Last week Amazon finally announced that New York and Northern Virginia would be the sites of its planned major expansion. While many are caught up in the excitement of Amazon bringing 25,000 high-paid jobs to both metropolitan areas, it is worth thinking through the costs that beauty contests like this one impose on state and local governments. Amazon extracted billions of dollars in concessions from the winners and could have extracted even more from some of the other cities courting them.

It is economically rational for companies to create such Hunger Games-type competitions among communities. These competitions reduce their costs and improve their bottom lines. But is it economically rational for the cities? As long as governments are acting independently, yes, it is rational for them to race to the bottom to secure a win. So long as they are a bit better off by snagging the prize than they would have been otherwise, they come out ahead. But the metrics that politicians use are unlikely to be limited to a hard-nosed accounting of costs and increased tax revenues. Positive buzz may be enough to satisfy them.

Consider Wisconsin Governor Scott Walker's deal with Foxconn. Just over a year ago, he was touting the $3 billion state subsidy for FoxConn's manufacturing plant. This was the year leading up to his hard fought election fight, a fight he ultimately lost. His public statements focused on Foxconn's promise to create 13,000 jobs. While that was a lot of jobs, it was a hell of a lot of subsidy—more than $230,000 per job, more than six times the largest amount Wisconsin had ever paid to subsidize a promised job. Walker got his campaign issue, FoxConn got its $3 billion and Wisconsin residents got … had. The $3 billion dollar subsidy has grown to over $4 billion at the same time that Foxconn is slowing down its investment in Wisconsin. So now taxpayers are subsidizing each job by well over $300,000 each. Nonpartisan analysts have determined that it will take decades, at the earliest, for Wisconsin to recoup its “investment.”

Likewise, hundreds of millions of dollars are thrown at stadiums and arenas even though economists have clearly demonstrated that those investments do not generate a positive financial return for the governments that provide these subsidies. Fancy consultants set forth all of the supposed benefits: job creation, direct spending by all of the people drawn to the facility, indirect spending by those who service the direct spenders. This last metric is meant to capture the increase in restaurant staff, Uber drivers and others who will cater to the new employees, residents and visitors to the facility. But as has been shown time and time again, these metrics are vastly overstated and willingly accepted at face value by politicians eager to generate some good headlines. They also ignore the opportunity cost of the direct subsidies—monies spent on attracting a company is money that can't be spent on anything else. While we don't know what it would have been spent on, it is likely to have been public schools, mass transit, roads or affordable housing in many communities.

A mutually beneficial and more economically rational approach would be for certain tiers of cities to commit to not compete against each other with direct subsidies to companies seeking to locate a new facility within their borders. For instance, top-tier metropolitan areas with a population of a million or more might agree amongst themselves to only offer indirect subsidies and incentives that benefit a broader constituency. These indirect subsidies could be infrastructure improvements to roads and mass transit. Incentives could include rezonings that rationalize the area surrounding the new facility. They could also include commitments to improve local public amenities such as parks. Smaller metropolitan areas could make a similar pledge amongst themselves, perhaps even offering a limited menu of options that they would consider for the right company.

Some might argue that governments live in a state of war of all against all when it comes to economic development—and that they will always seek to maximize their competitive advantage when a suitor comes calling. Maybe. State governments could help local governments stay firm by tying their hands so that they cannot offer certain types of incentives. If this is done far in advance of the next Amazon-like contest, it might just mean that local governments can get a better deal for their constituents than they have been able to get in recent years.

At the end of The Hunger Games, Katniss and Peeta both survive the death match imposed by their overlords by agreeing to live or die together. Imagine what might have happened if Virginia and New York City had the courage to do the same.

David Reiss is a Professor of Law at Brooklyn Law School and the Research Director of the Center for Urban Business Entrepreneurship. He is also the editor of www.REFinBlog.com which tracks developments in the rapidly changing world of real estate finance.