Industrial agglomeration and entrepreneurship

Edward Glaeser, William Kerr, 26 November 2008

a

A

Many academics, policy makers, and business leaders stress the importance of local conditions for explaining spatial differences in entrepreneurship and economic development. Behind the rhetoric, however, are many unresolved questions. Why do the plants of many industries group so tightly together? How do these groupings affect the location decisions of entering start-ups? Which industrial structures make entrepreneurship more likely to happen in a city? Everyone wants to foster the next Silicon Valley, but what is it about Silicon Valley’s industrial structure that is so attractive for new entry? Our recent research employs plant-level data from the Census Bureau to quantify these agglomeration forces generally and then specifically for entrepreneurship.

Determinants of industrial agglomeration

Alfred Marshall (1920) emphasised three different types of transport costs – the costs of moving goods, people, and ideas – that could be reduced by industrial agglomeration. First, he argued that firms would locate near suppliers or customers to save shipping costs. Second, he developed a theory of labour market pooling to explain clustering that takes advantage of scale economies associated with a larger pool of workers and firms. Finally, he began the theory of intellectual spillovers by arguing that in agglomerations, "the mysteries of the trade become no mystery, but are, as it were, in the air.” Firms, such as those described by Saxenian (1994) in Silicon Valley, locate near one another to learn and to speed their rate of innovation.

Although each of these determinants certainly contributes to agglomeration in some industries, it is challenging to assess their relative importance using data on which industries are agglomerated. Each Marshallian theory predicts that the same thing will happen for similar reasons: plants will locate near other plants in the same industry because there is a benefit to locating near plants that share some characteristic. One approach to this problem, pioneered by Audretsch and Feldman (1996) and Rosenthal and Strange (2001), is to examine cross-industry variation in the degree of agglomeration, such as regressing the degree to which an industry is agglomerated on the importance of R&D to the industry.

Our approach in Ellison, Glaeser, and Kerr (2007) instead takes advantage of coagglomeration patterns to tackle this problem. Plants are similar to the other plants in their industry in many dimensions. But when we look across industries, plants are similar in some dimensions and not in others. For example, some industry pairs exchange goods but employ very different workers. Other industries hire similar workers but never trade with each other. Hence, one can gain insight into which theories are more important by looking at which similarities are most predictive of whether industry pairs are coagglomerated. We also model shared dependencies for natural advantages that are not distributed evenly (e.g., coastal access, energy prices).

We measure the coagglomeration of plants through both the discrete index of Ellison and Glaeser (1997) and an approximation of the continuous metric of Duranton and Overman (2005). We find evidence to support the importance of all three Marshallian theories. Coagglomeration arising through shared natural advantages is estimated to be more important than any single Marshallian factor, but not as important as the cumulative effect of the three Marshallian factors. Input/output relationships have the largest effect of the Marshallian factors we consider, which is striking given the remarkable decline of transportation costs over the 20th century. Customer-supplier relationships are closely followed by similar labour needs. Our proxies for intellectual spillovers are weaker, but still economically and statistically important.

A concern is that plants of two coagglomerated industries may start employing similar types of labour as a consequence of their collocation, rather than clustering together due to similar initial labour needs. We develop two instruments to guard against reverse causality. The first employs industrial traits in the UK; the second calculates how similar two industries are through plants of each industry that are located where the other industry is mostly absent. These techniques yield comparable results.

Local industrial conditions and entrepreneurship

In Glaeser and Kerr (2008), we apply this lens to an analysis of how local industrial structures sponsor the entry of new firms. Some places, like Silicon Valley, seem almost magically entrepreneurial with a new start-up on every street corner. Other areas, like declining cities of the Rust Belt, appear equally starved of whatever local attributes make entrepreneurship more likely. We assess how much of the spatial distribution of manufacturing entrepreneurship can be explained through industrial structures and related theories of entrepreneurship.

We start with an application of our work on Marshall’s agglomeration economies. We summarise how the composition of a city’s industries provides needed goods, people, or ideas to new entrants. This greatly reduces the dimensions of local industrial structures that we must model. We also examine the Chinitz (1961) hypothesis. Chinitz argued that the presence of small, independent suppliers was particularly crucial for understanding why New York was so much more entrepreneurial than Pittsburgh. We test the Chinitz hypothesis by looking at whether new entry is more common when the relevant suppliers in a city are smaller in size.

