What's the Value of Social Capital? A Within-Person Job Attributes, Offer and Choice Test

Abstract

Whether and how social capital derived through social ties creates value has animated substantial research in management, sociology, economics, and political science. Scholars are now generally inclined to believe that social ties have an impact on labor market outcomes — that is, it is “who you know” that counts. Two mechanisms were initially identified to explain why this is the case: access to a better distribution of offers in terms of pay or non-redundant information that facilitates job offer. The evidence informing each theory, however, was often inconsistent and circumstantial. In this research we propose and test a unified model of these two theoretical approaches. We do so leveraging unique data on contemporaneous job search with detailed information about job search channels and offer characteristics in alternative-specific job attributes and choice models. We find that contrary to conventional wisdom, search through social networks typically results in job offers with lower total compensation (e.g., -18.0% for those derived through referrals from family and friends v. formal recruiting, -15.4% for those derived via alumni referrals). However, offers derived through weak bridging ties such as alumni contacts are more likely to be accepted because they are perceived as having better growth prospects, and students are willing to accept a significant total compensation discount to accept offers derived via school faculty and students they did not know prior to matriculation because they afford access to industries that are otherwise difficult to access. On balance, our tests are consistent with Granovetter’s argument that the value networks afford is information about opportunities that are otherwise difficult to access rather than a better distribution of offers in terms of pay.

Keywords: Social capital, networks, job search, hiring, MBA

JEL Classification: J00, J24, J31, J44, M55