Sarah Holder is a staff writer at CityLab covering local policy, affordable housing, labor, and technology.
A new economic analysis suggests that when cities and states offer tax deals for large companies, public education suffers and incomes eventually fall.
Are the economic incentives cities and states offer huge companies worth it for local residents? That depends on who you ask. But for the cities and states vying for Amazon HQ2 by bartering bigger and bigger tax breaks, the answer matters.
New research from Timothy J. Bartik, an economist at the W. E. Upjohn Institute, suggests that while incentives do indeed have benefits for local economic development in the short term, negative effects begin compounding as soon as 22 years into an agreement. It’s public education that suffers most drastically from budgetary reshuffling; and vulnerable low-income populations that are afforded the smallest gains.
“There’s no free lunch,” said Bartik. “The incentives have to be paid for, and the money comes from somewhere.”
Bartik used a model of a typical local economy to find out where, and to put real numbers to some conventional wisdom (and controversial assumptions) regarding economic development agreements.
The first: To what extent are economic incentives even responsible for new job creation at all? Incentives have long been the bread and butter of economic development operations. But they’re not the only thing most cities can or do offer. When companies choose a specific location they probably take a lot more than money into account (like cultural climate, local talent, transit access), and would create a lot of those jobs regardless of how much money it took to lure them. Bartik builds upon previous research to assert that only 10-15 percent of the new jobs companies create in incentive-offering cities and states can really be credited to the incentives they offer.
Once companies get there, however, jobs do follow. Each incentivized job has an average of a 2.5 percent “multiplier effect,” meaning 1.5 additional jobs are created locally for every one job the company itself adds.
Determining whether these jobs go to local or out-of-state workers is more complicated. In the short term, an estimated two thirds of the jobs are filled locally, and a third from out-of-state (or city) migrants. In the long run, however, the vast majority of the jobs created aren’t locally sourced: 85 percent of the jobs end up increasing the population via in-migration of workers, while only 15 percent increase the employment rates of local residents.
That has positive effects on internal tax revenue, local spending, and job growth—especially for those in lower income brackets—but since these community benefits are often paired with a jump in population, Bartik says, they all but even each other out. “Over time, as more people migrate in to fill the jobs,” he says, “the effects on the employment rate tend to go down.”
Research from the Economic Policy Institute earlier this year that looked specifically at the employment impact of Amazon Fulfillment Centers on their host communities came to a similarly sobering conclusion. “While warehousing jobs increased in the two years after,” wrote CityLab’s Tanvi Misra, “overall employment at the county level remained roughly the same.”
Those undersized (and eventually, waning) benefits could easily be dwarfed by the cuts cities have to make to afford incentive costs. Bartik says the most drastic cuts often come out of K-12 education—and skimping on schooling causes the worst damage to communities later. Every 10 percent siphoned from K-12 spending results in a long-term wage decrease in the community of 8 percent. Of course, those generational consequences only affect local wages if the students stay in-state. But ”people aren’t quite as hyper-mobile as you’d sometimes assume,” said Bartik. Sixty to 70 percent of Americans stay in the same state for most of their working careers, he says, a number that stays close to half in metro areas and drops to 40 percent among college-educated people.
It’s those already in the lowest education and income brackets whose long-term prospects are most affected, said Bartik, because their future wages depend on the existence of a high-quality public school system.
But if jurisdictions don’t funnel enough away from public services, that money has to come from tax-payers instead, Bartik says, again disadvantaging the already disadvantaged. “The lowest income group tends to bear more than its share … due to sales taxes,” state and local taxes, and state and local public spending. Meanwhile, gains in property value benefit upper-income residents.
In all, Bartik found that the net benefits of incentives on local incomes—at least on people living inside his model economy—amount to only 22.3 percent of incentive costs. And those benefits are spread unevenly among income brackets, due to the uneven burden of de-funding and taxing outlined above. In the end, the net income for those in the lowest income quintiles (and the second-highest, surprisingly) actually drops as a result of incentive policies. In places where incentives are explicitly paid for by cutting K-12 spending, state per capita income drops by more than $4 for each $1 spent on tax incentives.
So what are city and state leaders to do? Of course, not all incentives are created equal. And Bartik doesn’t suggest getting rid of them entirely. “There are things you can do to more carefully target your tax incentives so they have more impact, more bang for the buck,” he said. Strategies include targeting businesses with a high “multiplier effect,” and those that hire more local employees at higher wages. (Amazon has suggested that Seattle HQ’s multiplier is 3, while critics suggest HQ2’s will be closer to 1.) To ensure that incentives have a better short-term impact and a less detrimental long-term tail, Bartik also suggests front-loading the incentive timeline so cities pay more sooner, but less overall.
“If you did that targeting, you would be turning down those bigger deals,” he said. “Then that frees up funds you can use for manufacturing, customized job training, setting up business incubator programs, and doing a variety of things to foster the growth of locally owned businesses through services.”
Experts disagree on just how much cities, counties, and states across the U.S. offer companies each year. In 2012, The New York Times estimated $80 billion; in 2017, Bartik himself calculated a more conservative $45 billion. But most can agree that they’re accelerating. According to Bartik’s 2017 report, incentives roughly tripled as a share of the economy from 1990-2015 (most of the growth occurred between 1990 and 2001, and stabilized between 2001 and 2015). Foxconn’s $4.5 billion deal from Wisconsin—roughly 10 times the cost per job of typical U.S. incentives—and Amazon HQ2’s public offers of up to $7 billion ratchet up the stakes even higher for future uneven deals.
“The context is that there’s a threat these costs will escalate out of control,” said Bartik. “If they do, it might impinge even more on educational spending and on the potential for services to small business.”