Hatch Outlines Importance of Dynamic Scoring for Major Policy Reforms in Speech at American Action Forum
Washington, DC – In a speech today at the American Action Forum, Finance Committee Ranking Member Orrin Hatch (R-Utah) called for dynamic scoring to be used to evaluate any major reform, including an overhaul of the tax code, so that the proposal’s macroeconomic effects can be fully assessed.
“Tax reform has been and will continue to be a long and difficult process. I believe the expanded and sensible use of dynamic analysis can, if done correctly, be an important tool to help us achieve our goals.” said Hatch.
Below is the text of Hatch’s full speech delivered today:
Let me begin with a story I heard about an upstanding Utahn named Jim who recently visited a park on his way to see his friend Lisa. Jim happens to be keen on law-and-order and was very unnerved when he observed two people in the park apparently engaged in an illegal drug deal.
After leaving the park, Jim drove to pick up Lisa, who has a strong libertarian streak. When he arrived, Jim could barely contain himself and immediately shrieked: “Lisa, I just saw a drug deal take place in the public park, can you believe it?”
With strong indignation, Lisa cried out: “Oh my goodness, you were in a Public Park?”
The story is an example of how orientation influences what we view as being important. When it comes to the effects federal spending or taxes have on the economy, views also often tend to shift depending on orientation.
Some seem oriented toward the demand side of the economy, and they focus mostly on effects of federal spending.
Others seem oriented toward the supply side of the economy, and they focus on the effects of effective marginal tax rates.
In my view, both are right: demand matters and supply matters.
What I’d like to talk about today is how we have and will, in the future, bring supply, demand, and macroeconomic analysis to bear in analyzing proposals involving significant changes in taxes, spending, and other policy matters.
To preview where I come out on these matters, I’ll note at the outset that while analysis of macroeconomic effects of proposed legislation, or what we sometimes call dynamic analysis, is challenging, it has a number of benefits.
Both Democrats and Republicans alike have acknowledged that dynamic analysis can be useful in a variety of areas. It provides valuable information that should not simply be ignored or discarded. And, it should be used, as it has already been used, to reach budget and revenue estimates associated with major legislative proposals, including tax reform proposals.
That said, use of macroeconomic analysis in scoring and revenue estimation – sometimes called dynamic scoring – is not a panacea.
For example, when applied to tax changes, macroeconomic analysis shows positive effects from reduced marginal effective tax rates on growth in productive inputs like labor and capital. Those effects are real and significant and they capture how a policy proposal would impact American workers and businesses.
However, those effects are not a magic elixir.
While I’d like to tell you that tax cuts always more than pay for themselves, or maybe even that tax cuts cure influenza, I’m sad to have to tell you that just isn’t the case.
Nonetheless, reductions in marginal effective tax rates on labor and capital can and do have positive macroeconomic effects that cannot and should not be ignored by Congressional scorekeepers.
There are statistical studies, simulations, and cross-country comparisons that show those effects are plentiful. It doesn’t take much effort to browse through the Tax Foundation’s website, for example, to find evidence of these positive results. And, I’m sure that Doug, in his more productive academic days, produced numerous studies that show such effects.
Economic growth will be key to moving the economy out of the rut it has been in over the past six years.
Since the end of the recent recession in the second quarter of 2009, GDP growth has averaged only 2.3 percent, a full percentage point below the long-run average we’ve seen since 1947. Projected over long periods, the difference in growth rates means significant differences in standards of living for future generations.
Put simply, more growth means a better future.
We also face significant underemployment in the economy.
While the top-line unemployment number has gone down, other indicators confirm significant weaknesses in the labor market. For example, since 2009 we have an unbroken downward trend in labor force participation, which has fallen from 65.7 percent at the beginning of 2009 to rates not seen since the 1970s, like the 62.8 percent we saw in October.
Increased participation, job growth, and enhanced opportunities in labor markets come hand in hand with stronger economic growth. Economic growth comes from growth in employment and investments leading to growth in physical, human and intangible capital, and from technological change.
Long-run growth does not come from deficit-financed government spending or redistribution.
The true drivers of economic growth, together with returns from work effort, capital formation, and innovation, get tied together in basic economic models. Growth and other models are used by the Joint Committee on Taxation (JCT) in analyzing macroeconomic effects primarily of tax policy changes and by the Congressional Budget Office (CBO) in analyzing effects primarily of spending or other changes.
And the returns to work effort, capital formation, and innovation that matter, both to Americans in the actual economy and in the macroeconomic models, are after-tax returns, where effective marginal tax rates determine decisions at the margin.
