002; Suchman 2000 studied Silicon Valley law firms), because of the data access constraints (a notable exception is the research on biotechnology particularly Powell et al. 1996, 2002). The quality and completeness of our data allows us to accurately ascertain the location of four (if the firm did not receive venture capital, three) constituents of the ESN and compare them across industries. The existence of clusters of entrepreneurial firms and their ESNs is well-established. More recent research and theorizing has suggested that firms in clusters are not self-sufficient (Malecki 2000; Oinas and Malecki 2002), but rather have important extra-local linkages. Bathelt et al. (2004) suggested that there was “local buzz” and “global pipelines.” By pipelines, Owen-Smith and Powell (2002) meant the information transmission mechanism by which firms in a local cluster accessed knowledge (and other resources) from outside the region. Even in the highly clustered semiconductor industry, there are extra-local linkages. For example, Kenney and Patton (2005) found that there was a complex relationship between local and extra-local resource acquisition that differed by the location of the startup. In locations having munificent resources (on environmental munificence, see Freeman and Hannan (1983)) there was great clustering. But even in the most munificent clusters, extralocal support was in evidence. Interestingly, such concentrations did not preclude entrepreneurship in other regions. Some entrepreneurs overcame their locational liability. Since we have the locations of more than one constituent we can compare the local versus 1

We use the singular for “entrepreneur” for convenience’s sake, but recognize that many, if not most startups, are established by entrepreneurial teams. 2 On social embeddedness, see Granovetter (1985). On the embeddedness of economic activity in a regional context, see Storper and Salais (1997). For a critique of the networks-and-embeddedness framework, see Peck (2005). 3 We have extracted the firm’s accountants, but we are unable to unambiguously establish their location. However, the information in the database could be used to infer whether certain accounting firms have established strong regional or industrial practices.

1

extra-local distinction for different support services thus providing insight into what support is more easily supplied through pipelines versus accessed locally. Most of the previously reported research has drawn upon qualitative interviewing which, while providing powerful suggestive insights, is limited to relatively small numbers of cases because of the time-intensive nature of the methodology. Additionally, nearly all of the studies that are quantitative do not capture the multiple constituents of the organizational network facilitating new firm formation. Outside the myriad studies in the biotechnology industry (see, for example, Bagchi-Sen 2004; Cooke 2004; Powell et al. 1996, 2002; Zucker et al. 1998), the geography of organizational ESNs supporting entrepreneurship has received less attention. This is particularly true in non-technology fields, such as retail and finance, where there has been little research on how a startup acquires the resources necessary to build a firm (a significant exception is the literature on the spatial dimensions of ethnic entrepreneurship, e.g., Razin 1989; Boyd 2003).4 The reasons for the relative lack of attention to the non-technology startups may be the belief that they are not as significant to the economy or that they evince little clustering. In terms of significance, this dismissal may not be justified when one considers the importance of earlier startups such as Wal-Mart or McDonalds. Our results will provide insight into whether these industries and their ESNs exhibit clustering. It may also be that the concentration of these non-technological industries is more strongly conditioned by what have been termed “urbanization economies,” which demonstrates that the urban environment is most conducive to new firm formation (Dumais et al. 2002; Henderson 1997; Jacobs 1969). This project extends previous technology firmcentric research to these largely neglected entrepreneurial firms. In addition to advancing our knowledge of the geography of entrepreneurship, this project will have policy-relevant results. Local and regional governments today understand that smokestack-chasing is not the best economic development policy and are encouraging entrepreneurship and cluster formation. Unfortunately, not all regions have strong ESNs. The results will suggest fields within which successful firms can be created despite an absence of support actors, and identify industries within which extra-local ESNs may overcome the lack of local networks. To illustrate, Kenney and Patton (2005) found that biotechnology firms going public that were from outside the traditional clusters were able to draw upon venture capitalists, investment bankers, and board members from outside their region suggesting that lack of ESN access were not binding constraints in biotechnology. The final result of the project stems from our belief that progress in the social sciences have been retarded in comparison to the biological and natural sciences because databases are often not publicly available. We disagree with this practice, and will build a web interface and user’s guide to the MySQL database enabling other bona fide scholars to utilize the data obtained by this effort. Literature Review and Propositions to be Examined The spatial sciences offer a promising route for understanding the process of entrepreneurship and the actors that support entrepreneurship. Concentrations of similar and related types of economic activity are most frequently referred to as clusters or industrial districts, and the relationship between innovation, entrepreneurship, and geography of these clusters has attracted the attention of academics from a variety of disciplines. Marshall (1890) 4

For ethnic entrepreneurship, only the explicitly geographical analyses are consulted. The enormous literature based on the obvious fact that ethnic entrepreneurship tends to be clustered in ethnic enclaves is not included in our discussions.

2

recognized many, but not all, industries exhibit clustering beyond those that simply occur in urban areas due to population density (see also, e.g., Jacobs 1969; Perroux 1950; Porter 1990; Storper and Walker 1989). The importance of industrial clustering for innovation and firm growth has been widely noted beginning with Alfred Marshall (1890), extending through Michael Piore and Charles Sabel (1984) to contemporary geographers (Malecki 1980; Maskell 2001; Scott 1993, 2006; Storper 1995). However, others have remarked that certain industries such as retail do not exhibit as strong clustering suggesting that the geography of their ESNs would be different. Krugman (1991), in a restatement of Alfred Marshall's observations from 1890 and the research of economic geographers, observes that there are three distinct reasons for localization. First, clusters allow for a large market of workers with highly specialized skills. In some cases, such skilled labor can only be found within a cluster. Second, a cluster supports a wide range of specialized local suppliers of inputs and services. Certain specialized inputs may be far more readily available in clusters. This study differs from others in that we examine the providers of services to the entrepreneurial process, or what might be termed the facilitators for the production of new firms. The third aspect of industrial clusters that has drawn the particular interest of academics and policy makers is the tendency for knowledge to spill over from firms and individuals within a cluster. Alan Scott and Michael Storper (2003: 583) recast Krugman’s three reasons with greater attention to the role of ambiguous information and face-to-face contacts within a geographically limited region. Each of these three factors underlying geographic concentration, they argue, has the effect of creating positive externalities for both firms and workers. A large number of empirical studies demonstrate that knowledge spillovers are geographically mediated, which is to say that proximity matters (Jaffe et al. 1993; Audretsch and Feldman 1996; Zucker et al. 1998). In general, these studies do not address the more complicated issue of entrepreneurship, whereby an individual or a team uses knowledge to establish a new enterprise. Drawing upon Polanyi et al. (1971) and, more recently, Nonaka and Takeuchi (1995), a good deal of recent theorizing explicitly recognizes the role of knowledge, in particular tacit knowledge as distinguished from codifiable knowledge, as being a key determinant in the geography of innovation (Gertler 2003: 79). This literature concludes that, in general, codifiable knowledge travels well, while tacit knowledge is related to a specific setting and a particular place, or geography (Maskell and Malmberg 1999). The entrepreneurial act catalyzes knowledge, explicit and tacit, locally and uses them as a basis for praxis. Pinch et al. (2003) utilize the knowledge-based view of the firm to distinguish between component knowledge, which refers to scientific and technical skills, and architectural knowledge, which refers to routines and organizational capabilities. In this view simple geographical proximity is not sufficient for knowledge transfer within clusters unless those firms within a cluster share a system for organizing component knowledge (Pinch et al. 2003: 382). This concept of a region differs from the learning regions’ perspective where spatial proximity is essential for the effective transmission and production of tacit knowledge (Gertler 2003). In more recent work, Storper and Venables (2004) argue that all of the reasons originally cited by Marshall as to why economic activity tends to cluster depend critically on the most fundamental aspect of physical proximity -- face-to-face contact. All of the various theories of agglomeration, beyond those based on simple scale and scope economies, ultimately rely on the importance of human interaction and the face-to-face contacts that provide a richness of information sharing and a type of emergent knowledge not available through other media. Depending on the nature of the information exchanged, tacit, codified

3

or transparent, and the market environment in which these exchanges take place, stable or fluid, the need for geographical proximity ranges from low to very high. However, the general expectation from this literature would be that in the case of supporting entrepreneurship the face-to-face component would be great. These recent works point to different implications for clustering depending on the nature of the knowledge exchanged among economic actors. Our project is meant to enhance and expand an existing database of a set actors involved in the initial public offerings of new firms to examine these implications. This database will provide information on the members of a firm's ESN; venture capitalists, law firms, investment bankers, independent directors. Clusters, Innovation, and Entrepreneurial Support Networks The literature investigating clusters has found that both traded and untraded interdependency benefits are responsible for the success of these regional economic agglomerations (Storper 1995; Porter 1990). Michael Porter (1998), in conclusions not very different from those of economic geographers such as Walker (1985) and Storper and Walker (1989), identified three broad ways in which clusters affect competition. First, the externalities present in a cluster operate to increase the productivity of all member firms. Many of these are what Storper (1995) terms “untraded dependencies.” Second, the cluster accelerates the innovative capacity of its firms. Third, the concentration of specialized skills and knowledge within the cluster reduces the barriers to entry and facilitates new firm formation. Baptista and Swann (1998) found evidence to suggest that all of these factors are at work and that innovation, firm entry and growth are all stronger in clusters. In qualitative work directed at particular industrial clusters, Saxenian (1994) and Kenney and von Burg (1999) suggest that these benefits were significant contributors to the success of innovative regions such as Silicon Valley and Route 128. In these Italianate industrial districts, a term used by Markusen (1999) though she recognizes their emphasis on firm growth for an eventual stock offering or merger distinguishes them from the traditional Italian system, an entire ecosystem has developed to assist entrepreneurship. Perhaps the most notable feature of these types of clusters is the degree of innovation associated with them. As early as Malecki (1980), it was observed that there was regional variation in R&D and as a result there were significant differences between the ability of regions to innovate. Cooke et al. (1997) among others, term these regional systems of innovation (see also Cooke and Morgan 1990), but the ESN actors were combined with various other actors, such as suppliers. Focussing on entrepreneurship, Florida and Kenney (1988b) conceptualized these locations as hosting “social structures of innovation.” Feldman (1994), using data collected by the Small Business Administration, found that innovations in particular industries were highly concentrated in states such as California and Massachusetts for electronics and New Jersey and New York for medical instruments. Jaffe et al. (1993) found that patents will cite other patents originating in the same location more frequently than patents outside the location when controlled for the existing geography of related research. Audretsch and Feldman (1996) show that even after the geographical concentration of production is accounted for, innovations are found to cluster in industries where industry R&D, skilled labor, and university research are concentrated. Zucker et al. (1998) find a strong relationship between the location of biotechnology human capital in the form of star research scientists and startup biotechnology firms. Almeida and Kogut (1997) obtained similar results in studying patents in the semiconductor industry, indicating that patent citations are localized. These studies demonstrated that knowledge spillovers are geographically mediated and implies that innovatory activity clusters.

4

ESN actors are service providers. The literature has clearly noted the importance of local business services, and that small firms are more likely to depend on local service providers (Porter 1998; Bennett et al. 1999; Muller and Zenker 2001; Scott 2002). Bennett and Smith (2002) found that nearly all business service firms catering to small firms were located within a 50-kilometer radius. And yet, many studies examining the geography of business or producer services conclude that they are increasingly concentrated in larger cities (Gillespie and Green 1987; Tickell 1999; Keeble and Nachum 2002). All services are, of course, important to the small firm, but some service providers such as financial supporters provide legitimacy beyond simply the funds to continue operations. All firms are embedded in social and business networks, both local and extra-local (Fritsch 2001) and the character of this embeddedness varies across and within industries (Hendry et al. 2000). A number of industrial districts such as the film industry have networks of co-located agents with differing skills that cooperate in making a movie (Christopherson and Storper 1986; Scott 2002). Sydow and Staber (2002), drawing upon structuration theory, examine project networks in two different regions producing content for the German television industry. They find in the more successful region an infrastructure of public support contributed to its success. The significance of the ESNs in Silicon Valley is illustrated by the fact that the region has the largest concentrations of high technology firms in both biotechnology (Kenney 1986; Powell et al. 2002; Zucker et al. 1998) and a number of information technology sectors such as the Internet (Zook 2002; Kenney 2002), hard disk drives (McKendrick et al. 2000; McKendrick 2004), and semiconductors (Almeida and Kogut 1999). There is ample qualitative work that unambiguously argues that clustering benefits small startup firms, however there have been fewer studies attempting to map the ESNs in these clusters and measure their impact quantitatively. In one attempt to quantify the benefits of location, Deeds et al. (1997) found that for a biotechnology firm not located in one of the eight regions with the highest concentration of biotechnology firms, relocation to San Francisco would have on average allowed it raise $6.3 million more capital at its IPO. In the case of biotechnology, location appears to have a concrete wealth effect. Eisenhardt and Schoonhoven (1990) found that Silicon Valley semiconductor startups that had received venture capital funding grew significantly more rapidly than those located in other regions. Because of the broad range of industries in our database we can test propositions regarding the impact and interaction of variables such as location in a cluster and the types of ESNs for firms in various industries. In reviewing the recent literature on location and innovation Maryann Feldman (2000: 389) concludes that, "The consensus is that knowledge spillovers are geographically bounded within a limited space over which interaction and communication is facilitated, search intensity is increased, and task coordination is enhanced." This was echoed in a study of the concentration of the British motor sport industry in a small region of England that did not find thick layers of institutions as described by Cooke and Morgan (1990), though they found a highly competitive, but also interactive environment that they characterized as a community of knowledge (Wenger and Lave 1993; Pinch and Henry 1999), thereby echoing Brown and Duguid (2000) who theorized that Silicon Valley’s strength was based upon interfirm “networks of practice.” This project maps the spatial aspects of the institutional infrastructure of the intermediaries involved in supporting entrepreneurial startups. Previous studies have highlighted the role and proximity of various institutions such as venture capitalists, specialized legal firms, accounting firms, members of scientific advisory boards etc.

5

Most, but by no means all, of the quantitative studies have either used aggregate data such as regional R&D expenditures or venture capital investments and thus were not focused on the individual firms, or they have focused on single firms and the spatial dimensions of their relationships with financial agents such as venture capital or scientific advisory board members. A limit to these studies is that they did not examine the multiple intermediaries assisting the new firm formation and growth process. The Constituents of an Entrepreneurial Support Network The natural assumption from the literature is that ceteris paribus a startup would most benefit from having access to ESN members in close proximity. The face-to-face contact should facilitate the transfer of difficult to communicate tacit knowledge, thereby reducing information asymmetries. On the basis of the literature and our previous research, we examine three general propositions: General Proposition 1: The proximity of ESNs to the startups they are serving will be directly related to the concentration of firms of the same industry in a given location. The literature on industrial clusters, when applied to the constituents of a new firm's ESN implies a trade-off between proximity of a firm to its ESN, and the expertise of the members of the ESN. As a general proposition we expect that clusters would, other considerations being equal, have more experienced ESN actors serving the needs of start-ups, so that their expertise and proximity would make them the obvious choice of start-ups within the cluster. Startups located outside a cluster, on the other hand, must choose between less experienced but physically close service providers, and more experienced but distant providers. The extent to which proximity and expertise are valued determines this choice. General Proposition 2: The proximity of ESN actors to their focal firms will differ by industry. General Proposition 3: The proximity of ESN actors to their focal firms will differ by actor. In the U.S., legal assistance is usually the first type of support an entrepreneur seeks. As Suchman (2000) shows in the case of Silicon Valley firms, lawyers are likely to not only provide assistance in the technicalities of incorporation and intellectual property law, but also provide introductions to venture capital firms and other business services. In this sense, the entrepreneurial firm’s outside legal advisor might more properly termed a “counselor.” Because of their roles as counselors, these lawyers can be expected to have the most knowledge of the firm – a fact that would argue for their close proximity to the firm. Kenney and Patton (2005) found that in three high-technology industries the firm’s counsel was invariably more proximate to the firm than any other ESN constituent. In keeping with the results of our previous research, we would propose that for all industries: P1: A focal firm's legal counsel will be the most proximate of the four ESN actors across all industries.

6

For firms in more remote locations from the center of entrepreneurial activity legal counsel having sufficient expertise to take a firm public may be unavailable locally. Therefore, we would propose in this case that: P2a: The focal firm will accept a local law firm having limited experience with IPOs, or P2b: The focal firm will seek counsel from an experienced legal firm in a distant cluster of startups in the same industry. The expertise of a ESN actor will be measured by their frequency of having provided services to a firm going public in this population of IPOs. Access to capital is vital for most startups. Two financial intermediaries, venture capitalists and investment bankers, are included in this study. The role of spatial and network proximity for financial intermediaries has attracted scholarly attention. Agnes (2000) in a study of the interest rate swaps industry found that "different financial services have differing informational contents, with implications for the local embeddedness of financial services firms." This is confirmed by the finding that formal institutional networks are actually embedded in informal relationships through which transactions and information flows (Clark and O’Connor 1997; Pryke and Lee 1995; Thrift and Leyshon 1994). In other words, as Uzzi (1999) illustrates, formal relationships such as the lender-borrower relationship are embedded in a social context, and this social embeddedness, what Garud and Jain (1996) in their study of technological change refer to as “just-embedded,” actually reduces the cost of loans and reduces risk. Abolafia (1997) confirms this finding qualitatively by showing that the necessity of social and physical proximity differs by the nature of the financial product. So, for highly standardized products such as listed equities and government bonds, traders need not be proximate, whereas for other more idiosyncratic financial instruments proximity is of greater importance. These studies of finance are important for our study because venture capitalists and investment banker involved in IPOs are dealing with idiosyncratic uncertain transactions. There is an ample literature suggesting that venture capital investing, being an idiosyncratic financial practice with high uncertainty, is a locally embedded practice.5 This is true, because of the importance of monitoring and informal assistance functions that go beyond simply providing capital (Florida and Kenney 1988a; Gompers and Lerner 1999; Black and Gilson 1998). Indeed, Greenwald and Stiglitz (1992) observed that the venture capital industry shares the same aspects of localism that characterized early financial market communities (Lamoreaux et al. 2006). Because venture capital firms operate in a tightly knit community and have detailed information on the projects they fund and the industries in which their entrepreneurs operate, there is a strong reliance upon referral and reputation in the relationship between venture capitalists and the firms they fund. The critical venture capitalists in a startup are what are termed the “lead” venture capitalists who are the board members and those most intimately involved in monitoring and assisting the firm (Gompers and Lerner 1999). For this reason they are the most important venture capitalists in a startup and the ones that one would expect to be local, though Kenney and Patton (2005) found that this did differ by industry. Therefore, we propose:

5

Unfortunately, with only a few exceptions regional scientists have not actively studied the geography of the venture capital industry (see, Mason and Harrison 2002; Zook 2002).

7

P3: Regardless of its location, the focal firm will have at least one local venture capitalist on its Board of Directors. P4: The proportion of local VC directors serving focal firms will be correlated with the concentration of firms in its industry in the focal firm's location. Investment bankers are central to the IPO process as the lead investment banker is responsible for organizing the syndicate that takes the firm public (Ritter 1998). They provide advice to the startup firm on the most auspicious time to go public among other services. In the U.S., New York City is the center of the investment banking industry, however the larger firms have branches, some of which specialize in the needs of certain industries, and there are many smaller local investment banks. From our previous research, we found that in the case of the more dispersed biotechnology industry, many of the lead investment bankers were located in New York City despite the fact that there were few biotechnology startups in New York City. The financial cluster in New York City appeared to have a greater attractive force than the lesser clusters and isolated biotechnology firms. For this reason, we will examine whether: P5: Focal firms within the Silicon Valley and Boston regions will be served by investment banks/bankers in those regions. Those outside these regions will tend to be served by New York. There are industries such as retail and banking that appear to not be as clustered geographically as some others in terms of startups. For this reason it is likely that their ESNs will not be as concentrated close to the startups. The thinking behind this proposition is that if there are large numbers of similar startups in one location, it will attract or cause to come into existence the financial intermediaries. Here, we can explore the attractive power of vertical clustering, i.e., close to customers, versus horizontal clustering, i.e., close to similar firms. The initial proposition is: P6a: In more dispersed industries, the financial intermediaries, i.e., lead investment banker and venture capitalists, are more likely to be located in the New York financial cluster. P6b: In concentrated industries, the financial intermediaries, i.e., lead investment banker and venture capitalists, are more likely to be located in close proximity to the industry (ies) they are servicing. The final ESN members are the non-venture capital board members. This is an important, but polyglot group that can include representatives from lead customers or suppliers, academics, etc. The eclectic make-up of this group suggests that it should be the most widely dispersed. In locational terms: P7: These board members will be the most dispersed ESN members. P7a: Academic board members are likely to be located in close proximity to the focal firm.

8

P7b: Corporate board members are likely to have the greatest dispersion of any members. Since non-venture capitalist board members are likely to be the least integrated group in the operations of the firm, they may also be an important source of external information. There is a large literature on board of directors, and while our propositions related to these board members are not fully established we believe that this group is sufficiently important so as to justify further exploration. The literature on clusters suggests that proximity matters (Eisenhardt and Schoonhoven 1990). This database will allow us to test this proposition: P8. Focal firms whose ESN members are in close proximity will reach an IPO more rapidly than those whose ESN members are more dispersed. The previous propositions are generated from the existing literature and our previously reported research. The richness of the data and variety of industries being explored suggests during data analysis, entirely new insights will emerge. This is particularly true in the case of retail, banking, and other services, as the literature has focused on venture capital-financed, technology firms because of the accessibility of information through the VentureXpert and other databases. In the case of the Internet industry, we will have a particularly unique opportunity because we have the entire population of firms that made a public listing. Comparing the Geography of Entrepreneurial Support Networks Across Industries and Over Time Studies of clusters have explicitly recognized horizontal clustering, i.e. between competing firms, and vertical clustering, i.e. separate segments of the value chain (for example, a supplier and an assembler).6 Unfortunately, there have been few empirical studies of the location of business services supporting entrepreneurship. Our previous work with just three high technology industries produced important results. In examining the semiconductor industry ESN it was found that the geographical proximity between these actors and the firms they support varied significantly, with a firm's legal counsel being the most proximate, followed by investment bankers, venture capitalists, and independent directors (Patton and Kenney 2005). This was followed by a comparison of the semiconductor industry with biotechnology and telecommunications equipment (Kenney and Patton 2005). The results obtained indicated that generalizations about ESNs drawn from single industries can be misleading. The economic geography of the biotechnology ESN differed significantly from the ESNs in semiconductors and telecommunications equipment, in that biotechnology had a far more dispersed ESN structure than did the two electronics related industries. Clustering in biotechnology (and, perhaps, in other industries, such as medical devices) may be inhibited because the source of entrepreneurs is not as concentrated in existing firms, but rather more dependent upon university research. Also, it is possible to speculate that biotechnology firms may not be as dependent upon each other and on proximate suppliers, thus limiting the centripetal forces drawing them together. The normal cluster effects may not be sufficient to create the winner-take-all regional dynamics that scholars have predicted. If the centripetal forces are not as strong for the firms and the 6

In economic geography, see such classic texts as Storper and Walker 1989 and Perroux 1988.

9

sources of knowledge are more dispersed, then it is not surprising that the ESNs are not as concentrated in close proximity to the firms. Oddly enough, this suggests that the literature on biotechnology clustering overestimates the significance of the centrifugal forces in biotechnology, while explaining why every local jurisdiction believes it can build a biotechnology cluster, i.e., they will be able to encourage the development of biotechnology firms, but are unlikely to receive as strong cluster synergies. This database makes it possible to conduct cross-industry comparisons at the same moment in their life cycle. To the extent industries rely upon university research they, and their ESNs, will tend to be more dispersed. If the sources of knowledge that are the seeds of new firms are dispersed, so will the startups other things being equal. Start-ups in industries whose clusters are coincident with New York, Boston, and Silicon Valley may incorrectly imply a necessary proximity to the venture capitalists and investment bankers found in these regions. Separating out these other influences on the geography of new firm formation cannot be made without the cross-industry comparisons suggested here. The Data and Database Every firm wishing to go public must file a prospectus with the U.S. Securities and Exchange Commission (SEC) prior to its initial offering of stock (IPO). For our database, we compiled a list of all startups making an initial public offering from June 1996 through 2005. Excluded from this list were spin-offs, mutual funds, real estate investment trusts, unit offerings, reverse LBOs, foreign ADR listings, and so-called small business IPOs. The resulting population is approximately 2,400 startups. The scheme for categorizing these firms by industry group includes the complete range of SIC codes. However, this database can be parsed by other categorizations, such as, for example, Internet-related firms, which includes a number of SIC codes. The IPO is a defining event in the history of a firm. In return for being able to raise capital from the public, the firm must conform to the reporting and transparency requirements imposed by the SEC. The primary objective is to require companies making a public offering to publicly disclose relevant information so that potential investors can make an informed investment decision. These filed disclosure documents provide a detailed snapshot of the firm, and are the source of the data used in this study. The sheer magnitude of the task is worth a brief description. A PERL script must parse over 100 gigabytes of data that is not entirely standardized. The sheer magnitude of data points means that the database must be in the open source database program, MySQL. The strength of MySQL is that it will be easy to provide a web front-end user access website for other academics to use when we provide open access to other scholars. The resulting data when cleaned will include over 25,000 managers and directors (each with their appropriate title) and over 13,000 full educational histories for a total of approximately 30,000 separate title entries. There are nearly 4,000 individual lawyers, their law firms, and their complete address. The database contains variables that pertain to about 2,400 firms including the firm name, state of incorporation, address, stock exchange, ticker, offering size, initial share price, shares outstanding, underwriter discount, and company auditor. There are two law firms in this database; one is the law firm representing the firm in the offering, and the other is the law firm representing the underwriters. The information includes each lawyer's name, their law firm, and the address of the law firm. Because of this inclusion of the lawyer’s name, we do not have to attribute the lawyer’s address to the corporate headquarters, rather we know

10

the lawyer(s)’ mailing address, i.e., their office. This is a major improvement upon earlier studies. The variables related to the underwriters include: all underwriters of the IPO, and the number of shares each underwriter has agreed to purchase. Of particular importance to us is the location of the lead underwriter who is often very influential in the development of the startup in terms of raising institutional funds. Unfortunately, neither that person’s name nor address is given in the SEC filings. However, the investment bank’s lawyer’s address is given. To establish whether this might be used as a proxy for where the lead investment banker (an employee of the investment bank), we interviewed three investment bankers in Silicon Valley, and they stated that this was a good proxy for the location of the investment banker. We also found newspaper articles that identified the lead investment banker for a few IPOs, we then examined the SEC filings and found that the lawyer was indeed located in close proximity. For example, Frank Quattrone of Credit Suisse First Boston (CSFB) handled many technology deals. His office was in Silicon Valley while CSFB is headquartered in New York. The law firms advising his deals were located in Silicon Valley not New York. Despite the fact that this proxy for the location of the lead investment banker is not deterministic, it does contain information that is superior to any other method of locational attribution. The database contains information on each manager and independent director including their name and age, their current position, their term of office, whether they were a firm founder, and a prospectus biography that describes their business experience over the previous five years. The employment of each manager prior to their joining the firm is recorded, as is the year in which they first established a relationship with the focal firm. To measure the distance between various actors Microsoft MapPoint software is used. It converts each address in the database into latitude and longitude coordinates so that the distance between any two points on the globe can be determined. This allows the construction of a matrix of distances for any two addresses. Methodology The empirical data developed in this research project is descriptive, and the population is unique in the sense that they are among the most successful firms because they were able to make a public offering. For this reason the results cannot be uncritically generalized to all startups. Unfortunately, there is no similarly rich database available for all startups. The data being used is not available through any other data source, and the accuracy and detail are unique because of the legal requirements for truthful reporting. It is possible to test comparative propositions such as the ESN for biotechnology firms differs from that of say semiconductor firms, or that firms with ESNs in close proximity to the focal firm require less time to IPO. These results will provide important input and suggest hypotheses to be tested to researchers studying the larger populations of startups. To provide some control for our results, we will use other data sources for comparison. For example, it possible to gather the location of all venture capital-funded firms from the VentureExpert database. There are also measures of all new firm formations from other available databases. These can be used to ascertain if there is any systematic geographical bias that might exist in our more censored population. Most of the statistically testable propositions in this study are based upon nominallevel data. In consideration of the geography of startup firms and the members of their ESN our data assumes the form of classifications; whether the firm is located within a major cluster or not, or whether a venture capitalist is close to its portfolio firm or not. What

11

constitutes closeness may vary with the issue at hand, but we have found that a distance within 50 miles is reasonable. Because we have the exact addresses, it is possible to allocate firms to other geographical classificatory schemes such as SMSAs, counties, or states. In these cases chi-square analysis is the most applicable of tests. As a nonparametric statistic, it makes no assumptions as to the underlying distribution of spatial characteristics. The chisquare test only requires reasonably large sample sizes, which our database provides, and that there is no bias in sample selection. Because our database is the population of all IPOs there is no bias, at least among the firms that have gone public. This type of analysis has been used to examine the spatial proximity of actors in a variety of studies such as Audretsch and Stephan (1996). Table 1 taken from Patton and Kenney (2005) shows results for the semiconductor industry in our database. A low value of 0.96970 for the chi-square statistic in comparing firm lawyers and investment bankers indicates that there is no statistically significant difference in the proximity of these actors to their focal firms. In the case of both firm lawyers and investment bankers, around 3 out of 4 were within 50 miles of all semiconductor firms going public during this period. In the case of venture capitalist directors and non-VC directors around one half of these ESN member were within 50 miles of their focal firms and, contrary to our expectations, there was no statistical difference between their proximity patterns. Table 1. Proximity of Semiconductor IPO Actors to Firms (1996 - 2000) Firm Lawyers Inside 50 miles Outside 50 miles Total Chi-square Significance

35 79.5% 9 20.5% 44

Investme nt Bankers 31 70.5% 13 29.5% 44

Total 66 75.0% 22 25.0% 88 0.96970 n.s

Non-VC Director s 55 50.9% 53 49.1% 108

VC Director s 45 54.9% 37 45.1% 82

Total

4 Actor Total

100 52.6% 90 47.4% 190 0.29202

166 59.7% 112 40.3% 278 13.5674 3 n.s. 0.01 Source: Patton and Kenney (2005) p.10

Without being able to compare the semiconductor industry with other industries it is not clear whether these patterns of ESNs apply generally, or are being driven in part by the semiconductor industry's concentration in Silicon Valley. Over 60 percent of all newly public firms during this period were founded in Silicon Valley, a uniquely munificent cluster that is a major center of venture capital and investment banking for the industry. It is only through comparison with other industries that we find that the proximity pattern of law firms to semiconductor startups is statistically no different from the proximity of biotechnology and telecommunications startups to their legal counsel. Table 2 below shows the proximity counts from three high technology industries from our database.7

7

For a detailed discussion of the proximity patterns, and associated chi-square analysis, of these industries and their ESNs see Kenney and Patton (2005) pp. 218-223.

12

Table 2. Proximity of Entrepreneurial Support Network Actors to Firms going Public (1996 2000) Semiconductors Telecommunications Biotechnology Firm Lawyers Inside 50 miles 35 79.5% 40 75.5% 44 67.7% Outside 50 miles 9 20.5% 13 24.5% 21 32.3% VC Directors Inside 50 miles 45 54.9% 44 44.0% 28 25.9% Outside 50 miles 37 45.1% 56 56.0% 80 74.1% Investment Bankers Inside 50 miles 31 70.5% 26 49.1% 17 26.2% Outside 50 miles 13 29.5% 27 50.9% 48 73.8% Non-VC Directors Inside 50 miles 55 50.9% 50 39.7% 66 34.0% Outside 50 miles 53 49.1% 76 60.3% 128 66.0% Source: Kenney and Patton (2005) p. 219. Another result of comparative analysis is the realization that biotechnology startups are more reliant on services provided extraregionally. The well documented co-location of venture capital and biotechnology startups is only relative when compared to certain other industries. In considering the general propositions advanced above in General Propositions 1 through 3, we argued that clusters would support experienced ESN actors so that they would be the obvious choice to provide services for startups within the cluster. For startups outside of a cluster, though, a choice may be necessary between inexperienced but proximate service providers, and experienced but distant providers. Which choice is made will provide insight into the role of proximity in conveying knowledge. Proximity between a firm and a member of its ESN is readily determined by distance in miles. Measuring expertise is a challenge. One actor, investment bankers, can be ranked on the basis of reputation derived from underwriters' relative placement in stock offerings as "tombstone" announcements developed by Carter and Manaster (1990) and others.8 However, such measures for venture capitalists and lawyers are not available. We propose to generate an experience index for three actors; venture capital firms, law firms, and investment banks, on the basis of their relative frequency in participating in an IPO in the population of IPOs from 1996 through 2005. All actors can be ranked along this dimension and then bifurcated into a high experience and low experience group. At the most simply level a two by two contingency table can be produced for each of three actors across all industry groups as shown in Table 3 below. Table 3. Close (within 50 miles) Far (over 50 miles)

High Experience

Low Experience

A

B

C

D

8

See Carter et al. (1998) for a detailed discussion of this and other reputation-based measures of investment bankers.

13

In previous work in high technology industries we found that firm lawyers are the most proximate actor in our data on ESNs. Given the counseling role of a firm's lawyer we hypothesize that this holds for other industries as well. If so, we would expect that the count in cell B would be higher than in cell C in Table 3. Investment bankers, on the other hand, may very well be valued more for their expertise in which case we would expect that the count in cell C would be higher than in cell B. Since the count in cell A depends on the geography of the particular industry, insight into the importance of proximity of actors, relative to experience, will be provided by the distribution between cells B and C. This will allow us to not only compare actors over all industries, but allow us to discover if these spatial patterns vary across industries as well. Another series of questions addressed in Proposition 8 above is based on the time it took for a firm to go public from its founding. An examination of this issue is based on the application of a duration model where the particular estimated equation has the length of time for a firm to go public as the dependent variable. The STATA statistical package has a Cox proportional hazards model that will be used for this purpose. Explanatory variables, such as the proximity of ESN actors and venture capital funding, enter into the model by multiplying a hazard function by a scale factor producing what is known as a proportional hazard specification: h(t) = h0(t) eβ1x1+ β2x2+...+ βkxk Where h(t) is the probability that a firm will go public in time interval t, h0(t) is the baseline probability in time t, and x1, x2, ... xk are the independent variables influencing the time elapsed between a firm's founding and its initial public offering. One notable phenomenon of initial public offerings is their high variability in number over time. Jay Ritter (1998: 11) observed that the IPO volume shows a strong tendency to be high following periods of high stock market returns. This was the case in the time period examined here, with the bull market driving very high volumes of IPOs from 1996 through 2000, followed by a collapse in IPOs and the market in 2001 - 2003, and then a recovery in 2004 - 2005. In addition to variation in IPO volume, the length of time between founding and going public also varies.9 Besides controlling for the year of the IPO we control for venture capital involvement through two measures; major VC involvement indicated by a VC director on the board of the firm, and VC funding as determined by the VentureExpert database. In cases in which the focal firm, i.e., the startup, did not receive venture capital, it is likely there will be no venture capitalist on the board. For firms without venture capitalists, we will examine the prospectus to identify whether the firm was loan financed and whether a representative of the loan provider is on the board. The firm variables of interest will be whether the startup is located in a cluster of firms in the same industry or not, whether the firm is located in an entrepreneurial munificent setting such as Silicon Valley, and an index indicating the proximity of the members of the firm's ESN. The general thinking here is that firms will go public more quickly if they are founded in a cluster, and have greater proximity to members of their ESN. How this may vary across industries and over time remains to be seen. 9

The length of time for an average venture backed firm to go public has increased from around five or six years in the late 1990s, to around nine years currently. (Buckman 2006)

14

Database Preparation Tasks Remaining There are two major tasks remaining to complete the database: 1.) Internet Search of the Location of Support Network Actors The directors’ addresses are occasionally given in the prospectus. The remainder must be found through Internet searches. These searches are effective. For high technology firms less than 5% of the VC directors could not be located by name and office with which they are affiliated. Our data on the VC’s location is superior to the VentureExpert database, which attributes the investment to the VC firm’s headquarters. The addresses we attribute to the non-VC director is of their employer (legally this must be reported unless they are not employed). Collecting these addresses poses difficulties: First, some few directors report no employer (these are almost always retired executives). This event is coded as missing data. Second, the address found on the Internet in 2006 may no longer be the address of the employer (i.e., the employer’s office may have changed in the interim). We control for this possibility by using contemporary directories where possible but cannot discount that it will introduce a small measure of error into the database. Third, we attribute the director to the an address of that person’s corporate headquarters when another address is not clearly stated. This may not be quite as serious a data reliability issue as it at first seems, because invariably these individuals are elite corporate executives (usually Chieflevel or right below), and attributing them to their corporate headquarters (unless, of course, on the corporate website they are attributed to another location, e.g., the President of Microsoft China is located in Beijing) is unlikely to introduce significant bias, though it does introduce some error. The trade-off is to lose much information to eliminate what is likely to be a small amount of error. We feel the trade-off is justified. Experience from previous work is encouraging as a precise address was found for 95.6%, 86.9%, and 81.9% of the non-VC directors in semiconductors, telecommunications equipment, and biotechnology (biotechnology firms had a number of retired corporate executives on their boards of directors), respectively. 2.) Extend the Database from 2001 through 2005. We will collect and parse the data for the approximately 500 IPOs since 2000. Project Outcomes Three project outcomes are planned: 1. Publishing articles in scholarly journals in the spatial sciences (e.g., Economic Geography etc.), small business studies (e.g., Small Business Economics etc.), and research management (e.g., Research Policy) journals. 2. Through a password-protected system, the final database will be placed online, so that bona fide researchers can use it. 3. An online user’s guide to the database will be created.

15

Abolafia, M. 1997. Making Markets: Opportunism and Restraint on Wall Street (Cambridge, MA: Harvard University Press). Acs, Z. and C. Armington. 2006. Entrepreneurship, Geography, and American Economic Growth (Cambridge: Cambridge University Press). Agnes, P. 2000. "The 'End of Geography' in Financial Services: Local Embeddedness and Territorialization in the Interest Rate Swaps Industry." Economic Geography 76: 347366. Aldrich, H. 1999. Organizations Evolving (London: Sage Publications). Almeida, P. and B. Kogut. 1999. “Localization of Knowledge and the Mobility of Engineers in Regional Networks.” Management Science 45 (7): 905-917. Almeida, P. and B. Kogut. 1997. "The Exploration of Technological Diversity and the Geographical Localization of Innovation." Small Business Economics 9: 21-31. Audretsch, D. B. and M. P. Feldman. 1996. "R&D Spillovers and the Geography of Innovation and Production." American Economic Review 86 (3): 630-640. Audretsch, D. B. and P. E. Stephan. 1996. "Company-Scientist Locational Links: The Case of Biotechnology." American Economic Review 86 (3): 641-652. Bagchi-Sen, S. 2004. “Firm-specific Characteristics of R&D Collaborators and Noncollaborators in US Biotechnology Clusters and Elsewhere.” International Journal of Technology and Globalisation 1 (1): 92-117. Baptista, R. and P. Swann. 1998. “Do Firms in Clusters Innovate More?” Research Policy 27: 525-540. Bathelt, H., A. Malmberg, and P. Maskell. 2004. “Clusters and Knowledge: Local Buzz, Global Pipelines and the Process of Knowledge Creation.” Progress in Human Geography 28 (1): 31-56. Bennett, R. J. and C. Smith. 2002. “The Influence of Location and Distance on the Supply of Business Advice.” Environment and Planning A 34 (2): 251-270. Bennett, R. J., D. J. Graham, and W. Bratton. 1999. "The Location and Concentration of Businesses in Britain: Business Clusters, Business Services, Market Coverage and Local Economic Development". Transactions of the Institute of British Geographers 24: 393420. Black, B. S. and R. J. Gilson. 1998. "Venture Capital and the Structure of Capital Markets." Journal of Financial Economics 47: 243-277.

1

Boyd, R. L. 2003. “Ethnic Competition in the Retail Trade in Northern U.S. Cities in 1900: A Test of the Saturation Theory of Entrepreneurship.” Urban Geography 24 (1): 45-60. Brown, J. S. and P. Duguid. 2000. "Mysteries of the Region: Knowledge Dynamics in Silicon Valley." In The Silicon Valley Edge C. Lee et al. (eds.) (Stanford: Stanford University Press): 16-45. Buckman, R. 2006. "The Venture-Capital Yard Sale." Wall Street Journal, July 18. Carter, R. B., F. H. Dark, and A. K. Singh. 1998. "Underwriter Reputation, Initial Returns, and the Long-run Performance of IPO Stocks." The Journal of Finance 53 (1): 285-311. Carter, R. and S. Manaster. 1990. "Initial Public Offerings and Underwriter Reputation." Journal of Finance 45: 1045-1068. Christopherson, S. and M. Storper. 1986. 'The City as Studio; the World as Back Lot: The Impact of Vertical Disintegration on the Location of the Motion-picture Industry', Environment and Planning D: Society and Space 4 (3): 305-320. Clark, G. L. and K. O'Connor. 1997. "The Informational Content of Financial Products and the Spatial Structure of the Global Finance Industry." In Spaces of Globalization: Reasserting the Power of the Local K. R. Cox (ed.) (New York: Guilford Press). Cooke, P. 2004. “The Molecular Biology Revolution and the Rise of Bioscience Megacentres in North America and Europe.” Environment and Planning C: Government and Policy 22 (2) 161 – 177. Cooke, P., M. G. Uranga, and G. Etxebarria. 1997. “Regional Innovation Systems: Institutional and Organizational Dimensions.” Research Policy (26): 475-491. Cooke, P. and K. Morgan. 1990. Learning through Networking (Cardiff: University of Wales Press). Deeds, D., D. DeCarolis and J. Coombs. 1997. “The Impact of Firm-specific Capabilities on the Amount of Capital Raised in an Initial Public Offering: Evidence from the Biotechnology Industry.” Journal of Business Venturing 12 (1): 31-46. Dumais, G., G. Ellison and E. L. Glaeser. 2002. “Geographic Concentration as a Dynamic Process.” The Review of Economics and Statistics 84 (2): 193–204. Eisenhardt, K. M. and C. B. Schoonhoven. 1990. "Organizational Growth: Linking Founding Team, Strategy, Environment, and Growth among U.S. Semiconductor Ventures, 1978-1988." Administrative Science Quarterly 35: 504-529.

2

Feldman, M. P. 2000. "Location and Innovation: The New Economic Geography of Innovation, Spillovers, and Agglomeration." In The Oxford Handbook of Economic Geography G. L. Clark, M. P. Feldman, and M. S. Gertler (eds.) (New York: Oxford University Press): 373-394. Feldman, M. P. 1994. The Geography of Innovation. (Boston: Kluwer Academic Publishers). Florida, R. L. and M. Kenney. 1988a. "Venture Capital, High Technology and Regional Development." Regional Studies 22: 33-48. Florida, R. L. and M. Kenney. 1988b. “Venture Capital-Financed Innovation and Technological Change in the U.S.” Research Policy 17 (1): 119-37. Freeman, J. and M. T. Hannan. 1983. “Niche width and the dynamics of organizational populations.” American Journal of Sociology 88: 1116-1145. Fritsch, M. 2001. “Co-operation in Regional Innovation Systems.” Regional Studies 35 (4): 297-307. Garud, R. and S. Jain. 1996. "Technology Embeddedness." In J. Baum and J. Dutton (eds.) Advances in Strategic Management (13): 389-408 (Connecticut: JAI Press). Gertler, M. S. 2003. "Tacit Knowledge and the Economic Geography of Context, or The Undefinable Tacitness of Being (There)." Journal of Economic Geography 3: 75-99. Gompers, P. A. and J. Lerner. 1999. The Venture Capital Cycle (Cambridge Mass. : MIT Press). Granovetter, M. 1985. "Economic Action and Social Structure: The Problem of Embeddedness." American Journal of Sociology 91: 481-510. Greenwald, B., and J. E. Stiglitz. 1992. "Information, Finance and Markets: The Architecture of Allocative Mechanisms." Industrial and Corporate Change 1: 37-63. Gillespie, A. E. and A. E. Green. 1987. “The Changing Geography of Producer Services Employment in Britain.” Regional Studies 21: 397–411. Henderson, V. 1997. “Externalities and Industrial Development.” Journal of Urban Economics 42: 449-470. Hendry, C., J. Brown, and R Defillippi. 2000. “Regional Clustering of High Technologybased Firms: Opto-electronics in three countries.” Regional Studies 34 (2): 129-144. Jacobs, J. 1969. The Economy of Cities (New York: Random House).

3

Jaffe, A. B., M. Trajtenberg, and R. Henderson. 1993. "Geographic Localization of Knowledge Spillovers as Evidenced by Patent Citations." Quarterly Journal of Economics 108: 577-598. Keeble, D. and L. Nachum. 2002. “Why Do Business Service Firms Cluster? Small Consultancies, Clustering and Decentralization in London and Southern England.” Transactions of the Institute of British Geographers 27 (1): 67-90. Kenney, M. 2002. "The Growth and Development of the Internet in the United States." In The Global Internet Economy B. Kogut (ed.) (Cambridge: MIT Press): 69-108. Kenney, M. (ed.). 2000. Understanding Silicon Valley (Stanford: Stanford University Press). Kenney, M. 1986. Biotechnology: The University-Industry Complex. (New Haven: Yale University Press). Kenney, M. and U. v. Burg. 1999. "Technology and Path Dependence: The Divergence between Silicon Valley and Route 128." Industrial and Corporate Change 8 (1): 67-103. Kenney, M. and D. Patton. 2005. "Entrepreneurial Geographies: Support Networks in Three High-Technology Industries." Economic Geography 81 (2):201-228. Krugman, P. 1991. Geography and Trade. (Cambridge: The MIT Press). Lamoreaux, N. R., M. Levenstein, K. L. Sokoloff. 2006. “Financing Invention during the Second Industrial Revolution: Cleveland, Ohio, 1870-1920.” Unpublished working paper. Malecki, E. 2000. “Knowledge and Regional Competitiveness.” Erdkunde 54 (4): 334351. Malecki, E. 1990. “New Firm Formation in the USA: Corporate Structure, Venture Capital, and Local Environment.” Entrepreneurship and Regional Development 2: 247265. Malecki, E. 1980. "Dimensions of R&D Location in the United States." Research Policy 9: 2-22. Markusen, A. 1999. "Sticky Places in Slippery Spaces: A Typology of Industrial Districts." In The New Industrial Geography: Regions, Regulation and Institutions T. Barnes and M. Gertler (eds.) (London: Routledge): 98-126. Marshall, A. 1890. Principles of Economics. (London: Macmillan & Company). Maskell, P. 2001. “Towards a Knowledge-based Theory of the Geographical Cluster.” Industrial and Corporate Change 10 (4): 921-943

4

Maskell, P. and A. Malmberg. 1999. "Localized Learning and Industrial Competitiveness." Cambridge Journal of Economics 23: 167-185. Mason, C. M. and R. T. Harrison. 2002. “The Geography of Venture Capital Investments in the UK.” Transactions of the Institute of British Geographers 27 (4): 427-451. McKendrick, D. G. 2004. “Leveraging Locations: Hard Disk Drive Producers in International Competition.” In Locating Global Advantage M. Kenney (ed.) (Stanford: Stanford University Press): 142-174. McKendrick, D. G., R. Doner, and S. Haggard. 2000. From Silicon Valley to Singapore: Location and Competitive Advantage in the Hard Disk Drive Industry (Stanford: Stanford University Press). Muller E. and A. Zenker. 2001. “Business Services as Actors of Knowledge Transformation: The Role of KIBS in Regional and National Innovation Systems.” Research Policy 30: 1501-16. Nijkamp, P. 2003. “Entrepreneurship in a Modern Network Economy.” Regional Studies 37, (4): 395-405. Nonaka. I. and H. Takeuchi. 1995. The Knowledge-Creating Company (Oxford: Oxford University Press). Oinas, P and E. J. Malecki. 2002. “The Evolution of Technologies in Time and Space: From National and Regional to Spatial Innovation Systems.” International Regional Science Review 25 (1): 102-131. Owen-Smith, J. and W. W. Powell. 2002. "Knowledge Networks in the Boston Biotechnology Community." Presented at the Conference on Science as an Institution and the Institutions of Science in Sienna, 25-26 January. Patton, D. and M. Kenney. 2005. "The Spatial Configuration of the Entrepreneurial Support Network for the Semiconductor Industry." R&D Management 35 (1): 1-16. Peck, J. 2005. "Economic Sociologies in Space." Economic Geography 81 (2): 129-176. Perroux , F. 1950. "Economic Space: Theory and Applications." The Quarterly Journal of Economics 64 (1): 89-104. Perroux, F. 1988. "The Pole of Development’s New Place in a General Theory of Economic Activity". In Regional Economic Development B. Higgins and D. J. Savoie (ed.) (Boston: Unwin Hyman): 48-76.

