Following an influential article by Angrist and Krueger (1992) on two-sample instrumental variables (TSIV) estimation, numerous empirical researchers have applied a computationally convenient two-sample two-stage least squares (TS2SLS) variant of Angrist and Krueger's estimator. In the two-sample context, unlike the single-sample situation, the IV and 2SLS estimators are numerically distinct. We derive and compare the asymptotic distributions of the two estimators and find that the commonly used TS2SLS estimator is more asymptotically efficient than the TSIV estimator. We also resolve some confusion in the literature about how to estimate standard errors for the TS2SLS estimator.

Previous studies of recent U.S. trends in intergenerational income mobility have produced widely varying results, partly because of large sampling errors. By making more efficient use of the available information in the Panel Study of Income Dynamics, we generate more reliable estimates of the recent time series variation in intergenerational mobility. Our results, which pertain to the cohorts born between 1952 and 1975, do not reveal major changes in intergenerational mobility.

This note examines the implications of mean-reverting mea-surement error for two influential literatures based on longitudinal survey data: (1) the literature on real wage variation over the business cycle and (2) the literature on intertemporal substitution in labor supply. Accounting for mean-reverting measurement error suggests that real wages may be even more procyclical than indicated by recent longitudinal studies. We also find that the instrumental variables estimator commonly used in intertemporal substitution studies is inconsistent if changes in earnings and hours of work are measured with different degrees of mean reversion, but the magnitude of the resulting inconsistency appears to be small.

This study proposes using correlations between neighboring children in their later socioeconomic status to bound the proportion of inequality in socioeconomic outcomes that can be attributed to disparities in neighborhood background. We apply this approach to educational attainment data from the Panel Study of Income Dynamics, which has sampled neighboring children and followed them into adulthood. We find that, once the effects of a few readily observed family background characteristics are accounted for, the correlation between neighboring children in their eventual educational attainment is only about 0.1. Given that even this figure is inflated by neighbors' similarity in unmeasured aspects of family background, the results suggest a limited role for neighborhood factors in accounting for inequality in educational attainment.