Wife vs. Husband: Does It Matter Who Answers the Survey?
In the media: World Bank Development Impact Blog
Information on assets is used to conduct empirical research and to guide public policy. When using these variables, practitioners assume that they are less susceptible to misreporting. To test this assumption, I use data from poor households participating in Mexico's PROGRESA program. The same questions on assets were asked to the wife and to the husband. I find: (1) important discrepancies in the information reported between the spouses. For example, there is disagreement in 24% of the couples on the possession of a washing machine; (2) this result has consequences for identifying families living in poverty. For example, if asked to husbands 10.1% of the households would be classified as non-poor, but as poor if asked to wives; and (3) the discrepancies observed are partially explained by unanswered questions to a cognitive test applied at the beginning of the interview. This result is robust to a bounding argument for omitted variable bias implemented by Oster (2017). Overall, these findings suggest that the information on assets is not free of misreporting and who answers the survey matters.
Socio-Economic Development and the Empowerment of Women Within the Household
This paper analyzes the role of GDP, income inequality, and sex ratio on women's empowerment measured by domestic violence, personal freedom, gender roles, and participation in household decisions. Using three waves of a national-state representative survey specialized in women's empowerment in Mexico and state fixed-effects models, I find: (1) an increase in GDP is associated with improvements in personal freedom; (2) a better income distribution improves the participation of woman in household decisions; (3) sex imbalances affect the perception of gender roles; and (4) economic growth reduces the likelihood of suffering sexual violence. To check the robustness of these results to unobserved time-variant variables, a novel bounding technique is implemented. The results suggest that the estimates are robust to omitted variable bias.
Monetary Transfers and Domestic Violence
One of four women experience intimate partner violence in Mexico. I provide evidence of the effects of cash transfers (remittances and conditional cash transfers) on domestic violence by exploiting a national-state representative survey specialized in domestic violence in Mexico. I find that receiving remittances increases the likelihood of domestic violence by 6.2 percentage points; yet, I do not find evidence that conditional cash transfers affect domestic violence. In addition, I find a strong association between households that receive remittances and husbands who do not work. This suggests that spouses, in order to compensate the lack of income within the household, potentially exercise violence to extract part of the remittances in possession of the wives.
War on Drugs vs. Natural Disasters: Which Affects More Household Decisions?
This paper estimates the effects of Mexico's war on drugs and natural disasters on consumption, assets, credit, children's employment, school attendance, and use of time. Using fixed-effects models, the results suggest that: (1) there is no effect of violence and natural disasters on consumption; and (2) families smooth their consumption affecting their children's human capital instead of using assets or getting more credits. Yet, the way they affect the children's human capital depends on the type of shock. The violence increases children's labor supply (particularly males) and the natural disasters decrease school attendance. Finally, I analyze the effects of these shocks on the use of time. While the violence increases the time spent on household activities, the natural disasters decrease the time spent on household activities.
The Impact of Earthquakes on Mental Health
(with Juan Enrique Huerta-Wong, Julieth Santamaria, and Isidro Soloaga)
This paper analyzes the effects of earthquakes that impacted Mexico on women's mental health and substance abuse in 2017. Using a difference-in-differences approach we find: (1) the earthquakes have negative consequences on women's mental health; (2) we do not find evidence that the earthquakes increased the consumption of alcohol or cigarettes; (3) we find evidence of factors that worsen women's mental health (such as perception of insecurity), and others that help women to be more resilient (such as family size); and (4) women who received psychological support improved some measures of mental health, yet we do not observe this result for all the measures analyzed. It is estimated that 27\% of the population in Mexico is exposed to earthquakes. To face this situation, the Mexican government has implemented the Fund for Natural Disasters (FONDEN). This fund is used to distribute food and money for reconstruction. Yet, it is necessary to analyze the possibility of extending its support to the attention of mental health.