5

Pinch, S., N. Henry, M. Jenkins, and S. Tallman. 2003. "From 'Industrial Districts' to 'Knowledge Clusters': A Model of Knowledge Dissemination and Competitive Advantage in Industrial Agglomerations." Journal of Economic Geography 3: 373-388. Pinch, S., and N. Henry. 1999. "Paul Krugman’s Geographical Economics, Industrial Clustering and British Motor Sport Industry". Regional Studies 33: 815-27. Piore, M. and C. Sabel. 1984. The Second Industrial Divide (New York: Basic Books). Polanyi, K. et al. (eds.). 1971. Trade and Markets in Early Empires. (Chicago, IL: Henry Regnery). Porter, M. 1998. “Clusters and the New Economics of Competition.” Harvard Business Review (November-December): 77-90. Porter, M. 1990. The Competitive Advantage of Nations (New York: Free Press). Powell, W. W., K. W. Koput, J. I. Bowie, and L. Smith-Doerr. 2002. "The Spatial Clustering of Science and Capital: Accounting for Biotech Firm-Venture Capital Relationships." Regional Studies 36 (3): 291-305. Powell, W. W., K. W. Koput., L. Smith-Doerr. 1996. "Interorganizational Collaboration and the Locus of Innovation: Networks of Learning in Biotechnology." Administrative Science Quarterly 41: 116-145 Pryke, M. and R. Lee. 1995. "Place Your Bets: Towards an Understanding of Globalization, Socio-Financial Engineering and Competition within a Financial Centre." Urban Studies 32: 329-344. Razin, E. 1989. “Relating Theories of Entrepreneurship among Ethnic Groups and Entrepreneurship in Space -The Case of the Jewish Population in Israel.” Geografiska Annaler, Series B, Human Geography 71B: 167-181. Ritter, J. R. 1998. "Initial Public Offerings." Contemporary Finance Digest 2: 5-30. Rocha, H. O. 2004. “Entrepreneurship and Development: The Role of Clusters.” Small Business Economics 23: 363–400. Romanelli, E. and C. B. Schoonhoven. 2001. "The Local Origins of New Firms" In The Entrepreneurship Dynamic: Origins of Entrepreneurship and the Evolution of Industries C. B. Schoonhoven and E. Romanelli (eds.) (Stanford: Stanford University Press): 40-67. Saxenian, A. 1994. Regional Advantage. (Cambridge, MA: Harvard University Press). Schumpeter, J. A. 1947. “The Creative Response in Economic History.” Journal of Economic History (November): 149-159.

6

Scott, A. J. 2006. “Entrepreneurship, Innovation, and Industrial Development: Geography and the Creative field Revisited.” Small Business Economics 26, (1): 1-24. Scott, A. J. 1993. Technopolis: High-Technology Industry and Regional Development in Southern California (Berkeley: University of California Press). Scott, A. J. 2002. "A New Map of Hollywood: The Production and Distribution of American Motion Pictures." Regional Studies 36: 957-76. Scott, A. J. and M. Storper. 2003. "Regions, Globalization, Development." Regional Studies 37 (6): 579-593. Sorenson, O. and P. G. Audia. 2000. "The Social Structure of Entrepreneurial Activity: Geographic Concentration of Footwear Production in the United States, 1940-1989." American Journal of Sociology 106: 424-461. Sorenson, O. and T. E. Stuart. 2001. “Syndication Networks and the Spatial Distribution of Venture Capital Investments.” American Sociological Review 106 (6): 1546-1588. Storper, M. 1995. "The Resurgence of Regional Economies, Ten Years Later: The Region as a Nexus of Untraded Interdependencies." European Urban and Regional Studies 2 (3): 191-221. Storper, M. and R. Salais. 1997. Worlds of Production: Frameworks of Action in the Economy (Cambridge: Harvard University Press). Storper, M. and A. J. Venables. 2004. "Buzz: Face-to-Face Contact and the Urban Economy." Journal of Economic Geography 4: 351-370. Storper, M, and R. Walker. 1989. The Capitalist Imperative: Territory, Technology, and Industrial Growth (London: Basil Blackwell). Suarez-Villa, L. 1989. The Evolution of Regional Economies: Entrepreneurship and Regional Change (London: Praeger). Suchman, M. C. 2000. "Dealmakers and Counselors: Law Firms as Intermediaries in the Development of Silicon Valley". In Understanding Silicon Valley M. Kenney (ed.) (Stanford: Stanford University Press): 71-97. Sydow, J. and U. Staber. 2002. “The Institutional Embeddedness of Project Networks: The Case of Content Production in German Television.” Regional Studies 36 (3): 215227. Thornton, P. H. 1999. “The Sociology of Entrepreneurship.” Annual Review of Sociology 25: 19-46.

7

Thrift, N. and A. Leyshon. 1994. "A Phantom State? The Detraditionalization of Money, the International Financial System and International Financial Centres." Political Geography 13: 299-327. Tickell, A. 1999. “The Geographies of Services: New Wine in Old Bottles.” Progress in Human Geography 23 (4): 633-639. Uzzi, B. 1999. "Embeddedness in the Making of Financial Capital: How Social Relations and Networks Benefit Firms Seeking Financing." American Sociological Review 64 (4): 481-505. Walker, R. 1985. "Technological Determination and Determinism: Industrial Growth and Location." In High Technology, Space and Society M. Castells (ed.) (Beverly Hills, CA: Sage Publications): 226-264. Wenger, E. and J. Lave. 1993. Situated Learning: Legitimate Peripheral Participation (New York: Cambridge University Press). Zook, M. A. 2002. “Grounded Capital: Venture Financing and the Geography of the Internet Industry, 1994–2000.” Journal of Economic Geography 2: 151-177. Zucker, L. G., M. R. Darby, and M. B. Brewer. 1998. "Intellectual Human Capital and the Birth of U.S. Biotechnology Enterprises." American Economic Review 88 (1): 290306.

8

Proposal Title: The Geography of Entrepreneurial Support Networks (FUNDED) Context Statement NATIONAL SCIENCE FOUNDATION GEOGRAPHY AND REGIONAL SCIENCE PROGRAM CONTEXT STATEMENT -- FALL 2006 ADVISORY PANEL MEETING The Geography and Regional Science (GRS) Advisory Panel met on November 2-3, 2006, to evaluate proposals that were submitted in response to the August 15, 2006, target date. The GRS Advisory Panel considered 152 individual proposals in this round. Not all of these proposals required separate action by the panel, however, as 18 were secondary components of collaborative research proposals, with principal investigators from two or more institutions proposing collaborative work together through separately submitted proposals. These secondary components did not require independent consideration. Of the remaining 134 proposals, 114 were "regular" research proposals, 18 were proposals seeking Faculty Early-Career Development (CAREER) awards, and 2 were submissions for conferences, workshops, institutes, group travel, and/or some other community-building activity. A total of 51 of the regular proposals were jointly reviewed with one or more other NSF programs. All other proposals were evaluated solely by GRS. The GRS Advisory Panel evaluated proposals using the NSF criteria of intellectual merit and broader impacts as articulated in the Grant Proposal Guide (NSF 04 23). All proposals were sent to at least six external reviewers and to at least two advisory panel members. Every proposal was evaluated using at least three written reviews; most had five to seven written reviews. All written reviews (from both external reviewers and panelists) were made available to panelists prior to the panel meeting. On the basis of written reviews and extensive discussion during the panel meeting, the GRS Advisory Panel made recommendations regarding GRS and NSF support for each proposal. For the 114 regular proposals, the panel judged 65 proposals to be "fundable," and it declined to recommend funding for 50 proposals. The panel deferred making a recommendation on one proposal. Of the "fundable" regular proposals, the panel recommended that 3 be given high priority for support, 8 be given high-medium priority, 20 be given medium priority, 18 be given medium-low priority, and 16 be given low priority. For the CAREER proposals, the panel judged 6 proposals to be "fundable," and it declined to recommend funding for 12 proposals. Of the "fundable" CAREER proposals, the panel recommended that 2 be given high-medium priority, 3 be given medium priority, and 1 be given low priority. For the workshop proposals, the panel judged 1 proposal to be "fundable" with low priority for support and it declined to recommend funding for the other proposal. Depending on budgetary constraints and on the willingness of other programs to participate in joint funding of meritorious proposals, GRS program directors anticipate recommending funding for between 12 and 16 regular projects, 1 to 2 CAREER proposals, and possibly

1 workshop proposal. Verbatim copies of all completed reviews and any relevant panel summaries are made available to principal investigators via the FastLane system. GRS panelists and program officers do not necessarily agree with or endorse all statements by reviewers. Some reviews may contain irrelevant comments, which are not used in making final funding decisions. Panel members are asked to reflect on the substance of the written reviews and to draw generalizations that extend beyond the summary ratings alone. The reviews and panel summary contain evaluative material and constructive suggestions that may be used by the principal investigator in the conduct of future research regardless of whether a proposal is awarded or declined.

Panel Summary

The Geography and Regional Science (GRS) Advisory Panel met on November 2-3, 2006, to evaluate proposals that were submitted in response to the August 15, 2006, target date. The GRS Advisory Panel considered 152 individual proposals in this round. Not all of these proposals required separate action by the panel, however, as 18 were secondary components of collaborative research proposals, with principal investigators from two or more institutions proposing collaborative work together through separately submitted proposals. These secondary components did not require independent consideration. Of the remaining 134 proposals, 114 were "regular" research proposals, 18 were proposals seeking Faculty Early-Career Development (CAREER) awards, and 2 were submissions for conferences, workshops, institutes, group travel, and/or some other community-building activity. A total of 51 of the regular proposals were jointly reviewed with one or more other NSF programs. All other proposals were evaluated solely by GRS. INTELLECTUAL MERIT: The proposed research seeks to study the economic geography of entrepreneurial support networks (ESN), with the goal of verifying the role of network proximity to entrepreneurship. POSITIVE ASPECTS OF THE PROPOSAL AND PROPOSED RESEARCH: Positive merits include addressing reviewers' comments over the previous proposal. The propositions are reasonably formulated. Some of the rationales driving the propositions can be debated but it is overall difficult to quarrel with the PI's clarity in reasoning. SHORTCOMINGS AND WEAKNESSES OF THE PROPOSAL AND PROPOSED RESEARCH:

The independent variables of the hazards model are not clearly specified. The study may be biased towards IPO firms. ========================== BROADER IMPACTS: POSITIVE ASPECTS OF THE PROPOSAL AND PROPOSED RESEARCH: The PI is experienced and well-qualified to undertake the proposed work. He will likely be productive with the data as he has a good publication record. SHORTCOMINGS AND WEAKNESSES OF THE PROPOSAL AND PROPOSED RESEARCH: It is not entirely clear how innovative this work will be given the PI's long standing work on entrepreneurship. Face-to-face interactions are likely to be important in ESN networks as the PI observes but were not pursued further in the study. ========================== SYNTHESIS COMMENTS: Overall the proposal will build a good database, the sum requested is reasonable and the proposal also ties in the theory and empirical. However, the question is raised by the panel as to how this research will add to new insights in Economic Geography when a substantial literature exists on the importance of cluster economies. ========================== PANEL RECOMMENDATION (CHECK ONE): [ ] Fund with High Priority [ ] Fund with High-Medium Priority [X] Fund with Medium Priority [ ] Fund with Medium-Low Priority [ ] Fund with Low Priority [ ] Decline to Fund [ ] Defer Funding Recommendation In making this recommendation, panel members recognized that the Geography and Regional Science Program might well be unable to provide any funding for the project because of budgetary limitations.

Panel Recommendation: Fund with Medium Priority

REVIEW#1: What is the intellectual merit of the proposed activity? The theme on ESN deserves attention largely because the role of urbanization economies in clustering advantages is relatively unexplored compared to that of localization economies perhaps because of the lure of single industry studies. What are the broader impacts of the proposed activity? The proposal analysis is more detailed and tighter than other proposals I have read. Despite the literature review on the interplay of both local and extra-local dimensions, the propositions are much more focused on the local factors vis-a-vis the role of proximity in cluster advantages. P7 and P9 appear to address some extra-local effects but presumably this comes in the form of dispersion of ESN board members beyond a 50 mile distance û a threshold that will mean differently in the context of compact areas like NYC and Boston versus the Midwest. Table 1 is also interesting. It does not say much about localism if nearly half of the firms (VCs and non VCs) are outside 50 miles. Summary Statement Overall this is a worthwhile endeavor that attempts to understand the role of various ESN actors in supporting entrepreneurship. In particular, the proposal nicely ties the empirical with the theoretical.

REVIEW#2: What is the intellectual merit of the proposed activity? The proposal developes a database to disentangle the role of venure capital,lawyers and other members of the board. This builds on earlier work by the author and the present proposal will fill out the existing data base. This is important to help understant the entrepreneurial support network that exists or does not exist at the regional level.The authors should look at the AER in March 2005 for an exchange on this subject by Jaffe et al as well as the recent book by Saxenian. What are the broader impacts of the proposed activity? The availability of a data base for others to work from and expand. The public policy

issue of how regional entrepreneurship policy works is an ongoing area of research in the area of technology commercialization. Summary Statement I support this proposal.

REVIEW#3: What is the intellectual merit of the proposed activity? Merits: 1. Clustering/ESNs across all industries. Since only a finite number of "munificent" regions can be expected to attract the desirable high tech sectors (biotech, nanotech, etc.), a deeper understanding of the clustering/ESN patterns of other industries (i.e. services, etc.) will contribute greatly to economic development policy and practices. 2. ESN actors. This research examines a broader range of ESN actors beyond the usual suspect: venture capitalists. It also can tease out the ESN actors' roles from those of other cluster supporters including suppliers, collaborators, competitors, customers, etc. 3. Extends beyond disciplinary boundaries. The research contribution will extend beyond the geography/regional science disciplines to those of business management and entrepreneurialism among others. 4. The available and requested resources are adequate to conduct the work. Drawbacks: 1. The research proposes to "explore when and under what conditions the ESN actors are proximate or distant" from the entrepreneurial firm. While proximity factors can be extrapolated from the data, identifying the "conditions" under which this happens is not a direct outcome. 2. Study will explain how "some entrepreneurs overcome their locational liability." These are inferred conclusions, not explanations. 3. Further discussion is needed as to how a cluster will be defined across industries. How will the research determine which regions are "clusters" for a given industry? 4.The data relies exclusively on an elite group of companies, those offering IPOs.

What are the broader impacts of the proposed activity? 1. The accessible web interface/database will be invaluable to researchers seeking a more systematic approach to understanding clustering and entrepreneurship. A discussion group connected to it will enhance the possibility of research partnerships. 2. The empirical research will provide an important "corrective" to a policy debate that often valorizes clusters, particularly high-tech ones. It will offer a more nuanced approach for regions without a strong ESN presence by distinguishing between industries that need local presence to thrive and those that don't

3.Undergraduate students will play an important role on the research team. Summary Statement While reviewing this proposal, I had two competing and conflicting reactions. On most counts, this well-developed research agenda is both enticing and irresistible. The highlyqualified P.I. holds out the possibility of breaking the boundaries of a decade plus of research on industrial clusters. He proposes to do so by providing empirical evidence to confirm or deny a plethora of qualitative (and to some extent, quantitative) findings. However, the proposal contains a number of drawbacks. Most obviously, the data set is limited to an "elite" set of start-ups (i.e., those going public). This subgroup may possess significant success-related factors that lay outside of the ESN. To the P.I.'s credit, this issue is addressed to some extent in the proposal.

REVIEW#4: What is the intellectual merit of the proposed activity? The proposal is well organized and well written. Although the PIs relies on previous research conducted by a small circle of individuals, he asks important and recurrent questions in research on entrepreneurship. The questions posed, however, are rather conventional and this research will most likely prove what we already suspect. Numerous studies have been conducted on start-ups, without addressing the core of the issue, i.e., what drives entrepreneurship, and I worry that this is going to be yet another research that explores the periphery, but not the core, of entrepreneurship. I disagree with PI's abstract which states that "the role of social actors supporting entrepreneurship has recently been recongized as critical, yet their geography is not well understood." This research has in fact been pioneered by the PI as far back as 1988. Also, to answer the question of the geographical clustering of entrepreneurship by measuring the proximity between startups and auxiliary services would generate a partial story. What are the broader impacts of the proposed activity? The methodology is simple, and does not seem to include indicators that involve actual face-to-face interactions between lawyers and start-ups.

APPENDIX

iv. Lam, Nina; Campanella, Richard; and Pace, Kelley. Small Grants for Exploratory Research (SGER): DecisionMaking Among Businesses in PostCatastrophe Uncertainty: How Economic Geographies Re-Form in New Orleans..

PROJECT SUMMARY Decision-Making Among Businesses in Post-Catastrophe Uncertainty How Economic Geographies Re-Form in New Orleans Nina Lam, Geography, Louisiana State University Richard Campanella, Center for Bioenvironmental Research, Tulane University Kelley Pace, Finance, Louisiana State University

This project seeks to understand the commercial side of resettlement of residents in urban environments after a catastrophe. The overarching research question is: how businesses make spatial decisions on whether to return or relocate and how these decisions in turn impact the landscape and its economy. Specifically, for this SGER project, the objective is to collect and analyze time-critical data on what, where, how, why, and when businesses return to (and survive or fail in) New Orleans following the repopulation of the city after Hurricane Katrina. Both telephone survey and street survey of businesses will be conducted periodically. The survey data will then be mapped and integrated with census, flood damage, and other GIS (geographic information system) data layers. This will provide a valuable, time-critical spatial-temporal data set that can serve to answer a number of specific research questions. Intellectual Merits: Very little research has focused on collecting time-critical, empirical data on how businesses make spatial decisions on whether they remain or relocate after a catastrophe, especially a catastrophe as deep and wide as we have seen that affects an entire metropolis of New Orleans. The time-critical data that we collect for this project will provide unique information on how decisions among businesses are made in this unprecedented case. The coupling and tracking of street and telephone surveys over time will provide vital information for research on human-social-economic dynamics over space and time. The data we collect will also serve as an important benchmark dataset for subsequent research and for comparisons with other studies (e.g. studies on decisions made by individuals). Broader Impacts: The time-critical, integrated GIS data set collected in this project can be made available to other researchers and planners, and can be used as a basis for further related research, such as modeling the impacts of Katrina on health, poverty, and crime. Our preliminary analyses of the data will be published and widely disseminated. Our data will provide a first-hand, rarely captured view of how a city recovers, literally from ground zero, and how businesses make decisions in post-catastrophe uncertainty. This information will help governmental and planning agencies in devising effective policies for economic recovery in the region.

PROJECT DESCRIPTION Decision-Making Among Businesses in Post-Catastrophe Uncertainty How Economic Geographies Re-Form in New Orleans 1. Goal and Objective The proposed project seeks to understand the commercial side of resettlement of residents in urban environments after a catastrophe. Residents cannot resettle until commercial enterprises return with basic products and services, but commercial enterprises cannot provide basic products and services until residents return, as employees and as customers. Other researchers are focusing on the resident side of this conundrum; our goal is to understand the commercial side. The overarching research question is: how businesses make spatial decisions on whether to return after a major catastrophe and how these decisions in turn impact the landscape and its economy. The objective of this SGER project is to collect and analyze time-critical data on what, where, how, why, and when businesses return to (and survive or fail in) New Orleans following the repopulation of the city after Hurricane Katrina. Both telephone survey and street survey of businesses will be conducted periodically. The survey data will then be mapped and integrated with census, flood damage, and other GIS (geographic information system) data layers. This will provide a valuable, integrated spatial-temporal data set that can serve to answer a number of specific research questions, as outlined below. We will continue collecting the data within the funding period (one year) and plan to continue until New Orleans’ post-catastrophe recovery stabilizes. 2. A Brief Background and Relevant Literature There is a rich literature on decision making under uncertainty, and different theoretical decision models have already been suggested (Simon 1997; Edmonds 1999; Lam et al. 2003). However, decision making related to catastrophic events is not very well understood (Stewart and Bostrom 2002). These events are rare, highly uncertain, and often without timely data, hence making systematic studies using empirical data and statistical inference difficult. Historically, hurricanes have been a major threat to the coastal region in the U.S., causing many deaths and widespread damage (e.g., the Galveston hurricane in 1900 led to 8,000 deaths; Hurricane Andrew in 1992 caused $30 billion economic losses). However, none can compare with the recent disaster of Katrina, where a major city New Orleans was inundated and paralyzed. The disaster reveals many missteps or bad decisions made by various organizations involved. Among the various problems that need to be solved immediately for economic recovery is the need for better data to support better decisions. For example, during the height of the rescue phase, LSU GIS specialists and graduate students assisted 24-7 in the Baton Rouge command center, producing extremely useful maps for military personnel for the rescue operation. However, our first-hand experience shows that despite the voluminous satellite imagery and maps that are seemingly available, many of them are not useful (e.g., out of date or imagery with clouds). To this date, FEMA has no GIS ability to perform simple queries on map resources based on locations and to screen out noisy data. For better preparation, rescue, and recovery in the future, useful data will need to be collected, integrated, and used in a real-time basis. The collection of time-critical data on business decisions by this project will help to fill this void. In general, we know more about individuals making risky decisions (e.g., Slovic et al. 1974; Slovic et al. 1976; McAlister 1987; Averill 1987; Lopes 1987; Kunreuther 1992), but we 1

know very little about high-consequence decisions by groups and organizations. Based on the literature (e.g., Palm 1995; Palm and Hodgson 1992&1993; Cutter 2003), the decision factors for business to locate or re-locate could include: future hurricane risks, economic impacts, insurance rates, emergency support, and market potential. The empirical data collected from this project will verify which combination of factors are most important for spatial decisions. 3. Methodology and Research Questions The proposed research comprises two integrated approaches, one involving comprehensive field surveys of three selected commercial arteries in New Orleans, and one involving stratified telephone surveys of New Orleans businesses conducted from Baton Rouge. Finally, these data will be mapped and incorporated with existing GIS data layers to enable comprehensive, integrated analyses. Street Survey. The purpose of the street survey data is to answer spatial-temporal research questions with a comprehensive dataset. Three major commercial corridors will be selected: Magazine Street, a high-end retail and restaurant strip through uptown; St. Claude Avenue, an economically deprived commercial artery downtown; and Carrollton Avenue, which is socioeconomically positioned in between the other two corridors. The three study areas transect a wide range of social, economic, and physical conditions in the city, and their selection for this project is based on Co-PI Campanella’s research into the human and physical geography of New Orleans (Campanella and Campanella 1999; Campanella 2002; Campanella In press). The surveys will be conducted every two weeks, tracking what, where, and when businesses return, and of those that do, which survive and which fold during the research period. Every business will be recorded at the parcel level; there will be no sampling. A strategy will be developed to handle multi-tenant structures such as professional offices and upper-floor businesses (of which there are few). These data will be recorded along with their addresses and other basic information (locally owned versus national chain; approximate size of enterprise) in a database, and mapped out through a GIS. The GIS will contain other relevant data layers, such as elevation, infrastructure, areas which flooded during and after Hurricane Katrina, areas which suffered the worst damage, and socio-economic data at the block, block group, and census tract level from the 2000 census. Research Questions for Street Survey Data. The following descriptive statistics and maps will be developed from the street survey datasets: 1. Which businesses came back first, who owned them (local versus chain), what did they sell, and where were they located? 2. How did these patterns change over time? 3. Which ones closed, what did they sell, when, and where? 4. How did these patterns spatially relate to the businesses’ pre-Katrina socio-economic surroundings and their level of damage during Katrina? The street survey data will shed light on these questions: 1. Will those businesses located closest to wealthier areas that were not flooded return the fastest and survive in the highest numbers? 2. Will those businesses located near poorer areas that were flooded return slowest and go out of business in larger numbers?

2

3. Will those businesses located in other areas (wealthier but flooded, poorer but not flooded) return and thrive in numbers in the mid-range between the two aforementioned situations? 4. Will locally owned micro-businesses return first, and national chains later? 5. Will locally owned micro-businesses fold in higher numbers than national chains? 6. What is the relationship between the return and success rate of businesses and the pace and location of neighborhood repopulation/reconstruction? 7. Was total citywide evacuation a wise idea, from the perspective of rebuilding a customer base? 8. What can other cities do to foster the speedy return of the commercial sector in the wake of a catastrophe? 9. How does what we see along Magazine, St. Claude, and Carrollton affect the return-to-city and remain-in-city decisions of the residents of those neighborhoods? Telephone Survey. The biweekly street surveys will be complemented by telephone surveys conducted in 5 periods (every 2 to 3 weeks) by the LSU Public Policy Research Lab. Unlike the street surveys, the telephone surveys will target businesses throughout the entire city (New Orleans proper), based on a random sample stratified by geography (census tract) and by a few business categories (e.g., fast food, chain and non-chain grocery, essential services such as provided at physician offices, non-essential services such as provided by hair dressers). Each of the 5 rounds will seek to obtain 500 completed surveys. The survey will be identified as coming from a joint venture of LSU and Tulane, and this should increase the response rate. Consistently unanswered phone calls will be interpreted as a not-yet-reopened business; disconnected lines during any of the rounds will be interpreted as a never-reopened business, and an answered phone will indicate an open business or one that can potentially reopen. Since many businesses may not wish to spend much time with a survey, the interviewer will ask fewer and less probing questions than might be posed to individuals. The following questions will be asked to those that answer: 1. Are you open for business? If not, when do you plan to open? 2. What are your barriers to starting up or to succeeding? Rank on a scale of 1 to 5: Damage to Premises? Insurance? Lack of employees? Lack of customer base? Suppliers? Utilities? Communications? Crime? Environmental problems? Governmental problems? Other? 3. How optimistic are you about the future of your business? (Very optimistic, optimistic, unsure, pessimistic, very pessimistic). These data will be mapped at the block level and overlaid with the aforementioned GIS datasets. Telephone Survey Data Results and Research Questions. The following descriptive statistics and maps will be developed from the telephone survey datasets: 1. A map for each of the five survey rounds showing the status of the business (open, closed permanently, potential return). The map layer will show type of business, as well as chain or non-chain status. 2. A map for each of the five survey rounds showing the level of optimism. 3. A map for each of the rounds showing the response to barriers to success. The telephone survey data may shed light on these questions: 1. What is the relation between chain versus non-chain businesses and return? An individual business often has the flexibility to adapt, but may lack the non-local resources of chains.

3

2. What is the relation between socio-economic and Katrina-damage overlays and business returns? 3. What is the relation between the return and success rate of businesses versus the pace and location of neighborhood repopulation/reconstruction? Do businesses lag due to damage or lead due to opportunities? 4. Do levels of optimism show any systematic patterns with business type and location with respect to socio-economic and Katrina-damage overlays? 5. How do the responses to the barriers to success change over time? 6. What can other cities do to foster the speedy return of the commercial sector in the wake of a catastrophe? 7. How do these findings relate to return-to-city and remain-in-city decisions of the residents of those neighborhoods? 8. Do perceived barriers to success shift temporally from directly Katrina-related barriers (damage, insurance) to generalized barriers (lack of customer base, crime). Comparison of Street Survey and Telephone Survey. Results from the street survey and the telephone survey will be compared, to determine consistencies and understand contradictions. Deliverables. Deliverables will be an integrated GIS data set that consists of original survey data, other GIS data layers, maps, and descriptive analysis. This time-critical data set will benefit researchers and city planners in the future who want to understand how economic geographies re-form after catastrophes. A manuscript documenting the data and initial analysis will be prepared and submitted to a peer-reviewed journal in relevant fields. 4. Significance and Broader Impacts In terms of intellectual merits, very little research has focused on collecting time-sensitive, empirical data on how businesses make spatial decisions on whether they remain or relocate after a catastrophe, especially a catastrophe as deep and wide as we have seen that affects an entire metropolis of New Orleans. The time-critical data that we collect for this project will provide unique information on how decisions among businesses are made in this unprecedented case. The coupling and tracking of street and telephone survey over time will provide vital information for research on human-social-economic dynamics over space and time. The data we collect will also serve as an important benchmark dataset for subsequent research and for comparisons with other studies (e.g. studies on decisions made by individuals). This project has broad societal implications. The time-critical, integrated data set collected in this project can be made available to other researchers and planners, and can be used as a basis for further related research, such as modeling the impacts of Katrina on health, poverty, and crime. Our data will provide a first-hand, rarely captured view of how a city recovers, literally from ground zero, and how businesses make decisions in post-catastrophe uncertainty. Our preliminary analyses of the data will be published and widely disseminated. This information will help governmental and planning agencies in devising effective policies for economic recovery in the region.

4

5. Why This Research fits the SGER criteria This research will collect critical data that will soon be lost if not captured in time. The project is “exploratory and high-risk” because it has “a severe urgency with regard to availability of, or access to data, facilities or specialized equipment, including quick-response research on natural or anthropogenic disasters….” (NSF 04-23, Special Guidelines for Small Grants for Exploratory Research Proposals, page 32, July 2004). We are in the aftermath of an unprecedented catastrophe; an entire metropolis is relocated. Its return, and the return of businesses, will probably occur in an undocumented manner without an effort such as the one proposed here. 6. Project Management and Facilities GIS capabilities will be provided by both the Center for Bioenvironmental Research’s Environmental Informatics Lab (EIL) at the uptown campus of Tulane University and LSU. The Tulane facility was undamaged by Katrina and should be in operation soon. If not, the Computer Mapping Sciences Lab at the Department of Geography and Anthropology at Louisiana State University in Baton Rouge will be utilized. Both labs have the full suite of standard GIS and remote sensing hardware, software, and data, including ESRI ArcMap 9.1, Leica ERDAS Imagine 8.7, and others. The telephone surveys will be conducted by the LSU Public Policy Research Lab which has the capabilities of conducting large surveys. For example, the LSU Public Policy Research Lab conducts the consumer confidence surveys for Louisiana which become part of the national consumer confidence statistics. All PI’s participate in all aspects of the research, including the design of the survey questions, analysis of the results, and publications and presentations. Lam will oversee the entire project and focus on integrating the different components of the project (phone survey, street survey, GIS database integration). Campanella is primarily responsible for the street survey and GIS database development. Pace will provide a liaison between the LSU Public Policy Research Lab and assist in analyzing the data. 7. Relation to Previous Work of the Research Team PI Lam has conducted geographical analysis of health and environment at national, international, and local scales (e.g., HIV/AIDS, cancers, low-birth rates, hazardous waste sites), and has expertise in collecting and analyzing large-scale empirical data and developing GIS for spatial decision support. She has participated in a project on developing a GIS for the hurricane activity in the Caribbean region (see pending support). Co-PI Campanella has researched the human and physical geography of New Orleans for ten years, and has authored three books on the subject. Co-PI Pace has conducted extensive spatio-temporal econometric analyses on retail sales and real estate finance and economics. All PIs reside within or near the Katrina affected area and are very familiar with the region. One co-PI is currently displaced from New Orleans (temporarily residing in Baton Rouge), and another lost a house in Slidell, Louisiana. The qualifications and first-hand experience of this research team will ensure successful completion of this project.

5

SELECTED REFERENCES Note: Because of page limitation, we only cited a few in the project description, and they are marked with an *. Aerts, J., G. Heuvelink, and M. F. Goodchild. 2003. Accounting for spatial uncertainty in optimization with spatial decision support systems. Transactions in GIS 7(2):211230. Armstrong, M. 1994. Requirements for the development of GIS-based group decision support systems. Journal of the American Society for Information Science 45:669677. *Averill, J. R. 1987. The role of emotion and psychological defense in self-protective behavior. In Taking Care: Understanding and Encouraging Self-protective Behavior, ed. N. D. Weinstein. Cambridge: Cambridge University Press. Baker, E. J. 1995. Public response to hurricane probability forecasts. The Professional Geographer 47(2):137-147. *Campanella, R. In Press. Geographies of New Orleans: Peoples, Patterns, Perceptions, and Place. Center for Louisiana Studies, University of Louisiana at Lafayette. Subtitle to be changed to "Urban Fabrics Before the Storm." *Campanella, R. 2002. Time and Place in New Orleans: Past Geographies in the Present Day. Gretna, LA: Pelican Publishing Company. *Campanella, R., and M. Campanella. 1999. New Orleans Then and Now. Pelican Publishing Company, Gretna, Louisiana, 1999. Chen, J., P. P. Chong, and Y. Chen. 2001. Decision criteria consolidation: A theoretical foundation of Pareto principle to Michael Porter's competitive forces. Journal of Organizational Computing and Electronic Commerce 11(1):1-14. Chen, Y. 1989. Organizational strategies for decision support and Expert Systems. Journal of Information Science 15:27-34. Chen, Y. 1988. An entity-relationship approach to decision support and expert systems. Decision Support Systems 4(2):225-234. Cooksey, R. W. 1996. Judgment Analysis: Theory, Methods, and Applications. San Diego, CA: Academic Press. *Cutter, S. 2003. The vulnerability of science and the science of vulnerability. Annals of the Association of American Geographers 93(1):1-12. *Edmonds, B. 1999. Modeling bounded rationality in agent-based simulations using the evolution of mental models. In Computational Techniques for Modeling Learning in Economics, ed. T. Brenner, 305-332. Boston, MA: Kluwer Academic Publishers. Edwards, W., and J. R. Newman. 1986. Multiattribute evaluation. In Judgment and Decision Making: An Interdisciplinary Reader, eds. H. R. Arkes and K. R. Hammond. Cambridge: Cambridge University Press. Erlich, I., and G. S. Becker. 1972. Market insurance, self-insurance, and self-protection. Journal of Political Economy 80:623-648. Goodchild, M. F. 2000. Introduction: Special issue on ‘Uncertainty in Geographer Information system’. Fuzzy Sets and Systems 113:3-5

1

Graves, P. E., and P. D. Linneman. 1979. Household migration-theoretical and empirical results. Journal of Urban Economics 6(3):383-404. Heinz Center for Science, Economics, and the Environment. 2000. The Hidden Cost of Coastal Hazards: Implications for Risk Assessment and Mitigation. Washington, DC: Island Press. Jankowski, P., and T. Nyerges. 2001. GIS-supported collaborative decision making: results of an experiment. Annals of the Association of American Geographers 91(1):48-70. Jankowski, P., A. Lotov, and D. Gusev. 1999. Application of multicriteria trade-off approach to spatial decision making. In GIS and Multiple Criteria Decision Making: A Geographic Information Science Perspective, ed. J.-C. Thill, 127-148. London: Ashgate Publishing Ltd. Kann, A. 1998. A Comparison of Approaches for Performing Uncertainty Analysis in Integrated Assessment Models. Working paper of Energy Modeling Forum, Stanford University. http://www.icis.unimaas.nl/publ/downs/99_03.pdf. Klein, G. A., J. Orasanu., R. Calderwood, and C. E. Zsambok, eds. 1993. Decision Making in Action: Models and Methods. Norwood, NJ: Ablex. *Kunreuther, H. 1992. A conceptual framework for managing low-probability events. In Social Theories of Risk, eds. S. Krimsky and D. Golding. Westport, CT: Praeger Publishers. Kunreuther, H., R. Ginsberg, L. Iller, P. Sagi, P. Slovic, B. Borkan, and N. Katz. 1978. Disaster Insurance Protection: Public Policy Lessons. New York: Wiley. Lam, N. S.-N., Y. Chen, K. Liu, and K. Pace. 2003. A national research center for decision making under uncertainty in hurricane climate change. Proposal submitted to NSF (not funded). Lam, N. S.-N. 2001. Spatial analysis for reducing uncertainties in human health risk assessment. Acta Geographica Sinica 56(2):239-247. *Lopes, L. L. 1987. Between hope and fear: the psychology of risk. In Advances in Experimental Social Psychology, ed. Berkowitz, Vol. 20. San Diego: Academic Press. *McAlister, A. 1987. Social learning theory and preventive behavior. In Taking Care: Understanding and Encouraging Self-protective Behavior, ed. N. D. Weinstein. Cambridge:Cambridge University Press. Pace, R.K., and J. LeSage. 2004. Spatial econometrics and real estate. Journal of Real Estate Finance and Economics 29(2): 147-148. Pace, R. K., and R. Barry. 1997. Quick computation of regressions with a spatially autoregressive dependent variable. Geographical Analysis 29(3):232-247. *Palm, R. 1995. Catastrophic earthquake insurance: patterns of adoption. Economic Geography 71:119-131. *Palm , R., and M. J. Hodgson. 1993. Natural Hazards in Puerto Rico: Attitude, Experience, and Behavior. Boulder: University of Colorado, Natural Hazards Applications and Information Center Monograph #56. *Palm, R., and M. J. Hodgson. 1992. After a California Earthquake: Attitude and Behavior Change. Chicago: Chicago University Press. Pielke, R.A. Jr., and R.A. Pielke, Sr. 1997. Hurricanes: Their Nature and Impacts on Society. Chichester: Wiley.

2

Renas, S. M. and R. Kumar. 1979. Climatic conditions and migration: an econometric Inquiry. Annals of Regional Science 12: 95-104. *Simon, H. A. 1997. Models of Bounded Rationality, Vol. 3: Empirically Grounded Economic Reason. Massachusetts: MIT Press. *Slovic, P., B. Fischoff, and S. Lichtenstein. 1976. Cognitive processes and societal risk taking. In Cognition and Social Behavior, eds. J. S. Carroll, and J. W. Payne. Potomac, MD: Lawrence Erlbaum Associates. *Slovic, P., H. Kunreuther, and G. F. White. 1974. Decision processes, rationality, and adjustment to natural hazards. In Natural Hazards: Local, National, and Global, eds. G. F. White, I. Burton, and R. Kates. New York: Oxford University Press. Stewart, T. R. 2000. Uncertainty, judgement, and error in prediction. In Prediction: Science, Decision Making, and the Future of Nature, eds. D. Sarewitz, R. A. Pielke, and R. Byerly, 41-57. Washington, DC: Island Press. *Stewart, T. R., and A. Bostrom. 2002. Extreme Event Decision Making: Workshop Report. http://www.albany.edu/cpr/xedm/.

3

APPENDIX

v. Lawson, Victoria and Jarosz, Lucy. Geography and Regional Science: Interpreting Geographies of Poverty: Rural Gentrification and White Poverty in the American Northwest.

APPENDIX

v. Lawson, Victoria and Jarosz, Lucy. Geography and Regional Science: Interpreting Geographies of Poverty: Rural Gentrification and White Poverty in the American Northwest.

Interpreting Geographies of Poverty: Rural Gentrification and White Poverty in the American Northwest “People’s families fall apart when they can’t make a decent living….what we are really talking about is a way that we can live in our home places with some dignity, some honorable work and a sense of a beneficial ‘community’” (Mayor of Mountain View, WA 1999). All across the American West, rural families are dealing with rapid changes to their livelihoods and communities. Many rural places are experiencing dramatic social and economic transformations as urban middle and upper class migrants bring new politics, new aesthetics and new economic activities into contact with longer-term resident populations (Nelson, 1999; Beyers and Nelson, 1999; Rudzitis, 1997). These changes have brought the paradox of both income growth and rising poverty rates. Along with these new income distributions within rural counties, shifts in the class composition of communities and new social tensions between residents are emerging. While there is a growing volume of research on the demographic and economic dimensions of these changes (Rasker and Alexander, 1997; Nelson and Beyers, 1998; Shumway and Davis, 1996; Barrett and Power, 2001), less attention has been devoted to the social and cultural tensions surrounding this process of rural restructuring and how poverty is understood within this context of transformation and change (Rudzitis, 1993; Cloke, 1997; Nelson, 1999). Initial findings from our pilot research indicate several nonmetropolitan counties with both high white poverty and high rates of in-migration over the last decade. In Washington state for example, Asotin county has a white poverty level in the 18-24% range and net migration of 19% between 1990 and 1998, and Pend Oreille county has white poverty at 18-24% of the population and net migration of 25.5% for the same period (county average in-migration of 7.7% for WA). In Montana, Sanders county has a white poverty level of 18-24% and in-migration of 17.5% during 1990-1998, and Madison county has white poverty at 18-24% and net migration of 15.6% (county average of net migration is 2.5% for Montana). Our proposal builds from these suggestive initial findings to examine in more detail the recursive relationships between the socio-economic forces producing rural poverty and the cultural processes through which the poor and rural poverty are identified by diverse residents of rapidly gentrifying areas in the rural American Northwest. These dynamics are important because understanding how poverty is interpreted, perpetuated or transformed under rural gentrification is crucial for revealing how divisions emerge and are maintained in communities. These divisions – particularly those marked by economic status and gender - often make dialogue about development difficult. Beyers and Nelson (2000: 14) ask how rapid growth leads to divisive politics? Our project extends and deepens this question by asking how divisive politics are rooted in understandings of rural poverty and poor people in places undergoing rapid growth and change. Specifically, we focus on understanding the relationship between representations of poverty and the region’s political economy of rural development. We adopt this focus in order to understand the possibilities for building inclusive communities in places with substantial numbers of people who find it hard to earn a living wage and sustain their livelihoods. This project will help policy makers and community members

1

identify ways of encouraging inclusive dialogues about growth, conservation, and quality of life, while placing a consideration of rural poverty at the center of questions about rural development. This research has broader theoretical significance as well. We argue for the importance of bringing cultural and ideological processes into analyses of class that, within the economic restructuring literature, have been largely economistic. At the same time, we bring geographical political-economy analysis to bear on cultural studies, infusing largely identity-based analyses with political-economic insights. More specifically, despite the current interest in whiteness within cultural studies, there is little empirical work which explores place-specific constructions of identity and community conflict among whites in rural settings undergoing gentrification (Jarosz and Lawson, 2001; Duncan, 1999; Gibson, 1996). Since large numbers of America’s rural poor are white, it is vital to theorize the ways poverty is represented and lived within these uneven processes of development (Jarosz and Lawson, 2001; Cloke, 1997; Hooks, 2000)1. The central questions of this project examine how people understand class and gender differences in the context of high in-migration and rapid development and change. The major research questions are: 1. What is the geography of white poverty in rural regions in the American Northwest? How do patterns of white poverty vary in accordance with rural restructuring? 2. What are the cultural processes through which the poor and rural poverty are identified and understood? How do community groups and individuals talk about the poor and identify rural poverty and difference among them? 3. How are cultural and economic processes mutually intertwined and embedded in places (in geographies and economic histories) to produce both certain cultural understandings of white rural poverty and certain kinds of outcomes in terms of perpetuating or transforming white rural poverty? Our pilot research in Washington (Jarosz and Lawson, 2001) suggests that situated and scaled analyses of discourses of socio-economic difference can help us understand community dynamics as people understand, accommodate and resist the global and national changes shaping the ‘new’ West (a term signifying substantive changes in the economy and culture of the American West; Rasker, 1995; Reibsame, 1997). Through this linkage of economic restructuring in particular places with attention to the micro-level representations of rural poverty and poor people, we work to understand both the empirical and discursive processes that can limit or enable the emergence of inclusive communities in the rural Northwest. Our research design builds from important prior work on economic and demographic restructuring in the Northwest (Beyers, 1991; Rasker, 1995; Rudzitis, 1997: Markusen and Schoenberger, 1979; Krannich and Greider, 1990). Our project joins empirical analysis of economic change with discourse analysis of socio-economic difference within communities. The question of how these conjoined processes shape discussions of rural development remains open. Our project will be conducted over a three-year period and consists of two phases. First, we investigate patterns of white poverty and restructuring across nonmetropolitan counties and rural 1

In 1997, the poverty rate for whites (10.5 percent) was lower than that for other racial and ethnic groups. However, the majority of poor people in the US were white. Among people falling under the official poverty level, 68 percent were white (Dalaker, 1999).

2

census tracts across the American Northwest.2 Second, we conduct focus groups and in-depth interviews with residents in four rural case study sites across the region. We will compare booming and more stagnant places characterized by either high levels of white poverty or inequality to investigate the ways in which social and cultural tensions over rural restructuring and poverty are being worked out in the Northwest. Through these combined approaches we examine our theoretical arguments about the ways in which poverty and socio-cultural differences are co-produced. Investigator Qualifications: This collaboration brings together two experienced field-based researchers. Vicky Lawson has fifteen years of fieldwork experience on migration studies, economic restructuring and inequality in both Latin America and the American Northwest including survey research and in-depth data collection. Lucy Jarosz has specialized in qualitative and political economy research on rural community and agrarian change in both sub-Saharan Africa and the Pacific Northwest with a focus on the material and discursive productions of inequality. The PIs’ recently completed pilot study on rural restructuring and cultural discourses about white poverty in Washington state forms the basis for this proposal (Jarosz and Lawson, 2001). Theorizing rural change and poverty Theoretical explanations of poverty in the U.S. fall into one of two broad categories that can be characterized as socio-cultural or structural (Albrecht, Albrecht, and Albrecht, 2000; Collins, 1996). Socio-cultural theories explain poverty as the results of individual responsibility, behavior or cultural traits (Murray, 1984). Socio-cultural explanations of poverty range from arguments that people choose poverty and are empowered to be poor by the welfare state (Galloway and Vedder, 1985), to arguments that poverty is a question of individual agency, wherein people consciously choose simpler lifestyles. Critics call the former position ‘blaming the victim’ and point to key socio-economic structures that explain poverty such as economic restructuring and uneven development under capitalism (Bluestone and Harrison, 1982; Wilson, 1987; Jones and Kodras 1990; Peet, 1987); poor economic development characterized by economic stagnation and low wages (Eggers and Massey, 1992; Belzer and Kroll, 1986); social discrimination such as racism, sexism, ageism (Fitchen 1981); and government and institutional policy and discourse (Schwarz and Volgy, 1992; Ryan, 1974). Our project combines three strands from rural poverty research: work on regional economic and demographic change processes, the political-economy of rural restructuring and cultural dimensions of rural change. The first strand links regional economic and demographic dynamics, examining migration, community change, and the character and quality of employment under rural restructuring (see for example Albrecht, Albrecht and Albrecht, 2000; Markusen and Schoenberger, 1979; Leistritz and Murdock, 1981; Barrett and Power, 2001; Nelson and Beyers, 1998; Lasley, Leistritz, Labao, and Meyer, 1995; Brown, Fuguitt, Heaton and Waseem, 1997). This interdisciplinary research links sectoral economic shifts; socioeconomic processes of investment, production, and demographic shifts (especially labor migration) across spatial scales and this provides a backdrop to understanding poverty dynamics. 2

Researchers have regionalized the Northwest in various ways (Rudzitis, 1993). For this project, we identify rural Washington, Idaho, Oregon and Montana as part of the rural Northwest.

