Social Networks and the Co-Movement of Local Labor Markets: Evidence from Facebook Connections

Abstract

Are local labor markets more synchronized when they are socially connected? Using Facebook's Social Connectedness Index (SCI), which measures the intensity of social network ties between U.S. counties, I examine whether counties with stronger social connections to economically distressed areas experience correlated labor market outcomes. I find a robust positive relationship: a one-standard-deviation increase in network exposure to unemployment shocks is associated with a 0.28 percentage point larger own unemployment shock. This correlation persists after controlling for county characteristics and even with state fixed effects (coefficient attenuates to 0.14). However, the relationship loses statistical significance with state-clustered standard errors, and a leave-out-state exposure measure shows a negative coefficient, suggesting the correlation may reflect within-state spatial dependence rather than pure social network transmission. These findings highlight both the potential importance of social networks in transmitting economic conditions and the identification challenges inherent in separating social network effects from geographic proximity. \vspace{1em} JEL Codes: J64, R12, Z13 \\ Keywords: Social networks, labor markets, unemployment, Facebook, spatial correlation