Modeling Taxation and Growth to Assist the Tax Reform Effort
By James Carter
Note: This article was originally published by Tax Analysts in the publication, “Tax Notes” and is reprinted here with permission.
Senior members of Congress’s tax-writing committees have launched a serious effort to reform the individual and corporate income taxes. Enthralled by the promise of tax reform to make the code simpler, fairer, and more conducive to economic growth, they seek, as former Secretary of the Treasury William Simon once put it, “a tax system which looks like someone designed it on purpose.”
While most lawmakers are willing to pursue revenue-neutral tax reform, a small but highly influential contingent on Capitol Hill refuses to consider tax reform unless it is accompanied by a bipartisan agreement to increase federal revenue significantly above the current law baseline. This would seem the end of tax reform.
If we can’t have tax reform without additional revenue, and if we can’t unlock the jobs and income growth that comes with tax reform without tax reform, we need a reform that can meet both objectives. But what would it take to get both jobs done? In my view, economic growth is the key.
Without decent economic growth, it will be much harder to bring the long-run budget deficit under control. The economy is currently lagging about 15 percent below the growth trend of the last thirty years. Income growth is not adequate, and is reminiscent of the 1970s. The shortfall in income will cost the Treasury about $6 trillion in lost revenue over the next decade. To close the gap requires a better business climate, including a different tax structure.
We can no longer afford to lurch blindly from one set of tax policies to another without determining the impact the proposed tax changes will have on the growth of the economy. That involves determining what the tax changes will do to investment, employment, and wages, and what the resulting change in income will do to tax revenue under the revised tax base. Anti-growth “pay-fors” that would shrink GDP and AGI would not only collect less revenue than expected and raise the deficit, they would reduce the GDP against which the deficit is measured, making the deficit bigger as a share of GDP.
Recognizing and modeling the long run growth effects of the right kind of tax (and spending) changes — and the GDP losses from the wrong kind — would have several advantages. It would steer us away from ineffective, short run counter-cyclical “pump-priming” rebates (or stimulus spending) that should by now be completely discredited. It would lead to more helpful or less damaging tax changes, leaving people with higher incomes than under random, less benign tax changes. It would reduce the dollar amount of the static tax increase needed to achieve any given revenue target, increasing the number of winners and reducing the number of losers, making the reform easier to pass. It would more correctly reveal the full cost of government spending, which is the sum of the direct outlay and the dead weight economic loss from the tax imposed to pay for it. It would shift the focus of the reform effort away from the parochial concerns of Washington to the broader concern of what is best for the country.
In deciding on reforms to the tax system, we should not rely entirely on the revenue estimates produced by the Joint Committee on Taxation because they routinely omit the macroeconomic growth effects of tax changes. Former Senate Finance Committee Chairman Bob Packwood (R-OR) highlighted this years ago when he asked the JCT to estimate how much revenue a 100 percent tax on income over $100,000 would raise. The JCT dutifully replied that the measure would generate $289.1 billion the first year, $314.4 billion the second year, and ever more in subsequent years.
Admittedly, the JCT produced that score nearly twenty years ago. But as the Congressional Budget Office explained just last May in a letter to House Budget Committee Chairman Paul Ryan (R-WI), little has changed. CBO reiterated: “Conventional estimating assumptions hold overall economic activity constant.” (To be fair, the U.S. Treasury employs the same convention.)
The letter went on to explain that CBO and JCT would “relax” those conventional assumptions when scoring a comprehensive immigration bill (as they did with the Comprehensive Immigration Reform Act of 2006). Indeed, the JCT routinely produces macroeconomic analyses of major pieces of legislation. It’s just that these analyses are routinely ignored when it comes to producing the official score. As Thomas Barthold, the JCT’s staff director, testified in 2011: “The macroeconomic analysis we do is not part of scoring for Congressional scorekeeping and rule purposes.”
This is not a matter of poor staff work. The JCT is staffed by dedicated public servants known for their professionalism. The scoring rules that bind the JCT simply do not allow macroeconomic analyses to be used for anything more than supplemental information.
Even when the JCT attempts an estimate of the growth effects, it is tied to a limited number of models whose assumptions and performance have not been sufficiently tested by experience or vetted by the research community. And yet, the JCT’s estimates convey an unwarranted air of precision and “truthiness.” (For example, the JCT recently estimated that cutting the U.S. corporate rate to 25 percent would reduce federal revenue $1,233.7 billion over ten years. This brings to mind the late Edgar Fielder’s quip that economists express their economic growth “projections to the nearest tenth of a percentage point to prove they have a sense of humor.”)
Again, to be fair, the JCT freely admits that their estimates are only as good as the assumptions underlying them, but their point estimates hold sacred status on Capitol Hill. To quote Senator Chuck Grassley: “Do you question the work of the Congressional Budget Office and the Joint Committee on Taxation? Well you shouldn’t because they’re like God around here.”
Nor can we rely on the table of tax expenditures generated by the JCT, or the somewhat different presentation offered up by the Treasury, to be some kind of shopping list for the Ways and Means Committee and Finance Committee to pick from, for three reasons:
First, the cost estimates attached to the tax expenditures counterfactually assume no change in macroeconomic behavior.
