Agglomeration economies in entrepreneurship:  a not-so-clear picture

By Ranie Lin
Research Intern, McNair Center for Entrepreneurship and Economic Growth 

In January 2019, Rice University first announced plans for the transformation of Houston’s historic Sears building into the Ion, the future centerpiece of an emerging innovation hub located along Main Street between downtown Houston and the Texas Medical Center. The innovation district, which broke ground in July 2019, has generated considerable discussion about the connections between revitalization, investment and geography.

At one end of the spectrum, Edward J. Egan, a visiting professor at Georgetown University in Washington, D.C., claims that the innovation corridor, because of its size and location, could actually reduce venture investment in Houston by up to 5%. Egan contends that the innovation district fails to take advantage of what are known as agglomeration economies, i.e., the benefits from concentrating firms and output in small geographies. Economist Alfred Marshall originally developed the theory of agglomeration in his book “Principles of Economics” and laid out three sources of external economies of scale: input sharing, labor market pooling and knowledge spillovers. In brief, Marshall argued that the agglomeration of goods, people and ideas in a dense geographic region increases the productivity of industries and firms.

The precise applicability of Marshallian agglomeration to modern innovation and entrepreneurship is more complex. For example, Florida and Mellander find that dense urban centers, rather than suburbs and office parks, are the optimal modern geography for high-technology innovative firms. While Florida and Mellander use empirical evidence of start-ups concentrating in particular zip codes, Egan acknowledges that the choice of geography units — whether it be census blocks, zip codes or entire cities — has led to varying conclusions among researchers.

The difficulties in identifying the geographies of successful entrepreneurial ecosystems have made designating innovation districts a difficult task for city planners and legislators. In 2014, Katz and Wagner of the Brookings Institution put forth several models of innovation districts similar to the Ion model, where “leading-edge anchor institutions and companies cluster and connect with start-ups,” resulting in “multiple opportunities for neighborhood revitalization, quality employment, and poverty alleviation” for residents. In response, Egan dismissed these models as infeasible and based on weak hopes that “if you build it, they will come.”

The crux of Egan’s theory is that “entrepreneurial spawning requires a pre-existing population of venture-backed successes and other suitable firms to spawn from.” He argues that the Ion innovation district is too large and too distant from existing hubs of entrepreneurship. Instead, Egan insists that areas designated as innovation districts should “already be home to clusters of firms in high-technology industries that have intense agglomeration.”

While locating innovation districts in places where such firms already exist may make use of existing agglomeration effects, it can overlook other important benefits of entrepreneurial ecosystems, such as economic revitalization. For example, Lawrence, Hogan and Brown find that innovation districts can “improve the social life and networks of a city” by bringing “income, jobs, education, and safety into distressed neighborhoods that often lack economic and educational opportunity.” If a location is already home to a high density of firms, the benefits of a creating an innovation district may be less pronounced.

Lastly, the link between agglomeration and entrepreneurship is further complicated by the industrial structures within a geographic area, including input suppliers and skilled workers. For example, Glaeser and Kerr studied industrial agglomeration and found that the existence of small suppliers and abundant workers in relevant occupations explains between 60% to 80% of firm entry in manufacturing entrepreneurship. Egan further notes that it may be these industrial structures that are causing firm entry and agglomeration, rather than agglomeration causing entry and investment.

In sum, the role of agglomeration in stimulating investment, innovation and entrepreneurship is more complex than recent commentators may suggest, and economists continue to scrutinize the optimal geographic conditions that constitute successful entrepreneurial ecosystems. Houston’s innovation corridor, if examined carefully, may be able to help business leaders, academics and policymakers to further understand these crucial connections.