Well, what did you expect? With 238 entrants and 20 finalists, the Amazon HQ2 location tournament resulted in a resounding victory for Amazon: Billions of dollars in subsidies and binders full of detailed information on the contestants. Plus, we got a surprise twist at the end, when Amazon announced it would choose two “headquarters” instead of one. Of course, I never thought that having two headquarters made economic sense (“Doesn’t that defeat the idea of a headquarters as a central coordinating site?” I asked last year), and the same is even truer when you have three “headquarters.”
Leaving aside how Amazon plans to coordinate three headquarters’ operations, the subsidies boggle the mind and insult our intelligence. Let’s lay out what we know about the subsidies so far, remembering that there are other subsidy elements that are likely to be discovered as things play out. That is what happened with Foxconn, for example: Its subsidies in Wisconsin were originally reported as $3 billion in state subsidies plus local tax increment financing (TIF). By June of this year, Good Jobs First was reporting that further subsidies plus a huge TIF award brought the total to $4.8 billion (Megadeals spreadsheet, June 2018 update; download here). Something is likely to up the total incentives Amazon will receive, above what we know today.
So, which cities got half of HQ2? Amazon split the project in half, with 25,000 jobs set to go to Long Island City, Queens, New York, and 25,000 to Crystal City, Arlington County, Virginia. This much leaked out the week before the official announcement, but the November 13th official notification added that Nashville would get a 5,000 job consolation prize (for $138.7 million in incentives) as well as the incentive packages from each of these three jurisdictions — well, some of the incentives, anyway. As with Foxconn, this announcement was rapidly followed by the discovery of new incentives. I’ll skip the various updates and skip straight to what is currently known.
New York city and state both provided large incentives to the company. State benefits comprise mainly $1.525 billion in Excelsior employment tax credits, plus another $325 million based on the size in square feet of the Amazon offices, or a total state package of $1.85 billion. New York City will provide a job creation tax credit of $897 million over 12 years, plus a partial property tax abatement of $386 million over 25 years, according to a Good Jobs First analysis of the city’s press release on the project.
But wait, there’s more! Good Jobs First reports that the city will also provide a subsidy known as a payment in lieu of taxes (PILOT) that could itself cost another $100+ million. Last, as far as we know, but not least, the project will be located in a federal Opportunity Zone, which will provide further, though not-yet-estimated, benefits to long-term capital investors like CEO Jeff Bezos.
Total so far: $1.850 billion + 0.897 billion +0.386 billion = $3.133 billion, and likely more. While Amazon wants to emphasize the cost per job of its incentives, it only considers the first of these subsidies in its public calculation. But with just these three programs, we are already at a cost of $125,333 per job. While this doesn’t sound horrible compared to some incentive packages we’ve seen recently, it completely omits that with such large numbers of jobs, there are diminishing returns in the value of each job due to the increasing likelihood of dumping thousands of workers and their families on a locality’s infrastructure and educational system. Further, as I predicted in January, it normalizes the use of aid intensities* above 100%: $3.133 billion/$2.5 billion = 125%! Mind you, this is the nominal subsidy, not at present value, but with the 10-year Treasury note at 3.08%on November 16, the proper discount rate will be in that (low) vicinity, as has been the practice of the Organization for Economic Cooperation and Development in estimating the present value of U.S. subsidies for over 20 years.
To add a final insult to injury, the Amazon site will be in a federal Opportunity Zone, but the company’s project is destroying one of the things that would be most welcome there, affordable housing. Politico (h/t Daily Kos) reports that the New York outpost of HQ2 will displace a planned 1,500 units of affordable housing from two developers.
In Virginia, a second $2.5 billion investment, 25,000 job facility will be opened by Amazon as well. There, the company will receive $573 million in job creation tax credits. Virginia Tech University also plans to open a new “Innovation Campus” less than two miles from HQ2/Virginia. The $1 billion campus has been considered one of the biggest draws for Amazon in its location decision, and Good Jobs First includes the entire $1 billion as a subsidy for Amazon. I disagree; building new educational infrastructure will provide benefits to the students that they will always possess regardless of who their future employers may be, so I see the company as unable to capture much of the $1 billion as a subsidy. A university is a great economic draw, but this extends far beyond any single employer. If we exclude the new campus as a subsidy, the aid intensity in Virginia is only 22.92%.
Combining the two locations, we now find, even without the new Virginia Tech campus, that Amazon will receive $3.706 billion in subsidies for the HQ2 project proper. For the combined project, that brings us to an aid intensity of 76.1% of the investment.
Long-time readers know it’s time for a comparison with how this would be treated in the European Union under its regional aid guidelines. First, no region in the European Union is eligible for a 76.1% aid intensity, not even the poorest part of Bulgaria. Second, HQ2 is not going to the non-existent U.S. equivalent of Bulgaria, but to two of the richest places in the United States. What would Amazon get in state aid (=subsidies) for locating such a facility in London or Paris? Not one penny. Rich regions can’t give investment attraction incentives, period. So the entire $3.7 billion and counting subsidy for the company would be disallowed if rational regulation of the bidding wars existed.
Happily, these subsidies have come in for a great deal of criticism around the country. As The New York Times editorialized, “New York’s Amazon Deal is a Bad Bargain.” And how does it know New York overpaid? The same way I recommended back in 2014, comparing to a similar deal. And there’s no more similar deal than the other half of HQ2 that went to Virginia. Seeing that New York paid more than twice as much as Virginia for an equivalent project, the Times rightly concludes that New York paid more than it had to. Whether much political resistance to HQ2/New York develops or not, it’s great to see the press analyzing these deals well.
To end on a down note, though, we need to recognize the harm the Amazon auction did to transparency. The company made the finalists sign non-disclosure agreements, although a couple did make it into the public eye (Newark, New Jersey, and Montgomery County, Maryland, two of the largest finalist offers, though it appears Pittsburgh topped them all). Hopefully some of the losers will now come forward. However, that’s not the point; we need real-time transparency if there is to be any democratic oversight of the multi-billion giveaways that look to become more common than ever.
* Aid intensity is a metric, first used in the European Union, that allows us to compare the size of subsidies regardless of the size of the project. It is calculated as subsidy divided by investment. An intensity of 100% at present value means that the government is paying the entire cost of the investment. Thanks to reader TM for pointing out that I should have clarified this in the original article.