NYT: $80 Billion in State and Local Subsidies Annually (Updated)

In today's New York Times, Louise Story begins a series, "The United States of Subsidies," ten months in the making, with a story focusing on General Motors closures, the border war for investments between Kansas and Missouri in the Kansas City metropolitan area, and a new estimate of state and local incentives to business, $80 billion a year. Backing this up, and no doubt contributing to the long lead time, is a database of 150,000 state and local subsidy deals going back at least 20 years. Given its appearance in the country's newspaper of record, the series is sure to elevate the issue of state and local subsidies to a prominence it has never known before.

Since my 2011 estimate was $70 billion per year in total subsidies to business, and $46.8 billion in location incentives, the Times figure represents a substantial increase if accurate. Ever since David Cay Johnston reviewed my book when it first came out, he has argued that my $70 billion figure was probably an underestimate, and the new report would seem to back him up. Nevertheless, I will certainly be spending some time analyzing the database to see just what is in it. According to the story, $18 billion per year is accounted for by corporate income tax breaks, a whopping $52 billion by "sales tax relief," and the other $10 billion unspecified but most likely property tax breaks. I have some questions about these numbers, however.

First, it seems to me that property tax breaks likely exceed $10 billion a year. When California axed tax increment financing earlier this year, it was generating $8 billion in tax increment all by itself. Although California cities were by far the biggest user of TIF, municipalities in almost every other state still use it, as well as myriads of property tax abatements offered at the local level. Story is well aware of this. She writes:

The cost of the awards is certainly far higher. A full accounting, The Times discovered, is not possible because the incentives are granted by thousands of government agencies and officials, and many do not know the value of all their awards.

Thousands of local governments give subsidies, and these are overwhelmingly related to property tax. In my most recent estimate, there were several states in Missouriwhich local subsidies exceeded state subsidies, including Missouri and Michigan, so my default  assumption was that they were equal if I did not have adequate information on local incentives, as is usually the case due to the huge number of governments involved.

On the other hand, there is some chance that the $52 billion in sales tax subsidies could be an overestimate; it all depends on what The Times includes in this category. My own thinking about sales tax has changed since I first created the subsidy estimates in my 2000 book, Competing for Capital. My estimate for Minnesota, for example, included many hundreds of millions per year in sales tax exemptions for business services. Now, I tend to think of these tax breaks as methods to avoid tax cascading (paying the sales tax on a good more than once, by taxing the full value of every intermediate good) and not a subsidy at all. They have been removed from my estimate of total subsidies in my more recent work, which did not prevent my estimate for 2005 (published in 2011) from being $20 billion higher than that for 1995 (published in 2000). I do still count some sales tax breaks as subsidies, particularly those on plant and equipment, which apply to the initial investment rather than ongoing operations.

While this may seem like a sterile academic argument, in fact it makes a big difference whether incentives are $50 billion a year or $80 billion a year, approximately 600,000 public sector jobs paying $50,000 annually. The larger the true figure, the more pressing is the case for subsidy reform. The inauguration of this new series of articles, plus the database, will help us put a better number on the value, a critical first step toward galvanizing public opinion to force politicians to rein in subsidies.

I will be commenting more on this series over the course of this week.

UPDATE: Text corrected to reflect that although I had specific data for local incentives in Michigan, the total of local incentives was somewhat lower than that of state incentives. In addition, it is clearly true that TIF in California exceeded state subsidies, so obviously so did the total of local subsidies. However, I did not know this at the time I made the estimate.

Cross-posted at Angry Bear.