I am an Assistant Research Scientist at the Population Studies Center at University of Michigan.
My broad research interests are in health economics, labour economics, and demography, with a focus on public policy. My work explores two related topics: causes and consequences of fertility behaviours, and human capital development. I am particularly interested in childhood development and how the environment – both physical and social – influences fertility and health capital accumulation. I am also part of a long-term initiative to study effects of improving contraceptive access.
Fasting during pregnancy has adverse effects on fetal and adult health of the offspring. In the case of Ramadan, the Islamic month of fasting, these effects can be avoided through mothers timing pregnancy to avoid Ramadan during the gestation period. This paper examines whether mothers time pregnancy to avoid Ramadan, and the mechanisms behind the avoidance behaviour. Using fertility data from the Indonesian Family Life Survey, I find strong evidence of Ramadan avoidance in the timing of contraceptive use. The probability of using 3-month injectable contraceptives increases by 0.6 percentage points two months before Ramadan to ensure return to fertility during the optimal time to conceive. I also find a 0.2 percentage point increase of in the likelihood of conception after Ramadan. Moreover, my results suggest that expansion of Indonesia’s Village Midwife Program allowed mothers to use contraceptives, specifically birth control injections, to avoid Ramadan.
Dengue is one of the most wide-spread mosquito-borne viral diseases in the world. 390 million people are infected annually and 2.5 billion people live in areas where outbreaks occur every 3-5 years. Thus far studies on the consequences of dengue exposure during pregnancy have focused on more severe, symptomatic cases which account for only a quarter of total infections. Effects of a typical dengue infection, which is asymptomatic, are unknown. I combine data from the Passive Dengue Surveillance System and birth records for Puerto Rico for the years 1990- 2010 to examine the relationship between dengue and pregnancy outcomes. My identification strategy uses rainfall as an instrument for dengue and allows me to estimate the upper and lower bounds for effects of dengue on outcomes of interest. I find strong evidence that prenatal dengue exposure decreases birth rates. Pregnancy loss occurs in at least 15% of pregnancies with dengue, which includes both symptomatic and asymptomatic cases; annually there are 1-6% fewer births due to dengue in Puerto Rico. These effects are driven by exposure in the second trimester. Dengue exposure in utero also reduces gestation, and there is weak evidence that it reduces infant size even among-full-term pregnancies, although I cannot rule out null effects. Finally, this study illustrates the importance of accounting for measurement error, as specifications which ignore measurement error produce coefficients that may understate the detrimental effects of dengue exposure during pregnancy by several orders of magnitude.
This paper evaluates the consequences of workplace drug testing policies on marriage and fertility of black women. I exploit variation in the timing of state regulation on drug testing to evaluate the effects on the male-female earnings differential, marriage, childbearing, and children’s outcomes. The results suggest that pro-testing regulation leads to a 7.3% increase in earnings for black women and a 3.1% increase in earnings for black men, decreasing the male-female wage gap. Despite economically and statistically significant changes in economic conditions, the results suggest no change in marital status. In addition, black women with at least one child have 0.1 fewer children after the policy. I also show that the policy is associated with an improvement in living circumstances of households with a black household head, suggesting that the existing narrative that the War on Drugs – and drug testing – leads exclusively to a deterioration of black families does not capture the full extent of the consequences.