You might think that tax cuts that pay for themselves had disappeared from rational discourse ages ago. If so, you’d be sadly mistaken. Instead, “rationality” has been redefined to make such tax cuts “true by definition,” at least in one economic model that’s playing an influential role in state-level politics. As gridlock in Washington intensifies, and midterm campaigning focuses attention on statewide races for senator and governor, a new report paints a damning picture of a free market economic model that’s been used to push failed conservative economic policies — especially on taxes — at the statewide level all across the nation.
By now, everyone’s heard of ALEC, the organization that helps push such policies in all 50 states, but the Beacon Hill Institute, a free-market think tank located at Suffolk University, is ALEC’s much less well-known co-conspirator, and its State Tax Analysis Modeling Program is no better known than any other economic modeling program. (Quick, name three of them!) But STAMP has come in for some fairly withering criticism over the years, and now the Institute on Taxation and Economic Policy has integrated that criticism with its own analysis to produce a devastatingly critical report on STAMP’s biases and failings.
The main purpose of STAMP has been to argue for lowering taxes and against raising them — at least for those in the high-income brackets, and as ITEP’s own Who Pays? report documents, state governments are already inclined in this direction: Every single state taxes those in the bottom income quintile at a higher rate than those at the top. Overall, ITEP finds that “effective state and local tax rates by income group nationwide are 11.1 percent for the bottom 20 percent, 9.4 percent for the middle 20 percent and 5.6 percent for the top 1 percent.” But STAMP bolsters arguments for moving states constantly in an ever-more regressive direction — as well as simply cutting back government altogether, even on seemingly basic functions that even business leaders see the need for, like education and infrastructure spending.
The report has the snappy title, “STAMP Is an Unsound Tool for Gauging the Economic Impact of Taxes,” and the title is an indication of the dry, no-nonsense approach you’d both expect and want from an organization striving to provide unbiased economic analysis. But the report also cites a couple of reports (here and here) detailing STAMP’s reality-challenged use against renewable energy policies, where the model has both exaggerated the costs and underestimated or even ignored the benefits of renewable energy.
STAMP’s problems start with the kind of model it is, a “computable general equilibrium” model, which ITEP explains “is grounded in a concept of perfect economic efficiency that bears little resemblance to reality.” Its lack of realism is amplified by its detailed structure as well, ITEP notes: “Moreover, the thousands of linkages between economic sectors built into STAMP are in many cases not well-studied and not subject to statistical testing.” But STAMP also embraces unrealistic models of government (treating it basically as a pass-through device, without any distinctive characteristics of its own) as well as individuals (wealthy ones care more about taxes than poorer ones do, and everyone cares about taxes more than almost anything else). In short, STAMP has taken core elements of conservative economic ideology and incorporated them into its basic framework of “analysis,” so that conservative policy results invariably come out, no matter what questions are put to it, or what data it is given. To see how this works, let’s consider some major points of ITEP’s analysis.
First, the overall nature of the STAMP model is flawed. “STAMP is even more limited than some CGE models in that it assumes the economy is not only perfectly efficient, but also operating at its full potential,” ITEP notes, meaning there is no involuntary unemployment, and hence no need for government to do anything about involuntary unemployment. “Ackerman and Gallagher (2004) put this shortcoming into context: the fact that CGE models like STAMP find it ‘impossible to model unemployment and recessions’ is just one example of how ‘mathematical convenience … has … won out over realism about market imperfections’ in CGE modeling.” Of course, if you can’t model unemployment and recessions, it’s like having a model of the solar system that can’t model eclipses — or even sunrise, for that matter. It’s the very definition of “unrealistic.” ITEP goes on to quote another paper Ackerman co-authored, which pierces to the very heart of the absurdity:
In the world according to STAMP, the auto industry bailout of 2009 – or any other stimulus measure – couldn’t possibly save any jobs, because no one who wants a job is ever out of work. So why not save taxpayers’ money by letting the auto companies fend for themselves? Viewing public policy from this perspective, STAMP compares every proposed policy to an imaginary world of full employment. If you think you’re starting from the top of the mountain, there’s nowhere to go but down.
Next is the problem of how the model is fleshed with data:
STAMP seeks to quantify how roughly 6,000 relationships (or “flows”) are playing out between roughly 80 different economic sectors. Unfortunately, the quality of data available on these thousands of relationships often leaves much to be desired. Ackerman and Gallagher (2004) observe that CGE models are often “forced to rely on opaque and arbitrary approximations of numerous poorly understood relationships….”
But Ackerman and Gallagher were talking about national CGE models. If the national data quality leaves much to be desired, what can be said about state-level data? In many cases, such data simply won’t exist. It will have to be “interpolated” — in a purely unbiased manner, of course!
Next are the problems with how STAMP characterizes the role of government. This takes two forms. First, STAMP mischaracterizes government spending, because governments spend money quite differently than households: “Most importantly, states typically direct more of their spending toward labor-intensive services than households do, and they confine more of their spending to within the state’s border.” Thus, state spending does more to support in-state labor markets as well as the state economy as a whole, in contrast with household spending. Second, STAMP takes a nearsighted view of public services. Quoting from a prominent critic of STAMP in Arizona, Alberta H. Charney, an economist at the University of Arizona:
Unfortunately, there is complete silence in the [STAMP] document about investment in human capital. They model nothing about the long-term effects of public education expenditures on human capital or on the supply of different labor skills. This is a major shortcoming for a model that is all about incentives, investment decisions and long-run production effects. Education is a major portion of state and local government expenditures and yet its role in long-run accumulation of human capital is totally ignored.
