Connected and Disconnected? On the Impact of ... - Springer Link

7 downloads 0 Views 90KB Size Report
Swiss Federal Statistical Office (SFSO), Espace de l'Europe 10, 2010 ... in scientific and statistical circles we proceed first by taking a closer look at Putnam's use.
Computational & Mathematical Organization Theory, 10, 227–241, 2004 c 2004 Kluwer Academic Publishers. Manufactured in The Netherlands. 

Connected and Disconnected? On the Impact of Internet Use on Social Connectedness ¨ VOLKER G. TAUBE Swiss Federal Statistical Office (SFSO), Espace de l’Europe 10, 2010 Neuchˆatel, Switzerland email: [email protected]

Abstract Based on Durkheim’s idea that social differentiation in modern societies leads to division of labour and to increasing alienation between individuals some authors argue that the use of new technologies like the Internet will promote social isolation. Such a tendency towards a decline in Social Capital has been reported for the U.S. while citing decreasing numbers of membership in diverse organisations. The paper investigates similar tendencies for Switzerland and discusses the appropriateness of respective measures of social capital. Keywords: internet use, social capital measurement, social integration

Introduction “One major controversy generated from macro- versus relational-level perspectives is whether social capital is a collective or an individual good [...]. Most scholars agree that it is both collective and individual; that is, institutionalised social relations with embedded resources are expected to benefit both the collective and the individuals in the collective” (Nan Lin, 2001). Despite the backlashes on economic markets the dispersion of new information and communication technologies (ICT) is undoubtedly of central importance for future economies as they hold the promise to raise overall productivity and wealth. This fact is also highlighted through the widespread endeavours enforced by national and international politics in order to assure as well optimal diffusion of technologies—e.g. through the liberalisation of telecommunication markets—as well as adequate preparation of the actual and future labour forces with regard to ICT competence.1 Some scholars argue that these developments will effect social live profoundly and that the amount of actual social relations of persons will diminish with their increasing use of information technologies such as the Internet (Kraut et al., 1998; Nie and Erbring, 2000). In fact, the argument seems to be related to Putnam’s (1993, 1995, 1996) idea of a decline of social capital for which he tries to enumerate diverse empirical evidences. Putnam sees in the diminishing numbers of members in a wide array of institutions of public life “a pattern of civic disengagement”. Since the civic society is—in line with the ideas of Alexis de Tocqueville (1864)—understood as the basic ingredient that assures the common normative framework which makes a democracy work, the very existence of a tendency towards eroding social capital stocks might lead to serious existential problems for western democracies.

¨ TAUBE

228

Besides “community engagement” Putnam refers to “trust” as another, yet related indicator for social capital in the U.S. American society. The fact that both factors—community engagement and social trust—seem to decline over certain observational periods within the United States are thus interpreted as empirical evidence for the assumed social capital decay. With regard to the ongoing discussion on measurement procedures of social capital in scientific and statistical circles we proceed first by taking a closer look at Putnam’s use of the term “social capital”. As we will argue then, the fact that Putnam’s perspective on social capital is basically linked to the notion of collective and public goods impinges on the explanatory power of the concept since it remains an open question how such goods come into being in the first place. For such explanations one would have to rely on micro sociological aspects of social capital that might in the aggregate account for institutional changes over time or even the rise of new institutions. Since formal relations at work or loose contacts to acquaintances such as neighbours for instance are distinct in content and intensity from relations to family members or close friends it is necessary to account for different types of social contacts. Obviously, more distant social relations serve different purposes than do closer ones—the notions of “bridging” vs. “bonding” refer to such differences that will be examined in greater detail in the second part of this analysis. After having laid out some basic arguments on aspects of social capital measurement we subsequently turn to the empirical part of the analysis which proceeds in two steps: First, (in line with Putnam’s approach) we are going to check from a “collective good perspective” for the status quo of membership turnouts in some institutions like political parties, environmental groups, charitable organisations and sport or leisure groups in Switzerland. This will allow us to get an impression in how far Putnam’s “measures” on the social capital stocks of certain institutions in the United States correspond to the situation in Switzerland. Because the variables used for this kind of analysis are as we argue but vague indices on social capital stocks we turn in a second step to a more ego related network perspective on social capital while examining individual relations instead of institutional affiliations. By doing so we intend to gain a more thorough picture on the question if alleged fluctuations in social capital caused by Internet use exist and if they can be traced to certain sorts of social capital. The results from both analyses will then finally be compared in order to further stimulate discussion on measurement issues on social capital. 1.

Is the Internet Privatising Our Lives?

