Intra-household Resource Shares under Poverty Transfers [Draft]
This paper estimates a structural model of household behavior in the presence of cash transfers. I employ a GMM estimation technique and data from Ecuador to recover how much resources are allocated to men, women and children. Results show that there are intra-household inequalities, but the transfer induces a redistribution of resources among household members. Then, I explore the potential implications of this redistribution of resources in several domains: women's control of resources, measurement of poverty and inequality, consumption patterns and perception of living conditions. I find that women's control of resources changes strikingly when the household receive the transfer. I also show that the policy intervention generated welfare gains in terms of reducing poverty, but there are important misclassification errors. In the last part, I construct a proxy for women’s bargaining power to link women's control of resources and the household demand for food, education and health. Using the probability of receiving the cash transfers as an instrumental variable for the treatment of interest, I find that households where mothers have the majority of the control of resources increase the share of food (4%), decrease the share of education (0.9%) and do not have an effect on the share of health. I also provide compelling evidence that when women control the majority of household resources it influences how households react to unexpected shocks and affect the perception of standard of living.
Intra-Household Allocations and Consumption Inequality under Cash Transfers and Violence [Draft]
In this paper, I study how different types of households determine adult members' allocations of time and consumption. Household types are characterized by the presence of partner violence and cash transfers. Using a collective intra-household decision making model, together with data from an experimental evaluation of a cash transfer program in Ecuador, I structurally estimate the parameters of the model. Then, I perform a poverty analysis at the individual level for the different types of households and find that women are substantially poorer than men, and that income distribution is more unequal for women than it is for men. I also find that the policy intervention generated welfare gains in terms of reducing overall and individual poverty. However, these welfare gains are heterogeneous among the different types of households. Particularly, I find that transfers are effective in reducing the gender poverty gap mainly in households where there is no violence. Finally, I estimate indifference scales for the different types of households to measure how much income an individual living alone needs to have in order to be as well off as when living as a couple. I find that men need a higher share of initial household resources compared to women, and that indifference scales for women decrease with violence and increase when the household is a beneficiary of the transfer. This work contributes to understanding how intra-household allocation of resources takes place among different types of households, the importance of gender difference in poverty and inequality, and the effectiveness of poverty policies when there are factors that generate inequality in consumption.
Conditional Cash Transfers, Household Time Allocation and Bargaining Power: The Human Development Bonus in Ecuador [Draft]
Many developing countries provide cash transfers to low-income families. Exogenous changes in household income coming from a cash transfer program may alter the bargaining power of the recipient and the allocation of time within the family. Using a large-scale living standards measurement survey from Ecuador, I implement a fuzzy regression discontinuity design to show that conditional cash transfers affect the women’s freedom to decide (bargaining power) as well as women’s time allocation to certain activities. To complement the empirical results, I propose an intra-household decision-making model to study how heterogeneity in preferences over child quality induced by the conditionality of the transfer and differences in female income relative to male income affect the optimal decisions of household members related to time allocation and home production. In a further quantitative exercise, I compare how the household behaves in a scenario without transfer versus a scenario with a cash transfer equivalent to the amount provided by the program.
Experience, Tenure and Non-cognitive Skills: Evidence from the UK [Draft]
This paper examines the role of personality on workers' accumulation of experience, tenure and wages. Using a large-scale household longitudinal dataset from the United Kingdom, I am able to reconstruct labor market histories of individuals after leaving full-time education and merge this information with available data on measures of personality traits. Using this data, I find significant impacts of non-cognitive skills on the accumulation of workers' experience and tenure. The consistency and robustness of these effects are confirmed using an instrumental variable approach as well as a bounding strategy. To evaluate possible heterogeneous effects, I implement a quantile regression approach and a generalized propensity score method. I find that the effects of the traits differ along the experience and tenure distributions. Furthermore, the results suggest that differences in the intensity of a trait influences the expected level of accumulation of experience of a worker as well as his expected level of tenure. Finally, I construct and estimate a simple structural model that allows me to illustrate how an individual's endowment of personality affects earnings through three different channels: direct effects, schooling effects and tenure effects. By disentangling these different channels, this paper provides more detailed results of the returns to personality.
Child Labor, Schooling and Idleness in the Presence of Cash Transfers [Draft]
Developing countries provide subsidies to poor household to promote child education and mitigate child labor. This paper investigates the effect of a cash transfer on the household decisions regarding child activities (schooling, work, leisure), and the allocation of hours towards working activities. For this analysis, I use data from a randomized evaluation of a conditional cash transfer program in Ecuador. The empirical results suggest that the most prevalent behavioral shift caused by the program was a reduction in the probability that the household chooses child leisure, a reduction in the probability that the household sends the child to work and an increase in the likelihood that a household sends the child to school. On the other hand, the transfer reduced the allocation of time to certain child working activities, but these results are mainly driven by the effect of the transfer at the extensive margin. To rationalize these findings, I develop a theoretical model of parental decision regarding child activities. With this framework, I argue that a cash transfer attenuates the likelihood of parents choosing leisure and market work for their child and increases the likelihood that they send the child to school. By modeling the cash transfer as a subsidy of the human capital input and as lump sum transfer, this study also contributes to the discussion of whether cash transfers should be conditional (on school enrollment) or unconditional
Work in Progress
Pollution and Education: The Effect of Power Plant Closures on Students' Test Scores. [Joint with Julieth Santamaria]
Fuel-fired power plants emit millions of tons of air pollutants each year and have been found to affect segments of the surrounding regions. Recent work has demonstrated a link between pollution and health issues such as respiratory problems or cognitive performance. On the other hand, health problems could influence students' academic performance. In this paper, we study how exposure to environmental damage from power plants affect the human capital production process of students. Using different data sources, a panel of power plants and school districts throughout the United States is constructed. Relying on a quasi-experimental design that compare locations, we analyze if there is evidence of an effect on students' test scores, following the closure of power plants.
We construct a panel database for all states in the U.S. from 1980 to 2016, and document that the average corporate income tax rate has declined by approximately 40%. During the same period, we observe that most states have gradually shifted towards imposing a sales-only apportionment weight on multi-state firms. We ask whether these patterns are consistent with states competing in setting their corporate tax policy. Empirically, we find evidence of strategic interaction in setting tax policies between neighboring states. Theoretically, we show that moving towards a sales-only apportionment scheme is consistent with the prediction of a dynamic general equilibrium model of tax competition that incorporates the Formula Apportionment rule.
Pollution Inspection and Punishment
High levels of contamination continue to be a major issue in many countries even with strict environmental rules. Environmental inspections create a signal of whether the polluter firm has behaved dishonestly. I model this process to understand if an increase of punishment influences the probability of compliance with the regulation. In a scenario where the signal is perfect but costly, the regulatory agency and the polluter firm play a typical inspection game, whereas in a scenario where the signals are available without cost, but susceptible to errors, the regulatory agency and the polluter firm play a monitoring game. In both games it is assumed that the regulatory agency wants to promote compliance with the environmental regulation and at the same time it wants to correctly decide the application of punishment. Under the inspection game framework there is only one mixed strategy equilibrium, whereas in the monitoring game there is three perfect Bayesian equilibrium. The outcomes in the monitoring game can be characterized according to type of punishment (repressive, severe and tolerant). The analysis of the model shows that equilibrium probability of compliance with the environmental regulation is lowest under a repressive type of punishment, and highest under a tolerant type of punishment. Eradication of dishonest behavior is impossible. Moreover, an increase of punishment does not increase the probability of compliance with the environmental regulation.