ABSTRACT
This study estimates the impact of women’s college education on their fertility-related outcomes by taking advantage of a natural experiment known as the graduation quota program that massively expanded women’s opportunities to attend college in South Korea. After the 1979 military coup, the military regime in South Korea ordered all public and private colleges to expand their college admission levels by 30 per cent in 1981 and 50 per cent in 1982. This study implemented instrumental variable (IV) analysis using the Korean Longitudinal Survey of Women and Families (KLoWF). Specifically, our IV analysis uses the birth cohorts that were differently exposed to this sudden and exogenous policy change as an instrumental variable to identify the longer-term effects of college education on women’s fertility-related outcomes. It is found that college education caused those women who were induced to attend college by the graduation quota program to have fewer children. Furthermore, this study finds that this impact can be partially explained by women being less likely to get married.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 The Total Fertility Rate (TFR) generally mentioned in the paper refers to the period TFR. On the other hand, the TFR in our analysis is the cohort TFR that is created by using specific birth cohorts.
2 The natural population change is a population change determined by the number of live births and deaths and excludes migration flows.
3 Better quality of children refers to more-educated and healthier children, while the quantity of children indicates the number of children.
4 In South Korea, the academic year starts in March and ends in February. Thus, the 1981 academic year was from March 1981 to February 1982.
5 Please refer to an article in the Kyunghyang newspaper posted at: http://newslibrary.naver.com/viewer/index.nhn?articleId=1980070800329207010&editNo=2&publishDate=1980-07-08&officeId=00032&pageNo=7&printNo=10699&publishType=00020&from=news.
6 The estimate when we directly use the RD method for fuzzy RD cannot be interpreted to be the true impact of the policy intervention since the cutoff point is not deterministic in fuzzy RD.
7 The age-specific fertility rate is the annual number of births to women of a specified age or age group per 1,000 women in that age group.
8 We also created a table to examine college completion rates, and found that college completion rates are similar to college attendance rates. It is available upon request.
9 The IV analysis in this study is similar to the one in Jung et al. (Citation2016).
10 Equation (1) is the second-stage regression model and equation (2) is the first-stage regression model.
11 For details, please see the control variables presented in .
12 The cumulative number of overseas adoptions from South Korea is 167,547 since 1958. Most of those adopted children were born to single mothers; 99.7 per cent of children sent abroad for adoption were born to single mothers in 2018 (Ministry of Health and Welfare, Citation2018).
13 The related results are not included in the manuscript. They are available upon request.
14 Heckit is not used to correct the sample selection bias by excluding the single women in this analysis since using Heckit seems to do more harm than good in terms of analysis. Specifically, Heckit was considered to be innovative when it was introduced but is not commonly used nowadays because it is sensitive to the choice of variables included in the selection function and it is also sensitive to violations of underlying model assumptions, particularly the bivariate normality assumption (Briggs, Citation2004; Bushway et al., Citation2007).