Taking a Full Career Perspective on the Formation of Co-Founding Teams

Entrepreneurs often find co-founders among past colleagues. While previous work focused on the knowledge that entrepreneurs may transfer from their shared prior employer to the new venture, , this paper takes on a full career perspective to explore more comprehensively how shared affiliations affect the selection of co-founders. Using Crunchbase and LinkedIn data, we construct a detailed panel of 712 solo and team founders of innovative startups in Belgium from 1970 to 2019, including harmonized indicators of their prior affiliations (schools, firms) and expertise (study fields, job functions). We estimate logit models for the odds of co-founding for 379 actual co-founder pairs matched to a control group of counterfactual co-founders, accounting for each entrepreneur’s unobserved preference for founding solo. Our results confirm the shared employer effect on the odds of co-founding from prior research but show that it interacts negatively with shared schooling backgrounds. Drawing on theory of team formation strategies, i.e., resource-seeking and interpersonal attraction motives, we argue that co-founders weigh the complementarity of the network of their co-founder more heavily than interpersonal relations stemming from overlapping study and career trajectories. We rule out other explanations by controlling for other resource complementarities between the co-founders, for which we find they are not conducive to team formation, on the contrary. Our findings show the value of looking back further than only the recent shared work experience of entrepreneurs, contributing to the literature on team formation in new ventures.