The size of the union membership wage premium in Britain's private sector

The paper estimates the union wage premium in Britain’s private sector in 1998, after nearly two decades of union decline. It examines the performance of the linear estimator alongside a semi-parametric technique (propensity score matching (PSM)) – hitherto unused in the wage premium literature - which shares the same identifying assumption, namely that selection into membership is captured with observable data. Results using the two techniques are compared, and reasons for differences in results are identified and discussed. By altering the information set entering estimation the paper shows the sensitivity of OLS and PSM results to data quality.