In his article “The Strange Disappearance of Civic America” Robert Putnam (1996) presents his argument on a decline of social capital by referring to different cross sectional data sources on developments in the US. Among other things he points to evidence stemming from surveys of average Americans in 1965, 1975 and 1985 that reported on “time-budget” studies and that document a drop in time spent on informal socialising and visiting by nearly one quarter and time devoted to clubs and organisations that even sank to almost the half. In a similar manner participation in and membership records of diverse organisations (church

ON THE IMPACT OF INTERNET USE ON SOCIAL CONNECTEDNESS

229

attendance, League of Women Voters, the Red Cross, labour unions etc.) as well as measures of political participation, including attending a rally or speech show decreasing figures. As one of the usual suspects that are responsible for the changes observed by Putnam (1996, p. 3) the author identifies besides other factors “television, the electronic revolution, and other technological changes”. In fact, even if he discusses different other possible sources such as education, changes in family structure, the rise of the welfare state etc., it is especially television and new technologies in general that Putnam identifies as the major culprit for the increasing disconnectedness of American society: “In other words, each hour spent viewing television is associated with less social trust and less group membership [...]. An increase in television viewing of the magnitude that the U.S. has experienced in the last four decades might directly account for as much as one-quarter to one-half of the total drop in social capital [...]” (1996, p. 14–15). With regard to the fact that television has been in existence in the U.S. for over 80 years now, and taking into account that already since 1960, 90% of American households owned a television (“probably the fastest diffusion of a major technological innovation ever recorded”) one might wonder if Putnam’s conclusion “[...] that America still outranks many other countries in the degree of our community involvement and social trust” (1996, p. 2) is in line with his general argument: When almost every household in the U.S. has a television for about 40 years now—probably a dispersion much denser and already existing for a longer time than in other countries—, and TV consumption is the driving force behind the decline in social capital, how is it than that America is still a leading country when it comes to community involvement? Be that as it may, of greater importance for the enterprise at hand is the question if a similar tendency as the one described by Putnam for TV consumption might equally be observed with regard to the use of new media such as the Internet? Authors like Nie and Erbring (2000, p. 6) from the Stanford Institute for the Quantitative Study of Society point to empirical evidence for such a tendency stemming from the data of a panel study in the United States2 : “The more time people spend using the Internet the more they lose contact with their social environment. This effect is noticeable even with just 1–5 Internet hours/week (8 percent), and it rises substantially for those spending more than 10 hours/week, of whom up to 15 percent report a decrease in social activities. Even more striking is the fact that Internet users spend much less time of talking on the phone to friends and family: the percentage reporting a decrease exceeds 25 percent—although it is unclear to what extent this represents a shift to e-mail even in communicating with friends and family, or a technical bottleneck due to a single phone line pre-empted by Internet use”. Furthermore, time formerly spend on watching television is reported to be increasingly reallocated towards Internet use. Nie and Erbring present results showing that of those people using the Internet less than 1 hour/week 27 percent claim to have less time for watching TV whereas 65 percent of the people with more than 10 hours/week on the Internet report to have less time for watching television. Internet use seems to happen predominantly at home—about 64 percent respondents from the Stanford study deploy Internet mostly or exclusively at home whereas only 14 percent go online mostly or only outside their homes.3 Nie and Erbring (2000, p. 19) bring the connection between Putnam’s argument on TV

230

¨ TAUBE

consumption and use of Internet to the point stating that time for Internet use comes mainly out of time viewing television. Even though the key conclusion of this study are intriguing a certain scepticism seems to be justified: Internet use is reported to lead to a decrease of social contacts, based on self reported evidence of about 15 percent of heavy Internet users (more than 10 hours a week online) while about 85 percent of this group don’t seem to acknowledge such a trend—by the way, this user group consists of 14 percent of the total Internet user population. Opposed to Putnam’s conclusions on the connection between use of television and formal formation, the authors of the Stanford study comment on strong evidence for a correlation between higher education and Internet use (for the similar trend in Switzerland, see Harabi et al., 2000; T¨aube and Joye, 2002). The theoretical reasoning as well as the empirical facts cited by Putnam or by Nie and Erbring evoke the impression, that the related notion of social capital refers to well known aspects of social integration. Obviously standing in the modernisation theory tradition dating back to Durkheim (i.e. 1930) the common issue of these scenarios is that they assume in general a certain “top down” direction of causal flows: the cause is always located at the macro level and determines results at the micro level. Besides the fact that this view leaves open questions about the emergence of phenomena on the macro level,4 it is crucial for the manner in which the productive aspects of social relations are understood. With regard to social capital this leads often to an exclusive focus on the public and collective goods aspects of social relations while leaving the aspect of social capital as an individual resource aside. Although Putnam (1996, p. 1) states that “by ‘social capital’, I mean features of social life—networks, norms, and trust—that enable participants to act together more effectively to pursue shared objectives”5 he doesn’t provide a conceptualisation of social capital that accounts for its genuine network character. Indeed, the central aspect of social networks are individual relations and interactions and a measure of social capital has to take this into account6 —otherwise the notion is but a synonym for social cohesion and its analytical signification becomes blurred.7 This general problem of a misguided understanding of social capital as an alternative term for collective or public goods and norms becomes also evident in a different line of argument from Bowls and Gintis (1998, 2001). Using the term social capital in the same vein as Putnam—but this time pointing to evidence for the increasing chances for social cooperation stimulated by new technologies in the upcoming “cyber economy”—the authors conclude that the term social capital is but a fashion that they would like to substitute by the more precise term community (2001, p. 27). But let’s get back to Putnam’s argument and the empirical evidence reported by Nie and Erbring on the drop in the quantity and quality of individual social relations due to Internet use. As already mentioned, the questionnaire for the Stanford study entails with regard to information about a person’s social network the question “Has using the Internet changed the amount of time you spend with your family/friends?”. Obviously, this question points at the characteristic “time spend” with somebody and is not informative about the actual number of relations that a person has at a given point in time. Accordingly, the authors are more careful in their interpretation of the related data than the heading of the respective paragraph of their report (“social isolation is up”) might suggest:

