Dynamic Scoring: Evidence From the States

Daily Tax Report: State provides authoritative coverage of state and local tax developments across the 50 U.S. states and the District of Columbia, tracking legislative and regulatory updates,...

Tax Policy

Dynamic scoring seeks to measure the economic effects of changes in policy. The Congressional Budget Office uses dynamic scoring to estimate the effects of proposed tax legislation on the federal budget. States also use dynamic scoring to predict the effects of state tax policy changes. In this article, Georgia State University's Peter Bluestone discusses the real-world experience states have had in using dynamic scoring for tax policy making.

Peter Bluestone

By Peter Bluestone

Peter Bluestone is a senior research associate with the Center for State and Local Finance at Georgia State University. His research expertise includes urban economics, static and dynamic economic impact modeling, and state and local fiscal policy. His work has included modeling state and local impacts of policy changes and economic development using various economic models, including IMPLAN and Regional Economic Models Incorporated (REMI). He received his Ph.D. in economics and a J.D. from Georgia State University.

Dynamic scoring is a hot topic again in Washington. Key members of the Trump administration, including Treasury Secretary Steven Mnuchin, have stated that dynamic economic effects will diminish the projected deficits of proposed changes to national tax policy. While considerable scholarship exists modeling the effects at the national level of tax policy changes, it is worthwhile to reflect on the experiences of states with dynamic scoring. Not only do many states have considerable expertise working with dynamic economic modeling and scoring, but a handful of states have recently implemented tax cuts, with the stated belief that the resulting economic growth would offset part of the projected budget deficits. The results from these real-world experiments provide a cautionary tale to those who would rely on dynamic scoring to inform the budgetary process at either the state or federal level.

Tax Cuts, Economic Growth

Recent experience in several states shows that merely cutting taxes does not automatically result in sustained economic growth. Kansas, Louisiana, and Oklahoma all embraced ambitious tax cuts phased in over several years. But instead of unprecedented economic growth in the years that followed, these states found themselves facing budget deficits and difficult choices: either reverse course on the tax cuts or limit government expenditures in important areas such as education and health care. (Louisiana and Oklahoma faced additional economic headwinds as the price of oil plunged shortly after they announced their tax cuts.) Eventually, the budget cutting became too painful, and all three states reversed course and reinstated some of the previous tax policy and rates.

While not all of these states relied on dynamic scoring to justify their changes in tax policy, the example of Kansas, which did have a dynamic analysis done, is illustrative of the difficulties states face when trying to use dynamic scoring in the budgetary process. Some background on what dynamic scoring is and how states use it is necessary to illustrate these challenges.

State policymakers and their staffs in 21 states have experimented with dynamic scoring since the early 1990s. While many states regularly use dynamic models to assess the economic impact of infrastructure investments, almost all state-level efforts to dynamically score tax policies for official budgetary purposes have been discontinued. The reasons include unrealistic forecasts of revenue changes and the difficulty and expense of harnessing a highly imprecise policy tool in a balanced-budget environment.

Economic Ripple Effects

Dynamic scoring is largely concerned with the economic ripple effects from a tax change. Dynamic effects are often compared to the more traditional static revenue estimates, which typically measure only the direct effects of a tax change, though selected behavioral effects of tax changes (e.g., the effect of taxes on hours worked) may be incorporated. A dynamic estimate, on the other hand, attempts to account for economic growth (or decline) associated with reduced (or increased) taxes. Thus, a tax cut that has an estimated static effect of $100 million but that is dynamically scored might only reduce revenues by an estimated $90 million — a 10 percent dynamic effect.

In a joint report by Georgia State University's Center for State and Local Finance and the Fiscal Research Center, experts looked in detail at the range of dynamic effects in seven states that reported the results of dynamic modeling. States reported dynamic effects ranging from 1 to 20 percent of the static revenue estimate.

Despite the interest and some apparently large dynamic effects, almost all dynamic-scoring efforts at the state level have been discontinued. The reasons are twofold: First, even acknowledging economic-growth effects of tax cuts, the size of a dynamic effect — even a large one — is typically minuscule relative to the overall size of a state's budget. And second, given the complexity of these estimates and timing issues (when exactly does the dynamic effect occur?), the actual dynamic estimates are too imprecise and too uncertain to be built into a state's budget in any meaningful way.

