This paper studies differences in the labor markets for occupations with different automation risk, and how actual automation may induce changes in wages and employment. Using data from Colombia between 2009 and 2017, we compute wage disparities by automation risk. We find that 62% of the occupied people in Colombia are at high-risk of automation. In the same way, we find that 71% of informal workers are at high-risk, while 56% of formal workers are. The wage return to education is highest in the less automatable occupations. We then look at the effects of actual education, measured by ICT investment. On wages and employment, automation increases employment, decreases wages and the wages gap by skill. Education acts as a protection mechanism against new automation technologies.