When we refer to a piece of legislation’s budget score, as most here know, we are talking about projected changes in budget authority or outlays that will result from the legislation. And, revenue estimation refers to projected changes the legislation will have on federal receipts.
Of course, those changes are measured according to a “baseline,” which represents what outlays and revenues would be if we assume current law will remain in place or if we alternatively assume that some notion of current policy would hold. What policymakers choose for a budget baseline can matter a lot, as we saw in the so-called fiscal cliff episode at the end of 2012. In consideration of any proposed legislative change that will affect the budget, such as tax reform, policymakers need to arrive at agreement on the relevant baseline. The work of budget scoring and revenue estimation generally involves projecting how a legislative proposal will impact the federal budget relative to whatever is the chosen baseline.
There are two basic types of scoring that vary as to how they measure or predict the effects a legislative proposal will have on the general economy. The first, usually referred to as static scoring or conventional scoring in the case of JCT, assumes that a bill will not have any effect on important macroeconomic variables like employment, GDP, or national income. Static scoring does often incorporate some behavioral responses to policy changes, but not general economic effects. This is the most typical type of scoring employed by both CBO and JCT.
For many cases, like those that do not have large-scale effects on spending, tax rates, labor markets, or technology, a score with static assumptions is probably safe.
However, for large proposed changes to government spending, provisions in the tax code, or policies with significant labor force or technology effects, static scoring is downright dumb.
The second type of scoring, which I mentioned previously, is what some refer to as dynamic scoring or dynamic analysis, which simply refers to budget scores and revenue estimates that include analysis of a legislative proposal’s macroeconomic effects.
Both JCT and CBO can offer macroeconomic analysis of a proposal as either supplemental information accompanying a static analysis or as the principle score of interest to policymakers. Similar to a decision about what is to be the relevant baseline, it is up to policymakers to decide how to use results derived from macroeconomic analysis of a proposal.
Now, if you listen to some people discussing static versus dynamic analysis, you’d think that dynamic analysis is to be feared and is something that has never been used, ought to be avoided, contains mysterious features, is too hard to accomplish, or involves unmanageable uncertainties.
Those views are typically overblown and, in most instances, are also downright dumb.
But don’t take my word for it.
Take the word of CBO, JCT, the Social Security Actuaries, the IMF, the Federal Reserve, members of Congress on both sides of the aisle, or the administration. All of these groups have either produced or supported and utilized dynamic analyses of large-scale policy proposals to guide decisions, acknowledging that static scores would be – to paraphrase their views – downright dumb.
Some of the debate I hear concerning use of dynamic analysis by CBO and JCT seems remarkably uninformed. Hearing the debate, you’d sometimes think that dynamic analysis is some untested, never-before-used tool. But, anyone paying attention knows that is not the case. CBO and JCT have and will perform macroeconomic analysis of dynamic effects on the economy and the resulting budgetary implications of major legislative proposals.
And, not surprisingly, the Republic has survived.
Take, for example, the most recent immigration proposal put forward in the Senate.
That proposal involved policies that have the potential to generate significant effects on the labor market, including employment, earnings, and the skill mix of the labor force. Those opposing use of macroeconomic analysis in budget scoring would, if they were consistent in their arguments, say that CBO and JCT should have used static scoring, including an assumption that employment and GDP would remain fixed at values projected under current law. But, in explaining their dynamic analysis of the immigration proposal, CBO essentially wrote that with a proposal involving such a large possible labor market impact, the use of static scoring would be, in my words, not theirs, downright dumb.
Instead, CBO and JCT produced an analysis in which many macroeconomic effects were considered, which ran counter to a strict static analysis.
And, wouldn’t you know it, the Republic survived.
In fact, the positive economic results that CBO predicted were warmly embraced by a number of Democrats. Indeed, you can see the results of CBO’s dynamic analysis of the immigration bill touted on the White House website.
Oddly enough, these same Democrats, in other contexts, have written off dynamic scoring as some sort of fantasy used by Republicans to justify lower tax rates. But, when the same type of analysis could be used to bolster the case for the immigration bill, they were singing off a different song sheet altogether.
Immigration reform is not the only case in which Congressional scorekeepers have provided dynamic analyses of proposed changes to federal policy. While I don’t have time today to review them all, it does not take much effort to browse through the CBO and JCT websites to find many examples. JCT even has a special tab on its main web-page titled Macroeconomic Analysis.
The question is not whether CBO and JCT should use dynamic macroeconomic analysis. As I have said, they have done so, can do so, and will do so. The question is what role that analysis should play. And, as we continue to work toward tax reform, that question becomes all the more relevant.
Once again, I want to stress that dynamic scoring is not a magic elixir that solves all of our problems when it comes to tax policy.
