The Trump Tax Cuts Led to Record-Low, Not High, Revenues Outside of a Recession

The tax legislation that President Donald Trump signed in December 2017 significantly reduced federal revenues, with the largest tax cuts going to the richest Americans. Following the enactment of these tax cuts, federal revenues fell dramatically—as the Joint Committee on Taxation (JCT) and Congressional Budget Office (CBO) projected would occur at the time the law passed¹—and they remain below projections of federal revenues made prior to their enactment.

This issue brief provides new analysis of economic trends and federal revenues, finding:

• Federal revenues remain below the levels that were projected before the enactment of the tax law—regardless of whether measured as a percentage of the economy, adjusted just for inflation, or adjusted for both inflation and growth in the adult population.
• Revenues as a percentage of the economy are particularly low given the strength of the economy. Between 1986 and the enactment of the Bush tax cuts in 2001, every year in which the annual unemployment rate was below 5 percent saw revenues exceed 19 percent of gross domestic product (GDP). Since then, in only one of nine years with the unemployment rate that low did revenues exceed 19 percent of GDP—and it was an outlier year due to the COVID-19 recovery.
• Other politicians have cited increases in nominal revenues relative to pre-tax cut projections as evidence that the Trump tax cuts increased revenues. But much of this reflects COVID-19 policies, the expiration of the tax law’s business provisions that proponents want to extend, and higher-than-expected inflation. In 2023, nominal revenues were 6 percent below pre-pandemic projections after adjusting for inflation.

As Congress debates the future of the Trump tax cuts, it should work to ensure any legislation is at least revenues-positive, leaving the country better able to invest in people and their neighborhoods.

The best way to determine the effect of a law or policy is to compare what occurred to what would have happened in its absence. The CBO and the JCT are tasked with estimating the budgetary impact of legislation relative to if the legislation had not been enacted. In 2017, the JCT estimated that the Trump tax legislation would cut federal revenues by $1.6 trillion through 2025, relative to what would have been collected had the law not been enacted.² Recent estimates of the costs of extending the Trump tax cuts have also been significantly revised upward. Compared with estimates the agencies made in 2018, the CBO and the JCT now project that extending the Trump tax cuts is roughly a third more expensive.³

A somewhat less direct but still very helpful approach is to compare the revenue projections made prior to the enactment of the Trump tax cuts under the old tax system to what actually happened to federal revenues, as well as to the most recent projections. This approach has the drawback of not directly looking at the cost of the legislation, so it may also reflect the dramatic changes to the economy that have happened since 2017. Nevertheless, it shows revenues to be significantly below the CBO’s pre-Trump tax cut projected revenues. (see Figures 1, 2, and 3) This is true under any rigorous analysis.

In June 2017—six months before the tax cuts were enacted—the CBO projected that revenues would be 18.0 percent of GDP in 2023.⁴ Instead, they came in at only 16.5 percent of GDP, significantly below the pre-Trump tax cuts projections.⁵ Using these same projections, the CBO estimated that fiscal years 2024 and 2025—the last two years before large portions of the Trump tax cuts are set to expire—would not fare much better. In June 2017, the CBO projected that revenues in 2024 and 2025 would be 18.0 and 18.1 percent of GDP respectively,⁶ and its current projections have 2024 and 2025 revenues at 17.2 and 17.0 percent of GDP, respectively—meaningfully below the pre-Trump tax cuts projections.⁷ Similarly, revenues are currently projected to rise back up to 18.0 percent of GDP after the temporary individual portions of the Trump tax cuts expire.⁸

Revenues are also significantly below pre-Trump tax cuts projections when adjusted for both inflation and growth in the adult population, as well as when they are adjusted only for inflation.

When the economy is stronger, revenues as a percentage of GDP are higher. Taxes are levied on income generated by households. As more people have jobs, federal revenues rise as well.

Between 1995 and 2000, the unemployment rate fell from 5.6 percent to 4.0 percent⁹ and revenues rose from 17.9 percent to 20.0 percent of GDP¹⁰—the equivalent of taking in an additional $600 billion per year in today’s economy.¹¹ But today, revenues as a percentage of GDP stand much lower—lower even than they were in 1995. This is due to a series of expensive tax cuts that dramatically lowered federal revenues. Since the tax code was fundamentally rewritten in 1986, there were three years where the annual unemployment rate was below 5 percent, and in all three of those years, revenues exceeded 19 percent of GDP. And then, the 2001 Bush tax cuts were enacted.¹² Ever since, low revenue has been the norm regardless of the unemployment rate. In eight of the nine years since 2001 in which the unemployment rate was below 5 percent, revenues have stayed below 19 percent of GDP. And the only exception—2022—had revenues that were artificially high due to the COVID-19 pandemic recession recovery.

In other words, it didn’t used to be possible to have revenues this low with employment this high. And politicians cut taxes dramatically, leaving the country with record-low revenues given the state of the economy.

Revenues today, however, are even lower than that analysis would indicate. In addition to revenue growth due to a strengthening economy, revenues as a percentage of GDP should also generally grow as the nation gets richer. Tax brackets grow with inflation, and the United States has a progressive tax structure where higher income is taxed at a higher rate; as wages over time grow faster than inflation, more income is taxed at a higher rate.

