Comments Guidelines

All comments are pre-moderated. No spam, slurs, personal attacks, or foul language will be allowed.

Sunday, October 25, 2015

New Study Finds State Subsidies Go Overwhelmingly to Large Companies

Good Jobs First has just issued a new report analyzing state investment incentive programs open to small and large businesses alike. With the financial support of the Surdna Foundation and the Ewing Marion Kauffman Foundation, Shortchanging Small Business: How Big Businesses Dominate State Economic Development Incentives finds that 70% of the awards and 90% of the money goes to large companies. This is a big deal: The justification for many major incentive programs is that they benefit small business. This study is the first in a planned series of reports which show that this claim does not stand up.

If subsidy programs disproportionately benefit large businesses, they reduce market competition and thereby make the economy less efficient. As I discussed in Competing for Capital, subsidies to capital exacerbate income inequality (post-tax, post-subsidy). This effect will be magnified if the incentives are flowing primarily to large firms rather than smaller ones, as this new study suggests to be the case. The report's findings are relevant to the European Commission's ruling last week on Starbucks and Fiat, that subsidies created by tax havens harm the ability of small- and medium-sized enterprises (SMEs) to compete.

Shortchanging Small Business looks at 15 incentive programs in 13 states that are well-documented in Good Jobs First's Subsidy Tracker database, plus one Missouri program that is highly transparent online (and will soon be included in Subsidy Tracker), for a total of 16 programs in 14 states. Overall, these programs account for 4228 individual awards allocating over $3.2 billion.

Note that these are not one-off deals for a large company: For example, as I showed in my special report on North Carolina incentive packages, the deal Google received from the state in 2007 was worth $140.6 million at present value to the company. This dwarfs the $26.4 million over six years given by the One NC Fund, included in this Good Jobs First report, and is only one of a number of megadeals in North Carolina.

Moreover, neither are they apparently open programs with criteria that in fact rule out small companies through the use of large job creation or investment requirements. No, the 16 programs considered in this report are all genuinely available to large and small firms alike; that is what makes this such an important study. This report excludes programs directed solely to small businesses, but Good Jobs First has promised a separate analysis of those generally poorly funded measures.

What, then, is a small company? For the purposes of this study, it has to have fewer than 100 employees, it has to be an independently owned local firm, and it must have fewer than 10 establishments. If a company does not meet all three criteria, it is classified as a large company. Note that this cutoff is considerably below that of the U.S. Small Business Administration, which for most industries is 500 employees. On the other hand it is larger than the European Union definition of a small enterprise (50 workers) but smaller than the EU definition of a medium-sized enterprise (250 workers).

Despite the fact that small companies are theoretically eligible for the 16 programs analyzed, they receive only 30% of the awards and 10% of the money available through them. As I pointed out earlier, combined with one-off megadeals, programs that only appear to be open to small firms, and tiny programs specifically for small business, this adds up to a large bias in favor of big business, with all the consequences noted above.

What should be done? The report notes that many small businesses cannot benefit from the tax credit or tax abatement involved in the programs analyzed and, in a separate survey, many small business leaders said they would benefit more from public goods like job training, education, and transportation. Therefore, Good Jobs First proposes a reduction in incentive spending going to large companies, to be effected by using hard caps on each program's spending, on cost per job, and on the total amount any one company can receive under a given subsidy policy. While such caps are unusual in the United States, they are the main basis for the European Union's successful control over incentive spending there, elaborated further to have higher caps in poorer regions and a cap of 0 in the richest EU regions. In addition, the caps proposed by Good Jobs First could be augmented by using an EU metric known as aid intensity, which is simply the subsidy divided by the investment. While a cost per job cap is useful at resisting excessive capital intensity, an aid intensity cap is a valuable metric when substantial jobs are created but the government is paying for virtually the entire cost of the project (for example, Electrolux in Memphis).

I'm looking forward to further extensions of this research, and you should, too.

Cross-posted at Angry Bear.

Thursday, October 22, 2015

EU slams Starbucks and Fiat advance tax rulings as state aid; Is Apple next?

The European Commission decided two of its major tax subsidy cases on Wednesday, October 21, and the rulings could not have been worse for Starbucks and Fiat (h/t Chillin' Competition). These cases can be seen as a barometer of what is to come in the legally similar but much larger case of Apple, where potentially billions of euros could be at stake.

The gist of the three cases is that tax haven subsidiaries of each company (Starbucks in the Netherlands, Fiat in Luxembourg, and Apple in Ireland) were given advance tax rulings by each country that were so removed from economic reality as to constitute illegal subsidies ("state aid" in Euro-speak) under EU competition law. In the Commission's decision, it was emphasized that the artificially low tax bills created by the rulings gave them an unfair competitive advantage over competitors, especially small business ("small and medium-sized enterprises" or "SMEs" in Euro-terminology).

