Does Infrastructure Facilitate Social Capital Accumulation?　Evidence from Natural and Artefactual Field Experiments in a Developing Country

While social capital in general has been recognized as essential for economic activities, its accumulation mechanisms are largely unexplored. How does people’s trust toward others, one of the core dimensions of social capital, emerge? To shed new light on this largely unanswered question, we investigate the impact of physical infrastructure on social capital accumulation by comparing two hypotheses: the habit formation hypothesis and the repeated interaction hypothesis. We use a unique dataset from an irrigation project in Sri Lanka under a natural experimental situation where a significant portion of irrigated land was allocated through a lottery mechanism. Also, we look at the level of social capital using artefactual field experiments by a strategy method based on a within-subject design. By combining these two instruments, we find that physical distance embedded by irrigation systems explain variations in trust across irrigation communities, suggesting that the level of particularized trust is significantly higher than that of general trust. Also, within-community variation in particularized trust is driven largely by each individual’s years of access to irrigation and is not necessarily affected by social distance or repeated interaction among farmers. Our results indicate that social preference emerges from a technological environment set by physical access to irrigation, suggesting habit formation of pro-social behavior.