Members of the NBER's Entrepreneurship Working Group met in Cambridge December 7. Research Associates Josh Lerner of Harvard University and David T. Robinson of Duke University organized the meeting, which was sponsored by the Ewing Marion Kauffman Foundation. These researchers' papers were presented and discussed:
Bilal Zia, World Bank, and Patricio Dalton, Julius Rüschenpöhler, and Burak Uras, Tilburg University
Learning Business Practices from Peers: Experimental Evidence from Small-scale Retailers in an Emerging Market
Lack of growth aspirations can be an important psychological constraint to small-scale entrepreneurship in developing economies. Zia, Dalton, Rüschenpöhler, and Uras use a field experiment among urban retail entrepreneurs in Indonesia to provide an exogenous shock to their aspirations window. A randomly selected set of entrepreneurs are provided a handbook of key business practices implemented by successful local peers, complemented with two psychological and implementation nudges: A movie with business role models from the local area who demonstrate their path to success, and personalized business assistance to help entrepreneurs with individual implementation challenges. The researchers show that business growth aspirations respond strongly to these interventions, measured up to eighteen months afterwards. In line with the theoretical literature, the researchers find that the distance between the initial aspirations of the entrepreneurs and the business frontier is the key driver for the direction of the effects. Entrepreneurs with high business aspirations at baseline respond positively to the treatments and increase business aspirations, sales, and profits, while those with low initial aspirations respond negatively and decrease their business aspirations and performance. Zia, Dalton, Rüschenpöhler, and Uras find similar heterogeneity in complementary aspirations for children's education and satisfaction with household finances. These results confirm that initial levels of aspirations are crucial in determining how entrepreneurs respond to exogenous aspirational shocks.
Robert W. Fairlie, University of California, Santa Cruz and NBER, and Javier Miranda and Nikolas Zolas, U.S. Census Bureau
Job Creation and Survival among Entrepreneurs: Evidence from the Universe of U.S. Startups
Entrepreneurship is promoted around the world by governments and policymakers, but surprisingly, we do not know the answers to three fundamental questions important for basic welfare calculations: i) How many jobs does an entrepreneur create?, ii) Do these jobs disappear quickly?, and iii) How many entrepreneurs survive each year after startup? Fairlie, Miranda, and Zolas provide the first definitive answers to these three questions using a new compilation of administrative that captures the universe of U.S. startups. The average entrepreneur creates 0.56 jobs at startup (using the broadest definition possible). These jobs do not disappear quickly as the average startup employs 0.53 workers five years later. In total, the average annual cohort of 5.4 million startups in the United States creates 3.0 million jobs in the startup year and employs 2.9 million workers five years later. Without these jobs, net job creation by all other businesses would be negative. The researchers also find extremely low survival rates -- only 41 percent of startups survive two years and only 20 percent survive five years, with no industries being immune. Exploring the rich heterogeneity in the early-stage dynamics of startups, the researchers find that the negative influence from the high exit rates on job creation is mostly offset by rapid job growth over time among surviving startups. This rapid growth in job creation among survivors is not driven by a few extremely successful startups, but instead is driven by a continuous upward shift in the employment size distribution. Exploring heterogeneity across startup types, the researchers find that nonemployer startups make sizeable contributions to employment several years after startup (22 percent of jobs seven years later). Instead of excluding all non-employer startups, the researchers exclude sole proprietor non-employer startups to create a more restrictive definition of entrepreneurship. Using this upper bound measure, the study finds a job creation rate of 2.0 jobs per startup and a survival rate of 30 percent five years later. These findings have important policy, theoretical and welfare implications.
Jesse Davis and Xinxin Wang, University of North Carolina, Chapel Hill, and Adair Morse, University of California, Berkeley and NBER
The Leveraging of Silicon Valley
Venture debt is now observed in 28-40% of venture financings. Davis, Morse, and Wang model and document how this early-stage leveraging can affect firm outcomes. In the researchers' model, a venture capitalist maximizes firm value through financing. An equity-holding entrepreneur chooses how much risk to take, trading off the financial benefit against their preference for continuation. By extending the runway, utilizing venture debt can reduce dilution, thereby aligning the entrepreneur's incentives with the firm's. The resultant risk-taking increases firm value, but the leverage puts the startup at greater risk of failure. Empirically, the researchers show that early-stage ventures take on venture debt when it is optimal to delay financing: such firms face higher potential dilution and exhibit lower pre-money valuations. Consistent with this notion, such firms take eighty-two fewer days between financing events. This strategy induces higher failure rates: $125,000 more venture debt predicts 6% higher closures. However, conditional on survival, venture debt-backed firms have 7-10% higher acquisition rates. The researchers highlight the role of leverage in the risking-up of early-stage startup firms. Aggregation of these tradeoffs is important for understanding venture debt's role in the real economy.
