ISSN 1471-1427
Proceedings of the 4th DeHaan Tourism Management Conference
“MARKETING TOURISM: PRODUCTS, CHANNELS AND SEGMENTS” NOTTINGHAM UNIVERSITY BUSINESS SCHOOL 2006/1
Christel DeHaan Tourism and Travel Research Institute Nottingham University Business School Jubilee Campus, Wollaton Road Nottingham, Notts. UK NG8 1BB Email:
[email protected] Tel: 0115 84 66606
1
PROCEEDINGS CONFERENCE
OF
THE
4TH
DEHAAN
TOURISM
MANAGEMENT
“MARKETING TOURISM: PRODUCTS, CHANNELS AND SEGMENTS” NOTTINGHAM UNIVERSITY BUSINESS SCHOOL TUESDAY 13TH DECEMBER 2005 TABLE OF CONTENTS 1. Abstract 2. Online marketing channels of Devon & Cornwall hotels: An exploratory study Wai Mun Lim, and Rong Huang, University of Plymouth 3. Paradoxes of Consumer Independence: A Critical Discourse Analysis of the Independent Traveller Robert Caruana, University of East Anglia 4. Causing Tourism Wu Meng-Jun, Rob Young and Nottingham University Business School
Anita
Fernandez
Young,
5. The Impact of External Shocks on the Tourism Sector: the case of Turkey Murat KARAGOZ, Ali SEN, Ali KOCYIGIT Inonu University, Faculty of Economics and Administrative Department of Econometrics
TTRI,
Sciences,
6. Segmentation of Cultural Festival Visitors with Motivation Peter Schofield, University of Salford 7. Identity, Nostalgia and Diaspora Tourism Ranjan Bandyopadhyay, TTRI, Nottingham University Business School 10. An Exploration of Niche Market Festivals Elspeth Frew, La Trobe University, Victoria, Australia 11. New Approaches to Tourism Products Chris Cooper, The University of Queensland, Australia
2
ABSTRACT At a time when the tourism and travel industry is growing but that growth is threatened by terrorism, disease and natural disasters, tourism marketers face new challenges. The Internet enables travellers to create their own holiday packages, as the „no-frills‟ airlines are creating new markets. Thus, the value of traditional marketing approaches is being questioned and new strategic marketing approaches are being tested throughout the industry. As tour operators develop new products, many of them use the new channels of distribution, and try to target the market segments with which they are familiar. But which products will succeed? How do you match the product with the channel and the segment? Are new niche markets emerging, and is this the way the industry is going? What is the role of the Internet and how does it affect consumer choice? These are a few among the many questions tourism marketing professionals and academics are asking, and it is with these in mind that TTRI has taken Tourism Marketing as its theme for the 4 th DeHaan Tourism Management Conference.
3
ONLINE MARKETING CHANNELS OF DEVON & CORNWALL HOTELS AN EXPLORATORY STUDY Wai Mun Lim
Dr. Rong Huang
Lecturer in Hospitality Plymouth Business School, University of Plymouth, Plymouth, PL4 8AA Email:
[email protected] Tel: 01752 232845
Lecturer/Senior Lecturer in Tourism Marketing Plymouth Business School, University of Plymouth, Plymouth, PL4 8AA Email:
[email protected] Tel: 01752 232836
ABSTRACT According to the British Hospitality Association (2004), the independent hotels that make up the largest segment of the hotel sector represent approximately between 70-80% of all UK hotel establishments. Devon and Cornwall has long wielded a powerful attraction for holidaymakers, where tourists nights in 2004 make up a quarter of all room nights spent in England (StarUK,2004). Though various types of online marketing adopted by hotels have been identified in past studies, many of these channels were initially adopted by large chain hotels resulting in a plethora of debates and research in both academic and industrial circles. There is little research evaluating the independent hotel sector in these regions with regard to online marketing issues. This study therefore attempts to categorise online channels adopted by independent hotels in Devon and Cornwall. The conclusion discusses some of the possible implications of online marketing for the independent hotels. Keywords: online marketing, distribution channels, independent hotels, UK hotel industry INTRODUCTION Both industrial (Anon, 2002; Starkov& Price, 2001, 2004) and academic reports state (Buhalis & Main, 1998; O‟Connor, 2003; Sangster, 2001) that the use of the internet in the marketing of hotels now seem to be the norm rather than the exception. Various types of online marketing adopted by hotels have been identified in past studies. Many of these channels were initially adopted by large chain hotels resulting in plenty of debates and research in both academic and industrial circles. However, such large chain hotels do not represent the largest sector nor are they representative of the UK hotel industry in terms of total hotel numbers. According to the British Hospitality Association (2004), the independent hotels that make up the largest segment of the hotel sector represent approximately between 70-80% of all UK hotel establishments. Devon and Cornwall has long wielded a powerful attraction for holidaymakers. According to STARUK (2004), the majority of all tourist nights in the South West of England were spent in Devon and Cornwall, but there has been little research evaluating the independent hotel sector in these regions with regards to online marketing issues. This study therefore attempts to categorise online channels adopted by independent hotels in Devon and Cornwall. It will also attempt to identify online channels used by the independent hotels followed by an investigation of whether there is a correlation between a hotel‟s graded classification and the number of online channels used or whether an online booking facility is adopted. 4
ABOUT DEVON AND CORNWALL Devon and Cornwall have consistently been ranked the most popular counties in the South West of England in terms of tourist arrivals. According to Star UK (2004), see Table 1, the number of tourist nights spent in Devon and Cornwall for 2004 stood at 45 million nights, this accounts for 14% of England‟s total. In terms of tourists spend, the total between Devon and Cornwall accounted for more then half of all tourist spend in the South West of England.
DEVON CORNWALL SOUTH WEST ENGLAND
TRIPS 5.6 million 4 million 20.5 million 101.4 million
NIGHTS 22.5 million 22.5 million 80.1 million 314 million
SPEND £1,061 million £1,232 million £4,103 million £18,960 million
Table 1: Volume and spending of tourists in the South West 2004 Source: StarUK (2004) The tourism sector of both Devon and Cornwall‟s economy constitutes the largest industry, with staying visitors contributing, in 2004, some £2,293 million in expenditure to the region's economy. More significantly, 34% of this spending falls under accommodation. (StarUK, 2004). INDEPEDNENT HOTEL SECTORS IN THE UK The vast majority of tourism businesses around the globe are small in size, belonging to local entrepreneurs, are family run, predominately employ members of the host society and are rural based businesses (Middleton and Clarke, 2001; Main, 2002). Due to the hotel industry‟s fragmented and heterogeneous nature as a whole, it has been even more difficult to form a definition of an independent hotel. Adding to the complexity, there are a handful of private and public organisations within the UK travel industry which have an indirect stake on how hotels or independent hotels could be defined, they are private organisations such as Automobile Association (AA) and the Royal Automobile Council (RAC) whose central function within the hotel industry is to assess quality standards of hotels agreed between themselves and the English Tourism Council (ETC) (wholly government initiative) and thereafter to classify participating hotels in star categories. While features of independent hotels vary considerably, there are a few core characteristics identified by some academia that typify the entire independent hotel sector (Slattery, 1992; Alisau, 2002; Walsh, 2002; Stewart, 1996; Imrie & Fyall, 2000; Knutson, Beck & Yan, 2004). Up till as late as the early 90s, the independent hotel sector represented approximately 90% of all hotels (Main, 1995; Steward, 1996). According to a recent Keynote (2005) report, the number of enterprises involved in hotel operation in the UK fell by more than a third (37.3%) between 1990 and 2004 which indicates a drop from 14410 to 9030 units. There have been ongoing moves towards consolidation, with independent operators leaving the business and chains continuing to develop. However, the independent hotel sector is still the largest segment of the hotel sector in the UK (Keynote, 2005). It must however be emphasised that the above figures do not testify if the drop in percentage could be a result of independent hotels joining affiliations or representation firms (e.g. Best Western) to gain access to new distribution channels, being franchised to a brand or sold on 5
to a chain. One reason for this ambiguity could have stemmed from a lack of common definitive understanding of independent hotels. This is an important consideration because while such affiliations are subscribed to not unlike a membership, independent hotels which do so remain independent since the form of ownership and management control remain in the same hands. What becomes different when a hotel becomes affiliated is – it has simply added another marketing or sales channel to the hotel‟s portfolio. It can easily be misconstrued to think that every added membership to Best Western for instance, amount to an equivalent drop in the number of independent hotels. While this paper does not deny that the independent hotel sector is shrinking in the UK, there is no collective nationwide research to reveal any distinct trend, except that there is an indication of increased competitiveness within the independent hotel sector and the fervent expansion of international and local hotel brands, fuelled by operators eager to gain a foothold in the UK (Imrie & Fyall, 2000). Smith Travel Research (cited in Swig, 2000) also reported that the independent hotels are competing and performing well against the branded competition and acknowledged that although independents generally trail branded hotels in market occupancy share, they do achieve an average daily rate premium and can therefore equalise in terms of REVPAR (revenue per available room). Given their predominance, failure to engage such hotels in destination marketing initiatives may undermine the destination‟s potential and limit the range of quality experiences for visitors. E-COMMERCE PRACTICE The International Data Corporation predicted that worldwide e-commerce revenues will increase approximately eightfold in the next few years, growing from $634 billion in 2001 to more than $5 trillion in 2005 (Gantz, Glasheen & Emberley, 2001; cited in Wolfe, Hsu and Kang, 2004: 51). Travel and tourism accommodations have rapidly become the largest category of products sold over the Internet. Information, communication and technology (ICT) has frequently been demonstrated to be of supreme importance in the marketing and distribution of hotels (Connolly & Moore, 1995; Main, 1995; Mistillis, Agnes & Presbury, 2004; Christian, 2000; Marvel, 2001; Swig, 2000), ironically there is also an equivalent amount of lament (if not more) about the hoteliers lack of strategic implementation and management of ICT by the very same writers. Very often, implementation criticisms are directed at small medium sized hotels (Mistillis et al., 2004; Christian, 2000, Main, 1995) while large hotels have the effectiveness of their ICT strategies‟ evaluated, criticised and examined (Thomson & Failmezger, 2005; Anon, 2004; Hayes & Kontzer, 2004). The availability of an array of new technologies has brought about new interests in ICT adoption of hotels. These interests have regularly been discussed from two angles: one for the betterment of in-house service (such as having in-room internet access or speedy checkins etc) and second, a new channel for hotels, electronic business. It is with the latter that we are concerned with. A separate understanding into the mechanics of electronic channel management and adoption of the hotel sector would help it to discover the concepts that form the nucleus around how channels emerge, what happens with each channel and how the decision making process for adoption could be strategised. To begin investigating a hotel‟s online marketing and distribution strategy, it will be helpful to classify the levels of internet adoption open to hotels. According to research on tourism small medium sized enterprises (SMEs), Dholakia and Kshetri (2004) there are three levels of adoption. They are the pre-adoption stage where an SME owns at least a computer but no web site; the adoption stage where an SME owns a 6
website but does not sell on the internet and the routinisation stage where the SME sells on the internet. This finding is also similar to Duffy (2003) „1st, 2nd and 3rd generation Internet marketing respectively. In terms of types of online marketing channels adopted by hotels, Starkov and Price (2001) identified five, namely a hotel‟s own website, online travel intermediaries, travel and hotel directory portals, destination focused search engines, and property pages within a major brand website (Figure 1).
e-COMMERCE -Indirect Web-
Own web
Travel & hotel directories & portals
Property pages within a major brand website
Destination-focused search engine strategy & initiatives
Online travel intermediaries
Merchant model Services
Agency (Commission-based) Opaque rate model services model services
**Merchant model: Suppliers/clients retain a certain amount of control of room rates, availability and sale. Figure 1. Possible online channels of travel marketing and distribution Adapted from Starkov and Price (2001) THE ONLINE DIMENSION OF MARKETING AND DISTRIBUTION While deliberating on the online distribution or marketing strategies of independent hoteliers, the most immediate question would be, how much do these hoteliers know about the services offered by the various online channels. Also, one of the more critical results would be to discover if hoteliers recognize these as marketing or distribution channels or joint functions and if they are individually perceived to be operationally effective. Channel members themselves may not be aware of how many other organizations or individuals constitute the total channel. Managers making channel selection decisions should therefore have a comprehensive understanding of the underlying forces at work in a channel design (Bowersox et. al., 1980). According to Robert, Schurr and Oh (1987), the first of the five phases of relationships in marketing channels begins with awareness. Developed in the form of a life cycle, the five phases depict the possible stages a channel relationship could travel through. Prior to embarking on the life cycle of channel relationships, a hotel will have to deliberate over its channel design, described by Bowersox et. al. (1980) as a planning process where a firm either markets a new service or modifies existing distribution arrangements. For a channel design to succeed, hoteliers will have to be aware of which channel options are available. Where a firm markets a new service, 7
it is referred to as channel adoption and the modification of existing arrangements is referred to as channel creation (Walters, 1974). Figure 1 above has attempted to identify the various forms of online channels and intermediaries available to hoteliers. It is necessary to reiterate that the objective of the figure is to highlight some of the more prevalent online channel options (innumerable) that are available to hoteliers today. This study will attempt to categorise all online channels found in our data collected into the above five categories. Non-GDS/CRS based channels of distribution exist, but variety and more importantly online availability data of these variations may be unavailable to consumers. Hotels who develop pages within a major brand website ride on the site‟s wide consumer base and popularity, and in the majority of instances will not have a GDS link. A current example is latebeds.com. Hotels may also choose to place themselves within destination sites, which inform potential site visitors of hotel availability in the area. The hotel‟s pages would normally contain basic information such as address, amenities and room types. It may be insufficient for a hotel to simply have a domain name as potential guests may not be able to locate that hotel‟s website unless it has registered with a particular search engine. If not, the chances of potential guests obtaining information about the hotel via the web may be slim to none. In summation, there are a wide variety of online distribution methods available to hoteliers, but which ones s/he will employ is greatly dependent on what has been made available to the hotelier, how much an hotelier understands the various modes and how s/he came to know about them. Attempting to understand the hotelier‟s process of attaining knowledge vis-à-vis the many online possibilities through experiences and thoughts will be a major agenda following this research. By doing so, a more robust conceptualization of key parameters in the context of an hotelier‟s cognitive consciousness may perhaps help us to explore their dilemmas and struggles of online channel adoption. METHODOLOGY The research is exploratory in nature. For this research, independent hotels are typically, individual and personable small medium sized establishments providing accommodation and meals, whose management is free from outside control and are operationally flexible. The empirical research uses the Devon and Cornwall hotels listed in the AA (UK) website as samples. The reasons for adopting the AA website are due to the following constraints: i)
there are no compulsory registration schemes for accommodation establishments in England (Anon, 2003), therefore no such lists exist
ii)
looking at Yellow Pages (paper back) and under the category of Hotels, there is a mix of bed and breakfasts, hotels, hostels and lodges, making it impossible to distinguish a bona fide hotel- as some bed and breakfasts name themselves „ABC hotel‟ even though they are not operating as one. Making the task of distinguishing an independent hotel even more unattainable
iii)
a check with ten of England‟s tourism district councils revealed that, due to the data protection act, most were not able to provide the researcher with a list of bona fide hotels, the rest of the councils redirected the researcher to their website where classification of hotels and bed and breakfasts were not clear
8
iv)
the one and only alternative Hotel grading organisation, RAC (Royal Automobile Club) do not produce a commercial list of hotels, unless a specific location search of hotels is performed on their website but even then the list is a mix of graded bed and breakfasts and hotels.
There were a total of 210 bona fide Devon hotels and 150 bona fide Cornwall hotels extracted from the AA website. „Devon‟ was typed onto the AA location search box and „Hotel‟ as accommodation type was selected. This was performed on the 25th October 2005 and because it was decided that the research will only be evaluating 50 hotels as an exploratory cum empirical study, every 7th hotel on the list was selected for a search on the search engine Google, the most popular search engine according to Nielsen NetRatings (2005). The hotel‟s name and the county they are in (either Devon or Cornwall) are then entered into the Google‟s search box. The top 10 search results were then recorded. In the process, information about the individual hotel‟s star ratings, whether the hotel has their own Uniform Resource Locator (URL) and an online booking facility was recorded. In most instances, a more thorough search had to be carried out because a hotel‟s URL did not emerge in the top 10 search results. This was done by adding more keywords other then the hotel‟s name and county location into the Google search box. The task of ascertaining the availability of an online booking facility in each hotel was performed by scrolling through the web pages of the hotel. The number of marketing channels whether it is by subscription or complimentary, are taken into account only if information (however basic) of the hotel is present in the search result. Data analysis is performed by means of SPSS to discover the emerging frequency of online channels used by the hotels and to measure the correlations
FINDINGS & ANALYSES Of the 50 hotels searched, all but one hotel has a minimum of 4 valid marketing channels. It was apparent that all five possible modes of online marketing channels were present although a majority were accommodation wholesaler (travel intermediary) channels (Figure 2). There were a total of four hotels without its own website, but yet each had an average of 7 valid marketing channels from the top 10 search results on Google. With this finding, we are lured to ask the question of whether a hotel having its own website would necessarily contribute to increasing its online exposure.
