Abstract
This paper presents the Startup Cartography Project, which offers a new set of entrepreneurial ecosystem statistics for the United States from 1988-2016. The SCP combines state-level business registration records with a predictive analytics approach to estimate the probability of “extreme” growth (IPO or high-value acquisition) at or near the time of founding for all newly-registered firms in a given year. The results indicate the ability of predictive analytics to identify high-potential start-ups at founding (using a variety of different approaches and measures). The SCP then leverages estimates of entrepreneurial quality to develop four entrepreneurial ecosystem statistics, including the rate of start-up formation, average entrepreneurial quality, the quality-adjusted quantity of entrepreneurship, and entrepreneurial ecosystem performance over time. These statistics offer sharp insight into patterns of regional entrepreneurship, their correlation with regional economic growth and the evolution of entrepreneurial ecosystems over time. The SCP includes both a public-access dataset at the state, MSA, county, and zip code level, as well as an interactive map, the U.S. Startup Map, that allows academic and policy users to assess entrepreneurial ecosystems at an arbitrary level of granularity (from the level of states down to individual street addresses).