Financing new ideas : evidence from Kickstarter campaigns

In this dissertation, I study the crowdfunding market and the importance of access to financing for early stage entrepreneurs. Using a novel and rich dataset of crowdfunding projects from Kickstarter, the world's largest crowdfunding platform, I begin by showing that entrepreneurs who launch campaigns on days when many other entrepreneurs launch campaigns are less likely to reach their campaign funding goals and be funded. My evidence suggests that three main factors are responsible for this relationship: 1) capital supply, 2) capital allocation, and 3) investor limited attention. First, while the supply of capital increases with the number of campaigns launched, this increase is not enough to match the increase in the demand for capital. Second, crowdfunding campaigns that are very likely to succeed based on ex-ante characteristics are able to raise more financing at the expense of less promising campaigns. Third, this allocation is partly explained by investors' limited attention, which seem to cluster towards certain campaigns when their ability to screen projects suffers as the number of campaigns launched increases.

In the second part of the dissertation, I exploit the source of luck-driven variation in funding outcome introduced by the number of campaigns launched to examine how access to external financing affects entrepreneurial success and future entrepreneurial activity. Using machine learning to match crowdfunding campaigns with a self-collected dataset of Amazon searches, I estimate a significant causal effect of crowdfunding on the likelihood that an entrepreneur commercializes their subject idea, becomes a serial entrepreneur, and successfully raises capital for future projects. My findings suggest that limited attention and capital supply play an important role in determining which projects receive crowdfunding and that funding success has a causal effect on different entrepreneurial outcomes.