9
140 120 100 80 60 40 20 0
Travel Intermediaries
Travel & Hotel directories
Own Web
Destination initiatives
No. of different channels used
Property Web pages
No. of different channels used
Channels used
Figure 2: Valid online marketing channels
While only 4 hotels do not have their own website, 3 of those were 2 star hotels and the fourth a 3 star hotel. Although 46 hotels searched had their own websites, more then half of these hotels had their own URL ranked first in a search result. This could be due to a lot of factors, but the core indication this result provides is that a good majority of hotels seem to recognise that search ranking position is essential. 16 of the 46 hotels searched do not have their websites position in the top 10 of our search (Figure 3).
5
4 No w ebsite
14
Website on 1st Website not on top 10 Website on top 10 27
Figure 3: Status of web ownership
From the compiled data, within the top 10 search results, it became clear that the most common online marketing channel is a hotel‟s own website. See Table 2. The second most common occurrence in the search results was from Tripadvisor, which is a travel information and advice service that offers unbiased reviews and opinions of travellers. It is also important to note that where a hotel‟s own website is not ranked 1st in the search result, 4 of the hotels who do have their own website but are not ranked in the top 10 of the search has Tripadvisor as their number one search result. The third most common occurrence is via the channel Welcome Southwest. This channel is misleading as it re-direct surfers to a site that has no relation to the hotel nor to the South West of England. This could be due to the fact that until recently, welcomesouthwest.co.uk was the URL address for a destination-focused directory for travel information. Finally, the fourth most common occurring channel is information Britain. It is a travel
10
destination guide operated by a private organisation where subscription is not necessary but yet this channel was in the top 10 search result for 23 hotels. Rank 1 2 3 4 5 5 7 8
Site Own Website Tripadvisor (Travel Directory) Welcome Southwest (previously a Destination initiative) Information Britain (Destination initiative) Hotels.uk (travel intermediary) Britain Express (travel intermediary) Findukhotels.co.uk (travel intermediary) Chycor.co.uk (Destination iniative)
No. of times online channel emerged 46 29 25 23 9 9 8 7
Table 2: Most common online channels found Finally, SPSS was used to measure association between a hotel‟s graded classification and the number of online channels used. The calculated significance level (1-tailed) was 0.384 which was greater than 0.05, therefore there was no significant correlation between the hotel‟s grade and its number of online channels used. It should also be added that a strong 20% of the 50 hotels studied has an online booking facility on their hotel‟s web page. CONCLUSION From this exploratory study, it was apparent that all five possible modes of online marketing channels were adopted by the independent hotels in Devon and Cornwall regions although the majority of the channels were destination search channels and accommodation wholesaler (travel intermediary) channels. 46 hotels searched had their own websites. The other most common channels adopted by the hotels were tripadvisor, Welcome South West of England, and information Britain. From the statistical analysis, there was no significant correlation between the hotel‟s graded classification and the number of online channels adoption. Implications of this research are threefold. Firstly, online marketing channels are varied and although finite, no attempt has been made to consolidate and categorise them. Attempting to create categories of online marketing channels could introduce an organised construct for hoteliers when they are considering the modes of online marketing channels for adoption. Secondly, being able to identify the types of online channels the hotel appears in, a hotelier is in a better position to control the type of image their hotel aims to project. Finally, if a hotelier knows how many and what type of online marketing channels are used, the effectiveness (in terms of volume of sales generated or investment costs) of each channel category can then be measured. As this research is exploratory in nature and also that little or no studies have been done in relation to this study, it is important to note that there are limitations to this study. The study has only attempted to provide an initial glimpse into the types of online marketing channels for independent hotels in Devon and Cornwall regions. It is understood that further refinements to the research methodology is needed in our attempt to categorise the types of online marketing channels available and hopefully understand how these various online channels have an impact on consumer‟s search. The limitations to this study are:
11
i)
keyword search is limited to the hotel‟s name and county to provide some level of consistency in the study
ii)
if the search is performed on a different day or at a different time, search results will differ due to various search engine marketing (SEM) initiatives performed by the hotels.
Future research could aim to considerably increase sample size across counties, not only to improve validity but also its statistical significance where relevant. It must be acknowledged that although the study can be a time consuming effort, it is nevertheless a necessary step to enhance our understanding of online marketing channels that exist for independent hotels.
12
REFERENCES Alisau, P. (2002) Sense of autonomy, history aid growth of independents, H&MM, July 24th, U.S. Anon (2002) Travelocity introduces next generation technology with merchant hotel-offeringpioneering concept part of focused partner marketing program, Travelocity one, in http://www.hospitalitynet.org/news/4013606.print, accessed on 28th November 2004 Anon (2004) Now is the time for Independent Hoteliers to exploit the changing market and maximize their advantage over the global chains; hotel market embracing individuality, in http://www.hotelonline.com/News/PR2004_3rd/August04_Clustomers.html Accessed: 04 March 2005 British Hospitality Association (2004) Definitions and Classification of Accommodation Establishments, electronic mail response received on 14th April, 2004,
[email protected] Christian, R. (2000) Developing an online access strategy: Issues facing small to medium-sized tourism and hospitality enterprises, Journal of Vacation Marketing, Vol. 7, No. 2, pp. 170-178 Connolly, D.J. & Moore, R.G. (1995) Technology and its impact on global distribution channels in the hotel industry, in Proceedings of the annual meeting – Decision Sciences Institute, Vol. 3, pp. 1563-1565 Dholakia, R. & Kshetri, N. (2004) Factors Impacting the Adoption of the Internet among SMEs, Small Business Economics, 23, pp. 311-322 Duffy, P. (2003) eMarketing 101- What sales and marketing managers need to know, Sales and Marketing Excellence, Vol. 3, No. 2, February Hayes M. & Kontzer, T (2004) The Right Balance, Information Week, Issue 984, pp. 18-20 Imrie, R. and Fyall, A. (2000) „Independent mid-market UK hotels: Marketing strategies for an increasingly competitive environment‟, Journal of Vacation Marketing, 7, 63-74. Keynote (2005) „Hotel industry‟, Keynote report, November 2005 Knutson, B., Beck, J. & Ten, H. (2004) Marketing the mid-price independently owned resort: A case study with implication for managers, Journal of Hospitality & Leisure Marketing, Vol. 11 (4), pp. 65-79 Main, H. (1995) „Information technology and the independent hotel – failing to make the connection?‟ International Journal of Contemporary Hospitality Management, Vol. 7, No. 6, pp. 30-32 Marvel, M. (2001) Improving performance in small and medium sized hotels: The Swiss Experience, Travel & Tourism Analyst, No. 5, pp. 43-60 Middleton, V. and Clarke, J. (2001) Marketing in Travel and Tourism, Butterworth-Heinemann, London.
13
Mistillis N., Agnes, P. & Presbury, R. (2004) The strategic us of information and communication technology in marketing and distribution – a preliminary investigation of Sydney hotels, Journal of Hospitality & Tourism Management, Vol. 11, No. 1, pp.42-55 Slattery, P. (1992) Unaffiliated Hotel Chains in the UK, International Journal of Contemporary Hospitality Management, Vol. 9, No. 7, pp. 270-273 Starkov, M. and Price, J 92001) „Independent hoteliers: eBusiness levels the playing field, in http://www.hotelinteractive.com/hi_articles.asp?page_id=500&article_id=1084, accessed on 13th April, 2005 Starkov, M and Price, J (2004) „Hotelier‟s 2004 Top ten Internet strategy resolutions, in http://www.travelmole.com/printable.php?news_id=99460, accessed on 13th April, 2005 STARUK (2004) http://www.staruk.org//webcode/contents.asp?id=741&parentid=469&bg=white, accessed on 02 December 2005 Stewart, D. (1996) Hoteliers & Hotels: Case studies in the growth & development of UK hotel companies 1945-1989, Search Publications, Glasgow. Swig, R. (2000) Independent hotels: The new brand alternative, in http://www.hotelonline.com/Trends/Swig/Swig_Independents.html, Accessed 04,March 2005 Thomson, G.M. & Failmezger, A. (2005) Why customers shop around: A comparison of hotel room rates and availability across booking channels, CHR Reports, Cornell University, Vol. 5, No. 2 Walsh, J.P. (2002) Competitiveness lures independents to recognized brands, H&IMM, September 16th, U.S. Wolfe, K; Hsu, C. H and Kang, S. K (2004) „Buyer characteristics among users of various travel intermediaries‟, Journal of Travel & Tourism Marketing, 17(2/3), pp.51-62
14
PARADOXES OF CONSUMER INDEPENDENCE: A CRITICAL DISCOURSE ANALYSIS OF THE INDEPENDENT TRAVELLER
Robert Caruana University of East Anglia, UK Andrew Crane University of Nottingham, UK James A. Fitchett University of Leicester, UK
Address for Correspondence: Robert Caruana, University of East Anglia, Norwich, Norfolk UK Tel: 0115 9580166 Tel: 07886 226435
BIOSKETCH Robert Caruana is a Lecturer in Marketing at the University of East Anglia. His research focuses on tourism as a discourse of consumption and on the role of morality in consumption. Andrew Crane is Professor of Business Ethics at the University of Nottingham. His main research interests are business ethics, corporate social responsibility, and sustainability. James Fitchett is Reader in Marketing and Consumer Research at the University of Leicester Management Centre. His current research focuses on futurism, utopianism and myth in consumer and marketing discourses.
15
PARADOXES OF CONSUMER INDEPENDENCE: A CRITICAL DISCOURSE ANALYSIS OF THE INDEPENDENT TRAVELLER
ABSTRACT The „ideology‟ of independence lies at the very core of the marketing agenda. For the free market to operate as a legitimate means of social organisation, the right to be independent and to be free to enact ostensibly independent choices is for all intents and purposes sacred. Independence is an especially critical concept for marketing academics and practitioners to understand given the need to reconcile consumer demand for a sense of individuality, freedom and self, with an organisation‟s need to commodify consumption activities in order to realise market growth. This paper examines the ways in which a sense of independence is successfully offered to consumers within paradoxically mass-market communications. The study investigates what it means to be an independent traveller by implementing a critical discourse analysis of alternative guidebooks. Findings suggest that guidebooks construct independence by reifying inaccessibility, interpreting value, and constructing inauthenticity for consumers. This promulgates a powerful myth of the independent traveller as someone who defies inaccessibility, hunts for bargains, and avoids inauthenticity. Crucially, each of these cultural practices also acts to engender an implicit relation of dependency between the text and the tourist that is found to contradict, but ultimately not threaten, the whole notion of independence that the consumption experience itself is predicated on, orientated around, and indeed valorized by. KEYWORDS Independence; guidebook; commodity; critical discourse analysis; Spain; tourism
PARADOXES OF CONSUMER INDEPENDENCE: A CRITICAL DISCOURSE ANALYSIS OF THE INDEPENDENT TRAVELLER INTRODUCTION The „ideology‟ of independence lies at the very core of the marketing agenda. For the free market to operate as a legitimate means of social organisation, the right to be independent and to be free to make ostensibly independent choices is for all intents and purposes sacred. Without a well developed sense of independence citizens cannot become individuals, and without individuality the modern idea of the consumer as an autonomous agent of choice is untenable. There is of course no essential reason why people involved in market place behaviour should valorise independence over conformity for instance, and nor is there any concrete justification to sustain the belief that making product choices independently of others will necessarily lead to better consumer outcomes. The signification of independence as a positive personality trait has however been long recognised as important in consumer mythologies, from Benjamin‟s (2002) nineteenth century arcades flaneur all the way to contemporary images of American ruggedness (Hirschman, 2003).
