Tax Policy Center and Wharton School partner to score major tax plans

The Tax Policy Center and the Wharton School at the University of Pennsylvania are collaborating to analyze the economic effects of major tax proposals. This partnership will allow TPC to better understand how tax changes affect the overall economy and the federal budget. TPC and Wharton will look at tax policy in three different ways.

First, TPC will continue to produce a traditional revenue score and distributional analysis. The analysis assumes people adjust behavior in response to tax changes. For instance, investors would sell more stock in 2016 if they know that tax rates on capital gains rates will rise in 2017. But it does not account for changes in the broader economy.

Second, TPC created its own macroeconomic model that makes it possible to produce a dynamic score of tax plans that accounts for their effects on the overall economy. Finally, economists at Wharton are using their own macroeconomic model to produce a second dynamic score.

Dynamic scoring first estimates how a change in the tax law affects key elements of the overall economy, such as the Gross Domestic Product (GDP), labor supply, and investment.  Then, it includes those broad economic changes when calculating how those tax changes affect the budget deficit.

TPC and Wharton use different macroeconomic models. TPC’s new internal version is a Keynesian analysis that is based on two principles: In the short-run, changes in economic output and employment are driven by changes in overall demand; but in the long-run, output and employment are determined by the level of potential growth—controlled by fundamentals such as labor supply and productivity. The TPC model assesses the effects of those changes in demand.

The Wharton model, called an Overlapping Generations (OLG) Model looks at the world differently. It attempts to capture longer-term effects on the economy’s potential output by modeling how individual households would respond to tax changes over a long period of time. For instance, it models how a  change in tax law drives decisions to work, save, or invest, and then aggregates those responses to calculate how the entire economy would respond.

These two different perspectives can produce different results. Take for example a big tax rate cut. In the Keynesian model, a rate cut increases after-tax income and demand (though that rise in demand could be tempered, at least in part, by monetary policy). Over time, however, those effects wash out and economic output is determined by the economy’s potential.

In the OLG model, that rate cut would increase the labor supply and capital holdings over the long run, which in turn would boost economic growth. One key issue, however, is how those households would respond over the long-term to increasing budget deficits that may drive up interest rates.

To show how this works, TPC and Wharton analyzed three major tax proposals—those of Donald Trump, Hillary Clinton, and the plan issued by Speaker Paul Ryan and House Republicans in June.

Initially, TPC and Wharton were going to release their analyses of all three plans today. However, reality intervened. Because Trump announced major changes to his plan yesterday, TPC decided to delay releasing both the Clinton and Trump plans until it can work through Trump’s changes. That may take a few weeks.

However, we are able to look at a proposal consistent with the plan released by House Republicans in June. It provides a useful example of what a big package of tax changes would look like, using traditional and dynamic scoring.

The plan would cut individual and corporate tax rates, raise the standard deduction, and repeal the estate tax and the taxes associated with the 2010 Affordable Care Act. It would also repeal most itemized deductions and personal exemptions.

First, let’s look at this proposal in the traditional way, starting with a distributional analysis.

In 2017, it would cut taxes by an average of about $1,800, but high income households would enjoy the bulk of the benefits. The lowest income households would see an average tax cut of $50, a 0.4 percent increase in after-tax income. Middle-income households would get a tax cut averaging about $260, about 0.5 percent of after-tax income. The highest income 1 percent of households (who make $700,000 or more) would get an average tax cut of nearly $213,000, 13.4 percent of their after-tax income.

How would the House GOP plan effect federal revenues? Using traditional scoring, it would reduce them by $3.1 trillion over 10 years. With additional interest costs, it would add $3.7 trillion to the national debt over a decade.

What would it look like with dynamic scoring?

TPC’s Keynesian analysis finds the House plan would reduce federal revenues by slightly less—about $3 trillion over a decade.  Its big tax cuts would boost demand and increase economic output substantially in 2017, but after a few years, the economy would return to its long-term potential level of growth and tax revenues would slow.  

The OLG model shows a 10-year loss of tax revenue of about $2.7 trillion—slightly less than the Keynesian model, and about 14 percent less than the traditional estimate without macroeconomic effects. Interestingly, in the following 10 years, the OLG model finds the plan would reduce federal revenue by $2.4 trillion, slightly more than either the Keynesian analysis or the traditional estimate.

What happens? Initially, the tax cuts increase incentives to work, save and invest; and boost economic output and taxable incomes. But by the second decade, raising deficits crowd out more private investment, and reduce both economic output and tax revenues.

This is a first attempt by TPC and Wharton. While it is important to understand how tax cuts affect the overall economy, dynamic scoring is still relatively new and we are still learning about how it works.

TPC and Wharton will complete their analyses of the Clinton and Trump tax plans soon. Until then, consider this a case study in how different models produce different projections—though perhaps those dynamic scores are less different from traditional estimates than some backers of big tax cuts hope. 

This story originally appeared on TaxVox.