Structural Experimentation to Distinguish between Models of Risk Sharing with Frictions in Rural Paraguay

Abstract

Abstract. We conduct dictator-type games in rural Paraguay; different treatments involve manipulating players' information and choice sets. From individuals' choices in the games, we draw inferences regarding impediments to efficient risk sharing. Behavior in the games suggests that players in some villages are reacting to the kinds of incentives we would expect under private information, while in others players direct payoffs to recipients in a manner consistent with limited commitment. Overall we rule out full insurance, hidden investment alone, and limited commitment alone in favor of an environment with both hidden investment and limited commitment, though there is considerable variation across villages.JEL codes: C93, D85, D86, and O17.

Migration and Informal Insurance: Evidence from a Randomized Controlled Trial and a Structural Model

“…The results show that treatment also reduces the coefficient on log income for non-migrant households, suggesting that measurement concerns are not driving the treatment effects we observe in Table 4. These estimates are consistent with the experiment changing the risk-sharing equilibrium be- 17 Appendix Table 6 report the effects of various sub-treatments that could help discern the presence of hidden income (Townsend, 1982;Rogerson, 1985;Ligon, 1998;Kinnan, 2020). If migration makes it easier to hide income because some of the income is earned away from other villagers' watchful eyes, risk sharing can break down.…”

“…Households make migration decisions taking into account the returns to migrating, including risk-sharing transfers. We assume that risk sharing is constrained by limited commitment frictions (Kocherlakota, 1996;Ligon et al, 2002;Krueger and Perri, 2010). Other frictions income for control villages and shows that mean liquid and cash savings are around 5% of mean household income and 10% of the standard deviation of household income.…”

“…This problem is constrained by two sets of incentive compatibility constraints (one for each household at the before-migration stage, and one for each household at the after-migration stage), as well as a promise-keeping constraint that household one receives the utility promised to them. We follow the solution concept proposed in Ligon et al (2002) by solving for the conditional Pareto frontier that maximizes the utility of household two given a promised level of utility to household one. Because households make choices at two points in time during a period, we define two Pareto frontiers: first, the frontier that maximizes the before-migration utility of household two, V(h), given a state-dependent level of before-migration promised utility, U(h), to household one.…”

“…However, each of these mechanisms is imperfect. Informal risk sharing provides only partial insurance (Townsend, 1994;Ligon et al, 2002;Kinnan, 2020), and migration may be costly or itself risky (Bazzi, 2017;Beam et al, 2016). The two approaches may also interact with one another.…”

“…We reconcile these conflicting empirical results by showing that it is possible to get either crowd-in or crowd-out of risk sharing from the same theoretical model under different local conditions and preferences and also different assumptions about the permanence of the policy itself. Our model augments Morten (2019), which considers a model with limited commitment constraints on risk sharing (Kocherlakota, 1996;Ligon et al, 2002;Krueger and Perri, 2010) and an endogenous migration decision based on the net return to migrating (Harris and Todaro, 1970). To that we add the possibility that migrants develop contacts in the city, allowing them to better obtain jobs.…”