We analyze household and school survey data from poor counties in six Chinese provinces to examine the effects of poverty, intra-household decision-making, and school quality on educational investments (enrollment decisions) and learning outcomes (test scores and grade promotion). Unlike previous studies, we use direct measures of credit limits and women's empowerment. Drawing a distinction between the effects of wealth (measured by expenditures per capita) and credit constraints, we find that the former improves learning while the latter reduces educational investments. We find evidence of a story of gender bias in which academically weak girls are more likely to drop out in primary school while most boys continue on to junior secondary school. Women's empowerment reduces the likelihood of dropping out but does not affect other outcomes. Finally, our measures of school quality have some effect on the duration of primary school enrollment but not on learning.

This paper analyzes the relationship between a woman's intrahousehold bargaining position and her welfare within marriage. Simultaneity problems common to the literature are overcome by using dowry to proxy for bargaining position. Omitted variable bias is addressed by using grain shocks in the year preceding marriage and sibling sex composition as instruments for dowry. Instrumented dowry positively impacts several measures of a wife's welfare, including time allocation, household purchases, and the wife's decision-making authority, thereby offering strong evidence to support collective models of the household.

Since the dissolution of the Rural Cooperative Medical System at the end of the commune period, illness has emerged as a leading cause of poverty in rural China. To address the poor state of health care, the Chinese government unveiled the New Cooperative Medical System in 2002. Because local governments have been given significant control over program design, fundamental characteristics of the program vary from one county to the next. These differences may influence the decision to seek health care as well as the choice of hospital conditional on that initial decision. In this paper, we use a nested logit model to analyze household survey data from 25 counties to analyze the determinants of such health-seeking behavior. We find that age, the share of household expenditures allocated to food consumption (a measure of relative income), and the presence of other sick people in the household negatively affect the decision to seek health care while disability has a positive influence. Further, conditional on seeking treatment, the reimbursement scheme in place in each county and the average daily expenditure associated with hospitalization strongly influence hospital choice.

Despite two decades of rapid growth, indigenous Chileans are disproportionately poor. However, income data obtained from non-representative surveys yield imprecise estimates of poverty and inequality. This paper therefore estimates poverty and inequality using poverty mapping methods. In contrast to previous studies, however, we use ethnicity rather than geography as a basis for disaggregation. We find that indigenous Chileans are significantly poorer than non-indigenous people, but that inequality rates are also lower for indigenous groups. These reliable estimates of poverty and inequality may augment the antipoverty targeting criteria used in Chile, helping policy-makers to better identify poor households.

No abstract