3

In the American Northwest, work has been done on rural boomtowns looking at migration, job loss and the shift of employment from family-wage to minimum-wage jobs (Jobes, 2000; Rudzitis, 1997). Some authors link poverty to structural economic and demographic processes driving rural restructuring and others emphasize diverse pathways into poverty including choosing quality of life over higher incomes in other places (Barrett and Power, 2001; Jobes, 2000; Rasker and Alexander, 1997; Beyers and Nelson, 2000). The second strand of rural poverty research examines the political economy of agrarian change (Watts, 1983; Mackintosh, 1989; Friedland et al, 1991). Global in its scope, this work examines not only the changes in agrarian economies and rural environments due to colonization and capitalist development, but also links local changes in social relations and land use practices to the globalization of food and agriculture (Goodman and Redclift, 1991; Le Heron, 1993; Jarosz 1996; Goodman and Watts, 1997; Jarosz and Qazi, 2000). A key contribution of this research is its focus on the material forces that shape how rural class fragments are organized and how their interests articulate with other classes and interest groups. Recent work on the importance of gender difference in understanding rural change has also broadened this stream of research in important ways (Hoggart and Buller, 1987; Marston, Lowe and Whatmore, 1990; Sachs, 1996). The strength of this work comes from its analysis of material class and gender relations, however this political economy research tends to ignore the cultural politics shaping rural development trajectories and social conflict (Fagen, 1999; Moore, 1993; Barnes, 1995). We extend this stream of work by responding to Cloke’s (1997: 253) call for “(en)culturing of political-economy approaches” in order to reveal the ways in which material and ideological processes come together in the production of poverty. We argue that a detailed engagement with the ways in which economic restructuring and discourses of class difference work together can help policy makers to understand the political significance of poverty in rural areas. The third strand of research we draw upon examines rural restructuring and cultural change. As Nelson (1999: 39) argues about the American West “…the interwoven nature of economics and culture must be recognized because rural restructuring in the 1990s includes both changing economic characteristics of communities as well as new cultural productions, competing meanings and a changing sense of place based culture(s) in these growing rural areas”. This body of work begins from rural localities rather than from a broad focus on sectoral shifts in agrarian political-economic relations. Researchers in this stream integrate analyses of changing rural property relations, shifting settlement patterns and the relations between rural and urban places (Newby, 1985; Urry 1985; Marsden, Lowe and Whatmore, 1990; Page and Walker, 1991; Page, 1996) and link these to structural socio-economic processes (Marsden et al, 1993; Marsden and Murdoch, 1994). This work also engages with concepts of class, gender and race, taking seriously the ways in which rural societies are differentiated and so how some groups are marginalized through processes of rural restructuring. This third strand includes the rich tradition of contextual, cultural and economic research on rural poverty in the US which has analyzed particular localities in order to understand how local history, social structures, local meanings and values combine to shape the ways in which communities respond to economic change. Much of this research has focused on historically entrenched poverty in Appalachia and the deep South (Oberhauser, 1995; Cloke, 1997; Billings, Norman and Ledford, 1999; Duncan, 1999; Gibson, 1996; Deavers and Hoppe, 1992). There is

4

also comparative case study research on rural poverty focusing upon the northeast and south which seeks to uncover how regional cultures influence responses to economic distress (Scott, 1995; Ramsay, 1996; Duncan, 1999; Fitchen, 1981). Other work examines particular occupational groups such as farm-workers or miners in places facing economic decline (Griffith and Kissam, 1995; Scott, 1995). Our project by contrast, examines rising white poverty in the midst of dynamic economic growth in boomtowns in the American Northwest and so shifts the question from how historically entrenched poverty is understood and perpetuated, to an emphasis on how poverty is constructed and understood in relatively affluent places that have not been identified as core regions of rural poverty. Core regions of rural poverty are identified by Shaw (1996) as including the Texas-Mexico border, parts of Arizona and New Mexico and South Dakota, the Mississippi Delta and in central Appalachia. With the exception of Appalachia, rural poverty is often racialised as non-white and is regionalized, centering upon African Americans in the South, Hispanics in the Texas border region, and Native American populations elsewhere (Jones and Kodras, 1990). This regionalization has operated to associate white rural poverty primarily with Appalachia and this serves to regionally contain and to stereotype the white rural poor in popular culture (Sandell, 1997; Hooks, 2000). Popular culture representations of the ‘new West’ frequently emphasize quality of life, spectacular scenery, and the in-migration of urbanites to boomtowns (Egan, 1998; Limerick, 1987; White, 1996). However, there is rising white poverty in the American Northwest and cultural processes of representing the poor are engaging white identities in particular ways (Jarosz and Lawson, 2001). We know relatively little about these representations and experiences of white poverty, especially in relation to new boomtowns. We draw on recent work in cultural studies to explicitly theorize the connections between cultural representations of poor people and processes of economic restructuring – responding to Cloke’s call for an “(en)culturing” of political-economy analyses of rural change. ‘Enculturing’ Analyses of Rural Restructuring and Poverty Our theoretical goal is to extend political-economy research through the incorporation of cultural and ideological processes in order to build a richer understanding of how poverty and sociocultural differences are co-produced. Recent work in cultural studies examining the cultural construction of white difference can enrich geographical research. Specifically, cultural studies work theorizes i) why white poverty remains relatively neglected; ii) the ways in which representations of white poor rural folk obscure the material foundations of poverty; and iii) the ways in which this works against the emergence of a politics of engagement that can lead to inclusive rural communities. We argue that geographical research can reveal situated discourses about the white rural poor in the context of regionally specific geographies of political and economic change. In the context of rural restructuring a series of terms and themes echo in popular culture representations of white poverty. For example, Fitchen’s study (1981) of white poverty in upstate New York identifies a series of corrosive stereotypes held by dominant society about poor people in the community. She notes that these stereotypes not only shape how others see the poor, but that these representations become internalized by the poor themselves. White rural poor people are often referred to as lazy, trashy and resistant to change and these labels operate

5

to position the poor and to blame their poverty on themselves (Duncan, 1999; Gibson, 1996; Creed and Ching, 1997; Jarosz and Lawson, 2001). Watkins (1993) links these representations to shifts in the U.S. economy in the late twentieth century. He argues that as the U.S. moves ever more towards a transnational service economy based on information technologies and accelerating consumption, rural working class folks are constructed increasingly as obsolete. Watkins theorizes that class is being ideologically constructed in terms of consumption and style differences. Through this process, the discourse of class position as choice emerges. This move, simultaneously reassures middle classes that they are different from poor whites while obscuring the material processes of economic restructuring that produce the harsh realities of poverty. In recent decades, anxieties around rapid economic restructuring, coupled with concerns about the future potential for middle class upward mobility (driven by corporate downsizing and government disinvestment in the social sphere) emerge in discourses of ‘white trash’ and ‘rednecks’. Because white poor cannot be dismissed as a ‘non-white other’ we theorize, along with Watkins, that white rural poor folks are often understood in terms of class and gender difference which focus on class-as-lifestyle rather than on underlying political-economic processes of changing opportunities and displacements (Watkins, 1993). For example, in both the national and regional media and popular culture, we found evidence of class-as-lifestyle representations of rural poor folk. Recently, the Seattle Weekly, a middle class, urban paper ran a discussion of degentrifying Seattle. This discussion, in the wake of the most recent upturn in property values, concerned how to make the city affordable again. One of four spoofs defined the “Redneck Approach” which encourages residents “…to replace one set of curtains with a Confederate flag, another with aluminum foil, and a third with a yellowed bed sheet. Those with grassy front yards would have to park at least one vehicle on the lawn…” (Stusser, 1997).3 These kinds of images obscure class divisions with ridicule and laughter. We ask whether similar processes of representing poor people as obsolete, resistant to change, and unsophisticated are operating in Northwest communities. Much of contemporary development discourse on the American West also represents local people (cowboys, timber workers, small-scale fishers, farmers and ranchers) as stagnant, traditional and politically conservative (Cronon, 1996; Shanin, 1990). Richard White (1996) identifies the gap between the middle class views of the West as a place of leisure, recreation, aesthetics, and spiritual renewal as contrasted with the views of those who work on the land and depend upon it for their survival. Struggles over the meaning of nature and the environment in the West are embedded within debates about allowing horses on mountain trails and limiting the extent of logging and fishing. Cowboys who graze their herds on public range land are defined as obsolescent, relics of a bygone past by environmentalists who aim to preserve the ecological integrity of the vast public lands in the American West (Egan, 1998). Class-as-lifestyle representations emphasize cultural variables and consumption as the defining markers of difference and thus reinforce the notion that class position is purely one of individual choice, not structural economic processes. When 3

The use of humor to represent and distance working class, poor whites in the 1990s is also signaled by the enormous success of Jeff Foxworthy. He trades on his Southern accent and Georgia origins to self-identify as a redneck despite his middle class upbringing and his sumptuous Los Angeles lifestyle (Sack, 1996). The nationwide success of his stand-up routines, albums and books reinserts the term “redneck” into mainstream, middle class culture in ways that reproduce notions of poor whites as unsophisticated, outdated and comic. His jokes rely on themes of lack—of sophistication and education. For example, “you might be a redneck if…you’ve ever been too drunk to fish.”

6

represented as lifestyle “choices” the differences between the economic success of newcomers to Western counties and the dislocation of old-timers can be scripted as questions of willingness to change. One of the hallmarks of these discourses of white rural poverty is that they invoke imagined rural spaces which are constructed as wild, empty, and backwards. We found these images in the humor discussed above, on TV and in film.4 These abstract discourses of rural people and places obscure the material processes of impoverishment actually associated with economic restructuring and also serve to distance middle and upper class whites from poor whites. These mechanisms of obscuring and distancing allow middle and upper class whites to ignore their own role in the remaking of rural places through rising land prices and changing uses of public lands which prevent the poor from maintaining their livelihoods. Representations of the poor as different, of class position merely being a choice of lifestyle, are naturalized through characterizations of rural spaces as filled with old cars, unkempt homes, and unwashed children. By implication then, people could choose to ‘clean up’ their lifestyles or be willing to take service sector jobs instead of working in the woods. These homogenizing discourses of class as choice of lifestyle are not situated in particular political-economic contexts and histories of restructuring. While clearly, some people choose to remain in difficult economic circumstances for lifestyle, family or other commitments, they might also choose better paying work if it were available in those places. People’s choice to remain in a community often gets conflated with causes of poverty and results in homogenizing discourses of individual responsibility which obscure structural, economic causes. We argue for a geographical research agenda that situates these discourses of class difference in specific places and socio-economic circumstances in order to investigate the relationships between rural gentrification and poverty and how that poverty is understood, ignored or denied by different actors involved in community change. Our project focuses upon how cultural processes which construct lifestyles and occupations as obsolescent operate within the economic geography of rural restructuring in the West. We investigate the specific mechanisms by which these cultural processes interact with economic and demographic changes to constitute and reproduce poverty. We ask a series of empirical and theoretical questions. How are cultural constructions of class difference embedded in localized socio-economic processes of inequality? How is difference constructed among whites, when poverty cannot be construed as a racialized problem among non-whites? For example, do stigmatized representations of the poor shape who has access to new service economy jobs? Do discourses of poverty as lifestyle choice operate to obscure the roles of urban in-migrants and environmentalists in contributing to inequality? Do working poor construct identities (like redneck) in response to economic and cultural change that are bound up with masculine identities and jobs (logger, rancher, miner) and do these identities mitigate the kinds of work that rural poor are willing to do as the economic base changes? How important are these sorts of mechanisms, how do they operate, and how do they vary spatially?

4

The 1972 movie hit Deliverance is a classic example. The movie represents the Georgia woods and people as wild, hostile, lawless and uneducated. These violent people and places eventually emasculate one of the urban protagonists and this is grounds for destruction of the valley by hydroelectric development.

7

Rural restructuring and poverty in the American Northwest We analyze the human geography of these processes in an area characterized by both economic boomtowns and deepening rural poverty (Rasker, 1998). We build on a strong tradition of research on restructuring and poverty in the American West (Markusen and Schoenberger, 1979; Krannich and Greider, 1990; Cook, 1995; Cook and Jordan, 1997; Jobes, 2000). Rural poverty in the American West has been explained by economic restructuring, isolation and distance, and cultural traits and behaviors, immigration, and the obsolescence of particular sorts of occupations such as those of rancher, cowboy, and logger (Belzer and Kroll, 1986; Fossum, 1993; Taylor, Martin and Fix 1997; Egan, 1998; Wilkenson, 1991). Change is often characterized in terms of conflict between the 'old' West of miners, cowboys, timber workers, farmers and ranchers with the 'new' West of tourists, gentleman farmers, vacation homes, ski resorts, and upscale recreational amenities, retailers, and theme towns, but the degree of this conflict is disputed (Smith and Krannich, 2000; McCarthy1998, White 1996). Recent work has examined the ruraldestined migration flows of the last three decades, from the movement of ‘back-to-the-landers’ in the seventies to the wave of migrants from urban places into rural places during the nineties. For example, Jobes (2000) analyzes the aspirations and experiences of in-migrants to the Gallatin Valley, Montana and Nelson (1999) focuses on the conflicts emerging around the influx of ‘newcomers’ to four rural communities in the West. This body of work demonstrates that contemporary landscapes of the American West are being reworked through the shift from extractive resource and agrarian economies to a rise in the service sector (Mitchell, 1998; Brown, 1995; Nord, 1994; Lyson and Falk 1993). These shifts are driven by exhaustion of finite resources, changing environmental attitudes and pressures, changing farm subsidies, a reduced availability of public lands for grazing, and reduced access to public lands for logging. This complex of forces, coupled with labor saving technologies has lowered the demand for labor in these industries (Nelson and Beyers, 1998; Rasker and Alexander, 1997). This is coupled with the arrival of urban migrants and attendant changes to rural communities as wealthier populations bid up land and housing prices and work to change uses of public lands. The result has been rising poverty and inequality in recent decades. In the 1970s real pay per job in Montana was just 5 percent below the national average and this measure plummeted to 33 percent below in 1998 (Barrett and Power, 2001: 19). Barrett and Power (2001: 20) also report a trend that resulted in average incomes of the richest fifth of families in the Mountain West being ten times greater than incomes for the poorest fifth by 1996. Furthermore, in the 1990s population growth and in-migration increased more rapidly in rural than in urban areas. Numerous factors contribute to this influx, including issues of quality of life, the rise of telecommuting and urban flight (Beyers and Nelson, 2000; Rudzitis, 1993). These dynamics have created a paradox as Rasker (1998:63) poses it; “[W]hy, for example is the poverty rate in Eton County, Idaho and Ravalli County, Montana increasing in spite of a boom in population and aggregate economic activity”? This unresolved contradiction between aggregate economic growth and poverty, especially in cases where the poverty rate increased in the fastgrowing and seemingly prosperous communities, is central to our project. The local geographies and implications of poverty dynamics during the nineties have not been conclusively understood (Power, 1996).

8

Our project extends current research in three ways. First, it builds a new theoretical synthesis that conjoins cultural politics and the construction of identity and place to a political economy of rural development (Marsden, 1999; Fagen, 1999). Second, our project systematically compares socio-economic and cultural processes of poverty and inequality across localities with different histories and dynamics of economic change. Third, we argue for comparative, qualitative research that engages the ways in which economic and cultural processes are deeply embedded in places and place histories. We argue that a full understanding of these processes must emanate from research with residents and recent in-migrants themselves. Through their experiences and perspectives, community members are uniquely positioned to reveal processes of identity construction and their relation to poverty and wealth. Research Design We propose a two phase research design. Both PIs will be equally involved in the fieldwork and analysis and will conduct parallel procedures in different communities to expedite our comparative research design. The research focuses on the nonmetropolitan Northwest (Washington, Oregon, Idaho and Montana). First we conduct a descriptive analysis of the relationship between nonmetropolitan restructuring and poverty, employing county and tractlevel data, across the Northwest. The second phase involves fieldwork in four case study sites, selected in phase one. Fieldwork will include focus groups and in-depth interviews with long time residents and recent in-migrants and participant observation at community meetings. Phase One (April 2002 – December 2002): Investigates patterns of white poverty and inequality in nonmetropolitan places in the American Northwest (research question #1, page 2). Our first goal is to investigate geographical patterns of economic restructuring in relation to patterns of poverty in nonmetropolitan counties in Washington, Oregon, Idaho and Montana. We will employ data at the county and census tract scales to build a fine grained geography of poverty patterns overall, and white poverty in particular. Our second goal is to identify case study sites for Phase Two. a) Identifying geographies of nonmetropolitan restructuring: Scholars have developed new economic base models for capturing the complexity of forces shaping nonmetropolitan growth in the West (Nelson and Beyers, 1998; Rasker, 1995; Beyers and Lindahl, 1996; Nelson, 1997; Nelson, 1999; Power, 1996). Specifically, Nelson and Beyers (1998; 2000) employed data from the U.S. Bureau of Economic Analysis Regional Economic Information System (http://fisher.lib.virginia.edu/reis), to construct a typology of counties based on structural differences in the forces contributing to economic growth. This work clusters nonmetropolitan counties on their relative concentration, and lack, of certain sources of income (including farm earnings, non-farm earnings, dividends, rental income, pensions, transfer payments). Using data from 1985 to 1994 they produced a typology of four different structural types of economies which accounted for 60% of all nonmetropolitan counties in the West. Each cluster of counties is characterized by distinct forces of economic change. These clusters were then correlated with population growth patterns to identify growing and stagnating places across these structurally distinct economies. We will update their methodology in two ways. First, we will use data from

9

1989 to 1999 in order to capture change throughout the nineties.5 Second, for our cluster analysis, we will work with both the K-means algorithm and Wards hierarchical algorithm to determine which approach yields more meaningful clusters of variables and generate clusters that contain relatively large numbers of counties with distinctive structural characteristics. We will then correlate population growth and structural differences in county economies in order to select a subset of the 145 nonmetropolitan counties in Washington, Montana, Oregon and Idaho. For each of the four (we may change this based on our updated analysis) distinct economic structures identified (i.e. the clusters), we will select five counties exhibiting the greatest population growth and five counties exhibiting greatest demographic stagnation (yielding 40 counties total for the next step). b) Locating white poverty in areas of nonmetropolitan restructuring. Here we construct a human geography of poverty, initially comparing patterns of white and non-white poverty, for these 40 restructuring counties at the end of the nineties. We will then focus in more detail on white poverty in relation to nonmetropolitan restructuring. We will map white poverty in two ways. First we will map patterns of absolute poverty, using the percentage of white populations in poverty at the county and rural tract level. We will also construct an index of white inequality (or relative poverty) at the county level (calculated using the difference between the highest and lowest median income for white householders measured at the tract level within that county) for each of our 40 growing or stagnating counties. In addition, we will analyze poverty patterns in relation to age, gender and education of householder head at the county and rural tract level in order to construct a rich understanding of the patterns and characteristics of white poverty. We will conduct these analyses with year 2000 U.S. Census Bureau data, which will be available at the Center for Social Science Computing and Research (CSDE) at the University of Washington in August 2002. Two sorts of outcome emerge from phase I. First, these mappings of absolute and relative poverty will contribute to scholarship on the links between poverty and restructuring, by linking poverty patterns to growth and stagnation that emerge at different spatial scales (county and census tract). The latter issue is particularly important in the American West, where counties are very large, and rural tracts will disaggregate county-level patterns and provide a more fine-grained analysis of geographies of poverty. Second, from these mappings, we will then select four counties (two growing and two stagnating) from within which our study sites will ultimately be selected. c) Selection of case study sites. In phase II, we investigate how different groups and individuals in the nonmetropolitan Northwest understand and culturally represent white poverty. We compare poverty processes in growing and stagnating places because while we do not argue that restructuring and population growth cause white poverty, we do hypothesize that inequality among whites increases under gentrification and that poverty becomes more stigmatized as wealthy ex-urbanists are juxtaposed with long time rural residents. Our qualitative research is designed to reveal and analyze the cultural processes of stereotyping and differentiation or cooperation, set in motion by economic and demographic changes across the region. To understand the role of population growth and structural economic change, we compare one growing county with high absolute white poverty and one stagnating county with high absolute white poverty. We will also compare one county with high white inequality across rural tracts 5

Professor Beyers is a colleague at Washington and he has graciously agreed to provide access to his data sets and to consult as we rework this methodology.

10

and high population growth with a stagnating county with similar patterns of white inequality. Within each of the four counties identified above we will then select specific nonmetropolitan communities (populations of 2,500 or less) for in-depth qualitative analysis in Phase 2. Since these counties are part of larger clusters representing distinct structural types, insights from our case studies will have broader relevance. The final selection of study sites within these counties will be made after triangulating three additional sources of information. First, a detailed mapping of white poverty within each county for 1999 and 2000 employing census tract-level measures of total household income by race, gender, age, median household income and our measures of county-level white inequality (education variable not available on 1990 STF 3). Second, an investigation of research and media coverage of economic change within the county to understand the nature and specific sites of economic change and the issues it is raising. Third, a week-long visit to each county to conduct informational interviews with county officials, newspaper reporters, educators, and religious leaders. Phase Two (January 2003-October 2004): Here we focus on questions 2 & 3 (page 2), looking at how community members differentiated by class, gender rural/urban origins and length of residence understand poverty and the relationship between restructuring and poverty. We also examine the specific mechanisms by which cultural and economic processes interact to constitute and reproduce poverty in place. We employ a qualitative research design (building from Fitchen, 1981; Ramsay, 1996; Scott, 1995; Duncan, 1999) that focuses upon how cultural processes interact with socio-economic restructuring, within spatial and historical contexts, to maintain or reframe rural poverty (Cloke, 1997; Kushnick and Jennings, 1999). We will construct a richly textured historical analysis of economic, demographic and political restructuring in these places over the last thirty years during which a series of contemporary waves of migration into the nonmetropolitan West have occurred (Wyckoff and Dilsaver, 1995; Limerick, 1987; White, 1995). These analyses will provide a broader context for the in-depth data from focus groups and interviews. Each PI will spend two months in a separate study site, and so as a team, we will cover the four study sites over the course of two summers of fieldwork (summers 2003 and 2004). a )Gaining access to research participants in the study sites. In each community we will begin with a series of informational interviews with gate keepers such as mayors or city managers, social service agency directors, journalists, educational, business and religious leaders (n=8 per study site). We developed this approach in our pilot study and it was very effective for gaining access to a diverse list of community members who ultimately became research subjects, including both long-term residents and in-migrants of various class positions. Our approach is also effective for understanding community histories, economic change and social and political issues associated with in-migration and restructuring. In our pilot study, we were successful in gaining access to both wealthy urban in-migrants and long-term residents of the valley, many of whom were ‘stitching their lives together’ with an array of jobs and strategies. Following Jobes (2000), we will select participants from the community itself and a surrounding radius (Jobes used 15 miles and we will adapt this as appropriate). In our pilot research, these initial interviews generated substantial support from key leaders, resulting in invitations to attend planning meetings about the economic future of the valley. We will follow this same approach, asking gatekeepers for suggestions of people to contact who are both long term residents and recent migrants (those who have migrated to the community in the nineties) for possible

11

participation in our study. This snowball approach was very successful in our pilot project and is particularly important for identifying recent migrants to the community, since random sampling procedures typically suffer from the 'rare elements' problem of locating sufficient numbers of migrants (Bilsborrow, 1984; Lawson, 1999; 2000). b) Conducting focus groups and in-depth interviews. We will contact a diverse range of community members across various social locations (class, gender, migrant status, rural/urban origins). We will conduct 4 focus groups (n = 8 per group, 32 participants per community) in each field site for a dialogue about rural growth or stagnation and readings of conflicts and the nature of key differences among community members. Each focus group will be comprised of similar groups of people (put together in terms of combinations of migrant status, gender and class position). Our initial contact with potential participants will involve a screening conversation to determine first their willingness to participate, and second whether they meet our criteria for inclusion. We will select participants based on migration status, gender and class location (using for example, information about occupation, household income, education, prior place of residence). These focus groups will provide information about situated knowledge of rural change, migration, difference and conflict. Methodologically, focus groups have the advantage of fostering dialogue among individuals that often reveal points of tension or disagreement that sets the context for in-depth interviews (Krueger, 1993; Knodel, 1993). We will then select 4 participants from each focus group and 14 additional community members for in-depth, semi-structured interviews (total = 30 in each field site). These will provide more nuanced information, situated in oral histories and the lived experiences of community members. This research design builds on work in cultural geography which has focused on people who actively interpret their identities, make meanings about the places in which they live, and/or their mobility patterns (Chapman, et. al., 1985; Hugo, 1985; Skeldon, 1985; Silvey and Lawson, 1999). We have employed in-depth interviews both in Lawson’s (1999; 2000) work on migration and economic restructuring in Ecuador, and our pilot research in Washington state (Jarosz and Lawson, 2001). These interviews are an important tool for uncovering the links between understandings of poverty and rural restructuring and investigating whether these understandings are systematically linked to gender, class, ethnicity and migration status. This methodology yields a total of 128 focus group respondents and 120 in-depth respondents (with 64 of the latter having also been in focus groups) across the four field sites. In both focus group and in-depth settings we will gather data that reveal the specific mechanisms by which cultural constructions of class difference are embedded in localized socio-economic processes of inequality. We will pose questions to elicit the ways in which different community members talk about changes to their communities and their understandings of class difference. We will investigate respondents’ diverse interpretations of the causes and consequences of their own or others’ poverty. We anticipate that this will reveal diverse pathways into poverty (e.g. long-time residents or newcomers, choosing lifestyle over income or unable to find a decent living, etc.). We will also explore how respondents connect interpretations of economic change or relative stagnation to the presence of poverty in their communities. We will explore specific mechanisms shaping these processes such as how representations of the poor affect their access to local employment opportunities; or whether the poor construct identities in response to economic and cultural change that mitigate the kinds of work they will do, or other involvements

12

they have with the community. We will also pose questions about the ways in which gender difference is understood and how interpretations of community change and poverty are themselves gendered. Our in-depth interviews will examine respondent’s residential and migration history, socio-economic characteristics, work histories, personal accounts of economic restructuring, interpretations of the causes of restructuring, and accounts of social and political conflicts within each community. All interviews and focus groups will be tape-recorded and transcribed. c) Participant observation at community meetings. One of the challenges of analyzing social and cultural tensions arising around community change or stagnation is the sensitivity of these issues. One of our central purposes is to understand how material and cultural differences produce and perpetuate difference and inequality. Visiting community meetings dealing with economic restructuring, and addressing alternative visions of future development in our study sites can provide critical insight into the cultural and material differences that divide communities. In these public settings, as people struggle over land use issues, over tourism as an economic strategy, over regulation of resources and access to public lands, many of the dynamics that produce cultural and social difference are writ large. Through these visits in each research site we will deepen our understanding of the issues that communities struggle over, and how these struggles construct, resist, or reformulate discourses of difference within the community. d) Comparative data analysis. We will compare groups by migrant status, class and gender within each community and across the comparison sites. We will examine how differently situated subjects understand poverty and the connections between growth/stagnation and poverty. Drawing on results from our quantitative study, as well as our detailed historical analysis of economic, demographic and political restructuring in these places over the last thirty years, we will also examine contextual differences in these processes, by linking particular findings to the specificities of place and local experiences. Our transcribed data will be analyzed using Atlas.ti to code and search for specific concepts and patterns in the data. This software links specific terms to categories of respondents, and to the context in which terms or ideas were used. This is an invaluable tool for building a content analysis of our large volume of textual data. From this, we then construct a discourse analysis in which we analyze representations and pay special attention to linguistic forms such as metaphor and verbal imagery characterizing class difference and interpreting gendered poverty. Through the analysis of our interview data we aim to answer questions such as: Do systematic patterns (in terms of class/gender/migrant status etc) emerge in how community members understand and interpret poverty in the context of restructuring versus stagnation? Do discourses about the rural poor serve to obscure the relationship between restructuring and the causes of poverty? What are the diverse forms that white poverty takes in the Northwest? What are the pathways into poverty and what are the social and cultural mechanisms that reproduce or transform the lives of poor people? Our larger analytical goals are to build theoretical understandings of the links between economic restructuring and the cultural constructions of identity and difference in rural places. These theoretical and empirical insights will have implications for encouraging inclusive dialogues about growth and quality of life in rural boom towns. At the conclusion of our project (2005), we will develop and conduct workshops in our study sites to report our key findings back to community members. In addition, our quantitative and transcribed data sets and will be archived at CSDE at the University of Washington.

13

Contributions of this Research Our project makes three contributions. The first is empirical. Our work focuses on places where a large portion of those in poverty are white and so poverty cannot be racialized as a non-white issue. In these contexts, we examine how poverty is understood in the context of economic restructuring and migration in the Northwest. As the urban/rural divide becomes increasingly more porous and contested as urbanites move to rural spaces, it becomes ever more important to understand the ways in which marginalized populations are defined. Our second contribution is theoretical. This project extends a research agenda on the recursive relationships between political-economic restructuring and the socio-cultural construction of class difference. Our situated analyses of restructuring and poverty will contribute to our goal of bringing cultural and ideological analysis into creative tension with political-economy approaches – thus ‘enculturing’ political economy work. Finally, this research will be informative to community members and policy makers as they work to address the problems of poverty and to create inclusive communities in the face of rapid change. Our project will produce a nuanced geography of white rural poverty in the American Northwest as well as rich understandings of how this poverty is defined and explained. We will disseminate results from phases one and two at annual AAG meetings in 2003 and 2004. We will also attend the American Studies Association meetings to present our overall research findings in 2005. As noted above, we will also develop workshops with policy-makers and community groups in our field sites and then disseminate materials that are refined through dialogue with community members. As importantly, we will employ our findings in our undergraduate and graduate teaching at the University of Washington. This project will result in journal articles including a piece on the geography of poverty and restructuring in Economic Geography and another on “enculturing” political-economy in Society and Space as well as in popular outlets such as Atlantic Monthly. We will also develop a book addressing social and cultural change in the context of the ‘New West”. Results from Prior NSF Support Lawson’ previous NSF award from 1995 to 1998, was titled ‘Household Relations and Gendered Migration in Latin America’, grant SBR-9511129. The GRS Program award supported fieldwork in Ecuador at $102,742 (Lawson’s portion of a larger collaborative grant with Dr. R. Bilsborrow at UNC and Dr. A. Morrison at Tulane University, and subsequently Dr. E. Katz at UC Berkeley). The project focused on the role that household structures and power relations play in shaping migration decisions and gendered patterns of migrants’ labor market participation. The research also developed an innovative layered methodology, combining quantitative and qualitative data and analysis and also considered individual, household and community scales of analysis. Conclusions from the research have generated several types of product. The first is a comprehensive data base detailing household structures and employment relations and histories for both migrants and non-migrants over age 15 years and migration histories where relevant, in both Quito (their destination) and in key origin places. The destination survey data set,

14

completed in 1996 includes 632 households and 1406 individuals. The origin survey data set, completed in 1997 includes 1408 households and 1929 individuals sampled from the 5 provinces sending the most migrants to Quito. In addition, we have compiled a volume of 21 transcribed, in-depth interviews with in-migrants in Quito. These data are all available from Lawson, Department of Geography, University of Washington, and are archived at CSDE at the University of Washington. These data have also been distributed to policy-makers and researchers in Ecuador, through Centro de Estudios y Promocion para el Desarollo Social (CEPAR) in Quito. The second product is a series of journal articles in leading outlets (see short vita attached). The third is an educational product. Rachel Silvey worked as an RA on this project, co-authored an Annals article and earned her doctorate, looking at similar theoretical questions on gendered migration in Indonesia. She is now Assist. Professor of Geography at University of Colorado, Boulder. Kim Van Eyck also worked as an RA on this project, engaged in fieldwork with me in Quito and is completing her dissertation research after fieldwork in Colombia. In Ecuador, three local researchers collaborated on this project, Gustavo Rodriguez, Ana-Maria Albuja and Lastenia Rumbo. Both Ms. Albuja and Mr. Rumbo drew on this project to complete advanced degrees. There has been a broader impact on students in my undergraduate development classes and graduate seminars who have learned from my fieldwork experiences and articles that I use in my classes. Fourth is a learning and evolutionary project – the emergence of a new, but related research direction. While at first blush, this new research may seem unrelated, I am now pursuing a series of key themes that emerged from this prior grant. Initially funded by a Royalty Research Fund Award (1998-2000, $29,000, collaborative with Lucy Jarosz), I investigate the relations between migration, restructuring and discourses of poverty. These questions emerged from this prior NSF in which I investigated links between discourses of modernization, neo-liberal development and migrants’ understandings of restructuring and their own marginalization. his work has argued that migrants are complex subjects whose experiences of migration are socially constructed and politically contested. This project also revealed that a migrants’ social positioning (in terms of age, class, gender etc) profoundly affects destination work and household experiences and fortunes. Our current proposal brings these insights together and examines discourses of difference emerging in communities between migrants and long-term residents. Jarosz’ previous NSF award was ‘Electronic Information Skills for Collaborative Learning in an Undergraduate Geography Curriculum’ award SBR-9650348 from July 15, 1996 to June 30, 1999. This award from the Division of Undergraduate Education, Instrumentation and Laboratory Improvement Program provided $46,000 to develop and equip the Geography Collaboratory at the University of Washington. The Collaboratory provides space and equipment for students, faculty and librarians to collaborate and to enhance learning through synergistic exploration of course topics using electronic information skills. The Collaboratory is used for a wide variety of GIS and human geography courses and is the site for the Department’s annual Undergraduate Symposium which presents undergraduate research to members of the campus and Seattle metropolitan communities.

15

References Cited Albrecht, D.E., C.M. Albrecht, and S.L. Albrecht. 2000. ‘Poverty in nonmetropolitan America: impacts of industrial, employment, and family structure variables’ Rural Sociology. 65:1:87-103. Barnes, T. 1995. ‘Political economy I: 'the culture, stupid'’ Progress in Human Geography. 19:3:423-431. Barrett, R. and T. Power. 2001. Post-Cowboy Economics: Pay and Prosperity in the New American West. Washington DC: Island Press. Belzer, D. and Kroll, C. 1986. New jobs for the timber region: economic diversification for northern California. Berkeley: Institute of Governmental Studies, University of California, Berkeley. Beyers, W. 1991. ‘Trends in Service Employment in Pacific Northwest Counties: 19741985’ Growth and Change 22:4:27-43. Beyers, W. and D. Lindahl. 1996. ‘Lone eagles and high fliers in the rural producers services’ Rural Development Perspectives 11:2-10. Beyers, B. and P. Nelson. 1999. "Using Economic Base Models to Explain New Trends in Rural Income," Growth and Change, Vol. 29, No.3 (summer), pp. 295-318 Beyers, W.B. and P.B. Nelson. 2000. ‘Contemporary development forces in the nonmetropolitan west: new insights from rapidly growing communities’ Journal of Rural Studies forthcoming. Billings, D.B., G. Norman and K. Ledford, eds. 1999. Confronting Appalachian Stereotypes. Lexington: University Press of Kentucky. Billsborrow, R. 1984. Migration Surveys in Low Income Countries: Guidelines for Survey and Questionnaire Design. London: Dover. Bluestone, B. and Harrison, B. 1982. The deindustrialization of America. New York: Basic Books. Brown, B.A. 1995. In timber country: working people’s stories of environmental conflict and urban flight. Philadelphia: Temple University Press. Brown, D., G. Fuguitt, T. Heaton and S. Waseem. 1997. ‘Continuities in size of place preferences in the United States, 1972-1992’ Rural Sociology 62: 408-428. Chapman, M. and M. Prothero (eds.). 1985. Circulation in Population Movement; Substance and Concepts from the Melanesian Case. London: Routledge and Kegan Paul.

1

Cloke, P. 1997. “Poor Country,” in P. Cloke and J. Little eds. Contested Countryside Cultures. New York: Routledge. Cloke, P. and J. Little eds. 1997. Contested Countryside Cultures. New York: Routledge. Cloke, P and N. Thrift. 1990. ‘Class and Change in Rural Britain’ in Marsden, T., P. Lowe, and S. Whatmore, eds., Rural Restructuring: Global Processes and Their Responses. London: David Fulton. Collins, S.D. 1996. Let them eat ketchup!: the politics of poverty and inequality. New York: Monthly Review Press. Comer, K. 1999. Landscapes of the New West. London: UNC Press. Cook, A. K. 1995. ‘Increasing Poverty in Timber-Dependent Areas in Western Washington.’ Society and Natural Resources 8:97-109. Cook, A.K. and M. W. Jordon. 1997. Explaining Variation in Income between Hispanic and White Female-Headed Households in Washington. Hispanic Journal of Behavioral Sciences. 91:4:433-445. Creed, G.W. and B. Ching, eds. 1997. Knowing your place: rural identity and cultural hierarchy. New York: Routledge. Cronon, W., ed. 1996. Uncommon Ground. New York: W.W. Norton and Co. Cronon, W., G. Miles, and J. Gitlin. 1992. Under an open sky: rethinking America’s western past. New York: Norton. Dalaker, J. 1999. U.S. Census Bureau, Current Population Reports, Series P60-201. Poverty in the United States: 1998. Washington DC:U.S. Government Printing Office. Deavers, K. and R. Hoppe. 1992. ‘Overview of the Rural Poor in the 1980s’ in C. Duncan, ed., Rural Poverty in America New York: Auburn House. Duncan, C.M. 1999. Worlds Apart: why poverty persists in rural America. New Haven and London: Yale University Press. Egan, T. 1998. Lasso the Wind: Away to the New West. New York: Alfred A. Knopf. Eggers, M.L. and Massey, D.S. 1991. "The Structural Determinants of Urban Poverty: A Comparison of Whites, Blacks, and Hispanics," Social Science Research 20:3:217-255.

2

Fagen, G.H.. 1999. Cultural Politics and (post) Development Paradigm(s). In R. Munck and D. O'Hearn, eds. Critical Development Theory: contributions to a new paradigm. London and New York: Zed Books. Fitchen, J.M. 1981. Poverty in Rural America: A Case Study. Boulder: Westview Press. Fossum, H.L. 1993. Communities in the Lead: The Northwest Rural Development Sourcebook. Seattle: University of Washington, Northwest Policy Center. Friedland, W.H., L. Busch, F.H. Buttel and A.P. Rudy, eds. 1991. Towards a new political economy of agriculture. Boulder, CO: Westview Press. Gallaway, L. and Vedder, R. 1985. Suffer the Little Children: The True Casualties of the War on Poverty. Economic Committee, U.S. Congress War on Poverty: Victory of Defeat. Hearing before the Subcommittee on Monetary and Fiscal Policy, June 20. Gibson, J. 1996. The Social Construction of Whiteness in Shellcracker Haven, Florida. Human Organization 55(4): 379-389. Goodman, D. and M. Redclift. 1991. Refashioning nature: food, ecology and culture. London and New York: Routledge. Goodman, D. and Watts, M. 1997. Globalising Food: agrarian questions and global restructuring. New York and London: Routledge Press. Griffith, D. and E. Kissam. 1995. Working poor: Farmworkers in the United States. Philadelphia: Temple University Press. Hoggart, K. and H. Buller. 1987. Rural Development: A Geographic Perspective. London, New York and Sydney: Croom Helm. Hooks, b. 2000. Where we Stand. Class Matters. London; Routledge. Hubbard, M. and J. Elias. 1993. The Failure of Sustained-Yield Forestry and the Decline of the Flannel-Shirt Frontier. In T. Lyson and W. Falk eds. Forgotten places: uneven development in rural America. Lawrence: University of Kansas Press. Hugo, G. 1985. ‘Circulation in West Java, Indonesia.’ In R.M. Prothero and M. Chapman (eds.). 1985. Circulation in Third World Countries. London: Routledge and Kegan Paul. Jarosz, L. 1996. ‘Working in the global food system: a focus for international comparative analysis’ Progress in Human Geography 20:1:41-55. Jarosz, L. and V. Lawson. 2000. ‘Sophisticated People Versus Rednecks: Economic restructuring and class difference in America’s West’ Antipode forthcoming.

3

Jarosz, L., and J. Qazi. 2000. ‘The Geography of Washington's World Apple: confluences between the global and the local’ The Journal of Rural Studies 16:1-11. Jobes, P. 2000. Moving Nearer to Heaven. The illusions and disillusions of migrants to scenic rural places. London: Praeger. Jones, J.P. and J.E. Kodras. 1990. "Restructured Regions and Families: The Feminization of Poverty in the United States," Annals of the Association of American Geographers. 80:2:163-83. Knodel, J. 1993. ‘The Design and Analysis of Focus Group Studies: A Practical Approach’ in D. Morgan ed., Successful Focus Groups. London: Sage. Krannich, R.S. and Greider, T.R. 1990. ‘Rapid growth effects on rural community relations’. In Luloff and Swanson, Eds. American Rural Communities. Boulder, CO: Westview Press, pp. 61-73. Krueger, R. 1993. ‘Quality Control in Focus Group Research’ in D. Morgan ed., Successful Focus Groups. London: Sage. Kushnick, L. and Jennings, J., Eds. 1999. A New Introduction to Poverty: The Role of Race, Power, and Politics. New York and London: New York University Press. Lasley, P, L. Leistritz, L. Labao, and K. Meyer, eds., 1995. Beyond Amber Waves of Grain: An Examination of Social and Economic Restructuring in the Heartland. Boulder: Westview Press. Lawson, V. 1999. ‘Questions of Migration and Belonging: Understandings of Migration under Neoliberalism in Ecuador’ International Journal of Population Geography 5: 1-16. Lawson, V. 2000. ‘Arguments within Geographies of Movement: the theoretical potential of migrants’ stories’ Progress in Human Geography 24(2): 173-189. Le Heron, R. 1993. Globalized Agriculture. Oxford: Pergamon Press. Leistritz, L. and S. Murdock. 1981. The Socioeconomic Impact of Resource Development: Methods for Assessment. Boulder: Westview Press. Limerick, P. 1987. The legacy of conquest: the unbroken past of the American West. New York: Norton. Lyson, T. and W. Falk, eds. 1993. Forgotten places: uneven development in rural America. Lawrence: University of Kansas Press. Mackintosh, M. 1989. Gender, class and rural transition: agribusiness and the food crisis in Senegal. London and Atlantic Highlands, NJ: Zed Books.

4

Markusen, A. and E. Schoenberger. 1979. The Political Economy of Regional Development in the Western United States. Working Paper #315, Institute of Urban and Regional Development, UC Berkeley. Marriott, M. 1998. “As Big as Microsoft: Ribbing Gates,” New York Times. March 19. Marsden, T. 1996. Rural Geography Trend Report: The Social and Political Bases of Rural Restructuring. Progress in Human Geography 20:2:246-58. Marsden, T. 1999. Rural Futures: The Consumption Countryside and its Regulation. Sociologia Ruralis. 39:4:501-520. Marsden, T. and J. Murdoch. 1994. Reconstituting Rurality: class, power and community in the land development process. London: UCL Press. Marsden, T., J. Murdoch, P. Lowe, R. Munton, A. Flynn. 1993. Constructing the countryside. London: University College of London. Marsden, T., P. Lowe and S. Whatmore. 1992. Labour and Locality: uneven development and the rural labour process. London: David Fulton Publishers. Marsden, T., P. Lowe, and S. Whatmore. 1990. Rural Restructuring: Global Processes and Their Responses. London: David Fulton. McCarthy, J. 1998. "Environmentalism, Wise Use and the nature of accumulation in the rural West," in Remaking Reality: Nature at the millenium. Eds. B. Braun and N. Castree. London and New York: Routledge, pp. 126-149. McLaughlin, K.D., E.L. Gardner, E.L., and Lichter, D.T. 1999. ‘Economic restructuring and changing prevalence of female-headed families in America’ Rural Sociology. 64:3:394416. Mitchell, D. 1998. ‘Writing the Western: New Western History’s Encounter with Landscape’ Ecumene 5:1:7-29. Moore, D. 1993. ‘Contesting Terrain in Zimbabwe's Eastern Highlands: Poltical Ecology, Ethnography and Peasant Resource Struggles’ Economic Geography. 69:4:380-401. Murray, C. 1984. Losing Ground: American Social Policy 1950-1980. New York: Basic Books. Nelson, P.B. 1997. ‘Migration, Sources of Income, and Community Change in the Nonmetropolitan Northwest’ The Professional Geographer 49:418-430.

5

Nelson, P.B. 1999. Hegemony and the rural: economic and cultural perspectives on restructuring in the rural West. Doctoral Dissertation. Department of Geography, University of Washington, Seattle. Nelson, P.B. and W. Beyers. 1998. ‘Using economic base models to explain new trends in rural income’ Growth and Change 29: 321-24. Newby, H. 1985. ‘Locality and rurality: the restructuring of rural social relations’ Regional Studies 20:209-16. Nord, M. 1994. ‘Natural resources and persistent rural poverty: in search of the nexus’ Society and Natural Resources 7:3:205-220. NPR/Kaiser/Kennedy School Poll. 2001. Poverty in America. NPR Online. http://www.npr.org/progras/specials/poll/poverty/ Peet, R. 1987. "Industrial Restructuring and the Crisis of International Capitalism," in International Capitalism and Industrial Restructuring: A critical Analysis. Ed. R. Peet. Boston: Allen and Unwin, 9-32. Oberhauser, A.M. 1995. ‘Towards A gendered Regional Geography: women and work in rural Appalachia’ Growth and Change. 26:2:217-244. Page, B. 1996. ‘Across the great divide: agricultural and industrial geography’ Economic Geography. 72:4: 376-397 . Page, B. and Walker, R. 1991. ‘From settlement to fordism: the agro-industrial revolution in the American Midwest’ Economic Geography 67:281-315. Power, T.M. 1996. Lost Landscapes and Failed Economies. Washington, DC: Island Press. Ramsay, M. 1996. Community, Culture and Economic Development. Albany: State University Press of New York. Rasker, R. 1995. Home on the Range: Economic Realities in the Columbia River Basin. The Wilderness Society, Bozeman, Montana. Rasker, R. 1998. The New Challenge: People, Commerce, and the Environment in the Yellowstone to Yukon Region. Bozeman, MT: The Wilderness Society. Rasker, R. and Glick D. 1994. ‘Footloose entrepreneurs: pioneers of the new West?’ Illahee 10:34-43.

6

Rasker, R., and B. Alexander. 1997. The New Challenge. People, Commerce and the Environment in the Yellowstone to Yukon Region. Washington, DC: The Wilderness Society. Riebsame, W., ed. 1997. Atlas of the New West. New York: W.W. Norton and Company. Rudzitis, G. 1993. ‘Nonmetropolitan geography: migration, sense of place, and the American West’ Urban Geography 14:574-585. Rudzitis, G. 1997. Towards a Theory of Place in the “new American West”. Paper presented at the 1997 Association of American Geographers meeting. Ryan, W. 1974. "Blaming the Victim: Ideology Serves the Establishment, in The Poverty Establishment. Ed. P. Roby. Englewood Cliffs, NJ: Prentice-Hall, 171-179. Sachs, C. 1996. Gendered Fields: rural women, agriculture and environment. Boulder: Westview Press. Sack, K. 1996. “2,000 Ways You Might Be a Redneck,” New York Times, February 7. Sandell, J. 1997. “Telling Stories of ‘Queer White Trash’: Race, Class and Sexuality in the Work of Dorothy Allison,” in Wray, M. and A. Newitz, eds., White Trash. Race and Class in America. New York: Routledge. Schwarz, J.E. and Volgy, T.J. 1992. The Forgotten Americans. New York: W.W. Norton. Scott, S.L. 1995. Two Sides to Everything: The Cultural Construction of Class Consciousness in Harlan County, Kentucky. Albany: State University of New York Press. Shanin, T. 1990. Defining Peasants. Oxford: Blackwell. Shaw, W. 1996. The Geography of United States Poverty. London: Garland Publishing. Shumway, A., and J. Davis. 1996. ‘Nonmetropolitan Population Change in the Mountain West: 1970-1995’ Rural Sociology 61:3:513-529. Shuyler, K. 2000. Unpublished Research Notes for Lawson and Jarosz, pilot research study, University of Washington, Seattle. Silvey, R. and V. Lawson. 1999. ‘Placing the Migrant’ Annals of the Association of American Geographers 89(1): 121-132. Skeldon, R. 1985. ‘Circulation: a transition in Peru.’ In R.M. Prothero and M. Chapman (eds.). 1985. Circulation in Third World Countries. London: Routledge and Kegan Paul.

7

Smith, M.D. and R. Krannich, R.S. 2000. "'Culture Clash' Revisited: Newcomer and Longer-Term Residents' Attitudes Toward Land Use, Development, and Environmental Issues in Rural Communities in the Rocky Mountain West," Rural Sociology 65:3:396-421. Stusser, M. 1997. “Dear Seattle….” Seattle Weekly, November 26. Taylor, J.E., P.L. Martin and M. Fix. 1997. Poverty amid prosperity: immigration and the changing face of rural California. Washington, DC: Urban Institute Press. Urry, J. 1984. ‘Capitalist restructuring, recomposition and the regions’ In Bradley and Lowe eds. Locality and Rurality. London: Geobooks. Washington Department of Community, Trade and Economic Development. 1995. Defining Rural Character and Planning For Rural Lands. Washington State Growth Management Program. Olympia: State of Washington Department of Community, Trade and Economic Development Growth Management Services. Watkins, E. 1993. Throwaways. Work, Culture and Consumer Education. Stanford: Stanford University Press. Watts, M. 1983. Silent Violence. Berkeley: University of California Press. White, R. 1996. ‘Are you an environmentalist or do you work for a living?’ in Cronon, ed., Uncommon Ground. New York: W.W. Norton and Co., pp. 171-185. Wilkinson, K.P. 1991. The Community in Rural America. New York, London, Westport, CT: Greenwood Press. Wilson, J.W. 1987. The Truly Disadvantaged: The Inner City, the Underclass, and Public Policy. Chicago: University of Chicago Press. Wyckoff, W. and Dilsaver, L.M., Eds. 1995. The Mountainous West: Explorations in Historical Geography. Lincoln: University of Nebraska Press.

