The state of Michigan — like many other states — operates a wide array of incentive programs that proponents believe will produce more benefits than costs. These benefits are often sold first and foremost as new, retained and perhaps even better jobs. Such programs have been operated for decades and provide scholars with robust datasets which can let them study their efficacy.
The Mackinac Center for Public Policy has been researching and writing on government economic development programs since the late 1980s. It has published three rigorous statistical analyses — including the one contained in this study — specific to the state’s Michigan Economic Growth Authority alone. We have also previously performed similar analyses or reviews of the state’s Pure Michigan tourism subsidy program, the Michigan Business Development Program, the 21st Century Jobs Fund and the state’s film incentive program.
For this study, the Mackinac Center created a database of 7,300-plus deals found across 49 documents, all but two of which were obtained directly from the state library and the Michigan Economic Development Corporation. We then matched each company that received a deal with their employment records found in the National Establishment Time Series database.
The NETS database tracks firm employment — among other data — over time. This makes it easier to measure the jobs being created at incentivized firms and to better measure the direct impact from offering incentives to them. It also allows scholars to compare the performance of incentivized firms to a control group of nonincentivized firms.
In total, we found that the average incentive offered per job from our complete database of deals was $593,913. Furthermore, each incentivized firm created just over six jobs each, on average.
Across the nine program areas we examined, five showed no statistically significant impact and one produced a negative impact.
Three program areas demonstrated a positive impact but at a considerable cost, as measured by the amount of incentives offered per job. They included:
We also performed a separate analysis of the MBDP and found no statistically significant impact on job growth from MBDP incentive deals.
There is a broad and methodologically evolving literature on the role incentives play in business location decisions, and this study adds to that work. Incentives may influence the location of business. Several studies acknowledge they play a role in the decision of firms to locate or expand in a state or sub-state region. But the cost per job of incentivizing business location are often several orders of magnitude higher than the average annual wage of that job. While clearly identifying these effects present challenges, studies with careful identification strategies tend to report higher costs per job. This study provides a carefully identified analysis of the incentives offered per job in Michigan and finds similar results: the average cost per job of incentive deals is too large for these programs to pass a basic cost-benefit analysis.