ON THE IMPACT OF INTERNET USE ON SOCIAL CONNECTEDNESS

231

“The Internet could be the ultimate isolating technology that further reduces our participation in communities even more than television did before it” (Nie and Erbring, 2000, p. 19). From what has been said so far, certain issues have been marked that will be taken up in the empirical investigation later on. One of the central conclusions is that if we want to say something about the effect of the Internet on a person’s social contacts we need to stick to appropriate measures for the independent variable—e.g. data on actual aspects of individual relations—instead of using variables that are at best proxies for it. Since social relations should be differentiated due to their content we need to set out some central issues on measuring them before we are getting involved with the analysis of data on Internet use and social contacts in Switzerland. 2.

Distinctive Relational Aspects: Expressive and Instrumental Contacts

Even in our day to day language we differentiate between close social contacts like “family” and “friends” on one side and “acquaintances” or “colleagues” on the other side. Therefore, it seems to be even more appropriate to take such differences into account if we are looking with a scientific interest at social institutions. Are social contacts between friends that go bowling together the same as contacts between people joining the same political party? And does a change of relational quantity in one field means the same as such a change in the other field? The notion of social capital relates to the productive aspects of social relations. As empirical findings repeatedly suggest, there needs to be a distinction made with regard to social spheres in which social capital is productive. In this sense Fiedler (1967) differentiated between what he called “task oriented groups” and “social groups”. As opposed to task oriented groups Fiedler (1967, p. 16) describes social groups that “exist to promote the psychological well-being, enjoyment, or adjustment of the individuals who are members of the group”. The relational characteristics of such groups might against the background of social exchange theory (Homans, 1974) be understood as more or less intense due to their respective degree of reciprocity with the former case relating to expressive relations while less intense relations account often for instrumental contacts. In the context of Social Network Analysis the notions of closeness or betweenness account for similar characteristics. Burt (2001), using the term “structural hole” for to account of betweenness aspects of social networks, describes the distinct signification of the related sorts of social capital: “The closure argument is that social capital is created by a network of strongly interconnected elements. The structural hole argument is that social capital is created by a network in which people can broker connections between otherwise disconnected segments.” Therefore, as has been argued in the previous chapter, social capital needs to be distinguished with regard to its productive aspects: Whereas expressive or socio-emotional relations (support capital) mark an individual’s “close” environment such as the family or good friends instrumental relations (leverage capital) accounts for more distant social contacts like neighbours or colleges.8 Instead of immediately rejecting Putnam’s argument on the decline of social capital altogether, we aim at taking a more detailed look at it by looking for evidence with regard

¨ TAUBE

232

to such different relations. As we tried to show elsewhere (T¨aube and Joye, 2002) the closure aspect of social capital might well give some clues on the general orientation of people towards information and communication technologies in the sense that especially close social relations are decisive for adopting preferences for or against new technologies, whereas there seem to be reasons for the assumption that once a positive preference is established, people make increasingly use of more distant (instrumental) relations in order to tackle practical problems with ICT. The fact that Nie and Erbring in their Stanford study report on evidence for the assumption of a decline in social capital as it is constituted in the close environment (family and friends) underpins the necessity of taking the different qualities in social relations into account.9 In this manner the authors specify Putnam’s idea of a loss of social capital, stating that heavy Internet use leads to a decline in what we call here support capital. Because close social relations might be understood as fundamental for the cognitive development of an individual a general loss of such social contacts would indeed be an alarming sign for the future of civil society. In this vein also Becker (1993, p. 21) argues on the importance of support capital for the accumulation of human capital: “No discussion of human capital can omit the influence of families on the knowledge, skills, values, and habits of their children. [...] Large differences among young children grow over time with age and schooling because children learn more easily when they are better prepared”. Unfortunately, Putnam’s argumentation as well as the empirical evidence he presents don’t seem to make a difference between different sorts of social capital. Since we are interested in the question if a decline of social capital effects rather one sort of social capital than the other we first need to identify the two qualities empirically i.e. by referring to different fields of social contacts such as circles of close friends and neighbours and then testing for changes in the quantities of a person’s respective relations that might be linked to the person’s Internet use. With regard to the content of a relation the term “tie strength” is often used. Following Granovetter’s (1973) proposition that the strength of a tie is—besides other characteristics—also represented through the reciprocal services and the amount of time that people spend with one another it seems obvious that we interpret close friendship relations as strong ties or support capital, whereas social contacts to neighbours are interpreted as being comparatively weaker in the sense of leverage capital. As results of a hierarchical cluster analysis (single solution with 2 clusters) for the here used data of the Swiss Household Panel (SHP) show, neighbourhood contacts and contacts to close friends can be classified along their strength (quantity of contacts weighted by the respective frequency of the contact). In the second part of the following empirical analysis these two groups—friends and neighbours—will serve as the reference for testing on the effect of Internet use on support and leverage capital stocks in a cross sectional analysis. 3.

