Social capital has been a contested notion among economists since it came up in economic literature in the mid 1990s. It is not a theory as human capital theory is, nor is it a clearly distinguishable field of study, but it is a concept that has been embraced as the missing link in economic analysis. Basically, the concept addresses the social and cultural aspects of human behavior, recognizing that these dimensions have economic implications. Given the variety of views among economists on social capital, there are various definitions of social capital around. Most definitions emphasize a functionalist dimension, referring to collective action (in political science definitions), to social cohesion (in sociology), or to well-being or even economic growth (in economics). For example, the World Bank defines social capital in its research project on social capital and poverty as "the institutions, the relationships, the attitudes and values that govern interactions among people and contribute to economic and soci al development" (1998, 1). Earlier, I have suggested a more neutral definition of the concept, recognizing negative economic effects of social capital for the economy as a whole or for certain groups (van Staveren 2000, 7). In this definition, which I will extend a bit here, I understand social capital as a shared commitment to social values as expressed in the quantity and quality of social relationships, which may enable or constrain dynamic efficiency. Generally speaking, we find three different ways in which social capital is integrated in economic analysis:
* A preference in utility functions (Becker 1996; Glaeser, Laibson, and Sacerdote 2000).
* A resource next to other capitals (Fukuyama 1995a and 1995b; World Bank Social Capital Initiative 1998, summarized by Dasgupta and Serageldin 2000). (1)
* A mechanism to address market failures due to imperfect information and risk (Moore 1999; Dasgupta 2000; Grabowski 1998; Szreter 2000).
The economic literature on social capital and the closely linked notion of trust has become so extensive that I cannot even try to do justice to the literature on social capital and poverty in developing countries.2 My review below on the use of social capital in poverty research therefore necessarily draws on a small selection of the empirical literature, focusing on the more frequently cited publications, edited volumes, and evaluative publications.
Social Capital in Poverty Research
Economists have shown an eager interest in transforming the abstract idea of social capital into measurable variables, for which often secondary data are found outside the trodden paths of economic databases. A much-used macro-level data source is the World Values study (Inglehart 1990 and 1994). At the micro-level, we see primary data collection, predominantly through surveys. Empirical research on social capital employs one or more of the following three types of variables for the measurement of social capital:
* Trust and trustworthiness or credibility.
* Membership in formal and/or informal groups.
* Acceptance of moral rules and norms or adherence to certain values,
An empirical study at the macro level that I will briefly discuss here made use of all three of these types of social capital variables, drawing on the World Values study. (3) Stephen Knack and Philip Keefer (1997) did a sample of twenty-nine mostly developed countries for which data were available on trust and civic norms. They found that the variable of group membership has no impact on growth. But they also found that the variables of trust and civic norms have a significant and positive impact on the economic performance of countries, with more explanatory power for countries with more (income) equality. "A one-standard-deviation change in trust (fourteen percentage points) is associated with a change in growth of more than one-half (. …