Recent Research: Federal R&D boosts local economic development

The boost in federal R&D funds as a result of the 2009 stimulus package had a significant impact on local economic development, according to a new working paper from researchers at the University of Michigan. In “Local Fiscal Multiplier on R&D and Science Spending: Evidence from the American Recovery and Reinvestment Act” (ARRA), authors Yulia Chhabra, Margaret Levenstein, and Jason Owen-Smith look at changes in county-level employment in response to increased federal spending on R&D as a result of ARRA. The authors estimate causal effects of the ARRA R&D funding on local economic development, and find that, all else equal, every $1 million in new R&D spending due to the stimulus in a county led to 27 new jobs, with 25 of those being in the private sector. The authors estimate that the cost per each job-year was about $15,000, which is less than the reported costs of other types of federal stimulus programs.

Although most ARRA spending targeted areas that were hit hard by the recession, ARRA R&D spending was unrelated to economic conditions, the authors note. Instead, the allocation of ARRA R&D was based on merit – essentially, whether or not the region had a research university capable of carrying out the grants, or whether they had previously applied for R&D funding but did not receive it due to budgetary constraints. Between 2009 and 2013, every county received an ARRA award for a purpose besides research, while just more than half of all counties received ARRA R&D funding, the authors find. R&D funds represent a relatively small portion of overall ARRA spending, according to the authors. Although the average amount of ARRA R&D in a county was $34 per capita over the five-year period, the average amount of non-R&D ARRA funding in a county was far greater: $824 per capita during the same period. 

The authors note considerable differences between those counties receiving ARRA R&D funding and those that do not. Counties receiving ARRA R&D funding tended to be more populous and urban, have more people employed in scientific R&D services industries, and were twice as likely to have a research university. Seven of the eight counties with the largest ARRA research spending from 2009 to 2013 were in coastal counties, with four of the eight located in California. With the exception of Suffolk County in Massachusetts, home to Boston, large coastal counties are less prominent when ARRA R&D awards are standardized on a per capita basis.

Using several empirical models to evaluate the impacts of ARRA R&D, the authors conclude by noting that R&D spending has “significant stimulus effects and that those effects are larger than those that have been reported for many other types of federal stimulus.” Although the authors are careful not to draw too many conclusions from these findings, there are several potential implications for the regional economic development practice.

On the one hand, these results are encouraging. Advocates of technology-based economic development have long argued that increased federal funds for research and development would have positive reverberations on local employment in the private sector. The authors’ findings help confirm elements of this message. Not only was this stimulus package effective at supporting the overall R&D enterprise, it was also effective at creating new private-sector jobs at a cost much lower than other stimulus programs. As the federal government considers approaches to supporting economic development during the next recession, these findings should be considered.

However, the limitations of these findings are also worthy of consideration. Instead of funding research activities solely in regions well suited for success, perhaps more of an effort could be made at the federal level to help level the playing field for innovation and R&D.