MINNEAPOLIS — In a bid to lure Amazon's second headquarters to the Land of 10,000 Lakes, Minnesota pledged to give the company a bustling business community, lots of smart workers and a great quality of life for them.

Oh, and subsidies. The state suggested it could award Amazon $3 million to $5 million in subsidies, in addition to potential local incentives. (We know this because the state just released the bid which, despite a media lawsuit, was only made public after Amazon gave the go-ahead.)

Maybe it was our winters, maybe it was our taxes, or maybe it was that our incentive package was paltry compared to other locations'. But after being rebuffed in New York, Amazon ultimately chose to go to northern Virginia, which offered $573 million in performance-based incentives, and Nashville, which offered $102 million in city and state funding.

___

The nonprofit news outlet MinnPost provided this article to The Associated Press through a collaboration with Institute for Nonprofit News.

___

Hell hath no fury like a Minnesotan scorned, but being overlooked by Amazon might not have been a bad thing, suggests new research out of North Carolina State University.

By analyzing data on state tax breaks offered to companies in 32 states from 1990 to 2015, the forthcoming study found the financial incentives U.S. states give businesses have a detrimental effect on states' financial health.

There are lots of different packages these types of incentives come in: property tax breaks; research and development tax breaks, which reduce taxes to companies working on technological innovations; investment tax breaks, which give businesses deductions based on investments made; and breaks for job training grants and job creation.

Incentives like these have been around in some form or another for about as long as the U.S. has, said Timothy Bartik, senior economist at the W.E. Upjohn Institute for Employment Research, who compiled the database used in the NC State analysis.

The current wave started in the '30s, with southern states luring northern manufacturers by helping them build factories and buy equipment.

The trend ramped up after World War I as air conditioning made working in a factory in the sweltering heat more palatable, Bartik said. As northern states started feeling the competition, they jumped in with incentives.

States' use of incentives to lure companies tripled in dollars between 1990 and 2001, after which their expansion slowed on the whole, but some states increased them while others cut back, Bartik said.

But even as their use has become widespread, deployment of incentives often ruffles feathers. Proponents see them as a way to lure good jobs to a given area. Detractors see them as handouts to corporations.

Bruce McDonald, associate professor of public budgeting and finance at North Carolina State University and an author of the NC State study, got into the business of studying them after Raleigh, home to NC State, was named one of 20 finalists for the second Amazon headquarters.

"Everyone assumes financial incentives would have a positive effect because you're bringing new business," he said. But, "surprisingly little research definitively shows that. Most of the research has been on a case-by-case basis."

So, the team set out to determine the effects of incentives on states' fiscal health on the whole, attempting to account for different political characteristics, economic conditions, demographics and the structure of state governments and other factors that could affect fiscal health.

"Generally, financial incentives do a pretty bad job of impacting fiscal health, that is they're going to move the state into a worse financial situation," McDonald said. "The money that would have been coming into the government in some way is no longer coming in, but at the same time there's an expectation of increased expenditures: more roads to pave, more public services to offer, more people you're having to deal with."

Not all incentives were created equal, though, they found: The study's authors were surprised to find research and development tax credits hurt states' fiscal health the most.

"That surprised everybody. We're used to thinking about how much R&D has had such a positive impact," he said. But R&D costs a lot, and many new technologies fail, he said.

Next came investment tax credits and property tax abatements.

Property tax breaks can go a couple different ways, McDonald said: If a business gets a tax abatement to revitalize an area, there's a potential for positive benefit because it can increase the tax base in the long run. That's especially the case for places not generating much in property taxes from the get-go.

But these instances are few and far-between, he said.

"Most of the time, somebody comes in and they're being attracted to a certain area not because it needs revitalization but because that's where they want to move to," McDonald said.

Job training grants tend to foster dependence on the federal government, since the federal government often funds part of them, McDonald found.

Only job creation credits tended to have no significant effect, on average, on states' fiscal health, according to the study.

McDonald and his colleagues concluded that incentives can bring a return on investment, but also put the state's financial health at risk.

"As such, they are a financial tool that should only be used with caution," the study finds.

Some states appear to use them more than others.

Average the annual subsidies out over 10 years, from 2006 to 2015, and Minnesota ranks 28th among the 32 states in the study in terms of incentives relative to the size of the value added by its industries, at 0.4%, according to the study. New York ranked first, at 4.7%, while Washington comes in last at 0.2%.

Minnesota's tax incentives also tended to have a larger return on investment than other states', the study found.

In recent years, Minnesota spent the largest share of its incentives on jobs subsidies, followed by investment subsidies, training and then R&D subsidies, per the study.

A 2012 New York Times investigation found Minnesota's state and local governments spend $45 per capita on subsidies each year. Its agriculture, finance and telecommunications sectors were the biggest beneficiaries. Alaska, the heaviest per-capita subsidizer, spent $991 per capita, while Nevada, the lightest, spent $12 per capita.

Some of Minnesota's biggest recent subsidies include $585 million in infrastructure funding to Rochester's Destination Medical Center, contingent on private investment; $250 million in tax breaks for infrastructure approved by the Legislature in 2013 for Triple Five Group, the Mall of America's owner, to help expand the mall. Also of note: the Minnesota Vikings Stadium, which the state helped fund directly, at $348 million.

Bartik, who has studied the effects of incentives extensively, has a list of reforms states could enact to get more bang for their buck when they offer them.

He suggests incentives should be target firms that benefit from clustering in a given area, which tends to increase their cost-benefit ratio.

Incentives to big businesses can be balanced with programs to help small businesses. Bartik suggests expanding states' manufacturing extensions (similar to agricultural extensions, public agencies that do research on ag and advise farmers, but for manufacturing), and spending on customized job training.

He also suggests states set caps on incentives, making sure abatements for businesses don't diminish budgets in areas like education and infrastructure, which also help grow business.

Bartik also thinks governments should limit one-time deals to lure companies, which often don't actually influence their decision to locate.

When it comes to business incentives, "You need an economic development strategy, but you need to have one that's affordable and that's cost-effective," he said.

The nonprofit news outlet MinnPost provided this article to The Associated Press through a collaboration with Institute for Nonprofit News.