Throughout a surprising part of last week and the week before, a claim was making the rounds that, in order to get Amazon's second headquarters, Illinois had promised to give the salaries of its workers back to the company. Here's an example.

Over at Crain's, Steven R. Strahler rounds up a few more examples; there are plenty more where that came from, and they flooded my timeline for days.

This is not exactly how it works. But it's interesting how people immediately seized on it.

What has been reported is that Illinois would offer EDGE tax credits to Amazon over a long period of time. These credits apply to Amazon's corporate tax bill, equivalent to 50 percent of what the employees whose jobs were created would pay in personal income tax, over a period of 10 years (though that can be rolled forward for up to 15 years). The credits can go higher than 50 percent if a company locates to somewhere with high unemployment or if it incurs job-training expenses, but it doesn't appear Amazon would be eligible for those additional credits.

That's pretty much it. While nothing has been finalized, Amazon's offer appears to be different than what the state's offered other companies; they're specifically designed so that all companies get basically the same deal. It's just that the scale is unprecedented.

After that, it's kind of a matter of interpretation. The Reader's Ben Joravsky wrote this (emphasis mine):

For example, say Amazon hires someone at $100,000 to program computers at HQ2. Normally that employee would pay $4,950 to the state for things like road repaving, policing, education, and so forth. With the current tax rate of 4.95 percent, the hipster essentially pays up to $4,950 to his boss at Amazon.

That worker's taxes go into the same bucket as Amazon's corporate taxes, and the amount that Amazon pays into that bucket would be reduced by half of that for up to 15 years. So is it essentially like paying Amazon? I dunno, ask Richard Thaler. But Illinois isn't cutting a $2,500 check per worker to the company, which is how Joravsky's column came to be read, in a bloggy version of the word-of-mouth game Telephone.

I decided to run the numbers, and came pretty close to what has been reported that the state is offering Amazon. I based it on this breakdown reported by the Trib:

In the Oct. 16 proposal to Amazon, the state lays out a 17-year schedule of tax breaks based on the company’s estimate that it would add 50,000 jobs over 17 years. The schedule assumes 2,500 employees in the first year making an average of $100,000 per year, and concludes at 50,000 employees earning an average of $160,000 per year in the 17th year. Such a payroll would make Amazon a larger employer than the city and would result in a tax break totaling $1.32 billion.

Doing a back-of-the-envelope calculation required some assumptions: that the number of jobs would increase the same every year to get from 2,500 to 50,000; and that the average salary would increase the same every year to get to $160,000.

A bigger assumption I made is that the difference between the personal income tax rate and the effective tax rate would be similar to what it is now. For all Illinois taxpayers who made between $50,000 and $500,000 in 2015, the effective tax rate was about 0.46 percentage points lower. So I just used that. As a result, my estimate is about $120 million higher than $1.32 billion that's been reported, which isn't bad given my very elementary assumptions. (It's not entirely clear why Amazon's credits would run over 17 years instead of 15; I'm waiting on an explanation.)

Here's what it would look like (in millions of dollars):

In theory, the EDGE credit is supposed to make business incentives simpler and more transparent than individually negotiated deals. Amazon gets a predictable credit; the state gets a predictable amount of tax revenue above the break it's giving Amazon. Tax breaks and revenue scale up or down depending on the number of jobs and what they pay.

In reality it's more complicated. Illinois is promising Amazon a big tax break in exchange for the predictability of up to 50,000 jobs paying twice that break in income taxes—jobs that would not have been created but for the tax break. So, free money, right? Given that Amazon is hunting tax credits, it's likely that those specific Amazon jobs would not be created without it. But could Illinois attract thousands of high-paying jobs without such a tax break? (Actually, the research on tax breaks is pretty inconclusive; if you can answer that question the world will beat a path to your door.)

One of the reasons corporate tax breaks are so appealing is that the upsides are obvious: The state clears $1.32 billion in revenue even with the massive break. The downsides are more opaque: If Amazon gets a break because it can offer so many jobs, what happens to its competitors—for talent, if not in exactly the same industry—who have to pay full fare? What are the benefits compared to the increased cost of accommodating tens of thousands of new workers? How much is left to benefit other residents? And what happens when the next company wants even more?

The tax-credit offer Illinois has extended to Amazon has gotten a lot more attention than such deals have in the past, in part because of the scale of the headquarters, and in part because people have accidentally seized on a different way of framing the offer. Nothing indicates that the tax break itself is appreciably different from the way the state has done business in the past and will continue to do so. That's not to say people shouldn't be raising questions about it, but the answers to them are very, very hard.