Sibling similarities, differences and economic inequality

I use improved statistical approaches and much larger samples than previous studies to provide more robust estimates of the correlation in economic outcomes among siblings. A key finding is that more than half the variance in log wages among men is due to differences in family and community background. Slightly smaller estimates in the 0.45 to 0.5 range are found for earnings and family income. For women, the sibling correlation in family income is the same as that found for men. I estimate that the sibling correlation in years of schooling and AFQT test scores is higher than 0.6. In contrast, estimates for a variety of other non-economic outcomes (including physical attributes) are in the 0.2 to 0.4 range. Family and community influences are particularly important for those who start at the bottom of the income distribution. An analysis of the variance in outcomes within families, by quartiles of parent income provides a new set of facts that should inform theoretical models of family resource allocation. I also find that a large portion of the sibling correlation in some economic outcomes can be explained by observable characteristics.