16
In practice there are few mass-market situations where consumers are able to exercise unconstrained independence. In this regard, mass markets are unavoidably paradoxical. All too often consumers appear to independently choose very similar types of products to each other. One consequence of this is that it appears that highly individualised consumer cultures can tend to look very similar to highly conformist ones. Here consumers can find themselves involved in a continual struggle to sustain their own sense of independence and individuality where they frequently realise that others are trying to express their own independence in exactly the same way, using the same products and brands. In this respect the notion of independence in markets is problematic in a literal sense. The opportunity to realise „individuality‟ by exercising various freedoms from the influence of another or others is frequently undermined. Moreover, the notion of independence as a state or quality of being free from subjection or from the influence, control, or guidance of individuals, things, or situations is at best a contestable scenario. How then do marketers continue to retain resonance with a mass audience paradoxically characterised by consumers who seek to identify themselves as distinct from others around them and autonomous in their choices? In light of these underlying inconsistencies, tensions and contradictions this study finds that it is more analytically relevant to examine independence as a socially constituted discourse. From this perspective marketing is seen as both contributing to the construction of the idea of independence as a positive personality trait, as well as commodifying independence as a desirable quality and value of certain products and services. This approach is particularly apt for the study of goods and services which signify high social status and cultural visibility where it is often in the marketing of these commodities that the discourse of independence and individuality is especially prominent (Frank, 1999). Understanding the construction of discourses which promote the desire to stand out from the crowd, the longing for distinction (Bourdieu 1984), the desire to be different, and the need to feel that ones choices are unique or special to ones self (Belk 1988) is therefore of particular relevance to marketing and for this reason worthy of detailed study. This paper examines how one particularly prevalent discourse of consumer independence - that of the independent tourist traveller, is communicated and constructed for mass market consumption. We begin by introducing the independent travel industry and consider the primary importance of independent travel guides in structuring the market. Guidebooks, we suggest, must be understood as more than purely descriptive artefacts. In additional to providing various types of information about travel destinations guidebooks are also important as performative devices which direct and command the social parameters and identity of those consumers involved in the independent travel market (Jacobsen, 2004). We then employ critical discourse analysis as a method to deconstruct various aspects of independent traveller discourse contained in guide books to show where and how market value is maintained. Focusing on one such text, the Rough Guide to Spain, findings suggest that the construction of independent travel relies on three main elements: the reification of inaccessibility; an interpretation of value; and a construction of inauthenticity. Crucially, we contend that the effect of this construction is to position the independent traveller as independent whilst simultaneously engendering dependence on the guidebook to sustain that independence. We conclude that as a manuscript for practicing independent travel, alternative guidebooks promulgate a narrative of risk, adventure, guile, and autonomy within the context of a mutually dependent relation between tourist and product. 17
UNDERSTANDING INDEPENDENT TRAVEL THROUGH GUIDEBOOKS There has been relatively little academic attention paid to the independent travel phenomenon leading to a lack of breadth and maturity in definition and, significantly, conceptualization of the pivotal element of independence. One of the better definitions comes from Hyde and Lawson (2003), who have characterized independent travellers as: experiencing an evolving itinerary; willing to take risks in selecting vacation elements; and having the desire to experience the unplanned. This definition of who independent travellers are is derived from their understanding of what independent travel is. In their view, independent travel is a segment of the tourism market that is defined by its distinction from the traditional forms of travel arrangement, namely the package tour market. Independence, therefore, implies a catch all term for any form of travel not purchased as a package. Whilst formulated along similar lines, Morrison, Hsieh and O‟Leary (1994) offer an extension to this package-versus-non-package view of independent travel. Seeking to go beyond what they see as narrow definitions they identify a multi-segmented international travellers market which is divided according to the form of travel arrangement of each market segment: (i) independent; (ii) escorted tour; and (iii) non-escorted package. Whilst this offers an extended conceptualization, they essentially mirror the view of the independent travel segment offered by Hyde and Lawson (2003). Consequently, both these views configure independent travel as a generic form distinguished by fragmented, ad hoc travel arrangements rather than a more centrally organized ones. In this view, the actual notion of independence, and what it might signify for the traveller, remains largely unexplored. Others offer definitions of independent travel that center more upon the notion of independence in relation to a specific social category of travel (Loker-Murphy and Pearce 1995; Sorensen 2003; Weber 2001). Backpacking, budget and even adventure travel have all been identified as forms of travel centered on a notion of personal independence, that is, in the liberated, self-reliant sense of being independent. For example, Loker-Murphy and Pearce‟s (1995) analysis of young budget travellers associates a notion of independence with the cultural dynamics of a specific travelling group. They contend that, “backpackers are travellers who exhibit a preference for budget accommodation; an emphasis on meeting other people (locals and travellers); an independently organized and flexible travel schedule; longer rather than brief holidays; and an emphasis on informal and participatory recreation activities,” (1995:830-1). Essentially, this identifies a type of traveller around the notion of what it means to be independent. Under previous definitions, the academic arranging her own trip for an overseas conference is parceled into the same category as the backpacker. In such cases, the notion of independence itself becomes a less significant and relatively unsatisfactory criterion on which to conceptualize independent travel. In an expanded notion of independence, Sorensen (2003) focuses on the way meaning and identity is produced through interactions within a “backpacker culture”. A key feature of this cultural dynamic is the maintenance of a distinction between the traveller (“us”) and the tourist (“other”). Significantly, this distinction is largely reproduced on the basis of the backpacker‟s perceived ability and desire to self-organize and be self-reliant. In addition, these notions of independence that delineate travellers from the tourist are also applied to intra-cultural dynamics. Sorensen (2003) identifies what he terms a “road status” culture whereby a social hierarchy based on discernible degrees of independence emerges as an underlying structure in the social order of backpackers. In this sense, independent travel becomes a practice of intra-cultural as well as inter-cultural signification. The notion of 18
independence becomes a representational device for constructing group members‟ selfidentity both in relation to each other and, externally, in relation to other types of tourist. Understanding independence as a cultural signifier alerts us to its essentially constructed nature within the tourism landscape (Urry 1990). The cultural construction of independence occurs through a variety of processes and artifacts; however, the present study was particularly concerned with examining the role of the guidebook in constructing the nature of independent travel. Indeed, there is an identifiable niche of alternative guidebooks dedicated to the independent traveller, led by the Lonely Planet and Rough Guide series, among others (Sørensen 2003). It is widely acknowledged that guidebooks have an important mediating role in assisting tourists both in what they do and how they see tourist attractions (Bhattacharya 1997; Siegenthaler 2002). In terms of independent travel, McGregor (2000) reveals that the reader‟s response to guidebook texts influences both the meaning of foreign destinations as well as the practice of independent travel itself. Finding that travellers very rarely visited sites that were not featured within the guidebooks he concludes that: “…guidebooks delineate a world of experience, making some foreign places open, attractive and accessible, while at the same time restricting and commodifying the extent, and variability, of traveller‟s explorations.” (McGregor 2000:35). In this sense, guidebooks are seen as defining not only the places to go and things to see for independent travellers, but also the way in which such sites must be engaged with, and indigenous populations related to (Bhattacharya 1997). This results in a generic “traveller gaze” that reflects a commonly shared identification of places and travellers. Alternative guidebooks are therefore significantly more than a set of guidelines about sites “not to be missed,” but also play a role in constituting both the subject and the object of the travel experience. Just as the Bible represents more than an accumulation of key religious figures, temples and pilgrimages to the practicing Christian, the alternative guidebook is as much an implicit doctrine, or a manuscript, for practicing independent travel itself. These insights give rise to two interrelated questions that formed the basis of the present study. First, it was deemed important to establish how independent travel came to have a particular nature in alternative guidebooks that both characterized it as, and distinguished it from other more obviously „dependent‟ forms of travel. And second, how had it been made possible to present independent travel as a coherent, logical and legitimate social practice in such texts given the potential tension between such autonomy and the dependence implicit in tourists‟ use of guide books. Essentially, what does it really mean to be independent within a huge industry of supporting and enabling service industries, and how do consumers maintain a sense of independence within this web of market interdependencies? STUDY METHODS The study employs critical discourse analysis (hereafter CDA) to understand the way in which the notion of independence is constructed and made meaningful for consumers of independent travel and to examine the underlying context of consumer dependency upon the market to achieve this. Accordingly, this next section attempts to define a critical discourseanalytic approach, locate it within a growing text-based consumer research agenda and consider some of the methodological implications for its mobilisation within this current study, before going on to illustrate the implementation of CDA. Critical discourse analysis allows researchers to understand how the identities and cultural practices of consumers are reified in communications to them whilst enabling critical 19
analysis of the relations of power implicated within those communications. It is therefore dually concerned with the constructive properties of text and in its underlying dynamics of ideology and power (Fairclough, 1992). Discourse analysis, the „family‟ into which CDA is subsumed, is underpinned by a social constructivist epistemology. Researchers have examined texts for structural sequences, linguistic features, discourse, narrative structures and metaphorical devices that are seen to convey something meaningful about the nature and organisation of aspects of social life such as consumption. In most part due to the fledgling status of discourse analysis across the social sciences, for better or for worse there is no hard and fast template for its application. Within consumer research, text-based investigations are diverse both in terms of the vehicle of analysis (mode) and the primary concern of the researcher (purpose). Studies of advertisements (Stern, 1995), interview transcripts (Thompson et al, 1994), TV programmes (Hirschman, 1998), consumer‟s stories (Shankar, et al, 2001), conversations (Thomas, 1992) and even combinations of these text genres (McQuarrie and Mick, 1999) vary in how text is seen to organise interactions, produce culture (Thompson, 1994) and reveal ideology and power (Thompson, 2004; Hirschman, 1990). Consequently, whilst discourse analysis may appear at first to represent a discernable cluster of methodological conventions based around a common interest in text, there are clearly some associated definitional issues. Phillips and Hardy (2002) have attempted to discern some key discourse analytic approaches that are open to researchers, suggesting that the distinction between the approaches is characterized by two important continuums; firstly, the extent to which methodology embodies a concern with context as opposed to text; and second, the extent to which the focus is critical of power relations or is concerned instead with the constructive dynamics of the text. However, they recognize that the different dimensions are not mutually exclusive and the approach adopted in this study occupies a position quite squarely within these continuums and more closely aligned to Fairclough‟s (1992) multi-dimensional approach to critical discourse analysis. Incorporating Foucault‟s concern for discourse and power into a more structured, textcentered approach, Fairclough‟s multi-dimensional framework requires an understanding of the relationship between linguistic properties and social processes of text and how, in the production of texts, implicit power relations are created that privilege certain actors whilst marginalizing others. In practice, a primary level of analysis exposes the linguistic dynamics involved in the text‟s fabrication. Interest at this initial stage is with how the text functions through explicit linguistic devices such as metaphor, discourse, narrative and sentence structure. This then provides a robust platform for elucidating the broader social processes at work in the text. Concerned more specifically with what the text was doing, the second order analysis highlights the way in which meaningful social roles central to a particular tourist identity are constructed for the reader. Analysis over these first two levels assumes the replicable identification and exposition of explicit and essentially „concrete‟ social processes. Here, a criterion for analytical rigour rests on the principle that another researcher trained in discursive techniques would expose broadly similar cultural processes at work (regardless of whether or not they viewed the processes as producing true or meaningful cultural identity). Finally, a higher order of analysis is enacted whereby the hegemonic relationship between text and consumer, the condition in which the meaning of independence is both „produced‟ and „consumed‟, is critically appraised. Most significantly, where other discursive approaches appear to be concerned with specific linguistic (Potter and Wetherell, 1987), contextual (Wodak, 1996) or ideological 20
features of text (Hirschman, 1990, 1988), CDA examines the interrelationships between all three micro, meso and macro levels. This multi-level approach raises a particular issue pertinent to the current study. If texts are seen to embody linguistic properties that in turn produce social realities and establish unequal power relations, there is an implication that texts merely (re-) construct and reflect concrete social processes and structures that exist “out there”. That is, by looking at the “concrete” processes that make up a particular social reality, researchers can then claim that the subject under investigation reflects universal social processes and, accordingly, structural regimes of domination. In response to this view the present study does not claim that consumers are passive to marketer‟s devices. In turn marketing managers are not seen to manipulate cultural texts that „produce‟ independent travellers. Nor, in specific relation to power, does it assume that texts are the embodiment of dominant discourses, grand-narratives and meta-myths grounded in human nature (Thompson, 2004). The current study therefore does not rest on a “strong” or “hard” social constructivist perspective. To extend Phillips and Hardy‟s matrix it is worth noting that discursive studies differ (and this current study is no exception) in terms of how texts are seen as a platform for developing theoretical truths as opposed to meaningful theoretical insights about the phenomenon under investigation. Reflected in Foucault‟s more structural view that contends there is nothing outside of discourse, some consumer researchers have adopted, for entirely legitimate reasons, what can be seen as a “hard” social constructivist perspective in their understandings of consumer texts. Thompson (2004: 162) notes of Hirschman (2000) and Stern (1995) a common structuralist view of the relationship between text, discourse and consumer reality: “A major structuralist tenet is that mythic archetypes are grounded in the most fundamental concerns of human experience, such as birth and death, kinship relationships…emotional conflicts, and humanity‟s relationship to the natural world.” In examining marketplace mythology and discourses of power, Thompson‟s (2004) critique of Holt‟s (2002) assertion that consumers are trapped within the ideological discourses of market capitalism, exposes a useful deviation between “hard” and “soft” approaches to discourse within consumer research. It would appear that critical approaches to discourse analysis, whilst generally defined as qualitative in nature, can legitimately adopt alternative, and sometimes complex epistemologies that, in turn, provoke alternative approaches to data, analysis and claims to “adequate knowledge” (Morgan and Smircich, 1980). In attempting to both define and implement CDA, it would seem that is simply not sufficient for researchers to understand the way in which metaphors, narratives and discourses work at a technical level. Resonating with Davies and Fitchett (2005), an „intricate‟ epistemological understanding of what such semantic devices can tell us about consumption are unavoidable, fundamental issues of sociological enquiry. CDA and discourse analysis more broadly should be viewed in terms of methodology rather than method (Philips and Hardy 2002). It is therefore open to complex issues of epistemology and ontology throughout the course of its implementation. The critical discourse-analytic approach performed in this study maintained a focus on the concrete, constitutive social processes of texts whilst linking them to and appraising them in terms of the meaning (rather than truth) of an ostensibly „independent‟ culture of tourist consumption. The focus of the discourse analysis was The Rough Guide To Spain (Ellingham and Fisher 2002), a self-proclaimed alternative guidebook for independent travellers. This is part of a larger series of books covering a wide range of county and city destinations around the world. The Rough Guide series is for “independent-minded visitors on any budget” 21
(www.roughguides.com), and the selection of this series helped us to develop a picture of independent travel somewhat broader than that depicted, for instance, in Lonely Planet, the “backpacker bible”. Each book in the Rough Guide series is written in a similar style and so a single text is largely illustrative of the series. However, the selection of the Spain title (“the ultimate handbook to one of Europe‟s most vibrant countries”) further consolidated our attempt to explore independent travel in its broadest sense. This is because the position of Spain as one of Europe‟s top holiday destinations suggests a very diverse readership of independent travellers for this particular title, compared to a more obviously backpackeroriented title such as The Rough Guide to India or The Rough Guide to Zanzibar. CONSTRUCTING INDEPENDENCE Independence is a central idea running through the sample text and our analysis suggested that it could be best understood in terms of three underlying themes that characterize what the book serves to achieve for the reader. By analyzing these themes– reifying inaccessibility, interpreting value, and constructing inauthenticity–we reveal that the independent traveller can be characterized as someone who defies inaccessibility, practices bargain hunting, and avoids inauthenticity. REIFYING INACCESSIBILTY An important theme in the text‟s discursive construction of the practice of independent travel can be identified as the construction of destinations as inaccessible. The text consistently articulates the destinations to which it sends its readers as being generally “hard to reach”, “tucked away”, and “remote”, where bus services are restricted and even accommodation is “thin on the ground”. The concept of inaccessibility is manifest throughout the entire text and is evident within clauses such as “Vacancies are extremely thin on the ground” (emphasis added, 959) and “Although the options are very limited” (emphasis added, 960). Evident right at the outset, the excerpt below contained in the introductory “basics” section of the guidebook is typical of the many sentences that centre on the problems of gaining access: “Unless you‟re travelling on a rail pass, buses will probably meet most of your transport needs; many smaller villages are accessible only by bus, almost always leaving from the capital of their province” (emphasis added, 30). The continued manifestation of such clauses and extended sentences prompted the coding of inaccessibility as a potentially significant theme. A fuller picture of the purposefulness of these inaccessibility clauses and extended sentences can be gauged by seeing them in the context of the process in which they are part. For example, in the following paragraph, note how the destination is constructed as challenging and difficult and in particular how, through this, a certain metaphor is generated for the reader: “Mallorca is at its scenic best in the gnarled ridge of the Serra de Tramuntana, the imposing mountain range which stretches the length of the island‟s western shore, its rearing peaks and plunging sea cliffs intermittently intercepted by valleys of olive and citrus groves and dotted with the most beguiling of the island‟s towns and villages. There are several possible routes which take in the best of the region, but perhaps the most straightforward if you‟re reliant on public transport is to travel up from Palma to Soller, in the middle of the coast, and use this town as a base, making selected forays along the coastal road, the C710; not far away to the southwest lies the mountain village of Deia and the 22
monastery of Valldemossa, while within easy striking distance to the northeast are the monastery of Lluc, the quaint town of Pollenca and the relaxing resort of Port de Pollenca” (emphasis added, 959). In the above passage, key words and clauses have been highlighted that constitute the presence of a particular type of language-use in the text. On their own and taken out of context words such as “imposing” or “intercepted” may seem like an ad hoc poetic play or dramatization of words; however, they appear to represent a kind of military discourse. At the outset, the word “imposing” articulates a daunting, oppositional force standing, as it were, in front of the reader‟s path. The use of a similarly military word “intercepted” further suggests that the imposing area lying in the path of the reader is like a battlefield. Then, the use of the clause, “possible routes” compounds the emerging battlefield scenario by suggesting that most routes are impossible and that finding a way through the front line will be a difficult feat. The presence and usage of a military discourse becomes even clearer as the passage unfolds. The text then suggests that towns should be used “as a base” from which “selected forays” can be made across the metaphorical front line. Finally, the use of the clause “striking distance” makes a potential journey to the destination Lluc seem like mounting a short military mission. The text here is not literally informing readers that there is a war going on and that you need to bring along a private army just to get around; its intention is plainly metaphorical. The representation of an ostensibly military discourse in the text functions to suggest that travelling to this destination is like waging a war or going into battle. Significantly, a direct corollary of this is that the process of moving through the area and reaching the given destinations is presented to the reader as an inherently difficult and challenging task. A sense of impenetrability or inaccessibility has been conveyed. However, the text, having erected this metaphorical barrier does not steer the reader away from the battlefield for it is actively directing the reader to towns, monasteries, and resorts on the other side of the apparent front line. Significantly, through the invocation of a military discourse to convey a strong sense of inaccessibility the would-be traveller is at the same instance semantically woven into the metaphorical position of something like a guerrilla fighter. Indeed, although it is not explicit in the text as to whom, we can assume that implicitly the reader or would-be traveller is supposed to be the militarily disposed person using towns as bases and launching strikes and forays. Therefore, the reader is offered the role of guerrilla fighter, the very character that could tackle the metaphorical battlefield and gain access to the otherwise inaccessible destinations. In the context of the passages analyzed above the text functions to do two essential things. First, the use of implicit metaphor and sentence structure can be seen as part of the text‟s working-up of the issue of access, or more aptly the lack of accessibility. That is, by suggesting, for example, that it takes a military strategist to reach certain destinations they can be seen as an integral part of a process of reifying inaccessibility. Second, the way in which passages are constructed, and particularly through metaphor, offers to the reader an interpretive position by which the possibility of surmounting or defying such inaccessibility becomes tenable. In sum, through the process of reifying inaccessibility the text has offered a subject position to the reader of the text as someone whose role as an independent traveller it is to defy inaccessibility.