Social Connectedness in Switzerland: Some Empirical Evidences

Instead of basing our study for Switzerland on a selective sample of certain user groups as is quite often the case10 we are going to test some of the arguments we presented so far through an examination of the data from the Swiss Household Panel (SHP/“Living in Switzerland”).

233

ON THE IMPACT OF INTERNET USE ON SOCIAL CONNECTEDNESS

The target population of the SHP consists of all private households throughout the national territory. The national stratified random sample is representative of each of the seven major statistical regions and of Switzerland as a whole; it was drawn randomly from the electronic phone register (TERCO) by the section specialised in methods/sampling at the Swiss Federal Statistical Office (METH). The data set used in the first part of our analysis consisted of all three waves covering the years 1999, 2000 and 2001.11 In the second part the analysis turns to cross sectional analysis of the first wave (1999) and the third wave (2001) of the SHP in order to check for changes in the size of social networks of Internet users compared to non users. Since we are interested in differentiated information for denser and looser networks we will look for evidence in the network of close friendship contacts as well as in the network of neighbourhood contacts. By examining membership fluctuations of certain associations in Switzerland we intend to provide comparable results to those provided by Putnam.12 But instead of solely reporting on membership of Internet users in certain organisations we aim at a more differentiated picture by looking in addition to the membership status—that is, if a person is an active or passive member of an association which might give us a more fine grained impression on the member’s participation. We present results on numbers of memberships for Internet users in four sorts of associations that stand for diverse interests in Swiss society: Political parties, environmental groups, charitable groups and sporting and leisure associations. Turning first to memberships of Internet users in political parties in Switzerland (see Table 1) we see by and large a fairly stable situation: The total number of memberships decreased from 11.6% in 1999 (1st wave) to 11.0% in 2000 (2nd wave) and climbed after to 12.3% in 2001 (3rd wave). The number of active members almost stayed the same between 1999 and 2001 (increase by 0.2%) while the number of passive members sank from 1999 to 2000 by 0.8% (from 4.4% to 3.6%) and climbed again from 2000 to 2001 by 1.8% to 5.4%. Table 1.

Active and passive memberships in diverse Associations in Switzerland, 1999–2001.

Association

Membership Membership Membership Membership Membership Membership Total % 1999: 1999: 2000: 2000: 2001: 2001: Membership Active (%) Passive (%) Active (%) Passive (%) Active (%) Passive (%) per Year

Political parties

7.2

4.4

7.4

3.6

6.9

5.4

1999: 11.6% 2000: 11.0% 2001: 12.3%

Environmental groups

3.5

19.1

2.6

21.8

2.7

22.9

1999: 22.6% 2000: 24.4% 2001: 25.6%

Charitable organizations

6.5

18.2

6.6

20.1

5.9

23.3

1999: 24.7% 2000: 26.7% 2001: 29.2%

42.8

7.2

39.4

8.4

41.1

7.5

1999: 50.0% 2000: 47.8% 2001: 48.6%

Sport/Leisure groups

234

¨ TAUBE

Even if the number of Internet users actively participating in Switzerland’s political parties between 1999 and 2001 seem to show a slight tendency in the direction of Putnam’s argument the increase in participation between 1999 and 2000 as well as the increase in overall participation in political parties in Switzerland doesn’t seem to do so. The fact that active membership only went down by 0.5% between 2000 and 2001 while passive membership increased by 1.8% in the same period is quite interesting because it seems to contradict the expectation of a general drop out sequence in which a decrease in active membership is followed by an increase in passive membership which is itself but the last station on a person’s way out of the organisation. A test of proportions showed no significance for the changes between the membership fluctuations in political parties over all three waves. Still it remains unclear if the evidence presented so far for political parties has to be interpreted as a characteristic of support or leverage capital. Since political parties are by definition institutionalised interest groups one may be inclined to conclude that this is a rather weak evidence for a decline in leverage capital. The second row in Table 1 presents results for membership turnouts from Internet users in environmental groups in Switzerland: Overall membership turnouts from Internet users in environmental groups increased by 1.7% from 22.6% in 1999 to 24.4% in 2000 and once more by 1.2% in 2001 to 25.6%. If we only take a look at active participation in Swiss environmental groups we see a decline of 0.9% from 1999 to 2000. Between 2000 and 2001 active membership stayed almost constant (increase of 0.1%). For passive membership we see an increase from 1999 to 2000 by 2.7% (from 19.1% to 21.8%). However, neither for the overall participation rates nor for active or passive membership turnouts there seems to be clear evidence for a decline which is supported by the results from a test of proportions showing no significance in the fluctuations of membership turnout over all three waves. As is the case with political parties the question what kind of social capital is constituted through being member of an environmental group also remains unanswered. Because of the closeness to political institutions one might assume that the relations between members of environmental groups represent also leverage capital. Another sort of institution for which we present numbers is often regarded as being a corner stone of civic engagement in society: charitable institutions. The third row in Table 1 presents respective numbers of memberships over the years 1999–2001. Active membership of Internet users in Charitable Organisations dropped by 0.5% between 1999 and 2000 (from 6.5% to 6.0%) and stayed almost stable between 2000 and 2001 (decrease of 0.1%). With regard to numbers of passive memberships there’s a continuous increase over all three waves showing a total increase between the 1st and the 3rd wave of 5.1% (from 18.2% to 23.3%). So is the case with overall membership turnout which increases between 1999 and 2000 by 1.4% (from 24.7% to 26.1%) and again from 2000 to 2001 by 3.1% (from 26.1% to 29.2%). The fact that overall numbers for membership in charitable organisations of Internet users have clearly increased leaves no room for an explanation that would interpret the increase in passive membership as part of a drop out sequence. Again, a test of proportions showed no significance for membership fluctuations for the time period under consideration. As is the case with all here reviewed kinds of organisations the kind of social capital represented through membership in charitable organisations remains an open question.