8

APPENDIX

vi. Patel, Reena. Doctoral Dissertation Research: Working the Night Shift: Women’s Employment in the Transnational Call Center Industry.

Doctoral Dissertation Research: Working the Night Shift: Women’s Employment in the Transnational Call Center Industry Project Description In a recent review, Tuan (2004) contends that “cultural geography remains almost wholly daylight geography” (Tuan 2004,730) and that more attention needs to be given to the “after hours.” This contention makes particular sense as the “second shift,” namely a night shift labor force, emerges in the global economy. The hyper-growth of the transnational call center industry in India is the quintessential example of this nightscape. The nightscape is primarily an exclusive, male domain that often represents danger for women. With the infusion of educated, middle and upper-class women into the urban nightscape, via transnational call center employment, the nightscape also brings forth class and caste connotations that mark women’s bodies as sites of transgression (Cresswell 1996; Domosh 1998b) – aka “working the night shift.” Such processes evoke new questions about the spatial construction of social identities. I propose to examine how the demand for night shift workers in India is re-configuring women’s physical, temporal, social, and economic mobility. Another key concern is how households respond to changes in women’s mobility. In particular, I examine the variety of spaces occupied by women as a result of call center employment to articulate how notions of gender are inscribed in the nightscape. This research draws from globalization discourses and feminist geography, and utilizes qualitative research methods, such as participant observation and in-depth interviews, to investigate the ways in which space, place, and mobility of women working in transnational call centers shapes their identity at a variety of scales. In addition, the social and cultural norms that mediate women’s mobility are linked to broader processes, such as the rise of middle class and the role of economic globalization in the transformation of Indian society. I argue that notions of “middle-class morality” mark women’s bodies as the site of family honor, purity, and chastity (Bondi and Domosh 1998; Bondi and Rose 2003; Nagar 1998; Nast 1998). This in turn has significant consequences when thinking about the urban nightscape. Mobility and access to public space in the urban domain is particularly important in terms of call center employment because it requires physical mobility in terms of transportation to and from work and temporal mobility in the form of working at night to coincide with the office hours of U.S.based customers. For a woman to be out and about in the middle of the night is generally considered improper and unsafe. However, women are participating in this industry, and corporate strategies, such as the use of private shuttle vans to transport women to and from work in the middle of the night, reflect the ways in which both the industry and its female employees negotiate a presence in urban, public spaces. In light of these dynamics, my project will address the following research questions: 1) How does the global demand for night shift workers re-codify women’s physical and temporal mobility?; 2) What spatial and temporal barriers, as well as opportunities, do women face both in the household and urban public space?; and 3) How does call center employment translate into social and economic mobility or immobility? The emergence of an IT revolution (Castells 2000; Greenspan 2004), in combination with disparities such as gender inequity and the digital divide, expands the range of feminist geography to areas such as global technology and development. Currently, extensive research on women’s participation, or lack thereof, in the IT sector is available (Fountain 2000; Hafkin and Taggart 2001; Kelkar, Shrestha, and Nagarjan 2002; Patel and Parmentier 2005; Poster 1988; Webster 1996). However, despite the globalizing aspects of transnational call center operations – as demonstrated by a Delta Airlines customer in Chicago talking to a Delta travel agent in India about flight reservations to London – there is little attention given to its impact on geographical concerns in relation to gender issues.

Background Since the late 1990’s, advances in telecommunications technology combined with the globalization of IT services have directly contributed to the growth of transnational call center operations. Fortune 500 companies, from IBM to American Express, have become reliant on transnational call centers and over the past five years various data processing functions such as inputting medical transcripts and credit card applications/billing have transferred to India (Economist 2004; Patel 2002; Stitt 2002).Companies are moving call center operations to India because it provides a cheap, English speaking labor force. For instance, the starting salary for a full-time call center employee at TYJ Corp1., a top-ten call center in India, is 10,000 rupees per month (approximately 225USD). The global processes that fuel the emergence of this industry are embedded in national policies of both India and the United States. The restructuring of U.S. immigration policy in terms of reducing the number of H1-B visas, along with the economic downturn of the United States IT sector, serve as an impetus for companies to offshore what were previously U.S.-based positions (Rudrappa 2005). Essentially, because the Indian worker cannot migrate to the U.S., the work migrates to India. The tightening U.S. economy has led companies to offshore not only high wage engineering positions, but also low-wage call center positions. In this instance, the protectionist policy of limiting immigration in order to bring economic security to the American worker has created the opposite effect, as more and more jobs are transferred overseas. However, this trend should not be exaggerated because Pandit (2005) finds that the actual number of U.S. jobs that have moved overseas thus far is not significant. Simultaneously, the emergence of transnational call centers in India is contingent on a national policy that welcomes the presence of multi-national corporations (MNC’s). Historically, this was not the case. After independence, India went from being dubbed the “British Raj” to the “Permit Raj” (Greenspan 2004). India was considered an impenetrable market because of its unending bureaucracy, notorious corruption, and protectionist policies that sought to shield India from the outside economy. This dramatically shifted in 1991 when the government, under Minister Rao, removed import licensing requirements and sought to undo more than four decades worth of bureaucracy under the Permit Raj. As of 2002, a decade after the pivotal 1991 reforms, the National Association of Software and Service Companies (NASSCOM) estimates that 336 call centers have emerged throughout India and as of 2005, approximately 348,000 women and men are working in this industry (NASSCOM 2002; NASSCOM 2005). Currently, call center operations locate primarily in Gurgaon, Mumbai, Chennai, Bangalore, and Hyderabad. These major cities actively compete against one another in order to attract companies (Patel 2002). In addition, call centers have recently begun to emerge in tier two cities such as Ahmedabad and Jaipur. Call centers expand to tier two cities because human resources in major cities are exhausted and the wages in tier two cities are lower. The emergence of call centers in tier two cities represents a new and important contrast to the current migration of skilled workers to major cities such as Mumbai. In this instance, instead of workers migrating to Mumbai for call center employment, the work is now migrating to workers in tier two cities throughout India. In regards to the gendered aspects of this industry, Kelkar et. al (2002) explain that call centers in India prefer hiring young, educated women because they are viewed to be more loyal and compliant in comparison to their male counterparts. Yet the majority of executive positions are held by men, suggesting the existence of occupational segregation. Singh and Pandey (2005) argue that women often plateau at mid-level positions while men tend to progress further. This disparity is linked to men’s lack of participation, and responsibility, in household labor and childcare. 1

Names of companies and employees cited in this proposal are changed to protect identity

One gets a glimpse of such segregation in an India Today article (Chengappa and Goyal 2002). The cover story, labeled “Housekeepers to the World,” focused on call centers in India. In terms of visual presentation, the image on the front cover, along with photo-ops of customer service workers were primarily young women. Yet photo-ops involving high level positions, such as Chairman or President, were represented by older men. Within this article, it was the men who were presented as leaders of the industry and experts in terms of discussing future growth and challenges surrounding its development. In contrast, the women interviewed were mostly entry-level workers. The one exception to this was a female vice-president. However, she worked for a company that trains women on how to be effective customer service representatives, and so she was not in a direct position of power in terms of owning a call center or serving as an agent for influencing policy surrounding the development of this industry. The reader gets the sense that women are seen and heard by distant customers, but are rarely participants in corporate decision-making. Mirchandani’s (2004) and Poster’s (2004) research on call centers suggests a possible shift in the gendered aspects of call center employment. In the United States women are the majority of call center employees and this industry is considered to be a “pink collar” field (Bonds n.d.). In contrast, Poster’s and Mirchandani’s study of call centers in New Delhi found that 50% - 70% of the employees are men. This coincides with a call center I visited in Mumbai in which 60% of the employees are also men. Such initial findings contradict the media representation of call center employment in India as being primarily held by women (Chengappa and Goyal 2002). Unlike maquiladora labor, such as garment production, which was predominantly female in the United States and remained the same upon transfer to countries such as Mexico, Bangladesh, and the Philippines, it is unclear if call center employment will follow a similar path. The idea of a modern-day labor force that goes from female to male is rare in comparison to the integration of women into historically male-dominated positions such as clerical work (Boyer 1998; Elias and Carney 2005) and pharmacy (Ehrenreich and Hochschild 2003; Tanner and Cockerill 1996). Focusing on women’s mobility in the urban nightscape provides an understanding of whether spatial and structural barriers impact women’s employment opportunities in the call center industry. In summary, the emergence of transnational call centers is based upon a complex set of local and global dynamics, as demonstrated by its dependency on national policy as well as its intersection with gendered notions of a woman’s place in the urban nightscape. Although considerable attention is given to this industry by the mainstream media, the focus is primarily on corporations moving backoffice work overseas (Delaney 2003; Economist 2004) and the subsequent loss of U.S. jobs (Nichols 2004). By investigating how community norms of mobility and spatial access intersect with the night shift requirement of transnational call centers; my research contributes to new understandings of women in space. Disciplinary Significance Globalization Discourse Globalization is the subject of criticism and debate both within geography and other disciplines. Discussion abounds as to whether or not globalization is leading to cultural homogenization (Appadurai 1996; Cox 1997; Jackson 2003; 2004) or reducing the power of the nation-state (Agnew 2005; Castells 2000; Mitchell, Marston, and Katz 2003; Sassen 1998; 2000; 2005; Scott 1998). Geographers focus primarily on the political economy of globalization (Rankin 2003a). Traditionally, the focus is on how macro/global processes conflict with and/or impact local communities (Cox 1997; Harvey 2000; Herod 1997; 2001; Jackson 2004). Although the literature is vast, this project draws from globalization discourse as it relates to the spatiotemporal landscape and gender.

The Spatiotemporal Landscape Harvey (1989) coined the term “time-space compression” to explain the increased mobility and internationalization of capital in the early 1970's. In recent years, there has been an emergence of literature focusing on the impact of globalization on time and space (Adam 2002; Katz 2004; Sassen 1998; 2000; 2004; Scott 1998; Sheppard 2002; 2004). Sassen (2000) argues that globalization leads to the creation of new spatialities and temporalities in which “economic globalization itself can already be seen to contain dynamics of both mobility and fixity” (Sassen 2000,217). For instance, the mobility afforded to MNC’s seeking to offshore their customer service labor force is contingent on fixed structures at the national level, such as material and technological infrastructure, as well as national policies that promote, rather than hinder, the expansion of the global economy. Such overlap and interaction between the global and national, in combination with advances in telecommunications technology, reshapes the spatiotemporal landscape of globalization. A local U.S. workforce gone global now operates on a 24-hour timeframe that shifts the work space and work time of customer service employees worldwide. Twenty five years ago who would have expected that on any given afternoon an American residing in Texas would dial a 1-800 number that is instantly re-routed to the suburb of a major Indian city, and at 3a.m. India time, an employee with a “neutral” accent would answer “Good Afternoon. American Express, this is Julie speaking”? Adam would label this dynamic a “colonization of time” whereby the western clock is commoditized, set as the standard, and exported throughout the world (Adam 2002,21). This transformation of time into a global resource is reorganizing an employee’s identity, neutralizing accents, and temporally adjusting the normative nine-to-five work schedule. Indeed, the night shift labor force represents a new level of social and spatial interaction between industrialized and developing nations. Gender and Globalization Geographers investigate the globalization process from the perspective of how the production of local and global space is not only relational, but also constructed upon social, political, and economic interactions that remain inherently gendered (Elias and Carney 2005; McDowell et al. 2005; Nagar et al. 2002; Pratt 2004). From the household to the global economy, Haraway’s (1988) concept of situated knowledge offers a new way of thinking about globalization. Situated knowledge rejects notions of detached neutrality and the quest for universal findings (Cope 2002). In fact, Haraway (1991) argues that the idea of detached neutrality is an attempt by modern science to perform a “god-trick” by viewing itself as a disembodied Other that can produce objective findings. This veil of neutrality is problematic because it conceals the complexities of research. Haraway also uses the concept of vision to explain that subjects which are deemed feminine are not given sight (Holloway 2004). They are viewed primarily through a lens of being observed, described, or conquered. Essentially, they become outsiders explained away by a so-called objective, detached, scientific gaze. Gibson-Graham (1996) offers an interesting perspective on re-envisioning globalization through the lens of gender. By critiquing the ‘penetration of capital’ script that frames workers as rape victims who are powerless to the dominant forces of MNC’s, they argue that we need to explore ways of “making globalization lose its erection” (Gibson-Graham 1996,146). Such analysis reflects Haraway’s (1988) quest for partial perspective because it re-situates globalization in a context-specific framework of feminist theory. Oza (2001) looks at the 1996 Miss World Beauty Pageant in Bangalore from the perspective of how various scales, such as the gendered body, national politics, and globalization, interact with and inform each other. She contends that the nation is represented as enclosed and feminine while globalization is constructed as masculine and free floating. By linking gender to the nation, different groups in India formulated a critique for or against globalization. Opponents, for instance, constructed the nation as a symbol of desexualized motherhood that must be protected from the polluting forces

of globalization (i.e. a world pageant). In this case, resistance to globalization is represents a reinscription of control and suppression of women’s sexuality. Proponents in Oza’s study argued that the pageant demonstrated to the world that India is an advanced, liberal nation. Yet even within this framework, the pageant used raising funds for children’s causes as a means to detract opponents. By linking contestants to motherhood, women’s sexuality is performed within acceptable boundaries. Oza states “women’s bodies and sexualities became the material and discursive sites where nation was performed, values were contested, and border and boundaries were policed and controlled” (Oza 2001,1068-1069). Oza’s research provides insight into the ways in which women, traversing the urban nightscape as call center employees, may become the bodily site upon which arguments for or against globalization are framed. Feminist Geography Social change, or lack thereof, is reflected in the mobility and spatial access afforded to women (Boyer 2005; Hapke and Ayyankeril 2004; Kantor 2002; Law 2002; Mandel 2004; Rankin 2003b; Silvey 2005; Yeoh and Huang 1998). The built environment not only reflects status and power, but can also serve to reinforce existing gender divisions (Law 1999; Massey 1994; McDowell 1999). In turn, feminist geographers are actively engaged in deconstructing essentialist notions of “woman” by articulating how gender is socially constructed in space and place at multiple scales By examining space through a lens of “difference”, feminist geography also provides an understanding of how representation is a “…mediator and medium through with identities and spaces are (re)produced” (Oberhauser et al. 2003,746). In the House Historically speaking, Domosh (1998a) explains that geographers, in general, “did not move past the front stoop” (1998a,275). Yet, the construction of gender relations within the household provides a powerful understanding of how the spatial and temporal landscape of both public and private space remains gendered (Marston 2000). Instead of viewing the household as what Brenner terms “relatively stable background structures” (Marston and Smith 2001,618), feminist geographers argue that the household is a complex site where material practices are made and remade. Chant (1998), in particular, defines the household as a “geographically and historically dynamic social institution in which gender is embedded and negotiated” (5). Night shift employment provides a new context for gender relations in the household because it represents a shift in the temporal landscape of women’s economic opportunities and thereby intersects with the “second shift,” namely management of the household. Rankin’s (2003b) research on gender and socio-spatial change in Nepal finds that access to credit and income does not necessarily contribute to gender equity in the household. And night shift employment may serve to merely shift domestic responsibilities to other women within the household or, as illustrated by the research of Pratt (2004), Yeoh and Huang (1998), and Tyner (1999), lead to an increased demand for maids and servants, often from foreign countries. Women on the Move Regardless of the actual travel distance, changes in mobility represent a renegotiation of prescribed gender roles (Domosh and Seager 2001; McDowell 1999). As MNC’s redistribute labor on a global scale, feminist geographers are interested in how the global hyper-mobility of MNC’s reshapes the mobility and spatial access of its workers, particularly in relation to categories such as gender and class. Similar to Boserup’s (1970) argument that economic development can further marginalize women, Massey’s (1994) seminal research on gender, space, and place finds that “the mobility and control of some groups can actively weaken other people” (150). Although Hägerstand’s (1975) work is not within a framework of feminist geography he also clearly explains that “one

individuals use of his freedom influences what other individuals are able to do with theirs” (Hägerstand 1975,5). Massey’s assertion is reflected in Cravey’s (1998; 2005) comparative research in Nogales, a Mexican maquila border town, and Cuidad Madero, an interior town that emerged from a previous state-led industrialization model of import substitution. Her findings indicate that as paid employment becomes more connected to the global economy, the social geographies of women’s livelihoods become increasingly privatized and individualistic (Cravey 2005,119). Access to social services (daycare and medical insurance) that were considered a social right in Cuidad Madero became a social privilege in Nogales under the maquila framework. Ong’s (1987) research on transnational electronic factories in Malaysia uncovers how factory work re-shapes women’s physical, social and economic mobility. A key finding in her ethnography is the assertion that transnational corporations are operate within a hierarchy of domination and subordination that not only controls a worker’s body, but also the ways in which young female employees view themselves. Both studies provide an understanding of how the hyper-mobility of MNC’s reshapes the livelihood of its female-dominated workforce. Body Politics From being marked as a site of reproduction to a source of provocation, feminists are on the forefront of critiquing the social and biological construction of women’s bodies (Bordo 1993; Faludi 1991; McDowell 1999; Wilson 1992). Haraway (1988) argues that conceptualizing sex under the guise of biological determinism threatens the space of emerging work in critical social theory. Using figures such as the cyborg and trickster, Haraway (1991) refutes the dualistic notion of science versus nature/culture and urges us to re-envision how categories of unitary identity, such as global ‘woman’, “are made and remade within a shifting network of determinants and desires” (Jacobs and Nash 2003,266). By the mid-1990s, feminist geographers argued that the body, as scale of analysis, provides a powerful understanding of how space and place are constructed on categories such as gender, race, and class (Longhurst 1995; McDowell 1999; Moss 2002; Silvey 2005). Wright’s (1997a; 1997b) work is compelling in this area. Within a framework of gender, social geography, and global power relations, she examines how Mexican women’s bodies are marked as docile, submissive, and tradition-bound in the maquiladora. By linking this perceived embodiment to the regulation and control of women’s bodies, she uncovers how gender inequality is spatialized in the maquiladora. The embodiment discourse is linked to Butler’s (1993) argument that gender is an act of performativity. By ‘performing’ gender, Butler suggests that notions of womanhood are inscribed on the body and marked as a biological site of difference. Drawing from Butler’s work, feminist geographers provide new insight into the relationship between women’s bodies and space at multiple scales. Secor (2003), for instance, examines how “regimes of veiling” (2003,5) in Istanbul allow women to detract the male gaze, but at the same time reinforce the notion that women’s bodies are a source of provocation that must be controlled and concealed in urban public space. Conceptual Approach By examining globalization in conjunction with feminist geography, what emerges is a critical understanding of the interplay between space, gender, and identity. My investigation of transnational call center employment is based on a framework of social and economic embodiment, spatial access, and a concept I’ve termed temporal entrapment. Although the physical location of a call center remains fixed, the meaning embedded in the urban space it occupies shifts depending on the time of day. In addition, the meaning that is embodied in dayshift versus night shift employment intersects with the social construction of gender and class. In contrast to their male counterparts, the presence of

women in India’s urban nightscape is often linked to prostitution and questionable moral values. Ashini, a 23 year old employee, explains that her father’s response to call center employment was: “call center job equals call girl job.” Ashini’s co-worker Kavita goes on to argue that family concern for young women working the night shift is less about physical safety and more about how a woman’s presence in the urban nightscape will negatively impact a family’s reputation. “What will people think?” is a common response women received from family members expressing hesitancy about night shift employment. Such remarks can be linked to notions of middle-class morality, which mark women’s bodies as the site of family honor. In this instance, changes in women’s temporal mobility are viewed as potentially having a negative impact on the social status of both the worker and her family. In regards to spatial issues, shuttle transport at TYJ Corp. consists of vehicles carrying 6-8 employees and can take up to two hours each way, depending on where the employee resides. If all the employees are female, then, in addition to the driver, a security guard is on board. If one of the employees is a male, then a security guard is not required. However, the male employee is always the last one dropped off. This measure is necessary due to reports of the Mumbai police stopping the vans and accusing the female passengers of prostitution. Employees carry identity cards as proof of employment, but it is not considered enough by police. In order for a woman to justify her existence in the urban nightscape she requires the presence of a male counterpart. Although traversing the nightscape may represent new levels of spatial access for women, it is based on a continuum of protection and surveillance of women’s bodies. Temporal entrapment is conceived of from two dynamics. First, night shift labor may serve as a time-trap that marginalizes women. By working the night shift, and inevitably sleeping thru the day, it is possible that women become further excluded from social and economic opportunities within the larger community. Shubhika and Sonia complain that since taking a call center job their social life has diminished because they have lost touch with close friends and hardly have time for their families. Second, working the night shift for a call center, albeit a position that is viewed as secondary and having limited prospects for upward mobility (Kelkar, Shrestha, and Nagarjan 2002), may create new opportunities and spaces for women upon which they re-invent their identity in the formal economy and within their households. Instead of constraining her social life, Drasti explains that call center employment allows her to befriend people from various walks of life that she otherwise would not have access to. And perhaps shifting the dynamic of arranged marriage, Nazia exclaims “the call center becomes our marriage pool!” during an interview about how call center employment has impacted her family and social life. Linking women’s employment to embodiment, spatial access, and temporal entrapment provides a context for, and an understanding of how, night shift labor practices are reshaping the lives of transnational call center employees. Furthermore, the conceptual framework I propose to use allows me to examine how spatial practices, such as the journey to work, not only operate as a singular quantified event (i.e. two hours each way on a shuttle bus), but also how night shift employment reshapes the social construction of gender in a variety of spaces. Research Site: Description and Justification Dissertation fieldwork will take place in Mumbai, India. Mumbai is the financial center of the Indian economy and serves as the largest income tax base for the country. Call centers are primarily located in two outlying areas of Mumbai, Powai and Navi Mumbai, because of the exorbitant real estate prices in the central business district. Powai is the epicenter of call center operations and is home to Wipro-Spectramind, India's largest call center. Located over 30 miles away from Mumbai proper, Navi Mumbai is 212.3 square miles and is India’s largest urban planning project to date (Shaw 2004). Development of Navi Mumbai began in 1971 with the goal of creating 14 towns that will hold a population of approximately two million people. It was conceived of as a satellite town to

slow down the expansion of downtown Mumbai and to serve as a counter-magnet to draw new incoming migrants and re-settle some of its current population (Shaw 2004, 4). Arguably, as call center operations emerge throughout India, the question becomes why focus on Mumbai versus Bangalore, Chennai, or Hyderabad. When asked “Why setup in Mumbai versus Bangalore?” Sharon, a call center executive, contends that Bangalore is the IT hub, but not necessarily the call center hub. The presence of an educated, English-speaking population and the space available to build call centers in the outlying areas of Mumbai are the key magnets drawing companies to this area. Mumbai is also viewed as more cosmopolitan and professional, and is ahead of Delhi in terms of fiber-optic connectivity and its electricity infrastructure (Patel 2002). At the same time, during pre-dissertation research I discovered that some families are hesitant about women working for a call center in Mumbai because, unlike Bangalore, Mumbai is viewed as a city of illrepute, danger, and sin. In this context, focusing on Mumbai provides a complementary understanding of how the local conception of a cityscape intersects with the global demand for a night shift labor force. Research Methods and Timeline An inductive, qualitative research methodology based on participant observation, in-depth interviews, and surveys are used to gather data for this project. The methodology I use draws from the grounded theory approach by Strauss and Corbin (1990; 1997). Unlike hypothesis testing which compares findings to predetermined set of outcomes, grounded theory is an inductive approach in which new findings emerge from an ongoing interplay between collecting and analyzing data. (Cresswell 1998; Russell 2002). This methodology is well suited for research that: 1) examines the meaning and context of a phenomenon; 2) focuses on the processes that lead to an event or action; 3) seeks to identify unanticipated influences; and 4) seeks to uncover causal relationships (Maxwell 1998,75). My research will be carried out in the three stages discussed below. Stage One (January 2006 to April 2006) - Completed During the first four months of fieldwork, I networked with existing contacts to pre-test interview questions and to obtain entry into call centers. Entry into call centers is particularly difficult because the companies have strict contractual agreements with their U.S. and U.K. clients that forbid access to outsiders. However, I obtained entry into KMA Corporation and conducted semi-structured interviews with 20 employees. In addition, I conducted participant observation of KMA’s anniversary celebration with approximately 350 employees. Thereafter, I attended the Partners for Urban Knowledge Action and Research (PUKAR) Gender and Space Conference. Founded by Arjun Appadurai in 2001, PUKAR created the Gender and Space Project to investigate how the cityscape of Mumbai is gendered in ways that regulate women’s mobility in public space (PUKAR n.d.). I also interviewed Anand Dalmia, an outsourcing expert and contributor to Avendus, a Business Process Outsourcing newsletter. Both PUKAR and Dalmia provided an understanding of the emergence of transnational call centers as well as women’s access to the spaces they occupy. Stage Two (May 2006 to October 2006) – In Progress During this six month timeframe the focus is primarily on conducting structured and semistructured interviews with employees and participant observation in call centers. Participant observation allows me to evaluate the physical and temporal mobility of women as well as obtain research participants. In-depth interviews are conducted outside the call center so as to alleviate concerns or apprehension that findings from my interviews will be divulged to the employing organization.

I will interview approximately 50 female call center employees and will seek a range of employees from married versus single, living at home versus living in a hostel or company dorm, and newly hired employees versus long term employees (one year or longer). A snowball sampling technique is used to identify individuals to interview. This technique is based upon asking interviewees to identify other potential research participants (Cresswell 1998; Russell 2002). Although this method is not a random sample, it allows me to gain in-depth knowledge of the experiences of call center employees. Interviews will also be conducted with approximately 20-25 family members of call center employees and 10-15 managers, trainers, former employees, and government officials who are involved in IT policy. This sample will provide a different understanding of how various actors inform and impact the experiences of female night shift workers. In total, approximately 85-90 interviews will serve as the foundation for this study. As of August 2006, 43 interviews consisting of employees, managers, and family members, were conducted, 180 surveys were collected, and participant observation was conducted in two call centers and various locales, such as malls, cafes, and bars, that employee’s frequent. In addition, newspaper and television accounts relating to women’s participation in the call center industry are being collected in order to assess how night shift employment is portrayed in the media and community reaction to this relatively new industry. The interviews I seek to conduct, in combination with participant observation and media accounts of the industry, will serve as the foundation for uncovering the spatial livelihoods of call center employees. Stage Three (November 2006 to March 2007) In addition to wrapping up interviews in Mumbai, the NSF DDRI will allow me to expand the scope of my study to include interviews with employees, managers, and family members in Chennai and tier two cities. Chennai is a rapidly expanding call center hub. Although Mumbai and Chennai are both major cities in India, Mumbai is considered a “fast,” progressive city, whereas Chennai is far more conservative, particularly in terms of gender relations. Investigating night shift employment in Chennai, as a contrast to Mumbai, will provide an understanding of the role space and place has in reshaping women’s mobility. The tier two cities for this project are Jaipur and Ahmadabad. They were selected because Mumbai call centers are currently recruiting employees from both these cities. By investigating night shift employment in Mumbai and tier two cities, my project will uncover the similarities and differences in the social, spatial, and economic mobility experienced by women who migrate to Mumbai for call center employment versus those who remain in tier two cities to pursue call center employment. Contacts have been secured and permission is granted to visit call centers in both Chennai and Ahmadabad. Currently, I am currently working to secure permissions in Jaipur. Research Protocol and Analysis The interviews I conduct with call center employees serve as the foundation for this project. Some interviews will be quite short and focus on gathering factual data, others may focus on particular topics. For all interviews, the goal is to record some of the basic information under the following categories: 1) physical and temporal mobility; 2) household and community response to women working in call centers; 3) social and economic spaces. Following are some of the questions I draw from in my interview instrument: Physical and temporal mobility Describe your journey to work and home everyday. What challenges do you face? How does changing your work shift impact your daily routine? How often did you go out at night prior to working for a call center? What have been your experiences in terms of going out at night?

Have you experienced harassment when you’ve been out? Day or night? Explain. Describe the protective measures taken by yourself or your company to ensure your safety Does call center employment give you access to areas of the city that before were unavailable to you? Describe. Define places that you can now go to that were limited before? Or the degree to which you travel un-escorted regularly. Describe the positive experiences you’ve had since you began working for a call center Household and community response to women working in call centers What did your family think about you working for a call center? Has their opinion changed? How has working the night shift impacted your relationship with your family and friends? From your experience, what perceptions do people have about women working in call centers? Where do such perceptions come from? How are women who work at night perceived in your neighborhood? How are call center employees portrayed by the media? What is your opinion of their portrayal? Social and Economic Mobility How does working for a call center impact your spending habits? How has this job changed your day-to-day expenses? Where do you hang out when you are not working? What opportunities does this job provide you with that you didn’t have before? In general, if the respondent is willing to talk about a subject, they will be encouraged to do so for as long as they like (through verbal and non-verbal cues) and I will provide follow-up questions when necessary. In all cases, I inform participants that their identity will remain confidential. IRB has waived the informed written consent requirement and pseudonyms are used on paper and digital records to protect the identity of the participants. Data is stored on a password protected laptop. In the event that the laptop breaks or is stolen, data is also maintained on Digital Data Storage, an online backup server. Interviews are taped when permission is granted by the interviewees. Subsequently, tapes will be transcribed and a narrative analysis based on coding and categorizing interviewee responses onto a matrix will be performed (Rubin and Rubin 1995; Strauss and Corbin 1990). This narrative analysis will be compared to findings from participant observation as well as newspaper and television accounts relating to women’s participation in the call center industry. The NUD*IST, a qualitative data analysis software program, will be utilized for automated coding, text search and retrieval, and pattern discernment. Conclusion Call center employment provides women with relatively high paying jobs that were previously unavailable. Yet it is unclear whether working the night shift work will impede or enhance women’s mobility and spatial access. On the surface, call centers represent a marked shift in women’s access to employment in the paid labor force. At the same time, initial findings suggest that the emergence of this industry is not shifting patriarchal relations of power in a significant way due to social and spatial constraints on women’s mobility in the urban nightscape. The result of my research will be a dissertation that I plan to defend in May, 2008 and will eventually seek to publish into a book. Findings will be presented at the December, 2006 Women in Technology conference held in India and the April, 2007 Association of American Geographers conference. I will also use this research to pursue a career in the area of science and technology policy. Over the long-term my goal is to advise both governmental and corporate entities on genderbased linkages in technology, a critical issue as manufacturing expands in traditional societies.

Reference List Adam, Barbara. 2002. The gendered time politics of globalization: Of shadowlands and elusive justice. Feminist Review 3, no. 70: 3-30. Agnew, John. 2005. Hegemony: the new shape of global power. Philadelphia: Temple University Press. Appadurai, Arjun. 1996. Modernity at large: cultural dimensions of globalization. Minnesota: University of Minnesota Press. Bondi, Liz and Mona Domosh. 1998. On the contours of public space: A tale of three women. Antipode 30, no. 3: 270. Bondi, Liz and Damaris Rose. 2003. Constructing gender, constructing the urban: a review of AngloAmerican feminist urban geography. Gender, Place & Culture: A Journal of Feminist Geography 10, no. 3: 229-46. Bonds, Anne. n.d. Under review. Calling on Femininity? Gender, Call Centers, and Restructuring in the Rural American West. ACME: An International E-Journal for Critical Geographies . Bordo, Susan. 1993. Unbearable weight: Feminism, Western culture, and the body. Berkeley, CA: University of California Press. Boserup, Ester. 1970. Woman's role in economic development. New York, St. Martin's Press. Boyer, Kate. 1998. Place and the Politics of Virtue: clerical work, corporate anxiety, and changing meanings of public womanhood in early twentieth-century Montreal. Gender Place & Culture: A Journal of Feminist Geography 5, no. 3: 261-76. ________. 2005. Spaces of change: gender, information technology, and new geographies of mobility and fixity in the early twentieth century information economy. In A companion to feminist geography Lise. Nelson and Joni Seager, 228-56. Malden, MA : Blackwell Publishing. Butler, Judith. 1993. Bodies that matter: on the discursive limits of "sex". New York: Routledge. Castells, Manuel. 2000. End of Millennium. Oxford: Blackwell Publishers Ltd. Chengappa, Raj and Malini Goyal. 18 November 2002. Housekeepers to the World. India Today 27, no. 46: 36-49. Cope, Meghan. 2002. Feminist epistemology in geography. In Subjectivities, knowledges, and feminist geographies: the subjects and ethics of social research Liz Bondi, 43-56. Lanham, MD: Rowman & Littlefield. Cox, Kevin R. 1997. Spaces of globalization: reasserting the power of the local. New York: Guilford Press.

Cravey, Altha J. 1998. Women and work in Mexico's maquiladoras. Lanham, MD: Rowman & Littlefield. ________. 2005. Working on the Global Assembly Line. In A companion to feminist geography Lise Nelson and Joni Seager, 109-22. Malden, MA: Blackwell Publishing. Cresswell, Tim. 1996. In place/out of place: geography, ideology, and transgression. Minneapolis: University of Minnesota Press. ________. 1998. Qualitative inquiry and research design: Choosing among five traditions. Thousand Oaks, CA: Sage. Delaney, Kevin J. 13 October 2003. Outsourcing Jobs -- and Workers -- to India. Wall Street Journal - Eastern Edition 242, no. 73: B1. Dirsuweit, Teresa. 2005. Bodies, state discipline, and the performance of gender in a South African women's prison. In A companion to feminist geography Lise Nelson and Joni Seager, 350-62. Malden, MA: Blackwell Publishing. Domosh, Mona. 1998a. Geography and gender: home, again? Progress in Human Geography 22, no. 2: 276-82. ________. 1998b. Those 'gorgeous incongruities': Polite politics and public space on the streets of... Annals of the Association of American Geographers 88, no. 2: 209-27. Domosh, Mona and Joni Seager. 2001. Putting women in place: Feminist geographers make sense of the world. London: Guilford Press. Economist, 2004. A world of opportunity. 373, 8401: 19. Ehrenreich, Barbara and Arlie Russell Hochschild. 2003. Global woman: nannies, maids, and sex workers in the new economy. New York: Metropolitan Books. Elias, Marlene and Judith Carney. 2005. Shea butter, globalization, and women of Burkina Faso. In A companion to feminist geography Lise Nelson and Joni Seager, 93-108. Malden, MA: Blackwell Publishing. Faludi, Susan. 1991. Backlash: the undeclared war against American women. New York: Crown. Fountain, Jane. 2000. Constructing the information society: women, information technology, and design. Technology in Society 22: 45-62. Gibson-Graham, J. K. 1996. The end of capitalism (as we knew it):a feminist critique of political economy. Cambridge, MA; Oxford [U.K.] : Blackwell Publishers. Greenspan, Anna. 2004. India and the IT revolution: networks of global culture. New York: Palgrave Macmillan.

Hafkin, Nancy and Nancy Taggart. 2001. Gender, Information Technology, and Developing Countries: An Analytic Study. Washington D.C.: United States Agency for International Development. Hägerstand, Torsten. 1975. Space, Time, and Human Conditions. In Dynamic allocation of urban space Anders Karlqvist, L. Lundqvist, and Snickars, 3-14. Lexington, MA: Saxon House; Lexington Books. Hapke, Holly M. and Devan Ayyankeril. 2004. Gender, the work-life course, and livelihood strategies in a South Indian fish market. Gender, Place, and Culture 11, no. 2: 229-57. Haraway, Donna. 1988. Situated knowledge: the science question in feminism and the privilege of partial perspective. Feminist Studies 14: 575-99. ________. 1991. Simians, cyborgs, and women: the reinvention of nature. New York: Routledge. Harvey, David. 1989. The condition of postmodernity: an enquiry into the origins of cultural change. Oxford; Cambridge: Blackwell. ________. 2000. Spaces of hope. Berkeley: University of California Press. Herod, Andrew. 1997. Labor as an agent of globalization and as a global agent. In Spaces of globalization: reasserting the power of the local Kevin R. Cox, 167-200. New York: Guilford Press. ________. 2001. Labor geographies: workers and the landscape of capitalism. New York: The Guliford Press. Holloway, Lewis. 2004. Donna Haraway. In Key thinkers on space and place Phil Hubbard, Rob Kitchin, and Gill Valentine, 167-73. London;Thousand Oaks: Sage publications. Jackson, Peter. 2003. Space, theory, and hegemony: the dual crises of A sian area studies and cultural studies. Sojurn 18, no. 1: 1-41. Jackson, Peter. 2004. Local consumption cultures in a globalizing world. Transactions of the Institute of British Geographers 29, no. 2: 165-79. Jacobs, Jane and Catherine Nash. 2003. Too little, too much: cultural feminist geographies. Gender Place & Culture 10, no. 3: 265-80. Kantor, Paula. 2002. Female mobility in India - The influence of seclusion norms on economic outcomes. International Development Planning Review 24, no. 2: 145-59. Katz, Cindi. 2004. Growing up global: economic restructuring and children's everyday lives. Minneapolis: University of Minnesota Press. Kelkar, Govind, Girija Shrestha, and Veena Nagarjan. 2002. IT Industry and Women's Agency: Explorations in Bangalore and Delhi, India. Gender, Technology and Development 6, no. 1: 63-82.

Law, Robin. 1999. Beyond 'women and Transport': Towards New Geographies of Gender and Daily Mobility. Progress in Human Geography 23, no. 4: 567-88. Law, Robin. 2002. Gender and daily mobility in a New Zealand city, 1920-1960. Social & Cultural Geography 3, no. 4: 425-46. Longhurst, Robyn. 1995. The body and geography. Gender, Place & Culture: A Journal of Feminist Geography 2, no. 1: 97-106. Mandel, Jennifer. 2004. Mobility matters: women's livelihood strategies in Porto Novo, Benin. Gender, Place, and Culture 11, no. 2: 257-388. Marston, Sallie. 2000. The social construction of scale. Progress in Human Geography 24, no. 2: 219-43. Marston, Sallie and Neil Smith. 2001. States, scales and households: limits to scale thinking? A response to Brenner. Progress in Human Geography 25, no. 4: 615-20. Massey, Doreen. 1994. Space, place, and gender. Minneapolis: University of Minnesota Press. Maxwell, Joseph. 1998. Designing a qualitative study. In Handbook of applied social research methods Leonard Bickman and Debra J. Rog, 69-100. Thousand Oaks, CA: Sage Publications . McDowell, Linda. 1999. Gender, identity and place: a feminist perspective. Cambridge, UK: Polity Press. McDowell, Linda and others. 2005. The contradictions and intersections of class and gender in a global city: placing working women's lives on the research agenda. Environment and Planning A 37, no. 3: 441-61. Mirchandani, Kiran. 2003. Gender Eclipsed?: Work Relations in Transnational Call Centers. Paper presented at the ASAC 2003 Conference, Halifax, Nova Scotia Unpublished manuscript. ________. 2004. Practices of global capital: gaps, cracks,and ironies in transnational call centres in India. Global Networks 4, no. 4: 255-373. Mitchell, Katharyne, Sallie Marston, and Cindi Katz. 2003. Introduction: Life's Work: An Introduction, Review and Critique. Antipode 35, no. 3: 415-38. Moss, Pamela J. 2002. Feminist geography in practice: research and methods. Oxford, UK; Malden, MA : Blackwell Publishers. Nagar, Richa. 1998. Communal discourses, marriage, and politics of gendered social boundaries among South Asian immigrants in Tanzania. Gender, Place and Culture 5, no. 2: 117-39. Nagar, Richa and others. 2002. Locating globalization: feminist (re)readings of the subjects and spaces of globalization. Economic Geography 78, no. 3: 257-85.

NASSCOM. Indian software and services exports. 2002. [available from http://www.nasscom.org/artdisplay.asp?cat_id=408.]. 20 June 2004. NASSCOM. NASSCOM's ranking of third party players. June 2005. [available from http://www.nasscom.org/artdisplay.asp?Art_id=4400]. 3 October 2005. Nast, Heidi. 1998. Unsexy geographies. Gender, Place, and Culture 5: 191-206. Nichols, John. 30 August 2004, Global Fights Go Local. Nation, 279, 6, 22. Oberhauser, Ann and others. 2003. Geographic Perspectives on Women. In Geography in America at the dawn of the 21st century Gary L. Gaile and Cort J. Willmott, 736-58. Oxford; New York: Oxford University Press. Ong, Aihwa. 1987. Spirits of resistance and capitalist discipline. Albany: State University of New York Press. Oza, Rupal. 2001. Showcasing India: Gender, Geography, and Globalization. Signs: Journal of Women in Culture & Society 26, no. 4: 1067-95. Pandit, Kavita. 2005. Elite migration from 'body shopping' to 'reverse migration': The restructuring of the flows of Indian technology workers to the United States. Geography and the Environment Colloquium. University of Texas at Austin. February 4. Patel, Meenakshi Dhar. 2002. The Economic times IT enabled services 2002. Mumbai: Bennett Coleman & Co. Patel, Reena and Mary Jane Parmentier. 2005. The Persistence of Traditional Gender Roles in the Technology Sector: A Study of Female Engineers in India. Information Technologies and International Development 2, no. 3: 29-46. Poster, Winifred R. 1988. Globalization, gender, and workplace: Women and Men in an American multinational corporation in India. Journal of Developing Studies 14, no. 1: 40-65. ________. 2004. Who’s On the Line? Indian Call Center Agents Pose as Americans for U.S. Outsourced Firms. unpublished manuscript: University of Illinois at Urbana-Champaign. Pratt, Geraldine. 2000. Geography and Body. In The dictionary of human geography Ronald John Johnston and others. fourth edition ed., 48-9. Oxford, UK; Malden, MA: Blackwell Publishers. ________. 2004. Working Feminism. Philadelphia: Temple University Press. PUKAR. n.d. [available from http://www.pukar.org.in/pukar/genderandspace.html]. 15 October 2005. Rankin, Katharine N. 2003a. Anthropologies and geographies of globalization. Progress in Human Geography 27, no. 6: 708-35.

Rankin, Katherine N. 2003b. Cultures of Economies: gender and socio-spatial change in Nepal. Gender, Place, and Culture 10, no. 2: 111-30. Rubin, Herbert and Irene S. Rubin. 1995. Qualitative interviewing: the art of hearing data. Thousand Oaks, CA: Sage Publications. Rudrappa, Sharmila. 2005. Working borders: Linking debates about insourcing and outsourcing of capital and labor: Proceedings of the inaugural conference of the Bernard and Audre Rapaport Center for Human Rights and Justice, February 10-11. In Texas International Law Journal, 691-805. Russell, Bernard. 2002. Research methods in anthropology: Qualitative and quantitative methods. Walnut Creek, CA: AltaMira Press. Samarasinghe, Vidyamali. 2005. Female Labor in Sex Trafficking: A Darker Side of Globalization. In A companion to feminist geography Lise. Nelson and Joni Seager, 166-78. Malden, MA : Blackwell Publishing. Sassen, Saskia. 1998. Globalization and its discontents: Essays on the new mobility of people and money. New York: New Press. Sassen, Saskia. 2000. Spatialities and temporalities of the global: Elements for a theorization. Public Culture 12, no. 1: 215-32. ________. 2004. Local Actors in Global Politics. Current Sociology 52, no. 4: 649-70. ________. 2005. The Global City: Introducing a Concept. Brown Journal of World Affairs 11, no. 2: 27-44. Scott, Allen John. 1998. Regions and the world economy: the coming shape of global production, competition, and political order. Oxford; New York: Oxford University Press. Secor, Anna. 2003. Belaboring gender: the spatial practice of work and the politics of 'making do' in Istanbul. Environment and Planning A 35, no. 12: 2209-27. Shaw, Annapurna. 2004. The making of Navi Mumbai. New Delhi: Orient Longman. Sheppard, Eric. 2002. The spaces and times of globalization: place, scale, networks, and positionality. Economic Geography 78, no. 3: 307-31. Sheppard, Eric. 2004. The Spatiality of The Limits to Capital. Antipode 36, no. 3: 470-80. Silvey, Rachel. 2005. Borders, Embodiment, and Mobility: Feminist Migration Studies in Geography. In A companion to feminist geography Lise Nelson and Joni Seager, 138-49. Malden, MA: Blackwell Publishing. Singh, Preeti and Anu Pandey. 12 February 2005. Women in Call Centres. Economic and Political Weekly: 684-8.

Stitt, Greg. 2002. Diverted to Delhi. Prod. and directed by Greg Stitt. 55 min. Filmmakers Library. DVD. Strauss, Anselm L. and Juliet M. Corbin. 1990. Basics of qualitative research: Ground theory procedures and techniques. Newbury Park, CA: Sage Publications. ________. 1997. Grounded theory in practice. Thousand Oaks: Sage Publications. Tanner, Julian and Rhonda Cockerill. 1996. Gender, social change, and the professions: The case of pharmacy. Sociological Forum 11, no. 4: 643-60. Tuan, Yi Fu. 2004. Cultural Geography: Glances Backward and Forward. Annals of the Association of American Geographers 94, no. 4: 729-34. Tyner, James A. 1999. The web-based recruitment of female foreign domestic workers in Asia. Singapore Journal of Tropical Geography 20, no. 2: 193-209. Webster, Juliet. 1996. Shaping women's work: Gender employment and information technology. London: Longman. Wilson, Elizabeth. 1992. The sphinx in the city: urban life, the control of disorder, and women. Berkeley: University of California Press. Wright, Melissa W. 1997a. Crossing the factory frontier: Gender, place and power in the Mexican Maquiladora. Antipode 29 , no. 3: 278-302. ________. 1997b. Third world women and the geography of skill. Ph.D. diss.,The Johns Hopkins University. Yeoh, Brenda and Shirlena Huang. 1998. Negotiating public space: strategies and styles of migrant female domestic workers in Singapore . Urban Studies 35, no. 3: 583-602.

APPENDIX

vii. Solís, Patricia. International Research Fellowship Program: Advancing Academe: A Multidimensional Investigation of Geography in the Americas (AAMIGA).

