This article clarifies the empirical source of the debate on the effect of technology shocks on hours worked. We find that the contrasting conclusions from levels and differenced vector autoregression specifications, documented in the literature, can …