The views expressed herein are those of the author and are not presented as those of the Congressional Research Service or the Library of Congress.
Revenue neutrality is an often-stated goal of tax reform proposals. Typically, revenue neutral tax plans are marketed as “lowering tax rates and broadening the base” in such away that the opposing revenue effects exactly offset each other. This post looks at several approaches for attaining a revenue neutral tax reform. The approaches discussed below are not necessarily mutually exclusive or exhaustive.
There are around 200 corporate and individual tax expenditures that will cost the government approximately $1.2 trillion per year1. A little over $1 trillion of this revenue loss is attributable to the individual tax system, with the corporate tax system responsible for the remaining $150 billion. To achieve revenue neutrality, a significant amount (if not all) of these tax expenditures would need to be pared back or eliminated to offset the individual and corporate rate reductions that have been proposed in varies tax reform plans thus far.
Politically, it appears that deciding which tax expenditures should be kept maybe difficult. The ten most expensive individual tax provisions combined with the ten most expensive corporate tax provisions account for nearly 70% of the $1.2 trillion revenue loss. On the individual side, this top-ten list includes the exclusion for employer-provided health insurance, tax-preferred retirement accounts, and the mortgage interest deduction. These items are commonly referred to as “sacred cows” in the tax world, and harming one could result in backlash from constituents.
On the corporate side, one provision – deferred taxation of overseas corporate income – accounts for $83 billion (or over 50%) of the $150 billion revenue loss. Ending deferral and moving closer to a worldwide tax system could raise a significant amount of revenue but would be viewed negatively by the corporate sector. As a result, most reform proposals include a move toward a territorial tax system with a repatriation requirement that can be used to offset the cost of lower tax rates. A potential problem with this is that a repatriation requirement would only result in a one-time revenue surge, as most overseas earnings going forward would not be taxed under a territorial system. This may allow for revenue neutrality over the traditional 10-year budget window, but beyond that such a reform would likely lose revenue.
The revenue effects of any tax reform proposal must be estimated. The skilled professionals at the Joint Committee on Taxation (JCT) are charged with this task2. The JCT is the official revenue estimator for Congress. It is their estimates (known as “scores”) that matter for budget rules and determining whether a tax change increases, decreases, or leaves unchanged future deficits. Other groups such as the U.S. Treasury and outside think tanks also compute revenue estimates which may be used by policymakers to inform debates, but it is only the JCT’s estimates that matter for budget rules.
When the JCT scores a bill, it uses what is known as a “static” economic model. Static scores only account for micro-level responses in individual behavior resulting from a tax change. For example, if a bill includes a tax incentive for saving, then the JCT’s model will incorporate the change this will induce over individuals’ consumption/saving decision, but it will not account for the potential increase in investment resulting from the greater supply of savings, which, in turn, can lead to an expansion of the economy.
Recently, the JCT has also been using a “dynamic” economic model to estimate revenue effects. Dynamic scores do account for aggregate economic effects from tax policy changes. Thus, dynamic scores allow for revenue feedback effects. In the example above, a tax incentive for saving can lead to greater investment, which then boosts the economy. The greater economic activity then allows for more tax revenue generally to be collected, lowering the initial cost of the saving incentive. Proponents of dynamic scoring often argue that the failure of static models to account for aggregate economic effects leads static models to inflate the cost of tax cuts.
Increasingly, there have been calls for the JCT to make dynamic scoring the standard by which tax bills are evaluated. However, there is no widespread agreement on exactly which dynamic model should be used. Embedded in any dynamic model (and static model as well) are assumptions about how responsive individuals and firms are to tax policy changes. Changing these assumptions changes the revenue estimate produced. For example, some believe that individuals are quite responsive to lower taxes, so that a tax cut will boost labor supply and saving significantly, which will carryover into higher economic activity. Others believe that the responsiveness is rather muted. The more responsive individuals and firms are to tax reductions, the greater the feedback effect and the lower amount of revenue that is lost. While it is impossible to say for sure, given past dynamic scores produced by the JCT, it appears that they use more middle-of-the-road assumptions about tax-induced responsiveness, which makes it more difficult to offset a rate reduction with growth effects. Additionally, dynamic scores do not always result in lower cost estimates as seems to be commonly believed.