Patricia Solís, PhD Advancing Academe: A Multidimensional Investigation of Geography in the Americas (AAMIGA) Background and Significance As producers of knowledge in a globalized society, scientists and scholars must reckon with a world where economies are increasingly connected, where country boundaries are evershifting, where communication technology enables information to traverse the globe with great speed, where workplaces are increasingly internationally distributed, and where cities and towns are growing in ethnic and racial diversity. These phenomena impact the activities of knowledge producing enterprises, a fact which has prompted research-oriented organizations such as universities, R&D business, and federal agencies to promote international research collaboration (IRC) as a means to build intellectual capacity and increase competitiveness, among other goals (See National Science Foundation NSB 03-69; NSB 00-217; Association for the Advancement of Science Teich 2000; Social Science Research Council SSRC 2000). Despite widespread recognition of the increasing importance of IRC, there have been surprisingly few systematic studies measuring or assessing international research collaboration within academe (Drake, et al. 2000:2). Those who have studied the issue have not come to a clear consensus on what IRC actually comprises, let alone on the current status of IRC within academe, and they often publish somewhat contradictory results (Hauger & McEnaney 2000; Abraham & Harris 2000). This lack of clarity may be due to the tendency of the existing research to compartmentalize subjects into natural versus social sciences, to underestimate the effects of place-specific or cultural context, and to fail to articulate connections among local-national-regional dimensions of collaboration. The literature furthermore all but neglects the influence that key institutional actors – disciplinary professional societies – exert toward facilitating productive IRC. Thus the purpose of this proposed research is to systematically and holistically investigate the dynamics of international research collaboration through a multi-dimensional study of geography in the Americas. The aim is to better understand the nature of IRC within a particular context, looking through the window of an “interdisciplinary discipline”, geography, onto a region, the Americas. The project itself will be conducted as an internationally collaborative research project, in conjunction with the City of Knowledge (CoK) in Panama and with input from the Association of American Geographers (AAG) in Washington, DC. Such knowledge may then be applied, particularly by professional associations or for NSF international programs, to advance the intellectual and developmental impact of academe in our increasingly interconnected world. Why Investigate International Research Collaboration (IRC)? Too little is known beyond the general outlines of the status and current trends of IRC in academe and information that is available is at worst sometimes contradictory and at best often not conclusive or descriptive enough to guide institutional efforts at promoting collaboration. One study shows that although there has been an absolute increase in the number of joint publications indicating IRC within disciplines in the social sciences and humanities, the proportion so measured is actually falling (SSRC 2000). Others tout a growing trend within these same fields, especially because they represent “less expensive” theoretical research that takes advantage of increasing communication and mobility of scholars (Glanzel 2002; Luukkonen, Persson, & Silvertsen, 1992). Similarly, Wagner’s (2005) survey of a number of fields in the natural sciences show an increase in IRC generally, while highly variable among disciplines. Still, these studies rely upon 1

joint authorship as a proxy measure for IRC, an index which misses many of the activities that count as international collaboration but do not result in published journal articles. Leaders in academe agree almost unanimously that IRC as a concept is a worthy goal. The ability to initiate and sustain IRC over time is both a means and an indicator of increased competitiveness as well as creating value through the development of social capital (Rogers 2000; Fountain 1998). Indeed, faculty members with some experience with international collaboration view the practice as professionally important for themselves and educationally significant for their students and they see themselves as more capable researchers (Solem & Ray 2005). For these professors, internationalization promises “new theoretical constructs, new ways of doing research, the use of foreign data to test hypotheses, and the emergence of a global perspective” (Groennings 1990:22). However, there remains a gap in understanding between the assertion of the value of internationalization of research and actual practices and priorities. Despite some variety of perspectives on IRC, one opinion is common among them, that further analysis of the US scientific community is needed.† Why Geography as a Disciplinary Approach? A focus on a single discipline’s experience with respect to IRC offers numerous advantages. Researchers have found that the extent of multinational collaboration is strongly field–specific (Glaenzel & De Lange 2002) and that the forms that IRC takes may differ with respect to the field that collaborators belong to (Melin 2000). A focus on disciplines provides insights into the structure and processes of IRC, particularly in light of sociological studies of academic “tribes” (Becher 1989; Whitley 1984). Yet while a disciplinary-based approach might underestimate actual work practices that often grapple with multiple fields and sectors, geography could prove to mitigate this for three main reasons. First, geography’s practitioners have a long history of working with scholars in other fields, as an “interdisciplinary discipline” that is poised to continue to grow and contribute to society at large in coming years (Richardson & Solis 2004). This phenomenon includes collaboration within their home institutions: about 40 percent of departments in the US are double affiliations, e.g. Department of Geography and Anthropology or Department of Geography and Geology (AAG Guide, various years). The breadth of the discipline spans the spectrum of natural sciences, social sciences and humanities. Secondly, preliminary information indicates that the nature of IRC among geographers lends itself well to a holistic study of multiple forms of collaboration. While Abraham & Harris (2000) found no joint publications in their survey of geographic literature, Solem & Ray’s interviews (2005) revealed that fully 57% of geography faculty surveyed reported at least one experience collaborating with international colleagues, reporting a diverse range of research projects across the major geographic subfields. This research points to the importance of faculty training and support systems as factors facilitating IRC. Third, the experiences of geographers who have participated in the dialogue on international education can inform a study on international research collaboration as well, particularly about connections between education and research including the development of international networks for teaching and learning (Healey 1998; Shepherd, Monk, and Fortuijn 2000; Donert 2003), critiques of Western bias in internationalization of education (Garcia-Ramon and Monk 1997), and the role of ICT and distance learning (Foote 1999; Jain and Getis 2003). Why Focus on the Americas? While a disciplinary focus can offer a more nuanced perspective of IRC, the intellectual organization within specific disciplines does not fully explain dynamics of the phenomenon (Wagner 2005). Clearly the status and trends are also contingent †

But see Bond & Lemasson (1999) for a comprehensive treatment of the Canadian university system.

2

upon context: where the collaborations are literally taking place makes a difference. Hilbink and Drake (2000) note that by definition IRC always “involves a literal encounter or recognizes the influence of ideas from more than one place” and “that location matters.” While Wagner (2005) used visualization techniques to demonstrate that the growth of international networks among scholars in different countries are independent of the organizing motives of distinct disciplines, she neglected to produce an actual geography of where IRC occurs. What geographers and others have long known about international migration – namely, that flows are not random but demonstrate distinct spatial patterns with clear source and destination areas, called “migration streams” – may also be applied to an investigation of international research linkages. Thus this study seeks to understand connections within one region rather than treating IRC as a global phenomenon indiscriminate of the context of place. The American continents provide a compelling starting place, beyond the obvious reason that it already includes the US. This region contains a set of countries at a variety of economic developmental levels and a diversity of cultural contexts and sub-regions. Despite this diversity, it also comprises relatively fewer official languages than other broad regions of the world (Europe, Africa, Asia), facilitating the logistics and cost of the proposed study. Furthermore, over the past 40 years, scholars of and from Latin America in particular have contributed to some of the more influential social theories at the forefront of research (Coatsworth 1989). Geography as a discipline, too, has experienced a fertile landscape of collaboration among the Americas, in small part evidenced by the growth of regional geographers focused on Latin America, who became the first and only specialists of the association to form their own sister organization, the Conference of Latin American Geographers in 1970 (www.clag.org), now also active with the AAG’s Latin Americanist Specialty Group. To the North, Canada is already considered within the AAG’s governance structure. More broadly, the incredible growth of the Hispanic population in the US (Census 2000) provides further impetus for a focus on the Americas. Why from a Professional Association Perspective? Because the disciplines are the intellectual home of faculty members, disciplinary professional associations can play a central role in any effort to understand and enhance the international dimensions of academic research (Lawson 2005). However, research on IRC has focused on primarily two sets of actors, the individual scholar and universities or groups of institutions and seems to neglect the influence that disciplinary professional societies wield or potentially could bear upon facilitating productive IRC. Such an omission paints an inaccurate picture of the reward structures that may motivate, support, or promote IRC. These associations are also often at the forefront of specific research initiatives that are timely and inherently global or international in nature (e.g. the AAG’s Geographic Dimensions of Terrorism; Human Dimensions of Global Climate Change; responses to natural disasters, South Asian Tsunami and Hurricane Katrina). Association arenas are highly flexible and can often be more responsive with rewards or recognition that may motivate IRC than the bureaucratic university system, and as such, they may provide or may already be providing effective and relatively straightforward means of facilitating international research collaboration. What is the intellectual merit of the proposed activity? AAMIGA proposes to systematically study IRC through a multi-dimensional perspective of one discipline in one region. As such it will contribute to advancement of knowledge about collaboration within geography, but will also apply findings and provide a model for understanding other fields. What are the broader impacts of the proposed activity? While the project is principally a research investigation, it incorporates elements and methods that will ultimately facilitate IRC.

3

As such, participants from the region will not only serve as informants and survey respondents, but will also be actively engaged in the advising process. The AAG is poised to implement AAMIGA’s findings, seeking to understand IRC for the very purpose of facilitating it. How is Integration of Research and Education addressed? Because internationalization of education has already long received attention within geography, it provides one arena for understanding IRC. AAMIGA specifically seeks to learn how multi-country efforts to achieve global learning may also be setting up collaborations of scholarly inquiry. In the process, we hope to discover how collaborative geographic research and education are interrelated. How Might the Project Contribute to Enhancing Diversity? Because the study area contains a large number of Spanish-speaking countries, the Spanish-speaking minority population of geography scholars, current and potential students living within the US can benefit from the outcomes and deliverables of AAMIGA. To this end, members of the AAG’s Diversity Task Force will be engaged on the project advisory board.

Project Goals The overall goal of the project AAMIGA is to understand the current status and assess trends of international scientific collaboration of research activities throughout the Americas within the discipline of geography. The Association of American Geographers, as the largest scholarly society of geographers and related professionals, represents a locus for catalyzing communities toward this effort to advance academe. The AAG is the current professional home of the PI, and has expressed its support for this proposal (see letter, Richardson). The foreign host institution, the City of Knowledge (CoK), represents one of the most outstanding centers for international collaboration in the Americas, and has a mission that aligns squarely with the goals of the AAMIGA project (see below and letter, Tarté). Specific objectives and questions that drive the research fall into three dimensions (regional status, local contextual dynamics, and broad facilitation assessment): 1. To discover, articulate and communicate collaboration and collaborative opportunities among geographers and closely related scholars in the US, Canada, Central America and the Caribbean, and South America. What is the state of geography in the Americas as a discipline? What is the state of geographic research in the Americas? Its connection to geographic education? Where are the major sites and axes of collaboration with respect to participation in global or regional scale projects/networks, intergovernmental agreements, exchanges of personnel (students, researchers, experts), exchanges of information (curriculum, data, etc.), and/or joint conferences, meetings? Who are participating in international research collaboration (gender, ethnicity/race, location, subfield, age, etc.)?† 2. To explore how geographic contexts within the Americas impact upon actual IRC activities and how such considerations should be addressed by means of a national level case study. What are some of the common regional peculiarities within the Americas that must be considered when building collaborative efforts? Using Panama as an example, how do various geographic sectors interact within IRC settings? What are the needs of geography professionals in terms of capacity building for international collaboration? How do differing reward systems and inequalities in resources devoted to research impact IRC‡? How can †

Solem & Ray’s survey of geography faculty’s involvement in internationalizing education (2005) found no significant differences regarding how respondents view the professional value of IRC on the basis of gender, institutional type, academic rank, or academic specialization but did find a generational gap, with the oldest generation opining that it does not offer substantive benefits for professional development. ‡ See Bunders & Mukerjee (1997).

4

language barriers be overcome (English, French, Spanish, Portuguese)? By what means can translations of a highly academic character requiring specialized vocabularies be facilitated? What roles do ICT and the digital divide play in realizing or inhibiting connections? 3. To develop an appropriate assessment of the role of professional societies for IRC using the AAG as a model, culminating in an action plan and identification of resources for the association to promote international research collaboration within the discipline as a means to strengthen geography as an academic and scientific enterprise. What is the most appropriate role of a professional society in facilitating connections and strengthening the discipline through IRC? How should the various audiences be engaged (students, undergrads, grads, teachers, professors, private sector geographers, public sector geographers, related professionals)? What specific activities would be most productive for strengthening and internationalizing geography / geographic science in the US and the other regions of the Americas? How should the professional community approach capacity building for all participants? How should institutional and sustainability issues be addressed, such as service provision and formal cooperation between interested universities and other organizations in a way that accounts for contextual considerations? In light of research results, what recommendations can be made regarding organizational structure for internationalizing the AAG and/or professional societies in general? What specific benefits, roles, responsibilities, and forms of participation are possible and appropriate?

Technical Approach To answer these questions, this research necessarily combines quantitative and qualitative methodologies to perform its multi-dimensional objectives. To acquire data to answer the questions of general disciplinary status, a series of surveys of departments (to determine where are geography programs) and scholars (to determine where is IRC occurring) will be conducted. The surveys will be electronically administered using a Web-based system, although paper letters of invitation and telephone interviews will be used to boost response rates where needed. Each survey will undergo external review and revision before being sent to respondents. Respondent information will come from the AAG’s database of 30,000 contacts and internet research of universities in the Americas (e.g. Yahoo Universities Directory). Information will be geo-referenced and stored into a Geographic Information System (GIS) for spatial analysis of resulting patterns. The case study will be designed to address the contextual considerations outlined above, and the fieldwork methodology will use historical research, interviews, focus groups convened at international conferences and online via the AAG’s e-communities, and participant observation at the CoK. In focus group settings, the moderator commonly encourages open discussion of an agenda of questions as a means of encouraging participants to share ideas and experiences (Krueger and Casey 2000). Mostly unstructured, phenomenological questioning will be used regardless of face-to-face or electronic communication. Accepted evaluation methodology based on Effectiveness Models (hybrid comprehensive-stakeholder) will be used to determine what role professional societies do and should play (Vedung 2000). All activities will be informed by input from an international advisory board that includes broad geographic and topical representation, and liaisons from the AAG Diversity Task Force, the AAG-Latin Americanist Specialty Group, the Conference of Latin Americanist Geographers, the Pan-American Institute for Geography and History, and others. The purpose of the board is to ensure scientific quality of the research and to catalyze leadership for future facilitation of IRC in geography. All research and dissemination activities relate to ongoing and

5

planned research efforts at the CoK in Panama. In addition to providing a local setting for the case study, the broader research plan will: 1) assist in efforts to make geographic research available in the region by identifying the topics of actual collaborations; 2) inform efforts to amplify programmatic definitions of geographic research beyond applied or technological agendas; and 3) support capacity building and IRC-enabling activities at CoK (see letter). In reports, results will be presented in a manner that permits verification by the reader and considers the full range of positive, neutral, and negative comments elicited from the proceedings, in the original language with translations, ensuring scientific transparency and building rapport with the participants (Guba and Lincoln 1994). The qualitative data generated by interviews, focus groups, and participant observation, when considered alongside the quantitative survey results, will help advance understanding of the dynamics of international research collaboration among the various dimensions where it occurs. Evaluating results with respect to professional societies will help discover how IRC in geography can be effectively facilitated. The timeline in the appendix offers further detail.

Expected Results and Project Deliverables •

• • • • •

A Directory of Geography in the Americas (based upon an expanded AAG Guide) published in hard copy and in digital format using GIS to organize and analyze data; A Case Study of the State of Geography in Panama; A Geography Summit bringing together scholars from the Americas to serve initially as the AAMIGA research project advisory board and to catalyze a longer-term effort for major disciplinary advancement; A Set of Best Practices for engaging with fellow geographer communities (addressing contextual, language, technology, and other issues) that can also apply to other knowledge-workers and professional societies; A Sustainability Plan indicating funding sources and including development of additional proposals to sustain select activities; and An International Model that can be applied to other disciplines and other regions to advance scientific cooperation in research and scholarly inquiry, to be broadly disseminated through AAG and CoK networks, such as with COSSA, ACE, AAAS, etc.

Future Plans of the Principal Investigator I accepted my position as Director of Research and Outreach at the AAG immediately after completing my doctorate at the University of Iowa in 2002. My motivation in seeking employment with the association was develop my career within the discipline, but in a way that bridges the academic and research realms with the broader national community. In less than three years, this opportunity has opened new possibilities and granted me this wish. However, much of my work is necessarily project design and administration, prompting me to seek this opportunity to further my professional research agenda and hone my skills as a researcher in my own right. This fellowship presents a unique opportunity at this juncture of my career – one that will allow me to continue contributing to the discipline while opening up essential space for undertaking serious research. Upon completion of the fellowship, I will continue work at the AAG implementing the findings. My role would be enhanced in that I will be positioned to lead a new direction and initiative for the organization. My longer term career vision may include select posts at geography research institutes or perhaps within a university setting, so continuing my ability to produce and public significant research is critical for whatever the future may hold. 6

Rationale for International Host Site A serious study of the state of IRC in geography would do well to actually be itself the product of international research collaboration, as the AAMIGA project proposes to be. A set of seven criteria was identified for selecting the most appropriate host collaborator. (1) Should be located in the South to indicate the intended sense of joint ownership and to provide immersion in a non-US perspective of IRC in order to experience inequality of material resources or of power in setting research agendas, or other factors of difference which are best understood in context. (2) Should provide sufficient facilities to support international work at the scale proposed. (3) Should be focused on IRC itself, or at a minimum include it within its mission and programs. (4) Should be a broad-based institution, like a consortium of universities or research centers, as opposed to a single university or university system. (4) Should include a research focus need amenable but not necessarily restricted to geography and should engage in or present opportunities conducive to geographic research. (5) Should posses an academic core, but be active in connecting to public and private entities, given that IRC is often multi-sectoral. (6) Should offer a local environment conducive for the case study. (7) Should be organizationally disposed to participate in plans for the future sustainability of activities. Within the region, the host institution that meets and excels in all of these criteria and thus provides the maximum potential as host is the City of Knowledge (CoK) in Panama City, Panama (www.cdspanama.org). The CoK is an international complex of academic, research, technological development, industrial and cultural institutions, which converge within a context of integration of efforts toward promoting the generation, dissemination and use of knowledge for achieving integrated human development. Located within a former US military base in the Canal Zone area near Panama City, CoK forms a complex of more than 300 buildings and housing on a campus of 120 hectares. It provides high technology facilities and a capacity to establish strategic alliances among its members. The CoK has fully equipped rooms of assorted sizes and an Intelligent High Technology Center with facilities for videoconferences, ISDN and IP communications through broadband, intelligent classroom, and distance learning. Its membership includes several national and indigenous governments and most of the Panamanian government’s federal agencies; UNEP and UNDP; cooperation institutions such as USAID, CATHALAC, IICA, and GWP-CA; more than three dozen universities and research centers; scores of national and international NGOs, and numerous private sector firms and associations. The mission of the academic component of the City of Knowledge is to identify and facilitate links and associations among academic, research, and innovative technological institutions from around the world, in areas where scientific and humanistic knowledge can promote human development. A dedicated academic department is made up of an international consortium of institutions of excellence whose interests coincide with advancing scientific and academic collaboration in both research and education. The Academic Director, Dr. Rodrigo Tarté will serve as host collaborator on this project, ensuring that the research will be able to fully take advantage of the unique characteristics of the CoK and its academic activities in particular (see letter) as well as the high-level experience and academic networks that Dr. Tarté possesses (see CV). As such, the CoK represents an institutional base most appropriate for the AAMIGA research project because it provides effective infrastructure support as well as regional legitimacy since it is both located within the region and has strong connections across the Americas, ensuring that this research project upholds high standards of scientific quality and achieves the broader impacts intended.

7

AAMIGA Project Timeline - Dr. Patricia Solis (Tentative Start Date of March 2006) ACTIVITY Initial Advisory Board & Planning Session at AAG Annual Meeting in Chicago

Broad Facilitation Assessment

Local Contextual Dynamics

Regional Status

Gather departmental data Develop survey respondent listings Draft and pilot survey instruments; translate Implement surveys Mail reminders to nonrespondents Conduct follow-up telephone interviews to nonrespondents Create and input data into GIS Analyze data (spatial and statistical) Publish data in Directory and online Research local case study information sources Participant observation at CoK Investigate disciplinary history Devise focus group methodology/participants Conduct focus groups Data analysis of focus groups Devise interview methodology/respondents Conduct interviews Data analysis of interviews Research presentation at US geography conference Plan Geography Summit Send invitations for Geography Summit and Advisory Board meeting Conduct effectiveness model evaluation analysis Write project reports / NSF report

PURPOSE

M

To review research plan, identify additional collaborations To inventory geography in the Americas To identify IRC activities and participants

X

X

A

M

J

J

A

X

X

X

X

X

S

O

X

X

X

X

N

D

J

F

Deliverable Advisory Board

X

X

X

Directory of Geography in the Americas

X X X

To increase response rates

X X

To visualize and understand spatial patterns To share information and facilitate linkages To discover and understand qualitative context of IRC

X

X

X

X

X

X

X

X

X

X X

X

X

X

X

X

X

X

X

X

X X

X X

To test assumptions, determine extent or quantify findings To share results and gather geography community feedback, seek input for recommendations To develop recommendations To disseminate results and catalyze future IRC

X X X

Case Study

X X

X X

X

X

X X

X

Hold Geography Summit

X

Dissemination to geography community and other disciplinary societies

X

Set of Best Practices; Sustainability Plan Geography Summit International Model

Proposal Title:

International Research Fellowship Program: Advancing Academe: A Multidimensional Investigation of Geography in the Americas\(AAMIGA)

Context Statement NATIONAL SCIENCE FOUNDATION CONTEXT STATEMENT The International Research Fellowship Program (IRFP) is a National Science Foundation program designed to introduce scientists and engineers in the early stages of their careers to research opportunities abroad, thereby furthering NSF's goal of creating a diverse, competitive, and globallyengaged U.S. workforce of scientists, engineers, technologists and well-prepared citizens. For the 2006 competition, 208 proposals were received. They were evaluated using the NSF review criteria of intellectual merit and broader impacts as articulated in the Grant Proposal Guide (NSF 0423), as well as qualifications of the proposed host and host institution and complementarity, and qualifications of applicant, including applicant's potential for continued growth. On March 1-2,2006, an NSF panel met to discuss the reviews and proposals and placed them in the categories of Highly Recommended, Fund if Funds are Available and Not Recommended for Support. Panel discussions took into account not only the technical merit of the proposal but also the likely impact of an award on the applicant's career. Based on the discussion, the reviewers' comments, letters of reference, and relevance to program priorities, the panel placed 25% of the proposals in the category of Highly Recommended. NSF expects to fund approximately 20% of the 208 proposals received in this competition. Proposers should keep in mind that individual reviews are advisory; decisions to fund or decline are made by the National Science Foundation. These decisions are often difficult, and factors other than reviewer comments play a role. Such factors include the total funds available to the program, balance among fields and geographic regions, previous international research experience, and general NSF policies. Among proposals of comparable merit, those with no previous international experience and/or those who have not yet secured a tenure-track position will take precedence.

REVIEW #1: What is the intellectual merit of the proposed activity? This is an exciting and important proposal. The applicant argues convincingly for the need to study the best practices and potential for international research collaboration. The intellectual merit of the proposal seems to me to be that we don't really have a clear idea of what collaboration means (or could mean); as the author notes, collaboration is often equated with co-publication of research articles, when, in fact, the range of possibilities is much broader. Although ambitious, the project is well designed. Concentrating on research networks in Latin America is a well chosen direction: research networks are well developed in Latin America both between US-based geographers and Latin Americans and within the region. I hope the PI will focus on both of these aspects. What are the broader impacts of the proposed activity? (In addition, please consider the prospective benefits to the applicant, the scientific discipline and the United States.) The project has tremendous potential benefits to geographers in Latin America, to Latin Americanist geographers based in the United States, and as a model for other disciplines. The applicant delineates clear project "deliverables," including a publication of "best practices" for research collaboration. Qualifications of proposed host and host institution and complementarity.

The proposed host institution, the City of Knowledge in Panama, is located inside a former UA military base in the canal zone. The host institution seems to have an excellent technical base to support this project, and connections with government and international agencies. Less clear, however, are the connections with universitites in Panama and with intellectual networks outside the nexus of development agencies. The PI might do well to make connections with regional networks such as the Central American Council of Universities (CSUCA). Qualifications of applicant, including applicant's potential for continued growth. The applicant has the requisite skills to carry out this project. She is experienced at the management of large-scale projects, and she is very well connected to Latin America and to the Geography networks she will need to access. The project would allow the applicant to combine her expertise in policy and project execution with the further development of her research skills. Her supporting letters are quite strong. Summary Statement I urge the selection committee to give high priority to funding this project. I am convinced it will lead to tangible and substantial benefits not only for Latin Americanist geography but for internationally oriented geography in general. The applicant is well positioned to carry out the project and build upon its results.

REVIEW #2: What is the intellectual merit of the proposed activity? I feel that this proposal builds a framework for facilitating continued intellectual collaboration across the Americas and is therefore extremely important. I would have preferred to have seen more specifics on how the methodology would be employed (e.g. survey questions, points of discussion for the focus groups, ), but I suspect this is still being formulated. I think conceptually, the project has great intellectual merit and will provide an important contribution to the further internatationalization of the discipline of Geography. What are the broader impacts of the proposed activity? (In addition, please consider the prospective benefits to the applicant, the scientific discipline and the United States.) The broader impacts of the proposed project are explicit. The author seeks to build greater intellectual collaboration among Geographers across the Americas. Personally, I think this is vital to the continued intellectual growth of the discipline at large. I feel that the minimal linkages that currently exist are a lost opportunity for important intellectual engagement and I'm glad to see that steps are being taken to improve these relationships. Qualifications of proposed host and host institution and complementarity. It seems that the City of Knowledge is an ideal pilot study location for this project. They have the infrastructure to facilitate the multiple aspects of the project and the author seems to have strong support from the Director. In sum, I feel that the host institution will prove to be an important asset in the success of this project. Qualifications of applicant, including applicant's potential for continued growth. The applicant seems to have the appropriate background, work experience and linguistic skills to carry out this project. She has very strong letters of support and her grant writing record speaks for itself. Her extensive experience in Latin America is also a valuable asset. I see nothing in her background that would inhibit her from carrying out the proposed research. Summary Statement As stated above, I would have preferred to see a bit more detail on the methodology, but all in all I feel that this is a timely and important research project that could pay dividends for many years to come.

REVIEW #3: What is the intellectual merit of the proposed activity? Though the proposal mentions international research collaboration (IRC) as a key ingredient of building intellectual capacity, this proposal is little more than an effort to identify who is doing (or has done) what in geography in (selected?) countries of Latin America, and the various roles played by national and international organizations and geographical societies. While the mixed methodology to be used (questionnaires, surveys, interviews, etc) is described its products can be little more than compilations of persons, events, joint publications etc. The role of professional associations and organizations within the geography discipline is extremely limited; they do little more than provide clearing houses of information (both printed and electronic), and/or organize regional and annual meetings. Very rarely have they done anything as research entities: that is left to universities and individual geographers. Indeed since they have very limited budgets they can do little in the fields of investigation and teaching. That holds true for both the USA and Latin American contexts. The limited role of the International Geographical Union (IGU), for example, is further evidence of the long-standing lack of interest in international collaborative work amongst the geography community. Among the European community countries, which is where one could argue, given their level of development, one is probably most likely to find such international collaboration, national boundaries are still dominant and IRC is negligible. There also already exist mechanisms for obtaining information from each country as to the status of geography as a discipline, as well as many essays published (or shortly to be published) describing the same. Panama is not one of the most significant countries for examining the status of geographical research given its limited size and relative lack of contact with other more developed centers in Mexico, Colombia, Brazil, Argentina, Chile etc. Thus its selection as the base for the study is difficult to understand. What are the broader impacts of the proposed activity? (In addition, please consider the prospective benefits to the applicant, the scientific discipline and the United States.) One must assume that the analysis to be carried out will lead to the implementation of systems to ensure further development of IRC within the geography discipline, but a resistant problem, not sufficiently dealt with in the proposal, will be future funding. Qualifications of proposed host and host institution and complementarity. The selected institution has not, to date, had any role in geographical IRC and I would have thought that an entity such as the Pan American Institute of Geography and History (Mexico) would have been a better base, since it already has key linkages to all the countries (save Cuba) of the hemisphere. Qualifications of applicant, including applicant's potential for continued growth. While the applicant has a well-developed linkage to the AAG, and experience in its developmental initiatives, and has clearly worked very efficiently there, she has very limited international experience other than her dissertation work in Patagonia. I am also worried about her lack of fluent Spanish and Portuguese. Summary Statement I cannot support the proposal in its present form. I would advise the applicant to carry out a more limited sample study before embarking on such a major enterprise.

REVIEW #4: What is the intellectual merit of the proposed activity? I find this proposal quite intellectually meritorious in the sense that it is likely to not only break new ground, but also produce solid results, that will in turn potentially stimulate further research and action/application on the questions and issues surrounding international research collaboration. What are the broader impacts of the proposed activity? (In addition, please consider the prospective benefits to the applicant, the scientific discipline and the United States.) To my knowledge there has been relatively little research on IRC in the Americas, especially paths and patterns between Latin American and North American researchers. Within the discipline (geography) that this study will focus on, there is a very meagre literature on the topic. 1. Therefore, the first broader impact will simply be to make the individuals and institutions that are contacted and interviewed in the course of this research, aware of the topic. As suggested in my comments on the intellectual merit of this proposal, it should stir up a good deal of interest, and presumably some spin-off or replicative efforts by individuals that might not have thought of doing research in this area before. In a word or two, I think this has consider potential to take on a life of its own, once underway. 2. Another impact that I would hope to see, is that this project will act as a stimulus for work on the history of geography as discipline in Latin America. There have been few (if any) regional wide surveys of the state of geography within Latin American universities and government institutes. Some of this institutional history exists at the national level, but putting it together at the present time at the bi-continental scale of the Americas would be a big step forward for the eventual collation and excavation of this kind of information in historical time. 2. I can foresee immediate positive impacts of this study for the discipline of geography. The disciplinary collectivities that constitute geography's professional corpa in the Americas have co-existed in relative isolation (mostly segmented by national identities and boundaries) for a century or more. Major exceptions have been the U.S.-Canada interchanges, and some of the regional networks that exist in Latin America, but taken on a continental level as this study seeks to do, the picture is one of limited interactions. My impression is that in the past decade, however, there is beginning to be a much greater degree of interchange between geographers within the Hemisphere. This of course, has been greatly facilitated by e-mail and internet advances. This study has the potential to further coalesce this trend, and perhaps put the whole interaction field on new levels. As for the prospective benefits, in each of the categories (applicant, discipline, and country) I can envision a number of benefits. 1. Dr. Solis and her associates at the AAG and City of Knowledge among others, will undoubtedly benefit. The applicant will become known to a wide number of geographers throughout the Americas and depending on the degree of success of the study, will be in a position to provide services and information to these individuals and their respective institutions. This could accrue career-long benefits to the researchers involved. There are any number of other possibilities here, but in the interest of brevity, I will simply say that Dr. Solis has chosen a research topic that offers great potential for producing reciprocal benefits to a large number of individuals and institutions. 2. Discipline of Geography - I've already touched on some of the obvious benefits that could be generated for academic geography: building transnational networks, facilitating individual-to-individual exchanges, and perhaps more importantly, institution-

to-institution exchanges. Beyond this, this increased interaction should help strengthen the discipline's standing within given national academic arenas. U.S. geographers and geography should profit from learning more about geographic communities in other countries -- ones that perhaps offer examples of how the discipline is more fully represented in some nations' national life and academia. 3. United States. Here, the prospective benefits would seem to be particularly transparent and compelling. Given recent national policies that have tended to marginalize or even abandon earlier ideals (ironically especially those forged during the Cold War) of participating in international fora, treaties, and initiatives that address questions and problems with direct geographical content and implications, any efforts by geographers that mitigate such trends should find a warm reception abroad. This may be doubly true of Latin America at a time when U.S.-Latin American relations have reached something of a nadir in many observers' eyes. It would be hard to imagine an academic-geographic initiative with greater "bang-for-the-bucks" potential AND national interest benefits. Qualifications of proposed host and host institution and complementarity. I do not have direct knowledge of the host or the host institution on the Latin American end (The City of Knowledge) other than being invited to review some of their translations (English-Spanish) of classic & key U.S. Latin Americanist works. I was greatly impressed not only with the quality of the translations, but even more with the choice of materials. If this one venture (mentioned in the host's comments) is indicative of the quality and perspicacity of the institution's overall operations, then I would say that they are extremely well qualified. I have a good deal more familiarity with the U.S. host and host institution (the AAG and its Executive Director, Doug Richardson). I have the highest regard and confidence in the AAG at this juncture to provide the necessary support and at a highly competent level. I don't have a informed opinion on the question of complementarity (AAG/City of Knowledge), but as with all innovative enterprises -- "you have to start somewhere" -and my bet is that the fit will be pretty good, or if not at the outset, then it will become so. Of course, this is yet another aspect/benefit of this project -- to forge these interinstitutional links that have been largely lacking between North American and Latin American geographic institutions. Qualifications of applicant, including applicant's potential for continued growth. [Material redacted per PAM Chapter XI G.2] As for potential growth, I can only speculate here. But, having immersed herself the proposed project for several years, all sorts of opportunities will present themselves. Thus, it would be hard to imagine the applicant not experiencing a great deal of both personal and professional growth through this project. Summary Statement I rarely rank NFS proposals in the "excellent" category, but in this case I certainly will. The qualifications of the researcher, host individuals and institutions aside (though I would characterize them as excellent as well), it is overarching importance and probable tangible results and benefits of this project that makes it simple -- even imperative -- for me to urge that it be funded, and to rank it as "excellent.

APPENDIX

viii. Song, Conghe. Geography and Regional Science: Scaling Up Forest Ecosystem Carbon Budget from Stand to Landscape: Impacts of Forest Structures.

Project Summary Broader Impacts: The overall goal of this project is to improve understanding of the impacts of the multi-dimensional forest structures on scaling up forest ecosystem carbon cycle across spatial and temporal scales through the integration of remote sensing, ecological modeling and ground observations. The research expands the scope and depth of the existing projects at Duke Forest. The project will take the necessary steps to transform the understanding of mass and energy exchange between forest ecosystems and the atmosphere at Duke Forest to a regional understanding. Currently the global FLUXNET has over 200 towers worldwide. The up-scaling strategy investigated in this project will provide tools to scale up from FLUXNET measurements to a global understanding of carbon cycle. Forest ecosystems have been identified as the primary terrestrial ecosystem that may be responsible for the “missing sink” of global carbon cycle in 1990s. Plant communities in the Duke Forest are representative in the US Southeast, where 50 percent of the land area is covered with forests, and is believed to be one of the potential locations for the North America carbon sink. The results of the project can help further narrow down the location of the terrestrial carbon sink in the US and the mechanism responsible for the carbon uptake. The overall goal of this research contributes to the goals of North America Carbon Program. Intellectual Merits: The project proposes to use remotely sensed data from multiple sensors to map the multi-dimensional forest structures, including tree size and density, stand ages, leaf area index, and subpixel tree cover. The remotely sensed data include high-resolution (≤1m) digital orthopoto quads and space-born images from Ikonos/QuickBird, medium resolution (30 m) Landsat images, and coarse-resolution (250m) MODIS/MISR images. The project will develop algorithms to use information from spatial, spectral/temporal and directional domains of remotely sensed data. The project can substantially enhance the use of remote sensing to extract detailed spatial vegetation information. A series of well established ecological models will be used in the project, each of which will take forest structure at the appropriate scale to simulate terrestrial ecosystem carbon cycle. The project will take advantage of the rich data collected at Duke Forest by multiple past and on-going projects, including the permanent sample plots dating back to as early as the 1930s and data from the AmeriFlux and the FACE projects. The observed ground data will be used to provide model parameters as well as to evaluate the model simulations. In addition to quantifying the errors caused by omitting forest structures in simulating carbon cycle, the project will lead to major improvements to these models. The project will foster collaboration in teaching and research between the University of North Carolina at Chapel Hill and Duke University beyond the project lifetime. The project will improve research, teaching and advising abilities for the junior faculty leading the project, provide research opportunities for both graduate and undergraduate students, and enhance cross-campus student interactions. The expertise and research facilities from the two campuses are complementary. The project will strengthen the Geographic Information Science/Earth System Science programs at UNC Chapel Hill, and will also add a geographic dimension to the research projects at Duke Forest run by the School of Environment at Duke University.

Project Description 1 INTRODUCTION There is growing evidence that during the past decade the terrestrial biosphere of the Northern Hemisphere has acted as a carbon sink (Fan et al., 1998; Schimel et al., 2001; Hurtt et al., 2002). However, the magnitude, location, and mechanism responsible for the carbon uptake are highly speculative. Some studies suggest that the sink may be a result of forest regrowth on abandoned agricultural land and land management practices (Caspersen et al., 2000). Others argue that elevated CO2 concentration in the atmosphere, increased nitrogen deposition, and changes in regional precipitation patterns may lead to the sink (Fan et al. 1999; Schimel et al., 2000, Nemani et al., 2003). There is an urgent scientific and societal need for a solid understanding of sources and sinks of carbon on continental and regional scales and the development of these sources and sinks over time. Forest ecosystems hold the key to an improved understanding of terrestrial carbon cycle because forests contain up to 80 percent of terrestrial above ground carbon and 40 percent of below ground carbon (Dixon et al., 1994). Numerous studies indicate that a multi-dimension of forest structures, such as stand age, plays critical roles in the gaseous exchange between the forest ecosystems and the atmosphere on a stand scale (Yoder et al. 1994; Ryan et al. 1997; Law et al., 2000; Law et al., 2001; Klopatek, 2002; Katul et al. 2003; Litvak et al. 2003). Yet the landscape heterogeneity of forest structure is largely ignored in ecological models simulating terrestrial carbon cycles over large areas. The “big-leaf” model is the fundamental strategy to estimate regional ecosystem carbon fluxes (Kull and Jarvis, 1995). Scaling up from stands to the landscape for a definitive characterization of carbon fluxes remains a major challenge in carbon cycle studies. We propose to expand the scope and depth of existing studies at Duke Forest by scaling up carbon fluxes from stand to landscape through integration of remote sensing, ecological modeling and ground observations. Our scaling up strategy differs from the traditional “big-leaf” model as we explicitly incorporate spatial vegetation heterogeneity into ecological models to simulate landscape carbon cycle. The in-situ measurements of carbon fluxes at Duke Forest provide data needed for model parameterization and error assessment. The overall goal of the research is in-depth understanding of the impacts of forest structure on estimating carbon fluxes. The major science questions we are seeking answers for include: How well can the various remotely sensed data help us characterize forest structure over the landscape? How does the importance of forest structure changes in estimating landscape carbon budget as the spatial and temporal scales change? What are the appropriate spatial and temporal scales in order to capture the impacts of vegetation spatial heterogeneity on carbon cycle? 2 SIGNIFICANCE Change in the atmospheric CO2 concentration is a major source of uncertainty in predicting future global climate change. Due to the complexity and the importance of global climate change, the US Global Change Research Program (USGCRP) was created through the Global Change Research Act of 1990. In the summer of 2001, the Climate Change Research Initiative (CCRI) was launched to enhance the ongoing USGCRP elements that would rapidly lead to critical decision support information. Understanding of the uncertainties of carbon sources and sinks in North America has been identified by the National Research Council as one of the three key areas for CCRI. The proposed research will lead to better understanding of the spatial variation of carbon sources and sinks in the forest ecosystems of US Southeast, thus directly contributes to CCRI research objectives.

1

Increasing carbon storage in the terrestrial ecosystem is a major strategy in mitigating the increase of CO2 in the atmosphere. The Kyoto Protocol of the United Nations Framework Convention on Climate Change set a target for developed countries to reduce greenhouse gas emission levels by at least 5 percent below 1990 levels in the commitment period from 2008 to 2012. Under the Kyoto Protocol, industrialized countries can use the carbon sinks from certain forests to meet their greenhouse gas emission reduction commitments. Results of this study shall help reduce the uncertainties in estimating terrestrial carbon budget, thus contribute to effective implementation of Kyoto Protocol. At present, a global network (FLUXNET) consisting of over 200 micrometeorological tower sites operates on long-term and continuous basis measuring the exchanges of carbon dioxide, water, and energy between terrestrial ecosystems and the atmosphere (Baldocchi et al. 2001). Though much has been learned about the factors controlling the gaseous exchange between vegetation and the atmosphere, we still cannot transfer the tower measurements into regional estimation. Duke Forest area is home to three tower sites, each representing a typical vegetation type in the US southeast. Successful up scaling of carbon cycle in this region has global significance on scaling up measurements at many other FLUXNET sites. In addition, this project will foster synergistic research and education collaborations between UNC Chapel Hill and Duke University beyond the lifetime of the project and offer research opportunities for graduate and undergraduate students. Both the expertise and research facilities at Duke University and UNC Chapel Hill are complementary. The advanced facilities at Duke Forest enabled scientists at Duke University to study the biophysical, biological and biogeochemical processes that control the mass and energy exchange between the forest ecosystems and the atmosphere at the cutting edge. The geographic perspective in understanding and modeling theses processes and the state-of-the-art digital technology offered by scientists in the Department of Geography at UNC Chapel Hill are the exact components needed to transfer what we have known at Duke Forest to a regional understanding. In addition, Dr. Song’s research, teaching and advising capabilities will be significantly enhanced through the collaboration with three well-established senor scientists. Dr. Song is currently teaching advanced remote sensing and will teach a new course of ecological modeling at UNC Chapel Hill, both of which will be integrated with the research. The project will strength the Geographic Information Science and the Earth System Science programs at UNC Chapel Hill. 3 OBJECTIVES The overall goal of this research is to improve understanding of the impacts of the multidimensional forest structures on terrestrial carbon cycle across spatial and temporal scales through the integrative use of remote sensing, ecological modeling and ground observations. Specifically our research has two major objectives: (1) to map landscape forest structures with remotely sensed data from multiple sources, including digital orthophotos, Ikonos/QuickBird, Landsat, MODIS/MISR. The unprecedented quality and quantity of remotely sensed data available today have high potential to map detailed spatial vegetation information. The forest structure information we plan to derive from remote

2

sensing includes tree size and density, stand age, leaf area index, composition of vegetation types and subpixel tree cover. (2) to quantify errors caused by omitting forest structure in estimating carbon cycle across spatial and temporal scales using ecological models. We plan to use RHESSys (Band et al. 1993; Tague and Band, 2003) and flux tower measurements to evaluate the impacts of tree size and density, RegCarb (Song and Woodcock, 2003a) to evaluate the effects of stand age, and a light-use-efficiency model to evaluate subpixel tree cover effects on carbon cycle. Fig. 1 illustrates the steps to achieve our objectives. A series of errors will be generated at different spatial and temporal scales from model cross comparisons and observations at stand (flux tower) and regional (radiosonde) scales. The observational component of this project will directly taken from the on-going projects at Duke Forest led by Drs. Oren and Katul.

DOQQ/Ikonos (≤ 1m)

Landsat (30m)

Tree size/density

Stand ages

RHESSys

RegCarb

NEP Model Parameters

MODIS/MISR (250m)

Errors

LAI/Tree Cover

LUE

NEP

Errors

Remote Sensing

Objective 1

Ecological Models

NEP

Errors

NEP

Errors

Errors

Objective 2

NEP

Flux Towers

Radiosonde Model Parameters

3

Observations

Fig. 1: Project flow chart. There are three components in this project: remote sensing, ecological modeling and observations. The integration of the three components leads to two objectives: (1) spatial information of forest structures, and (2) errors in carbon cycle caused by omitting forest structure across spatial and termporal scales.

4 STUDY AREA Duke Forest, which was established in 1931 by Duke University, is located in the Piedmont Plateau of North Carolina. Soils in this region are characterized by relatively coarse-textured surface soils (sandy loams to loams) with clay subsoils derived from sedimentary rocks, except the bottomlands where new alluvium is deposited frequently. The climate is mild with a long growing season (≥200 days). Annual precipitation averages 47 in, and it is in general evenly distributed throughout the year with a somewhat higher amount in the summer. Plant communities in the Duke Forest are typical in the US southeast where a large portion of the landscape is covered by forests (~50 percent). Most of the carbon accumulation in the US is believed to occur in this region (Goodale et al., 2002). The Blackwood Division of the Duke Forest (35o98'N, 79o8'W, elevation=163 m) is home to three AmeriFlux towers and a FACE site operated by Duke University (Fig. 2a). The three flux towers are continuously monitoring the gaseous exchange with the atmosphere for three vegetation types: a hardwood forest, a pine plantation, and an old field. Oosting (1942) systematically studied the plant communities in the Piedmont of North Carolina (Fig. 2b). The general course of plant succession in this region is that herbaceous plants dominates an abandoned old field for the first three years, then pines began to establish. Oak-hickory dominated hardwoods eventually replace pines to become the climax community. Topography, available soil moisture, soil chemical properties (e.g. PH value, nutrient availability) and light environment, and species ecophysiological characteristics all act together to determine the course of plant succession at a particular site (Oosting, 1942; Kozlowski, 1949; Keever, 1950; Bormann, 1953; Christensen, 1977). (b) (a)

PP HF OF

Fig. 2: (a) Duke forest landscape with old field, pine (NIGEC sponsored AmeriFlux site) and hardwood forests. OF - old field, PP – pine plantation, HF- hardwood forest. (b) The geographic extension of Piedmont of North Carolina and the location of the three AmeriFlux tower sites.