23
INTERPRETING VALUE The analysis also demonstrated that a theme of cost was persistent the text, and that the guidebook would effectively interpret value for money for the independent traveller. The text frequently articulated accommodation, eating and drinking, transportation and most other activities featured in the book in terms of the financial value imparted to the traveller. Reflected in the example below, this theme was invoked right from the beginning of the guidebook in the „basics‟ section and repeatedly played itself out in the rest of the book: “Simple, reasonably priced rooms are still very widely available in Spain, and in almost any town you‟ll be able to get a no-frills double for around €15-21, a single for €9-15” (emphasis added, 42). In line with this example, typical clauses would read: “inexpensive tapas bars”, “good food at low prices”, “for just €70”, “prices are surprisingly low” and “very well-priced” (956). Interestingly, these clauses appear to resemble a kind of “good value” marketing discourse. They work to suggest that the reader can expect to find food of a good quality but importantly, that it will not cost too much, generating for the reader an interpretive position equivocal to the social role of the bargain hunter. For example: “The best place to eat is at the Hotel El Guia. It may be a little formal for some, but the prices are very reasonable, with a delicious menu del dia for around €15.” (emphasis added, 960). This sentence can be seen to embody a concession and counterargument structure whereby one piece of discourse is tactically marginalized and subjugated by the other. The opening clause begins with the factual statement; “The best place to eat is…” However, the text then places the discourse marker “may” signaling a concession in regards to this factual statement – “It may be a little formal….”. Immediately following this concession is the second discourse marker “but” which signifies the counter-argument whose purpose it is to return the reader to, and harmonize them with, the initial factual statement – “but the prices are very reasonable.” The overall discursive effect of this structure is to resolve an anticipated contradiction in the text where readers might associate eating in a formal hotel with conspicuousness and high prices. If left unresolved, this statement could be seen to contradict the prevalent discourse and the bargain hunter role implicit in the text. The statement, “but the prices are very reasonable” acts to resolve this tension and return the reader in the direction of the initial statement “The best place to eat is…” reaffirming the thematic proposition of good quality at low prices. What is slightly different about this sentence from the earlier sentence and clauses is that there is an extra dimension regarding the discursive context. In other parts of the analysis the emphasis fell on the relative cost between different travel, eating and accommodation options, however, here there is evidence of a struggle between two alternative discourses seemingly in conflict – perhaps good value versus conspicuous tourism. The point is that in line with the normal, prominent discourse forwarded in the text one would not expect readers to eat in formal and potentially expensive hotels – beach shacks serving inexpensive tapas are clearly the order of the day. What this sentence says is that travellers are prepared, on occasion, to utilize establishments that appear to deviate from the norm so long as they resonate with the interpretive repertoire offered by the role of bargain hunter. Therefore, we can extend this interpretive position to incorporate a notion of independent traveller as utilitarian – someone who is prepared to use a formal setting of a hotel to obtain good food at low prices. Essentially, the text works to interpret for the independent traveller a notion of value based on getting the most for your money. 24
CONSTRUCTING INAUTHENTICITY The emergence of authenticity as a key theme in the text‟s construction of independent travel was identified through the reoccurring manifestation of words such as “traditional,” “authentic,” and “real”. These words were often found to work in conjunction with the key cultural identities “Spanish,” “Mallorcan,” “Balearic,” or “Catalan”. Significantly, many of these clauses such as “authentically Mallorquin” (961) appeared juxtaposed against an inauthentic reality of Spain. Not only did the text appear to invoke a tension between the inauthentic and authentic, but in all cases it formulated authenticity as the preferred option for readers. For example, this excerpt, taken from the introduction of „The Balearic Islands‟ chapter: “Mallorca, the largest and best-known Balearic, also battles with its image, popularly reckoned as little more than sun, booze and high-rise hotels. In reality you‟ll find all the clichés, most of them crammed into the mega resorts of the Bay of Palma and the east coast, but there’s lots more besides: mountains, lovely old towns and some beautiful coves and the Balearic‟s one real city, Palma” (emphasis added, 929). The analysis of the excerpt below reveals more fully the purposefulness of the linguistic functioning of authenticity-related text: “In central Palma, especially along the harbour front and around Placa Llotja, many restaurants are unashamedly geared to the tourist trade, with menus in a babble of Euro-tongues. Most serve perfectly reasonable food, mainly grilled meats and fish, but away from these enclaves you‟ll find that prices are a little lower and menus more exclusively Catalan and Spanish” (emphasis added, 955-956). This paragraph embodies a linguistic structure in the form of a concession and counter-argument device. As mentioned previously, the general effect of this structure is to harmonize the reader with one piece of discourse by weaving in and then subjugating an anticipated contradictory piece of discourse. In this case, the opening statement can be seen as the formulation of the dominant proposition: “many restaurants are unashamedly geared to the tourist trade, with menus in a babble of Euro-tongues.” This statement essentially proposes that because of the (shameful) tourist trade, restaurants in this area do not reflect indigenous cuisine. This can be seen as an opening problematising strand of discourse. Then follows something of a concession–the food is of a reasonable quality–which represents an anticipated contradiction to the previous sentence. Swiftly following this concession is the counter-argument, the effect of which is to subjugate this previous concession and return the reader to the initial proposition. Here, the discourse marker “but” cues the counter-argument which, in line with the opening proposition, maintains the central proposition that “authentic” (Spanish) cuisine can only be found away from these enclaves that appear to be unashamedly geared towards the tourist trade. On a different level, the text can be seen to constitute a process by which inauthenticity is actively constructed as a problematic issue. Within the text, this issue is elevated to constitute a potential barrier to travellers experiencing the “real”, “authentic” Spain. Significantly, mirroring the discursive dynamics identified under the theme of “reifying inaccessibility”, the text simultaneously offers the reader the opportunity to surmount this perceived inauthenticity. This possibility for transcendence is evident where the texts actively directs readers away from “staged” cultural experiences and towards ones that seem to represent more closely the cultural practices of the indigenous Spaniards. Consequently, the repeated textual situations in which the authentic and inauthentic are juxtaposed and which 25
see, in the end, the reader‟s transcendence over inauthenticity, implicitly suggest the reader as someone whose social role as an independent traveller is to avoid inauthenticity. DISCUSSION: ENGENDERING DEPENDENCY Critically reflecting upon the explicit social processes now identified as central to a market discourse of independence, this next section discusses the relations of power implicit within those processes in the context of the broader hegemonic relationship between consumer and text. The discussion is focussed squarely upon critical reflection of the nature of independence as presented to the consumer, casting light on the underlying implications for the consumer and the paradoxical nature of the market ideology of independence. In the previous section, we saw how the alternative guidebook helped to construct independent travel through three main explicit themes, and explored the concomitant practices that were identified as constitutive of the role of the independent traveller. We now turn to look at how these themes can be seen to engender relations of power by implicitly promoting a form of dependency between the reader and the guidebook itself. The key argument presented here is that by referencing the explicit instructions of the guidebook, the tourist, as a text practitioner, effectively „becomes‟ an independent traveller, and is therefore inherently reliant upon the product for this transformation. Therefore, whilst apparently reflecting the market‟s ideology of independence, the three social processes highlighted by the analysis come to represent, in the end, authorization, control and power over what independent travel is about and what independent travellers should do in this role. As a result of this, the consumer of independent travel is implicitly situated in a position of inherent subjectivity to and disempowerment by these mediated market processes, leading to the conclusion that independent travel is enacted, albeit contradictorily, in a context of security, reassurance and ultimately, dependency. The following passage helps to demonstrate how the text functions to engender this dependence between tourist and text: “Arriving by train at Soller the obvious option is to continue by tram….down to the seashore, a rumbling, 5km-journey ending at Port de Soller. If you pass straight through, however, you’ll miss one of the most laid-back and enjoyable towns on Mallorca, an ideal and fairly inexpensive base for exploring the surrounding mountains.” (emphasis added, 960). The key to understanding the functioning of the passage lies in the use of the central words “If” and “however.” The use of the word “If” in the first conditional tense serves to link together two key events: “passing through” and “missing out.” As a result, this sentence structure makes the event of “missing out” a condition of “passing through.” The position of the discourse marker “however,” in the middle of this conditional clause, can be seen to work in conjunction with the word “If.” As a contrasting discourse marker “however” acts to emphasize that the second part of the passage contrasts the first. In the context of the passage, this marker makes sure the reader understands that passing through and continuing down the “obvious” route to the sea will mean missing out on what is an “ideal” town in which you can be a bargain hunter (“inexpensive base”) and also defy inaccessibility (“exploring the surrounding mountains”). Significantly, it must be noted that not only does “however” contrast the two events–going to the sea or staying in Soller–but it can also be seen to contrast two potential roles for the reader: the traveller who is ignorant to advice and eventually loses out by doing only what is obvious to himself is set against a traveller who, by heeding to the book‟s advice, is subsequently rewarded. In some ways then, through its linguistic functioning, the text can be seen to implicitly weave the reader into a social role equivocal to a disciple–one who listens well, follows 26
instructions, and is duly rewarded. Reifying inaccessibility, interpreting value, and constructing authenticity are explicit social processes that, in defining the appropriate meanings for the consumer act to promote dependency between reader and guidebook. In this sense, The Rough Guide text can be seen both to make independent travel meaningful through a certain set of tourism practices and, at the same time, cement the tourist‟s dependency upon the text itself. This resonates with other studies that contend that guidebooks play a key role in shaping the traveller‟s engagement with destinations (Bhattacharyya 1997; McGregor 2000; Siegenthaler 2002), but extends this line of reasoning to show how the identity of independent travellers, and the meaning of independent travel itself, are inextricably bound up in the discourse of alternative guidebooks, thereby challenging the notions of freedom and autonomy at the centre of the market ideology of independence. How, though, is this contradiction between independence and dependence successfully maintained for the independent traveller? It would seem that the alternative guidebook offers tourists a self-image of risk, adventure and autonomy–an independent travel identity (Sorensen 2003)–whilst also offering them protection from the concomitant uncertainty, unfamiliarity and fear that independence in travel might be expected to bring. The modes of practice that confer this identity are characterized by a specific “tourist gaze” (Urry 1990) upon bargains, inaccessibility and authenticity, yet underlined (and undermined) by a dependency embedded in the implicit instructions and normative values that construct this identity. As such, the notion of the independent traveller represents a powerful cultural myth to anchor understandings of tourist experiences and to give meaning to otherwise potentially contradictory tourism behaviors (Belk and Costa 1998; Stern 1995; Thompson 2004). Here then, we see identity wrapped up not so much in tourist possessions (e.g. Belk 1988) but in a set of meaningful, but in many ways quite mundane, social practices. It should be noted that because CDA is concerned with implicit relations of power, the critical level of analysis could have generated different kinds of insights from the analysis of a text, perhaps to do with gender, ethnicity, class or morality. Taking, for instance, a critical interest in the way texts marginalise certain actors, explicit processes such as “defying inaccessibility” could be seen as implicitly subverting disabled groups where „good access‟ is often a condition of ones participation in consumption. At least at a critical level, the purpose and nature of the study largely influences the nature of the findings. It was our motivation towards gaining a better understanding of the paradoxical nature of consumer freedom and autonomy within a largely commodified market discourse of independence that shaped the current study. CDA facilitated this understanding through a structured examination of a mass produced consumer text, leading to the conclusion that because it is the market that ultimately mediates the „myth of independence‟ for consumers, consumers here are implicitly less empowered, autonomous or free than the notion of independence they seek in their consumption would imply. CONCLUSION In this paper, the nature and meaning of independent travel and the independent traveller have been explored. Although consumption of independent travel represents a substantial and increasing proportion of the travel and tourism industry, it was found that the existing literature relied on fairly unsophisticated, and rather unhelpful conceptualizations of the practice. As a result, the conceptualization of independent travel developed in this paper is not so much based on a set of tourist preferences constituting a particular niche, or still less a specific form of distribution of holiday products, but rather as a cultural practice. This 27
represents an important, and potentially illuminating, new direction for identifying and understanding independent travellers that has yet received only scant attention. Whilst further research on travellers is needed to refine this alternative conceptualization, the study reported here has offered some important first steps in delineating the nature of independent travel as a cultural practice. Identifying the alternative guidebook as a key repository of cultural knowledge and signification about such practices, the study reported here used discourse analysis of a representative text to explore how guidebooks worked to construct independence and give meaning to the notion of an independent traveller. The analysis suggested that alternative guidebooks construct independence by reifying inaccessibility, interpreting value, and constructing inauthenticity, promulgating a powerful myth of the independent traveller as someone who defies inaccessibility, hunts for bargains, and avoids inauthenticity. However, the analysis of the text played out not at one but at two levels. What was perhaps most significant about this though was that each of these practices also acted to engender an implicit relation of dependency and thus power between the text and the consumer that contradicted, but ultimately did not threaten, the whole notion of independence communicated by the text. These observations are particularly significant given that independent travel, by definition, is typically identified as a tourism practice that is autonomous, empowered and free. Yet, our analysis offers an alternative interpretation that suggests independent travel is simultaneously a controlled and constrained tourism practice. Hence, in contrast to Hyde and Lawson (2003), who characterize independent travellers as experiencing an evolving itinerary, a willingness to take risks, and having a desire to experience the unplanned, the discourse of alternative guidebooks suggests a tightly controlled set of predetermined itineraries that enable the traveller to defy inaccessibility and avoid inauthenticity; a comforting security in knowing how this can be achieved, and in the most cost-effective way; and offering a thoroughly mediated travel experience. It was only through CDA‟s dual concern with the explicit and implicit meanings of text that we came to understand and were able to synthesise the hegemonic nature of the relationship between the producer and consumer of independence at a deeper, more critical level; challenging the nature of consumer freedom within a market predicated upon and largely justified through an ideology of independence. These findings certainly beg further research as to how tourists themselves manage this relation with the guidebook, and how (or even if) they sustain an independent traveller identity in this context. Future studies could therefore usefully explore the symbolic role that alternative guidebooks play in the on-going identity construction of independent travellers, both in terms of their positive and negative contributions to a sense of personal independence. Furthermore, research into the mediated nature of other consumption activities is required that considers, in particular, the context of market hegemony and the corresponding degrees of „freedom‟ in which consumption practices take place (Holt, 2002). Here, more critical approaches may support such inquiry, especially where there is call for closer examination of the negative, disempowering and marginalizing features of marketbased relationships both for the consumers who take part in them and for society more broadly (Denzin, 2001). To conclude, the study reported here suggests that the independent traveller constructed in alternative guidebooks defies inaccessibility, hunts for bargains, and avoids inauthenticity, but whether the traveller accepts this version of themselves, or even if they do, how exactly the guidebook shapes their enactment and meaning, remains to be seen. 28
The analysis serves to further highlight the socially constructed nature of independence in marketing discourse. In this, it also throws light on the power of the market over consumers, even within consumption practices that are most obviously about being autonomous and free. The myth of independence is shown to operate as a powerful marketing differentiator that allows consumers a viable space in which to construct a positive and valued sense of self. What is interesting is the extent to which highly opposing experiences are able to co-exist in a kind of mutual consumer symbiosis, whereby the myth of the independent traveller, and the romantic search for the authentic (Campbell, 1986) are able to sit alongside the realities of mass produced independent tour guides and largely commodified destinations. As this contradiction, which lies at the heart of many consumer encounters, becomes more prolific, critical discourse analysis is able to illustrate how consumers are able to mediate tensions in such a way so as not to diminish the value and meaning that can be enacted.
29
REFERENCES Belk, R. W. 1988 Possessions and the extended self. Journal of Consumer Research 15 (2):139168 Belk, R.W. and J. Costa 1998 The mountain man myth: a contemporary consuming fantasy. Journal of Consumer Research 25:218-40. Benjamin, W 2002 The Arcades Project. Harvard: Belknap. Bhattacharya, D. 1997 Mediating India: an analysis of a guidebook. Annals of Tourism Research 24:37189. Bourdieu, P. 1984. Distinction. London: Routledge. Campbell, C. 1986 The Romantic Ethic and Spirit of Modern Consumerism. London: Blackwell. Denzin, N. K. 2001 The seventh moment: qualitative inquiry and practices of a more radical consumer research. Journal of Consumer Research 28:324-330. Dicks, B. 2003 Culture On Display: The Production Of Contemporary Visitability. Maidenhead: Open University Press. Ellingham, M. and J. Fisher 2002 The Rough Guide To Spain. Tenth Edition. London: Rough Guides. Elsrud, T. 2001 Risk creation in travelling - Backpacker adventure narration. Annals of Tourism Research 28 (3):597-617. Fairclough, N. 1992 Discourse And Social Change. Cambridge: Polity Press. Frank, R.H 1999 Luxury Fever. Money and Happiness in an Era of Excess. Princeton University Press. Hirschman, E.C. 2003 Men, Dogs, Guns, and Cars: The Semiotics of Rugged Individualism. Journal of Advertising 1: 9-22. Hyde, K. F. and R. Lawson 30
2003 The nature of independent travel. Journal of Travel Research 42:13-23. Jacobsen, J.K.S. 2004 Roaming Romantics: Solitude-seeking and Self-centredness in Sightseeing. Scandinavian Journal of Hospitality & Tourism 4: 5-23.
