Abstract

Micro-enterprise development has become one of the most important approaches to reduce poverty in the LDCs. Like other LDCs, in Ethiopia micro-enterprise development forms the major component in the promotion of broad based growth and improvement of the well-being of the poor by providing income generating opportunities. Accordingly, formalizing informal sector activities has become one of the priorities of micro-enterprise programs. The main aspect of micro-enterprise programs is the use of social capital as a substitute for collateral in providing credit and forming enterprise groups. Despite the significance of social capital in micro-enterprise programs in particular and the informal economy in general, its nature and potential contributions remain under-investigated in Ethiopia. The purpose of this study is, therefore, to examine the configuration of social capital among the poor street vendors in Addis Ababa. The study has employed the network approach to social capital. Data were collected from 154 street vendors living in Addis Ababa. Multi-stage sampling procedures involving purposive and systematic random-walk techniques were used to draw samples. The study applies a mixed-methods research design. Accordingly, quantitative and qualitative data were collected through name and position generator surveys and in-depth interviews. While the quantitative data were analyzed through social network analysis procedures and statistical techniques, the data from interviews were transcribed, classified, and presented in a narrative form. Two-sample T-test, one-way-ANOVA, and OLS and Instrumental Variable regression models were used as statistical tools for the study. The results of the study reveal that homophily in religion and ethnic lines forms the strongest divide among street vendors’ personal networks followed by sex and marital status homophily. However, street vendors exhibit heterophilous networks regarding income, age, and occupation. Street vendors demonstrate dense, less effective, less efficient, and highly constrained network structures. They also exhibit greater proportion of strong ties in their personal networks. Street vendors have most of their relationships with people of lower occupational prestige. In addition, they have low access to high prestige positions, low resource heterogeneity, and low social capital volume. Comparisons of networks between gender and among ethnic groups show the presence of significant differences. Women’s network exhibits larger percentage of strong ties and more ethnic and religious homogeneity than men. Also, women exhibit small network size, less effective, and highly constrained networks. Network characteristics by ethnic group shows that the Gurages exhibit high proportions of strong ties and high levels of ethnic homophily but embedded in networks of diverse education, occupation, and income compared to the Amharas and the Oromos. Conversely, the Amharas have diverse ethnic and religious contacts and demonstrate relatively high proportions of weak ties than others. Structurally, while the Gurages exhibit large network size with dense and less effective networks; the Amharas display small network size and less dense networks. The overall heterogeneity index shows that the Gurages exhibit more heterogeneous networks than the Amharas and the Oromos. By examining network dynamics, the study also reveals significant changes in the number and nature of ties kept, ties lost, and new ties created over the phases of enterprise development. There have also been changes in network composition and structure over the three entrepreneurial phases. The study further investigated the effect of social capital in enterprise performance. Four separate regression models were fitted to predict the effect of relational, structural, and embedded resources dimensions of social capital on enterprise performance. After controlling the potential endogeneity problem of social capital, the estimation results revealed that the resources embedded in networks contribute positively to enterprise performance compared to the strength of ties and the structural constraint. Human capital measures, on the other hand, do not significantly predict enterprise performance. The implications of the outcomes of the study is that in providing credit and establishing enterprise groups, micro-enterprise programs should evaluate the trade-off between strong versus weak ties and homogeneity versus heterogeneity of networks. While religion, ethnicity, gender, and marital status homophily as well as strong family and friendship ties are worthwhile for credit delivery and forming enterprise groups, network heterogeneity is central for enterprise success. Overall, it is unlikely that social problems can be resolved without analyzing the social ties of individuals in particular and the community in general. Thus, it is imperative to conduct further studies in a broader scope to advance the significance of social capital for poor targeted development interventions in Ethiopia.