Second, many of the items on the list represent “loopholes” relative to a particular vision of an ideal income tax, but are natural and integral parts of a more saving-consumption neutral, and potentially more pro-growth tax systems, such as the Bradford X tax or a cash flow tax. Until we know the growth effects of the alternative tax systems and the roles of these “loopholes” that are natural features of them, we cannot make an informed decision what type of tax system to prefer, and whether to keep the tax expenditures, expand them, or trade them in for lower tax rates.
Third — and this an extension of the previous point — the entire exercise assumes it is “normal” to double, triple, and quadruple tax saving and investment.
I would note that the JCT makes it clear that tax expenditure estimates are not revenue estimates. This is a critical distinction often ignored by some policy makers.
We need meaningful briefing material for the tax writing committees and the Congress to enable them to select a set of tax reforms that will lift incomes on a permanent basis by encouraging long run growth. The question is, where is such briefing material to come from? There is an extensive tax literature on the inner workings of the income tax, and on the concepts behind various alternative tax bases. However, there is less literature on what tax changes have done for, or to, the economy in the past. This brings us to the question of what information is available to guide the tax reform discussion.
Economic forecasting models used to predict GDP over the next few quarters are not necessarily geared to simulating the long run effects of changes in federal tax policy (or spending policy) on economic growth and gains in potential output or economic capacity. Current tax simulation models may focus more on who pays the tax, assuming no change in GDP. What is needed is a means of forecasting the effects of the tax changes on capital formation, labor force participation and hours worked, production, government revenue, and federal safety net spending, including the interactions among them.
Before the Committees report out their tax reform drafts, it would be helpful to have a frank discussion of how taxes affect GDP, what the different tax simulation and economic models are based on, what their assumptions and capabilities are, where they agree and disagree, and what we need from them to assist the Congress in its tax reform efforts. The models can be useful in guiding the reform if they can be made to reflect the real world, and if the Members of Congress can be convinced that they may be relied upon. They can help to design a beneficial proposal, and evaluate whatever the Committees propose.
The issue of economic growth in the context of tax reform has become urgent. All concerned state that growth is a major objective, but those designing the reform exercise have imposed some unfortunate restrictions that may render the growth objective unattainable. For example:
Spending cuts appear to be off the tax reform table.
The overall tax reform package must be revenue neutral, not just budget neutral. Furthermore, the corporate and individual tax reforms must each separately be revenue neutral, with all changes to be paid for with offsetting revenues from within their own areas.
The reform must be distributionally neutral, at least, or even more progressive than the current system.
The reform must arrive at rates on subchapter S corporations and other pass-throughs that are comparable to the rates on subchapter C or “regular” corporations to be organizationally neutral.
And all of this must be accomplished in a static sense before the inevitable economic repercussions.
Meeting these requirements is a tall order that may neutralize the entire effort. Let’s see why.
Disallowing spending cuts is a big obstacle to reform. “Corporate welfare” in energy, transportation, and agriculture are under the jurisdictions of their own separate Congressional Committees, but this is not a good reason for them to be out of bounds. Even those spending areas within the jurisdiction of the Senate Finance Committee and House Ways and Means Committee, primarily the retirement and health care entitlements, are to be dealt with separately from tax reform. Yet the JCT and many other researchers have found that some types of spending restraint do less economic damage in connection with tax rate reductions than any other form of budget offset.
Why revenue neutrality? This may be a bow to political reality, in that many Members of Congress may be unwilling to vote for a bill that might risk any widening of the budget deficit. Or perhaps it is an indication that the reformers are more interested in protecting federal revenue and spending than in achieving optimal growth of incomes for the people. In either case, it is in sharp contrast to the eagerness with which the Congress recently spent one-and-one-quarter trillion dollars on stimulus spending, with no budget offsets and no measurable economic gains to show for it. This is an amount, by the way, that would have completely funded a corporate rate cut to 25 percent for the entire ten year budget window.
Sectoral neutrality is a problem because there are not enough true “loopholes” in the corporate arena to fund the desired rate cuts to 25 percent on a revenue neutral static basis without additional sources of funding. To make this work, the effort must rely to some extent on the understanding that there would be significant real growth with an attached revenue feedback from taxes of all types, not just the corporate tax. This feedback must be credited toward the corporate rate cut.
Distributional neutrality? Is that to be measured before or after the growth-related gains in pre-tax income? One essential goal of tax reform is more jobs paying higher wages made possible by increased worker productivity. To achieve that, we need a tax system that is more neutral across economic activities, and less punishing of saving and investment relative to consumption. That goal may conflict with distributional neutrality in the short run, until all incomes rise due to added capital formation. It is far easier to achieve a reasonable distributional outcome if people recognize from the start that there will be gains in wages and employment at all income levels from a pro-growth reform.
The growth of the pass-through business form has moderated the double-tax sting of the corporate tax for those able to reorganize. Some view this as an erosion of the corporate tax base. Others may view it as a back-door approach to corporate tax integration. Before we tamper with the arrangement, it would be nice to know what the availability of S corporations has done to the cost of capital and the incentive to invest, and whether restoring a double tax is compatible with a pro-growth reform effort.
It is not too late to put the tax reform effort on a sound footing. Let us have a thorough discussion of what we can do with the tax system (and the spending side of the budget) to get the economy moving forward, and let the results of that conversation inform the legislative effort.
James Carter was a deputy assistant secretary of the U.S. Treasury under President George W. Bush and served on the staff of the Senate Budget Committee. Mr. Carter also serves on the Board of Directors of Associated Industries of Missouri.