Mitigating Earnings Imputation Bias: Evidence from the CPS

Some participants in income and labor market surveys fail to report their earnings. If the procedure used to impute missing earnings is biased, then imputation bias is substituted for nonresponse bias. We use data on imputed and reported wages for the same workers, taken from the 1989 change in the CPS processing system, to distinguish and measure three biases in earnings equations: reporting bias, erasure bias and model bias. We find that, for white males, the effect of correcting for reporting bias dramatically increases the returns to education. In addition to stressing the importance of reporting bias in the pre-1989 data, our results indicate that failure to correct for nonparticipation is a serious and continuing problem in the current CPS processing system.