As a result, ITEP notes, “In STAMP, the services being paid for with taxes, at the margin, are simply unimportant to the economy. And tax plans that BHI admits would require eliminating as many as 20,000 positions filled by teachers, firefighters, construction workers, and other government employees are depicted as being of great benefit to state economies.”
Indeed, ITEP quotes from a 2013 BHI report opposing tax increases to pay for education and infrastructure spending in Massachusetts, in which BHI argued, “Infrastructure and education spending are important but beyond a certain level both initiatives meet head on with the law of diminishing returns … increased government spending on transportation and education is not only inefficient, but is also susceptible to politically vested interests, mismanagement, and cost overruns.”
STAMP also has no idea how state spending is reflected in employment. At different times it has equated state spending of $1 million with anywhere from one to 37 public sector jobs.
But perhaps the most glaring assumption driving STAMP’s conclusions is the assumption that taxpayers are hypersensitive to taxes, as ITEP notes:
STAMP is designed so that the individuals and firms contained within the model will engage in dramatically more productive activity if their taxes are lowered, and dramatically less if their taxes are raised.
Nowhere is this issue more apparent than in STAMP’s assumption of how workers will react to a change in income tax rates.
Indeed, for both individuals and businesses, STAMP assumes that taxes have significantly more impact on decision making than is indicated by the economic literature. Thus, it overestimates how much raising labor costs will lead to laying workers off — a useful assumption for opposing minimum wage laws.
STAMP likewise overestimates how much higher taxes will discourage high-income earners from working. Indeed STAMP assumes that high-income earners ($100,000/year and more) are almost three times as sensitive to increased taxes as low-income earners (under $25,000/year). Consequently, ITEP explains:
STAMP assumes that income tax cuts will lead to a relative boom in private sector economic activity among high-income taxpayers, while income tax increases will quickly dampen those same households’ work effort . In explaining why they made this assumption, BHI simply cites “the literature” and their own “professional judgment”
In contrast, the Congressional Budget Office assumes a very modest decrease in sensitivity from low-income to high. The reasons for these very different assumptions are as revealing as the assumptions themselves. ITEP reports:
CBO, by contrast, released a thorough review of more than two dozen scholarly articles—thirteen of which were published after BHI first released their current elasticity estimates in 2004.
So there you have it, in a nutshell: ideological assumptions on the one hand, telling BHI that tax cuts on high earners will automatically grow the economy vs. empirical evidence used by CBO, saying, no, not so much.
So what does the historical record say? The report cites examples in Arizona, Massachusetts and Rhode Island where STAMP’s projections of the impact of tax changes have been wildly out of line with more standard models. But in the case of Kansas we have more than contradictory models to point to:
In Kansas, a STAMP analysis recently concluded that sizable income tax cuts enacted in 2012 will result in the creation of between 33,430 and 41,690 new jobs (Davidson et al., 2012). While it is too early to know what the precise economic effects of this “real live experiment” in tax policy will be, the results so far cast serious doubt on STAMP’s conclusions. Leachman and Mai (2014) observe that growth in Kansas jobs, incomes and business establishments has lagged the country as a whole following the implementation of the tax cuts. These trends have not gone unnoticed inside or outside of the state. In April 2014, Moody’s Investors Service downgraded the state’s credit rating, citing both the tax cuts and the lack of robust economic growth in their wake.
Kansas is also now facing a massive budget deficit with revenues down $480 million so far this fiscal year: The tax cuts did not pay for themselves, after all. The state Supreme Court has ordered lawmakers to restore funding to poor school districts, saying the spending levels they set were unconstitutionally low, and as Chris Hayes reported as part of his Kansas series on “All In” last week, small town schools are starting to close entirely. Because education, who needs it, right? Education spending is “is not only inefficient, but is also susceptible to politically vested interests, mismanagement, and cost overruns.” All in all, Kansas is better off without it, don’t you think?
The Kansas tax cuts were just what STAMP ordered. A separate analysis ITEP provided showed that the state went from a typically regressive tax structure — with the top 1 percent paying 5.5 percent in state and local taxes compared to 9.7 percent for the lowest 20 percent — to one much more regressive structure: 3.6 percent on the top 1 percent compared to 10.4 percent on the bottom 20 percent. Before the tax cuts, low-income Kansans paid taxes at 180 percent of the rate paid by the top 1 percent. After the tax cuts, they paid taxes at 290 percent of the rate paid by the top 1 percent. This is precisely what right-wing economics calls for. The resulting economic boom? It’s never shown up before, why should it now?
In fact, ITEP points out, this reflects all too well the record of what happened in the 2000s, quoting from a report from the Center on Budget and Policy Priorities:
Of the six states that enacted large personal income tax cuts in the years before the recession, three states saw their economies grow more slowly than the nation’s in subsequent years, and the other three saw their economies grow more quickly. The three that grew quickly are all major oil-producing states that benefitted from a sharp rise in oil prices in the years after they implemented their tax cuts.
In the end, ITEP’s conclusion is quite blunt:
STAMP’s flimsy foundation, biased assumptions, and highly questionable results are ample reason to avoid using it as a tool for understanding how changes to a state’s tax system will affect its economy. STAMP is designed in such a way that it almost invariably portrays tax cuts as being good for state economies, despite the fact that more mainstream economic models, academic studies, and states’ actual experience with tax cuts do not support such a finding.
But, if your whole point is to destroy government, then STAMP is exactly what you want. Just ask ALEC.