by Serge Coulombe and Jean-François Tremblay.
This paper presents an empirical analysis on the determinants of aggregate levels of training across fourteen OECD countries. Training data comes from the 1994 International Adult Literacy Survey (IALS), which provides highly comparable cross-country data on the percentage of employed individuals that received job-related training. We use a panel data structure to explain the levels of training across country and age groups from the average literacy skills of the corresponding population subset and other cross-country variables, including indicators of compression in the wage structure, the rate of unionization, the unemployment rate, the level of innovation activity and measures of industrial structure. We find that the average level of literacy skills in each age-groups has a positive and highly significant effect on the proportion of workers that receive training, which is consistent with microeconomic evidence on the effect of educational attainment on the probability that workers participate in training. More importantly, our analysis shows that compression at the bottom of the wage distribution increases training but compression at the top has the opposite effect. These effects are robust and highly significant across gender and for different age-group samples. Potential policy implications are discussed.