|
Abstract:
|
"Social capital is an important element in the successful management of a common pool resource. User groups which share high levels of trust with one another reduce the costs of monitoring and enforcing rules. While theory and games highlight the importance of repeated interactions, empirical work is limited to cross-sectional analysis in which socioeconomic attributes proxy for social capital. I focus on the role of repeated interaction and estimate the impact of introducing new users (not additional) to well-established common property institutions in Taos County, New Mexico. I build a panel data set of 25 communal irrigation systems, known as acequias, from 1984 to 2008. Using satellite imagery to assess the output of each acequia in each year, fixed effects regressions are ran to measure the impact of new users being present. Other user group characteristics are included, specifically, the number of users, the economic heterogeneity of the users, as well as the cultural homogeneity of the group. The panel data and fixed effects analysis allows me to focus on the impact of the disturbance within a system. The structure of the data also permits me to correct for the omitted variable bias which likely plagues cross-sectional analysis. Fixed effects control all variables which vary across systems but not within a system, in addition to those which vary uniformly over time. I find that the introduction of new users, lowering the social capital, does negatively impact the system, though smaller shocks have less impact and the systems are resilient overtime, returning to normal levels of output after 4 years. Findings on the number of users and cultural homogeneity support prior work in that more users and less cultural homogeneity reduces output within a system. Concerning economic heterogeneity, as measured by the Gini coefficient of land holdings, within the systems increased inequality increases production whereas across systems those more equal do better. Given the resilience of the systems in this setting to small disturbances of new users, follow up work is needed to understand what about the institutional setting facilitates this in order to apply the lessons to other settings. While difficult to come by, the importance and distinction of panel data is highlighted by the Gini coefficient result and more research should address these disturbances within systems. It is important to expand our knowledge of how a single system can respond to a given disturbance."
|