Scenic
Loker-Murphy, L. and P. L. Pearce 1995 Young budget travellers - backpackers in Australia. Annals of Tourism Research 22:819-43. May, J. 1996 In search of authenticity off and on the beaten track. Environment and Planning D-Society & Space 14:709-36. Morrison, A. M., S. Hsieh, and J.T. O‟Leary 1994 A Comparison of the Travel Arrangements of International Travellers from France, Germany, and the UK. Tourism Management 15 (6):451-63. Noy, C. 2002 “You must go trek there”: The persuasive genre of narration among Israeli backpackers. Narrative Inquiry 12 (2):261-90. Phillips, N. and C. Hardy 2002 Discourse Analysis: Investigating Processes Of Social Construction. London: Sage. Siegenthaler, P 2002 Hiroshima and Nagasaki in Japanese guidebooks. Annals of Tourism Research 29:1110-1137. Sørensen, A. 2003 Backpacker ethnography. Annals of Tourism Research 30:847-67. Stern, B. A. 1995 Consumer myths: Frye‟s taxonomy and the structural analysis of a consumer text. Journal of Consumer Research 22 (2):165-85. Thompson, C. J. 2004 Marketplace mythology and discourses of power. Journal of Consumer Research 31 (1):162-80. Urry, J. 1990 The Tourist Gaze: Leisure And Travel In Contemporary Societies. London: Sage. Weber, K. 2001 Outdoor adventure tourism - A review of research approaches. Annals of Tourism Research 28:360-377.
31
Causing tourism
Robert Young Meng Jun Wu Anita Fernandez Young
Business School University of Nottingham England
Abstract The event of an individual taking a particular holiday, or experiencing any tourism outcome, can be regarded as the end of a causal chain involving potentially several causal factors and sometimes extending over a prolonged period of time. This raises a number of questions about what it means to cause tourism, especially when there is uncertainty within the causal chain. When we say that a stimulus causes a tourist event do we mean that if the stimulus occurs then the event will follow, or do we mean that but for the stimulus the event would not have taken place? Given uncertainty in the causal connection, how are we to evaluate the contribution of a stimulus to the occurrence of the tourist event. This paper adopts a general formal model of causing recently introduced into the economics literature and adapts it to the particular features of tourism outcomes. The resultant model gives rise to a three-phase dynamic of the evolution of the causal relationship between causes of tourism and tourist events. The model has general applicability to the evolution of the effects of and on markets channels. Examples are used to illustrate the generality of the model as a unified explanation of empirically observed phenomena relating to destination marketing, advertising and branding.
32
1.
Introduction
Consider a destination manager rebranding a destination, a package tour operator cutting the price of a package with an imminent departure date, an airline advertising a new route. These are all examples of purposive acts. In each instance, there is an intended outcome in terms of tourism – potential tourists visiting a destination, buying a package tour or an airline ticket. Behind each intention is an underlying proposition to the effect that the act in question causes tourism. For any tourism outcome we care to consider there is a multiplicity of factors which contribute to its occurrence. The various decision makers, from government through players in the tourism industries to the tourist herself, rely on a causal underpinning in making their decisions. In any instance, there may be differences of opinion as to what the principal causal factors are, how large an effect each of them has, perhaps even disagreement as to why a particular factor has an effect. What is beyond dispute is that we can do an act A to bring about an outcome D if and only if the act actually causes the outcome. This much at least is common ground among us all and we know it to the extent that we don‟t even have to talk about it. Or do we? When we say that a particular act causes a tourism outcome, do we mean that whenever the act is done the outcome follows? If A then D? Or do we mean that but for the act the outcome would not have occurred? If not A then not D? Consider the following sequence of events. Shug is a devoted husband who always takes the advice of his wife Senga. One evening Shug sees a television travel program featuring the island of Bermados, a destination he had never heard of before. He thinks that this looks like a nice place for a holiday. Later that evening, Senga returns from her Creative Accounting class. Shug mentions Bermados to her. Senga says that she has heard good things about Bermados and that they should go there. The following morning Shug books a holiday package to Bermados. What caused Shug to book the holiday? In our account there are two events we can identify as causes. The first is the television item, the second is Senga‟s advice. But in respect of each of these, what do we mean by „caused‟? The television feature caused Shug to book the holiday in the „if not … then not‟ sense. But for seeing the feature Shug would not have known about Bermados and, therefore, would not have booked the holiday there. The television program did not cause the outcome in the „if … then‟ sense. In spite of seeing the feature on Bermados Shug might still very well not have made the booking. In contrast, when Senga expressed her advice that they should go to Bermados she did cause Shug‟s booking in the „if … then‟ sense, because Shug invariably follows her advice. However, Senga‟s advice did not cause the booking in the „if not … then not‟ sense. Even if Senga had expressed no prescription Shug might have decided to book the trip anyway. This little tourism tale illustrates two points to do with causing. First, the two senses in which we might understand causing are distinct: one may be the case while the other is not. Secondly, it is by no means obvious which of the two possible interpretations of causing we have in mind when we think about causing tourism.
33
In this paper we consider a formal model of causing applied to tourism outcomes, with particular reference to its implications for tourism marketing. A general model is set forth in Section 2 and applied to the special features of causing tourism in Section 3. In Sections 4 and 5 we consider the implications of the model for assessing and optimizing tourism marketing, first in the short run and then in the long run. Section 6 concludes. 2.
A general model of causing
In Section 1, we raised two different meanings of the verb “to cause”: the „if … then‟ sense and the „if not … then not‟ or „but for‟ sense. The first of these is the scientist‟s concept of causality. The second is the lawyer‟s concept of causation. Applying to causing tourism, the event of an individual taking a particular holiday is the final effect of a causal chain from potentially several causal factors. Below this general formal model of causing will be illustrated in the sense of causal certainty and then causal uncertainty towards what it means to cause tourism. The scientist‟s concept of causality: Without uncertainty, suppose an act A caused D, it implies that occurrence of A directly leads to the occurrence of D: AD. However, this notion involves the further ingredient of necessity: in the philosophical sense, one could say that D necessarily follows from A; or in every instance D always follows A. For instance, in the instance of tourism advertising, one could say that the act of advertising necessarily leads to potential tourists buying the product. The purchase follows from the act of advertising every time. The lawyer‟s concept of causation: A caused D if, but for A, D would not have happened: ~A ~D. That is to say, under the circumstance of certainty, it is certain that without the act A, D would not have happened. Continuing the advertising example, therefore, here one could say that the product would not be purchased if there is no advertising. However, in all instances of real interest, the assuming away of uncertainty in this way is unjustified. Corresponding to the two concepts of causing, there are two types of causal uncertainty. First, the event (such as advertising) does not make the outcome inevitable. An intended outcome is not always a guaranteed outcome. Secondly, in the absence of the event the outcome is not, in general, impossible. It might occur anyway. Causal uncertainty implies that D follows from A is only true in some cases while not in others. Suppose there is a cause of A* of D, which is true in every instance that D follows from A*.1 As a result, in any sense if A causes the occurrence of D, then A must be part of A*. There must also be another part say X of A*, so A* = A∩X, so that A is an insufficient but necessary part of a sufficient but not necessary condition for D.2 Since A* is a strict cause of D, it follows that D←A∩X. However, this notion does not construct as this A∩X ←D, that is to say that we cannot exclude the possibility of other causes of D other than A*. It implies that a tourist event could follow from several causal factors rather than only constrained in 1
The following analysis and model derive from R J Young, M Faure and P Fenn, ‘Causality and Causation in Tort Law’, International Review of Law and Economics (2004). 2 This logical relationship between A and D corresponds to a general definition of causing set forth in J L Mackie ‘The Cement of the Universe’, Oxford (1974).
34
advertising. It is also important to note that such an event X could be recognised as „implicit‟ events as it may not be immediately obvious what X is, not is it necessarily the case that X could be discovered. The idea that A is a cause of D if and only if A is a necessary but not sufficient part of a sufficient but not necessary antecedent to D, is illustrated as Figure 1 Figure 1 A
A
O
C
D
B In Figure 1, the point A represents the act or the cause of D. O stands for the initial state of the world as the origin of the events we are considering. The outcome D occurs if and only if there occurs a combination of events which completes a path from O to D. A, B and C denote „implicit‟ events. These allow for the causal uncertainty. It is possible for D to fail to happen when A occurs because a path from O to D is completed only if A and C also occur. The outcome D can happen in the absence of the event A because there is a path from O to D via B and C which does not involve A. In other words, from the existence of causal uncertainty we can infer that there must (in general) be events A, B and C in the causal chain leading from A to D. Algebraically, Figure 1 is equivalent to D [(A∩A‟) B] ∩C There are two sufficient conditions for D: D A∩A∩C D B∩C A is a necessary but insufficient part of the sufficient condition A∩A∩C, which is not a necessary condition for the occurrence of D because B∩C is an alternative sufficient condition. The uncertainty in this model derives from the fact that each of the implicit events A, B and C may or may not occur. Let the probabilities of occurrence of A, B and C be respectively q, 1- and p. Now, since there is causal uncertainty we cannot say unambiguously that A causes D in either causal sense. However, based on the probabilities of occurrence of A, B and C we can derive a pair of causal probabilities which show the extent to which A causes D in each of the two senses. We define the probability in causality (k) to be the probability that D occurs following A occurring. The probability in causation (c) is the probability that but for A D would not have occurred.
35
Young et al (2004) derives the following expressions for k and c in terms of the probabilities of the implicit events. k= (1-+q)p c= q/(1-+q) So, in general, when A and D are causally related there is an element of each of the two senses of causing in the relationship. The causal probabilities k and c, respectively, express the extent to which causality and causation are present. 3.
Causing tourism
When we consider the question of causing tourism from the perspective of this model, two particular features emerge. The first of these is to do with the timing of cause and effect. When D follows from A, D occurs only once a path from O through A to D has been completed, i.e. only once A and C have occurred. Consider any element, say E, of A or of C. D results from A only once every such E has occurred. The temporal relationship between A and any E may be any one of the following: (a) E may pre-exist; (b) E may occur contemporaneously with A; (c) E may follow A after some given time lag; (d) the timing of E may be unrelated to the occurrence of A. Where the outcome is a potential tourist taking a holiday there is the particular question of leave from work. Let E be the event that the potential tourist has leave from work. In the instance of a tourist who first takes leave and then looks for a holiday we have situation (a). If the tourist can take leave instantly we have situation (b). If a fixed amount of notice is required we have (c). If leave can be taken only when circumstances permit we have situation (d). The implication here is that the time gap between A and D may well be substantial and unrelated to the occurrence of A.3 Secondly, in tourism there is often an effect of the outcome on the causal chain. This arises from two related effects. Once an outcome has occurred for a particular tourist this may well affect the probability of it occurring again in the future. Correspondingly, at the level of aggregate occurrence of the outcome once one tourist has experienced the event this may well increase the probability of the outcome occurring for other tourists. An obvious example is the case where the outcome is a visit to a new destination. So the phenomenon here is that the occurrence of the outcome makes the future occurrence of the outcome more probable. In terms of our model, the probability of the outcome derives from the probabilities of the implicit events in the causal chain, p, q and . We illustrate this in Figure 2.
3
In particular, the time interval between A and D need not accord with any conventional times series (vector ARIMA) model.
36
Figure 2 A
A
O
C
D
B
q
p
When the outcome D occurs it alters the probabilities in the causal chain in such a way that the probability of D is now increased. This effect follows whether A is repeated in the future or not. From the model in Section 2, if A is done the probability of D is (1-+q)p. If A is not done the probability of D is (1-)p. If both of these are increased when D has occurred, the change in the causal chain must be either an increase in p or a decrease in . In other words, once D has occurred, there is an increase in either: (a) the probability of D flowing from A in the future; or (b) the probability in the future of D flowing from some background source, i.e. some cause other than A. These considerations have implications for the long term promotion of tourist outcomes and we consider these in Section 4 below. But first, we consider now the short term implications of our causal model. 4.
Short term implications of the model
Consider the decision of a party in the position to exert some influence as to whether or not to do some act that would be causative of a tourism outcome. If there were no causal uncertainty then the outcome would occur if and only if the act is done. We could then ascribe the value of the outcome, whatever that might be, to the act. A decision could then be based on a comparison of the cost of doing the act and the value of achieving the outcome. However, in the presence of causal uncertainty the full value of the outcome cannot be ascribed to the act. The two reasons for this are that the act might not produce the outcome and that the outcome might have happened anyway. This raises the question of how we should evaluate the productive aspect of an act intended to cause a tourism outcome.
37
In terms of the probability of the outcome, the effect of doing A is that the probability of D is increased from (1-)p to (1-+q)p, i.e. by the amount qp. Referring back to Section 2 and the causal probabilities in causality (k) and in causation (c) we have the result kc = qp In other words, the contribution of the act A to the probability of the tourism outcome happening is kc, the product of the probability in causality and the probability in causation. This offers some insight into how we should evaluate the contribution of a causative factor. If we are to ascribe value to a causative act, deriving from the potential outcome, then the act in question must be causative of the outcome in both senses, i.e. in terms of causality and in terms of causation. An act attracts a large value (ceteris paribus) where p, q and are large. This can be interpreted as being where: (a) qp is large so that there is a large probability of the outcome flowing from the act; and (b) p is large so that there is a large probability of the outcome happening whether it flows from the act or otherwise; and (c) is large so that 1- is small and there is (in that) only a small probability of the outcome happening when the act is not done. In particular, the value of an act depends not just on the probability that it will bring about the outcome, but also on the probability that the outcome would have happened anyway. The probability of bringing about the outcome contributes positively to the act‟s value. The probability that the outcome would have happened anyway contributes negatively to the value of the act. In short, a high value act is one which creates a large probability of a tourism outcome which would otherwise be unlikely to occur. It may be interesting to note that there is an inherent distinction in this between the probability of the outcome when the act is done and the probability of the outcome flowing from the act. The latter probability is qp. The former probability is (1-+q)p. This includes the possibility that the act is done and the outcome happens but not because of the act. To illustrate these points, consider the example of film tourism. Let the act A be the featuring of some particular location in a film. Let D be the outcome that a tourist visits the location. How are we to evaluate the contribution of the film to tourism to that location? Our model is to the effect that the contribution is proportional to the causal probabilities k and c. k is the probability that a tourist visits the location, given that it has been featured in the film. Prima facie, this might appear to be what we‟re after – the contribution of the film is proportional to the number of tourists who visit after the location has been featured. However, the implication of our causal analysis is that we must take into account also c, the probability that but for the film the tourist would not have visited the location. At the aggregate level, the product of causal probabilities kc tells us the proportion of potential tourists who visit the location as a result of the film and who would not otherwise have visited. 5.
Long term implications 38
In Section 3 we concluded that the achievement of the outcome D makes its recurrence in the future more likely and that we can infer from this a change in the probabilities in the causal chain, specifically an increase in p or a reduction in . There are two primary stimuli to D: the act A and the background possibility B. An increase in p means that D is now more likely to occur irrespective of which of the primary stimuli is present. In contrast, a reduction in means that the background stimulus is more likely to be present. Whichever of these may be the reason for the increase in P(D) the implication is that there will be a long term effect of A on D beyond the initial effect. The initial effect comes about through the causal chain. The long term effect comes about because of a change in the causal chain. Although both versions of the long term effect result in an increased occurrence of the outcome, their implications for the effectiveness of the act A differ diametrically. For ease of reference, let us refer to the case in which p increases as Case P and that in which decreases as Case . Recall (from Section 2) that the contribution of A to the outcome D is proportional to qp. In Case P, the effectiveness of A in causing D has been increased. The reason for this is as follows. Whether the primary stimulus to D is A or B, the implicit event C is necessary to convert the stimulus into the outcome. p is the probability of C occurring. Therefore, an increase in p implies that if A is now done there is an increased likelihood of it being converted into the outcome. This creates an increased incentive to doing A because of the increased probability in causality. In Case the effectiveness of A causing D has not been increased – it remains unaltered. The reason that D is now more likely to occur is that the background stimulus B is now more likely to occur. This has the effect of reducing the probability in causation. In the absence of A being done, it is now more likely than it was that D will occur anyway. In this, the incentive to doing A has been reduced. Reverting to our example of film tourism, consider the following distinction. Suppose that a potential tourist T sees a film in which a particular location is featured. T visits the location. She is favourably impressed by her experience of the visit. Now consider two alternative continuations of the story. Version 1. T‟s friend S now sees the film. She discusses it with T. T remarks that she has actually been to visit the location where the film was made and it was just like in the movie. This influences S to make a visit to the location. Here we have an increase in p – the probability of the stimulus A being converted into the outcome D. Version 2. T‟s friend S has not seen the film and she is unaware of the location as a potential tourist destination. T tells S about visiting the location. S now has this location in mind as a potential destination. Here we have an increase in 1- (reduction in ) the probability of S being stimulated towards a visit other than through seeing the movie. In the first version of events the effectiveness of the film in causing visits to the destination has been increased. Future screenings, video release etc. will have an increased effect on visits as a result of the now greater probability in causality. In the second version of events, 39
the need for the film to cause visits has been reduced – there is now a smaller probability in causality. Finally, in the long run there is potential for the activity level (i.e. repetition of A) to be altered in response to its value. Suppose that A is a burst of advertising for a destination. If the long run effect of a burst is as in Case P, there is now greater value attached to a potential second burst. On the other hand, if we have Case the value of a second burst has been reduced. These possibilities are illustrated in Figures 3 and 4 respectively. Figure 3
A
A
O
C
+
D
B
+ p
Figure 4
A
A
O
+
C
D
B
In Case P, Figure 3, the long term effect of the outcome being achieved is to increase p. The larger value of p implies an increase in the value of A. Correspondingly, there is more incentive to do A and the activity is increased. In other words, there is positive feedback from A to itself. 40
In Case , Figure 4, there is negative feedback from A to D to to A. Although the long run effect increases the outcome level D, it makes A less valuable because there is less need to do A in order to achieve D. These long run possibilities are summed up in the following schematic and illustrative time path diagram. Figure 5
A
D
Secondary increase in D Case
Case P Increased A
Tertiary increase in D
Decreased A
Tertiary reduction in D
Figure 6 Case P Outcome level Case
time The point of bifurcation in the time paths of Figure 6 corresponds to the branching into Case P or Case in Figure 5. If the effect of the outcome is to make the outcome more likely in the future because of an increased background stimulus level, then the rational response to this will be a reduction in the activity level A. Once the causal chain parameters have been fully 41
adjusted by the increased outcome level, the reduction in advertising will lead to a dominant reduction in the outcome level. These alternatives are reminiscent of the possibilities in the development of a tourist destination. After an initial period of expansion there is a period of consolidation and further growth followed by either continued growth or the onset of decline. However, the difference here is that the patterns emerge from a completely general model which depends only on the circumstance of there being a factor which causes the tourism outcome and feedback from the outcome to the parameters of the causal chain. 6.
