Still Not Ready for Prime Time: Tax Reform and Dynamic Scoring

Representative Dave Camp, the Republican Chairman of the House Ways and Means Committee, recently unveiled his long awaited comprehensive tax reform proposal. It is one of the very few serious congressional tax reform proposals in several years (Senators Wyden and Coates are probably the only others with a serious tax reform plan in recent years). Like any proposal of substance, there is much to like in Camp’s proposal and much to dislike. Much has been written on what is good and bad about it and more will undoubtedly follow. In this post, I’ll focus on the macroeconomic growth outcomes of the plan which are, after all, the whole point of what is supposed to make politically difficult “tax reform” worth it in the end. The bottom line is, estimates of the plan’s macroeconomic outcomes mostly tell us that dynamic scoring models are still not ready to be used as guides to policy.

The Camp plan reduces the statutory tax rate on corporations from 35 percent to 25 percent, reduces the statutory tax rate for most individuals (the top tax rate is reduced from 39.6 percent to 35 percent—a 12 percent reduction—but the rate reductions are far from simple to quantify), and eliminates many tax loopholes in order to reduce tax rates (all under the current Washington mantra of “lower rates and broaden the base”). The plan is revenue neutral over the next 10 years, but most likely increases deficits after that, because of various gimmicks that shift tax revenue inside the 10-year budget window. In addition to revenue neutrality, the plan aims to be distributionally neutral as well. (Non-economists may well wonder what the point of a revenue and distributionally neutral tax reform could possibly be. The answer suggested by many economists is the tax rate reductions made possible by tax reform will dramatically increase economic growth and employment.)

Little has been written on the macroeconomic analysis that accompanied the proposal, except for a piece by the Wall Street Journal editorial board, a group who, quite frankly, are not known for careful, thoughtful, or unbiased analysis of economic issues. They got their “facts” from the Ways and Means Committee GOP majority’s website. The summary of the proposal states that this plan will “make our economy stronger” and result in $3.4 trillion in additional economic growth and 1.8 million new jobs.

It is worth taking a close look at these numbers, particularly since they derive from an exercise in “dynamic scoring” undertaken by the Joint Committee on Taxation (JCT). Dynamic scoring of tax cuts lets policymakers account for the greater economic activity potentially spurred by these cuts, thereby reducing the cuts’ overall costs. In theory, there is nothing wrong with dynamic scoring. But in practice, empirical estimates of the macroeconomic impact of tax reforms have proven nowhere near precise enough to use to make informed policy. The results critically depend on the value chosen for parameters, and assumptions on future monetary and fiscal policies—different parameters values and assumptions can produce a very wide range of results for the economic effects of policy changes.

The nonpartisan JCT produced a macroeconomic analysis of the plan using two models: their macroeconomic growth model (or “MEG” model) and a complex overlapping generations model (or “OLG” model). These “black box” models are too briefly summarized by JCT staff. The necessary first step to make these models relevant to policy is to expose them to the close scrutiny of the research community. The lack of transparency so far argues strongly that even this necessary first step hasn’t been taken.

This lack of transparency is even more troubling because the two models produce very different results. The large numbers cited by the Wall Street Journal and the Ways and Means GOP website come from the OLG model—the results from the MEG model are largely ignored.

So let’s start with the OLG results trumpeted by Mr. Camp and the Journal. The JCT model indicates that under these results, GDP by 2023 would be 1.5 percent higher than it would be without the plan—about $240 billion in today’s dollars. Employment would also be 1.4 percent higher in 2023—about 1.8 million workers in today’s economy. The Ways and Means Committee’s claim of $3.4 trillion in “additional growth” has the strong potential to mislead—they are actually adding up the difference between nominal GDP (not inflation-adjusted) forecast by the OLG model under the Camp plan and nominal GDP forecast without it for each year between 2014 and 2023. But, in 2023, GDP is absolutely not $3.4 trillion larger under the Camp plan. In fact, in inflation-adjusted terms, it is not even a tenth this amount higher, and so it seems wrong to say that the Camp plan would lead to this much extra growth in GDP in the decade—a point made by Chad Stone of the Center on Budget and Policy Priorities. Mr. Camp, therefore, was being misleading, at best, and intellectually dishonest, at worst, when he claimed that the plan “could increase the size of the economy by $3.4 trillion—that’s equivalent to 20 percent of today’s economy.”

JCT states that the positive growth effects of the Camp proposal are due primarily from its effects on labor supply and saving. In a comprehensive examination of the complex models used for dynamic scoring, Jane Gravelle of the Congressional Research Service concluded the “effects of tax cuts on capital income can be large in these more complex models, reflecting shifting of consumption and leisure to periods far in the future. These shifts, which can induce large short-run increases in labor supply and saving, are generally not supported empirically and may be unlikely. One question is whether the benefits of formal theory in these models outweigh their empirical weaknesses.” In other words, the models may be elegant from the perspective of economic theory but don’t seem to capture what goes on in real life.

Now, all else equal, increasing GDP by 1.5 percent would be great, to be sure. But it was not the only estimate of the economic effect of the Camp tax reform plan produced by JCT. The largely ignored MEG results (from the same nonpartisan JCT that produced the large number referenced above) show the proposal would increase GDP by about 0.1 to 0.5 percent. This is equal to an additional $17 billion to $86 billion in today’s dollars (2014 GDP is projected to be about $17.3 trillion), which would be associated with 400,000 additional jobs by 2023, or about the number of jobs created in a good month. This result suggests that the Camp tax reform proposal would have a negligible effect on the economy and the labor market. These results are a big difference from the OLG results. The question is which results should we believe?

Since there is little available from the JCT assessment to help make this judgment, we can turn instead to past instances of tax reform. Fortunately, one particular case seems directly relevant to assessing the Camp plan. In 1986, President Reagan signed the Tax Reform Act of 1986 (TRA86) into law, which was a revenue neutral tax reform proposal that reduced the top tax rate from 50 percent to 28 percent (a 44 percent reduction). Two noted economists, Alan Auerbach and Joel Slemrod, reviewed the voluminous research literature on the economic effects of TRA86. They conclude that the “estimated behavioral responses suggest that the overall efficiency effects of changes in marginal tax rates facing labor supply and saving were small.” If the large marginal rate changes from TRA86 did not do much to increase economic growth, then it’s doubtful that the relatively small rate changes in the Camp proposal could result in the large effects predicted by the OLG model.

Overall, the results from the OLG dynamic scoring model showing that the Camp tax reform plan will significantly increase economic growth and employment should not carry much weight. The Camp plan crystallizes the inside-the-Beltway obsession with lowering tax rates, even in the face of substantial evidence that lowering these rates provides very little economic gain.

Stepping back from the weeds of dynamic scoring, it’s worth focusing on other issues not addressed by the plan. It does not increase revenues (which any serious fiscal plan should do), it does not reduce the budget deficits that preoccupy so many policymakers, and it does nothing to address the staggering rise in income inequality in recent decades. And against this, it cannot be credibly claimed to increase economic growth or employment nearly enough to justify its downsides.

Reforming the tax system just for the sake of reforming the tax system just does not cut it when our infrastructure is crumbing, our schools are underfunded, teachers are underpaid and overworked, government research is being slashed, and budget deficits are projected to soon start growing again. It is unfortunate that congressional leaders are encouraged to devote so much effort to produce tax plans that fail to address America’s many fiscal issues, and then to sell the plan with the results of a deeply uncertain exercise in dynamic scoring.