Amazon’s much-anticipated HQ2 announcement was predictable. Not just because the tech giant went with two obvious choices to add thousands of tech jobs in — New York City and Arlington, Virginia in the Washington, D.C. metro area — but because of how much these two cities were willing to give away to get them.

For two cities that are supposed to be home to superior talent pools and have thousands of tech workers already, New York and Arlington were still willing to pay Amazon roughly $60,000 per job and $22,000 per job respectively (that’s if Amazon provides the promised 25,000 jobs).

Another predictable development, as is the case with most corporate incentive package announcements: Some state and city officials, as well as local residents, are trying to thwart the incentive packages before the ink dries. A New York state lawmaker announced a plan to block the proposed incentive package for Amazon and instead use it to buy up and cancel some of the student debt held by state residents.

At the heart of the argument against providing Amazon with subsidies are a few points: One, why does a company that generates more than $175 billion in revenue annually need hundreds of millions in subsidies? Two, if Amazon already downgraded the projected number of jobs it would give to its HQ2 city (well, now HQ2a and HQ2b), how can the company be trusted to deliver on its promised number of jobs?

Amazon’s bait-and-switch may have caused a lot of grumbling, but it won’t be enough to get cities to forgo giant subsidy packages. It will take a much more forceful movement instead to reform the way cities and states negotiate with corporations looking to relocate.

Economic development groups and corporations in favor of corporate subsidy packages like to argue that giving more money to corporations now will actually result in cities taking in more tax revenue in the future, by increasing the tax payer base. But giving away millions or billions to one company to create jobs limits the amount of incentives to other companies — it creates a system that makes cities or states too reliant on one company to provide jobs. Then, if that company shuts down or relocates somewhere else, that city is screwed.

Offering corporations giant subsidy packages is as American of a business practice as celebrating the entrepreneurs who start their companies in their parents’ basement. Before it was Amazon who was vilified for taking subsidies, it was Walmart.

One of the worst parts of this whole song and dance is that offering a higher subsidy package doesn’t typically get cities or states better or higher-paying jobs. Just take a look at the amount the state of Wisconsin was willing to offer Foxconn to open a factory in the southeast corner of the state — upwards of $230,000 per job. That’s over $200,000 more than Arlington is paying for 25,000 Amazon jobs with an average annual wage of $150,000. Foxconn is supposed to bring 13,000 jobs with an average annual wage of $53,875. Wisconsin was willing to offer Foxconn all this despite its history of reneging on job promises.

And if Wisconsin’s Foxconn deal looks bleak to you, take a look at the hundreds of millions of dollars states and cities are still willing to offer companies like Facebook, Apple, and Amazon to build datacenters, despite a host of case studies showing that datacenters don’t have that profound of an effect on the local economy.

The reality is, governors and mayors find short-term job announcements too irresistible. They want to be able to say that they’re the ones who brought thousands of jobs to their respective state or city come election time — even if it means that, by the time they’re out of office, these deals won’t look so good anymore.

Derek Thompson over at The Atlantic wrote that the only way to stop this madness is to make large corporate subsidy packages illegal — either by Congress outlawing them outright (though that might not survive a legal challenge) or by state governments passing laws to tax incentive packages at such a high rate that they would no longer be economic viable.

I don’t think that that’s necessarily the only way to wean states off their reliance on corporate subsidy deals. I think it would be effective for a group of mayors and/or governors to collectively commit to not offering a subsidy package above a certain amount to companies (urbanist Richard Florida suggested that cities do this during the HQ2 bidding process). Outside of that, I think it would take some political challengers basing a campaign against incumbents on a pledge to not offer corporate subsidy packages — and winning — to really make politicians think twice.

It’s also worth noting that it’s not just the tax breaks offered by cities and states that need reforming — it’s also the terms offered by cities once these companies set up shop. According to Amazon’s deal with Arlington, state officials and the Virginia Economic Development Partnership will “actively encourage” authorities in the state to adopt a “forward-looking regulatory framework” that when, possible, “limits prescriptive conditions on the design and operation of new technologies or devices.”

Arlington is also supposed to alert Amazon when there’s a Freedom of Information Act request so that Amazon can “limit disclosure … to the maximum extent permitted by applicable law.” In New York, Amazon won’t have to adhere to the public land review process that other corporations do.

As the HQ2 process proves, it doesn’t matter how much a corporation is truly “in need” of subsidies, or how many warning signs a company gives off that they’re not going to provide the number of jobs initially promised. If the leadership in places with plenty of jobs like New York City and Washington, D.C. are going to give millions in tax breaks to a company like Amazon, who won’t the mayors of smaller cities in greater need of jobs give subsidy packages to? The only thing that will put an end to this practice is a collective, sustained movement against corporate subsidy packages.