Job tasks and wages across the world: Evidence from PIAAC

This paper addresses for the first time differences on job tasks across a harmonized sample of 20 countries across the world by using very precise information on job contents at the worker level. In addition, the relationship between job contents and individual and job characteristics as well as with computer use at work, both across and within occupations, is assessed. Then, we estimate the empirical relationship between job tasks and wages, assuming that workers self-select into occupations based on comparative advantage.

Our results indicate, first, that New Zealand and Great Britain display the largest values in non-routine cognitive and personal tasks, whereas routine contents are found to be highest in Lithuania and Turkey. Second, computer use at work is a key factor in explaining differences in job contents, both between and within occupations. Finally, job tasks exert substantial predictive power for wage differences across and within occupations. Our individual-based job tasks measures are validated by comparing them with those previously used with O*NET database.