Impact of the Corrections to the Monthly Labor Survey Data on Productivity Analysis (Japanese)

This study investigates the impact of recent correction to the Monthly Labor Survey on the measurement of productivity analysis at the firm-level. According to the analysis, first, quantitative impact of the correction on measured productivity is insignificant on average, but relatively large revisions are found for a very small number of firms. Second, measurement errors of productivity growth rate resulting from combining the uncorrected years' data with the corrected years' data are sometimes larger than those obtained solely from the uncorrected data. Third, according to our analyses, the qualitative results of past studies on the relationship between firm characteristics and productivity using the uncorrected series of working hours taken from the Monthly Labor Survey are unlikely to be overturned, although it is hard to draw definitive conclusions. In light of the importance of policies in improving productivity, correction to the Monthly Labor Survey data before 2011 is highly expected to facilitate evidence-based policy making.