Even if we agree to use dynamic scoring on major tax reform proposals, there are a number of questions we must consider, including: what economic models to use; what so-called parameter values to choose for things like sensitivities of labor and capital suppliers to after-tax returns on their efforts and investments; what assumptions to make about possible behavior of the Federal Reserve or foreign policymakers; and what assumptions to make about how any proposal fits with the government’s long-run budget constraint.
But, while these issues are certainly challenging, they have not prevented JCT or CBO from arriving at informative projections regarding large-scale spending or tax policy changes in the past, and they shouldn’t hinder such efforts in the future.
Macroeconomic analysis providing projections of future effects of policy changes are, of course, subject to uncertainties. And no matter what models, parameter values, and assumptions we use, that will remain. However, the argument that we should not use information from dynamic analysis of policy proposals because the analysis is uncertain and difficult is almost comically misguided to me.
Oddly enough, many who argue against use of macroeconomic analysis of tax proposals because of some uncertainties are the same people who argue for the embrace of point estimates from global climate-change models that are flat-out rife with uncertainties.
Go figure.
So where do I stand on use of dynamic scoring for tax policy changes?
To me, it is clear that we should continue using dynamic analysis and work with CBO and JCT to ensure that those efforts proceed and are accelerated and refined.
As with the economics profession, the work at CBO and JCT should adapt to the advancement and development of the tools of analysis used by economists. I recall seeing a picture of Milton Friedman and Anna Schwartz as they were analyzing data for use in their historic work on the Great Depression and Monetary Policy. The picture shows the two scholars at a large table looking over a grid of data points, and they were trying to fit a trend line to the data using a long piece of string. Well, since that time, there have been major developments in the tools that economists use, including the development of dynamic programming and computing. Arguing against use of dynamic analysis by CBO and JCT is like harkening back to the days of fitting data with a string.
I’d like to end with a brief note on what is at stake with respect to our efforts on tax reform, especially given that I believe that there is a lot of misinformation being peddled out there about dynamic scoring.
Let’s look back at JCT’s projections from its dynamic analysis of Chairman Camp’s proposal. Of course, Chairman Camp deserves a lot of credit for putting his plan out there and showing how hard it is to engage in comprehensive tax reform, especially when constrained by static revenue neutrality and distributional neutrality.
According to JCT’s results, Chairman Camp’s plan could produce, from positive macroeconomic effects, upwards of $700 billion of revenue relative to a static analysis.
If you accepted the $700 billion positive revenue effect – which was, once again, at the high end – and hypothetically plowed it back in for further rate reductions, you’d probably be able to lower tax rates in the various brackets by less than one percentage point.
While that would be real change with real impact on Americans, it would hardly be the supercharged supply-side miracle that many, including some on my side of the political spectrum, have argued we would see.
Don’t get me wrong, incentives, the supply side, growth, and capital formation matter. And they matter a lot for future living standards, and cannot be ignored. Indeed, as tax writers, we would be committing malpractice if we undertook an exercise as far-reaching as tax reform and ignore macroeconomic data. However, especially with the type of exercise like Chairman Camp undertook, we should not expect dynamic scoring to produce outsized miracles from either the supply side or the demand side.
As always, there have been and continue to be political and economic debates about economic analysis, but that is healthy and useful. What is not healthy is to ignore useful information or to try to bias or distort analytical work done by CBO or JCT. And, if anyone thinks of purposefully writing legislation to distort the way bills will be scored or to dupe the American people, well…don’t get me started.
Dynamic analysis of major spending, tax, regulatory, and other proposals should be valued. It is intellectually dishonest to praise and accept such analysis for spending or labor-market related proposals like immigration, while arguing against the use of that same type of analysis when it comes to taxes.
Tax reform involves changing many margins along which Americans make important resource allocation decisions. Dynamic analysis can help us analyze how changes in the margins, distortions, and deadweight losses will affect people, for better or worse.
That being the case, such analysis should be used with regard to tax policy changes.
We need to stay focused on what is important. The primary goal of tax reform and analysis of any such reform must be growth in opportunities, incomes, and jobs for Americans, as well as fairness and simplicity in the tax code. If we limit ourselves to static analysis, we take attention away from things that really matter. What matters to American households and businesses is not simply how much a proposal changes a revenue estimate or budget score; what matters is what tax policy will do to or for them.
Tax reform has been and will continue to be a long and difficult process. I believe the expanded and sensible use of dynamic analysis can, if done correctly, be an important tool to help us achieve our goals.
I look forward to working with all of you as this effort continues.
Thank you, once again, for having me here today.
