Estimating the intergenerational correlation of incomes: an errors in variables framework

The estimation of the intergenerational correlation of incomes is usually carried out by proxying permanent incomes using suitable indicators of economic status, and by treating the resulting measurement error problem using averaging or instrumenting procedures. Here we take the permanent income of the parents' family to be unobserved, but we assume that its determinants are known to the researcher. A two-stage procedure as well as a MIMIC type covariance estimator applied to a US sample of parents and children entail estimates of the order of 0.61 to 0.64 for the coefficient of intergenerational income transmission. OLS estimates this parameter at 0.5. The variance ratio of permanent to total income is also estimated to be in the range of 0.77 to 0.8, implying a correction factor of 1.25 to 1.3 for OLS estimates.