Since the alleged subsidies were not notified to the European Commission in advance as required by EU law, the Commission has ruled that Starbucks and Fiat have to repay the illegal aid to the granting countries, with interest. The decision states that each company will owe 20-30 million in aid repayments.

 Of course, both of these cases will be appealed to the European court system, all the way to the Court of Justice of the European Union (CJEU), the highest court in the EU. Tax haven shenanigans are built into the economic structure of both Luxembourg and the Netherlands, and the two countries will do everything they can to maintain the status quo. The Apple case is much bigger, because it goes back all the way to 1991, and some estimates have put its tax savings at billions per year. If Apple loses, and I think it will, we can again be assured that the case will be appealed to the CJEU.

If the Commission makes these decisions stand up on appeal, it will dramatically change the shape of tax havens in Europe (including Switzerland, which the EU holds as being subject to the state aid rules through its free trade agreement). It won't put them out of business, because the decisions pertain to corporate income tax rather than personal income tax, but the amount of revenue lost on the corporate tax alone is a very big deal.

Cross-posted at Angry Bear.

Friday, October 2, 2015

Time for a new incentives estimate!

It's been over four years since I last published an estimate of U.S. state and local government subsidies in Investment Incentives and the Global Competition for Capital, and the data there represented the situation circa 2005. So I think it's time to begin a new estimate, don't you?

We have a great place to begin. As you may recall, in December 2012 Louise Story of the New York Times published a series of articles, "The United States of Subsidies," which was accompanied by two searchable databases, one of individual deals and one on subsidy programs. The latter was far more comprehensive than anything previously published.

Unfortunately, the program database was flawed because $52 billion of the reported $80 billion in annual subsidies were for sales tax breaks that were for the most part not subsidies at all, but ways to prevent sales tax from being charged for inputs into goods that would pay full sales tax on their final sale. Upjohn Institute economist Timothy Bartik documented that $4.80 billion of the $4.83 billion in sales tax "subsidies" in Michigan fell into this category. There were smaller errors in the program database as well, such as wrongly including net operating loss provisions and omitting the cost of single sales factor apportionment (h/t Timothy Bartik).

Now, dear reader, I'd like to solicit your help. The first step is to review the sales tax breaks in the database, so as to include those relatively few that are genuinely subsidies, such as exemptions for capital goods, which by nature are targeted at investment. If you have insights on this for any of the sales tax programs, I would be extremely grateful if you would share them with me. I would also appreciate any assistance on step 2, which is updating from the 2011 data Story used. If you have the numbers or know the sources, please send them to me. Finally, the same is true for step 3, gathering information on local subsidies, which is much sparser than state-level program data. I will, of course, acknowledge your help in any publications based on this work.

For this project, you can email me at

Thanks in advance!

Thursday, September 3, 2015

Obamacare hasn't killed full-time jobs, either

When we last looked at Obamacare as an alleged "job-killer," Matt Yglesias had just pointed out that 2014, the first full year of insurance on the exchanges, was also the best year for job creation since 1999. But recently a non-blogging friend reminded me of a related anti-Obamacare meme, the idea that employers have been cutting their workers below 32 hours per week so they would not have to provide them with health insurance. His argument was, logically enough, that this would mean a loss of full-time jobs.

As with so many other anecdotal Obamacare horror stories, this one does not stand up to even simple inspection. Just like total job creation, it turns out that full-time (BLS uses 35 hours/week, not 32, by the way) job creation has quickly increased since December 2013, just before exchange insurance went into effect. Not only that, part-time employment has fallen slightly. The Bureau of Labor Statistics' monthly "Employment Situation" (Table A-9 in both cases) tells the tale.

Date            Full or Part Time     Not Seasonally Adjusted Jobs          Seasonally Adjusted Jobs

December 2013   Full-time          116,661,000                                      117,278,000
July 2015             Full-time         123,142,000                                     121,589,000
Change                                      + 6,481,000                                      + 4,311,000

December 2013   Part-time           27,762,000                                        27,372,000
July 2015             Part-time          26,850,000                                        27,265,000
Change                                          - 912,000                                          - 107,000

I included both seasonally adjusted and not seasonally adjusted data for completeness sake, but when we are comparing a summer month to a winter month, surely the seasonally adjusted figures are the correct ones to use. For those of you keeping score at home, then, full-time jobs have increased by 4.3 million since Obamacare exchange insurance went into effect, whereas part-time jobs have fallen by 107,000. Neither of these fits the anecdotes of workers being shunted from full-time to part-time work to avoid providing insurance. This increase in full-time work has been accomplished in the span of just 19 months, or an average of over 226,000 new full-time jobs per month.

