سرمایه اجتماعی و عملکرد بازپرداخت وام در جنوب شرقی آسیا

This paper analyses the effects of different forms of social capital on credit repayment behaviour. In the context of development economics an innovative data collection approach is used that originates from the field of sociology. A personal network survey is carried out to measure the individual social capital of borrowers. We distinguish four social capital variables between the respondent and his/her network members according to tie strength (bonding/bridging) and social distance (bondinglink/bridginglink). The effects of social capital vary according to socio-cultural context. For instance, in Thailand bonding and in Vietnam bridginglink social capital has a positive effect on loan repayment.

Social capital can facilitate access to credit, but also affect repayment behaviour of borrowers (van Bastelaer, 2000). Many studies link repayment performance to social capital (Cassar et al., 2007) or to social ties (Ahlin and Townsend, 2007). However, these scholars use joint liability credit groups as the unit of analysis. Little research exists on repayment performance of individual borrowers and social capital or social ties. Most studies focus on social ties crossing power differentials in which powerful individuals exploit the credit program. In addition, their methodological design is often anecdotal e.g. Vaessen (2001). Moreover, the literature on loan performance within credit groups concentrates on intra-group ties, but excludes ties to persons outside the group. An important aspect of social capital is thus left out. Research in development economics has only recently begun to pay heed to social relations or ties. Often, measurement of these ties has been rather crude, focusing for instance on role relationships like friends, relatives, or neighbours. Our approach to measuring social ties is more elaborate. We use a survey tool from the field of sociology hitherto rarely used in development economics.1 This technique involves the use of instruments referred to as ‘name generator’ and ‘position generator’ to measure the personal network, or preferably a sample of the respondent's personal network. These network data are then used to create measures of the individual social capital of borrowers. The empirical part of this work focuses on the influence of social capital on repayment behaviour of rural borrowers. While Ahlin and Townsend (2007) have applied the concept of social capital, albeit using a different definition, to joint liability groups in Thailand, no research so far has applied the concept of social capital to repayment behaviour of individuals in the rural credit market in Southeast Asia.2 The case-study countries here are Thailand and Vietnam. In both research countries, the Government offers various credit lines intended to foster the socio-economic development of the rural population. In Thailand most of these credit lines are administered by the Bank for Agriculture and Agricultural Cooperatives (BAAC). Furthermore, the Thai government manages different lending schemes such as the One Million Baht Village Fund and the Poverty Eradication Fund. About 80 percent of all villages in Thailand have access to the One Million Baht Village Fund (NSO, 2003). The Vietnamese Government offers various credit lines to support rural households. Most of these are issued by the Vietnam Bank for Social Policies (VBSP) or the government based mass organizations. Another major player in the rural loans market is the Vietnam Bank for Agriculture and Rural Development (VBARD). In both countries, formal and semiformal lenders have enormous outreach but with a trade-off: many of their borrowers are not creditworthy. Rescheduling of loans is therefore common, ultimately adding to the amount of nonperforming loans. Our two hypotheses are: • H01: Regardless of the form of social capital, it does not affect individual loan repayment. • H02: The basic socio-economic principles underlying loan repayment behaviour in both countries are different.

The literature on repayment performance in rural areas has so far focused predominantly on credit group and intra-group ties. Our measure of social capital, meanwhile, went beyond intra-credit-group ties and our unit of analysis was the individual credit. In our analysis we distinguish three different forms of social capital: bonding, bridging, and linking and use four different measures of social capital: 1. bonding, 2. bridging, 3. bondinglink, and 4. bridginglink. Bonding social capital has a significant and positive influence on repayment performance in Thailand. But we found no significant impact of bridging and linking social capital on loan performance. This is surprising, as most of the literature suggests the opposite, following the famous argument of Granovetter (1973) on ‘the strength of weak ties’, which highlights the informational gains to be made through weak ties. This could be another example from Asia supporting the finding from Bian (1997) in China, who brought the positive effects of strong ties back into the discussion. However, in Vietnam bridginglink social capital is positively connected to repayment performance. Obviously social capital is context-specific. Measures of social capital are therefore difficult to compare in a cross-country context. However, the underlying social phenomena, such as informational gain or peer pressure are at work in both countries, but of course to a varying degree, depending on the institutional setup. One explanation for the difference in the social capital variable in the two country regressions may be that in Thailand borrowing is a well known everyday procedure. Thus, information gains received on the procedure may be less important. In Vietnam however, the process sometimes is still cumbersome and new to rural borrowers. Furthermore, the difference in bonding social capital may hint to more intact village structures in Vietnam. Help is still given to everybody who is in need regardless of how many people the person knows and how well he is connected to them. In contrast to that, Thai villages seemed to have lost some of their social cohesion during the rapid economic development in the last decades. Thus, bonding social capital based on strong ties is more important to receive help in times of need. What policy conclusions can be drawn from these findings? The most obvious approach for credit institutes in Thailand to foster bonding social capital among their clients would be an increased use of credit group schemes to strengthen the relationships between different borrowers. However, and as other research has shown (Ahlin and Townsend, 2007), too close ties in credit groups can be counterproductive as they hamper or even impede social sanctioning and promote collusion among group members. Thus, when credit group schemes are employed, credit institutes must find a balance. Ties must be close enough to allow for credible social sanctions and peer pressure – but not too close to avoid collusion or impede social sanction. An alternative approach could be the use of credit groups without joint liability or a further extension of the loan guarantor scheme. The aim would be to increase the ‘pressure to conform’ without social sanctioning. But even individual borrowers could be encouraged to engage in a variety of joint activities to further strengthen their relationships with other people. Training provision by credit institutes seems to be a good measure as this might not only improve the human capital of borrowers, but also their social capital, which is the basic idea behind the credit-plus approach in micro-finance. In a few words, credit-plus implies the supply of financial and social services. The pitfall of this approach lies in itself, as any move away from core areas of operation, which are normally financial services, may weaken the financial sustainability of the financial institute. Finally, the cost of such an approach could outweigh any gains resulting from reduced losses. In Vietnam the monitoring of credit groups seems to work well. Thus, one could recommend including individual loans into this monitoring procedure. However, credit institutes in Vietnam are heavily drawing on local government officials for this work and those officials must not be overloaded.