Search

Using Korean data, this study investigates whether son-favoring ideas or the preference for sons affect fertility decisions. Son-favoring fertility behavior in Korea is of interest because the sex ratio at birth has recovered to a natural level after having been very skewed. To isolate the effects of the preference for sons from the effects of the surrounding environment, we compare the fertility behavior of individuals living in the same region but who were born in different regions or years. Exploiting the male-female gap in educational achievement at the parents’ time and place of birth as exogenous variation in the 2000 Census Korea 2% sample, we find that the strength of son preference formed at an early age is associated with the strength of son-favoring fertility behavior as adults. Our results indicate that parents are more likely to have a third child if they happen to have only daughters as their first two children. More importantly, this tendency is stronger if parents were born in a spatiotemporal region with more skewed gender gap in educational investment. These findings are robust against various alternative specifications, including endogenous migration issues.

In credit-constrained households, some family members who would otherwise not work participate in the labor market. This particular supply of labor, however, is expected to decrease when their families overcome credit constraints. To examine this family investment hypothesis, we develop a test which exploits the dynamic features of the labor supply of married foreign-born women in the United States. Empirical findings, based on the matched March Current Population Survey, are consistent with the hypothesis. This paper also finds that previous results which refute the hypothesis are reversed when the sample is confined to women working in low-skill jobs.

Count models often best describe the nature of data in health economics, but the presence of fixed effects with excess zeros and overdispersion strictly limits the choice of estimation methods. This paper presents a quasi-conditional likelihood method to consistently estimate models with excess zeros and unobserved individual heterogeneity when the true generating process is unknown. Monte Carlo simulation studies show that our zero-inflated quasi-conditional maximum likelihood (ZI-QCML) estimator outperforms other methods and is robust to distributional misspecifications. We apply the ZI-QCML estimator to analyze the frequency of doctor visits.

This paper takes a new approach to testing whether employer learning is public or private. We show that public and private learning schemes make two distinct predictions about the curvature of the wage growth path when a worker changes jobs, because less information about the worker's productivity is transferred to a new employer in the private learning case than in the public learning case. This prediction enables us to account for individual and job‐match heterogeneity, which was not possible in previous tests. Using the National Longitudinal Survey of Youth 1979, we find that employer learning is public for high‐school graduates and private for college graduates.