Relevant ties matter: The Causal Effect of Work Ties on Unemployment Duration

What is the effect of social networks on labor market outcomes? The tradition following Grannovetter

claims that social ties have a positive effect on wages, job prestige and job satisfaction. Although a number

of studies have found positive correlations between friendship characteristics and labour market status, there

is still disagreement on whether or not the effect is causal. One objection is that the correlations could be

caused by social homophily--the fact that similar people tend to become friends with each other. This paper

uses unique labor market data from Denmark to test for the effect of a very specific form of social capital--

the number of former work colleagues--on respondents' unemployment duration. To address the possible

role of social homophily and reverse causality, we use an indirect test of causality proposed by Mouw (2003)

based on the probability of using contacts to find work. Results indicate that the number of industry-specific

colleagues is both associated higher employment rates with the probability of using contacts to find work.

We argue that this is consistent with a causal role of social networks in the labor market, as former

colleagues provide information about job openings.