Networks and Growth in India's Startup Ecosystem

Regional clusters, like Silicon Valley, are the engines of economic growth. Emerging economies from Chile to China have devoted resources to building their own innovation ecosystems. Yet the efficacy of these strategies remains unclear (Chatterji et al. 2013). In particular, researchers have argued that the true source of an ecosystem’s success is the existence of rich networks that connect people and firms (Saxenian 1996). Despite the multitude of reasons to believe that networks promote growth, the mechanisms remain under-theorized and poorly tested, especially at the inter-firm level. This is primarily because existing research in this area has been exploratory, producing limited understanding of the causal mechanisms underlying network effects even in developed contexts. On the other hand, the experimental literature on the functioning of inter-personal networks in both developed and developing countries has grown substantially in recent years. Much progress has been made in understanding how social interactions among individuals, affect network formation across different types of ties (Hasan and Bagde 2015), individual learning (Hasan and Bagde 2013), and the demand for co-founders (Koning and Hasan 2015). We are proposing a study that builds on the progress of recent experimental studies of inter-personal networks to study inter-firm networks among growing startups in India’s product ecosystem. In collaboration with iSPIRT, a leading Indian non-profit working to encourage entrepreneurship in the software product sector in India we will conduct a study of how networks shape firm learning, strategy, hiring, and other outcomes. The basic structure of the study will consist of a pre-treatment data collection effort where we will collect detailed firm-level data on a group of 150-200 startups in India’s ecosystem. Next, we will run a field experiment in collaboration with iSPIRT where we will introduce random variation into the social networks among the founders and CEOs of the startups in our sample. Finally, we will measure how our network randomizations affect whether formal arrangements between firms emerge, whether and what firms learn from each other, and whether variation in the connections that firms have to each other affect the outcomes of these startups. The study will provide a novel experimental method for understanding the impact of network connections on the outcomes of organizations.