The spatial distribution of plant communities is closely related to the land-use history in this region. Much of the land area was under cultivation during the early days of settlement. Due to poor agricultural practice, the fertile topsoil in the fields soon was washed out, and the fields became 4

nonproductive. It was common practice to abandon the nonproductive fields and clear new land. Such agricultural practice has led to the spatial pattern of plant communities at present, an irregular patchwork of fields and forested areas of all sizes and ages. This project can benefit tremendously from a rich collection of long-term and recently collected data. Duke Forest has an extensive set of permanent sample plots, some of which date back to 1930s. Every individual in the sample plots with stem greater than 1 cm is measured for diameter at breast height (DBH) and height on a 5-year cycle. Prof. Robert Peet in the Department of Biology at UNC Chapel Hill is primarily responsible for the recent data collection for permanent sample plots. These data are available to this study upon request (Dr. R. K. Peet, personal communication). Data collected at the AmeriFlux tower sites and the FACE sites are standard to all AmeriFlux and FACE sites (Ellsworth, 1999; Oren et al., 2001; Katul et al. 2001; Lai et al., 2002). The observed data can provide key parameters for model simulation as well as error assessment. 5 Task 1: Mapping Multi-dimensional Forest Structures with Remote Sensing Task 1.1 Mapping Tree Size and Density with High Resolution Remotely Sensed Data Tree size and density are the primary canopy structure on the stand scale. Tree size and density provide knowledge on how leaves are clumped in the canopy. Most ecological models use Beer’s law to predict Photosynthetically Active Radiation (PAR) transmission through forest canopies. It is the gaps created by leaves clumping into crowns that make forest canopies violate the assumption for Beer’s law: uniform spatial distribution of leaves. With tree size and density in a stand, we can quantitatively characterize both the between- and within-crown gaps, which are critical to predicting transmission of solar radiation through forest canopies (Kucharik et al. 1999). Tree size and density also determine the surface roughness, which influences the dynamic exchange of mass and energy between the atmosphere and the canopy. Given species, we can also derive LAI and the initial carbon content in a stand from tree size and density based on allometry. Canopy LAI and initial carbon content are required as initial conditions for carbon cycle simulation by many models. Though important, the primary means to obtain tree size and density to date is through field measurement, making it impossible to obtain such information over an extensive landscape. Here we plan to develop an operational algorithm to map tree size and density using high resolution remotely sensed data based on the theory by Song and Woodcock. (2002) Numerous studies found that the spatial characteristics of remotely sensed imagery contain the information of the structure of the underlying scene. Woodcock et al. (1988a,b) found that the height of the sill of the semivariogram is related to the density of the objects. The range of the semivariogram is related to the size of the objects, and the shape of the semivariogram is related to the variance of the size of the objects. Cohen et al. (1990) showed that semivariograms can provide useful information for stand structures, but the amount of information is highly dependent on image spatial resolution. St-Onge and Cavayas (1995, 1997) further demonstrated that good empirical relationships can be established between the spatial properties of high-resolution images and tree size and density. However, attempts using Landsat imagery to map tree size and density have generally failed, as there is no spectral signature consistently related to them and the spatial resolution of Landsat is insufficient to make extensive use of spatial signatures (Cohen et al., 1990; Woodcock et al., 1997). The effect of pixel size is critical as the sill of a semivariogram decreases with increase in pixel size. Recently Song and Woodcock (2002) identified an additional link between the image spatial properties and tree size and density, i.e. the rate of change in the sill of

5

the semivariogram with spatial resolution is diagnostic of tree size and density. Due to spatial averaging, the sills of the semivariograms decrease as the pixel size increases. The sill of the semivariogram decreases much more slowly for images containing larger trees than that with smaller trees. Based on this finding, an analytical model that estimates tree size and density from the sills of multiresolution imagery is developed (Song and Woodcock, 2003b). The spatial model by Song and Woodcock (2003b) has been tested in the Western Cascades of Oregon where trees are predominantly coniferous. The model has not been applied to broadleaf and mixed forests. We believe that the spatial theory is applicable to broadleaf and mixed forests. We plan to further test the theory and operationalize the algorithm to map tree size and density for the forests in the Piedmont of North Carolina. Such an algorithm will revolutionize the use of highresolution imagery to obtain forest structure over large areas. The State of North Carolina collected high-resolution digital orthophotos twice for the state, once in 1993 as B&W and once in 1998 as color infrared. We will further test the algorithm against the permanent sample plots in Duke Forest. We propose to use 1993 and 1998 digital orthophotos, and a recent Ikonos or QuickBird image to map tree size and density. Recently two major disturbances caused wide spread damage to the canopy: one was Hurricane Fran in 1997, and the other was the ice storm in 2002. The three high-resolution images in sequence are suited to test further whether our algorithm can capture the extent and severity of these disturbances. Similar efforts to map canopy structure using Lidar technology were planned for this region. Though Lidar can provide the vertical profile of canopy structure, it is very expensive to obtain and data are not widely available at present. There are numerous sources of high-resolution optical imagery, such as QuickBird, Ikonos and high-resolution digital orthophotos, that are much easier to obtain. Our products for canopy structure can well serve as a benchmark for the more expensive technology, and provide fundamental substrate information for future North America Carbon Program operations. Task 1.2 Mapping Forest Successional Stages with Multi-temporal Landsat Imagery Forest succession is closely related to tree size and density, but it also incorporates the individual replacement process. Stands at different forest successional stages are usually composed of trees at different ages, and sometimes different species composition. Though the mechanisms are debatable (Yoder et al., 1994; Ryan and Yoder 1997), the fact that forest productivity strongly depends on its successional stage is well accepted as it is confirmed in numerous studies from different perspectives (Birdsey et al., 1993; Law et al., 1999; Barford et al., 2001; Law et al., 2000). Therefore, knowledge of forest successional stages over space should lead to improved estimation of carbon budget over the same area. Due to its unique land-use history, the landscape of the study area is composed of vegetation in diverse successional stages. In this project, we propose to map forest successional stages in the Duke Forest and its surrounding areas using multitemporal Landsat images based on a recent theory by Song et al. (2002). Numerous studies have shown the potential of mapping forest successional stages with remote sensing. Past studies are primarily focused on separating stands into a few discrete stages (Hall et al., 1991; Fiorella and Ripple, 1993a,b; Cohen et al., 1995; Jakubauskas, 1996). Though mapping the discrete succession stages with remote sensing is relatively successful, mapping the continuous successional stages, i.e. stand ages, is much more challenging. Mapping stand ages with remote

6

sensing is mostly limited to young stands (Kimes et al., 1996; Helmer et al., 2000). More recently, Cohen et al. (2001) mapped forest ages in Western Oregon with an empirical model based on wetness and brightness indices of the Tasseled Cap Transformation from a single date Landsat image. The model had an R-square of 0.51 and a RMSE of 107 years, indicating further improvement of mapping stand age is needed. A common characteristic in previous efforts of mapping forest succession is that they are generally based on empirical linear analysis. The spectral/temporal history associated with forest succession has not been fully utilized to map forest succession. Recently, Song et al. (2002) revealed that forest successional trajectories in the spectral space are highly nonlinear, and the assumption of linear relationship between remote sensing signals and forest succession is only valid for early stands under certain conditions. To account for the nonlinearity of forest succession in the spectral space, multiple points are needed to characterize the status of forest succession. In collaboration with Dr. W. B. Cohen at USFS Pacific Northwest Research Station, Dr. Song is working on extracting stand ages in the Pacific Northwest using multitemporal Landsat TM/ETM+ images and the Forest Inventory Analysis (FIA) data. Initial results from over a thousand stands with known ages show that multitemporal images significantly enhance our ability to predict stand ages (work in preparation). In this study, we propose to expand the approach Song et al. (2002) developed in the Pacific Northwest to North Carolina Piedmont, mapping forest successional stages using multitemporal Landsat images. While the plant communities in the US Southeast are dramatically different from those in the Pacific Northwest, we believe the theoretical basis still applies, i.e. multitemporal Landsat imagery enhances our ability to monitor forest successional stages. The difference in the species composition in the US Southeast may generate dramatically different patterns of successional trajectory. We plan to first reparameterize the GORT-ZELIG model developed by Song et al. (2002) for the forests in North Carolina Piedmont to understand the spectral/temporal signatures associated with forest succession in this region. Successional trajectories for stands in the Pacific Northwest are driven primarily by changes in canopy structure. We believe successional trajectories for stands in North Carolina Piedmont to be driven by both change in canopy structure and change in vegetation types, i.e. hardwoods vs. conifers. We plan to modify GORT model so that it can simulate canopy BRDF for mixed forests in the US Southeast. We expect the successional trajectory for the typical successional course in North Carolina Piedmont will be very different from those in the Pacific Northwest (Song et al., 2002), and we also expect that the successional trajectory will be also be highly nonlinear. Therefore, we need multiple images to determine the stand ages in the region. To map stand ages over a large area, we propose to use the recently developed neural networks, ARTMAP (Liu, 2001), to map the stand ages. The advantage of using ARTMAP is that it is learning based without normality assumption required by traditional statistical techniques. We propose to train ARTMAP with successional trajectories generated by GORT-ZELIG. The trained ARTMAP will be used to map stand ages using multiple Landsat images through time. As the course of succession trajectory in the Piedmont of North Carolina strongly depends on a number of environmental conditions, such as topographic position, soil moisture conditions, we will adopt a stratified approach to map stand ages accordingly. In contrast to traditional empirical approach, our approach is built on mechanistic models driven by ecological processes and physics of radiation transfer through plant canopies. If successfully implemented, this effort would lead to the first operational approach to map stand ages over large areas. Landsat program has been operating for three decades now, and multitemporal Landsat images are archived for almost anywhere around the globe. Successful mapping of stand ages in the US Southeast

7

holds the potential to further clarify the mechanism responsible for the increased carbon uptake in the conterminous USA. Several studies suggest that the regrowth of young forest on abandoned agriculture land is responsible for the carbon sink (Caspersen et al., 2000; Hurtt et al., 2002). This effort can help us find where these forests are located and how much area these forests occupy. Task 1.3 Mapping LAI and Subpixel Tree Cover with MODIS and MISR Data A. Mapping Leaf Area Index (LAI): Leaves are the primary interface for gas, water and energy exchanges between the terrestrial ecosystem and the boundary layer atmosphere. Leaf area is perhaps the most important canopy structural parameter for the terrestrial ecosystem. Leaf area in a vegetation canopy is measured by Leaf Area Index (LAI), which is defined as the one-sided leaf area divided by the ground area the canopy occupies. Scientists have been working to derive LAI from remotely sensed data for many years (Tucker and Sellers, 1986; Peterson et al. 1986; Spanner et al., 1990; Chen and Cihlar, 1996; Tian et al., 2003). Numerous previous studies have found that remotely sensed signals are sensitive to LAI variation only at low LAI values, and often saturate at high LAI. One of the primary products from MODIS/MISR missions is global LAI. The algorithm is based on the theory of radiative transfer (Knyazikhin et al., 1998; Myneni et al., 2002). Vegetation heterogeneity in the landscape remains a problem to the effort of mapping global LAI where each 1×1km footprint is represented by one and only one of six biome vegetation types. Initial validation results by the BigFoot project (Cohen et al., 2001) indicate that the global LAI product from MODIS may be seriously biased (Cohen 2003, personal communication). By taking advantage of the high resolution remote sensing products generated earlier in the projects, we propose a simple new approach to derive landscape LAI from remotely sensed data collected by the MODIS sensors on board both Terra and Aqua spacecrafts, and multi-directional MISR sensors. Our approach will estimate LAI from the Spectral Vegetation Indices (SVI) by combining the morning and the afternoon MODIS data, and the 9 directional MISR data. High LAI stands are usually associated with larger trees that cast larger and darker shadows on the ground or neighboring tree crowns. We believe that forest canopy BRDF captured by MODIS morning and afternoon data with the sun at different positions and MISR data from 9 viewing directions should improve mapping LAI. To help facilitate mapping of LAI using MODIS/MISR data, we first produce an LAI surface using the high-resolution images based on tree size and density based on allometry. The high resolution LAI surface will be embedded in the MODIS/MISR 250×250 meter Spectral Vegetation Index (SVI) surfaces to establish an empirical relationship of LAI with MODIS/MISR SVI images. Then we will generate an LAI map from the empirical relationship for the Duke Forest and its surrounding areas and compare our products with MODIS LAI product. We expect that our approach will generate a more accurate LAI map within the local areas. B. Mapping Subpixel Vegetation Cover: Vegetation cover is a key land surface variable in terrestrial ecosystem processes. Vegetation mapping at global/continental scales with remote sensing can only be achieved with pixel size on the order of several kilometers. There are few, if any, pixels covered entirely by forests at these scales. Due to the nonlinear nature of ecosystem processes, it is essential to derive the subpixel tree cover information with coarse resolution remotely sensed data in order to properly estimate terrestrial ecosystem carbon budgets at these scales. We propose to map subpixel tree cover through spatial analysis using MODIS data. Our spatial analysis indicates that the ability of mapping object size and density using remotely sensed

8

data depends on the ratio of pixel size to the object size. The spatial signatures are most sensitive to object size and cover when the ratio approaches unity. With 250×250m MODIS data, we cannot obtain tree size and density as we did with the high-resolution imagery, but we expect we can obtain the patch size and subpixel tree cover. We propose to extend our spatial analysis from highresolution imagery to coarse-resolution images for subpixel tree cover information. Traditionally there are three approaches available for deriving subpixel vegetation information from coarse resolution data, such as data from AVHRR (DeFries et al., 1997). One approach is linear mixture modeling, which assumes that the remotely sensed signal for a pixel is a linear combination of a few cover types (end members). The relative fraction for each cover types is calculated based on the spectral signature of a pure pixel for each cover type. DeFries et al. (2000) used a linear mixture model applied to multi-year 8×8 km AVHRR data to estimate global vegetation cover with three end members, woody vegetation, herbaceous vegetation, and bare ground. The second approach for deriving subpixel vegetation cover is based on adjusting the classification of AVHRR data with classification of fine-resolution images. The third approach used to determine subpixel vegetation cover involves two steps: (1) the development of an empirical relationship between vegetation cover and AVHRR data based on some “calibration centers” where vegetation cover is derived from fine resolution images, and (2) the extrapolation of the empirical relationship between vegetation cover and AVHRR data over large areas. The linear mixture modeling assumes constant end member signatures over space. In fact, the spectral signature of end members may vary from pixel to pixel. The second and third approaches have similar problems due to the fact that the fine resolution images often cannot match in time with the coarse resolution images to map subpixel tree cover. The approach we propose here is unique as it only depends on spatial structure of the underlying scene, the patch size and cover, not the spectral signatures. Thus our approach does not suffer from the problems caused by end-member signature variation from pixel to pixel. To further reduce the potential uncertainty of modeling carbon assimilation by forests, we will continue to use the 250×250m MODIS image to separate the subpixel vegetation cover into conifer and deciduous forests for the conifer and deciduous forests have quite different light-use efficiencies that could cause relatively large errors in carbon cycle simulation (Turner et al., 2002). Wintertime Landsat ETM+ imagery will be used to derive green vegetation cover information based on NDVI. The wintertime green vegetation cover is assumed to be conifer only in this region. Then the Landsat image will be imbedded in MODIS images to establish a relationship between the percentage of annual NDVI change with percentage of deciduous forests in a pixel. 6 Task 2: Simulating Landscape Carbon Cycle There are dozens well established models simulating terrestrial ecosystem carbon cycle at different spatial and temporal scales. However, we still cannot account definitively and quantitatively for the net ecosystem exchange of carbon with the atmosphere over the landscape (Schimel et al., 2000). These models remain as research tools, which often focus on gaining better understanding for certain aspects of the whole problem. Therefore, it is not surprising that independent estimations of global/continental scale carbon budget do not agree with each other (e.g. top-down vs. bottom-up approaches). For example, Schimel et al. (2000) found that the carbon sink for the terrestrial ecosystem of the conterminous USA is three times larger based on inventory data as compared to model simulation. The most consistent conclusion over global carbon cycle to date based on

9

different approaches is the finding of a significant terrestrial carbon sink in the Northern Hemisphere during the 1990s, while the land biosphere was approximately neutral in the 1980s (Fan et al., 1999; Schimel et al. 2000; Myneni et al. 2001; Pacala et al., 2001). However, it is highly debatable with regard to whether the carbon sink is located in North America or in Eurasia, how big the carbon sink is and how the extra carbon was uptaken. Schimel et al. (2001) estimate that the terrestrial carbon sink in the 1990s in the Northern Hemisphere is roughly split between North America and Eurasia, while Myneni et al. (2001) estimates that nearly 70 percent of the carbon sink in the Northern Hemisphere occurred in Eurasia. Some believe that CO2 fertilization and nitrogen deposition are responsible for the carbon uptake in the Northern Hemisphere (Erickson, 1999; Kicklighter et al., 1999). Recent studies suggest that forest regrowth on abandoned farmlands is primarily responsible for the carbon uptake on the conterminous USA (Houghton et al., 1999; Caspersen et al., 2000; Hurtt et al., 2002; Goodale et al., 2002). Nemani et al. (2003) suggested that climatic change, which enhanced plant growth in the northern mid- and high-latitudes, may be the cause. We believe one of the major factors that lead to the uncertainty in balancing the carbon cycle is the lack of representation of spatial heterogeneity of vegetation in ecological models. The spatial heterogeneity of forest ecosystems is manifested as changes in forest structure over the landscape, such as tree size and density, LAI, stand ages etc. The second task of the project is to quantify the errors in characterizing carbon cycle caused by omitting forest structures in the landscape. We propose to integrate the remotely sensed information of forest structure and ground observations into ecological models that explicitly account for the impacts of forest structure. The following section elaborates on our plans to investigate the impacts of tree size and density, stand age and subpixel tree cover on scaling up carbon fluxes. Task 2.1 Modeling Tree Size and Density Effects on Landscape Carbon Cycle We propose to use the RHESSys model (Running et al., 1989; Band et al. 1991, Band, 1993; Band et al., 2001; Tague and Band, 2003) to study the impacts of tree size and density on the landscape carbon cycle. RHESSys is a GIS-based hydro-ecological modeling framework that combines a set of physically based process models and a methodology for partitioning and parameterizing the landscape to simulate carbon, water and nitrogen fluxes. RHESSys architecture simulates the spatial distribution and spatial-temporal interactions between the different processes at the watershed scale. Carbon uptake from atmosphere in RHESSys is simulated based on Biome-BGC, which uses the Farquhar model to simulate photosynthesis. Nitrogen cycle in RHESSys is simulated based on the Century approach (Parton et al. 1993). Soil moisture redistribution and runoff production is based on either the TOPMODEL (Beven and Kirkby, 1979) or an explicit routing model. The hierarchical representation of the landscape in RHESSys allows different processes to be modeled at different scales within the model and allows basic modeling units to be of arbitrary shape (Fig. 3). We propose to superimpose the tree size and density as separate GIS layers on top of the landscape represented in RHESSys so that each spatial unit simulated by RHESSys has unique tree size and density information. The advantage of using RHESSys is that this is a well-established model, and many other hydro-ecological processes that are critical to carbon cycle have already been accounted for. To use RHESSys to study the impacts of tree size and density on carbon cycle over a landscape, we can focus on the problem without other unnecessarily distractions.

10

GIS Input Layers DEM LAI Land-cover Soil Streams

Landscape Representation

Canopy

Meteorology

Spatial Output

Hydrology

Temporal Output RHESSys Simulator Fig. 3: The model structure of RHESSys. The spatial heterogeneity of the landscape is represented by hierarchical spatial units. The model fully couples canopy, hydrologic and meteorological processes and produce both spatial and temporal outputs

Tree size and density in a stand first directly influence the amount of photosynthetically active radiation that a canopy absorbs (APAR). APAR is the total amount of energy that vegetation can use for photosynthesis. In the current version of RHESSys (v5.8), APAR is estimated based on the Beer’s law with a biome-specific constant gap fraction in the canopy. In fact, leaves in the forest canopy are highly clumped, and light interception simulated with Beer’s law can be highly over estimated (Liu et al., 1997). With the tree size and density superimposed over the landscape as additional GIS layers in the RHESSys model, we can predict much better the amount of APAR in forest canopies with a more advanced canopy radiative transfer model, MVP, which was recently developed by Song and Band (2003). MVP assumes leaves in a canopy clump into crowns, forming two types of gaps: the between- and within-crown gaps. There is no attenuation of light when passing through the gaps. Canopy gaps and the proportion of sunlit and shaded leaves are simulated as a function of sun angle. Validation of measured light under forest canopy with multiple PAR sensors indicates that MVP can capture light penetration through the canopies effectively (Song and Band, 2003). We believe improved prediction of APAR in forest canopies will further capture the spatial variation of carbon uptake in the forest ecosystem due to changes in tree size and density. Though carbon models using advanced canopy radiative transfer approach to simulate APAR exist in the literature, such as Maestro (Wang and Jarvis, 1990), they cannot be used for an extensive landscape as large as a watershed. Carbon cycle models based on simulation of photosynthesis usually face the problems of providing parameters for many other related processes that usually cannot be obtained in a single project, such as autotrophic respiration and below ground carbon allocation. At the Blackwood Division of Duke Forest, many interrelated projects are active simultaneously, each of which focuses on a different aspects relating to terrestrial carbon cycle (Ellsworth, 1999; Oren et al., 2001; Katul et al. 2001; Lai et al., 2002). All components of carbon cycle for the three vegetation types are being monitored continuously, making the site as an ideal location for process-based simulation of carbon

11

cycle. The observations at the tower sites, such as dark respiration, soil respiration and below ground carbon allocation, will be used to parameterize the model. The model will be calibrated with Flux tower measurements of Net Ecosystem Production (NEP) for the stands. After it is calibrated, the model will be run for the whole area covered by the high-resolution remotely sensed data (approximately 16×16 km) with the Blackwood Division of Duke Forest at its center. This will directly provide us with the error in carbon cycle caused by omitting the spatial variation in tree size and density. Task 2.2 Modeling Stand Age Effects on Landscape Carbon Cycle Due to the land-use history, the study area is covered by forests at diverse successional stages, each of which may sequester carbon at a rate different from others. We propose to continue to build on the RegCarb model (Song and Woodcock, 2003a) to study the effects of stand ages on landscape carbon cycle. RegCarb is a regional carbon cycle model developed especially for forest ecosystems. Most simulations of carbon cycle over large areas are based on regular grid with its size on the order of several kilometers. The heterogeneity in forest successional stages within the grid is ignored in these models. RegCarb is unique compared to most other models as RegCarb explicitly account for the effects of stand ages on regional carbon cycle. The core of RegCarb is an individual-based model (IntCarb) to simulate carbon cycle on a stand basis. IntCarb is developed from the ZELIG model (Urban, 1990), which in turn was based on the FORET model and developed as a generalized simulator that could be modified and extended easily for new applications. IntCarb estimates Net Primary Production (NPP) as the sum of increase in biomass and litter fall for all individuals in a stand. Decomposition of the dead biomass from litter fall and mortality is simulated based on the framework of the Century model (Parton et al, 1993). The IntCarb model is then imbedded in RegCarb to estimate the regional carbon budget as a result of land-use change and forest succession for the 10 million hectares of commercial forests in the Pacific Northwest of Oregon and Washington. Simulation results by IntCarb for a typical stand in the H. J. Andrews Experimental Forest compared well with observed data based on the permanent plots (Song and Woodcock, 2003a). Individual-based modeling of carbon cycle offers significant advantages over the conventional “big-leaf” approach. First, population dynamics of a stand through time can only be simulated through the use of individual-based models (Huston and Smith, 1987; Tilman, 1988; Peet, 1992). Shifts in species composition through time due to succession are associated with major changes in ecosystem productivity. For example in the US Southeast early successional stands like Pinus taeda (loblolly pine) have NPP as much as 50 percent higher than those of the climax oak stands (Peet, 1981). Second, the change of canopy structure through time can be readily derived from individual-based simulation. Many process-based model cannot account for the transient status of the terrestrial ecosystems, but only for equilibrium ecosystems. The individual-based simulation can be used to capture disturbances that cause significant damage to canopy structure. In addition, an individual-based model is suitable to a mixture of species with significant different ecophysiological properties, which is characteristic for the forests in the US Southeast. In this project, we plan to reparameterize RegCarb for the study area, and provide the stand age information mapped with the multitemporal Landsat imagery to RegCarb for the study area covered by high-resolution remotely sensed data. Water and nutrient availability in the soils is equally important to plant growth in this region (Oosting, 1943; Kozlowski, 1949). To account for

12

the spatial variation in plant growth due to topographic and edaphic differences, we will adopt a stratified simulation for IntCarb to account for these effects. Task 2.3 Modeling Subpixel Tree Cover Effects on Landscape Carbon Cycle Much emphasis is given to LAI in simulating terrestrial carbon cycle due to the critical role that leaves play in gaseous exchange with the atmosphere. Much less emphasis is given to the tree cover, which may also exert significant influence on carbon assimilation, particularly for simulations with a large footprint. For a give LAI within a footprint, how the leaves are distributed within the footprint not only influence the amount of solar radiation that the leaves absorb, but also influence how much direct solar radiation the leaves absorb. We propose to integrate the subpixel tree cover information together with LAI into an ecological model to understand its impact in estimating carbon cycle over large area. We propose to a model that is based on the light-useefficiency (LUE) theory, a different modeling paradigm that recently received a lot of attention in carbon cycle studies. The algorithm that the MODLAND group used to estimate global gross and net primary production using data from MODIS is based on the LUE theory (Running et al., 2000), but the subpixel tree cover is not accounted for. Simulation of carbon cycle over large areas is significantly simplified based on LUE models as the maximum LUE is a conservative parameter for plants within a biome. In a thorough literature review, Goetz and Prince (1999) found theory and available evidence support the convergence of maximum LUE for Gross Primary Production (GPP). In addition to its simplicity, the direct linkage between remotely sensed data and LUE models makes this type of models even more attractive. Remotely sensed data can be used to estimate the fraction of photosynthetically active radiation absorbed (fAPAR) by plant canopies (Sellers, 1987; Prince and Gower, 1995). For example, CASA (Potter et al., 1993), GLO-PEM (Prince and Gower, 1995), and 3-PGS (Coops et al., 1998) all use a simple empirical linear relationship to predict fAPAR from Normalized Difference Vegetation Index (NDVI) in the models. However, subpixel tree cover within the large footprint (>1×1km) is not accounted for in estimating fAPAR. In this project we plan to quantify the errors caused by omitting subpixel tree cover in estimating carbon cycle. The essential effect of tree cover is its impacts on the amount of PAR absorbed by the canopy. At a given LAI, the estimated fAPAR with subpixel tree cover is f APAR = Q × exp(−kΩL′ / cosθ ) (1) while fAPAR without subpxiel tree cover is f APAR = exp(−kΩL / cosθ ) (2) where Q is subpixel tree cover, and L′ is the adjusted LAI (L′=L/Q). Ω is leaf clumping factor in the canopy, and k is light extinction coefficient. The solar zenith angle is θ. Obviously, Eqs. (1) and (2) are very different. We will use Eq. (1) to estimate fAPAR in our LEU model. A secondary subpixel tree cover effect is the proportion of deciduous and needleleaf forests in a footprint. We will use the flux tower measurements to obtain the LUE for the conifer and the hardwoods as they have very different LUE. We propose to estimate a weighted average LUE for a pixel based on the proportion of deciduous and needleleaf forests using data from the flux tower measurements at Duke Forest. Due to its large pixel size, we cannot compare the tower measurements with simulated fluxes using MODIS/MISR data. We will use flux data derived from radiosonde measurement from a separate project lead by Drs. Oren and Katul to evaluate the simulation, and we will use the MODIS GPP and NPP products to evaluate the impacts of subpixel tree cover on carbon simulation. 13

The LUE model can simulate a much larger area than RHESSys and RegCard can do, as the footprint of the remotely sensed data is much larger. There are two sources of data we will use to evaluate errors caused by omitting subpixel tree cover in the LUE model. One is model simulation at higher spatial resolution from RHESSys and RegCarb. RHESSys can produce both spatial and temporal simulation results at different scale, making it very flexible for model comparison. The other is regional carbon flux observation from radiosonde in a separate project led by Drs. Oren and Katul. Overall, the inter-model comparison and comparison of model simulation from flux tower and radiosonde observations can substantially enhance our understanding of the impacts of forest structure on scaling up carbon fluxes from stands to lsandscape. 7 EXPECTED RESULTS

•

•

• •

• •

• •

Maps of tree size and density will be produced for the study area on a stand basis from high-resolution remotely sensed imagery, and an operational algorithm that derives the information. The algorithm to map tree size and density will enhance the use of highresolution remote sensing for forest inventory analysis, assessment of damage from disturbances to the forest, such as hurricanes and ice storms. The tree size density map can also provide surface roughness information to inverse analysis. The project will produce a stand age map for the study area using multitemporal Landsat imagery. Stand age is directly related to the land-use history, which is critical to assess the long-term potential of carbon assimilation by the terrestrial ecosystem. The operational algorithm developed to map stand age can be used to narrow down the locations of carbon sinks in forest ecosystems. LAI and subpixel tree cover map will be produced from MODIS/MISR data. These products can be used as input to carbon cycle models, and be used as an independent validation to the existing MODIS products as well. We will produce a new version of RHESSys model that explicitly accounts for the spatial heterogeneity of tree size and density. The current version of RHESSys simulates photosynthesis based on the “big-leaf” theory. The new version of the model will work with a gappy canopy consisting of individual tree crowns. We will produce a new version of RegCarb model that can be used for the US Southeast. RegCarb is a relatively new type of carbon cycle model as it explicitly accounts for effects of stand ages in estimating regional carbon budget. The most interesting products of this project are the errors in carbon cycle caused by omitting a hierarchy of forest structures, including tree size and density, stand ages, and subpixel tree cover at coarse spatial resolution. Lack of uncertainty information is a major weakness in studies characterizing terrestrial carbon cycle over large areas. We expect one graduate student to finish his/her Ph.D. degree through this project. The project will provide research opportunities for other graduate and undergraduate students through the educational program. We expect multiple peer-reviewed joint publications will appear on professional journals through this project in addition to conference presentations and proceedings.

14

8 Management Plan and Timeline

Dr. Song will be responsible for the overall project management and progress. He is also responsible for the remote sensing of forest structure and the RegCarb model development. Dr. Band will be primarily responsible for proper use and improvement of the HESSys model as he is a primary author of the model. Drs. Oren and Katul help with providing data from AmeriFlux and FACE projects and advising correct use of the data. In addition, the project will recruit a Ph.D. student with Dr. Song as the primary advisor. Drs. Band, Oren and Katul will serve in the doctoral committee for the student, co-advising the student with Dr. Song. The proposed project will require 3 years to complete. The following table provides a general project progress schedule in quarters. Table 1: Project progress timeline. 1st Quarter 2nd Quarter Year one

3rd Quarter 4th Quarter 1st Quarter 2nd Quarter

Year two

3rd Quarter 4th Quarter 1st Quarter 2nd Quarter

Year three

3rd Quarter 4th Quarter

Data acquisition and preprocessing, including geometric correction and radiometric calibration, to build a GIS database Finish preprocessing and GIS database development, test running RHESSys for the Blackwood Division of Duke Forest Start developing spatial algorithm to map tree size and density with high-resolution remotely sensed data. Validation of tree size and density derived from 1993 B&W DOQQ, 1998 DOQQ and recent Ikonos/QuickBird data Modify RHESSys to take tree size and density over the landscape to simulate carbon cycle. Calibrate RHESSys with flux tower measurement and error analysis. Start transferring GORT-ZELIG to the local area Continue work on model simulation with GORT-ZELIG and develop algorithm to map stand ages with ARTMAP. Validating stand age map and integrating stand age with RegCarb to simulate carbon cycle Error analysis for age effect on carbon cycle; mapping subpixel tree cover and LAI with MODIS/MISR. Evaluate subpixel LAI and tree cover products; test running a light-use-efficiency model Error analysis for subpixel LAI and tree cover with Radiosonde measurements and other simulations. Publication preparation, and project report.

15

REFERENCES 1. Baldocchi et al. 2001. FLUXNET: a new tool to study the temporal and spatial variability of ecosystem-scale carbon dioxide, water vapor, and energy flux densities. Bul. Am. Meteor. Soc. 82:2415-2434. 2. Band et al. 1991. Forest ecosystem processes at the watershed scale: basis for distributed simulation. Ecol Mod., 56:151-176. 3. Band, L. E. 1993. Effect of land surface representation on forest water and carbon budgets. J. Hydrol. 150:749-772. 4. Band et al. 2001. Forest ecosystem processes at the watershed scale: Hydrological and ecological controls of nitrogen export. Hydrol. Proc. 15:2013-2028. 5. Barford et al., 2001. Factors controlling long- and short-term sequestration of atmospheric CO2 in a mid-latitude forest. Science. 294: 1688-1691. 6. Beven, K. J. and Kirby, M. J. 1979. A physically based, variable contributing model of basin hydrology. Hydrol. Sci. Bull. 24:43-69. 7. Birdsey et al., 1993. Past and prospective carbon storage in United States Forests. For. Ecol. Manag. 58:33-40. 8. Bormann, F. H. 1950. Factors determining the role of Loblolly Pine and Sweetgum in early old field succession in the Piedmont of North Carolina. Ecol. Monogr. 23:339-358. 9. Caspersen, J. P. et al. 2000. Contributions of land-use history to carbon accumulation in U.S. forests. Sci., 290:1148-1151. 10. Chen, J.M. and Cihlar, J. 1996. Retrieving leaf area index of boreal conifer forests using Landsat TM images. Rem. Sens. Environ. 55:153-162. 11. Christensen, N. L. 1977. Changes in structure, pattern and diversity associated with climax forest maturation in Piedmont, North Carolina. Am. Midl. Nat. 97:178-188. 12. Cohen et al., 1990. Semivariograms of digital imagery for analysis of conifer canopy structure. Rem. Sens. Environ. 34:167-178. 13. Cohen et al., 1995. Estimating the age and structure of forests in a multi-ownership landscape of western Oregon, USA. Int. J. Rem. Sens. 16:721-746. 14. Cohen et al. 2001. Modelling forest cover attributes as continuous variables in a regional context with Thematic Mapper data. Int. J. Rem. Sens. 22:279-2310. 15. Coops, N. C. et al. 1998. Assessing forest productivity in Australia and New Zealand using a physiologically-based model driven with averaged monthly weather data and satellitederived estimates of canopy photosynthetic capacity. For. Ecol. & Manag., 104:113-127. 16. DeFries, R, et al. 1997. Subpixel forest cover in central Africa from multisensor, multitemporal data. Rem. Sens. of Environ., 60:228-246. 17. Defries, R. et al. 2000. Global continuous fields of vegetation characteristics: a linear mixture model applied to multi-year 8 km AVHRR data. Int. J. of Rem. Sens., 21:13891414. 18. Dixon, R. K. et al. 1994. Carbon pools and flux of global forest ecosystems. Sci., 263:185190. 19. Dubayah, r. O. and Drake J. B. 2000. Lidar remote sensing for forestry. Journal of Forestry. 98:44-46. 20. Ellsworth, D. S. 2000. Seasonal CO2 assimilation and stomatal limitations in a Pinus taeda canopy with varying climate. Tree Physiol. 20:425-435.

21. Erickson, D. J. 1999. Nitrogen deposition, terrestrial carbon uptake and changes in the seasonal cycle of atmospheric CO2. Geophy. Res. Let. 26:3313-3316. 22. Fan, et al. 1998. A large terrestrial carbon sink in North America implied by atmospheric and oceanic carbon dioxide data and models. Sci., 282:442-446. 23. Fan et al. 1999. Terrestrial carbon sink in the Northern Hemisphere estimated from the atmospheric CO2 difference between Manna Loa and the South Pole since 1959. Tellus B, 51:863-870. 24. Fiorella, M. and Ripple, W. J. 1993a. Determing successional stage of temperate coniferous forest with Landsat satellite data. Photo. Eng. Rem. Sens. 59:239-246. 25. Fiorella, M. and Ripple, W. J. 1993b. Analysis of conifer foorest regeneration using Landsat Thematic Mapper data. Photo. Eng. Rem. Sens. 59:1383-1388. 26. Goetz, S. J. and Prince, S. D. 1999, Modeling terrestrial carbon exchange and storage: the evidence for and implications of functional convergence in lightuse efficiency. Adv. Ecol. Res., 28:57-92. 27. Goodale, et al. 2002. Forest carbon sinks in the northern hemisphere. Ecol. Appl. 12:891899. 28. Hall et al. 1991. Large-scale patterns of forest succession as determined by remote sensing. Ecol. 72:628-640. 29. Helmer et al. 2000. Mapping montane tropical foest successional stage and lans use with multi-date Landsat imagery. Int. J. Rem. Sens. 21:2163-2183. 30. Houghton et al. 1999. The US carbon budget: Contributions from land-use change. Science. 285:574-578. 31. Hurtt, G. C., et al. 2002. Prohecting the future of the US carbon sink. Proc. of Nat. Acad. Sci. USA, 99:1389-1394. 32. Huston, M. and Smith, T. 1987. Plant Succession: life history and competition. Amer. Nat. 130:168-98. 33. Jakubauskas, M. E. 1996. Thematic Mapper characterization of lodgepole pine seral stages in Yellowstone National Park, USA. Rem. Sens. Enviorn. 56:118-132. 34. Katul et al. 2001. Quantifying the complexity in mapping energy inputs and hydrologic state variables into land-surface fluxes. Geophy. Res. Let. 28:3305-3307. 35. Katul et al. 2003. Relationship between plant hydrailic and biochemical properties derived from a steady-state coupled water and carbon transport model. Plant, Cell, Enviorn. 26:339350. 36. Keever, C. 1950. Causes of succession on old fields of the Piedomont, North Carolina. Ecol. Monogr. 20:229-250. 37. Kicklighter et al. 1999. A first-order analysis of the potential of CO2 fertilization to affect the global carbon budget: a comparison of four terrestrial biosphere models. Tellus B, 51:343-366. 38. Kimes et al. 1996. Extracting forest age in a pacifi northwest forest from Themtic Mapper and topgraphic data. Rem. Sens. Environ. 56:133-140. 39. Klopatek, JM. 2002. Belowground carbon pools and processes in different age stands of Douglas-fir. Tree Physiol. 22:197-204. 40. Knyazikhin et al. 1998. Estimation of vegetation canopy leaf area index and fraction of absorbed photosynthetically active radiation from atmosphere-corrected MISR data. J. Geophy. Res-Atmos. 103:32239-32256.

41. Kozlowski, T. T. 1949. Light and water in relation to growth and competion of piedmont forest tree species. Ecol. Monogr. 19:209-230. 42. Kucharik et al. 1999. Charaterization of radiation regimes in nonrandom forest canopies: theory, measurements, and a simplified modeling approach. Tree Physiol. 19:695-706. 43. Kull, O and Jarvis, PG. The role of nitrogen in a simple scheme to scale-up photosynthesis from leaf to canopy. Plant, Cell, Envion. 18:1174-1182. 44. Lai et al. 2002. Modeling nighttime ecosystem respiration by a constrained source optimization method. Global Change Biol. 8:124-141. 45. Law et al., 1999. Seasonal and annual respiration of a ponderosa pine ecosystem. Global Change Biol. 5:169-182. 46. Law et al. 2001. Carbon dioxide and water vapor exchange by young and old pondersosa pine ecosystems during a dry summer. Tree Phys. 21:299-308. 47. Law et al. 2000. Measurements of gross and net ecosystem productivity and water vapor exchange of a Pinus ponderosa ecosystem and an evaluation of two generalized models. Global Change Biol. 6:155-168. 48. Li, X. and Strahler, A. H. 1985. Geometric-optical modeling of a conifer forest canopy. IEEE Trans. Geosci. Rem. Sens., GE-23:705-721. 49. Litvak et al. 2003. Effect of stand age on whole ecosystem CO2 exchange in the Canadian boreal forest. J. Geophy. Res. –ATM. 108(D3). Art. No 8225. 50. Liu, J. et al. 1997. A process-based boreal ecosystem productivity simulator using remote sensing inputs. Rem. Sens. of Environ., 62:158-175. 51. Liu, W. 2001. ARTMAP neural network for land cover classification and multisensor fusion in remote sensing. Dissertation. Boston Unviersity. 52. Miline, B. T. and Cohen, W. B. 1999. Multiscale assessment of binary and continuous landcover variables for MODIS validation, mapping, and modeling applications. Rem. Sens. Environ., 70:82-98. 53. Monteith, J. L. 1972. Solar radiation and productivity in tropical ecosystems. J. Appl. Ecol., 9:747-766. 54. Myneni, R. B. et al. 2001. A large carbon sink in the woody biomass of Northern forests. Proc. Nat. Acad. Sci. USA, 98:14784-14789. 55. Myneni R. B. et al. 2002. Global products of vegetation leaf area and fraction absorbed PAR from year one of MODIS data. Rem. Sens. of Environ., 83:214-231. 56. Nemani et al. 2003. Climate-driven increases in global terrestrial net primary production from 1982-1999. Science. 300:1560-1563. 57. Oosting, H. J. 1942. An ecological analysis of the plant communities of Piedmont, North Carolina. Am. Midl. Nat. 28:1-126. 58. Oren et al., 2001. Soil fertility limits carbon dequestration by forest-ecosystems in CO2 enriched atmosphere. Nature. 411:469-472. 59. Pacala et al. 2001. Consistent land- and atmosphere-based US carbon sink estimates. Science. 292:2316-2320. 60. Parton et al. 1993. Observations and modeling of biomass and soil organic-matter dynamics for the grassland biome worldwide. Global Biogeochem. Cyc. 7:785-809. 61. Peet, R. K. 1981. Changes in biomass and production during secondary forest succession. In: Forest Succession: concepts and application. West et al. (Edts). Springer-Verlag. 324338.

62. Peet, R. K. 1992. Community structure and ecosystem function. In: Plant Succession: Theory and prediction. Glenn-Lewin et al. (Eds). Chapman and Hall. London. 63. Peterson et al. 1986. Analysis of forest structure using Thematic Mapper simulator data. IEEE. Trans. Geosci. Rem. Sens. 24:113-120. 64. Potter, C. S. et al. 1993. Terrestrial ecosystem production: a process model based on global satellite and surface data. Global Biogeochem. Cycles, 7:811-841. 65. Prince, S. D. and Goward, S. N. 1995. Global primary production: a remote sensing approach. J. Biogeography, 22:815-835. 66. Raich, J. W. et al. 1991. Potential net primary productivity in South America: application of a global model. Ecol. Appl., 1(4):399-429. 67. Reich, P. B., Turner, D. P. and Bolstad, P. 1999. An approach to spatially distributed modeling of net primary production (NPP) at the landscape scale and its application in validation of EOS NPP products. Rem. Sens. Environ., 70:69-81. 68. Running et al. 1989. Mapping regional forest evapotraspiration and photosynthesis by coupling satellite data with ecosystem simulation. Ecol 70:1090-1101. 69. Running, S. W. and Hunt, E. R. 1993. Generalization of a forest ecosystem process model for other biomes, Biome-GBC, and an application for global scale models. In: Scaling Physiological Processes: Leaf to Globe. Academic Press, San Diego, CA. 70. Running et al. 1994. Terrestrial remote sensing science and algorithms planned for EOS MODIS. Int. J. Rem. Sens. 15: 3587-3620. 71. Running et al. 2000. Global terrestrial gross and net primary productivity form the Earth Observing System. In: Methods in Ecosystem Science. Sala et al. (Eds). Springer-Werlag, New York. 72. Ryan, M. G. and Yoder, B. J. 1997. Hydraulic limits to tree height and tree growth. BioSci. 47:235-242. 73. Schimel, D. et al. 2000. Contribution of increasing CO2 and climate to carbon storage by ecosystems in the United States. Sci., 287:2004-2006. 74. Schimel DS, et al. 2001. Recent patterns and mechanisms of carbon exchange by terrestrial ecosystems. Nature, 414:169-172. 75. Sellers, P. J. 1987. Canopy reflectance, photosynthesis, and transpiration. II. The role of biophysics in the linearity of their interdependence. Rem. Sens. Environ., 21:143-183. 76. Song, C. and Woodcock, C. E. 2002. The Spatial manifestation of forest succession in optical imagery: the potential of multiresolution imagery. Rem. Sens. Environ., 82:272285. 77. Song, C., Woodcock, C. E., and Li, X. 2002. The Spectral/temporal manifestation of forest succession in optical imagery: the potential of multitemporal imagery. Rem. Sens. Environ. 82:285-302. 78. Song, C. and Woodcock, C. E. 2003a. A regional forest ecosystem carbon budget model: impacts of forest age structure and landuse history. Ecol. Model., 164:33-47. 79. Song, C. and Woodcock, C. E. 2003b. Estimating tree crown size from multiresolution remotely sensed images. Photo. Eng. Rem. Sens. In Press. 80. Song, C. and Band, L. E. 2003, MVP: a model to simulate the spatial patterns of photosynthetically active radiation under discrete forest canopies. Can. J. For. Res. Submitted.

81. Spanner et al. 1990. Remote sensing of temperate coniferous forest leaf-area index: the influence of canopy closure understory vegetation and background reflectance. Int. J. Rem. Sens. 11:95-111. 82. St-Onge, B. A.. and Cavays, F. 1995. Estimating forest stand structure from high resolution imagery using the directional variogram. Int. J. Rem. Sens. 16:1999-2021. 83. St-Onge, B. A. and Cavays, F. 1997. Automated forest structure mapping from high resolution imagery based on directional semivariogram estimates. Rem. Sens. Environ. 61:82-95. 84. Tague, C. L. and Band, L. E. 2003. RHESSys: Regional Hydro-Ecologic Simulation System: An object-oriented approach to spatially distributed modeling of carbon, water and nutrient cycling. 85. Tilman, D. 1988. Plant strategies and the dynamics and structure of plant communities. Monograhs in Population Biology 28. Princeton Univ. Press, Princeton, NJ. 86. Turner, D. P., et al. 2002. Effects of spatial variability in light use efficiency on satellitebased NPP monitoring. Rem. Sens. Environ., 80:397-405. 87. Tucker, C. J and Sellers, P. J. 1986 Satellite remote-sensing of primary production. Int. J. Rem. Sens., 7:1395-1416. 88. Wang, YP and Jarvis, PG. 1990. Description and validation of an array model-MAESTRO. Agri. For. Meteor. 51:257-280. 89. Woodcock et al. 1997. Inversion of the Li-Strahler canopy reflectance model for mapping forest structure. IEEE. Trans. Geosci. Rem. Sens. 35: 405-414. 90. Yoder et al., 1994. Evidence of reduced photosynthetic rates in old trees. For. Sci. 40:513527.

Facilities Department of Geography University of North Carolina at Chapel Hill The Department is wired internally to support the latest network technology. Building wide wiring was installed in 1993 with Category 5 UTP cable to support data transfer speeds of 155mb/sec and all offices and labs are supported with the minimum of 10BaseT switched Ethernet. With the help of our on campus networking support group, the Department have been able to take advantage of Fast Ethernet 100BaseT technology for those platforms. Academic Technologies and Networking (ATN) of UNC provide networked AFS disk space service to projects with extended disk space need. In addition to this resource, the Department has recently included services provided by ATN via the Distributed Computing Initiative and now gains access to RAID 5 file service and a large base of public domain software, compilers and software development tools, and statistical and mathematical packages. In addition to internal departmental resources any student in the lab may make use of other on-campus resources such as public computer labs and the statistical and compute platforms (a Sun Enterprise 6000 with 12 CPUs and 5GB of RAM and a SGI Origin 2000 with 16 CPUs and 8GB of RAM respectively) maintained by the University. Sun Microsystems Blade 1000 (2), Sun Microsystems Blade 2000 (1) Sun Microsystems workstation Ultra 60 (1), IBM PCs (1.5Ghz) (10), Tektronix Phaser 360 Color Ink Jet Printer, Hewlett-Packard HP 4000TN Monochrome Laser Printer, Hewlett-Packard DesignJet 750C Large Format Ink Jet Plotter, Calcomp 9500 Large Format 36" x 48" Digitizing Tablet, Polaroid Slide Scanner, Polaroid Film Printer, 4mm and 8mm Tape Drives, Recordable CD-ROM Drive, Epson 836XL Large Format Color Scanner, 1 Calcomp 1025 pen plotter, 1 Sun SPARCprinter E B&W Laser Printer, 1 Sun CD-ROM reader, 1 Sun SPARCprinter E Black and White Laser Printer, Exabyte 210 8 mm Tape Library System, Epson 836 XL color image scanner. Trimble Navigation GeoExplorer III GPS Unit (2) IBM Portable NoteBook Computers (3), Li-Cor 1800 Spectroradiometer, Li-Core Plant Canopy Analyzer (2), Li-Cor and Campbell data loggers, LiCor quantum sensors, Nikon Cool Pics 5000 Digital Camera, Sony 8mm Video Camera, JVC TV, JVC VCR, Topcon Mirror Stereoscope, Walkie-talkies, Canon printer/copier/scanner, CD Writeable System. PCI 6.2 , ESRI Institutional Site License for UNC (e.g. ARC/INFO and ARCVIEW), ERDAS Imagine 8.3 (23 seats); ERDAS Orthomax and Sub-Pixel (1), ER Mapper 5.1 (10 seats), IPW, AVS 5.0.2 (1 seat), GRASS 4.1 (5 seats), SAS 6.0.9 (25 seats), S-Plus 3.2 (25 seats), Adobe Photoshop, Adobe Premiere, Claris Works, Access database, Fragstats Software (1).

DURHAM NORTH CAROLINA 27708-0328

NICHOLAS SCHOOL OF THE ENVIRONMENT AND EARTH SCIENCES BOX 90328

TELEPHONE (919) 613-8000 FACSIMILE (919) 684-8741

August 12, 2003 Dr. Conghe Song Department of Geography University of North Carolina at Chapel Hill Chapel Hill, NC 27599 Dear Conghe We are writing to confirm our interest in your proposed work involving the use of mircometeorological data to estimate impacts of spatial vegetation heterogeneity in regional scale carbon fluxes. We will provide all the necessary measurements from the three Duke Forest AmeriFlux sites, including all the core and desired measurements specified by the AmeriFlux science team. Furthermore, we will work with your student on parameterzing the model with the necessary physiological inputs such as light-use efficiency, maximum carboxylation capacity, leaf area dynamics, etc…. Your proposed approach regarding landscape heterogeneity provides unique synergies with our recently supported DOE-TCP proposal at the Duke AmeriFlux sites as well as our DOE-NIGEC effort. We look forward to a continued exciting collaboration with you and your student. Please let us know if there is anything further we can do to assist you.

Gabriel Katul School of the Environment Duke University Durham, NC 27708-0328

Ram Oren School of the Environment Duke University Durham, NC 27708-0328

APPENDIX

ix. Ward, Lucas. Doctoral Dissertation Research: Risk, Vulnerability, and Sustainability – Translating Integrated Water Resource Management into Action on the Paraguay River.