Of course, it's theoretically possible that using sophisticated statistical controls might uncover a hidden negative relationship; that we'd have even more full-time jobs than we do if the exchanges hadn't gone into effect. Even if that were true, it's obvious that everything else going on in the Obama economy is having a much bigger effect on full-time employment, so there's no justification for using the epithet "job-killing" on the off chance that it's true.

Cross-posted at Angry Bear.

Wednesday, August 19, 2015

Basics: Tax Increment Financing Subsidies

Last week I made a presentation to the Colorado Assessors Association on tax increment financing (TIF) subsidies. With the organization's permission, I am sharing the PowerPoint presentation for my talk, as well as adding this introduction.

The talk begins by putting TIF in the context of subsidies generally. As a subsidy, TIF is subject to the three main potential drawbacks of their use: Inefficiency, inequality, and environmental harm. These differing harms often call forth coalitions that prove the adage that "Politics makes strange bedfellows." One example was a joint appearance by Ralph Nader, Rep. John Kasich, and Grover Norquist in 1997. I was in a similar coalition in 2003 fighting a TIF that would have leveled downtown O'Fallon, Missouri, then the fastest-growing city in the state. Our leadership included people from liberal Democrats to future Tea Party members. Similarly, the mayor and TIF-supporting council members were bipartisan. It's easy to find single-party governments that abuse subsidies, too, whether it's now-Governor Kasich offering $400 million to move Sears from Illinois, or Democratic stronghold Kansas City, MO, offering multiple multi-hundred million dollar TIFs.

I then go on to discuss the legal particularities of TIF, which gives it another level of controversy. Tax increment financing typically allows governments to capture the property tax revenue of other districts (something especially hard on school districts), and these revenue fights are matched by the intense hatred the use of eminent domain usually brings forth.

The very first state to adopt TIF (in 1952), California axed TIF in 2011 due to its budget crisis, caused in part because the state reimbursed property tax revenue that school districts lost to TIF. Last year, a new version of TIF was passed, but it completely removes the option to take money from school districts, and lets other taxing jurisdictions opt out as well (see below).

Also of note in TIF history is the constant pressure to expand the locations able to use it (i.e., progressively less economically deprived areas want it in their subsidy arsenal) and the gross overuse of TIF for retail projects. As I mention in the link, St. Louis-area municipalities gave $2 billion in retail TIF from 1990 to 2007, creating all of 5400 jobs, no more than we'd expect on the basis of the area's income growth. That's why the relevant slide says $2 billion = 0 jobs.

Thanks to Karen Miller of the CAA for inviting me and agreeing to let me post the PowerPoint.

Cross-posted at Angry Bear.

Saturday, August 15, 2015

Final subsidy accounting rules published!

On Friday, the Government Accounting Standards Board (GASB) published the final version of its new rules requiring governments to make reporting on subsidies a standard part of their financial reports (known as Comprehensive Annual Financial Reports, or CAFRs). Since GASB determines the content of "Generally Accepted Accounting Principles," its new rules will have to be widely adopted.

Subsidy reformers such as Good Jobs First did not get everything they wanted in the new rules, but their publication represents a gigantic step forward in subsidy transparency. In particular, every government that gives tax-based subsidies will be required to report the amount they give each year. Moreover, every government that loses revenue to subsidies given by another government must report the amount as well. Most commonly, this will be school districts and other political units that lose property tax revenue to tax increment financing (TIF) or property tax abatements.

GASB continued its incomprehensible use of the term "tax abatement" as its catchall for all tax-based subsidies. As I pointed out in my own GASB comments, the term properly only refers to property tax abatements, a smallish proportion of local subsidies used in fewer states than TIF (see Greenbaum and Landers, "The TIFF over TIF," 2014, no ungated version available). On the plus side, as Good Jobs First points out, the final rules make clear that subsidies such as tax increment financing are included in their definition of "tax abatement," although a strict reading of the draft definition might well have excluded it and other subsidies. I considered this a worry in my GASB comments, because too much focus on legal fictions (TIF recipients are legally deemed to have paid their taxes, for instance) would obscure the fact that a subsidy exists. Happily, GASB opted to be inclusive in the definition, which will make it more difficult for governments to evade the new rules.

The real drawbacks of the new rules are that they don't require that the recipients, even the largest ones, be identified (this is optional); they don't require governments to say how many subsidies are in place (just the total cost of subsidies for that year); and they don't require a listing of future subsidies already committed. In addition, as I pointed out in my comments, they do not require a cross-listing of non-tax subsidies such as grants and loans so that people reading a CAFR can determine how much a government is spending and committed to spend on total subsidies.

