Causal inferences: identical twins help and clarity about necessary assumptions is critical.
Vikesh Amin,Jere R Behrman,Hans-Peter Kohler,Yanyan Xiong,Junsen Zhang
DOI: https://doi.org/10.1016/j.socscimed.2014.12.012
IF: 5.379
2015-01-01
Social Science & Medicine
Abstract:Parental schooling is widely thought to improve child outcomes. But most studies on parental-child relations are associative, without control for estimation problems, such as unobserved intergenerationally-correlated endowments, if causality is of interest. The few exceptions are relatively recent studies that focus on high-income countries (HICs), with their much different contexts than the low- and middle-income countries (LMICs) in which the vast majority of children globally are growing up. This paper estimates the causal (conditional on the assumptions for the model) relationships between parents’ schooling and their children's schooling in the most populous LMIC, using adult identical (monozygotic, MZ) twins data from urban China. Our ordinary least-squares estimates show that one-year increases in maternal and paternal schooling are associated, respectively, with 0.4 and 0.5 more years of children's schooling. However, if we control for genetic and other endowment effects by using within-MZ fixed effects, the results indicate that mothers''and fathers’ schooling have no significant effects on children's schooling. Our main results remain with various robustness checks, including controlling for measurement error. These results suggest that the positive associations between children's and parents’ schooling in standard cross-sectional estimates in this major LMIC are mainly due to the correlation between parents’ unobserved endowments and their schooling and not the effects of their schooling per se.