Project Description Doctoral Dissertation Research: Risk, Vulnerability, and Sustainability - Translating Integrated Water Resource Management into Action on the Paraguay River 1. Introduction In the northernmost tip of the Paraguayan Pantanal, the most ecologically diverse and poorest portion of the world's largest continuous wetlands area, inhabitants of the region use the Paraguay River for fishing, animal husbandry, irrigation of crops, drinking water, and other life sustaining activities. Approximately a decade ago, the Paraguayan Congress passed a water management law, Ley 799/96, which outlawed net fishing on the Paraguay River during droughts and required those who fished the river to pay for and possess fishing licenses. However, water managers charged with upholding 799/96 rarely translated the “no net fishing” and fishing license requirements into effective management or enforcement activities (JICA 2002, SEAM 2003). Consequently, Ley 799/96 has had little impact on Pantanal fishing practices, local livelihoods, or fish populations. This is all about to change. In September of 2006, as part of its World Bank-required National Environmental Action Plan, the Paraguayan government began the final stages of reforming its water laws according to the principles of Integrated Water Resource Management (IWRM). An IWRM approach to water governance incorporates the water needs of various users into a management arrangement in which participation from multiple stakeholders is supposed to ensure that water use is socially, economically and environmentally sustainable. Guided by IWRM, experts from the World Bank’s Global Environmental Facility and the Paraguayan government are working together with non-governmental organizations (NGOs), scientists, and civil society groups to reform 799/96. The participation of these new actors in the impending reform of 799/96 raises the following research questions: (i) How will the adoption of IWRM affect the ability of the Paraguayan state to set resource management priorities? (ii) How will new and existing managers translate the IWRM–influenced version of 799/96 into regulatory action at state, regional, community, and scales? (iii) To what extent will implementation of IWRM make households more vulnerable? By gathering data on how different water managers, including inhabitants of the Pantanal themselves, perceive and use 799/96, the proposed research seeks to understand how IWRM is translated into action in the Pantanal and how this translation affects the Paraguayan state and Pantanal livelihoods. The co-PI will use multiple methods of data collection, including semi-structured interviews, participant observation, timelines, and Q-sort, and several methods of data analysis, including comparative analysis, inverted factor analysis, and cluster analysis to address these questions. Comparative analysis of the translation practices of water managers in charge of implementing IWRMbased 799/96 at state, regional, community, and individual scales will allow the co-PI to identify and isolate the contextual factors that influence how managers translate the reformed version of 799/96 into action as well as how households adapt to the new water law. Data produced by the Q-sorts of the state and sub-state scale water managers as well as Pantanal households will be analyzed using inverted factor and cluster analysis to produce variable-wise and case-wise interpretations of the dimensions distinguishing water managers’ views on the risks associated with water management and use, Ley 799/96, and scale-specific management conditions. 2. Conceptual Framework Net fishing has traditionally been a particularly important source of food and money for many of the poor isolated communities that populate the biodiverse yet impoverished Pantanal (c.f. Renshaw 1996). Because of lack of enforcement of Ley 799/96, the law that prohibits net fishing in times of drought and requires those who fish the Paraguay river pay for and possess a fishing license, Pantanal households have previously been able to access and place fishing nets in the Paraguay river in order to meet their needs. However, recent critical reviews of Paraguayan environmental standards published jointly by uni-and multilateral development agencies and the Paraguayan government suggest that the situation will soon change. These reports state that regular and effective enforcement of Ley 799/96 is deficient and promote an “integrated” approach to water resource management as a more effective means to properly allocate water (JICA 2002; c.f. SEAM 2003; GEF 2003). At the same time, these reports

1

indicate that enforcement of the “no net fishing” and fishing license requirements will be a priority for Paraguay’s new, reformed, IWRM-based version of 799/96 (Ibid). As of September 5, 2006, when the Paraguayan Senate held a mandatory audiencia publica (public notice and comment session) designed to allow various NGOs, environmentalists, and scientists to comment on the IWRM-based version of its new water law, such restrictions are one step closer to being implemented in Paraguay (Paraguay 2006). At the abstract level of multilateral environmental agreements like the Dublin Principles for water management, IWRM is recognized as the best management strategy for coordinating the water requirements of humans and non-humans, including, fishermen, farmers, owners of industry, terrestrial and aquatic wildlife, and riparian vegetation (Garcia 1998, Wescoat and White 2003; WB 2006). 1 IWRM incorporates the water needs of various users into a management arrangement in which participation from multiple stakeholders ensures that water use is socially, economically and environmentally sustainable (Saleth and Dinar 2000, GEF 2005). IWRM frameworks are especially attractive because in balancing the water needs of multiple users, they ostensibly "avert water crisis and reduce poverty" (GEF 2003). To do so, however, IWRM systems require the cooperation of water users, planners, managers, and policy makers who are spread over large geographic areas and whose water needs are different and sometimes contradictory. At the same time, as seen in the example of the audiencia publica, IWRM requires the participation of new actors in the implementation of country, region, and community-specific management rules. Research from Africa shows that at some point in the IWRM participatory process, the water needs and associated benefits of certain water users must be sacrificed so that sustainable use by all users can be realized (Mazungu 2002; Madula 2003;c.f. Junk 2000). Related geographic research suggests that the adoption of management schemes like IWRM in the Pantanal may be damaging to local economies, political power, and environments (c.f. Leach and Mearns 1996; Zimmerer and Basset 2003, Peet and Watts 2004). This research is substantiated by the 2003 pilot study conducted in the Pantanal by the coPI (Ward 2003). In interviews, heads of Patanal households explained that in times of drought, benefits from all income earning and food procuring activities – from subsistence agriculture to hunting to wage labor on cattle ranches and in neighboring communities – decreased. In the past, they claimed, many households had responded to hazards like droughts by diversifying their livelihoods to rely more heavily on activities such as net fishing, which yield more protein and income than activities like subsistence agriculture or animal husbandry (Renshaw 1996). Pantanal heads of household insisted that license requirements and dry-season restrictions on net fishing – one of many activities in each household's "portfolio" of livelihood practices – would make it harder for them to survive droughts (Ward 2003, c.f. Appendini 2001). In other words, IWRM initiatives for sustainability would increase their vulnerability to natural hazards. By requiring new actors, each with different management goals, expertise, and values, to participate in governing the Paraguay River, IWRM changes how water in the Paraguay River is allocated. This change in governance systems also raises the questions of how the IWRM approach affects the Paraguayan state, which had previously enjoyed exclusive authority over its water resources, and how it will affect Pantanal households, who had previously enjoyed unregulated access to the river. To address these questions, the proposed research conceptualizes IWRM as a governance system whose concrete effects – on the Paraguayan state and on Pantanal livelihoods – are the product of a series of translations wherein managers at different scales translate IWRM into country-, region-, community-, and finally individual-scale management action. In order to understand the effects of IWRM, this research proposes to track the translation of IWRM-based 779/96 through these different scales of management. To investigate how IWRM affects the Paraguayan state and Pantanal livelihoods, this study will focus on (i) how the adoption of IWRM affects the ability of Paraguay’s federal environmental agency to set resource management priorities, (ii) how managers at different scales translate 799/96 into action and (iii) how practical use of these translated versions of 799/96 affects Pantanal households’ access to resources, income, and services. 1

The concept of IWRM is extremely popular in World Bank and GEF rhetoric as well (c.f. GEF 2003, WB 2004).

2

3. Research Questions Research Question One: How does the adoption of IWRM affect the ability of Paraguay’s Ministry of the Environment (SEAM) to set resource management priorities? Hypothesis One: By forcing SEAM to accept input from “local” NGOs, scientists, and environmentalists about the substance and wording of Paraguay’s new water law, IWRM reduces the ability of the state to set resource management priorities. Research Question Two: The proposed research asserts that understanding how these managers translate laws into action is critical to understanding the outcomes of these laws like the IWRM-based 799/96. This raises the second research question: How do managers responsible for implementing IWRM at state, regional, and community scales translate the IWRM-based 799/96 into regulatory action? Hypothesis two: Investigations of governance systems in post-structural political ecology and anthropology suggest that managers translate standards into action according to a variety of factors, including their own expertise as well as their perception of risks, local management conditions, and the standards themselves (c.f. Bowker and Star 2000, Kull 2005). The proposed research assumes that all four of these factors interact with each other as they influence how managers implement Paraguay’s new ley de agua, the IWRM-based 799/96. Hypothesis two is divided into four sub-hypotheses in order to trace out the individual effects of each factor on managers at different scales. Sub-hypothesis One: Managers translate the IWRM-based 799/96 into action according to their different expertise. Sub-hypothesis Two: Managers translate the IWRM-based 799/96 into action according to their different perceptions of IWRM. Sub-hypothesis Three: Managers translate the IWRM-based 799/96 into action according to their perceptions of risks associated with water use and management. Sub-hypothesis Four: Managers translate the IWRM-based 799/96 into action according to their perceptions of local conditions. Research Question Three: Preliminary research suggests that if these new managers translate the IWRM-based Ley 799/96 into regulatory action in a way that forces Pantanal inhabitants to observe net fishing and license regulations, it will make many Pantanal households more vulnerable to natural hazards, such as droughts and flooding. This raises the question of how individual water users use 799/96 in their day-to-day activities and what its concrete effects are. The third research question asks: How do Pantanal households’ adaptive responses to the IWRM-based 799/96 make them more vulnerable to natural hazards? Hypothesis Three: The third hypothesis is that Pantanal households whose practical use of 799/96 forces them to alter the ways that they access protein and income and interact with communityscale water managers and service providers will be the most vulnerable to droughts and floods. 3. Theoretical Framework Contemporary governance systems and the state: Post-structural studies from geography and anthropology suggest that by identifying the cause of poor countries’ development problems as managerial (the state is a poor manager of its resources because it is partisan and unscientific in its approach to management) and determining the solution to be participatory, scientific management, multilateral development agencies create space for themselves and other non-state actors to participate in “integrated” governance (c.f. Foucault 1991; Ferguson 1994, Escobar 1995, 1998; Forsythe 2003; Goldman 2004). As a result, many geographers have argued, states lose exclusive authority to make decisions about the management and improvement of ecological spaces that had previously been considered their sovereign territory (c.f. Peet and Watts 2004; see Agrawal 2005 for an opposite position on state power). New governors & how they govern: Geographers and anthropologists have produced a great deal of research into how civil society groups (c.f. Escobar 1998; Sundberg 2003; Latour 2004),“local” and international non-governmental organizations (Bebbington 1993; Young 2003; Peet and Watts 2004),

3

as well as existing state agencies (Scott 1998; Agrawal 2005; Yeh 2006; Li 2005; c.f. Dunn 2003) exercise power in contemporary environmental management systems. These studies suggest that new actors govern by negotiating, contesting, and compromising with each other, existing state agencies, and multilateral development agencies to shape governance systems so that they meet their different needs and objectives (Forsythe 2003; Robbins forthcoming). As a whole, these studies suggest that in a given site or scale, different practices – from cultural to economic to technological to expertise-based practices allow individuals or groups to influence both the structure of overarching governance arrangements, such as IWRM, and the spatial organization of the different landscapes they seek to manage (Zimmerer and Basset 2003, Robbins 2004, Peet and Watts 2004). Similar research suggests that expertise about a location or group of people is critical to the exercise of power because it permits a potential manager to define a problem, plan and execute a re-action to the problem, and to deal with unforeseen consequences of the re-action (c.f. Foucault 1977, Scott 1998; Mitchell 2002; Forsythe 2003, 2004; Goldman 2004). While multilateral development agencies possess a great deal of the expertise required to govern these spaces (Peet and Watts 2004; Goldman 2004), they are by no means the only purveyors of it. Foucault (1977; 1991), Scott (1998) and others have argued that state agencies possess and strategically use country-specific or regionally specific expertise in order to influence the economic, political, and cultural organization of their territory (c.f. Dunn 2003, Agrawal 2005; Yeh 2006, Li 2005; see Bebbington 1993; and Young 2003 for a similar argument about NGOs). At the same time, Escobar (1998, 1999) argues that there are infinite subdivisions of expertise - scientific expertise, cultural expertise, or bureaucratic expertise, to name a few - and therefore an infinite set of possible actions that a group or individual might engage in in order to exercise power (see Escobar 1998, 1999 for a case study of such a scenario; c.f. Ellen and Harris 2000; Scott 1985). Geographic and anthropological research into the politics of spatial organization insists, however, that it is important to recognize that while these seemingly discrete practices may be deployed at different scales (Zimmerer and Basset 2003) or be based on different forms of knowledge (Ellen and Harris 2000), they generally take place within the broader context of an overarching management system administrated by the state or multilateral development agencies (c.f. Goldman 2004; Agrawal 2005). This research proposal argues that resource management systems like IWRM serve as the politicized context in which these various actors’ practices play out (Agrawal and Gibson 1999; Agrawal 2005). In lieu of focusing exclusively on the Paraguayan state (c.f. Scott 1998; Yeh 2006) or the World Bank (c.f. Ferguson 1994, Escobar 1994, Peet and Watts 2004) as a means to investigate how power works or how expertise is deployed in contemporary conservation and development settings, this research takes up the translation of the IWRM-based Ley 799/96 into a practicable standard as a way to analyze how state agencies, civil society, and multilateral development agencies all compete to control water use and ecological space. This line of inquiry is informed by geographic and anthropological theories about standards. What standards are: Standards, for the purposes of this proposal, are resource management laws intended to lead to economically and ecologically sustainable development by enforcing a particular organization of space as well as the productive behavior of those actors involved in its construction (c.f. Busch 2000; Strathern 2002; Li 2005). Standards describe and attempt to create particular governable and improvable organizations of space and in doing so create new political spaces and actors (c.f. Gupta and Ferguson 1992, Scott 1998, Bowker and Star 2000; Li 2005). These actors are expert managers whose responsibility to translate the written text of a rule into a practicable standard and to enforce the standards makes them key players in contemporary governance projects (Bowker and Star 2000). In the context of decentralized, participatory governance systems, the individual to whom standards apply can also be considered a manager (c.f. Agrawal 2005). Because the spaces (rivers, factories, hospitals) onto which standards seek to inscribe a particular organization of space are inhabited and used by humans, understanding how individuals manage and explain their practical use of environmental standards is critical to understanding how space is organized, controlled, and improved in contemporary governance arrangements (Luke 1997; Goldman 2004; Agrawal 2005). The degree to which the IWRM-based 799/96 affects fishing and other water-based practices as well as the Paraguayan

4

state, then, depends on how state managers – state agencies, NGOs, scientists, civil society groups – and sub-state managers – regional, community, and ultimately individual managers – translate the IWRMbased 799/96 into action. This approach thus rejects the notion that the process of translation takes place in a technocratic realm devoid of power relations, and focuses on the way that local context shapes implementation and the law’s eventual effects. Translation of standards – “nonstandard categories”: Geographic and anthropological research informed by science studies and governmentality literature shows that practical use of standards entails that those to whom standards will apply develop “work-arounds” and “ad-hoc nonstandard categories” in order to negotiate the political, economic, and ecological processes that also constrain their daily practices (Bowker and Star 2000, 15; cf. Scott 1985, Watts 1987, Strathern 1998; Dunn 2004). As both managers and the intended beneficiaries of resource management and sustainable development projects respond to the standards for acceptable practice embedded in projects like IWRM, their “ad-hoc” responses are mediated by local, place-based knowledge structures and determinate economic and social conditions (c.f. Bourdieu 1977, Agrawal 1995, Fairhead and Leach 1995, Ellen and Harris 2000). Variation in responses to standards is also a function of the heterogeneity of those charged with enforcing the rules (Kull 2004, Agrawal 2005). These debates suggest that in Paraguay, the material outcomes of IWRM will be ultimately determined by how Pantanal water users adapt Ley 799/96 into their daily routines. The second half of the proposed research seeks to add risk perception to the discussions of what factors shape ad-hoc responses to standards. At the same time, by focusing on Pantanal households’ adaptive responses to 799/96, this research seeks to understand to what degree decentralized, science-based sustainability initiatives like IWRM transform makes certain households more vulnerable to natural hazards. Vulnerability: In natural hazards literature, vulnerability is used to describe and investigate the susceptibility of specific groups of people living in specific locations to a natural hazard (Blaikie et al. 1994, Smith 2004). People’s constantly changing ability “to compete for access to rights, resources, and assets” is a critical component of their vulnerability to natural hazards (Pelling 2003, 75; c.f. Blaikie et al. 1994; Cahn 2002). Blaikie et al. (1994) argue that the human aspects of vulnerability emerge not from natural causes but from social ones, including access to power structures and resources as well as economic and political ideology (c.f. Wisner et al. 2004). Dynamic pressures, including macro-forces such as population growth and debt, as well as more case specific conditions, such as lack of institutions, markets, or access to resources, translate these socially-constituted root causes into “particular forms of insecurity that have to be considered in relation to the types of hazards facing people” (Blaikie et al. 1994: 24). Thus people’s ability to access the resources, rights, and assets that would decrease their vulnerability to natural hazards, such as droughts and flood, is influenced by “intervening sociopolitical structures” (Pelling 2003, 77). In the Pantanal, the implementation of IWRM-based 799/96 represents an intervention that may, in attempting to regulate fishing practices, actually corrupt the existing mechanisms through which Pantanal households access and distribute essential resources and decrease their vulnerability to natural disasters. 4. Methods and Research Schedule – Hypotheses One and Two To test the first hypothesis,2 which addresses the question of how IWRM affects the Paraguayan state, and the set of four sub-hypotheses,3 which address the second question of how managers at different scales translate IWRM-based laws into action, the co-PI will spend three months in Asunción, the capitol of Paraguay, and three months in Fuerte Olimpo, the headquarters for regional governance of the Pantanal. In Asunción, the co-PI will collect data by conducting interviews at and reviewing documents prepared by the following groups: (i) Paraguay’s federal environmental agency, Secretaria del Ambiente (SEAM);(ii) departments of Agroforestry and 2

The first hypothesis is that IWRM reduces the ability of the state to set resource management priorities by forcing SEAM to accept input from other actors about the substance and wording of Paraguay’s new water law. 3 The four sub-hypotheses for research question two are that managers translate the IWRM-based 799/96 into action according to their different expertise, perceptions of IWRM, perceptions of risks associated with water use and management, and their perceptions of local conditions.

5

Human Ecology at the national university in San Lorenzo, Paraguay (iii) the Paraguayan NGO Sobrevivencia, which has been part of the “public audience” for state-scale reform of 799/96; (iv) civil society groups including the national indigenous association INDI; (v) the World Bank’s Global Environmental Facility, which works closely with SEAM; and (vi) other relevant actors identified using network techniques and through contacts from the 2003 pilot study. This research will be conducted by the co-PI – who speaks Spanish and Guaraní fluently – and a graduate assistant from the department of Human Ecology at the National University of Asuncion as a field assistant. Approximately 60 interviews, including follow-ups, will be conducted in these first three months (from 6/2007 – 9/2007). Data collected from these state-scale managers with diverse expertise will provide the co-PI with baseline information about perceptions of the IWRM-infused 799/96. This will allow the co-PI to compare perceptions of critical differences in how the reformed 799/96 describes and proposes to govern and improve vital spaces in the Pantanal, such as fishing net placement sites, to how management was coordinated by the 1996 version of 799/96. This data will allow the co-PI to determine the extent to which the new participatory, science-based decision making component of IWRM creates new managers (for example Water User Associations –(c.f. WB 2006)), new scales of management (for example, watershed-scale managers (c.f. SEAM 2003)), or alters the ability of existing state-agencies to manage and improve the Paraguayan Pantanal (for example, by giving watershed scale managers enforcement responsibilities formerly reserved for the state). As a whole, this baseline information will provide the co-PI with a conceptual map of who water managers are under IWRM systems, what their responsibilities are according to the text of the revised Ley 799/96, and what their perceptions of IWRM are in practice. In Fuerte Olimpo, the co-PI will spend three months conducting similar research with regional-scale managers, “local” branches of NGOs, regional fishing organizations, and law enforcement officials in order to understand how they translate the IWRM-based 799/96 (hereafter referred to as “799/96”) into regionally-specific management action. Approximately 60 interviews, including follow-ups, will be conducted during these second three months (from 9/2007 – 12/2007). Data Collection: Semi-structured interview questions will be designed to elicit data about the factors that influence how different managers translate 799/96. In order to better understand how different managers in Asunción and Fuerte Olimpo seek to implement net fishing and license laws the co-PI will examine printed material (pamphlets, web pages, maps, and meeting summaries) that describe the expertise, motives, and values that inform their translation practices. In follow-up interviews with these managers the co-PI will also conduct Qsort style risk-mapping exercises in order gather additional data on the values and perceptions of risk that inform different managers’ translation of 799/96. Q-sort: Q-sort approaches aim at uncovering the ways in which respondents bundle together different attitudes, beliefs, and understandings rather than allowing the interviewer to impose a priori assumptions about these relationships (Robbins and Krueger 2000). Q-sort, or Q methodology, is a way to analyze people’s subjectivity or perceptions. In a Q study, a subject ranks a series of statements (from agree to disagree) according to a different conditions or instructions provided by the researcher. These statements represent the range of things people think or say about a topic, not “facts” or “truths” per se. Because the subject is ranking the statements from his or her point of view, this activity provides insight into the respondent’s subjectivity. The rankings are recorded and subjected to factor and cluster analyses. The resulting factors indicate segments of people’s subjectivities (Robbins and Krueger 2000). Interview Questions: To establish a baseline on these groups’ positions during the translation of 799/96 into a state-specific law, as well as how these new (NGOs, scientists) and old (SEAM) managers viewed the positions of other potential governors, the co-PI will ask various representatives about the key arguments they and other groups have made the in “public audiences,” such as the one held in September of 2006, about the spatial organization of the Pantanal. State and regional-scale managers will also be asked to explain what they and their group/agency do – write letters, organize meetings, make videos, take boat trips, follow rules/orders - to influence how 799/96 is implemented. To test whether scientific expertise is a key element in the IWRM approach to governance, as suggested by post-structural political ecologists, representatives of these different groups will be asked how they validate their claims about appropriate use of the Paraguay River. Semi-structured interviews and Q-sort: To understand how state and regional water managers’ values, political economic conditions, and perceptions of risks associated with water use and management influence how

6

they attempt to implement 799/96, the co-PI will use Q-sort risk-mapping techniques (Smith et al. 2000, Robbins and Krueger 2000). Using statements drawn from interviews, newspaper articles, and other local sources to represent views on the key dimensions of this study — water management priorities, perceptions of risk associated with water use and management, perceptions of local conditions, and expertise — a set of 40-45 quotes will be sampled, and paraphrased where necessary, to represent differing policy priorities, such as claims about the impacts of net-fishing laws, claims about past and current water laws, claims about risk, and the type of evidence that respondents use to support these claims. The statements will be printed on a set of cards to administer in a follow-up survey with 20 key representatives of the development community. Respondents are to rank the provided quotes from most highly agree to highly disagree with the relative position of each quote recorded for further analysis. This will provide a measure of the respondents’ tradeoffs of values, goals, and perceptions, including the intensity with which they respond to the statements (Smith et al. 2000). This information will allow the co-PI to compare patterns in values, perceptions, and demographic information to how respondents report that they will attempt to translate 799/96 into action. Data Analysis Analysis of Q-sort: The data analysis techniques described below, for analyzing data from the development community, will also be utilized to analyze data from the Pantanal communities. In the context of investigations into state-scale translation, the Q-sort will provide evidence that documents both the range and intensity of respondent views with regard to key elements of the process by which competing groups attempt to impose their own water values into the reformed799/96. At regional, community, and individual scales of analysis, the Q-sort will also allow the co-PI to map out the different values that respondents give to statements related to water use priorities, risks and other fundamental elements of this study, across respondent sets. Results of the Q-sort will be subject to an inverted factor analysis (Robbins forthcoming) to determine common patterns of response and consistent relationships between specific dimensions of the decision making process – for example, between differing specific claims about net fishing in times of drought, what risk is, what acceptable use of water is – within respondent sets. In the context of data from the actors involved in state- and sub-state-scale translation of IWRM, this will elicit coherent subject positions on the spatial organization and management of the Paraguay River and key values and strategies that influence how the IWRM is translated into country-scale regulations that can be controlled for comparison (Robbins and Krueger 2000). The results of the inverted factor analysis will serve as a "check" of the patterns observed within respondent sets using the Q-sort. In doing so, this analysis technique will allow the co-PI to compare the degree to which there is agreement within respondent sets – and particularly within the group of respondents representing the state - as to how space in the Paraguayan state is (and ought to be) governed and organized. This comparison will permit the co-PI to determine whether management systems like IWRM affect the Paraguayan state by making it more difficult for its environmental agency (SEAM) to set priorities for water management. Testing the significance of agreements/ disagreements across subject sets will also indicate which aspects of water management and sustainable development processes different respondents find particularly controversial, and therefore likely to require negotiation and compromise. At the same time, this analysis will indicate which aspects of translating IWRM into a new version of 799/96 are more likely to generate alliances between governance groups. This information on strategic alliances between actors with different political objectives, or “discourse coalitions” (Robbins forthcoming; c.f. Forsythe 2003), will allow the co-PI to address the possibility that the state responds to its loss of direct control over space by assuming control of the discourses through which other actors exercise their authority. Cluster Analysis - The interview responses will also be analyzed with cluster analysis to identify the different clusters of viewpoints and the extent to which configurations of views and values are associated with different multilateral development agencies, state agencies, or other groups as well as values (risk perception) or are better characterized by nonorganizational attributes. 5. Methods and Research Schedule – Hypothesis Three To test the hypotheses that Pantanal households whose practical use of 799/96 forces them to alter the ways that they access protein and income and interact with community-scale water managers and service providers will be the most vulnerable to droughts and floods, the co-PI will need data about

7

people’s past and present livelihoods with specific emphasis on changes that they have made to their daily practices in response to water laws and why those changes were made. To collect data on how, why, and to what effect community and individual managers translate 799/96 into a practicable standard, the co-PI will spend 6 months conducting semi-structured interviews, participant observation, and ethnographic data collection activities, including Q-sort, with 120 households in three different Pantanal communities.4 He will also travel to sites where state, regional, and community-scale management meetings are held in order to conduct participant observation of these participatory processes in the field. By linking patterns in risk perception within and between Pantanal households to action that these water users take in response to 799/96 this analysis will test whether there is a risk-profile of households who are made more vulnerable to droughts and floods as a result of 799/96. Data Collection Participatory Risk Mapping: This component of the research project will be used in conjunction with the Q-sort to collect data on whether there is a change in risk perception in communities located at different points along the Paraguay River. This data will be analyzed with data about practical use of standards in different communities as well as perception of water management in order to ascertain whether there is a relationship between how different groups perceive risk and how they translate 799/96 into a practicable standard. Step one of participatory risk mapping is to get people to free-list things they are worried about. This will be done in focus groups of approximately 5-15 male and female adults upon arrival to each of the three communities. Focus group participants will be chosen through self-selection. Upon arrival in each community, the co-PI will issue invitations (at church meetings and community events) to participate in the risk mapping exercise. A list of potential focus group participants will be compiled and all of those on the list will be invited to participate in the focus group on a given date. If more 15 people respond to invites, multiple groups will be formed. Where possible these groups will be split up according to gender to test whether male and female water users perceive risk and respond to 799/96 differently. Focus groups will be asked to free list things that pose a threat to their well-being. Based on these lists, a number of different perceived risks will be compiled and printed, one-per-card, on a set of cards. Steps two (ranking the risks) and three (collecting household economic and demographic histories) of participatory risk mapping will be conducted with the Q-sort activity (see below), following the semistructured interviews and participant observation. Semi-structured interviews and participant observation: In the semi-structured interviews in Guaraní and Spanish the co-PI will ask community water managers and Pantanal water users questions about how they have changed their livelihood and water management strategies over time.5 Questions will be designed to elicit specific information about how income and food earning activities have or have not changed as a result of new water laws. Questions will also be designed to collect information on whether practical use of 799/96 at the community scale has affected households’ ability to “compete for access to rights, resources, and assets,” a critical component of their vulnerability to natural hazards (Pelling 2003, 75; c.f. Blaikie et al. 1994; Cahn 2002). In order to gain a long-term perspective on people’s daily practices, the co-PI will work with households to construct a timeline that shows the various activities that the heads of household have engaged in in order to sustain themselves and their dependents. Since individuals’ claims alone rarely provide clear insight into their daily behavior, participant observation will be conducted with Pantanal households, recording their daily interactions with other community members, water law authorities, and the nature and location of their work. 4

The co-PI will be in each Pantanal community for 2 months so he will work with approximately 40 households in each community, using multiple methods of data collection, for a total of 120 households during the research project. The communities are: the mestizo community of Bahia Negra, the indigenous Chamacoco community of Puerto Diana, and the indigenous community of Catorce de Mayo. The co-PI selected these three field-sites because he has prior research experience in each community and because their contrasting ethnic compositions will allow him to check for cross-cultural variation in how households respond to and are affected by the IWRM-based 799/96. 5 In households where only Chamacoco is spoken the investigator will work with a local translator.

8

Q-sort: As with data collection with actors involved in state and regional-scale translation of 799/96, selected statements from interviews and participant observation with Pantanal households and community-scale water managers, along with newspapers and other local sources will be used to construct a Q-sort. Selected statements will reflect a variety of different claims about key elements of household experiences with water laws such as participation in IWRM planning, community-scale management, perspective on water laws and risk, livelihood changes as a result of water laws, and interaction with water managers. These statements will be printed on a set of cards to administer in a follow-up survey with 30 representative households and community-scale water managers in each Pantanal community. This follow-up survey will collect basic demographic information from interviewees, and then will require respondents to sort the provided quotes from most highly agree to highly disagree with the relative position of each quote recorded for further analysis. The Q-sort of 30 Pantanal households’ and community managers’ cognitive maps of water law experiences & risk perception will provide evidence that indicates both the range and intensity of their views with regard to their claims about their water knowledge, daily work habits, risks, and 799/96. Data Analysis Comparative analysis: Comparative analysis across three Pantanal villages will allow the co-PI to identify and isolate the contextual factors influencing household adaptive behavior of response to 799/96. Comparative analysis of data collected during timeline, participant observation, semi-structured interview, and participatory risk-mapping activities will also provide an extensive examination of the ways that Pantanal water users respond to water laws and how, when, and why they adapted the law to fit their daily practices. Analysis of Q-sort: The analysis of the Q-sort of 25 Pantanal households’ cognitive maps of their water law experiences will provide evidence that indicates both the range and intensity of different respondents’ views with regard to their claims about changes in their knowledge, daily work habits, risks, and 799/96. Data from the Pantanal households’ and community water managers’Q-sorts will be analyzed using inverted factor analysis and cluster analysis to determine common patterns of response and consistent relationships between household water use, work, risk, and landscape change to produce a coherent profile of how different households translate 799/96 into behavior that affects both their wellbeing, and their landscapes. As with the analysis of data collected from state-scale and sub-state-scale managers, inverted factor analysis and cluster analysis will be used to check patterns observed within and between respondent sets (for example, within and between different communities, different political affiliations, and different community water managers). Along with analysis of timelines and other ethnographic data gathered during participant observation, the results of the Q-sort with Pantanal households will be used to examine patterns within the Pantanal communities of households who claim that practical use of 799/96 has forced them to alter the way that they access resources, earn their living, recover from shocks. Q-sorts with Pantanal households, and community, regional, and state scale managers will contain several identical items (such as perceptions of risks associated with water use and management), analysis of the Q-sort will allow the co-PI to check for cases where Pantanal households’ perceptions of key elements of this study cluster with NGOs, scientists, or SEAM representatives. 6. Intellectual Merit This research project will make theoretical contributions to geographic and anthropological research on the changing practices of the state (and other actors) in exercising power over the environments of developing countries. By situating these theoretical debates in an empirical analysis of how state and non-state managers in Paraguay translate science-based sustainability initiatives (IWRM) into action at different scales, this study seeks to understand how risk perception, expertise, and local conditions influence how managers at different scales translate the substance of laws like 799/96 into actions that fit their personal and managerial needs. In doing so, this research will build upon inquiries from geographic political ecology, indigenous knowledge, and science-management literature into the role of local values and socio-economic conditions in shaping practical use of standards in response to sustainability initiatives. By investigating how practical use of 799/96 affects different households’

9

abilities to “compete for access to rights, resources, and assets” (Pelling 2003, 75; c.f. Blaikie et al. 1994; Cahn 2002) the proposed study attempts to link empirically driven models of the geographies of risk and vulnerability (c.f. Blaikie et al. 1994; Wisner et al. 2004), in which the nation-state plays a key role in shaping the “specific conditions” of vulnerability, to theoretically driven debates about the effects of the decentralized exercise of power in the governance and production of particular forms of spatial organization. While geographers who have adopted a livelihoods approach to sustainable development recognize that a constitutive relationship exists between vulnerability, livelihoods, poverty, and participatory sustainable development, the links between vulnerability and required participation in development projects remain under-explored (c.f. Blaikie 1994, Bebbington 1999, Zimmerer and Basset 2003). Simultaneously, while natural hazards literature on vulnerability acknowledges that people may learn from their negative experiences during disaster relief interventions not to rely on services provided by authorities (c.f. Blaikie et al. 1994; Pelling 2003), the degree to which experiences with resource management systems organized for sustainability actually may increase vulnerability is poorly understood. Lastly, this research adds the Paraguayan Pantanal, a region relatively unstudied in Latin American Geography, to these debates. 7. Broader Impacts The proposed investigation will generate new basic data on the nature of World Bank-sponsored resource management and improvement systems and how they affect livelihoods and land- and waterscapes in the Paraguayan Pantanal that will be immediately useful in academic and policy circles in Paraguay and beyond. The use of context-specific research methods such as Q-sort and cluster analysis makes the research findings more accessible and understandable to policymakers. On an institutional level, because this research project will be carried out in collaboration with Paraguayan graduate students in Human Ecology, researchers, and NGO representatives, it will make practical contributions to their research goals and to broader discussions of how the translation of resource use standards shapes the governance and improvement of landscapes and livelihoods within Paraguay. Depending on their contribution, the co-PI will offer these collaborators co-authorship on an article on the collaborative fieldwork design. The co-PI will also publish findings from this research in Spanish language journals. The findings of this research will be disseminated at professional conferences such as the Association of American Geographers, in academic journals and in the classroom. The co-PI will integrate project-based learning modules based on this research into future courses he designs and teaches. The findings of this research will also be summarized, translated into Spanish, and distributed to the affected Pantanal communities, Paraguayan NGOs and state agencies, and Peace Corps Paraguay. Finally, the project will build the research experience of the doctoral student who will conduct it, fostering the training of a student who plans to pursue an academic career and train future generations of scholars. 8. Qualifications of Researcher The co-PI is well prepared for the proposed fieldwork. He lived and worked in two separate communities in Paraguay for three years as an apiculture extensionist for the Peace Corps (08/98-12/01). He speaks, reads, and writes Spanish fluently and speaks Guaraní (which is not a written language) fluently. From July to September of 2003, to gather data for his Master’s thesis the co-PI conducted a pilot study on what factors influence household participation in NGO development projects in all of the Pantanal communities where he will conduct the proposed fieldwork. Based on this experience, the co-PI gained a number of personal and professional contacts and offers of support and collaboration from the director of the NGO Sobrevivencia, faculty in the departments of Agroforestry and Human Ecology at the National University in Paraguay, and Peace Corps Paraguay, which provided access to computer terminals, contact information, and logistical support during the 2003 pilot study. In addition to the required graduate level quantitative methods research class; the co-PI successfully completed the required graduate level qualitative research methods class in which a unit was devoted to the ethical responsibilities of qualitative researchers. An application for a Human Research Committee (HRC) protocol will be submitted to the University of Colorado at Boulder HRC for the proposed research by November of 2006.

10

APPENDIX

x. Wasklewicz, Thad. NSF CAREER Award: Alluvial Fan Form Quantification to Advance Geographic Science and Education.

Project Summary Form analysis in geomorphology has languished for decades, mired in studies of 2D shape as opposed to interpreting 3D and 4D landform characteristics. This produced a situation whereby our current understanding of continuous terrain is not equated with repeatable, measurable form attributes, but rather with qualitative observations or simplified empirical interpretations. Recent developments in geomorphometry have promoted the concept of numerically characterizing form by analyzing geomorphometric structures. Structure arises from a quantitative understanding of the spatial arrangement of morphometric point data and represents a numerical signature of the topographic form. Historical analyses of alluvial fans, which have produced a solid literary foundation, the ubiquitous nature of fans, and an exposed surface expression of fans make them an ideal feature for establishing a morphometric structure approach. The proposed research will use a combination of GPS surveying and GIS analyses to advance a methodology for quantifying fan form using geomorphometric structures. Fans generated from varied formative processes will be considered in the initial design of morphometric structures. The coarse-scale fan structures will lead to further consideration of the concept of simplexity in geomorphology. The development of a methodology to measure geomorphometric structures will then be used to assess several conceptual and theoretical ideas in geomorphology and specific to fan research. GPR, GPS, and GIS will be used in conjunction to assess the surface topographic response of fans to 3D spatial patterns of depositional units in the fan. This same track of research will be expanded to determine if form and depositional pattern relationships hold under different regional environmental conditions. Fan structures developed from ALSM data will also be used to test the long-standing assumption of the ergodic hypothesis in geomorphology. Finally, fan structure will be used to provide a new interpretation of drain basin and fan relationships. USGS 10m DEM data, GPS, and GIS will examine catchment vs. recent fan segment relations to examine the most recent component of fan evolution. The integration of technology with theoretical constructs of geography is critical to student learning and development. The proposed research is readily transferable to the classroom and skills promoted in the research proposal will permit students to progress and eventually answer “big picture” questions using innovative solutions required in scientific, government, and corporate research. While many articles have discussed a need for linking technology and theory as an effective teaching methodology, only recently has a true working design been developed. However, the PI believes course-and curricular-based approaches do not take the learning experience far enough. Integrative courses, taught by the PI, will assist students in developing experimental designs based on theory and technology. The PI proposes to advance the learning experience by having undergraduate and graduate students carry over the coursework to develop original research projects for Honor’s, Masters, and Ph.D. theses or the generation of articles from class experiments. The active learning experience will provide students with a greater appreciation for acquisition of scientific knowledge. Students will be assessed using a set of rubrics. The rubrics serve as a contract, with detailed criteria to outline the exact expectation of the students. There are no hidden agendas in the syllabus. The criteria are determined by proficiency in technology, analysis, theory, independent and critical thinking, participation, and communication (written, oral, and visual). This exercise will provide valuable feedback for the students and the PI.

Project Description PI’s PRIOR NSF RESULTS The PI, in conjunction with Professor Adam Burnett, has received prior NSF support while at Colgate University. The project was entitled “Integrating Global Positioning System Technology within the Geography and Environmental Studies Program at Colgate University.” The global positioning systems have been fully integrated into fieldwork and GIS analyses in an entry-level Environmental Studies course. Majors and non-majors have been involved with learning the technology behind the GPS, basic mapping techniques, and navigating skills in the introductory geography courses. Students in the capstone environmental studies course have been involved in mapping land-use change in the Adirondack Mountains. The GPS units have also been used in the development of spatial data for modeling the movement of contaminated groundwater from an abandoned garbage site in the Hamilton area. An ongoing course of action has been to get students to use GPS technology for honor’s theses research. Wasklewicz was also part of a team of researchers (3 ecologists) that successfully acquired funding through the NSF Biological Field Stations and Marine Labs Program. The proposal was to develop a regional large river ecology research station at the University of Memphis’ Meeman Biological Field Station. NSF funded ~$181,000 of the project and the University of Memphis provided matching funds to build student sleeping quarters, a dining facility, which doubles as a conference area, and long-term living quarters for research faculty members. This is the only facility of its kind along this section of Mississippi River. It is also located in close proximity to the city of Memphis and provides unique experiences for inner-urban youths and teachers. Currently, a bidding process is under way and we anticipate a project start date of Sept. 1, 2002. INTRODUCTION AND PURPOSE Form Analysis: Scientific progress in assessing form has lagged considerably behind quantitative analyses of process mechanics and greatly reduced possibilities for developing process-form relationships. Analysis of form in geomorphology has gone through numerous transitions that began with qualitative assessments of form through uncanny field observations (Gilbert, 1909) and an attempt to unravel historical landscape development (Davis, 1899). Emphases in form analyses changed during the “quantitative revolution” in geography to planimetric measures of landform shape (classic examples: Horton, 1945; Strahler, 1950; Rapp, 1957; Chorley, 1959; Yasso, 1964). Recent technological advances for collecting surface topographic data and subsurface assemblages have led to a revival of geomorphometric analyses (Pike, 2000). However, the numerical characterization of form remains in its infancy despite geomorphometric studies that have begun to more closely approximate the 2.5D and 3D form of geomorphic features (Schimdt et al., 2000). Geomorphologic research must move beyond the oversimplification of form pervaded by qualitative assessment and planimetric studies of shape and focus on form quantification in higher-dimensions (2.5D to 4D). The field of geomorphometry is burgeoning with new experimental designs and digital techniques to quantify geomorphologic forms and patterns (Pike, 2000). Recent developments in global positioning systems (GPS) and other advanced remotely sensed data technologies have revolutionized high-resolution topographic data production required for form quantification

1

(Pike, 2000). Innovative technologies hold the greatest promise for production and analysis of fine-scale elevation data (Pike, 2001). Geomorphometry will not reach its potential without highresolution data, but continual emphasis on the development of advanced technologies runs the risk of technological progress at the expense of scientific knowledge (Schmidt and Dikau, 1999). This is already evident in many university curricula as well as research projects that utilize computer-based technologies not grounded in theoretical and conceptual ideas (Brown and Olson, 2001). A logical progression for this integrative task begins with quantitative characterization of landforms. This step assures consistent names and similar quantities will be assigned to analogous forms. The precise morphometric data can be combined with quantitative studies of processes to enhance process-form relationships and models. Alluvial fan form analyses have only occurred relatively recently. Lecce (1990) provides an overview of the considerable research in fan morphometry that transpired during the 1960s. Planimetric drainage area vs. fan area relationships were derived in numerous studies with results leading to gross 2-D morphometric interpretations (e.g. Bull, 1964). Recent research on morphologic parameters has combined processes with planimetric shape measurements (Mukerji, 1990; Sorriso-Valvo et al., 1998) or through the analysis of modeled data (De Chant et al., 1999). A promising area of research is automated landform extraction (Giles, 1998; Sowmya and Trinder, 2000). Miliaresis (2001) has used a combination of remote sensing data and digital elevation models to extract fan forms from a bajada. However, despite recent advances, numerical characterization of fan morphology is not equated with repeatable, measurable attributes that express the nature of the 2.5D and 3D form (Pike, 1995). The qualitative observations and planimetric shape analyses of fans has limited geomorphic research and modeling (see De Chant et al., 1999). It also has inhibited work of other scientists and engineers reliant on boundary conditions supplied by geomorphic data. Objectives. The main objective is to quantify external (2.5-D) alluvial fan structure using data derived from digital elevation model (DEM) and spatial analytical techniques. Schmidt and Dikau (1999) proposed a conceptual framework whereby aggregation of geomorphometric point (basic building block for all morphometric studies and consists of x, y, and z values) data can be used to develop morphometric structures. Structure arises from a quantitative understanding of the spatial arrangement of point data and represents a numerical signature of the topographic form in specific and general geomorphic features. A combination of global positioning systems real-time kinematic surveying (GPS-RTK) and remotely sensed topographic data will provide x, y, and z data for the development of DEMs. GISys will serve as a tool to assist in extracting form parameters and uncovering spatial patterns. Derivation of external structure in this manner presents a way to produce a numerical signature of 2.5D fan form (Pike, 1988; 1995; Schmidt and Dikau, 1999). To date, the use of geomorphologic (geomorphometric) signatures has been limited to: classifying landforms (McDermid and Franklin, 1995; Giles, 1998); automated simulations for extracting landforms (Chorowicz et al., 1989); and as an exercise to derive distinct glacial erosion landforms from DEM (Mackay et al., 1992). The current proposed research intends to utilize the conceptual framework for establishing structure and move beyond classification and extraction to quantitatively compare and contrast alluvial fan form. The investigation of fan form by means of morphometric structure will permit new insights into geomorphic concepts and ‘meta-theories’ as well as impact current knowledge

2

of fan evolution. Research within the current fan dynamics paradigm (Lecce, 1990) has assessed controls of alluvial fan sedimentation (e.g., Harvey, 1988; Blair, 1999). Only cursory linkages have been made between sediment supply, erosion, deposition, and fan morphometry (Calvache et al., 1997; Mills, 2000). A second objective will be to use 2.5D fan structure to provide new insights into fan morphometry as it relates to depositional events leading to external fan form. The 2.5D structure of alluvial fans will be related to the internal spatial arrangement of depositional environments. Ground penetrating radar (GPR) and GISys will be used to assess relations between fan internal structure and surface morphometric characteristics. The process and form relationships identified from the GPR and surface structural analyses will also provide a more concise understanding of how environmental conditions (humid vs. arid fans) impact fan structure formed from similar processes. The humid fan literature, not unlike the arid fan research, is replete of quantifiable analyses of anything except planimetric shape. Humid fan studies have been caught up in searching for similar depositional processes to arid fans. For example, comparisons made between arid and humid fans from the most often cited literature (Kochel and Johnson, 1984; Kochel, 1990) were based on planimetric measures and field measurements of texture and stratigraphy of the humid fans. While there is no debating similar formative processes can occur in humid and arid conditions, there has not been quantifiable evidence for similarities in fan form or its relation to the spatial arrangement of depositional conditions. A third objective will use external and internal structure relations to quantitatively assess fan from in various environmental settings. The final objectives of this proposed research are to first use 2.5D fan structure to test the ergodic hypothesis and second to expand the key concept of drainage area to fan area relationships. The use of ergodicity in geomorphology is often assumed and expressed by phrases such as "location-for-time" or "space-for-time" substitution (Thorn, 1988). The numerical external structure of varying aged alluvial fans can be used to test this hypothesis and determine mathematically whether this often assumed concept is valid. Finally, external fna structure will be used to present new insights with regard to drainage area vs. fan area relationships. Research on this topic has focused on the entire fan surface (e.g. Hooke and Rohrer, 1977; Lecce, 1991). However, the active connection occurs between the recent fan segment and the drainage basin rather than the entire fan surface. The structure of recent fan segments in bajadas will be compared with the structure of the drainage basin to examine links between current drainage basin form and most recent fan deposits. Significance. Quantification of alluvial fan form and its relation to formative processes are topics with both scientific and applied merit. Advancement of geography and geomorphology as fields of science depends on a synergism of form and process (Dorn et al., 1991; Bauer, 1996). A true numerical characterization of 2.5D and 3D form has not been accomplished. Many studies claiming to assess form have only scratched the surface by examining planimetric (2D) shape (boundary conditions). The construction of numerical fingerprints plays a key role in unraveling a characteristic form(s) of alluvial fans. Furthermore, it has direct applications to other geomorphic features. The use of an arithmetical signature will provide a repeatable and comparative value that will advance the understanding of form in geographic and geomorphic science (Pike 1995; Schmidt and Dikau, 1999).

3

Research with the capability to produce accurate form information can be applied to improving conditions of human settlements and protect them from natural hazards (Pike, 1995). For example, urban environments have expanded onto fan surfaces, which are produced under dynamic physical conditions. Inhabiting these environments places people at risk and puts a burden on society to pay for damages from flooding. Science is beginning to unravel some of the causal mechanisms associated with flooding and channel avulsion, but it does not know exactly how they link to fan morphometry. Therefore, precise information on alluvial fan form will supply better tools for hazard mitigation techniques and regulations. It also creates factual knowledge base from which civil engineers can develop safer building pads and structures. PRESENT STATE OF KNOWLEDGE AND FUTURE RESEARCH NEEDS Fan Morphometry. It’s widely recognized source area characteristics (lithology, sediment supply, climate, and tectonics) are a major controlling factor in fan morphology (e.g. Blair, 1999). This recognition led researchers to hypothesize fans with similar catchment attributes exhibit comparable gross morphology, which has been an underlying theme in fan morphometry research (Harvey, 1997). Fan morphometric parameters that have received considerable attention are shape, area, and slope. Several researchers have attempted to describe fan shape accurately using simple equations (Bull, 1964; Hooke, 1968). Many studies of fan shape have relied on planimetric analyses using area or topographic profiles (e.g. Denny, 1967; Hooke and Rohrer, 1979). Recent studies have continued to perpetuate this approach and have produced similar findings whereby fan morphology was controlled by constructive processes and closely tied to basin morphology (Sorriso-Valvo et al., 1998). The most promising approach from planimetric analyses has been a little cited work on the use of radii to describe fan morphometry (Bull, 1968). Radii equations, originally designed by Troeh (1965), provide a close approximation of the 2.5D fan form. The most widely examined morphometric parameter has been fan area and its relationship to drainage basin area (e.g. Lecce, 1988; Whipple and Traylor, 1996). These empirical studies have led to several common associations whereby larger fans tend to have larger drainage basins and smaller fans have smaller drainage basins. This trend is consistent across various environmental settings (Mills, 1983; Kochel and Johnson, 1984), but some regional variability exists (Oguchi and Ohmori, 1994). Lecce (1991) discovered large fans in the White Mountains of California and Nevada were produced by drainage basins formed from resistant bedrock and smaller fans were developed in areas where bedrock was less resistant. Hooke and Rohrer (1977) described a similar pattern whereby fan area was determined by varied erosion rates. The investigation of slope in conjunction with fan morphometry has produced important findings with regard to the spatial distribution of energy and sediment transport capacities (Harvey, 1997). Fan slope was inversely related to drainage basin area (Bull, 1964). Denny (1965) found a correlation between declining fan slope and decreasing particle size in a down fan direction. Whipple and Dunne (1992) and Whipple et al. (1998) found fan slope was sensitive to physical controls of sediment type and flux, water discharge, and sediment transport mechanisms in the basin area, while Milana and Ruzycki (1999) have revealed similar trends postulating fan slope was be a function of sediment transport efficiency.