Still, the new rules represent a tremendous advance over the status quo. Every state and local government will now have to report the value of the subsidies it gives every year. CAFRs are audited reports, meaning that multiple sets of eyes will look at the documents even before the public sees the reports and can cross-reference them with everything else known about a government's subsidies. Good Jobs First will be expanding its Subsidy Tracker database to include the new data in governments' annual reports. I'm looking forward to all these developments, and you should be, too.

Cross-posted at Angry Bear.

Wednesday, July 29, 2015

Basics: Is trade zero-sum between workers in different countries? had a long, interesting interview with Senator Bernie Sanders covering a large number of political and economic issues. In this post, I want to focus on just one issue he raised: Whether rising incomes for Chinese workers have to come at the expense of U.S. workers. Here is what Sanders told Vox's Ezra Klein:
I want to see the people in China live in a democratic society with a higher standard of living. I want to see that, but I don't think that has to take place at the expense of the American worker. I don't think decent-paying jobs in this country have got to be lost as companies shut down here and move to China.
What Sanders doesn't mention is that the market, left to itself, will indeed force a tradeoff between U.S. and Chinese workers. We can see this via the Stolper-Samuelson Theorem, which says that increasing trade will raise the real incomes of a country's abundant factors of production and reduce the real incomes of the scarce factors of production. The reason is that abundant factors of production (relative to the rest of the world, of course) will find new markets abroad as trade increases, while scarce factors of production will face increased import competition. Since China is a labor-abundant country and the United States a labor-scarce one, the theorem implies that real wages will rise in China and fall in the United States as they increase trade (all trade, not just with each other). And this effect can be sped up if U.S. companies close factories in the United States and open them in China, just as we have seen happen.

To disable the tradeoff requires political intervention in the market. If you want to preserve gains from trade that are predicted by the theory of comparative advantage, and you want to not worsen income inequality in the United States, you need to find a way, as Ronald Rogowski pointed out, for  the winners to compensate the losers from trade. This isn't easy: As Rogowski also noted, the winners increase their clout in the political system while the losers see their influence decrease (look at the long-declining influence of unions here). As I've discussed before, the increased mobility of capital exacerbates this problem in the U.S., since capital is much more mobile than workers. And so we have seen a steady decrease in the tax burden paid by corporations and the rich, more trade agreements signed, and a constant drumbeat to cut Social Security (despite the coming retirement crisis) and "phase out" Medicare.

What would compensating the losers from trade look like? Most obviously, and most focused, is trade adjustment assistance, which is often criticized as inadequate. Yet it does not really make sense to compensate only those who lose their jobs directly to foreign competition, because those workers then spill into other sectors of the economy, driving down wages as they go. Thus, we need to go beyond trade adjustment assistance.

To raise wages in the economy more generally, we need broader measures. One would be to raise the minimum wage: It pushes up workers' pay, but it also reduces turnover and training costs for employers, and puts money into the hands of people with a high propensity to consume, creating multiple channels to counteract the seemingly self-evident fact that raising something's price means people will buy less of it.

Another broad-spectrum approach to raising wages is to restore the power of unions. As I have pointed out before, the United States has the fifth-lowest union density in the 34-member Organization for Economic Cooperation and Development (OECD). Senator Sanders, in the interview linked above, notes that the increased power of unions in Nevada's gambling industry has enabled house-cleaning staff in the state's casinos to earn "$35,000 or $40,000 a year and have good health-care benefits." Having a National Labor Relations Board that is not in the pocket of industry is critical for us to see this take place.

Third, less targeted still but having the political benefit of universal coverage, an expansion of the social safety net would make it possible for people to simply refuse to take crappy jobs. Yes, this is about bargaining power! It would also encourage entrepreneurship because failure would not mean the loss of one's health insurance, for example. Medicare for all has long been one of Senator Sanders' standard prescriptions, a program that benefits from having far lower overhead costs (it avoids outrageous executive salaries, the need for profit, and does not have to advertise much) than private insurance. We could do a lot worse than considering it -- and we have.

Finally, to pay for these programs, it's necessary to raise taxes on corporations and rich individuals. Thomas Piketty, in his monumental Capital in the Twenty-First Century, suggests that the top marginal income tax rate should be 82% for individuals in the top 1/2% or top 1% of income. He notes that this will not raise much money, in part because it will reduce various lucrative but economically unproductive financial shenanigans. Instead, he thinks a tax of 50-60% on the top 5% of incomes would produce substantial revenue to create what he calls a "social state" for the 21st century. One could go further, of course, by adding a financial transactions tax (I hope to write about this soon) and shutting down tax havens.

To return to our original question, there is no reason that Chinese workers and U.S. workers can't both prosper from trade. But to make it possible in the United States requires a great deal of rule rewriting that will not be achieved overnight.

Cross-posted at Angry Bear.