Manufacturing Output and College Graduates: Is There A Relationship?

An increase in a city's share of college educated workers results in an increase in manufacturing output in that city, according to a report released by the National Bureau of Economic Research. Human Capital Spillovers in Manufacturing: Evidence from Plant-Level Production Functions, authored by Enrico Moretti, bolsters the notion that spillover of knowledge through human capital is important to tech-based economic development (TBED) and that knowledge spillovers play a critical role in many theories of regional growth (i.e. endogenous growth theory, clusters).

Moretti is interested in exploring the effect the education of a workforce has on human capital spillovers at the U.S. city level. The author compares the productivity of manufacturing plants to the city's share of college graduates as an indicator of human capital spillovers. Moretti finds that a 1 percent increase in a city’s share of college graduates is associated with a 0.5 percentage point increase in manufacturing output.

Moretti's work is based on the notion that if spillovers exist, plants in cities with a higher level of human capital can be more productive than similar plants in cities with less human capital utilizing the same inputs. The author concludes that the productivity of plants in cities with a large share of college graduates increases more than the output of plants in cities with a small contingent of college graduates.

The presumed existence of spillovers has provided a critical foundation for many TBED policies: research such as Moretti's supports arguments for why state and local governments should be involved in many aspects of encouraging the growth of a tech-based economy. For instance, the policy implications of Moretti's work would argue for increasing academic research capacity, and the number of college graduates that reside in an area, as well as working with manufacturers and industrial facilities through programs such as the Manufacturing Extension Partnership, run through the National Institute of Standards and Technology.

Moretti analyzes the consistency of his results through four specification checks. The author separately analyzes components of his results to reinforce his conclusions regarding the effect of human capital spillovers on manufacturing productivity. The results of some of these checks have additional policy implications for TBED practitioners.

For example, one of Moretti's specification tests relays that spillovers between plants that rarely interact are zero, while spillovers between plants that often interact are significant. This finding is particularly encouraging for the field given much of the work or benefit of many technology-based economic development programs, regional technology councils, and industry-university research consortia is intended to encourage collaboration and strategic alliances among local businesses and industries.

Moretti also supports his results by verifying that the estimated spillovers are greater in single-unit plants than multi-unit plants. Multi-unit establishments may have plants in cities across the country, and decisions affecting that firm are probably made where the headquarters is located, Moretti observes. In addition, spillovers are greater for high-tech plants and more knowledge intensive industries, while they are virtually zero for low-tech plants.

The fourth specification test conveys that the density of physical capital outside a plant has no effect on a plant's productivity in the model. The author is interested if the stock of physical capital outside of a plant is associated with an increase in productivity. Since the author’s results show no effect of physical stock on plant productivity, then the estimates are theoretically capturing only human capital externalities in the author’s research.

The complete paper can be downloaded at: http://papers.nber.org/papers/W9316