Analysing the effect of training on wages – using combined survey‐register data

This article analysis the impact of training on wages using combined survey‐register material. The objective is threefold: to combine survey and register information to estimate individual return from training on wages; to estimate the impact from both individual and firm level training on individual wages; and to confront the results with control for unobserved individual and firm level heterogeneity. Results using register data show that the return from individual training is approximately 3.1 per cent per year. The return from the average training level at the firm is also positive and significant. After control for unobservable individual and firm level heterogeneity a positive impact from individual training is found, but evidence of a positive relationship between the mean level of training at the firm level and individual wages is no longer found. Our interpretation of these findings is that training has a positive individual effect, but has no positive firm‐specific external effects on other workers at the plant.