Estimation of the distribution of hourly pay from household survey data: The use of missing data methods to handle measurement error

Measurement errors in survey data on hourly pay may lead to serious upward bias in low pay estimates. We consider how to correct for this bias when auxiliary accurately measured data are available for a subsample. An application to the UK Labour Force Survey is described. The use of fractional imputation, nearest neighbour imputation, predictive mean matching and propensity score weighting are considered. Properties of point estimators are compared both theoretically and by simulation. A fractional predictive mean matching imputation approach is advocated. It performs similarly to propensity score weighting, but displays slight advantages of robustness and efficiency.