How Would Fundamental Tax Reform Affect the Time Burden of Filing Taxes?

In a Brookings Institution report titled Radical Simplification of the Income Tax (March 2024), William Gale argues that the existing individual income tax is needlessly complex and, as a result, imposes undue compliance costs on taxpayers. These costs take the form of time spent filing returns, money paid to professional tax preparers,¹ economic efficiency losses from tax avoidance, and, in Gale’s view, political economy risks associated with notions of unfairness.

In this context, Gale proposes a menu of four reform options with goals of tax simplification, increased progressivity, and revenue-neutrality in 2023. The proposals share several core design features, including base broadening and the consolidation of tax credits. Preferential tax rates and the alternative minimum tax would also be eliminated. The proposals vary, however, on the precise degree of progressivity – and whether a new value added tax is used as a revenue raiser.

In addition to budgetary and distributional estimates for each proposal, Gale offers qualitative descriptions of how each reform would reduce tax complexity. We add to his analysis by generating quantitative tax burden estimates – specifically, how the proposals would affect the time burden of filing taxes and the number of families who could benefit from pre-filled tax returns. We also offer a more general budget score of the proposals, highlighting their budgetary, distributional, and microeconomic effects.

For 2026, we project that average time to file under current law will be 14.2 hours. This average is higher than that of 2025 and reflects the expiration of several TCJA provisions which reduced time burdens, such as limitations on itemized deductions, the higher standard deduction, and a higher AMT exemption. The reforms presented in this analysis would reduce the time to file by an estimated minimum of four hours over current law in 2026, with the Future option cutting average time burdens by half.

These numbers reflect the burden of filing the individual income tax only. But both the “Future” and “UBI” proposals would add a VAT as part of the revenue-raising mix. A VAT is a kind of consumption tax levied at each stage of the production process. While VATs are understood to be a comparatively administratively efficient way to raise revenue, the introduction of a new tax system does impose an additional time burden through its recordkeeping and filing requirements. However, this new burden is borne by businesses rather than individuals.

How much of the time savings from income tax simplification would be shifted to businesses under the new VAT? We can calculate a back-of-the-envelope estimate based on data from other countries. While no country imposes a VAT as broad and simple as Gale’s proposed design, the time burden of a VAT in peer countries with relatively simple designs ranges from 15 to 30 hours per year; refer to the appendix for additional details on these estimates. About 35 million businesses file taxes in the United States.⁴ The proposed VAT, then, would increase time burdens on businesses by between 525 million and about 1 billion hours per year. These figures can be compared to the estimated total individual time savings in 2026, which ranges from 670 million hours under the “UBI” option and 1.2 billion hours under the “Future” option. Note that most of the 35 million businesses are sole proprietorships, complicating the interpretation of these reforms as shifting time burden from individuals to businesses. Still, an overall per-business average VAT time burden estimate masks the fact that smaller businesses (including sole proprietorships) would have lower-than-average time burdens.

We estimate that the set of provisions common to each reform option would increase revenues by more than half a trillion dollars annually after the expiration of TCJA. Itemized deduction repeal and the elimination of preferential tax rates on dividends and capital gains are the largest contributors.⁵ This estimated revenue gain is somewhat larger against current law than against current policy. All else equal, this difference in baselines pushes Gale’s reform options from revenue-neutrality into deficit reduction.

A portion of this revenue is invested into a small UBI under each reform option. The share-of-GDP cost of these UBI provisions falls over time as productivity growth exceeds population growth. We estimate that the two non-VAT plans, “Simple” and “Modified Simple”, would raise around 1-to-1.2 percent of GDP by the third decade of enactment.

The other two scenarios, “Future” and “UBI”, also include the introduction of a 10 percent VAT assessed on nearly all domestic consumption in the US. We estimate this new tax would raise more than 4 percent of GDP per year, even after accounting for its negative effects on other sources of revenue.⁶ Because the Budget Lab assumes that the Fed would allow for a one-time price level increase in response to the enactment of a VAT, we express dollar-amount revenues both in “real” terms (i.e. baseline dollars) and nominal terms. The former presentation allows for a more intuitive comparison to baseline; the latter is helpful for government budgeting, which is generally done in nominal terms. We do not account for any effects of higher prices on government spending, which would somewhat reduce the net budgetary savings from a VAT.

We decompose the overall distributional effects overall into provision-level contributions shown in the figures below. The first of these charts breaks down the effects of base-broadening provisions common to all reform options; the latter four figures show how each proposal’s design difference would stack on top of those common features.

The burden of the proposed revenue-raising provisions would fall disproportionately on low- and high-income households. The loss of itemized deductions and preferential tax rates on dividends and capital gains would generate substantial tax increases on high-income families. (The loss of the QBI deduction has no effect in 2026, as current law already repeals it.) The loss of personal credits, the largest of which are the EITC and the CTC, reduces after-tax incomes most for the bottom two quintiles in relative terms. As shown in the revenue estimates, AMT repeal has a limited effect on liability when stacked after the repeal of provisions it is designed to claw back.

