The origins of entrepreneurship and innovation clusters

Aaron Chatterji, Edward Glaeser, William Kerr, 4 June 2013

a

A

For decades, the conventional wisdom among local officials pursuing employment growth was to attract a large firm to relocate. ‘Smokestack chasing’ often leads to zero-sum games where regional governments bid against each other to provide substantial incentives to large companies making location choice decisions (e.g. Greenstone et al. 2010).

The success of ‘entrepreneurial clusters’ in recent decades has challenged this wisdom. Now many policymakers state that they want their regions ‘to be the next Silicon Valley’. This new emphasis on bottom-up strategies has led to extensive efforts to seed local entrepreneurship (e.g. Lerner 2009). Today’s policymakers are especially eager to announce the launch of an entrepreneurial cluster in a hot industry, such as biotechnology, nanotechnology, or advanced manufacturing.

In recent research, we explore the rationale for and efficacy of policies to promote local entrepreneurship and innovation and reflect on recent initiatives in this domain (Chatterji et al. 2013).

Why the shift in attention to entrepreneurial clusters?

Our review begins with a conceptual introduction to industrial/innovation policy. We first discuss potential rationales for intervention – externalities, redistribution, and financing constraints – and the degree to which cluster strategies have followed these rationales. We then highlight how policymakers must choose the spatial scope of the intervention:

  • From neighbourhood to national.

And the degree to which it targets specific firms:

  • From general to industry-/sector-specific to firm-specific.

Entrepreneurial clusters occupy a particular spot in this framework. On one hand, they provide policymakers a tool for targeting a specific industry/sector that is deemed important, often either due to historical comparative advantages or to future growth prospects, but they keep policymakers out of the business of choosing specific firms to support. Moreover, while it is possible to have pro-entrepreneurship policies that are not tied to cluster-making, supporting a cluster of small-scale entrepreneurs allows policymakers to affect many entrepreneurs simultaneously, providing important scale to their policy interventions, and respecting the empirical tendency of economic activity to cluster.

Beyond the conceptual foundations, the move away from chasing smokestacks to fostering entrepreneurship is understandable given the strong correlation between small establishment size and local economic development. Figure 1 provides a representative graph from taken from Glaeser et al. (2012). The horizontal axis of each panel provides a potential way to measure entrepreneurship around 1982, and the vertical axis of each panel measures the employment growth of the city during 1982-2002. There is a very strong connection between initial entrepreneurship and subsequent employment growth.

Figure 1. City employment growth and initial entrepreneurship: Cross-sectional plots of urban growth 1982-2002 vs. initial traits

Notes: Taken from Glaeser, Kerr, and Kerr (2012). Figure presents city employment growth over the 1982-2002 period against measures of local entrepreneurship in 1982. Panels A and B use average establishment size, with places with smaller establishments associated with greater subsequent growth. Panels C and D use start-up entry rates, with places with greater start-up shares showing greater subsequent growth.

Discussion of this dates back to Chinitz’s 1961 paper. While striking, academic work is only beginning to provide a causal assessment of these features. Our review describes papers that have made the most progress towards empirical identification in the US context with respect to various forms of innovation and entrepreneurship. We have reasonable evidence at this stage for the causal role in local economic growth of venture capital investment, breakthrough inventions, and university spillovers. We are making progress towards the general entrepreneurship features shown in Figure 1, and we have the farthest to go with respect to innovation very broadly defined.

What can be done?

While the literature is making slow steps towards causal assessments of the links between local entrepreneurship and economic growth, the literature on spatial determinants of entrepreneurship is more established. We review factors that have been found important at the city level: education, age structure, local entrepreneurial culture, and physical infrastructure. We also consider evidence regarding industry linkages within cities, such that the local industrial composition favours start-ups in one sector over another, and the role of skilled immigrants for spatial variations in entrepreneurship for technical fields. Most of the drivers identified for the US appear to be shared with other countries.

