Last month, the state of Delaware announced it was giving Amazon $4.5 million to help set up a fulfillment center that would create 1,000 jobs. This was a more expensive deal than it looked: Delaware was to pay cash, even though most incentives are tax abatements over 10 or more years.

It wasn’t clear which other locations Delaware was competing against. Of course, cities offered much bigger incentives in fall 2017 when Amazon, instead of quietly shopping a few locations for its new headquarters, publicly announced it was accepting proposals from cities in a call that included 23 mentions of “incentives.” (The Washington Post is owned by Amazon chief executive Jeff Bezos.)

Voters don’t seem to pay much attention to the specifics of the deals politicians offer to attract businesses to their cities or states, except when things go bad (as they did with Wisconsin’s offer of billions to Chinese manufacturer Foxconn). However, as we argue in our book, “Incentives to Pander,” politicians use these incentives to take credit for investments, even if they weren’t very important to decisions.

This has created a complicated political economy in which political consultants, and sometimes government agencies, generate revenue from these programs. Here’s what you need to know about the incentive programs that localities use to try to attract businesses.

Officials say incentive programs create a ‘prisoner’s dilemma’

The unusual feature of Amazon’s program wasn’t that it tried to extract big incentives from local governments. It was that it did so publicly. Many Amazon HQ2 finalists actually hid their bids from the public, using public records exceptions or sending their proposals through nongovernment bodies such as local chambers of commerce. Amazon required states to sign nondisclosure agreements, but the fact that cities were openly competing over Amazon wasn’t a secret.

Because incentive programs are often shrouded in confidentiality, we often find out about them only when things go wrong — and there’s a public investigation. For example, a scathing audit of New Jersey’s economic development programs recently led to public hearings. However, many state and local officials argue that incentive programs need reforms. They call the incentive war a “prisoner’s dilemma” — a competition in which everyone ends up worse off.

According to this analysis, businesses are in the prime position, because they have the information and the bargaining power. Cities and states are forced into a ruthless competition for capital, so that they have to pay large amounts of money to attract businesses to locate their activities there.

But it turns out to be more complicated

This sounds like a standard political economy story, in which businesses have the bargaining power and cities and states have to pay to attract them. But the actual story is more complicated.

First, it isn’t clear that incentives drive businesses’ location decisions. In the case of Amazon’s headquarters, Bezos already had his short list of locations. He chose to locate in Virginia and accept a much smaller incentive than had been offered to locate a few miles away in Maryland, where the Amazon deal would have been over $8 billion.

Thanks to the New Jersey investigation, we know that economic developers knowingly offered incentives to companies that they knew were coming to New Jersey anyway. In Georgia, the economic development agency vastly overstated the benefits of its film incentive program and were willing to pay enormous incentives to attract activities of questionable economic value.

Incentive programs have generated their own clandestine political economy

What this tells us is that incentive programs don’t just put states and cities at a disadvantage vis-a-vis business. They can sometimes generate their own hidden political economy. The combination of high secrecy, high discretion and big budgets sometimes allows well-connected insiders to cream off the benefits.

For example, the hearings and subpoenas in New Jersey revealed juicy details about how politically connected consultants and their client firms had literally written the incentive legislation. However, perhaps the most important finding was the role of the economic development agency in offering incentives to firms that were already operating in New Jersey.

To receive New Jersey incentives, a company had to claim it was willing to relocate out of state. The investigation revealed 12 companies all threatened to move to the same New York address to get $100 million in incentives. Agency employees were told to “not ask questions” and carried out no basic due diligence. The economic development agency had strong incentives to pass money along without raising awkward problems — it received a fee for the incentives it handed out.

Georgia offered close to $1 billion dollars in the “Georgia Film Tax Credit” program, offering movie companies a tax credit for up to 30 percent of their expenses, including high-paid actors, which they could then resell to other companies that paid taxes in the state. This isn’t a tax abatement; it is an indirect way to offer cash.

States such as Maryland, Massachusetts and Virginia have evaluated similar programs and found they are major loss makers. However, Georgia allowed a lot of money to circulate. The audit found the state was extraordinarily careless in record keeping and basic administration of the program, allowing companies to claim lots of expenses, including a company that had $150,000 in “petty cash” expenses.

Businesses and state officials alike benefit

This suggests the usual story about incentive programs is off target. It is true that there is some competition between states and localities for investment, but it is not a simple prisoner’s dilemma, in which clashes of self-interest leave everyone worse off.

Instead, some parties do very well, indeed. Claims about the competition for investment are sometimes little more than an excuse for spending large amounts of money in ways that benefit the agencies that allocate the money, the businesses that receive it and the specialized consultants who help businesses apply for it, sometimes under legislation that they themselves have helped to write.

Nathan M. Jensen (@natemjensen) is a professor of government at the University of Texas at Austin, a senior fellow at the Niskanen Center and the co-author (with Edmund J. Malesky) of “Incentives to Pander: How Politicians Use Corporate Welfare for Political Gain” (Cambridge University Press, 2018).