The “Simple” and ”Modified Simple” reforms are designed to offset tax increases from base-broadening measures using the UBI and the wage subsidy. Universal per-person transfers, as in the case with the “Simple” option, are steeply progressive: $1 of tax relief for a low-income family is a greater share of income than for a high-income family. The means-tested version of a UBI in “Modified Simple” adds progressivity by concentrating a similarly size aggregate tax cut among those below a certain income threshold. Note that these are average estimated tax changes; some number of low-income families would still be worse off, especially under the “Simple” option wherein the UBI does not offset the loss of child-contingent benefits for certain parents.⁸

Both the “Future” and the “UBI” options trade a more generous UBI, which is more expensive from a budget perspective, for a VAT, which raises additional revenue. We estimate that a VAT, in its second year of enactment, would be slightly progressive. This finding, which may challenge common intuitions about the distributional impact of consumption taxes, stems from two factors:

• First, most major elements of the broader tax and transfer system are indexed to inflation. We assume that the Federal Reserve would allow consumer prices to rise in response to a VAT (as opposed to the VAT’s burden being “passed backwards” in the form of lower incomes). A disproportionate share of the bottom quintile’s income stems from transfer programs (including Social Security), most of which are indexed and thus would automatically rise in proportion to a VAT-driven increase in consumer prices. For example, we project that the $3,900 UBI in the “UBI” reform option would rise to $4,300 in 2026 in response to the VAT. This cushion against rising prices contrasts with the expected response in labor income (which would not rise in response to a VAT) and capital income (which would only rise over time as new investments are made), both of which comprise a greater share of income for higher-income groups.
• Second, VATs are more progressive in the immediate years following enactment than they are in the long run when they become regressive. It is widely understood that VATs function as a kind of tax on wealth, burdening returns to “old” (i.e. pre-enactment) investments while exempting the normal return to new investments from tax. In the first year of enactment, all capital income is “old” capital; in the long run, all capital income is “new” capital. Because capital income is skewed towards high-income families, this dynamic means that a VAT will become more regressive over time. A similar “vintaging” effect happens with Social Security recipients. Beneficiaries who retired prior to VAT enactment would see a cost-of-living adjustment for the VAT-driven price increase, whereas those who retire after enactment would see no such adjustment. Thus, over time, a greater share of Social Security recipients, the benefits of whom disproportionately contribute to total income of lower-income groups, would be burdened by the VAT. As such, the VAT’s effect would shift from progressive to regressive over time – our distribution tables in the 30^th year of enactment, which can be found in the data download file, show the VAT provision would be somewhat regressive.

We emphasize that quantifying the burden of indirect taxes involves making more assumptions than is required for individual income and payroll taxes. Other approaches, such as distributing the burden in proportion to current-period consumption, would generate a more regressive result. Please refer to our VAT methodology page for a full discussion of these modeling considerations.

A final noteworthy finding is that increasing the standard deduction is less progressive when done in the context of a tax system with no itemized deductions, as demonstrated in the provision-level breakdown of “Future”. Overall, the “Future” and “UBI” reform options are the most progressive, highlighting an established result in public finance that a flat rebate paid for by a tax on consumption is a progressive fiscal program.

Current law is characterized by negative EMTRs for low-income parents as child benefits phase in with income. Average EMTRs quickly rise with higher earnings to more than 30 percent for low- to middle-income parents as these benefits phase out and income exceeds the standard deduction. For non-parents, who don’t benefit from the CTC or a more generous EITC, average EMTRs are always positive and rise steadily. In other words: the existing tax system explicitly encourages parents to participate in the labor force.

Gale’s reforms would eliminate the credits that give rise to these dynamics, replacing them with a UBI and an individual-level wage subsidy credit. The UBI has no phase-in, and the wage subsidy phases in at a rate of 20 percent. For parents, the net effect is an increase in EMTRs relative to current law; for non-parents, the net effect is a decrease. That is, for those who are marginally attached to the labor force, parents’ work incentives would be weaker, whereas non-parents' incentives would be stronger.

For workers in the $30,000 to $60,000 range, the effect on work incentives varies by reform option. In this range, current law generates high marginal rates for parents as child benefits phase out. Gale’s reforms, which have no such programs and thus no similar effect, would increase work incentives for this group. The exception is the “Modified Simple” option, which includes a sharp phase out of the UBI and thus creates EMTRs of more than 50 percent. The large standard deduction increases in the “Back to the Future” option means that most workers in this range would pay no income tax and thus face lower EMTRs.

How would these changes in work incentives translate to employment? Our approach to estimating labor supply responses assumes that different kinds of workers are more responsive to tax policy than others – and thus accounts for the fact that Gale’s reforms would have heterogenous effects on low-income, middle-income, parents, and non-parents. We find that, on net, each reform would reduce employment by about 500,000 workers, about 0.3 percent of the labor force. The “Modified Simple” reform, with its strong disincentive effects created by the UBI phase out, would induce a slightly higher number of labor force exits; “Back to the Future”, whose large standard deduction means that married workers face no income taxes until their income exceeds $100,000, would induce a slightly lower number of labor force exits.

Another potential impact of tax policy is its effects on the future economic outcomes of children. By partially determining the amount of financial resources available to families with children, tax policy may might impact later-life outcomes through various channels including improved nutrition and health, better education, the ability to move to higher-opportunity geographic areas, and more. We simulate these effects using a parsimonious approach based on observed correlations between parents’ incomes and children’s future incomes. Our strategy maps changes in today’s income caused by a tax reform to changes in future earnings, with effect sizes varying based on a child’s parent’s rank in the income distribution. A key determinant of how fundamental reform affects future earnings, then, is to what extent it transfers resources from non-parents to parents.

The figure below shows estimated effects on wages in 2054 among workers who were exposed to these reforms as children, organized by parent income rank. We find that the “Simple” reform option would modestly reduce future wages for children who grew up in the bottom quintile. This negative effect stems from the $1,000 UBI and work credit being insufficient to offset the repeal of current-law child benefits, a finding which Gale’s own distributional analysis confirms. The “Modified Simple” reform attempts to remedy this effect and as such would have a negligible impact on future wages. In contrast, the larger UBI values in the “Future” and “UBI” reforms would leave lower-income children better off on average compared with current law. We estimate these children would see a 1.5 percent increase in earnings under the latter reform – an annual earnings increase of about $500 in today’s terms.