ON THE IMPACT OF INTERNET USE ON SOCIAL CONNECTEDNESS

235

After having seen some results on what we thought could be interpreted as being examples for leverage capital we shift our analysis towards inspection of at least one sort of association that seem to account rather for support capital since the size of sports and leisure groups is expected to be smaller with more local references that involve rather close friendship contacts.13 For memberships of Internet users in sports and leisure groups we see an overall decrease in participation rates by 2.2% from the first wave in 1999 (50.0%) to 47.8% in 2000 followed by an increase of 0.8% to 48.6% in 2001 (row four in Table 1). A downward trend is obviously traceable for active membership turnout even if we see between 2000 and 2001 an increase in active memberships. Passive memberships show an increase from 1999 to 2000 of 1.2% (from 7.2% to 8.4%) followed by a decline of 0.9% to 7.5% in 2001. If we assume that for some members a drop out sequence “active membership”, “passive membership” and finally “drop out” exists, the rise of passive membership from wave one to wave two appears not to be a contradictory characteristic to the assumption of a general decline in memberships of sporting/leisure organisations; but since overall membership numbers increased between again between 2000 and 2001 there is no unambiguous sign of diminishing participation rates in such groups in Switzerland due to Internet use. A test of proportions showed no significant changes of membership numbers for the entire period. The analysis presented so far compared in Putnam’s manner numbers of membership turnout of Internet users for certain social organisations in Switzerland. As has become evident the results are far from being conclusive with regard to an alleged tendency of a decline in social capital described by Putnam for the U.S. Having said before that the notion of social capital is basically related to actual relations of persons we conclude that the hitherto presented numbers do not account for aspects of networks of social relations, and, thus, may be regarded as deficient for an appropriate description of social capital. For to test nevertheless on the validity of the argument on a social capital decline related to Internet use, we conduct some analysis on cross sectional data from the 1st and 3rd wave (1999 and 2001) of the SHP. In the first instance a one way analysis of variance has been calculated with a 3-tiered independent variable “Internet use”14 and the dependent variables “social contacts to neighbours” and “social contacts to close friends”.15 The differentiation between contacts to neighbours and contact to close friends allows us to control separately on the effect of Internet use on leverage capital and support capital as two sorts of social capital. Even though the overall model for both dependent variables appeared to be significant for the 1999 data as well as for the 2001 data the Duncan Test for correlation between each tier of the independent variable and the dependent variables never showed to be significant. Before we are going to calculate some linear regressions in order to gain further insights into the relation between an Individual’s Internet use and her social contacts we take a look at the mean values stemming from the ANOVA which might give us an idea of how Internet use and social contacts to neighbours and close friends may be related. Obviously, persons with no Internet use show a higher average volume of social contacts to their neighbours than do people with Internet use. In line with our differentiation between different sorts of social capital this could point to a decline in leverage capital caused by Internet use even if the fact that persons with a higher intensity of Internet use seem to have a higher volume of social contacts to neighbours than do people with a lower intensity rests

¨ TAUBE

236 Table 2. Mean values from one way ANOVA on the connection between Internet use and social contacts to neighbours and friends.a Mean number of contacts to neighbours (weighted) 1999/2001

Mean number of contacts to close friends (weighted) 1999/2001

No Internet use

42/40

45/29

Weak Internet use

34/31

51/35

Heavy Internet use

35/33

71/50

number of contacts = mean number of contacts to neighbours/close friends of contact; no Internet use = 0 minute per week; weak internet user = 1–180 minutes per week; heavy internet user = more than 180 minutes per week.