Steven Nafziger, Williams College, and Amanda G. Gregg, Middlebury College
The Births, Lives, and Deaths of Corporations in Late Imperial Russia
Understanding the birth, growth, and death of firms in the early stages of industrial development is a relatively unexplored area of economic history, yet these processes are at the heart of transitions to modern economic growth. Nafziger and Gregg investigate the competitiveness and financial development of the Imperial Russian economy by examining patterns of entry and exit into the corporate sector. Their analysis relies on a newly developed panel database of detailed annual balance sheet information from every active corporation in the Russian Empire between 1899 and 1914. In the data, firms entered the corporate sector as new firms or partnerships newly transformed into corporations and exited when they shut down. The researchers examine the industrial distribution of firm entry, document how new and newly transformed corporations evolved over their life cycles, and construct a proportional hazard model that includes balance sheet and governance characteristics to predict firm exits.
Yael Hochberg, Rice University and NBER; John M. Barrios, University of Chicago; and Livia Hanyi Yi, Rice University
The Cost of Convenience: Ridesharing and Traffic Fatalities
Hochberg, Barrios, and Yi examine the effect of the introduction of ridesharing services in U.S. cities on fatal traffic accidents. The arrival of ridesharing is associated with an increase of 2-3% in the number of motor vehicle fatalities and fatal accidents. This increase is not only for vehicle occupants, but also for pedestrians. The researchers propose a simple conceptual model to explain the effects of ridesharing's introduction on accident rates. Consistent with the notion that ridesharing increases congestion and road utilization, the researchers find that the introduction of ridesharing is associated with an increase in arterial vehicle miles traveled, excess gas consumption, and annual hours of delay in traffic. On the extensive margin, ridesharing arrival is also associated with an increase in new car registrations. These effects are higher in cities with higher ex-ante use of public transportation and carpools, consistent with a substitution effect, and in larger cities and cities with high ex-ante vehicle ownership. The increase in accidents appears to persist (and even increase) over time. Back-of-the-envelope estimates of the annual cost in human lives range from $5.33 billion to $13.24 billion per year.
Christopher Geczy and David Musto, University of Pennsylvania; Jessica Jeffers, University of Chicago; and Anne M. Tucker, Georgia State University
Contracts with Benefits: The Implementation of Impact Investing
Impact investing private equity and venture capital funds are a rapidly emerging force in capital markets, premised on the service of two goals at once: a financial goal as well as a social-benefit goal. The addition of this second objective complicates the already challenging problem of aligning incentives across layers of agency, and raises the question of how contracting practices should adapt. Geczy, Jeffers, Musto, and Tucker draw on contract theory and a unique set of legal documents from impact funds to answer this both normatively and positively. Contracts struck by impact funds, both forward to portfolio companies and back to investors, use new terms to directly operationalize impact, and also adjust the use of existing terms on governance, investor protection, and other concerns to facilitate it. Moreover, funds' direct contracting on impact with investors passes through to their contracting with portfolio companies. For the most part, observed contracting terms align with theory, though they also differ in interesting ways, such as on compensation and covenants. Finally, the researchers find evidence that different forms of contracting serve complementary roles in supporting impact.
Kevin Boudreau, Northeastern University and NBER
In many instances, platforms have "crowds" of amateurs working on them as complementors, in other cases professional entrepreneurs or a mix of both. What precisely can a platform owner do to implement these outcomes? Boudreau clarifies conditions shaping participation by amateurs and professionals, and the role played by platform design. Key empirical predictions are shown to be consistent in the analysis of data on app developers. Small incremental shifts in the bare minimum cost necessary to develop a minimum viable product lead the bottom to fall out to amateurs. This more than doubles the total number of complementors and produces a flood of lowest-quality complementary goods — leading to a "long tail" that is considerably longer than that predicted by standard theory. The added developers are longer-lived, thus tending to accumulate on the platform, despite their low quality. Nonetheless, in these data, overall product quality increases where the bottom falls out to amateurs. Boudreau discusses possible interpretations.