Perhaps the strongest message provided to policymakers from this work on spatial determinants is the key importance of ‘supply-side’ factors for local entrepreneurship, especially among the local population that disproportionately constitute local entrepreneurs (Michelacci and Silva 2007). While local policymakers may already be predisposed to emphasise the importance of cultivating local entrepreneurs, the historical emphasis on smokestack chasing is practically the opposite extreme of an intervention (the firms are large, and the firms are imported from elsewhere). The recent literature emphasises caution about this latter approach. Other studies have also raised awareness of the extreme localisation of spillover benefits in cities and clusters (on dimensions ranging from spatial distance to genders of business owners), and this may suggest an important role for inclusionary criteria in intervention design or some alternative form of dissemination of benefits to ensure that positive spillovers are realised throughout the local area.

We close the review by exploring recent US policy initiatives at the federal and local level to encourage entrepreneurship and innovation. We highlight the variety of recent efforts to increase the local supply of entrepreneurs, especially through education initiatives, mentorship programmes, and increasing the availability of entrepreneurial finance. We also review the varied role of policy in the development of three well-known innovative clusters: Silicon Valley, Boston’s Route 128 corridor, and North Carolina’s Research Triangle Park. A common theme in many of these discussions is the importance of correct baseline business environments, or ‘setting the table’ activities, relative to targeted interventions (Porter 1990, Lerner 2009). Finally, we discuss the recent emphasis on promoting high-growth entrepreneurship at the federal level, long prevalent in Europe but becoming more important in the US.

Conclusion and call for experimentation

Even though entrepreneurship is a powerful force that engenders local and economic growth, it is not obvious that government policy can create entrepreneurship. Even if entrepreneurs naturally cluster in tight geographic units, it is not obvious that the government should use public policy to support such clusters. While the entrepreneurship policies discussed in our review generally do not require large funding streams, it is still important to consider the downsides of ineffective policies. At this point, we are still just beginning to acquire enough wisdom to create sound policies that internalise the externalities that can come from innovation and new start-ups.

While we believe that there are conclusions that can be drawn at this stage – research universities powerfully impact local development; focusing on large-scale employers can crowd out small-scale start-ups – we need much more information before entrepreneurship policy can attain the relatively mature status enjoyed by, for example, policies towards international trade and monopoly. The primary need is for experimentation and evaluation, especially with randomisation. Our review provides a variety of ideas for how greater experimentation can be built into existing policy initiatives and proposals.

It would be ideal to conclude that paths towards better entrepreneurship policies are clear, but they are not. Some policies do seem to have many upsides and few downsides, such as allowing more skilled immigrants, strengthening education systems, and eliminating unwise regulations. But when we move beyond such simple broad policies towards specific entrepreneurship strategies like clustering, our ignorance becomes obvious. The best path forward involves experimentation and evaluation. In their absence, we cannot be confident that policies to promote entrepreneurship will have their intended impact.

References

Chatterji, Aaron, Edward Glaeser and William Kerr (2013), “Clusters of Entrepreneurship and Innovation”, NBER Working Paper Series, No. 19013, prepared for the Innovation Policy and the Economy conference, April.

Chinitz, Benjamin (1961), “Contrasts in Agglomeration: New York and Pittsburgh”, The American Economic Review 51(2), 279-289.

Glaeser, Edward, Sari Kerr, and William Kerr (2012), “Entrepreneurship and Urban Growth: An Empirical Assessment with Historical Mines”, NBER Working Paper 18333.

Greenstone, Michael, Richard Hornbeck and Enrico Moretti, “Identifying Agglomeration Spillovers: Evidence from Winners and Losers of Large Plant Openings”, Journal of Political Economy 118, 536-598.

Lerner, Josh (2009), Boulevard of Broken Dreams, Princeton, NJ, Princeton University Press.

Michelacci, Claudio, and Olmo Silva (2007), “Why So Many Local Entrepreneurs?”, Review of Economics and Statistics 89(4), 615-633.

Porter, Michael (1990) The Competitive Advantage of Nations, New York, Free Press.

Topics: Industrial organisation
Tags: entrepreneurship, innovation

Associate Professor at the Fuqua School of Business, Duke University
Professor of Economics, Harvard University
Assistant Professor, Harvard Business School

Vox Talks