Of the metrics we test, industrial structures consistently have the strongest predictive power for entry. Input-output relationships are again very important, but particularly through the relationship that Chinitz suggested. Small suppliers predict new entrants, while general proximity to suppliers or customers is less important. We also find that the presence of industries that use the same type of labour is robustly important. Looking across the entry size distribution, we find that Chinitz factors are most important for smaller entrants, while larger entrants more equally weight general input conditions. Technology and idea sharing also appear most important for smaller start-ups. Labour mix theories, on the other hand, receive equal emphasis throughout entrant size categories.

Beyond industrial structures, we find less evidence to support the importance of local demographics for explaining entry beyond the specific labour pooling economies modelled above. In contrast to self-employment metrics (e.g., Glaeser 2007), manufacturing start-ups are not more common in places with older or better-educated citizens. This makes sense given the scale and investment required, as well as the use of self-employment by some older workers as a transition to retirement. Entry is, however, higher in cities with more workers between 20 and 40 years old. We also find little evidence for a "culture" of entrepreneurship or Jacobs (1970) industrial diversity arguments. These findings hold when contrasting entrepreneurship with facility expansions by existing firms (e.g., Kerr and Nanda 2008).

While the correlation between local industrial conditions and entry is impressive, it is certainly possible that firms and industries cluster in cities in anticipation of large amounts of entry. To partially address these endogeneity concerns, we return to the natural cost advantages modelled in our earlier work. Following Ellison and Glaeser (1999), predicted city-industry employment shares are developed by interacting 16 local cost advantages and natural resources with factor intensities of industries in a non-linear least squares framework. This predicted spatial distribution of employment also predicts entrepreneurship well, highlighting the importance of basic cost considerations and natural advantages for explaining entry patterns. We then use these cost measures to predict Marshallian agglomeration economies by city-industry. We again find substantial evidence for the labour pooling and output markets rationales, while results for inputs and technology spillovers are not robust across specifications.

All told, our metrics of industrial structure explain between 60% and 80% of the spatial structure of manufacturing entrepreneurship. Much of this explanatory power comes through existing industry agglomerations and natural cost advantages. Many of the powerful forces identified by Marshall (1920) remain very relevant today. As manufacturing is a decreasing share of US employment, future research needs to explore other industrial sectors, too.

References

Audretsch, David, and Maryann Feldman, "R&D Spillovers and the Geography of Innovation and Production", American Economic Review 86 (1996), 630-640.
Chinitz, Benjamin, "Contrasts in Agglomeration: New York and Pittsburgh", American Economic Review 51:2 (1961), 279-289.
Duranton, Gilles, and Henry Overman, "Testing for Localization Using Micro-Geographic Data", Review of Economic Studies 72 (2005), 1077-1106.
Ellison, Glenn, and Edward Glaeser, "Geographic Concentration in U.S. Manufacturing Industries: A Dartboard Approach", Journal of Political Economy 105 (1997), 889-927.
Ellison, Glenn, and Edward Glaeser, "The Geographic Concentration of Industry: Does Natural Advantage Explain Agglomeration?", American Economic Review Papers and Proceedings 89 (1999), 311-316.
Ellison, Glenn, Edward Glaeser, and William Kerr, "What Causes Industry Agglomeration? Evidence from Coagglomeration Patterns", NBER Working Paper 13068 (2007).
Glaeser, Edward, "Entrepreneurship and the City", NBER Working Paper 13551 (2007).
Glaeser, Edward, and William Kerr, "Local Industrial Conditions and Entrepreneurship: How Much of the Spatial Distribution Can We Explain?", NBER Working Paper 14407 (2008).
Jacobs, Jane, The Economy of Cities (New York, NY: Vintage Books, 1970).
Kerr, William, and Ramana Nanda, "Democratizing Entry: Banking Deregulations, Financing Constraints, and Entrepreneurship", HBS Working Paper 07-033 (2008).
Marshall, Alfred, Principles of Economics (London, UK: MacMillan and Co., 1920).
Rosenthal, Stuart, and William Strange, "The Determinants of Agglomeration", Journal of Urban Economics 50 (2001), 191-229.
Saxenian, AnnaLee, Regional Advantage: Culture and Competition in Silicon Valley and Route 128 (Cambridge, MA: Harvard University Press, 1994).

Topics: Productivity and Innovation
Tags: agglomeration, entrepreneurship, localisation economies, spatial concentration

Professor of Economics, Harvard University
Assistant Professor, Harvard Business School