| 摘要: | 研究期間：10108~10207;The twinning approach has been widely used in economics, sociology and behavioral genetics to examine the role of genetics on a broad range of demographic, social and economic outcomes. However, the focus of the literatures has been distinct. The economic researchers have been primarily concerned with the need to control for unobserved characteristics in interpreting the causal relationship (e.g., the impact of education on earnings, fertility, health, and children’s economic outcomes). By using data on pairs of identical twins, one can effectively “difference out” genetic and fixed family characteristics. On the other hand, behavioral genetic analyses have mostly been concerned with decomposing the variation in the outcomes into genetic, shared environmental and non-shared environmental components. The standard model requires a linear and additive function form, no allowance for assortative mating as well as the interaction between genes and environment. Therefore, the behavioral genetic approach is subject to more restrictive assumptions than the within-MZ approach in economics. In the economics literature, studies based on twins data are concentrated on examining the causal impact of education on earnings, health, fertility and children’s schooling (Amin, 2009, 2011; Bonjour et al., 2003; Behrman and Rosenzweig, 2002; Behrman et al, 2011; Bingley, Christensen and Jensen, 2009; Holmlund, Lindahl and Plug, 2011; Pronzato, 2011). The implications for this line of evidence are important for designing education policies. Further, under the assumption that twins share the same family environment, the differences between fraternal and identical twins can be attributable to genetic contributions. Using the Taiwanese twins data, we examine the causal effect of female schooling on fertility, and the parental schooling on children schooling. Despite that the negative association between female education and fertility and the positive association between parental and child schooling has been well documented in the social sciences and intergenerational-mobility literature, only recently, however, have studies emerged identifying the causal effect of education on fertility and that of parental backgrounds on children’s outcomes. For the first topic, the vast majority of studies use the instrumental variable approach; for the second topic, researchers have employed three strategies to identify a causal effect of parental education on children’s education: adopted children, identical twins, and instrumental variables (IV approach is mostly employed). As studies differ in context, data and identifying strategies, the evidence is quite mixed. It is worth noting that the IV estimates tend to be local average treatment effects (LATE) that are relevant for individuals who are at the margin to be affected by the instruments used; however, they are not average treatment effects (ATE) for the broader population beyond this margin. Because within-twin schooling differences exist over most schooling levels, the twin fixed-effect estimates are more likely to be closer to ATE rather than LATE. To date, most twin studies exist in the U.S., Denmark, Sweden, Norway and Australia. The evidence from the developing countries is scant, and also, to our knowledge, there is no study on this subject based on such a nationally representative and comprehensive databank in Asian countries. The twins data in this study is created by combining the household data, birth data as well as the college entrance examination data in Taiwan. Based on the same-sex twins, we use the within-twin approach to difference out the genetic endowments and unobserved family characteristics. Controlling for assorative mating and other controls, we will regress the difference in schooling between the children of twin parents on the difference in schooling between twin parents. As regard to the impact of female education on fertility, we will regress the difference in schooling of female twins on the difference in the number of children and the age at first birth. The findings in this study would provide important implications on designing the education policies. |