a Mean

∗ frequency

contradictory in this regard. Taking a look at the second column of Table 2 we see that the situation for social contacts to close friends differs markedly. Persons which claimed not to use the Internet show a lower average volume of social contacts to close friends than do people with medium or intense Internet use. From this point of view social capital as support capital seems to be positively related to the intensity of Internet use. This result contradicts the findings from the Stanford Study of Nie and Erbring stating that Internet use leads a person to neglect her close social relations. But since the Duncan Test showed no significant results we conduct a linear regression analysis in order to check for the validity of these observations. This time the independent variable is introduced as dummy variables.16 Table 3 shows the regression coefficients for the respective analysis. As can be seen from Table 3 almost all coefficients for Internet users in general are significant although the Beta values are rather small. The signs of the Beta values correspond exactly with the results presented in Table 2 for social contacts to neighbours showing higher volumes of social contacts for non users of the Internet compared to Internet users. But as has been stated for the findings than, a higher volume of social contacts to neighbours for intense Internet users compared to medium users puts a question mark on the interpretation of decreasing leverage capital due to Internet use. Also for the regression model on Internet use in general and social contacts to close friends almost all coefficients turn out to be significant. The Beta values are again rather small implying only weak effects of Internet use on volume of contacts to close friends. However, the higher Beta values for more intense Internet use imply a greater positive relation between this characteristic and Internet use. Taking the signs of the Beta values into account the findings confirm the tendency already shown in Table 2 indicating a rise in volume of social contacts to close friends with increasing use of Internet. The results presented in Table 3 suggest the assumption that those persons who use the Internet dispose over larger networks of close friendship contacts than do people without Internet use.17 If this outcome is itself the result of Internet use or if it is rather the case that those people using the Internet are in general more communicative than others remains an open question. For to control for eventually differences with regard to sex Table 3 presents furthermore the regression coefficients for social contacts to neighbours and close friends for male and

ON THE IMPACT OF INTERNET USE ON SOCIAL CONNECTEDNESS Table 3.

237

Regression coefficients for different groups of Internet users and distinct social contacts. Contacts to neighbours

Contacts to close friends

1999 Beta/Sig.

2001 Beta/Sig.

1999 Beta/Sig.

2001 Beta/Sig.

Internet use all 180 minutes Internet use per week

–.033/.010

–.051/.001

.016/.226

.052/.000

More than 180 minutes Internet use per week

–.026/.047

–.023/.113

.059/.000

.102/.000

Internet use men 180 minutes Internet use per week

–.035/.065

.048/.032

–.010 /.603

–.009/.693

More than 180 minutes Internet use per week

–.023/.222

.012/.604

.034/.072

–096/.000

Internet use women 180 minutes Internet use per week

–.034/.051

–.050/.010

.036/.040

.074 /.000

More than 180 minutes Internet use per week

–.039/.028

.000/.985

.037/.000

.074/.000

female Internet users. As can be seen for male Internet users with regard to social contacts to neighbours, almost all regression coefficients are not significant from which we conclude that Internet use has virtually no effect on a male person’s number of such social contacts. As is the case with contact to neighbours Internet use doesn’t seem to have an effect on the quantity of friendship relations of male persons either as we conclude from the respective regression coefficients which almost all turn out to be not significant. Obviously, these results show no substantive clue for an effect of Internet use of male persons with regard to the quantity of social relations. Consequently, we turn to a similar analysis for female Internet users next. As can be seen for contacts to neighbours two from four coefficients for the years 1999 and 2001 cease to be significant even though one does only slightly so. Trying to interpret the results it seems as if female Internet users in general appear to have fewer contacts to neighbours no matter if they use Internet at a lower or higher intensity. Comparing the results for male and female Internet users with regard to their neighbourhood contacts that we interpret as being indicative for leverage capital the findings appear to support the well known thesis that men in general dispose over larger networks of weak relations. If this argument holds we should expect larger volumes of social relations of female Internet users for closer relations compared to male Internet users. From the respective results for contacts to close friends we conclude that this is exactly the case: for lower intensity of Internet use females appear to show a Beta value three times higher than that of male users while the Beta values for female Internet users with a higher intensity of use are more than doubled compared to their male counterparts: From what has been shown in the empirical part of this analysis there appears to be no evidence for a decline in social capital related to Internet use in Switzerland. This holds for the findings based on comparisons of membership status in different organisations in Switzerland as it does for results based on aspects of people’s actual social networks.

¨ TAUBE

238 4.

