Cities and Entrepreneurship

May 1-2, 2009
Edward Glaeser, Stuart Rosenthal and William Strange, Organizers

Ajay K. Agrawal, University of Toronto and NBER, Iain M. Cockburn, Boston University and NBER, and Carlos Rosell, Department of Finance, Canada
Not Invented Here: Creative Myopia and Company Towns

Agrawal and his co-authors examine variation in the market structure of innovation across 72 of North America's most highly innovative locations. In 12 of these areas, innovative activity is particularly concentrated in a single large firm; the authors refer to such locations as "company towns." They find that inventors in these locations tend to draw disproportionately from their firm's own prior art relative to what would be expected given the underlying distribution of innovative activity across all inventing firms. Furthermore, they find that inventors in company towns are more likely to draw upon the same prior art year after year as opposed to those in more industrially diverse locations. They refer to this tendency as "creative myopia." However, even though a greater fraction of the impact from company town inventions is realized by the inventing firm itself, they find no evidence that inventions from this type of location have any less impact on subsequent innovation overall, nor do they find that the geographic scope of their influence is in any way diminished.

William Kerr, Harvard University
How Important Is Local Innovation for Entrepreneurship? An Assessment through US Scientific Immigration

Kerr investigates the extent to which immigrants facilitate the spatial reallocation of invention and related entrepreneurship across U.S. cities. A 10 percent increase in ethnic invention for a technology is associated with a 1 percent increase in the rate of spatial reallocation of invention for the technology across U.S. cities. The causal direction of this association is confirmed through regressions that interact national immigration trends with initial dependency on immigrants by technology. Similar results also are found when looking at an exogenous surge in Chinese and Indian scientific immigration after the Immigration Act of 1990.

Henry Chen, Harvard University, Paul Gompers, Harvard University and NBER, Anna Kovner, Federal Reserve Bank of New York, and Josh Lerner, Harvard University and NBER
Buy Local? The Geography of Successful and Unsuccessful Venture Capital Expansion

Chen, Gompers, Kovner, and Lerner document geographic concentration by both venture capital firms and venture capital financed companies in three elite cities: San Francisco, Boston, and New York. They find that firms open offices based on the success rate of venture capital-backed investments in an area. Geography is also significantly related to outcomes. Venture capital firms based in the elite locales outperform, regardless of the stage of the investment. Ironically, this outperformance arises from outsized performance outside of the venture capital firms' office locations, including in peripheral locations. Outperformance of non-local investments suggests that policymakers in regions without local venture capitalists might want to mitigate costs associated with established venture capitalists investing in their geographies, rather than encouraging the establishment of new firms.

Jed Kolko, Public Policy Institute of California, and David Neumark, UC, Irvine and NBER
Does Local Business Ownership Stabilize Employment?

Local economic policies sometimes favor the creation and growth of locally-owned businesses. Kolko and Neumark's goal in this paper is to assess one of the prominent arguments in favor of such policies. In particular, the argument is that locally-owned firms are more likely to internalize the costs to the community of decisions to reduce employment - including closing or relocating. This may stem from economic factors, such as the costs that decisionmakers bear affecting their own economic well-being via effects on their communities. It may also stem from loyalty toward the headquarters' hometown or a desire for better public relations in the headquarters' hometown, perhaps for political reasons. An empirical implication of this argument is that, when faced with an unfavorable "shock" that reduces the demand for labor, firms are less likely to reduce employment in the areas in which they are headquartered. The same would hold true for single-establishment firms relative to multiple-establishment firms with headquarters elsewhere. To assess this hypothesis, the authors examine how the establishment-level employment response to labor demand shocks is affected by establishment ownership. They compare the employment responses of single-establishment firms, headquarters of multi-establishment firms, establishments in locally-headquartered multi-establishment firms, and establishments in non-locally-headquartered multi-establishment firms to both industry shocks (nationally - that is, across all regions) and regional shocks (across all industries). They use data from the National Establishment Time Series, covering the entire United States for the period 1992-2006.

Olav Sorenson, University of Toronto, and Michael Dahl, University of Aalborg,
The migration of technical workers

Using panel data on the Danish population, Dahl and Sorenson explore the revealed preferences of scientists and engineers for the places in which they choose to work. The results indicate that these technical workers exhibit substantial sensitivity to differences in wages but that they have even stronger preferences for living close to family and friends. The magnitude of these preferences, moreover, suggests that the greater geographic mobility of scientists and engineers, relative to the population as a whole, stems from more pronounced variation across regions in the wages that they can expect. These results remain robust to estimation on a sample of individuals who must select new places of work for reasons unrelated to their preferences-those who had been employed at establishments that discontinued operations.

Mark Doms, Federal Reserve Bank of San Francisco, Alicia Robb, UC, Santa Cruz, and Ethan Lewis, Dartmouth College
Local Labor Market Endowments, New Business Characteristics, and Performance

It is often asserted that a highly educated workforce is vital to improving the competitive position of American businesses, especially by boosting entrepreneurship. To examine this contention, Doms and his co-authors use population Census data and a rich, new, nationally representative panel of startup firms, to examine how the education and skill level of the local labor force are related to the creation and success of new businesses. They find that areas that possess more skilled labor also possess higher rates of self-employment and more skilled entrepreneurs. As in previous studies, they find that education of the business owner is strongly linked to improved business outcomes. Potentially consistent with the popular view, they also find that, conditional on owner's education, higher education levels in the local market are positively correlated with improved business outcomes.

