Social Networks and Labor Market Outcomes: Occupation Matters

We study how the influence of social networks on individual labor market outcomes
varies across occupations, specifically between manual and cognitive jobs. Using data
from over fourteen million Brazilian workers and exploiting exogenous job termination
due to mass layoffs, we confirm that social networks reduce unemployment duration and
increase wages in the new job, but show that these effects are heterogeneous depending
on workers’ occupations at the time of displacement. Manual workers benefit more from
networks in terms of job reentry but less in terms of wages compared to workers performing
cognitive tasks. We argue that these different patterns are due to the fact that networks
reduce the likelihood that manual workers find new jobs in the same occupation, given that
occupational change is associated with reductions in wages.