Summary

One of the central critiques broad forward in the introduction of this investigation referred to the lack of conceptual clarification in the use of the term social capital deployed by authors like Putnam or Nie and Erbring. Because social capital is in this context almost exclusively treated due to its importance for the provision of collective goods it becomes somehow but a metaphor for the problem of how social integration leads to the provision of socially desirable norms such as trust or civic engagement. From this perspective the utility of the notion of capital which refers to aspects of productivity seems questionable and social capital appears to be just another buzzword for what has been known already for a long time as social integration. Another yet related point of critique refers to an appropriate measurement of social capital and its dimensions. If one is to accept that social capital can equally be understood as a resource to the individual the notion of the productivity of social contacts becomes a distinct concept. Although authors like Putnam concede that the notion of social capital is fundamentally linked to networks of social relations the social capital measures he presents—counts of memberships in diverse social organisations—are quite apart from reflecting genuine characteristics of an individual’s social network which would include such information as i.e. quantities of relations or contact frequency for to depict different relational intensities. With regard to the productive aspects of such differing intensities or weak and strong ties we introduced the notions of leverage and support capital as two fundamental dimensions of social capital. Leverage capital in our empirical analysis has been regarded as being represented through weaker relations such as neighbourhood contacts whereas the notion support capital was assigned to close friendship contacts. After having first reviewed some empirical results presented by Putnam as well as by Nie and Erbring for a general decline of social capital in the U.S. and a discussion on the meaning of the differentiation between relational contents we turned to some related analysis for Switzerland based on data from the Swiss Household Panel. For to test on the possibility to differentiate empirically between support and leverage capital a cluster analysis was conducted which showed that contacts to neighbours differed clearly from close friendship contacts over all three waves (1999, 2000 and 2001) based on the number of respective contacts multiplied by their monthly frequency. Because our main goal was to clarify the assumption of eroding social capital stocks due to Internet use we included in the next step of our analysis only those respondents that claimed to have Internet access from their homes and traced membership turnouts in several social organisations in Switzerland for the years 1999 to 2001. After having found no clear evidence for Putnam’s respectively Nie and Erbring’s assertions on the effect of Internet use on social participation rates we calculated a number of statistics based on cross sectional data from the 1st and the 3rd wave (1999 and 2001) of the SHP. The coefficients for the overall population as well as figures calculated separately for men and women showed no substantial effect of Internet use on leverage and support capital. With regard to the directions of those effects found the results seem to contradict a tendency towards eroding social capital stocks for Internet users so that on the basis of the findings presented there is no empirical evidence for the assumption of a general decline of social capital in Switzerland due to the use of Internet.

ON THE IMPACT OF INTERNET USE ON SOCIAL CONNECTEDNESS

239

Unfortunately, up till now the data available on Internet use in Switzerland doesn’t allow for application of social capital measurement methods based on concepts from Social Network Analysis—what one needs in order to gain more thorough insights into the dynamics of social affiliations are data about ego centred relational aspects of persons. Therefore, the inclusion of such instruments as sociometric name generators in standard social surveys would be a desirable first step towards reliable measurement of social capital stocks.

Acknowledgments I would like to thank Wolfgang Sodeur and Giampiero Beroggi for comments on an earlier draft of this paper. Comments from Maja Huber and Harald Katzmair are also gratefully acknowledged. This study has been realised using the data collected by the Swiss Household-Panel, a project financed by the Swiss National Science Foundation Program, SPP, “Switzerland Towards the Future” (Grant no. 5004-53205).

Notes 1. The European Community stressed repeatedly the importance of these targets in the eEurope Action Plans 2000 and 2005. 2. The data for the study were collected in December 1999, from a national random sample of 4113 individuals in 2689 panel households. Questionnaires were completed independently by each member of a panel household using their television and their WebTV controls to answer the questions displayed on the screen. The results on Internet use presented in this study are based only on the responses of the participants who had Internet access (at home or elsewhere) prior to and independent of the WebTV access (see Nie and Erbring, 2000, p. 5). 3. See Huber et al. (2002, p. 22) for similar tendencies in Switzerland. 4. In functionalist theory the idea is that the unintended consequences of individual actions assure the functioning of the overall system–how the system comes into being in the first place remains from this perspective an open question. 5. Also Coleman (1988, 1994)—more concerned with the reconciliation of structural analysis and the Rational Choice paradigm—stresses the collective good aspects of social capital while leaving open the aspect of an individual use of social capital. 6. Discussions on measures for social capital based on concepts from Social Network Analysis may be found in Borgatti (1998), Lin (2001), and T¨aube (2002, 2004). 7. See also Lin (2001, p. 12): “These debates and clarifications lead to the suggestion that social capital, as a concept, is rooted in social networks and social relations, and must be measured relative to its root. Therefore, social capital can be defined as resources embedded in a social structure which are accessed and/or mobilised in purposive actions.” 8. See also Lin (2001, p. 19): “To operationalize explicitly the critical elements, we may sharpen the definition of social capital to investment in social relations by individuals through which they gain access to embedded resources to enhance expected returns of instrumental or expressive actions.” 9. Nevertheless, Nie and Erbring’s (2000, p. 19) conclusion that individually reported statements of Internet users on diminishing time commitments in close social relations can be interpreted as increasing tendency towards “social isolation” seems to be somehow problematic. 10. See also Wellman et al. (1996). 11. For all waves N = 1484 persons using the Internet. 12. See Jungbauer-Gans (2002) for results on membership fluctuations in Germany.

