Other studies using NETS data to track establishment and employment changes over time have been performed for Maine, New York, Illinois, Oklahoma, Vermont and California.[21] Two consistent themes jump out from most of these analyses, including those involving Massachusetts, mentioned above, and Pennsylvania, referenced below.
The first theme is that business relocations into and out of states has a relatively small net effect, particularly when compared to each states’ homegrown businesses. In other words, a state’s economic performance appears to hinge more on how many new businesses are formed within it — and how much incumbent businesses expand — than on how many businesses relocate from other states. The second theme is that when relocations do occur, they typically come from nearby — often bordering — states. This makes sense because there are smaller moving and management costs associated with relocating over a nearby border. The rest of this section describes some of this literature.
“Do Some Enterprise Zones Create Jobs?” was published in the Journal of Policy Analysis and Management in 2010.[22] Enterprise zones represent attempts to bring more economic development to a geographic area than would otherwise occur by providing targeted fiscal favors, such as tax incentives. These zones are typically used in poor or otherwise economically depressed areas.
The authors of this paper — Jed Kolko and David Neumark— employed NETS data drawn from between 1992 and 2004 to track employment and other differences within regions that included designated enterprise zones in the state of California.[23] A total of 31 zones were included in the analysis.[24] The authors’ approach is nuanced, attempting to account for different characteristics of zones, including “locational factors and variations in implementation of administration.”[25] These include such things as public relations work to ensure businesses know of incentives available in the zone, offering worker training or “encouraging the building of additional infrastructure.”[26]
Kolko and Neumark found that, on balance, enterprise zones in California “do not increase employment.”[27] The authors attribute this finding to the unique traits of California’s program during the period of their study.[28]
There was, however, better outcomes for employment in those zones that had a lower percentage of manufacturing businesses relative to other types of employers.[29] The authors also found that some zones had better employment outcomes, or “job-creating effects.”[30]
Another 2010 study, “Boon or Boondoggle? Business Incubation as Entrepreneurship Policy,” used the NETS database, among other sources, to examine the impact of business incubators (many of which were publicly subsidized) as a factor in economic growth. The author compared incubated firms with nonincubated firms on growth in sales and employment, as well as firm survival between 1990 and 2009. Of the incubators included in the study, 32 came from Michigan.[31]
Alejandro S. Amezcua, now an assistant professor of entrepreneurship at Syracuse University, found that “incubation is not associated with a major increase in the survival, employment growth, or sales growth of new ventures on average.”[32] In fact, incubated firms have somewhat lower survival rates than their nonincubated equivalents.[33] Sales and growth rates are higher in incubated firms but arguably not by enough to justify special treatment.[34] Notably, incubated firms that are underwritten by universities outperform other incubators.[35]
The 2010 study, “Growing Pennsylvania’s High-Tech Economy: Choosing Effective Investments,” published by Good Jobs First, also uses the NETS database. It does not use a treatment and control group analysis, as other studies in this literature review have, but it finds targeted tax incentives to be ineffective. It also compares the Keystone State’s use of incentives to lure high-tech jobs and business to those of six competing states.[36]
Doug Hoffer, one of the authors, finds that the in- and out-migration of jobs to and from Pennsylvania is tiny, relative to those establishments started in the state, along with those that died, expanded or contracted there. Specifically, between 1990 and 2006, there was a small net loss of 2,850 high-tech jobs from firm migration, despite the state’s efforts to offer incentives to grow tech-related employment.[37] The number of businesses moving in slightly exceeded the number moving out, by just 43.[38] Hoffer finds that establishment relocations into Pennsylvania are typically to and from neighboring states, a theme echoed in other studies using the NETS database.
Alan Peters and Peter Fisher authored a section of the report titled “How Taxes and Economic Incentives Affect Returns on New Manufacturing Investment in Pennsylvania and Surrounding States.”[39] They use what is known as a representative firm analysis to measure the importance of both taxes and targeted tax incentives to the post-tax profits of high-technology manufacturing businesses.
Their Tax and Incentive Model, or TAIM, was used to “compute how corporate taxes and incentives in an average city in each of the seven states [surrounding Pennsylvania] interact with typical financial statements of actual firms in eight manufacturing sectors.”[40] With their model, the authors try to measure the impact that both taxes and tax incentives have on location decisions for manufacturing firms in Pennsylvania.
They find in their first analysis — which focused solely on tax rates — that the “after-tax rates of return vary little among the states (with other factors held equal).”[41] When incentives are included in their model, little changes, except in two states — Maryland and New York — that offer very generous tax credits in particular geographic zones.[42] These states performed better.[*] They conclude, “For the vast majority of companies, tax breaks are windfalls, not determinants, and are therefore wasted.”[43]
The studies above are only a few of the many involving NETS data, and those are only a subset of the larger literature universe on state and local economic development incentives. A larger review of peer-reviewed, academic studies and others on this topic goes beyond the scope of this paper.[†]
[*] It is worth mentioning that Ohio may be an “anomaly” in Peters’ and Fisher’s overall review, but they explain that its performance may have been a function of the tax and tax year (which included some reforms) they were modeling.
[†] “The Failures of Economic Development Incentives” by Alan Peters and Peter Fisher is just one example of a larger review. It was published in 2004, though little academic research has been produced since then, in our opinion, to challenge their conclusions. The authors conclude: “The most fundamental problem is that many public officials appear to believe that they can influence the course of their state and local economies through incentives and subsidies to a degree far beyond anything supported by even the most optimistic evidence.” Alan Peters and Peter Fisher, “The Failures of Economic Development Incentives,” Journal of the American Planning Association 70, no. 1 (2004): 35.