Gender Bias and The Indonesian Financial Crisis: Were Girls Hit Hardest?

We analyze how the financial crisis affected a wide range of investments in Indonesian children and children’s outcomes including school enrollment, immunizations, and mortality. Our dataset is the National Socio-Economic Survey (Susenas), a large nationally representative sample. We build on past research by differentiating outcomes for boys and for girls, and by separating regions heavily affected by the financial crisis from others that were relatively unhurt. Along most dimensions, children were well protected. Contrary to some theory and press reports, girls did not fare worse than boys during the crisis. ______________________________ We appreciate comments from Garrick Blalock, Paul Gertler, Daniel Levine, Jack Molyneaux, and Jules Reinhart as well as price deflators from Kai Kaiser. CIDER papers are produced by the Institute of International Studies and the Institute of Business and Economic Research. This paper can be found online at the new UC eScholarship Digital Repository site: http://repositories.cdlib.org/iber/cider with links to the CIDER publications page: http://www.haas.berkeley.edu/groups/iber/wps/ciderwp.htm The East Asian financial crisis of 1997-1998 had a devastating effect on the Indonesian economy. Substantial research in developing countries suggests that female children often bear a disproportionate share of the burden during economic crisis such as this one. In addition, press and policy-oriented reports on the effects of the crisis in Indonesia support this prediction. This study examines the effects of the crisis on children, focusing on whether its effects were more negative for girls than boys. This paper makes five contributions to the results of past studies: 1. Most studies to date have analyzed data from half or fewer of Indonesia’s provinces. We analyze nationally representative data from all provinces and districts with reliable data. 2. Most studies have used relatively small samples; we analyze the Susenas living standards survey, which provides data on over 200,000 households per year. 3. Most studies have analyzed only a single outcome, such as education, yet existing studies have shown widely varying effects of the crisis on different outcomes. Thus, it is important that we examine a wide range of investments in children and children’s outcomes, including school enrollment, immunizations, and mortality. 4. Although the crisis had very different effects in different regions, most studies have examined national trends (perhaps breaking out urban versus rural regions or Java versus the rest of the nation). We separately analyze trends in regions with higher or lower exposure to the financial crisis. 5. Finally, most studies have looked at children’s average outcomes. We examine not only average outcomes but also whether the crisis had disproportionate effects on girls and whether they were disadvantaged before the crisis. The size of our data set is beneficial for studying multiple outcomes by various age and gender groups, as sampling variation implies that roughly one in 20 of the changes in outcomes we examine will be statistically significant at the 5 percent level. With our large sample size, we