240

¨ TAUBE

13. Since we haven’t really operationalised the concept of social capital here, these assumptions rest of course arbitrary. 14. Variable “Internet use”: 0 = Does not use Internet, 1 = 1–180 minutes Internet use per week, 2 = 181 minutes and more Internet use per week. 15. The dependent variables were calculated by multiplying the recorded number of respective social contacts by frequency of contact per month. 16. Dummy variables for “Internet use”: first dummy = 1–180 minutes Internet use per week; second dummy = 181 and more minutes Internet use per week. 17. Even though it would have been interesting to control for the effect of internet use on social contacts with regard to different income groups this aspect has been left out here since we controlled with regard to instrumental contacts only for the neighborhood which is probably a network too limited for to reflect the effect of income disparities on instrumental social contacts.

References Becker, G. (1993), Human Capital, 3rd edition. University of Chicago Press, Chicago, Ill. Borgatti, S., Jones and M. Everett (1998), “Network Measures of Social Capital,” Connections, 21(2), 15–36. Bowls, S. and H. Gintis (1998), “The Moral Economy of Community: Structured Populations and the Evolution of Prosocial Norms,” Evolution and Human Behaviour, 19(1), 3–25. Bowls, S. and H. Gintis (2001), “Die Gemeinschaft als Regelmechanismus: Das ‘Soziale Kapital’ zwischen Markt und Staat” Neue Z¨urcher Zeitung, Ressort Fokus der Wirtschaft, 14 July, 161, 27. Burt, R. (2001), Structural Holes. Harvard University Press, Harvard. Coleman, J.S. (1988), “Social Capital in the Creation of Human Capital,” American Journal of Sociology, 94, 95–120. Coleman, J.S. (1994), Foundations of Social Theory. The Belknap Harvard Press, Harvard. Durkheim, E. (1930), De la division du travail social. Presses Universitaires de France, Paris. Franzen, A. (2000), “Does the Internet Makes us Lonely?” European Sociological Review, 16(4), 427–438. Fiedler, F.E. (1967), A Theory of Leadership Effectiveness. McGraw-Hill, New York. Heintz, B. and Ch. M¨uller (2000), Virtuelle Vergemeinschaftung–die Sozialwelt des Internet, Final Report for the Swiss National Science Foundation, Program “Zukunft Schweiz”. Huber, M., F. Cosandey, V.G. T¨aube, et al. (2002), La soci´et´e de l’information en Suisse: Etat de lieux et perspectives. Neuchˆatel: Office f´ed´eral de la statistique. Jungbauer-Gans, M. (2002), “Schwindet das Soziale Kapital?” Soziale Welt, 2, 189–208. Kraut, R. et al. (1998), “Internet Paradox: A Social Technology that Reduces Social Involvement and Psychological Well-Being?” American Psychologist, 53, 1017–1031. Lin, N. (2001), “Building a Network Theory of Social Capital,” in N. Lin, R. S. Burt and K. Cook (Eds.) Social Capital: Theory and Research. Aldine de Gruyter, New York. Nie, N.H. and L. Erbring (2000), Internet and Society. Report of the Stanford Institute for the Quantitative Study of Society. Putnam, R. (1993), “The Prosperous Community: Social Capital and Public Life,” The American Prospect, 13, 35–42. Putnam, R. (1995), “Bowling Alone: Americas Declining Social Capital,” Journal of Democracy, 6(1), 65–87. Putnam, R. (1996), “The Strange Disappearance of Civic America,” The American Prospect, 7(24), 7–24. T¨aube, V.G. (2002), Zur Messung des Sozialkapitals von Akteuren mit Einfluß in empirischen Netzwerken; Series: Europ¨aische Hochschulschriften, Frankfurt a. M.: Peter Lang Verlag. T¨aube, V.G. (2004), “Measuring the Social Capital of Brokerage Roles,” Connections, 26(1), 29–52. T¨aube, V.G. and D. Joye (2002), “Determinants of Internet Use in Switzerland: Structural Disparities and New Technologies,” in W. Glatzer (Ed.) Rich and Poor: Disparities, Perceptions, Concomitants. Kluwer Academic Publishers, Dordrecht, pp. 73–86. Tocqueville, A.D. (1864). De la Democratie en Amerique. Wellman, B. et al. (1996), “Computer Networks as Social Networks: Collaborative Work, Telework, and Virtual Community,” Annual Review of Sociology, 22, 213–238.

ON THE IMPACT OF INTERNET USE ON SOCIAL CONNECTEDNESS

241

Volker G. T¨aube received his degree in Sociology, Economics and Psychology (Dipl. Soz.- Wiss.) at the Gerhard-Mercator-University in Duisburg (Germany). After studies at the University of Groningen (Netherlands) and the University of Bern (Switzerland) he worked as scientific assistant at Essen University in Germany (Division: Empirical Social Research). He obtained his doctoral degree (Dr. phil.) at Essen University for a work on an empirical measurement procedure for social capital. Subsequently, he worked from 1998 to 2001 for the Swiss Information and Data Archive Service (SIDOS) before he switched in 2001 to the Swiss Federal Statistical Office, SFSO, in Neuchâtel where he is today the deputy head of the division “Economic Cycle and Structure”. His professional and scientific interests include: indicators for the Information Society in Switzerland, Social Network Analysis, and the measurement of social capital.