UC Berkeley

As the world's largest payments for environmental services program, China's Sloping Land Conversion Program has reforested vast areas of environmentally sensitive farmland since 1999 and provided subsidy payments to millions of poor farmers in mountainous western China. This dissertation analyzes the socio-economic effects of the program in Shaanxi Province.

Chapter 1 examines the effects of enrollment in the Sloping Land Conversion Program on household labor market outcomes. It uses the exact timing of enrollment to identify a causal link between enrollment and non-farm employment, and finds that enrollment has a small but statistically significant positive effect on non-farm employment at the household level. The probability of an adult without non-farm employment beginning such employment in a particular year increases from 1.4% in years without new household enrollment to 2.1% in years with enrollment. This amounts to an increase of national labor supply of at least 600,000 individuals. The analysis measures enrollment in different ways to distinguish among competing channels of the effects of enrollment on employment, and finds that employment effects arise not from alleviating constraints, as other researchers have suggested, but rather from simple farm to non-farm labor substitution.

Chapter 2 focuses on problems in the implementation of the program, including farmers not receiving subsidy payments to which they are entitled, and the over-reporting of areas eligible for subsidies on the part of local governments. On average, villages reported 72% more area enrolled in the program than was actually the case, and 15% of enrolled farmers received at least a portion of the subsidies to which they were entitled late or not at all. The chapter finds that both misaligned incentives and low managerial ability contribute to inefficient outcomes. Villages that are poor and remote (and are assumed to be less able to fund administrative costs without over-reporting and less likely to be audited) over-reported more, while farmers were less likely to receive subsidy payments to which they were entitled if the village leader had lower managerial ability and a larger village to manage. In the Sloping Land Conversion Program, finely tuned targeting that might be optimal in a smaller program is impractical due to administrative costs.

Chapter 3 compares determinants of enrollment at the parcel and household levels for parcels where farmers made the decision of whether to enroll to parcels where local governments made the decision. It finds no evidence that farmers place more weight on productivity relative to ecological factors, but rather that decisions made by local governments are more easily predicted by plot characteristics such as slope and soil quality, and to some extent by a desire to create contiguous forests, whereas farmers place more weight on land characteristics relative to land on the same farm and on education and other household characteristics. The most important difference between farmer and local government decision-making is the frame of reference, the scale within the landscape to which land under consideration for enrollment is compared, not the relative weights placed on different criteria of suitability for enrollment.