Conclusion
In this paper we have considered what we can mean when we say that an act or activity causes a tourism outcome. We adopt a completely general model of causing and adapt it to the special circumstances of tourism. In particular, we accommodate the long run effect of the outcome level on the probabilities in the causal chain and hence the activity level. The value of an activity as a contribution to the occurrence of the outcome depends on the conjunction of the two concepts of causing: causality and causation. Quantitatively, the contribution of the activity to the outcome is proportional to the product of the probability in causality and the probability in causation. Any consideration of the causal chain leading to a tourist outcome depends on an appreciation of causing from both these perspectives. In the short run, activities that are most worthwhile are those which create a large probability of the outcome happening, but only where the outcome would be otherwise unlikely. The essence of the matter is that where there is already a large probability of the outcome happening there is (correspondingly) little upside potential for increasing the probability through the activity. It follows that, in assessing a potential activity, it is essential to consider not only the probability that the outcome will occur, but also the probability that but for the act in contemplation the outcome would not occur. However, the question of the impact of the activity on the outcome is further complicated by the fact that the outcome occurs only once all the necessary implicit events or circumstances in the causal chain are in place. Such a situation may pre-date the act A, it may be coincident with A, it may follow A after a time lag or it may be independent of A. This implies that the timing of the act A and the outcome D may need careful consideration of the effect of A is fully to be captured and evaluated. These two considerations (value and timing) have management implications in terms of both assessment and decision making. The general implication is that there must be a full perspective on the potential consumer of the tourism product. A proper appraisal of the causal probabilities in causality and in causation can be approached by addressing to purchasers, visitors etc. questions to solicit whether they were subject to a particular stimulus and to what extent it was true to say that but for the stimulus they would not have participated in the outcome. The issues of timing of completion of a path through the causal chain could be approached through questions about when the factors necessary to the outcome in fact came together. In the long run tourist outcomes affect the probabilities in the causal chain so as to make the outcome more likely in future. However, there are two mechanisms whereby this may come 42
about. These alternative mechanisms alter the value of the activity in opposing ways. This gives rise to two alternative time paths for the outcome level, one of continued growth and the other following growth with decline. Although this is a completely general result in respect of any activity which causes tourism, specific examples of these alternative patterns are already in the literature. In essence, some activities increase the ease of conversion of a stimulus (any stimulus) into the outcome and this makes the outcome more likely in the future. Other activities lead to an increase in the background stimuli, i.e. they make it more likely that the outcome will occur other than as it flows from the activity. In the latter case, as the activity is pursued it becomes more probable that the outcome will occur even in the absence of the activity. This leaves progressively less upside potential to continue to use the activity to produce the outcome. Expressed alternatively, the activity becomes progressively less necessary because it is becoming increasingly probable that the outcome will occur anyway. It is important from a practical perspective to keep in mind the difference here, because an effect of the latter sort is less preferable. The more preferred long term development is where the activity increases the probability of conversion rather than the background probability. For example, feedback from consumer to consumer is more beneficial in the long run where it induces an inclination to purchase and not product awareness.
43
THE IMPACT OF EXTERNAL SHOCKS ON TOURISM SECTOR: THE CASE OF TURKEY
Murat KARAGOZ4
Ali SEN5
Ali KOCYIGIT2
PROPOSAL ABSTRACT The objective of this study is to assess the temporal impacts on Turkey‟s tourist arrivals. An econometric strategy is selected to determine the existence of unit roots in data series containing the quarterly number of tourist arrivals between 1993:1 and 2005:3. The present study finds that the data series contain deterministic trend and seasonal components together with detected structural changes. Hence any form of exogenous shocks, will not have permanent impact on Turkey‟s tourist arrivals. Key words: Temporal Effects, Turkey, Tourism Industry, Unit Root. INTRODUCTION Tourism is one of the most important economic pillars of the Turkish economy. It contributed around 5 % to Turkey‟s GDP over the last decade. However, in a span of 16 years, Turkey, together with other external events, struggled with many crises. In last decade, Turkish economy, alongside the other global crises, struggled with several internal crises. The 1994 financial crisis was caused by both the foreign exchange and stock markets. In August 1999 one of the most disastrous earthquakes of the history has taken place. In 2001, the terrorist attacks on the United States and its aftermath sent shock waves through out the region. Moreover, early 2003 the economic growth of Middle East countries threatened by US attack on Iraq. As a top global event all the phases of Iraq war were on the top of mass media headlines. The effect of such external events on the tourism will be temporary if the tourism demand series is trend stationary rather then difference stationary. In the first case, time series evolves around a deterministic time trend and deterministic seasonal 4
Associate Professor, Inonu University, Faculty of Economics and Administarive Sciences, Department of Econometrics,
[email protected] 5 Assistant Professors, Inonu University, Faculty of Economics and Administarive Sciences, Department of Economics,
[email protected], and
[email protected].
44
components. In the second case, data imitates an unpredictable random walk process and an external shock to this sort of series will have a permanent effect. We start with a prediction that the impact of external shocks on the tourism industry in Turkey would be temporary. There is no previous study that has attempted to forecast the impact of an exogenous shock on the tourism sector of Turkey. Nevertheless there are some studies for other countries. Enders et al. (1992), for example used the ARIMA model to predict the effect of terrorism on the tourist industry in Western Europe. Raymond (2001) examines issues relating to the impact of economic factors on tourist expenditure and hotel room occupancy rate. The article first presents an overview of the tourism industry in Hong Kong, an examination of the particular aspects of the tourism industry, and an estimate of tourist spending and hotel room occupancy rates. Using an expectations model, real tourism expenditure is found to depend on expected income, expected exchange rate and price level. The results of the article also reveal that the equilibrium hotel occupancy rate is a function of tourist flows, exchange rates, price level and length of stay. Au et al (2005) assessed the temporal impact of SARS on the tourists‟ arrival in Hong Kong. The analysis finds that data series of 24 countries contain unit roots and hence any form of exogenous shocks, like the SARS epidemic, can have permanent impact on the number of tourist arrivals. The paper recommends that authorities should take source-country-specific measures to overcome the negative effect of SARS. Bhattacharya and Narayan (2005) provide evidence on the random walk hypothesis for visitor arrivals to India using the recently developed panel unit root tests. Test results allow them to reject the random walk hypothesis, implying that shocks to visitor arrivals to India from the 10 major source markets have a temporary effect on visitor arrivals. Halıcıoğlu (2004) empirically examines aggregate tourism demand function for Turkey using the time series data for the period 1960-2002. The total tourist arrivals into Turkey are related to world income, relative prices and transportation cost. She employs bounds testing cointegration procedure to compute the short and long-run elasticities of income, price, and transportation cost variables. She also implements CUSUM and CUSUMSQ stability tests on the aggregate tourism demand function. The empirical results indicate that income is the most significant variable in explaining the total tourist arrivals to Turkey and there exists a stable tourism demand function. Specifically, the objective of this research is to determine if the impact of external shocks on tourist arrivals in Turkey is temporary or permanent in nature. It is expected that tourists from different countries would act differently to any form of external shocks, and so the impact would vary from one source country to another. By choosing an appropriate econometric strategy we evaluate the impact of external events on tourist arrivals in Turkey. This study also offers methodological contributions to tourism research and future research directions by suggesting a way to quantify arrival patterns with respect to the concept of stationarity and demonstrates the use of method in the case of structural change. Adopting the approach in the present study could empirically evaluate the effects of other tourism crises.
45
The rest of the paper is organized as follows: The following section explains the methodological strategy we have used. This is followed by a discussion of our findings and finally we end with some concluding remarks.
THE EMPLOYED TIME SERIES METHODOLOGY We start with a description of the data by a uni-variate time plot. An examination of the time plot may reveal trend and seasonal fluctuations together with possible brakes or structural changes in the process. In simple terms, a series is said to be stationary if its statistical properties in terms of mean, variance and autocovariances remain constant along the time path. Under such circumstances, any form of external shock will have a transient and diminishing effect on the series, implying that the series will naturally return to its original property over the time. Conversely, a series is said to be non-stationary if it has non-constant mean, variance and autocovariances over the time path. Time series econometricians refer to this type of series as a random walk or unit root process. Since this type of series does not possess property of stationarity, any exogenous shock will persist and hence the effect on the series will be long-term in nature. In order to determine whether a particular series is stationary or non-stationary, one has to detect whether the data series contains a unit roots in levels and seasonal terms. There exist a number of procedures for testing the presence of unit root in a data series. Dickey and Fuller‟s (1979, 1981) ADF approach is a well known approach. This procedure has been elaborated for different cases by some authors. Peron (1989) has offered a formal procedure to test for unit roots in the presence of structural change. A modification of this strategy used by Enders (2005, p.213) and another modification introduced by Hoffmann et al. (2003). In case of trend and/or seasonality we should start with an assumption of deterministic trend and seasonal components as an alternative hypothesis. In fact this is the course of action followed by Dickey-Fuller (1979) type tests for unit roots. We have employed Peron‟s procedure that has been outlined in Enders (2004, p.203). In Table 1 below we present a small modification of this procedure. Table 1. The Procedure of Unit Root Test with Structural Change Step 1: Estimate the most probable deterministic trend and deterministic seasonal model together with possible structural change elements in constant term and slope of the deterministic trend. Save the residuals as de-trended and de-seasonalized values. Step 2: Form the sample autocorrelation function (ACF): rk , k=1, 2, …, K. Here K is chosen so as to be equal to approximately one third of total observation numbers, T/4. If the overall coefficients are statistically significant, that is exceeds two standard error bound 2 / T and decreasing very slowly over the lag numbers, this is an indication of unit roots at levels. Significant coefficients at seasonal periods, s, 2s, 3s,…, on the other hand are the indicative of seasonal unit roots. Perform Box-Pierce Q T rk2 or Ljung-Box (1978) tests Q T (T 2) rk2 /(T k ) over the residuals to detect the autocorrelations in levels and in seasonal periods. Later test is robust for small samples. If Q exceeds the chisquare critical value of K degrees of freedom reject the null of no auto-correlation 1 2 K 0 . If there is no sign of autocorrelations conclude that there is no unit root, otherwise go to step 3.
46
Step 3: Perform HEGY test (Hylleberg et al 1990) for seasonal unit roots. To this end form the variables e1t 1 et 1 et 2 et 3 et 4 , e2t 1 et 1 et 2 et 3 et 4 , e3t 1 et 1 et 3 and
e4t 1 et 2 et 4 . Then estimate 4 et a1e1t 1 a 2 e2t 2 a3 e3t 3 a 4 e4t 4 u t and its restricted form for a3 a4 0 . Perform the restricted F test for the null hypothesis of seasonal unit roots. The test statistics Fcal [( SSEr SSEu ) / m] : [ SSEu /(n K )] is for the restriction given by null hypothesis, where SSEr and SSEu are the restricted and unrestricted regressions‟ residual sum of squares, m=2 is the number of restrictions, n is the number of observations, K= 4 is the number of parameters in the full model. In case of seasonal units, take the seasonal differences of the original series and repeat the whole procedure, otherwise go to step 4. Step 4: Perform ADF test to detect the unit roots in levels, add the lagged valued of dependent variable in case of autocorrelation in the errors as follows et et 1 i et i vt . Use Peron‟s critical values for the null hypothesis of unit root 0 , in case of structural change. If there is a unit root, take the difference of the original series and repeat the whole procedure, otherwise conclude that the present model is the best.
The results of this methodology can be classified in five categories as follows (Au et al 2005): (1) Stationary–the impact of external shocks would gradually diminish over time. This would imply that in the long run, the number of tourists would return to the original constant level. (2) Trend-stationary–the impact of external shocks would gradually diminish over time. In this case the number of tourists would return to its long-term trend. Thus, in these cases the impact created by shocks is temporary. (3) Random walk–the impact of external shocks on non-stationary series would not diminish over time. This would imply that there would be a permanent effect on the number of tourists. (4) Random walk with a drift–the impact of external shocks on a non-stationary series with a drift behaves exactly the same as (3) except that the series is either drifting upward or downward. The impact of shocks on such a series is permanent. (5) Random walk with a drift and a trend–the impact of shocks on a non-stationary series with a drift and a trend would not diminish over time. However, since the data series exhibits a trend, both the trend and the shock contribute to the changes in the number of tourists. Although the impact of these shocks on such a series is permanent, the growth in the number of tourists is governed by a trend. Thus, the impact is not as harmful as in (3) and (4).
DATA AND EMPIRICAL FINDINGS We have selected the number of overnight visits of foreign nationals as a most probable proxy for the tourist numbers of Turkey. The monthly data on foreign arrivals are collected by General Directory of Security. State Institute of Statistics has tabulated and presented to us. Quarterly data has been produced by us from the monthly data. We start with a plot of time series data.
47
10000000
8000000
6000000
4000000
2000000
0 1993
1995
1997
1999
2001
2003
2005
Figure 1. Time Plot of Quarterly Tourist Arrivals of Turkey (1993:Q1-2005:Q3) From a cursory examination of Figure 1 above, several points can be detected. (1) Needless to say the seasonal pattern. We should expect seasonality either deterministic or stochastic. (2) There is a slightly increasing trend. Therefore we should again expect a unit root at levels or a deterministic time trend. (3) There is an obvious non-constant (increasing) variance over time. To prevent the potential disruptions to be caused by non-constant variance, we should log transform the series. (4) The negative effect of tragic earthquake August in 1999 can be seen easily from the figure. A dummy variable representing this event is expected to be statistically meaningful. (5) Another most important point for the Turkish tourism can be detected from the figure is that there is a clear structural change beginning from 2003. That positive change could be attributed to the economic stability which is an outcome of the political stability gained by parliamentary majority of one-party government. (6) Another equally important point to be made is that there is no sign of negative effect due to Iraqi war out-broken at March 2003. This can be partly inferred as the event started and finished before the pick season. As a first step of the analysis we conduct a Ljung-Box chi-square test for sample autocorrelation function (ACF) to establish whether or not seasonal and non-seasonal unit roots exists. The results presented at Figure 2 below are carried out over the log transformed series. 1,0
,5
0,0
-,5 Conf idence L imits
-1,0
Coef f icient 1
4
7
10
13
16
Figure 2. The Sample Auto-correlation Function of Tourist Arrivals 48
The overall coefficients are statistically significant and decreasing very slowly over the lag numbers, a case which is an indication of unit roots at levels. Significant coefficients at seasonal periods on the other hand are the indicative of seasonal unit roots. Although the trend and seasonality are certain components of Turkey‟s tourist arrivals, we have to start with an assumption that both of these components being deterministic. In Table 1 below we present the most probable deterministic trend and seasonal model together with structural change components. The deterministic trend variable TIME runs as t=1, 2, …, 51, staring from the first quarter of 1993 till the third quarter of 2005. We have employed dummy variable D99 for the 17th August 1999 earthquake taking the value 1, for the third quarter in 1999, zero for all the other periods. Another dummy variable D03 stands for the structural change in constant term, starting from the year 2003 takes the value 1 for all the periods beginning from the first quarter of 2003 till the end of data, and zero for all the periods before 2003. The dummy variable TD03 is an outcome of multiplication TIME and D03 variables and it stands for the structural change in the slope of trend variable TIME. Seasonal dummies D1, D2 and D3 take the value 1 for the quarters 1, 2 and 3 respectively and zero otherwise.
