39 Pages Posted: 15 Jun 2017
Date Written: June 14, 2017
Research on geography and knowledge spillovers is premised on the proposition that proximity reduces the cost of search and coordination. Thus, learning from proximate parties is easier than from more distant ones. As a consequence, nearby individuals, teams, and firms share overlapping knowledge and correlated outcomes. In this paper we theorize that spatial spillovers fundamentally depend on the presence of existing relationships. Using multi-dimensional network formation data from the random placement of teams at a startup bootcamp, we show that spatial spillovers decline if team members have existing ties within a particular social setting. For teams with preexisting ties within the bootcamp, localized spillovers appear small or non-existent. For teams without preexisting ties we find that outcomes improve if neighbors are high performing, but that outcomes worsen if neighboring teams are low performing. Our findings suggest that existing relationships do affect spillovers, primarily by capping downsides, but also by limiting the upsides of being near a high-performing team.
Keywords: social networks, geography, entrepreneurship, field experiment
Suggested Citation: Suggested Citation
Hasan, Sharique and Koning, Rembrand, When Proximity May Not Be Destiny: The Role of Existing Relationships (June 14, 2017). Harvard Business School Strategy Unit Working Paper . Available at SSRN: https://ssrn.com/abstract=2964215