The only reason revenues as a percentage of GDP have hovered around their 50-year average, rather than continue to rise, is that Congress keeps enacting expensive tax cuts. Not only are revenues now much lower than they were in 2000; they’re dramatically lower than expected, given how much larger real GDP per capita is now than in 2000. The unemployment rate is roughly the same as it was in 2000, and average wages and salaries, after adjusting for inflation, are 22 percent higher.¹³

But when measured as a percentage of GDP, revenues are 2.8 percentage points lower than they were in 2000. After adjusting for the size of the economy, the United States is collecting $809 billion less per year than it was in 2000, despite being richer and having similar employment.

Looking only at nominal revenues yields an incomplete picture. Due simply to inflation, nominal revenues rise over time, even if the inflation-adjusted level of revenues falls. Nominal revenues also grow naturally as the population—and therefore the number of people paying taxes—increases, as well as due to real increases in economic activity per person. While it is true that nominal revenues are higher than they were prior to the passage of the Trump tax cuts, that is not a helpful analysis for assessing revenue trends. Nominal revenues are more than 100 times bigger than they were in 1944, but as a percentage of the economy, revenues are significantly lower.

Some politicians have instead used a hybrid version that compares nominal revenues relative to pre-Trump tax cut CBO projections, with a focus on 2021 through 2023.¹⁴ While this approach avoids the pitfall, described above, of assuming that prices, population, and GDP growth would not change after 2017, it poses another set of problems.

First, revenue numbers during this period reflect temporary factors related to the pandemic. For example, pandemic legislation gave employers the option to defer payroll taxes on 2020 wages.¹⁵ Half of the deferred payroll taxes were due December 31, 2021, falling into federal fiscal year 2022 and the other half were due on December 31, 2022, falling into federal fiscal year 2023.¹⁶ This had the effect of shifting tax revenues from fiscal years 2020 and 2021 to 2022 and 2023, inflating revenue collections in those years. Pandemic legislation also created an incentive for businesses to shift income from 2020 to 2021 since it allowed them to immediately write off losses on their 2020 income but not their 2021 income.¹⁷ These are just some of the many factors—such as capital gains realizations surging from about 4 percent to 6 percent of GDP—that make the years immediately after the pandemic difficult to use for assessing the cost of the Trump tax cuts.¹⁸

Second, the cost of the Trump tax cuts was always designed to fall significantly in the years leading up to 2025, which makes a focus on these years’ revenues misleading. The original JCT score projected that the cuts’ 2023 revenue loss would amount to $164 billion, compared with a 2019–2021 annual average cost of $254 billion.¹⁹ This is because over the period of 2022 through 2025, many Trump tax cut business provisions, such as bonus depreciation, were scheduled to begin expiring, while many tax increases on businesses kicked in, reducing the tax law’s cost over this period. This is, therefore, a misleading time period over which to assess the cuts’ permanent cost, especially since proponents of the Trump tax cuts support reversing those tax increases and tax provision expirations that reduced the law’s reduction to revenue.

Third, comparing nominal revenues with projections several years ago picks up macroeconomic factors unrelated to the Trump tax cuts. Since the enactment of the Trump tax cuts, the United States has experienced more inflation than was expected in June 2017, in large part due to supply chain disruptions resulting from the pandemic.²⁰ Inflation pushes up nominal revenues—higher nominal wages and incomes bring more money into the system—but about 90 percent of those positive budget effects disappear due to higher nominal spending through items such as increases to Social Security’s cost-of-living adjustment.²¹ In other words, increases in revenues driven purely by inflation are meaningless.

In 2023, nominal revenues were less than 2 percent above the CBO’s pre-pandemic projections, but after adjusting for inflation, they were 6 percent below these projections. For this reason, focusing on revenues as a percentage of GDP—which allows the analysis to abstract away from often unrelated macroeconomic phenomena—is far more revealing. And as discussed earlier, revenues as a percentage of GDP lag far behind projections made prior to the enactment of the Trump tax cuts.

The Trump tax cuts meaningfully reduced federal revenues. As a percentage of GDP, revenues are historically low given the high level of employment in the United States. And federal tax collections lag significantly behind projections made prior to the enactment of the Trump tax cuts.

With a significant portion of the Trump tax cuts set to expire at the end of next year, 2025 presents an opportunity for Congress to improve the tax code, both to make it fairer and to raise revenues to invest in American families and communities. As Congress considers the future of the Trump tax cuts, it should ensure the wealthy and corporations pay their share, especially by paring back some of the previously enacted corporate cuts.

The authors would like to thank Jean Ross, Madeline Shepherd, and Emily Gee for helpful suggestions and David Correa for research assistance.

When adjusting for inflation, the authors used the GDP price index. Figures are presented in chained 2017 dollars. For the adult population, the authors used the CBO projections of the noninstitutionalized civilian population of those ages 16 years and older.

For the June 2017 CBO baseline, the GDP forecast was rebased to match the latest actual value for fiscal year 2016. And for the April 2018 CBO baseline, the GDP forecast was rebased to match the latest actual value for fiscal year 2017.