Making the most of coresident data: Credible evidence on intergenerational mobility with sibling correlation
利用孟加拉国、智利等六国数据，发现同住样本中兄弟姐妹相关性（衡量代际流动性的指标）的偏差仅4.30%，远小于代际回归系数的10.25%，且能90%-95%正确排序国家，为研究者使用同住数据提供了依据。
Many available data sets are not used for estimating intergenerational mobility owing to concerns about sample truncation bias in coresident data. Using data from Bangladesh, Chile, Ghana, India, Indonesia, and Mexico, we report the first evidence that the bias in estimated sibling correlation, a broad measure of relative mobility, is small in coresident samples (4.30%), much smaller than that in intergenerational regression coefficient (10.25%). The low bias reflects offsetting effects of sample truncation on the numerator and denominator of the sibling correlation formula. Sibling correlation estimates from coresident samples preserve the correct cross-country ranking 90%–95% times. Our findings have far-reaching implications for researchers working on intergenerational mobility. • This paper estimates biases in intergenerational mobility measures in co-resident data in six countries. • We compare three measures of relative mobility, with a focus on sibling correlation. • Sibling correlation shows the lowest bias compared to Intergenerational Regression Coefficient (IGRC) and Intergenerational Correlation (IGC). • Bias in sibling correlation is lower for younger cohorts and in smaller families. • Sibling correlations in co-resident data provide 90%–95% correct cross-country ranking.