Peters and Fisher (2004) ask if economic development incentives are a cost-effective strategy for achieving economic growth.
They break the question down into one with three parts:
1. Do business incentives actually cause states or localities to grow more rapidly than they would have otherwise?
2. If so, is growth targeted so as to provide net gains to poorer communities or poorer people, or is it merely a zero-sum game?
3. How costly to government is the provision of these incentives compared to alternative policies?
They define business incentives as both tax incentives (abatements, credits, exemptions, etc.) and non-tax incentives (grants, loans, loan guarantees, etc.) that are provided to firms as the initial recipient.
They estimate that nationwide spending on business incentives is approx. $50 billion.
They review two broad and related justifications of incentive provision:
1) that incentives lead to business investment and new jobs, increasing demand for local goods and services, leading to further economic growth, and
2) economic growth increases public revenue, allowing the provision of more or better public services or lowering of tax rates.
They suggest that most other justifications are derivatives of these two.
One argument against incentives is that US workers are highly mobile so that new jobs are primarily filled by in-migrants, thus providing little or no benefit to the local unemployed.
They suggest that this argument is faulty because worker migration is slow and sticky; migration is time-consuming and has costs. They review Bartik's (1991) discussion of hysteresis effects in job markets: Local increases in employment allow employed workers to move up, creating entry-level vacancies for unemployed workers with fewer job skills. This employment allows marginally-skilled workers to gain new job skills that not only help them in the short-term but increase their long-run employment prospects. Thus, it is possible that stimulation of local employment can have positive welfare effects and larger effects in poorer areas.
Peters and Fisher suggest that incentives will produce these benefits only if they are targeted at the poorer areas or people.
Now they address their three questions:
Does [the use of] economic development [incentives] induce jobs or investment?
They review literature reviews. Up until the late 1980s, the consensus was that incentives had very little to no effect on business location decisions; that, on average, incenitves did not "tip the balance."
The reasoning was that taxes make such a small percentage of operating costs, their effects would be swamped by other costs such as labor or transport costs.
There are other reasons:
1) Firms pay income taxes on their incentives, lessening their value. Estimates are that 30-45% of incentives go to other governments, primarily as higher federal income tax.
2) Firms may be wary of locating in jurisdictions offering overly generous incentives -- they are wary of profligate governments.
3) Low taxes and large incetives may lead to poor quality public services.
Thus, the early consensus position, popularly stated in Eisinger (1988), was that economic development incentives had at best an ambiguous impact on growth, but probably no impact at all.
However, from the late 1980s through the late 1990s there was new research that found significant, although small, effects of incentives on economic growth. The new consensus was that lower taxes and higher incentives were likely to result in greater economic growth.
The shift in findings was attributed to better econometric techniques. However, the findings have been called into question because of flawed data and an inability to replicate results across time and geopgraphy.
The rationale against the existence of an effect is the same as before: the effect of incentives is too small compared to other factor costs -- notably labor.
Also, they note that even if statistically significant, incentives may not be practically significant. Estimates of interstate elasticity of economic activity is -0.3. (i.e. cut taxes 10% increase jobs or investment by 3%). Thus, the practical effect, even of relatively large incentives, is small.
In short, do incentives produce economic growth? Who knows?
Next, they focus on the distributional issue.
Who takes the jobs "created" by economic development?
Or: Is economic development policy appropriately focused on poorer people or poorer areas?
They argue that such a focus would be more efficient, referring to Bartik's (1991) discussion of reservation wages.
They consider three conceptualizations of the question:
* Do poorer place pursue economic development more vigorously than other places?
* Do states target incentives at more needy places or populations?
* Do poor people living in targeted areas benefit from targeted policies?
They claim the empirical literature on these questions is "skimpy."
One at a time:
*Do poorer places pursue economic development more vigorously than other places?
Research is mixed, with later studies showing poorer jurisictions not to be more vigorous. Why would this be the case?
Poorer places have less money to deveote to incentives.
All localities feel they are competing; wealthier ones can offer bigger incentives.
There is "enormous policy inertia." Even if a community starts offering incentives in hard times, it doesn't stop when times get better.
*Do states target incentives at more needy places or populations?
They note that states tend to target programs (such as enterprise zones) to poorer locations, but that doesn't mean that they will be effective in helping poorer people.
Because firms may not take advantage of incentives targeted to people with low skills; the incentive may not be great enough to justify hiring a low-skill worker.
Because labor incentives are often overshadowed by incentives that lower the cost of capital.
*Do poor people living in targeted areas benefit from targeted policies?
Available research (but there's not much) says, "No. Not really." Mainly because job markets are not strictly local. New jobs don't necessarily go to people already in the target area and not all of those people are poor or unemployed.
Furthermore, even when officials target programs to poor areas, other -- not necessarily poor -- areas want to have the same programs available. It is hard not to let the programs spread to the other areas.
The third question:
Are business incentives fiscally benficial?
In other words, aside from effects on the poor or unemployed, does providing incentives result in a net gain to the public treasury?
They review the available research and conclude that it is not likely, at least at the state level. They review a discussion by Bartik (1994) of why this is the case.
First, the elasticity is small (~ -0.3) so you don't get much bang for the buck. Second, by the time initial incentives end, the firm may have moved on elsewhere. [Or, I would add, they may have extracted additional incentives not to move on elsewhere.]
They note that incentives are more likely to have net fiscal benefits at the local level (rather than at the state level) since they may tip the balance between closely matched cities. They note that this gain is actually a loss at the regional or state level if it merely results in selection between neighboring cities in the same state. They don't discuss the possibility of competition between similar, but remote, cities.
They say that this literature is the "thinnest of all" and that this issue of fiscal benefit is far from settled. But that the available evidence suggests that incentives are costly.
Since locational effects are weak, states and localities lose more by granting incentives to firms that would have located anyway than they gain from the truly marginal firms. Cities may gain, but mainly (they say) at the expense of nearby cities in the same state. And if state funds are used for locally targeted incentives the state subisidizes moves from one place to another.
Finally, they summarize and offer alternatives.
They point out that available research doesn't look good for the answers to the questions they posed at the start:
1. Do business incentives actually cause states or localities to grow more rapidly than they would have otherwise?
2. If so, is growth targeted so as to provide net gains to poorer communities or poorer people, or is it merely a zero-sum game?
3. How costly to government is the provision of these incentives compared to alternative policies?
They suggest a need for a "radical transformation of ideas" on local economic development policy.
The main problem is that public officials believe that subsidies and incentives have a lot more effect than is supported by the evidence. They suggest that we need to make a case for "a more sensible view of the role of government":
... providing the foundations for growth through sound fiscal practices, quality public infrastructure, and good education systems -- and then letting the economy take care of itself.
They suggest there is still a role for programs aimed at improving worker employability and occupational and geographical mobility.
Posted by Chip on May 23, 2004 at 06:16 AM | TrackBack