What makes incentives effective? A recent article explains that “mediation,” or active, relationship-based, strategic, sector-specific recruitment and retention activity, makes a critical difference.  In other words, good economic development practices can enhance the job-creation potential associated with incentive offerings. 

“Mediation dramatically improves the ability of the state’s incentives to generate meaningful job creation.”
BTC Reports, October 2012


The May 2014 article, Mediating Incentive Use: A Time-Series Assessment of Economic Development Deals in North Carolina, asks whether incentives are more effective when used in conjunction with industry- or sector-wide supporting functions, such as industry targeting or “institutional mediation,” which means providing strategic support to specific sectors.

The study described in the article finds that, yes, incentives that “are coupled with sectoral economic development efforts . . . generate substantially stronger employment effects” at the establishment level.  State-led mediation activities “positively influence incentive effectiveness,” with greater job creation at incentivized firms than among those that were not incentivized and those that were incentivized but not mediated.

In contrast to many incentives studies, the authors examine job generation at the establishment level, rather than at the aggregate (community, regional or state) level.  They use a state announcements database and annual reports to identify companies receiving incentives between 1996 and 2008.  They use the NETS database (itself based on Dun & Bradstreet business listings) to obtain establishment characteristics for these businesses over time (including employment) and to build a control group of similar businesses that did not receive incentives. 

The dataset consists of 270 valid records, divided into retentions and recruitments.  One concern is that NETS data are generally considered to be adequate at the aggregate level, but employment data at the individual firm level may not be reliable.  The authors do acknowledge the fact that “some observers have been concerned about the measurement of employment levels” in the NETS database.

Another important point is that in this study, the specific sectors evaluated for “mediation” are limited to life sciences/biotech and advanced textile manufacturing/nonwovens.  It is not clear if selecting additional sectors would affect the findings.

This study is really more about mediation than incentives, with the incentives data serving as a way to get at the hard-to-quantify issue of whether industry support activities make a meaningful difference, especially in terms of job generation.  In fact, the incentives variables are binary – Y received incentives, N did not receive incentives – so there is not much insight into which incentives work well. 

The authors conclude that:

  • Incentives are one policy tool – not the only policy tool – available for economic development recruitment and retention.  Combining these tools in “mutually reinforcing ways” improves job creation performance.
  • Strategic investments, especially in workforce, create locational advantages and cost-saving opportunities, which diminishes the relative importance of the incentives.  (While this is sensible, the value of the incentive was not included in the analysis so it is not clear to me how they draw this specific conclusion.)
  • Incentives can have a greater impact when targeted to sectors that are the focus of additional state support. 

Mediating Incentive Use: A Time-Series Assessment of Economic Development Deals in North Carolina was published in the May 2014 edition of Economic Development Quarterly. A version of this paper is also available here on the W.E. Upjohn Institute for Employment Research website.