Experience in Kansas

Kansas' recent experiment with dynamic scoring illustrates these pitfalls. In 2012, Kansas adopted major reductions in its income tax. For fiscal year 2015, the state economist's static estimate was that revenues from the 2012 tax changes would decline from $6.466 billion to $5.642 billion (an $824 million loss, or 13 percent). However, a dynamic analysis from a pro-tax-cut research institute predicted that the state would actually lose only $714 million. That analysis forecasted that $110 million in additional revenues would be recovered through economic growth, a dynamic effect of 13.5 percent (as measured from the original static estimate of an $824 million loss).

The problem for the state was that even with such a large tax cut and a large estimated dynamic effect, the estimated value of the dynamic effect would still be only 2 percent of general-fund revenues. While dynamic models do not generate a margin-of-error estimate for dynamic effects, research has shown that traditional revenue estimates carry an error rate of around 3 percent.

Large Deficits, Massive Cuts

In the Kansas case, the state's revenues trended below the static revenue estimates, causing large state budget deficits and massive cuts to state spending on education and health care. Even several years after the implementation of the tax cuts, the Kansas legislature in 2015 faced the daunting task of closing a $400 million budget gap. By 2017, the Kansas legislature had seen enough; Kansas raised the top tax on wage income and ended the special treatment of business income, despite a veto of their earlier efforts by the governor.

Whether the dynamic effects from the Kansas tax cut failed to materialize, were incorrectly estimated, or were simply lost in the normal error rate around a traditional revenue estimate, is an open question. It is also a question that is likely to never be conclusively resolved. Despite the hardship, Kansas was better served by sticking with the more conservative static estimate —budgeting based on the dynamic estimate would have left the state with even larger budget gaps to fill.

Comparing Economic Tradeoffs

None of this is to suggest that dynamic scoring can't be useful in comparing economic tradeoffs among different tax-policy choices or even among different tax and expenditure mixes. A recent study done by Georgia State University's Fiscal Research Center modeled the effects on the Georgia economy of fundamental tax reform in which a broad-based tax on consumption (a sales tax) is substituted for the state's current personal income tax. The study uses the dynamic state economic model from Regional Economic Models, Inc. (REMI), to examine the effects of the tax reform on different sectors of the Georgia economy, including labor force participation, savings and investments, consumer spending, and industry sector employment, among others. The report finds that fundamental tax reform has a modest effect on some of the areas studied but is dependent on the assumed changes in the cost of capital in the state.

Distributional Effects

Another state that uses dynamic models to inform the policy debate but not as a budget tool is Nebraska, which uses a dynamic model of the state economy to show the potential impacts of various tax policy changes. In Nebraska, the state is particularly interested in modeling how tax policy changes effect taxpayers across the income spectrum, known as distributional effects. The experience of Georgia and Nebraska as well as other states show that these models can be customized to highlight the distributional effects of tax-policy changes across income classes or industry types.

Dynamic modeling has some interesting applications to policy analysis and provides potentially useful information on the different ways that the effects of tax policies ripple through the economy. For instance, noting that some tax changes may cause job losses or declines in wages even while growing the productivity of the economy is helpful information if a policymaker is largely concerned about job growth. Dynamic models may also be quite useful in comparing tax and expenditure tradeoffs.

Forecast Uncertainty

Where dynamic modeling falls short, and what is apparently often disappointing to policymakers, is that dynamic revenue analysis has not proved to be a particularly useful tool for budgetary decision-making. A state's economy is a vastly complex system. The results obtained from dynamic models rely heavily on assumptions made by the model builders as well as on the availability of data. Even with the advances in computing power and increased data availability, simplifying assumptions are needed, which increases the uncertainty of any forecast.

Even assuming that the dynamic models are highly accurate, relatively large dynamic effects, such as those estimated in Kansas, take time to materialize and are ultimately small when compared to a state's general fund revenues. The practical effect is that dynamic effects are likely to go unnoticed by the average citizen, state policymakers, or state budget staff.

In light of these concerns, states contemplating the use of dynamic models should consider several issues. First, what do policymakers want to learn from dynamic revenue estimation? Based on recent state experiences, policymakers and analysts need to recognize that dynamic revenue modeling can be useful for informing a policy debate, but policymakers should generally not expect large effects and because of the uncertainty of the estimates should avoid using these estimates in making budget decisions. Policymakers in states such as Massachusetts in the 1990s and more recently, Kansas, found that the dynamic effects take a long time to materialize. Second, states need to consider the resources required to develop, customize and then interpret the results from a dynamic model. These models are expensive to build and maintain and are complicated to use. More than a few states have decided that the added value of the information is simply not worth the money, time and effort required to purchase, develop, maintain, and use dynamic models.

Copyright © 2017 Tax Management Inc. All Rights Reserved.

Request Daily Tax Report: State