Tax reform designers could alter the timing of policy changes. For example, consider an across-the-board reduction in tax rates that will result in an annual revenue loss of $50 billion, or $500 billion total over the 10-year budget window. To keep the proposal revenue neutral, policymakers will have to curtail tax expenditures. But say for various reasons, only $400 billion in tax expenditures can be trimmed. How can the $100 billion revenue gap be closed? One way to do this is to schedule the rate reduction to take effect in year three of the budget window. Alternatively, the rate reduction could be phased in over the first several years. The latter would probably be the more likely approach taken, but either would push some of the revenue loss from the rate reduction outside of the 10-year budget window, and thus appear revenue neutral for official scoring purposes.
Likewise, policymakers could decide that abruptly removing particular tax expenditures would be too much of a shock to the economy. To ease the transition, tax expenditures deemed important could be temporarily extended for a year or two. Thus, the cost of leaving these expenditures in the tax code would only be a 10% or 20% of the cost of leaving them in the tax code under a 10-year budget window. As these expenditures expiration date approaches, however, there may be a good chance that letting them expiring would have political or economic ramifications that could not be tolerated, so they may be extended again, and again, effectively becoming permanent. But repeated temporary extensions would hide the true 10-year cost of the provisions.
How a tax proposal affects revenues depends on the period over which revenue effects are measured. This period is referred to as the “budget window.” The most commonly referenced budget window in Washington is the 10-year window. So when a debate is being had about whether a tax reform proposal is revenue neutral, it is typically over whether the proposal is revenue neutral over the next 10 years. However, the budget window could be shortened or lengthen to produced a desired revenue effect. For example, if Congress wanted to completely eliminate the mortgage interest deduction, they would have to phase it out within 10 years to maximize the revenue gained. This may be deemed too much of a shock to the housing market and economy. Extending the budget window to say 20 years, would allow for a more gradual adjustment while still reaping the revenue from eliminating the deduction.
Finally, the JCT could be required to use a different “baseline” when developing their estimates. Baselines provide a benchmark against which the JCT can determine the cost of a tax change. As of now, the JCT is required to use a “current law” baseline when estimating revenue effects. Using the current law baseline results in JCT assuming that the laws that are currently in place will remain in effect in the future. Thus, a temporary tax provision that is set to expire in two years will be assumed by the JCT to actually disappear in two years. Alternatively, the JCT could be required to use a “current policy” baseline. Under a current policy baseline, the JCT would assume that current policy, not current law, would remain in effect. Therefore a currently available tax provision that is set to expire in two years will be assumed to remain in effect even after the expiration date listed in the tax code. Which baseline is used can affect the overall score of a bill.
Consider the following example. Say there is a temporary tax provision that will expire next year and costs the government $10 billion annually. Policymakers would like to make this provision permanent as part of tax reform. Under a current law baseline, the cost of permanently extending the provision would be $100 billion over a ten-year window because the current law baseline assumes it would not be on the books in the future. However, under a current policy baseline, there would be no revenue cost from a permanent extension because the baseline already assumes that the provision will be extended permanently. Other disparities can arise depending on the baseline used.
There are a number of ways Congress could attempt to attain revenue neutrality. The task will not be easy. This post discussed several methods that could be used, and they may need to be combined to one degree or another to achieve the desired revenue effects. One recently proposed revenue raiser is switching to a border adjusted tax. Unofficial preliminary estimates suggest it could be a significant source of revenue. Alternatively, there is nothing written in stone that says tax reform must keep revenues constant. Some have argued that the long-term fiscal situation of the U.S. may require raising taxes and possibly cutting spending.
The information provided in this article is for general information purposes only. The information is not intended to be comprehensive or to include advice on which you may rely. You should always consult a suitably qualified professional on any specific matter.
Mark Keightley is an economist with the Congressional Research Service (CRS) in Washington DC. His research areas include corporate and business taxation, housing taxation, financial securities taxation, and macroeconomics. He serves as an adjunct professor in the School of Policy, Government, and International Affairs at George Mason University where he teaches graduate-level courses in macroeconomics, microeconomics, and investment. Mark was previously an associate at the Congressional Budget Office, a visiting scholar at the Federal Reserve Bank of Saint Louis, and an adjunct professor in the economics department of The George Washington University. He earned a BS in economics from the College of Charleston, and an MS and PhD in economics from Florida State University.
* The views expressed herein are those of the author and are not presented as those of the Congressional Research Service or the Library of Congress.
Maastricht University - IV Global Tax Policy Conference: Tax Policy after MLI and CCCTB, what can be expected? On 19 June 2017 at Hilton Schiphol in Amsterdam Prof. Dr Hans van den Hurk, chairman of the Annual Global Tax Policy Conference of the Maastricht Centre for Taxation (Maastricht University), with his esteem speakers are addressing the above question.Read more