Edward Glaeser, Harvard University and NBER, Giacomo Ponzetto, Harvard University, and William Kerr, Harvard University
Geographic Amenities and the Agglomeration of Innovation Entrepreneurship

Employment growth is strongly predicted by the number of establishments per worker, both across metropolitan areas and across industries within metropolitan areas, but there is little consensus on why this relationship exists. Glaeser, Kerr, and Ponzetti present a simple model that formalizes several different explanations for this phenomenon: lower fixed costs in some sectors in some places lead to both smaller firms and more entrepreneurship; higher profit margins lead to more competition and more entrepreneurship; a greater supply of entrepreneurial human capital leads to more firms and more growth; and, some places and sectors have evolved independent suppliers who abet competition and entrepreneurship. The evidence on returns does not support the higher-profit-margins hypothesis, but all three other hypotheses receive some support in the data. Yet none of them can significantly explain the powerful correlation between firm size and later employment growth.

Steven Klepper, Carnegie Mellon University,
The Origin and Growth of Industry Clusters: The Making of Silicon Valley and Detroit

Klepper uses data for all producers of automobiles and integrated circuits on their origins, base location, and performance to analyze the factors behind the historical clustering of the two industries in Detroit and Silicon Valley respectively. He elaborates on key ideas concerning organizational reproduction and heredity and explains how spinoffs from incumbent firms in the same industry can lead to clustering. His findings concerning the spawning of spinoffs, entry by firms in related industries, and firm performance suggest that organizational reproduction and heredity were the primary forces underlying the clustering of the two industries.

John C. Haltiwanger, University of Maryland and NBER, Ron S. Jarmin, U.S. Census Bureau, and C. J. Krizan, Bureau of the Census
Mom-and-Pop Meet Big-Box: Complements or Substitutes?

Big-Box stores are popular shopping venues offering consumers expansive product lines at low prices. However, there is a popular perception that they displace smaller, often family owned (a.k.a. Mom-and-Pop) retail establishments. Several studies have examined the empirical evidence on the effects of Big-Boxes on local retail employment but no clear consensus has emerged. The questions remain: how frequently do Big-Box stores displace or substitute for single unit or small chain stores, and when, if ever is the presence of a Big-Box complementary to smaller retail stores? Haltiwanger and his co-authors argue that the county-level data used in prior analyses is too coarse to address these questions. They use retail establishment data from the Census Bureau's Longitudinal Business Database (LBD) to more directly measure the impact of changes in Big-Box activity on the activity of nearby single unit and smaller chain stores. The authors attempt to quantify the impact of Big-Box store entry and growth on nearby single unit and localized chain stores operating in both the same and different retail sectors. They incorporate a rich set of controls for local retail market conditions by measuring key demographic and income variables of the population within a few miles of the stores' locations. Their main finding is that there is a substantial negative impact of Big-Box entry and growth on employment growth at both single unit and especially smaller chain stores - but only when the Big-Box activity is in the immediate area and in the same detailed industry. A general pattern is that this impact declines with distance. That is, the impact tends to be the largest if the Big-Box activity is within 1 mile or 1 to 5 miles as opposed to 5 to 10 miles of the store in question. The impact of increased big-box activity manifests itself through a substantial reduction in net employment growth and smaller retailers which is mostly accounted for by an increase in job destruction from store exit. The authors find no systematic relationship between the growth and entry of Big-Box activity in other sectors on single unit and smaller retail chain stores.

Stuart Rosenthal, Syracuse University, and Amanda Ross, Syracuse University
Violent Crime, Entrepreneurship, and Vibrant Cities

Although numerous studies have examined the causes of urban crime, relatively few have considered the impact of crime on patterns of urban development. Rosenthal and Ross add to that small literature by assessing the degree to which violent crime discourages retail and high-end restaurants, establishments central to a vibrant urban area and nightlife. Their research design compares local retail to wholesale activity within individual industries, and also high-end to lower-tier restaurants. Differencing in this manner helps to strip away common threats from property crime along with the influence of other unobserved factors. Their models also control directly for city fixed effects and a host of census tract sociodemographic attributes, as well as local employment and population density. The results indicate that higher local violent crime rates depress retail employment relative to wholesale, and that the magnitude of the effect is noteworthy. For the thirteen industries examined, an increase in crime from the 10th to the 90th percentile would reduce retail employment relative to wholesale by roughly 35 percent. Analogous estimates based on a comparison of high- to low-end restaurants are nearly twice as large. These findings indicate that efforts to make distressed portions of cities more vibrant must give consideration to the need to ensure that such areas are safe.

NBER Videos

National Bureau of Economic Research, 1050 Massachusetts Ave., Cambridge, MA 02138; 617-868-3900; email:

Contact Us