Table 2. Coefficients Estimates of Deterministic Trend and Seasonality Variablesa
b
Std. Error
beta
t
p
Const.
14.056
0.046
-
306.323
0.000
TIME
0.0160
0.002
0.426
10.658
0.000
D99
-0.330
0.115
-0.082
-2.876
0.006
D03
-2.028
0.485
-1.501
-4.183
0.000
D1
-0.449
0.044
-0.352
-10.203
0.000
D2
0.345
0.044
0.271
7.870
0.000
D3
0.716
0.045
0.561
15.962
0.000
TD03 0.0483 0.011 1.651 4.567 0.000 NOTES: a:Dependent Variable: LNTURS. b: Unstandardized Coefficients. Standardized Coefficients, Adjusted R-square: 0.962, DW Statistic: 1.556.
beta:
An examination of the estimation results presented in Table 2 above reveals that, all the individual parameter estimates are of high quality in that all the estimates statistically different from zero even in 1 % significance level. In spite of such a large number of parameters in the model, the adjusted coefficient of determination is still at a high level of 96th percentile. Durbin-Watson lower and upper critical values for the number of observation 50 and number of parameters 7 except constant term are 1.246 and 1.875 (Gujarati 1995, p.818). In the present model DW statistic exceeds the lower bound. We cannot comfortably accept the null hypothesis of no auto-correlation for the residuals of this model. DW statistic 1.556 is in between lower and upper critical values, that is, we are on the left inconclusive area. We have to elaborate on the autocorrelation or unit tests. Therefore we perform HEGY test for seasonal unit roots and DF test non-seasonal unit roots for the residuals of the estimated model. 49
To perform HEGY test for seasonal unit roots we have constructed variables e1t 1 , e 2 t 1 , e3t 1 and e 4 t 1 from the residuals of deterministic model as outlined in step 3 of Table 1. Then we have estimated the auxiliary regression of HEGY test and its restricted form for a3 a 4 0 . The results are presented in Table 3 below. Table 3. Parameter Estimates of HEGY Regressions Parameter Standard 1 2 Models Variables Estimates Errors Unrestricted e1t 1 -0.130 0.055 -0.388 0.090 e 2 t 1 0.046 0.100 e3t 1 Restricted
t Statistics -2.346 -4.292 0.468
p 0.024 0.000 0.642
e 4 t 1
-0.250
0.097
-4.370
0.000
e1t 1
-0.128 0.398
0.065 0.106
-1.971 3.747
0.055 0.001
e 2 t 1
NOTES: (1) Linear Regressions through the Origin (2) Dependent Variable: 4 et . Here only third parameter in the unrestricted model seems to be not significant. Dropping the third seasonal variable, the same model could be re-estimated. However, as far as restricted F test is concerned, test statistic will not be biased. In fact HEGY test does not have any provision for that sort of model estimates. Hence we go forward to perform the restricted F test for seasonal unit roots. Table 4. The Results of HEGY Seasonality Regressions Residual Degrees Durbin Regressio Sum of of Watson ns Squares Freedom Unrestrict 0.392 43 2.00 ed Restricted 0.569 45 1.90
R Square 0.51 0.29
Restricted and unrestricted residual sum of squares are 0.569 and 0.392, number of restrictions is 2, number of observations is 49, and number of parameters in full model is 4. Hence the calculated value of F statistic is Fcal [( 0.569 0.392 ) / 2] : [0.392 /(49 4)] =10.172
The critical value reported by Hylleberg et al (1990) is 3.08. As the test statistics exceeds the critical value, we reject the null hypothesis. Therefore no seasonal unit roots have been detected. According to the t-statistics given in Table 3 for the non-seasonal and semi-annual unit root parameters, null hypothesis of unit roots should be rejected. However these ordinary t-statistics are not robust. Hence we should apply DF test for non-seasonal unit roots. DF test results are presented in Table 4.
50
Table 5. Dickey-Fuller Unit Root Test for Levels Variablea Coefficient Std. Error t-Statistic Prob. DW AIC et-1 -0.780820 0.140049 0.0000 1.968608 -1.749835 b 5.575330 NOTES: (a) Dependent Variable: Δ4et , (b) DF Test Statistic, MacKinnon (1991) critical values for rejection of hypothesis of a unit root are -2.61, -1.95 and -1.62 for 1, 5 and 10 % significance levels respectively. Sample (adjusted): 1993:2 2005:3 Included observations: 50 after adjusting endpoints Durbin-Watson statistic being close to 2, shows that there is no autocorrelation structure in this auxiliary regression. As the test statistic -5.57 is less than critical value -2.61, we reject the null hypothesis of non-seasonal unit roots. Peron (1989) reports a critical value of -4.24 for λ=0.5, the proportion of the observations occurring prior to the break. This is the point when the maximum difference between the two statistics occurs at 5 percent significance level. Our test statistic exceeds the Peron‟s critical value in absolute terms as well, pointing out again that, there is no unit root. CONCLUSION Testing for unit root hypothesis, which asserts that a series is a non-stationary process, in the case of tourist arrivals has important implications for policy makers. If, for instance, visitor arrivals are characterized by a unit root, then it implies that shocks to visitor arrivals are permanent. However, if visitor arrivals are without a unit root, this implies that shocks to visitor arrivals are temporary. This study provides evidence on the unit root hypothesis for tourist arrivals to Turkey in case of structural change, using a modification of the previous methods. The test results allow us to reject the null of unit root hypothesis, implying that external shocks have a temporary effect on Turkey‟s tourist arrivals. Our results from Table 2, 3 and 4 highlight several important points. First, Turkey‟s time series of tourist arrivals have deterministic properties. This would imply that tourists are not vulnerable to any form of external shocks and make adjustments to their tourism destination whenever news of these shocks is received. As mentioned earlier, the extensive media coverage of Iraq war could have had a serious negative effect on the number of tourist arrivals if the process has been a random walk. It must be noted that among the 105 source countries are European countries together with Middle East make up nearly 80 percent of tourist arrivals into Turkey. Three structural changes within the deterministic components have been detected as well by dummy variables. The earthquake of August 1999 has a significant but temporary negative effect on tourism demand. The relative political stability started from 2003 has given a momentum to the growth in tourism demand for Turkey, by increasing the slope and decreasing the intercept parameters. It must be noted that our analysis uses total tourist arrivals. Future research could consider the effect of shocks to individual nationalities and on particular classification of tourists, for instance business travelers, backpackers, group travelers etc. For instance, using panel data for several important sources of demand for Turkey‟s tourism over a number of years, a panel unit root test may reveal more reliable results. 51
REFERENCES Au, A.K.M, B. Ramasamy and M.C.H. Yeung (2005), “The Effects of the Hong Kong Tourism Industry: An Empirical Evaluation”, Asia pacific Journal of Tourism Research, Vol. 10, Vo.1, March. Bhattacharya, M and P.K. Narayan (2005) “Testing for the random walk hypothesis in the case of visitor arrivals: evidence from Indian tourism”, Applied Economics, 37, 1485–1490. Dickey, D. A. & Fuller, W. A. (1979). “Distribution of the Estimators for Autoregressive Time Series with a Unit Root, Journal of the American Statistical Association, 74, 427–431. Dickey, D. A. & Fuller, W. A. (1981). Likelihood Ratio Statistics for Autoregressive Series with a Unit Root, Econometrica, 49, 1057–1022.
Time
Enders, W. (2004). Applied Econometrics Time Series, Second Edition, John Wiley, New York. Enders, W., Sandler, T. and Parise, G. F. (1992). An Econometric Analysis of the of Terrorism on Tourism. Kyklos, 45(4), 531–554.
Impact
Gujarati, D.N. (1995), Basic Econometrics, Third Edition, McGraw-Hill Inc., New York. Halıcıoğlu, F. (2004) “An ARDL Model of International Tourist Flows to Turkey”. Global Business and Economics Review 2004 Anthology, pp.614-624. Hoffmann, R., Lee, C. G. and Ramasamy, B (2003). “Shocks to Malaysia‟s exports: Temporary or Permanent?” Journal of Asia-Pacific Business, 5(1), 19–32. Hylleberg, S.R., R. Engle, C. Granger, and B. Yoo (1990), Seasonal Integration and Cointegration”, Journal of Econometrics 44, 215-38. Ljung, G. and G. Box (1979) “On a Measure of Lack of Fit in Time Series Models,” Biometrika, 66, 265–270. MacKinnon, J.G. (1991) “Critical Values for Cointegration Tests,” Chapter 13 in Longrun Economic Relationships: Readings in Cointegration, edited by R.F.Engle and C.W.J. Granger, Oxford University Press. Peron, P. (1989), “The Great Crash, the Oil Price Shock, and the Unit Root ypothesis”, Econometrica 57, 1361-1401. Raymond, Y.C. T. (2001),” Estimating the impact of economic factors on tourism: evidence from Hong Kong” Tourism Economics, 2001, 7 (3), 277–293 State Institute of Statistics (2005), Monthly Data for Tourist Arrivals of Turkey from to 2005:10.
1993:1
52
Segmentation of Cultural Festival Visitors by Motivation Karen Thompson Scottish Hotel School University of Strathclyde Glasgow GL4 0LG Tel: 0141 548 4801 Fax: 0141 552 2870
[email protected] Peter Schofield Management and Management Sciences Research Institute University of Salford Salford M6 6PU, U.K. Tel: (0161) 295 4579 Fax: (0161) 295 2020
[email protected]
53
Segmentation of Cultural Festival Visitors by Motivation Karen Thompson and Peter Schofield ABSTRACT There has been a significant increase in cultural festivals and events over the last decade following a growing recognition of their value for enhancing a destination’s image and appeal, improving recreational opportunities, contributing to local and regional economies and enhancing local pride and culture (Long and Perdue, 1990). The analysis of visitor motivation for attending festivals, as a basis for segmentation, is an important prerequisite for targeting markets, planning festival programmes and positioning these products (Crompton and McKay, 1997). This study identified five motivation dimensions for visitors attending the 2005 Naadam cultural festival in Mongolia, using factor analysis. A cluster analysis on the five factors produced five stable motivation segments: multipurpose seekers; indifferent; culture and sport seekers; togetherness, socialisation and sports seekers; and socialisation and local event seekers. Significant associations between motivation clusters and visitor age and type were identified, although there was no significant interaction between the clusters and visitor type with respect to overall satisfaction. The results are generally consistent with the outcomes from previous research on festival and event motivation in Europe and North America. This suggests that these core themes are universal and should underpin production and marketing models used to plan and manage all festivals and events, although the unique combinations of motivation dimensions identified in this study also suggests that further research is needed to develop our understanding of variable interaction. Keywords: motivation, segmentation, factor-cluster analysis, Naadam festival INTRODUCTION
The success of a festival or event is heavily dependent on the implementation of a strategic marketing plan; an understanding of the relationship between a destination event and its visitors and the identification of target markets are critical factors in this process. Within this framework, market segmentation is extensively used to understand the characteristics of visitors and identify distinct groups that might require separate experiences and marketing service mixes. However, the use of inappropriate segmentation techniques can result in destinations either missing strategic marketing opportunities or failing to reap the rewards of a tactical marketing campaign (Bloom, 2004). The analysis of visitor motivations for attending festivals as a basis for segmentation is an important prerequisite for targeting visitor markets, planning festival programmes and their positioning (Crompton and McKay, 1997). Moreover, the segmentation of markets based on visitor motivations facilitates both the identification of the strengths and opportunities of each market and their satisfaction (Lee and Lee, 2001). Dimensions of Festival and Event Motivation
54
Since Uysal et al (1993) conducted their groundbreaking study of the dimensions of event motivation, increasing attention has been paid in the literature to testing and refining the underlying motivations which influence attendance at festivals and events. Lee et al. (2004) conducted an extensive literature review in this area, and it is considered more useful here to present an updated summary of their review (presented below as Table 1) than to repeat their efforts. As Table 1 illustrates, a number of salient dimensions of motivation have been identified by a range of studies undertaken in a variety of festival settings, across a number of cultural groups and often using different scale items to measure the concept of festival motivation. Some of the dimensions appear to be specific to the event, its setting or the nature of the attendees. Other dimensions of motivation occur repeatedly across the various studies and will therefore be explored in more detail below. The desire to „escape‟ and „recover equilibrium‟ has been identified by a number of reports on festival motivation. Uysal et al. (1993), in one of the first studies in this area, found escape to be one of five factors, extracted from 24 motivational items, explaining motivation to attend the South Carolina Corn Festival in the USA. No significant between groups differences were found on this factor across the demographic variables within the sample. Later studies by Mohr et al. (2003), Scott (1996), Schneider and Backman (1996), Lee (200) and Lee et al. (2004) identified a similar escape factor among festival attendees in the USA, Jordan and South Korea. Of these studies, Scott (1996) found statistically significant differences on this and other factor scores between the three different festival populations which he surveyed, indicative of differences in the overall composition of motivation at different festivals. Moreover Lee (2000) found evidence of differences in motivation between Eastern (Korean and Japanese) and Western (American and European) national groupings. The novelty of the event was found to be one of the five dimensions in the original exploratory study of event motivation by Uysal et al. (1993) and, in common with the escape factor, exhibits similarities with one of Crompton‟s (1979) socio-psychological motives. The novelty pull of the event has since been identified as an underlying factor by all but three of the subsequent studies summarised in Table 1, namely Backman et al. (1995), Scott (1996) and Schneider and Backman (1996). Nonetheless, similar factors arguably exist within the above factor solutions. For example, Scott (1996) identified a curiosity factor on which he found statistically significant differences for first time and repeat visitors. Crompton (1979) indicated that curiosity appeared as a synonym for novelty in his research on pleasure travel motivation, citing an associated preference for going to previously unvisited destinations. Indeed, there appears to be some disparity between the novelty and excitement factors reported in later studies with the „site novelty‟ factor reported by Formica and Uysal (1998) expressing first time visitation, as opposed to the desire for adventure and excitement and to satisfy curiosity which contribute to Lee‟s (2000) novelty factor.. Moreover, Table 1 reveals that, whilst the early studies of festival and event visitors identified excitement as one of the salient factors underpinning motivation, the most recent studies (Lee 2000; Lee et al. 2004; Chang, 2005) failed to confirm this factor. Attributes expressing excitement in these latter two studies loaded respectively on the factors novelty and festival participation and learning. The authors do not attempt to account for this difference in the factor structures. The socialization factor appears in all of the studies reviewed in Table 1. Crompton (1979) argues that trip motivation can be people rather than place oriented and this dimension of motivation appears to be particularly important in the case of festivals and events. Socialization refers to the desire and willingness to meet with people from beyond the normal circle of acquaintance and to extend social contacts. Given the nature of festivals as places 55
where a large number of people with a common interest are gathered together, it is not surprising that socialization has repeatedly been shown to be a salient factor in event motivation. It should also be noted that in some studies (Crompton and Mackay, 1997; Lee, 2000) a distinction has been established between the desire to spend time with friends and associates (known-group socialization) and the appeal of meeting new people or observing others (external interaction/socialization). Crompton and Mackay (1997) also reported a gregariousness factor, which they believed to be closely associated to the latter two socialization factors. Furthermore, Formica and Uysal (1996) found statistical evidence, in the case of the Umbria Jazz Festival in Italy, that residents attending the festival were strongly motivated by the socialization factor than non-residents. The importance of being together as a family has emerged from the majority of studies into event motivation as a salient dimension, normally labelled family togetherness. Crompton and Mackay (1997) were surprised not to identify a factor, which represented Crompton‟s (1979) domain of enhancing kinship relations and ultimately concluded that their research instrument should be extended to incorporate this factor. It is notable that the only other study, of those reviewed in Table 1, which failed to identify the family togetherness factor (Chang, 2005) based the survey instrument on Crompton and Mackay (1997). In the other studies family togetherness has been found to represent an important motivational factor although, perhaps unsurprisingly, its importance has been found to differ according to matrimonial status (Uysal et al., 1993; Backman et al., 1995). The influence of exploring new cultures on motivation to attend festivals and events emerges strongly as the factor explaining the highest percentage of the variance in some of the more recent studies (Crompton and Mackay, 1997; Lee, 2000; Lee et al. 2004) and as a lesser factor in others (Formica and Uysal, 1996; 1998; Chang, 2005). The importance of culture in motivation to visit an event is clearly linked to the significance and interest of the culture(s) being celebrated by any individual event. It is therefore not surprising that this factor should emerge unambiguously in studies of motivation to attend the World Cultural Expo (Lee, 2000; Lee et al. 2004). It is hypothesised, and will be argued below, that one of the key attractions of the Naadam Festival is the traditional and unique culture that can be consumed there.
