Amazon CEO Jeff Bezos must have been smiling recently when his company revealed the 20 locations in contention for its corporate headquarters expansion, "HQ2." The announcement reignited a media frenzy over who will be selected for a $5 billion facility and 50,000 new jobs.

Local leaders around the country are falling over themselves to attract Amazon's attention — more than $22 billion in tax write-offs have already been offered, and that's only counting the half of the bids that are available to the public. There's just one problem: Subsidies won't determine where Amazon ends up.

History suggests that factors like a city's geographic location, industrial mix, or workforce — which politicians generally have little control over — matter much more over the long run than tax breaks.

Take Newark and the State of New Jersey, which are offering the largest combined incentive package: $7 billion in tax breaks over 20 years. To New Jersey folks, that's a huge subsidy — the state collects around $2.5 billion in corporate taxes each year, so they're offering Amazon about 14 percent of the taxes paid by other businesses for the next two decades.

But to Amazon, it's merely a nice perk that pales in comparison with the $175 billion it earned last year alone. That means the largest subsidy offered to Amazon amounts to only 0.2 percent of its current annual revenue. And this proportion will drop even lower given Amazon's 20 percent annual growth, with revenue poised to top $1 trillion in the next decade.

In fairness, tax breaks affect profits more than overall revenues. But a location that streamlines business supply chains, enables cooperation with other companies, or offers a skilled labor force is more profitable in the long run than a location that merely offers tax incentives.

Imagine that choosing the best location would allow Amazon to grow just 1 percent faster each year. Over 20 years that extra growth would translate into more than $1 trillion in extra revenue and more than $17 billion in extra profits. In other words, even the most generous subsidy can't beat the value of locating in the right place.

Adding insult to injury, subsidizing corporate profits hurts local residents: Public services suffer from a lack of funds, taxes are higher than they would otherwise be, or both.

Instead of offering Amazon subsidies, Chicago ($2.25 billion) and Philadelphia ($3 billion) could instead add 650 and 1,000 officers to their police forces — increases of 5 and 15 percent, respectively. Alternately, the tax breaks could pay for the education of 145,000 students in Chicago this year or they could completely repave every street in Philadelphia — four times over.

Similarly, the $3 billion Wisconsin gave away to attract the electronics manufacturer Foxconn last year could have instead reduced corporate income taxes by 21 percent for 16,000 businesses across the state.

The broader economy suffers when we subsidize one company at the expense of others, leading them to grow more slowly because of their relatively higher tax burden. And if economic development incentives did motivate companies to choose less-than-ideal locations, the resulting production inefficiencies would also lead to reduced economic growth. Lastly, economic development subsidies often have strings attached that limit business flexibility, harming the dynamism of the economy.

Moreover, the competition between local governments to attract jobs pushes politics toward cronyism. Giving special-interest groups undue influence runs against the ideals of democratic governance — that government authority should serve the people in general, rather than just those with power and influence.

Amazon's competition for HQ2 is a publicity stunt designed to pump up profits through taxpayer-funded handouts. It likely already knows where it's going and any tax break is just icing on the cake. But it's a convenient publicity stunt for politicians, too, who get to show voters they are "doing something" to improve the economy. In the end, Amazon's smile only comes at the cost of taxpayer frowns.

Michael Farren is a research fellow with the Mercatus Center at George Mason University.