Background:

The canonical model of the labor market (Acemoglu, 2002), with two skill groups performing two distinct and imperfectly substitutable tasks, has proven to be unable to explain some features of the labor market. We test whether there is enough evidence in favor of the polarization hypothesis on the Colombia labor market.

Methods:

We use occupations descriptions to build technology-use related tasks intensities, and match them to series of cross sections of household survey data in order to evaluate to what extent the reallocation of workers across occupations that require different tasks intensities are consistent with the Polarization hypothesis.

Results:

We find an increase of employments at the extremes of the wage or skill occupations distribution, particularly pronounced in the period when the availability of technologies increased, and their prices substantially dropped.

Conclusions:

Overall, our results provide support for the polarization hypothesis in Colombia.