Essays in Development Economics

This dissertation explores three important topics in the Development Economics literature. The first chapter provides further evidence on the consequences of illness episodes on earnings and consumption in Indonesia. We exploit a rich panel dataset that allows us to combine fixed effects and instrumental variables to address biases associated with the endogeneity of illness. Using deteriorations in physical functioning to measure illness for a sample of prime-age individuals we find that earnings are significantly reduced. However, the effects on consumption are relatively small and statistically insignificant. In addition, we explore informal mechanisms households may be using to smooth out the consequences of these episodes. In particular, we study whether households deplete liquid assets or receive transfers from non co-resident kin. Transfers are responsive to illness but there is no evidence of asset depletions. Overall, the findings suggest some ability to smooth the income effects of illness on average. The second chapter explores the measurement of expectations about future resources. Expectations about future resources play a key role in many models of behavior. However, few empirical studies include measures of expectations about resources simply because they are not typically measured in surveys. Two new, experimental questions that measure each respondent's expectations about his/her future economic well-being in the future are carefully evaluated: the ladder question (in which the respondent places him/herself on a six-step ladder representing SES), and future household income. Data are drawn from the Work and Iron Status Evaluation in Indonesia. We find that there is remarkable coherence across the questions and their relationship with current resources. We conclude that both measures seem to have content about the future. In particular, we highlight the potential value of the ladder question given its simplicity and low costs of implementation. Finally, the third chapter focuses on another important aspect of developing countries which is the migration of labor across international borders. Undocumented migrants are hard to enumerate in surveys. This chapter provides evidence on the extent and nature of this problem using uniquely rich data from the Mexican Family Life Survey. MxFLS, an on-going longitudinal survey of Mexicans, follows and interviews movers to the U.S. In the first follow-up, we interviewed 91% of movers to the U.S. We compare respondents in MxFLS with recent migrants from Mexico interviewed in U.S. surveys. MxFLS respondents are less educated, more likely to be working and have lower earnings. In addition, we carefully study the selectivity of movers who stay in the U.S. and those who return to Mexico. Contrasts are drawn with Mexicans who move within Mexico and those who do not move. These contrast allow us to inform about the selectivity of Mexican migrant central to understand the consequences of this phenomenon for both Mexico and the U.S.