Do Parents Selectively Time Birth Relative to Ramadam? Evidence from Matlab, BangladeshAdd to Calendar
12:00 pm – 1:00 pm
Studies examining health effects of maternal fasting during Ramadan find that in utero exposure to Ramadan has negative implications for child health, labor and education out-comes. One key methodological concern in those studies is that the identifcation is based under the assumption that parents do not selectively time birth relative to Ramadan. In Matlab, a region in Bangladesh, a family planning program was initiated in 1977 in which
women in treatment areas received free contraceptives door to door but those in control areas were not part of the program. Using the Matlab Health Socio-economic Survey 1996, I find women living in treatment areas are five to six percentage points less likely to give birth 8 to 9 months after Ramadan after exposure to the family planning program. Moreover, more educated mothers living in treatment areas, are less likely to overlap their pregnancies with Ramadan. This suggests that parents selectively time birth relative to Ramadan and selection in in utero exposure to Ramadan should be checked allowing for time varying changes. I also examine child height, which reects both genotype and phenotype in utero, and find evidence suggesting that presence of selection problems may lead us to wrong conclusions about the effect of maternal fasting on child height. To explore that negative association between mother years of education and Ramadan exposure is not limited to treatment areas of Matlab, I use the four waves of the Indonesian Family Life Survey (IFLS) and find similar parental selection in Ramadan exposure.