4

It is evident in three major research topics that alluvial fan form(s) has not been truly represented or quantified. Many of the investigations have relied on quantification of point data, transects, or 2D shape (boundary) as opposed to 2.5D and 3D form. The PI believes this has hindered studies of process-form relationships and reduced the potential to accurately model fan development (see De Chant et al., 1999; Coulthard et al., 2002). There is a clear need to develop geomorphometric structures to present an accurate representation of form (Schmidt and Dikau, 1999). The proposed development of fan structures will meet this need and present a methodology for characterizing form in other surface features. Evolution of Alluvial Fans. Several authors have provided contextual overviews of models used to unravel the complex history of fan evolution (e.g. Mukerji, 1990; Dorn, 1996; Blair and McPherson, 1994a,b; Mills, 2000). Many existing models are underlain by facies models and/or based upon interpretations of ancient alluvial fan sequences. However, facies models are unproven and no sound evidence exists for using ancient sequences to interpret recent processform relationships (Blair and McPherson, 1994b). The lack of a sufficient model to explain modern alluvial fans has prompted Blair and McPherson (1994a) to propose a four-stage alluvial fan evolutionary scenario. A precursor stage begins with talus cone development, but not all talus cones progress to form alluvial fans. The development of an incipient fan (stage 1) occurs when debris-flows, rock avalanches, or other gravity slides add material to the talus creating a more fan-like feature. Stage 2 involves development of a composite fan morphology characterized by gentle radial slopes. The final stage consists of fan progradation; characterized by radial enlargement of the fan and unique development of an incised channel that confine subsequent flows, forcing deposition at distal fan locations. Mills (2000) used a similar framework to examine fan evolution in humid regions, but the fans he examined did not conform to Blair and McPherson’s model (1994a). However, some similarities existed. The earliest stage of development appears to be the formation of an incipient fan. Most fans in the Appalachians formed away from mountain fronts and lacked talus cones. There was evidence for the production of large boulders in Appalachian drainage basins and this mechanism was perceived as important fan development and maintenance. The emplacement of large boulders would be analogous to a precursor stage and stage 1 of Blair and McPherson (1994a). The propagation of active depositional lobes outwards was uncommon to Appalachian fans and Mills (2000) proposed four alternative scenarios of fan evolution: telescopic pattern (Dorn, 1988); concentric pattern (Bull, 1964); random pattern (Hack and Goodlett, 1960); and lateral stream migration pattern (Mills and Allison, 1994). The aforementioned conceptual models of fan evolution require quantitative testing of the 3D spatial arrangement of depositional units and their relationship to surface morphometric characteristics. GPR has been used in a number of other geomorphic features to provide this type of information, but has only been undertaken twice on alluvial fans (Mills and Speece, 1997; Eckes and Hickin, 2001). Mills and Speece (1997) found GPR penetration (1-25m) was highly variable in the Appalachian Piedmont because of the debris flow origin led to high clay contents, which made the detection of internal fan structures difficult with GPR methods. Eckes and Hickin (2001) reported GPR achieved over 40m of penetration with a 50 MHz antenna and allowed the extraction of 10 radar facies. Radar facies indicated a massive, matrix-rich diamicton dominated lower units, with sheetflood gravel deposits to depths of 20-25m above these coarser

5

units (Eckes and Hickin, 2001). Although Mills and Speece (1997) met with less success than Eckes and Hickin (2001), there are indications that with a rigorous approach GPR can used to assess internal spatial patterns of fan deposits. The anticipated GPR results will provide new insights into fan evolution and produce a novel linkage between surface and subsurface data. The findings will produce the most comprehensive understanding of fan form and depositional units. Regional Environmental Controls. Much of the alluvial fan research has concentrated on arid and/or tectonically active regions (Blair and McPherson, 1994a). However, other humid regions contain significant amounts of research: Japan (Saito, 1988, 1999; Ono, 1990; Oguchi and Ohmori, 1994); British Columbia (Ryder, 1971a, b; Church and Ryder, 1972; Kellerhals and Church, 1990; Eckes and Hickin, 2001); and the Appalachian Mountains (Mills, 1982, 1987; Kochel and Johnson, 1984; Whittecar and Duffy, 1992). Regional climatic differences spurred debates over environmental controls on fan evolution. First hypothesized by Blackwelder (1928), it was widely believed that formative processes were different for “dry” (arid and semi-arid) and “wet” (humid) environments. Dry fans were characterized by debris-flow processes, whereas fluvial processes dominated humid fans. Numerous studies have refuted this idea (Kochel and Johnson, 1984; Whittecar and Duffy, 1992; Blair, 1999). However, Kochel (1990) discussed how Appalachian fans were often found in distinctive “interbasinal” or topographical confined settings. New data provided by Mills (2000) have permitted a reassessment of Kochel and Johnson’s (1984) work and supplied valuable insights into the role of regional variability in fan dynamics. Similarity was found in fan size of the Appalachian Mountains and those in arid regions, but there was a propensity for humid fans to be smaller. Mills (2000) also identified similar slope gradients between humid and arid regions. Drainage area to fan area relationships were similar for humid and arid environments. The research conducted to date has been based on qualitative observations, planimetric measures, and stratigraphic relations between fans in different environmental settings. There have been no studies conducted that used quantifiable measures of form to determine if relationships exist. There is a definite need to reassess relations between “humid” and “arid” fans constructed by similar formative processes. A combination of 3D GPR data and morphometric structure measures can be used to address this issue. Ergodicity. A common assumption in geomorphic research has been spatial distributions of landforms reveal information about their temporal evolution (Thorn, 1988). This is the classic ergodic hypothesis described in the methodology of many studies as substituting space for time. Ergodic properties can be attributed to landforms having simultaneously a certain spatial and temporal variability as well as certain spatio-temporal independence, such that the observation of numerous landforms over large spatial areas can overcome a short-time scale of field measurements (Paine, 1985). The ergodic hypothesis permits the use of a suitable spatial analysis as a substitute for temporal analyses of landform development that likely exist beyond an individual’s lifetime. Geomorphologists have made very few attempts to study this hypothesis quantitatively. A more rigorous study of the ergodic hypothesis in geomorphology is needed to provide a mathematical interpretation of space and time relationships. Alluvial fans because of their ubiquitous nature and the close association of fans of vary sizes lend themselves to testing this hypothesis. The

6

proposed numerical characterization of fan structure can be used to compare fans of varying sizes and further the understanding of the use of ergodic assumptions in landform analysis. Drainage Basin and Fan Segment Relations. As previously described in the fan morphometry section, numerous studies examined fan area vs. catchment area relationships. The goals addressed by past research have been to show long-term evolutionary links between fans and their source areas. This represents a relevant and logical endeavor, but does not characterize recent fan development. The recent catchment connection in many alluvial fans corresponds to the active fan segment (Denny, 1967, Dorn, 1988; Milana, 2000). A large portion of the fan surface is not actively connected to the drainage basin by current processes. Older surfaces often evolve through secondary processes (Blair and McPherson, 1994a). Climate variability assessments of fan evolution have shown this link, but not provided a quantifiable measure of the resulting form (Lustig, 1965; Wells et al., 1990). Ritter et al. (2000) provided an example of connectivity between drainage basins and fans from examples in the Buena Vista basin of Nevada. The authors proposed a drier period led to sediment deficient discharges from the basin, which entrenched proximal fan deposits and transported them downslope to form a secondary fan. The changing climate altered drainage basin dynamics generating a secondary fan, which would indicate current catchment dynamics are closely related to the most recent fan segment. The relationship between a recent fan segment vs. drainage basin structure will provide new insights into the evolutionary processes of fans and their catchments. Hooke and Dorn (1992) have shown long-term drainage basin changes have a definitive impact on the fan equilibrium slope, which led to fanhead incision and downfan deposition (segmentation). However, no study has assessed segment form vs. drainage basin form relationships. This line of research could also be expanded to understanding how segmentation varies between fans formed from different sedimentary conditions, dependent on outcome of the initial proposed research on 2.5D structure. SPECIFIC RESEARCH HYPOTHESES TO BE TESTED The PI believes advancement of geomorphic research depends upon quantitative analyses of form, which can be melded with existing mathematical interpretations of processes. Progression of this research relies on the development of a way to assess morphometric structure(s) and the use of this information to test key conceptual and theoretical constructs in geomorphology. This will be accomplished by addressing the following specific research hypotheses: 1. The initial phase of research investigates the structure of fans generated from various formative processes. It is hypothesized coarse-scale fan structures should reveal similarities regardless of formative processes. This finding would be in line with the concept of simplexity (topologic and ergodic equifinality), whereby features exhibit coarse-scale geometric and topologic structural dimensions that are independent from local-scale disorder and complexity (Culling, 1987, 1988; Phillips, 1997). 2. A critical component for interpreting form is knowledge of 3D depositional patterns as they relate to surface topography. Data gathered with GPR and the 2.5D surface structural data will be used in combination to assess the topographic response to spatial arrangement of subsurface deposits. The GPR work conducted by Eckes and Hickin (2001) identified stage 1 and stage 2 of the Blair and McPherson model (1994a) were present in a fan in British

7

Columbia. These results lead to the hypothesis that magnitude of the surface structures should be higher near the fanhead and midfan, while lower magnitudes should be found at the lateral extents. It is also likely the largest changes in surface structure will exist at the boundary between multiple depositional units. 3. The lack of quantifiable studies of form in different environmental settings provides an opportunity to expand the research hypothesis tested in #2. The GPR data and 2.5D structural data will be used to compare findings from #2 with new data in humid settings to test if the relationship holds for different environmental settings. The PI hypothesizes the relationships will not hold between the two environmental settings. The hypothesis is based on differences in models proposed by Mills (2000) and Blair and McPherson (1994a). It is further supported by evidence of limited internal structure identified in humid fans (Mills and Speece, 1997). 4. The ergodic hypothesis has been assumed to exist in many geomorphological studies. The assessment of alluvial fan structure will permit a quantitative verification of this hypothesis. 5. A new assessment of drainage basin vs. fan relationships will be broached by examining the drainage basin link to the most recent fan segment. The research will test the hypothesis that drainage basin size is inversely related to the fan segment size. This hypothesis is based on findings by Gomez and Mullen (1992) who have provided experimental modeled data that indicated sediment discharge declined exponentially with time. Sediment yield and budget studies have provided evidence that as drainage basins become larger, runoff becomes more localized in the basin, producing episodic sediment delivery (Cooke et al. 1993). Renard (1972) found that localized deposition becomes the norm in basin evolution because of high transmission losses in channels within the drainage basin. Furthermore, catchment slopes lessen with time and further reduces sediment yields (Phillips, 1990). METHODS OF INVESTIGATION Study Area. A central focus for this research will be Death Valley, CA and surrounding locations. These sites consist of Basin and Range topography created during regional extensional plate motion. In Death Valley, normal faulting has created the Funeral Mountains and the Black Mountains on the valleys eastern flank and the Panamint Range on the western margins. There is a greater tendency for alluvial fans on the eastern side of Death Valley to not coalesce, unlike fans on the western side. The fans on the eastern side provide an opportunity to establish distinct boundaries between individual fans, while the western side fans provide an opportunity to investigate fans in a bajada as well as tributary fans that form at proximal fan locations on the bajada surface. Debris flows, sieve flows, ephemeral stream flows, rock avalanches, and a combinations thereof have led to fan development. The multiple processes provide an opportunity to sample a wide variety of fans. The PI has research and field experience in Death Valley. Other proposed sites include Eureka Valley, Panamint Valley, and Saline Valley. The study will use data collected from fans in the Appalachian Mountains and fans located in Japan. The Appalachian Mountains are 10-12 hr drive from the Univ. of Memphis and will serve as a site for field research in the PI’s course(s). Common locales of fans in the eastern United States were where tributary streams join a master stream (Mills, 2000). The Blue Ridge Mountains, the Valley and Ridge province, and the Shenandoah Valley were often cited locations of fans in this region (Kochel, 1990; Whittecar and Duffy, 1992). The formative processes of these fans consist mainly of debris flows, but several large fans were dominated by

8

fluvial processes (Mills, 2000). Professor Oguchi’s extensive research in Japan will assist in developing study sites in the Kitakami, Yamagata, and Matsumoto regions. The majority of Japan’s alluvial fans are located in intermontane basins and coastal plains (Oguchi and Ohmori, 1994). Tectonic uplift and volcanic activity have produced rapid uplift and high rates of sedimentation. Many of the fans were fluvially dominated with intermixed debris flow processes. Sampling Terrain Data. The research begins with field measurements to collect raw data required for morphometric analyses. Three different tools will be used to acquire elevation data: (1) GPS surveying; (2) Airborne Laser Swath Mapping (ALSM); and (3) USGS 10m DEM. One or a combination of these techniques will be used to generate terrain data. The GPS surveying equipment will assist in developing high-resolution DEMs. Pearson and Wasklewicz (in progress) have used a method to sample breaks-in-slope along the fan length. At each break-in-slope, a high density of points were sampled across the fan width. After all of the break-in-slopes were sampled, a 5m grid of points over the entire fan surface and 20m beyond the fan boundaries was sampled to provide a greater number of points for grid interpolation. We attained accuracies of +5 to +10cm on the vertical and +1cm on the horizontal with the GPS unit. GPS surveying techniques will be used to sample fans less than 500m in length and width. Fan boundaries will also be established with GPS. A sample size of 20 to 25 alluvial fans will be evaluated for the study of fan structure. The GPS techniques will also be used to establish the topographic data in the combined GPR and structure analyses. The ALSM data collection will be farmed out to University of Florida, Geosensing Systems Engineering Center. ALSM is a system that emits rapid pulses of laser light to precisely measure distances from a sensor mounted in a port opening on the bottom of an aircraft's fuselage to the ground surface. A scanning mirror or prism is used to direct laser pulses back and forth across a wide swath underneath the airplane path. Swath width varies with aircraft altitude and scan angle. Horizontal accuracy and location (x, y coordinates) of the elevation point is determined by a GPS on board the aircraft. The GPS antenna is mounted directly overhead of the laser. The combination of onboard GPS and laser data with the ground control points should produce horizontal accuracies of less than 0.5m and vertical accuracies of 0.5m. The data will be used to sample 30 to 45 fans for the erogodicity study. The USGS 7.5-minute DEMs will be constructed through interpolation of digital line graph (DLG) hypsographic and hydrographic data to provide 10m resolution data for Death Valley, Eastern Panamint Valley, and Saline Valley. These data will be stored as profiles with a 10m square grid spacing along and between each profile. A UTM projection system will be referenced to the North American Datum of 1983 (NAD 83). This is the coarsest resolution topographic data acquired for this proposed study, but is essential because of the volume of data required to map a fan segment and its drainage basin. The resolution of ALSM is to fine and would be costly while GPS surveying would require an eternity to cover the area proposed for the USGS 7.5-minute DEMs. The 10m resolution data will used to compare 40-50 fan segments and their relations to drainage basins. The fans and drainage basins will be extracted from the DEM using on-screening digitizing. A subset of the digitized data will be tested against GPS measurements from the field.

9

Geomorphometric Structure Data. The structure data will be developed and used in each research phase. Morphometric structure data will begin with basic point data that will be kriged and stored in a raster DEM within a GISys. A series of geomorphometric algorithms included in GIS software, designed by the author, and developed by other researchers will be used to extract DEM derivatives (slope, aspect, curvature, topographic variability, etc) in a cell data format that can be assimilated to produce structure(s) (Schmidt and Dikau, 1999). The structures represent homogenous zones and will be used uncover patterns in surface topography. Development of geomorphometric structures will be accomplished by standardizing parameter data in the different ArcGIS layers and using overlay techniques to produce a composite measure of terrain (Pike, 1988; 2000). The layer produced by the overlay techniques will be analyzed using echelon analysis, which can be used to quantitatively determine spatial organization of the topographic structure of a single fan and then this data can be used for comparison between fans (Myers and Patil, 2002). Echelon analysis divides terrain into structural entities consisting of peaks, foundation of peaks, and foundations of foundations using organizational recursion. Algorithms identify peaks and local plateaus first, which represent a series of structural entities. These structural entities are numbered from 1 (highest peak) and then moving outward and downward the numbering increases by a single value. Eventually the algorithm reaches areas that yield a change in the response variable towards increasing values, which delineates the boundary of first order echelons (a foundation). Foundations represent a secondary series of entities, which are numbered in order of increasing value from the top level, starting with the next number after that assigned to the lowest peak. This process continues to produce echelons as long as data are present. Echelon characteristics (9 parameters that formalize the structural complexity of the surface) are recorded in a table and can be statistical analyzed to compare fan structures. GPR Sampling Design. GPR studies of alluvial fans in different environmental settings have shown potential for acquisition of subsurface depositional data (Mills and Speece, 1997; Eckes and Hickin, 2000). However, there remains a need to verify GPR interpretations in other locations. The initial phase of GPR data acquisition and analysis will be developing several experimental lines on tributary fans in Death Valley. Tributary fans form where tributary streams debouched onto the surface of a larger fan. Many tributary fans have been completely incised because of base-level changes resulting from fanhead entrenchment on the larger fan. Down cutting has exposed the tributary fan depositional sequence and provides an opportunity to calibrate GPR data with the different exposed stratigraphic layers. Experimental lines will be examined with different antenna frequencies, various gain control techniques, and 400V and 1000V transmitters. Analysis of the experimental lines will be used to determine the dimensions of the 3D grid. It is anticipated that 3 to 5 fans will be investigated using 3D GPR sampling techniques. A small sample size will be used because of the time required to develop 3D radar facies. GPS surveying will also be conducted to produce each fan’s morphometric structure. Three-dimensional GPR consists of collecting data along closely spaced (usually less than 1 meter) grid lines established on the fan surface (Junck and Jol, 2000). The lines of EM data are then composed into a 3D data volume that can be observed from any angle using any data subset. The 3D data will be post-processed using software supplied with the GPR equipment. 3D analytical software will be used to extract attributes from radar imagery. The Memphis Center

10

for Advanced Spatial Analysis (MCASA, directed by Wasklewicz) has been granted a license for Landmark software and Wasklewicz is learning the nuances of this program and other aftermarket 3-D visualization software (Fortner Slicer, Rockware 3-D Seismic, and Geographix Seismic package). The major focus in attribute analysis will be to quantify variations in amplitude, frequency content, phase of the reflected waveform, as well as volumetric measures. These quantities will be sliced horizontally to produce layers transferable to ArcGIS and correlated with a geomorphometric structure layer. Similar techniques will be used for assessing structure data in Japan to compare fans from different environmental settings. Ergodicity Analysis. Relative age dates will be assigned to several alluvial fans generated from similar formative processes. The relative age dating will follow guidelines in Christenson and Purcell (1985). Relative age dates are used as opposed to absolute ages for two reasons: (1) truly ergodic studies are postulated from theory rather than knowing absolute ages of the actual sequence (Brunsden and Kesel, 1973); and (2) the PI believes the uncertainty and errors associated with absolute dating techniques do not warrant the cost or time expenditure required to sample materials for this proposed study. Comparison of fan morphometric structures identified from different age fans over a wide spatial area will be used to test erogodicity. DEM data for the fans will be collected using ALSM techniques. The data will be kriged using geostatistical techniques and the geomorphometric structure extracted from this data. Echelon analysis will serve as the foundation for the comparison between fans of various sizes because of the scale independent nature of the technique. The details of these steps have been outlined in the previous sections. Further support for the echelon findings will come from analyses performed using geostatistical analyses, specifically semivariance analyses. The semivariance analysis can be used to assess the spatial dependence of the morphometric data and standardized to compare fans of various sizes. Catchment and Fan Segment Relations. USGS 10m DEM data will be used in a GISys to assess drainage basins relations with recent fan segments. This research involves several tiers of analysis. The most basic analysis of the data will involve planimetric measures to assess if the hypothesized inverse relation between drainage basin size and the most recent fan segment size exist. Analyses will follow established methods for comparability (e.g. Lecce, 1991; Oguchi and Ohmori, 1994, Whipple and Traylor, 1996). The second phase of the fan segment vs. catchement relationship study is to assess drainage basin structure in relation to the fan segment structure using techniques described for studies in previous years. A third phase of this research is to assess variability of morphometric structure between fan segments throughout the region, again using methods described above. Timeline of Events. The experimental design and educational experiences outlined above creates the following tasks: (1) equipment acquisition; (2) initial site selection; (3) fieldwork to collect terrain data and control points; (4) development of databases (terrain data, morphometric variables); (5) GIS and spatial statistical analysis of digital data; and (6) dissemination of results. Year 2003

Research STRUCTURE AND SIMPLEXITY

Spring (Mar-May) 1. Sem. Geomorphology 2. Fieldtrip to CA 3. Establish field sites for Structure and GPR studies

Summer (Jun-Aug) 1. Purchase Equipment 2. Develop website 3. GIS Database 4. Initiate lit. research

11

Fall (Sep-Nov) 1. Sabbatical 2. Structure Fieldwork 3. GIS database 4. GIS analyses 5. Write papers 6. GPR equipment

Winter (Dec-Feb) 1. GPR/GIS fieldwork: experimental line 2. Analysis of GPR experimental data 3. Establish 3D Sampling Scheme

2004

GPR AND STRUCTURE

2005

GPR, STRUCTURE, & ENVIRONMENTAL SETTING

2006

ERGODICITY

2007

DRAINAGE BASIN AND FAN SEGMENT

Time expenditure

1. GPR/GPS fieldwork: 3D grid 2. Geomorph. Course 3. Fieldtrip to CA 4. Present AAG 1. Present AAG 2. Data analysis 3. Sem. Geomorph. 1. Process ALSM data 2. Fan Boundary data 3. Geomorph. Course 4. Field trip to CA 5. Present AAG 1. Fan boundary data 2. Field check data 3. Sem. Geomorph. 4. Present AAG PI 20% - 100% RA 50% - 100%

1. Process 3D data 2. Data analysis 3. Site selection in Japan 1. GPR in Japan 2. Process GPR data 3. GISys database 3. Data analysis 1. GISys database 2. GISys Analysis

1. GIS database 2. Extract Parameters 3. GISys analysis PI 100% RA 100%

1. 3D analyses 2. Comp. Map Course 3. Appalachian data (field trip) 4. Write papers 1. Write papers 2. Present GSA 3. Update Website

1. Process GPR data 2. GIS database for Appalachian data 3. Update website

1. Computer Mapping Course 2. Write Papers 3. Update website 4. Present IAG 1. GISys Analysis 2. Present GSA

1. Site selection for fan segment study 2. Acquisition of DEM data 3. Process DEM data 1. Write Papers 2. Update website 3. Begin book

PI 20% to 60% RA 20%

PI 20% - 100% RA 50% - 100%

1. Site selection for ergodic study 2. ALSM data

EDUCATIONAL MERIT AND EVALUATION Education Goal and Significance: The proposed educational aim will be to integrate innovative data collection, theory, and analytical educational experiences. The proposed study results and data will be readily transferable to the classroom and can be used to provide active learning experiences. The PI teaches undergraduate and graduate level courses in computer mapping, geomorphometry, and geomorphology. Each course will emphasize certain topics centered on conceptual and theoretical issues, but investigate them through “hands-on” projects that rely heavily on technology for data collection, analysis, and communication. The proposed education and research activities are a significant shift away from the compartmentalization of techniques and concepts found in many disciplines. A holistic approach, like the one proposed, will compel students to develop sound habits in project design, fieldwork, data collection and management, analysis, synthesis, and articulation (written, spoken, and visual). A student immersed in this learning environment can provide innovative approaches to broaching integrative subject material. The application of these systematic practices will permit students to go beyond lecture and lab to conduct publishable original field- and computer-based research because they have a clearer view of science and its applications. Effective Teaching and Learning in Geography. Healy (2000) has attempted to broach effective teaching by returning to scholarship. Two major factors in scholarly activity, knowledge of current and acceptable teaching practices in the field and assessment and evaluation reflect topics inherent to teaching effectiveness. A major theme and practice in geographic and geoscience teaching has been how to successfully include technology to promote learning (Nellis, 1994; Hall-Wallace, 1999). Research suggests technology in the classroom leads to more active learning and adventurous teaching (Healy et al., 1998; Wentz et al., 1999). Active learning is often accomplished through hands-on projects. Secondary active learning comes from student-faculty contact as well as student-student interactions as questions arise from the use of technology. Adventurous teaching results from opening new doors for the students with the technology, allowing them to explore information and get prompt feedback. It is also adventurous for the instructor because students often come across new ideas or ways of accomplishing tasks that can open avenues for class discussion.

12

The development of technological skills runs a risk of moving away from theoretical underpinnings of the discipline (Nellis, 1994; Brown and Olson, 2001). To effectively teach students the instructors must not only be a facilitator of technological information, but must also imbed theory in these experiences. Many courses across the country have jumped on the technology bandwagon, but have forgotten the critical step of incorporating scientific concepts (Brown and Olson, 2001). Several models for effectively teaching a combination of technology and theory have been proposed. Wikle and Lambert (1996) have proposed introducing technology topics into non-technology courses and integrating it vertically within the curriculum. Brown and Olson (2001) suggest an alternative by developing a new integrative course designed to bring together both concepts and techniques in GISci centered around five major themes: (1) introduction to concepts and technology of terrain analysis; (2) data collection; (3) data storage and management; (4) analysis and synthesis; and (5) visualization. The assessment of student learning (proficiency) is perhaps the most important medium for communication between teacher and learner. Assessment is not purely a measuring device rather it serves to enhance an active learning environment by providing vital feedback to students. Many instructors in higher education, constrained by large class sizes or by too many commitments, tend to rely on predetermined grade distributions; a “norm-referenced” approach. The use of a normal distribution curve does not properly address student proficiency, but rather forces a set of scores or set of grades into a normal distribution predetermining student outcome. Marzano’s (2000) observations endorse student assessment using learning objectives. Student products and performances are assessed based upon a detailed set of criteria that are linked to standards clearly defined by the instructor. The “criterion referenced” assessment system sets benchmarks that define acceptable or unacceptable levels of performance in concrete and meaningful ways to gage a student’s level of proficiency. This method has begun to infiltrate the geographic discipline (NCGIA, 1999). The main benefits of criteria based standards are: (1) they set clear targets and expectations for students; (2) students improve their ability to monitor and evaluate their work and assess their progress; and (3) since products and performances are ongoing, proficiency is measured through time rather than a one time measure. The key to this process is to identify criteria and then assign a scale. An example of this is the NCGIA (1999) scale that indicates awareness, competency, and mastery of GIS skills and concepts. A letter or numerical grade can be generated from detailed information presented within each assigned scale. Present Education Needs. Learning is not a spectator sport. It is the responsibility of educators to involve students in: designing original field- or computer-based research; a discussion of what they are learning; writing results from their analyses; and presenting materials to peers and professionals using visual and oral communicative skills. This process begins in the classroom by integrating technology with theory (Brown and Olson, 2001). However, the approach proposed by Brown and Olson (2001) must be taken a step further. The knowledge and skills fostered in the classroom should be expanded by having students conduct original research or by furthering their expertise in experimental design through involvement in research topics proposed by the instructor. This is major step to advancing geoscientists and promoting translation of innovative problem solving skills from the classroom to “real-world” scenarios.

13

There are also educational needs at the University of Memphis. The Department of Geography is currently being merged into a Department of Earth Sciences (DES). The newly formed department provides an excellent opportunity to generate a new approach within the curriculum where strong integrative courses will be required. The integrative nature of geography places it at the front of developing such a learning environment. The coursework/research opportunities currently proposed by the PI will advance the standing of the geographic discipline in the DES. It is imperative that geography has a strong voice in undergraduate and graduate education as well as the development of a new curriculum within this program. Educational Effectiveness. The University of Memphis uses a standard series of multiplechoice questions and occasionally peer review to evaluate teaching performance. Students are evaluated by the instructor and a letter grade is assigned. Most instructors use tests, a series of practicum, writing assignments, and/or presentations to assess the educational effectiveness. The PI proposes to develop a detailed set of standards to assess students. At the broadest level, grading will be designed around five rubrics, which represent the A, B, C, D, and F grading system at the University of Memphis. The five rubrics consist of: Mastery (an “A”), Competency (a “B”), Awareness (a “C”), Needs Work (a “D”), and Little Evidence of Meeting Standards (a “F”). Detailed criteria will be assigned to the five rubrics and outline exact expectations of the student with regard to technology, skills, participation, presentation, and conceptual knowledge. A majority of the accountability for educational effectiveness lies in an assessment of the instructor. Beside the standard assessments used by the University, the PI proposes to have students present their final projects to a group of professionals. The professionals will include other faculty members, scientists from government agencies, and scientists from consulting agencies. The professionals will be asked to judge the merit of the student projects and to judge the effectiveness of the instructor in light of how well the students meet or exceed the standards. This approach will foster feedback and interactions on many levels for the students and the instructor. This approach will also permit students to network with future employers. RELATION OF PROPOSAL TO PI’S GOALS/RESPONSIBILITIES AND INSTITUTIONS GOALS AT THE UNIVERSITY OF MEMPHIS The PI was hired to bridge the gap between geographical techniques and theoretical geography. He is expected to provide learning and research experiences that integrate fieldwork, technology, and analytical methods for undergraduate and graduate students. The CAREER funding would support this goal by funding some of the technological infrastructure to make these integrated experiences possible. Funding student assistantships would assist underrepresented students in completing theses using integrative techniques, which would indicate the PI is fulfilling academic as well research responsibilities. Support of the research and education experiences would also foster scholarly publications. This is a major goal for the development of PI’s career. There is the potential for 10 to 12 peer-reviewed scientific and education publications. The PI would eventually produce a book on landform analysis built on current proposed studies. The PI’s research is expected to contribute to local, regional, national, and international understandings of how landforms are created and how they will change in the future. The PI has been conducting funded research with students and other faculty members on local rivers. The

14

next step is to develop a research niche nationally and internationally. The Career Grant would provide a major resource towards this goal. As proposed, the research experiences provide innovative twists on the burgeoning field of geomorphometry and provide new insights in geomorphology. Publications, conference presentations, and Internet retrospectives of the research would bring recognition on a national and international level. Presenting study results at national and international meetings will develop connections with other researchers and students. As an urban campus with a diverse community, the University looks favorably on projects that will enhance diversity in the composition of the student body. Geography and the geosciences have been attempting to increase enrollment of women, minorities, and individuals with disabilities in science, but there is room for improvement. The PI intends to get underrepresented students participating in research and assisting them by fostering a dynamic student-centered learning environment. Students from these groups need the structure proposed by the current project design as well as the research tools and financial support. The PI will make a point to use RA funding to employ underrepresented students. The proposed research sites for this study would also expose students to environments that they have never experienced. The travel to different environments alone would support incredible learning experiences. SUMMARY OF PRIOR RESEARCH AND EDUCATION ACCOMPLISHMENTS Wasklewicz has been conducting research in geomorphology. Specific research includes studies on sediment production, sediment transport and deposition, and interactions of sediment dynamics with biotic and abiotic features. Initial research examined the impact of environment on mineral weathering in basalt flows passing through different ecotones in Hawaii. Results indicated mineral weathering sequences, specifically the often-cited Goldich mineral susceptibility sequence, did not occur in Hawaii’s savanna and semiarid regions (Wasklewicz, 1994). A provenance study of sand dunes, a critical habitat for endangered lizard species in California, linked fluvial sediment with aeolian dunes using geochemical analyses of the sediment (Wasklewicz and Meek, 1995). Riparian vegetation patterns are dependent upon flood disturbance mechanisms. A major component of flooding is sediment erosion and deposition, which generate or remove geomorphic surfaces for vegetation. Recent articles have been published examining these interactions from central Arizona (Wasklewicz, 2001) and along channelized sections of the Mississippi River (Wasklewicz et al., 2001, in press). The most outstanding education accomplishment for the PI to date was being honored with an Arizona State University Teaching Assistant Excellence Award. This award was given in recognition of outstanding teaching service in the academic community at ASU. The PI's teaching skills are evolving through experiences at two visiting instructorships (Texas A & M Univ. & Colgate Univ.) and as a tenure-track member at Memphis. The PI has a history of developing electronic lectures, note, study guides, and projects for his classes. Evaluations of these approaches have shown students write less in class and can discuss in detail, relationships, and connections of the Earth’s processes and forms. Students from the computer mapping courses have found the experience permits them ask specific questions about techniques and concepts beyond materials discussed in class. The approach presented in this grant is a logical transition from past teaching/research experiences and will lead to a fruitful career grounded in innovative research and teaching.

15

References Cited Bauer, B.O. 1996. Geomorphology, Geography, and Science. In Rhoads, B.L. and Thorn, C.E. (eds.) The Scientific Nature of Geomorphology, Wiley, Chichester, p. 381-413. Blackwelder, E., 1928. Mudflow as a geologic agent in semi-arid mountains. Bulletin of the Geological Society of America, 39: 465-484. Blair, T.C., 1999. Cause of dominance by sheetflood vs. debris-flow processes on two adjoining alluvial fans, Death Valley, California. Sedimentology, 46: 1015-1028. Blair, T.C. and McPherson, J.G., 1994a. Alluvial fan process and form. In A.D. Abrahams and A.J. Parsons (eds.), Geomorphology of Desert Environments, Chapman and Hall, London, p. 354-402. Blair, T.C. and McPherson, J.G., 1994b. Alluvial fans and their natural distinction from rivers based on morphology, hydraulic processes, sedimentary processes, and facies assemblages. Journal of Sedimentary Research, Section A: Sedimentary Petrology and Processes, 64: 450-489. Brown D.G. and Olson J.M., 2001. Integrated teaching of geographic information science and physical geography through digital terrain analysis. Journal of Geography, 100: 4-13. Brunsden, D. and Kesel, R.H. 1973. Slope development on a Mississippi River bluff in historic time. Journal of Geology, 81: 576-598. Bull, W.B., 1964. Geomorphology of segmented alluvial fans in western Fresno County, California. U.S. Geological Survey Professional Paper, 352-E: 89-129. Bull, W.B., 1968. Alluvial fans. Journal of Geological Education, 16: 101-106. Calvache M.L., Viseras, C., and Fernandez J. 1997. Controls on fan development - evidence from fan morphometry and sedimentology; Sierra Nevada, SE Spain. Geomorphology, 21: 69-84. Chorley, R.J. 1959. The shape of drumlins. Journal of Geology, 3: 339-344. Chorowicz, J., Kim, J., Manoussis, S., Rundant, J.P., Foin, P., and Veillet, I. 1989. A new technique for recognition of geological and geomorphological patterns in digital terrain models. Remote Sensing of Environment, 29: 229-239. Christenson, G.E. and Purcell, C. 1985. Correlation and age of Quaternary alluvial fan sequences, Basin and Range Province, southwestern United States. Geological Society of America Special Paper 203, 115-122. Church, M. and Ryder, J.M. 1972. Paraglacial sedimentation: a consideration of fluvial processes conditioned by glaciation. Geological Society of America Bulletin, 83: 3059-3072. Cooke, R.U., Warren, A., and Goudie, A. 1993. Desert Geomorphology. UCL Press, London, pp. 526. Coulthard T.J.; Macklin M.G.; and Kirkby M.J. 2002. A cellular model of Holocene upland river basin and alluvial fan evolution. Earth Surface Processes and Landforms, 27: 269288. Culling, W.E.H. 1987. Equifinality: modern approaches to dynamical systems and their potential for geographical thought. Transaction of Institute of British Geographers, 12: 57-72. Culling, W.E.H. 1988. A new view of the landscape. Transaction of Institute of British Geographers, 13: 345-360. Davis, W.M. 1899. The geographic cycle. Geographical Journal, 14: 481-504. De Chant L.J., Pease P.P., and Tchakerian, V.P. 1999. Modeling alluvial fan morphology. Earth Surface Processes and Landforms, 24: 641-652.

Denny, C.S. 1965. Alluvial fans in the Death Valley region, California and Nevada. U. S. Geological Survey Professional Paper, 466. Denny, C.S., 1967. Fans and pediments. American Journal of Science, 265: 81-105. Dorn R.I., 1988. A rock varnish interpretation of alluvial-fan development in Death Valley, California. National Geographic Research, 4: 56-73. Dorn R.I., 1996. Climatic hypotheses of alluvial-fan evolution in Death Valley are not testable. In B.L. Rhoads and C.E. Thorn (eds.), The Scientific Nature of Geomorphology: Proceedings of the 27th Binghamton Symposium in Geomorphology, John Wiley & Sons Ltd., New York, pp. 191-220. Dorn, R.I., Dixon, J.C., and Orme, A.R., 1991. Integrating geomorphic process and landscape evolution: An editorial. Physical Geography, 12: 301-302. Eckes, C. and Hickin, E.J. 2001. Ground penetrating radar facies of the paraglacial Cheekye Fan, southwestern British Columbia, Canada. Sedimentary Geology, 143: 199-217. Gilbert, G.K. 1909. The convexity of hilltops. Journal of Geology, 17: 344-350. Giles, P.T. 1998. Geomorphological signatures: Classification of aggregated slope unit objects from digital elevation and remote sensing data. Earth Surface Processes and Landforms, 23: 581-594. Gomez, B. and Mullen, V.T. 1992. An experimental study of sapped drainage network development. Earth Surface Processes and Landforms, 17: 465-476. Hack, J.T. and Goodlett, J.C. 1960. Geomorphology and forest ecology of a mountain region in the central Appalachians. U. S. Geological Survey Professional Paper, 347: pp.66. Hall-Wallace, M.K., 1999. Integrating computing across a geosciences curriculum through an applications course. Journal of Geoscience Education. 47: 119-123. Harvey, A. M. 1988. Controls of alluvial fan development; the alluvial fans of the Sierra de Carrascoy, Murcia, Spain. Catena Supplement, 13: 123-137. Harvey, A.M. 1997. The role of alluvial fans in arid zone fluvial systems. In D.S.G. Thomas (ed.), Arid Zone Geomorphology: Process, Form and Changes in Drylands, John Wiley & Sons Ltd., New York, pp. 231-259. Healey, M., 2000. Promoting lifelong professional development in geographical education: developing the scholarship of teaching in higher education in the 21st century. Invited paper presented to the plenary session of the 29th International Geographical Congress, on "Developing Geographical Educators for the 21st Century”, Seoul, Korea. Healey, M., Robinson, G. and Castleford, J., 1998. Developing Good Educational Practice: Integrating GeographyCal into University Courses. In: Proceedings of the Institute of Australian Geographers and New Zealand Geographical Society Second Joint Conference, Hobart, Australia, Department of Geography, The University of Waikato. Hooke, R. LeB., 1968. Steady-state relationships on arid-region alluvial fans in closed basins. American Journal of Science, 266: 609-629. Hooke, R. L. and Rohrer, W. L., 1977. Geometry of alluvial fans; effect of discharge and sediment size. Earth Surface Processes, 4: 147-166. Hooke, R.L. and Dorn, R.I. 1992. Segmentation of alluvial fans in Death Valley California: new insights from surface exposure dating and laboratory modeling. Earth Surface Processes and Landforms, 17: 557-574. Horton, R.E. 1945. Erosional development of streams and their drainage basins: hydrophysical approach to quantitative morphology. Geological Society of America Bulletin, 56: 275370.

Junck, M. B. and Jol, H. M., 2000. Three-dimensional investigation of geomorphic environments using ground-penetrating radar. GPR 2000; proceedings of the Eighth international conference on Ground penetrating radar, SPIE Proceedings Series 4084: 314-318. Kellerhals, R. and Church, M. 1990. Hazard management of fans, with example from British Columbia. In A.H. Rachocki and M. Church (eds.), Alluvial Fans: A Field Approach, Wiley, New York, p. 335-354. Kochel, R.C., 1990. Humid fans of the Appalachian Mountains. In A.H. Rachocki and M. Church (eds.), Alluvial Fans; A Field Approach, John Wiley & Sons, Chichester, United Kingdom. p. 109-129. Kochel, R.C., and Johnson, R.A., 1984. Geomorphology and sedimentology of humid-temperate alluvial fans, central Virginia. In E. H. Koster and R.J. Steel (eds.), Sedimentology of gravels and conglomerates Memoir - Canadian Society of Petroleum Geologists-10, p. 109-122. Lecce, S.A., 1988. Influence of lithology on alluvial fan morphometry. White and Inyo Mountains, California and Nevada. M.A. Thesis, Arizona State University, pp. 130. Lecce, S.A., 1990. The alluvial fan problem. In A.H. Rachocki and M. Church (eds.), Alluvial Fans: A Field Approach, Wiley, New York, p. 3-24. Lecce, S.A., 1991. Influence of lithologic erodibility on alluvial fan area, western White Mountains, California and Nevada. Earth Surface Processes and Landforms, 16: 11-18. Lustig, L.K. 1965. Clastic sedimentation in Deep Springs Valley, California. USGS Professional Paper, 352F: 131-192 Mackay, D.S., Robinson, V.B., and Band, L.E. 1992. Classification of higher order topographic objects on digital terrain data. Computers, Environment and Urban Systems, 16: 473-496. Marzano, R.J., 2000. Transforming Classroom Grading, Assn for Supervision & Curriculum Development, Alexandria, VA, USA. pp. 148. McDermid, G.J. and Franklin, S.E. 1995. Remote sensing and geomorphometric discrimination of slope processes. Zeitschrift fur Geomorphologie Supplementband, 101: 165-185. Milana, J.P. 2000. Characterization of alluvial bajada facies distribution using TM imagery. Sedimentology, 47: 741-760. Milana, J.P., and Ruzycki, L., 1999. Alluvial-fan slope as a function of sediment transport efficiency. Journal of Sedimentary Research, 69: 553-562. Miliaresis, G.Ch. 2001. Extraction of bajadas from digital elevation models and satellite imagery. Computers and Geosciences, 27: 1157-1167. Mills, H.H., 1982. Piedmont-cove deposits of the Dellwood quadrangle, Great Smokey Mountains, North Carolina, U.S.A.: Morphometry. Zeitschrift fur Geomorphologie, 26: 163-178. Mills, H.H., 1987. Variation in sedimentary properties of colluvium as a function of topographic setting, Valley and Ridge province, Virginia. Zeitschrift fur Geomorphologie, 31: 277-292. Mills, H.H., 2000. Controls on form, process, and sedimentology of alluvial fans in the Central and Southern Appalachians, Southeastern U.S.A. Southeastern Geology, 39: 281-313. Mills, H.H., and Allison, J.B., 1994. Controls on the variation of fan-surface age in the Blue Ridge Mountains of Haywood County, North Carolina. Physical Geography, 15: 465480.

Mills, H.H. and Speece, M.A. 1997. Ground-penetrating radar exploration of alluvial fans in the southern Blue Ridge Province, North Carolina. Environmental and Engineering Geoscience, 3: 487-499. Mukerji, A. B., 1990. The Chandigarh Dun alluvial fans; an analysis of the process-form relationship. In A.H. Rachocki and M. Church (eds.), Alluvial Fans: A Field Approach, Wiley, New York, p. 131-149. Myers, W.L. and Patil, G.P. 2002. Echelon analysis. In A.H. El-Shaarawi and W.W. Piegorsch (eds.), Encyclopedia of Enivornmentrics (vol. 2), John Wiley and Sons, Ltd., Chichester, p. 583-586. NCGIA, 1999. GIS Core Curriculum for Technical Programs visited site 07/12/01 http://www.ncgia.ucsb.edu/education/curricula/cctp/Welcome.html Nellis, M.D. 1994. Technology in Geographic Education: Reflections and Future Directions. Journal of Geography, 93: 36-39. Oguchi, T. and Ohmori H., 1994. Analysis of relationships among alluvial fan area, source basin area, basin slope, and sediment yield. Zeitschrift fur Geomorphologie, 38: 405-420. Ono, Y., 1990. Alluvial fans in Japan and South Korea. In A.H. Rachocki and M. Church (eds.), Alluvial Fans: A Field Approach, Wiley, New York, p. 91-107. Paine, A.D.M 1985. “Ergodic” reasoning in geomorphology: time for a review of the term? Progress in Physical Geography, 9: 1-15. Pearson, M.E. and Wasklewicz, T.A., (in progress). Geomorphometry of Debris-flow Fans, Death Valley. Phillips, J.D. 1990. Relative importance of factors influencing fluvial soil loss at the global scale. American Journal of Science, 290: 547-568. Phillips J.D. 1997. Simplexity and the reinvention of equifinality. Geographical Analysis, 29: 1-15. Pike R.J., 1988. The geometric signature-quantifying landslide terrain types from digital elevation models. Mathematical Geology, 20: 491-511. Pike R.J., 1995. Geomorphometry - progress, practice, and prospect. Zeitschrift fur Geomorphologie, Supplementband, 101: 221-238. Pike R.J., 2000. Geomorphometry - diversity in quantitative surface analysis. Progress in Physical Geography, 24: 1-20. Pike R.J., 2001. Digital terrain modeling and industrial surface metrology: converging realms. Professional Geographer, 53: 263-274. Rapp, A. 1957. Studien uber Schutthalden in Lappland und auf Spitzbergen. Zeitschrift fur Geomorphologie, 1: 179-200. Renard, K.G. 1972. Sediment problems in the arid and semi-arid southwest. 27th Annual Meeting of the Soil Conservation Society of America (Portland, Oregon), p.225-232. Ryder, J.M. 1971a. The stratigraphy and morphology of paraglacial fans in south-central British Columbia. Canadian Journal of Earth Sciences, 8: 279-298. Ryder, J.M. 1971b. Some aspects of the morphometry of paraglacial fans in south-central British Columbia. Canadian Journal of Earth Sciences, 8: 1252-1264. Saito, K., 1988. Alluvial Fans in Japan. Kokon-Shoin, Tokyo, pp. 280. Saito, K., 1999. Development of mountains and alluvial fans in Japan, Taiwan, and Philippines. Geographical Review of Japan, 72: 162-172. Schmidt, J., and Dikau, R., 1999. Extracting geomorphometric attributes and objects from digital elevation models - Semantics, methods, future needs. In R. Dikau and H. Sauer (eds.): GIS for Earth Surface Systems, p. 153-174.

Schmidt J., Hennrich K., and Dikau R. 2000. Scales and similarities in runoff processes with respect to geomorphometry. Hydrological Processes, 14: 1963-1979. Sorriso-Valvo, M., Antronico, L. and Le Pera, E., 1998. Controls on modern fan morphology in Calabria, southern Italy. Geomorphology, 24: 169-187. Sowmya A. and Trinder J. 2000. Modelling and representation issues in automated feature extraction from aerial and satellite images. ISPRS Journal of Photogrammetry and Remote Sensing, 55: 34-47. Strahler, A.N. 1950. Equilibrium theory of erosional slopes approached by frequency distribution analysis. American Journal of Science, 248: 673-96, 800-814. Thorn, C.E. 1988. An introduction to theoretical geomorphology, Unwin Hyman, Boston. Troeh, F.R., 1965. Landform equations fitted to contour maps. American Journal of Science, 263: 616-627. Wasklewicz, T.A., 1994. Importance of environment on the order of mineral weathering in olivine basalt, Hawaii. Earth Surface Processes and Landforms , 19: 715-734. Wasklewicz, T.A. 2001, 'Geospatial analysis of riparian vegetation variability along perennial streams in central Arizona', Physical Geography, 22: 361-375. Wasklewicz, T.A. and Meek, N. 1995. Provenance of aeolian sediment: the Coachella Valley, California. Physical Geography, 16: 539-556. Wasklewicz, T.A., Franklin, S., Greulich, S., and Grubaugh, J., 2001. Historic stream flow periodicity along the Mississippi River and the impacts of engineered structures. 1st International River Basin Management Proceedings, Cardiff, Wales. Wasklewicz, T.A., Greulich, S., Franklin, S., and Grubaugh, J., 2001. 20th century trends in stage along the Mississippi River. Water Resources Research, (in review). Wells, S.G., McFadden, L.C., and Dohrenwend, J.C. 1990. Preliminary results of age estimations and regional correlations of Quaternary alluvial fans within the Mojave Desert of southern California. In R.E. Reynolds, S.G. Wells, and R.J.I. Brady (eds.), At the end of the Mojave: Quaternary Studies in the Eastern Mojave Desert, San Bernardino County Museum Association, Redlands, CA, p. 45-53. Wentz E.A., Vender J.C., and Brewer C.A., 1999. An evaluation of teaching introductory geomorphology using computer-based tools. Journal of Geography in Higher Education, 23: 167-179. Whipple K.X. and Dunne T., 1992. The influence of debris-flow rheology on fan morphology, Owens Valley, California, Geological Society of America Bulletin, 104: 887-900. Whipple, K. X. and Trayler, C. R., 1996. Tectonic control of fan size: the importance of spatially variable subsidence rates. Basin Research, 8: 351-366. Whipple, K.X., Parker, G., Paola, C., and Mohrig, D. 1998. Channel dynamics, sediment transport, and the slope of alluvial fans; experimental study. Journal of Geology, 106: 677-693. Whittecar G.R., and Ryter D.W., 1992. Boulder streams, debris fans, and Pleistocene climate change in the Blue Ridge Mountains of central Virginia. Journal of Geology, 100: 487494. Wikle T.A. and Lambert D.P., 1996. The global positioning system and its integration into college geography curricula. Journal of Geography, 95: 186-193. Yasso, W.E. 1964. The geometry and development of spit-bar shorelines at Horseshoe Cove, Sandy Hook, New Jersey. U.S. Naval Research Project NR 388-057, Technical Report 5, Department of Geology, Columbia University, New York, pp. 104.