Table 1: Summary of Research on Festival Motivation Researcher
Major Objective
Delineated Factors
Uysal et al. (1993)
Examine dimensions of event motivations
Mohr et al. (1993)
Identify dimensions of event motivations; examine variations of demographic variables; delineated factors and satisfaction by visitor types Examine dimensions of event motivation; analyze variation of
Escape; event novelty; excitement/thrills; socialization; family togetherness Socialization; escape; family togetherness; excitement/uniqueness; event novelty
Backman et al. (1995)
Excitement; external; family; socializing; relaxation
Event Name/ Location Corn Festival/USA
Balloon Festival/USA
Pleasure Travel Market Survey/USA
56
Scott (1996)
Formica and Uysal (1996)
Schneider and Backman (1996)
Crompton and Mackay (1997)
Formica and Uysal (1998)
Lee (2000)
Lee et al. (2004)
delineated factors and activities by demographic variables Determine differences among visitors‟ motivations to attend three festivals; examine motivational differences between first time and repeat visitors Identify dimensions of event motivations; compare differences between regional and extra-regional visitors Examine crosscultural equivalence of motivation scale; test motivation scale on festival celebrating Arab culture identify the set of motives stimulating festival attendance; develop an instrument to measure these motives; assess differences in motivation across different types of events within one festival; assess the extent to which motives validated the escape-seeking dichotomy. Determine principal event motivations; cluster festival visitors based on motivational behaviour Identify major driving motivation factors; examine motivation differences between Caucasian and Asian visitors identify underlying dimensions of motivations; segment festival market on delineated motivation factors; explore differences
Nature appreciation; event excitement; sociability; family togetherness; curiosity; escape
Bug Fest/USA
Excitement/thrills; socialization; entertainment; event novelty; family togetherness Family togetherness and socialization; social/leisure; festival attributes; escape; event excitement
Umbria Jazz Festival/Italy
Cultural exploration; novelty/regression; gregariousness; recover equilibrium; known-group socialization; external interaction/socialization
Fiesta San Antonio/USA
Socialization/entertainment; event attraction/excitement; group togetherness; site novelty; cultural/historical; family togetherness
Spoleto Festival/Italy
Cultural exploration; escape; novelty; event attractions; family togetherness; external group socialization; knowngroup socialization Cultural exploration; family togetherness, novelty; escape (recover equilibrium); event attraction; socialization
1998 World Culture Expo/South Korea
Jerash Festival/Jordan
2000 World Culture Expo/South Korea
57
between domestic and foreign visitors on segments; examine importance of motivation clusters and type of visitors as factors Chang (2005)
Profile festival attendees on basis of motive and demographic characteristics; development, test and apply existing scale of festival motivation
Equilibrium recovery; festival participation and learning; novelty seeking; socialization; cultural exploration
Wu-tai annual aboriginal festival/Taiwan
Adapted from Lee et al. (2004)
In summary, the studies discussed above broadly exhibit similar factor solutions across different events in a range of geographical and cultural settings, suggesting that there are a set of key factors (as identified above) that explain the motivation behind attendance at festivals and events. Moreover, statistical evidence has been found to suggest that some significant motivational differences exist between geographic market segments (Formica and Uysal, 1996; Lee, 2000; Lee et al., 2004) and demographic groups (Uysal et al., 1993; Backman et al. 1995) for the case of some cultural festivals. It should be stressed that the lack of differences found between the underlying dimensions of motivation across different festivals may, in part, be due to flexible interpretation of the factors, with many attributes appearing under different factors in different authors‟ solutions. Indeed it is worth reiterating that Scott‟s (1996) research found differences between motivations sought for different festivals within the same study. In a similarly constructed experiment, however, Crompton and Mackay (1997) found that the same factor structures existed across different events within the same festival, but that some of the factors applied more to some events than others. The research documented below therefore aimed to respond to Lee‟s (2004) call for further exploration of differences between motivation clusters on the basis of demographic and behavioural variables, in light of the increasing internationalisation of events. The Naadam Festival: Attraction and Motivation Eriin Gurvan Naadam (The Festival of the Three Manly Sports) is an annual sporting and cultural festival, more accurately a series of annual festivals, held throughout Mongolia. The event has survived in its traditional form for more than two centuries. The origin of the festival is traced to the skills of war, defense and hunting (Kabzińska-Stawarz, 1991), and traditionally celebrates the prowess of the male, but is also linked to folk-religious rites celebrating and giving thanks for health, wealth and prosperity (Pegg, 2001). During the period of Soviet occupation, Naadam was sponsored by the Mongolian government. The Naadam Festival is the most important sporting event in the Mongolian calendar, celebrating the three traditional game of men, wresting, horseracing and archery. Minor festivals are held in every province and county, with the largest event held in the capital, Ulaanbaatar on 11 th and 12th July. The festival attracts a large audience across these two days to the National Stadium, on the 58
outskirts of the city. Since the date of the festival coincides with the anniversary of the foundation of the state of Mongolia, the event takes place during an annual holiday. The calibre of the sporting events, and in particular the wrestling, attracts a sizeable audience to observe the best competitors from all over the country. There is an enormous amount of prestige attached to winning the sports competitions; the champion wrestler is regarded as a national hero. However, the traditions underlying these sports are arguably equally important in attracting visitors to Naadam. The sportsmen and women compete in traditional costume and there is an air of theatre surrounding the sporting events. Wrestlers, for example, are escorted by heralds who sing of their sporting talents before each bout takes place. Similarly, there are important traditions attached to the horseracing events. At the finishing line of these races across the Steppe, spectators crowd close to the competitors to anoint themselves with the dust and sweat of the winning horses. Traditional victory songs are sung to the horses and libation rituals are undertaken. The opportunity to observe and participate in these historic, indigenous traditions clearly holds appeal for tourists from overseas, but is also argued to be a valuable way of reaffirming Mongolian identity and culture for the native audience, some of whom travel great distances to attend. A further strong motivation for Mongolians in visiting the Naadam festival is the opportunity of spending time with family and friends. It is worth noting that, prior to collectivisation, Naadam celebrations presented one of the few opportunities for nomadic herders to assemble together, renewing old and making new acquaintances. In present day Mongolia, almost one third of the population lives in the capital city and there are a large number of Mongolian nationals living outside Mongolia. The Naadam holiday is a traditional time for families to meet. In identifying motives for attending Naadam, it is useful to stress that the Tourist experience of Naadam may be a very different one to that of Mongolian people. In the first instance, a considerable percentage of the international visitors to the Naadam Festival attend as part of an organised package tour. Because these tours have a tightly scheduled itinerary, tourists typically attend the opening ceremony on the first day, staying only to watch the first half hour or first round of the wrestling. They are then taken for lunch by the tour operator and, in the afternoon, may visit the horseracing, which is about thirty kilometres outside the city. Independent travellers and local people, on the other hand, are more likely to spend a full day at the stadium, watching the various sports competitions that are taking place as well as eating, drinking, shopping and socialising at stalls which are located outside the stadium. It was hypothesised that the key domains of motivation for attending the Naadam Festival would broadly encompass the key factors identified in the literature review above. Cultural exploration applies both in the sense of international visitors wishing to observe Mongolian culture and local people wishing to learn more about their own culture. Socialization is thought to be an important motivator for the independent traveller who wishes to meet and enjoy the event with like-minded people, and indeed the package tourist may be motivated by the experience of visiting the festival in a group. The novelty and escape factors could be argued to apply not only to the Naadam Festival but, for international visitors, also to Mongolia as a destination. In addition, though, with regard to the characteristics of the event, it is useful to embrace the probable importance of the sports events in motivation to attend Naadam. There is a growing body of literature investigating underlying motives for attending sports festivals. In particular investigations by Kim and Chalip (2004) shed some light on motivations for attending sports events which are associated with fan interest (identifying with a team or particular competitors) and aesthetic appreciation of the sport. This motivation factor is hypothesised to be of greater importance for Mongolians attending Naadam than for 59
international visitors who have little understanding of the sporting events included in the festival. Within this context, there are four objectives. 1. Identify the underlying dimensions of motivation for visitors attending the 2005 Naadam Festival 2. Segment the Naadam festival market on the basis of the motivation factors 3. Assess the differences between domestic and international visitors with respect to the motivation clusters. 4. Examine the influence of motivation clusters and type of visitors (domestic and international) on levels of overall satisfaction. METHODOLOGY Instrumentation A three page self-complete questionnaire was designed to survey visitors to the two day Naadam Festival in Ulaanbaatar, Mongolia in July 2005. The instrument contained a scale of 27 items measuring motivation for attending the Naadam Festival this was generated following a review of the literature on cultural festival and event motivation and motivation for attending sporting events. The 27 items ultimately selected were hypothesised to be the most appropriate for the Naadam festival, given its dual role as a cultural and sporting festival. Subjects were asked to rate their level of agreement/ disagreement with each statement relating to their visit motivation presented on 7-point Likert-type scales ranging from „Very Strongly Disagree‟ (1) to „Very Strongly Agree‟ (7) with a no response option; each option was clearly labelled and numbered (Orams and Page, 2000). In addition to the motivation statements, other sections of the questionnaire asked respondents about the characteristics of their stay, their levels of satisfaction with certain aspects of the Naadam Festival, and their overall satisfaction with the festival and likelihood of revisiting. The final section of the instrument collected socio-demographic information on respondents. A test of face validity was conducted by asking a number of local experts to comment on the suitability of the items. Some of these experts were tour operators, others were connected with the organisation of the festival. As a result of this exercise, some small changes were made, in particular to the items relating to the sporting events. The questionnaire was then translated into Mongolian by a native speaker. As a further test of reliability, a backtranslation was conducted by a different native speaker, who had not had access to the original English language questionnaire. Some slight differences necessarily existed between the Mongolian language questionnaire and the questionnaire for international visitors with regard to trip characteristics. Sampling Design A survey was undertaken in July 2005 at a number of sites across the festival including inside the main Naadam stadium, the archery stadium, the horse racing venue and outside the stadium in the area of the stalls and other entertainments. International visitors were targeted by researchers with English as their first language and with a number of other language skills. A team of Mongolian speaking researchers was recruited from the tourism department of Orkhon University in Ulaanbaatar and trained in the necessary interview skills through an 60
interpreter. A convenience sample of visitors produced 539 useable questionnaires; 182 (33.8%) were completed by international visitors and 357 (66.2%) by Mongolian nationals. The sample consists of 45% females / 55% males and the breakdown of age categories is as follows: 65+ (1.7%), 55-64 (6.7%), 45-54 (15.8%), 35-44 (21.3%), 25-34 (30.9%) and 18-24 (23.6%). Data Analysis The data were analysed using SPSS Version 13.0. A factor analysis, using principal components as the method of extraction, with oblique rotation was conducted on the subjects‟ ratings on the motivation scale items to delineate underlying dimensions. All factors with eigenvalues greater than or equal to 1.0 were retained, because they were considered significant (Kaiser 1974). The determinant of the correlation matrix, Cronbach's alpha coefficient, a Kaiser-Meyer-Olkin (KMO) test of sampling adequacy and Bartlett's test of sphericity confirmed the factorability of the correlation matrix. A reliability coefficient (Cronbach‟s alpha) was computed for each factor to estimate the reliability of each scale; all factors with a reliability coefficient above 6.0 were considered to be acceptable for an exploratory study (Churchill, 1979). Hierarchical and non-hierarchical cluster analyses were used to identify meaningful motivation segments from the factors. Independent samples ttests and ANOVA were employed to determine the characteristics of cluster profiles and validate the outcomes. RESULTS The Dimensions of Visitor Motivation Subjects‟ ratings on the motivation attributes were subjected to factor analysis to identify underlying dimensions. Table 2 presents the results; one attribute, „the Nadaam festival was part of an organised tour‟ was excluded because of its low correlation with other attributes in the construct. The 26 remaining Nadaam festival items were analysed using principal components as the method of extraction with an oblique rotation because the extracted factors were correlated (Pedhazur and Schmelkin, 1991). The minimum coefficient for factor items to be included in the final scale was .40, as recommended by Stevens (1992) for the sample size. The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy (.90) was „meritorious‟ (Kaiser 1974) and the Bartlett‟s Test of Sphericity reached statistical significance (X2 (325) = 4305.39; p1.0) accounted for 55.27% of the overall variance before rotation (Table 1). Factor 1 (.82 alpha) accounts for 29.83% of the variance and loads mainly on the motivational attributes describing cultural exploration and has been labelled as such. This demonstrates the relative importance of the cultural motivation. Factor 2 (.79 alpha) accounts for 11.33% of the variance in the data and loads on variables that seem to describe togetherness (with family and friends). Factor 3 (.64 alpha) accounts for 5.53% of the variance and loads on the attributes describing socialisation. Factor four (.72 alpha) accounts for 4.59% of the variance and has been labelled sports attraction because of its variable loadings. Factor 5 (.77 alpha) accounts for 3.99% of the variance and appears to describe local special events. The motivation dimensions that emerged from the factor analysis are generally consistent with the results from earlier festival and event motivation studies (Uysal et al, 1993; Mohr et 61
al, 1993; Backman et al, 1995; Schneider and Backman, 1996; Scott et al, 1996; Formica and Uysal, 1996, 1998; Crompton and Mackay, 1997; Lee, 2000; Lee et al, 2004; Chang, 2005). This supports the notion of universality in the range and grouping of event motives (Lee et al, 2004).
Table 2: Factor Analysis of Naadam Cultural Festival Visitor Motivation Visitor Motivation Variables Factor 1: Cultural Exploration I want to experience Mongolian culture I wish to learn more about Mongolian culture I would like my family to learn more about Mongolian culture I enjoy experiencing culture in its unique, historical setting I admire the talents of the sports competitors The Naadam festival is exciting
Factor 1
Factor 2
Factor 3
Factor 4
Factor 5
Communality
.783
.647
.715
.605
.671
.689
.617
.534
.472 .442
.568 .539
Factor 2: Togetherness I thought it would be fun to attend the festival with friends/group I want to spend leisure time with my family I am supporting certain competitors I know people who are competing in the sporting events I thought my family would enjoy the Naadam festival I like to be with people who enjoy the same things I do
.728
.582
.714 .618
.582 .641
.608
.654
.550
.564
.533
.422
Factor 3: Socialisation I am an adventure seeker I like to meet people from all over the world I like being with people who are enjoying themselves Factor 4: Sports Attraction I am a keen sports fan I enjoy the festival atmosphere I enjoy sports events I was curious about the Naadam festival Factor 5: Local Special Events I enjoy special events I thought the Naadam festival sounded like fun I like to experience local customs and cultures I am interested in local events I enjoy cultural experiences I enjoy the music, ritual and dance which accompanies the sporting events Eigenvalue 7.756 Variance (%) 29.832 Cumulative Variance (%) 29.832 Cronbach’s Alpha .82 Number of Items (Total = 14) 6 Note: only loadings above .4 are displayed.
.746
.596
.730
.542
.481
.499
.682 .535 .518 .416
2.945 11.327 41.159 .79 6
1.438 5.531 46.690 .64 3
1.193 4.590 51.280 .72 4
.600 .605 .623 .454
.735
.569
.684
.454
.662 .525 .524
.530 .581 .514
.420
.356
1.036 3.986 55.266 .77 6
62
Motivation Clusters To develop further our understanding of the motivation dimensions, a cluster analysis was performed on the five factors. Initially, a hierarchical cluster analysis was used to identify the number of clusters required for a K-means non-hierarchical algorithm (Hair, Anderson, Tatham and Black, 1998). Five distinct clusters emerged from this analysis. Initial cluster centres were selected by SPSS Version 13.0 and iterated until the Euclidean distance between centroids changed less than 2% to reduce the bias of designating initial cluster seeds and produce stable clusters when the criterion had been met. The „average linkage between groups‟ (unweighted pair-group using arithmetic means) method of clustering was used. Table 3 shows the results from a one-way ANOVA test to show that all five factors contribute to differentiating the five motivation clusters (p
