Using data from a two-year pricing experiment, we study the impact of subsidy policies on weather insurance take-up. Results show that subsidies increase future insurance take-up through their influence on payout experiences. Exploring mechanisms of the payout effect, we find that for households that randomly benefited from financial education, receiving a payout provides a one-time learning experience that improves take-up permanently. In contrast, households with poor insurance knowledge continuously update take-up decisions based on recent experiences with disasters and payouts. Combining subsidy policies with financial education can thus be effective in promoting long-run insurance adoption.

We use a randomized experiment in India to show that improved technology enhances agricultural productivity by crowding in modern inputs and cultivation practices. Specifically, we show that a new rice variety that reduces downside risk by providing flood tolerance has positive effects on adoption of a more labor-intensive planting method, area cultivated, fertilizer usage, and credit utilization. We find that a large share of the expected gains from the technology comes from crowding in of other investments. Therefore, improved technologies that reduce risk by protecting production in bad years have the potential to increase agricultural productivity in normal years.

In many developing countries property rights over rural land are maintained through continuous personal use instead of by land titles. We show that removing the link between land use and land rights through the issuance of ownership certificates can result in large-scale adjustments to labor and land allocations. Using the rollout of the Mexican land certification program from 1993 to 2006, we find that households obtaining certificates were subsequently 28 percent more likely to have a migrant member. We also show that even though land certification induced migration, it had little effect on cultivated area due to consolidation of farm units.

Subjective performance evaluation could induce influence activities: employees might devote too much effort to pleasing their evaluator, relative to working toward the goals of the organization itself. We conduct a randomized field experiment among Chinese local civil servants to study the existence and implications of influence activities. We find that civil servants do engage in evaluator-specific influence to affect evaluation outcomes, partly in the form of reallocating work efforts toward job tasks that are more important and observable to the evaluator. Importantly, we show that introducing uncertainty about the evaluator’s identity discourages evaluator-specific influence activities and improves bureaucratic work performance.