Proponents of a federal value‑added tax (VAT) argue it would raise substantial revenue at comparatively low economic cost. A VAT is a type of consumption tax levied on the value added at each stage of production and distribution. The idea has recently resurfaced in Washington policy circles. At an October 20 Brookings Institution event on the nation’s fiscal outlook, panelists discussed potential solutions to the worsening federal budget situation, with a VAT being pitched as a possible remedy.
Drawing on cross‑country evidence and micro‑studies, I show that VATs tend to (1) increase the overall tax take and the size of government (“money‑machine” dynamics), (2) depress small‑firm growth and entrepreneurship through threshold notches and compliance costs, (3) impose heavy administrative burdens relative to a retail sales tax, (4) ratchet up prices with asymmetric pass‑through, (5) collide with U.S. federalism and reduce state sales‑tax bases, and (6) fail as a path to durable fiscal consolidation. The American tax and federalist architecture makes these problems more acute than in countries where the VAT successfully replaced cascading turnover taxes. A U.S. VAT would most likely become an additional layer of taxation that expands government while dampening business dynamism.
The renewed interest in a federal VAT springs from a familiar premise: The United States faces persistent fiscal deficits and needs a durable, “efficient” revenue source. Because the VAT taxes value added at each stage of production while allowing firms to credit tax paid on inputs, it is often described as a neutral consumption tax that can raise large sums with little economic harm. That description is accurate in the classroom, but it is only loosely connected to the way VATs work in practice or to how one would interact with American federalism and the structure of U.S. business.
What VATs Actually Do: Revenue, Rates and Efficiency
In theory, the VAT’s credit‑invoice structure ensures that only the final consumer bears the cost of the tax: Each firm remits VAT on its sales and claims a credit for VAT paid on inputs. In textbooks, that design looks like a broad, uniform and neutral consumption tax. Real‑world VATs depart from this ideal in two ways that matter for policy.
First, VATs raise on average about 7% of GDP, with country experiences ranging from roughly 2.5% to 12% of GDP. Among OECD members the average statutory rate is about 17.7%. Yet statutory rates only tell part of the story; VAT performance hinges on “C‑efficiency,” the ratio of actual VAT revenue to the product of the standard rate and aggregate consumption. On this score, the typical OECD VAT collects only about half of what a perfectly broad tax base would yield. Reduced rates, exemptions and compliance gaps erode the base, and governments often respond by raising rates.
Second, VATs do not simply displace other revenue sources. Economists Michael Keen and Ben Lockwood’s cross‑country work finds that the long‑run increase in total revenue associated with VAT adoption is about 2.4% of GDP in one specification, implying that roughly two‑thirds of the VAT’s gross revenue comes at the expense of other taxes, and about 4.5% of GDP in a later paper. The common element is not the exact magnitude but the direction and the durability: VAT adoption raises the overall revenue‑to‑GDP ratio. If the United States adopted a VAT, these findings suggest it would not be a clean swap for existing federal taxes. It would be an additional stream of revenue and, as experience shows, an easily expandable one.
Money‑Machine Dynamic and the Growth of Government
That the VAT tends to increase the aggregate tax take is not a bug; for many advocates, it is the point. As early as 1980, professor of finance at University of Oregon Richard Lindholm presented the VAT as a tool for coordinating tax reform with the “rapidly expanded and expanding level of social expenditures.” The ease with which a VAT can be adjusted by one or two points, the opacity of its incidence, and the administrative machinery that makes compliance routine all contribute to what critics have called a money‑machine dynamic. Historically, statutory VAT rates drift upward: The average rate at adoption was just under 10%, and by 2010 it had reached nearly 16%—a 63% increase. Rate creep is not an accident. Once the infrastructure is in place and the tax is normalized, small hikes look painless to legislators and diffuse to voters.
The connection to government size is visible in the data. Across 15 European countries that adopted a VAT between 1967 and 1995, government spending as a share of GDP grew by roughly twice as much as in the United States over the long run, as shown in the figure below.[1] While U.S. government outlays rose by about 10 percentage points of GDP since the mid‑1960s, the adopters saw an increase just under 20 percentage points.
To probe whether this is merely a trend difference or a discrete shift around adoption, consider a regression‑discontinuity design using a five‑year window on either side of each country’s adoption date, as shown in the figure below. The scatter with local linear fits shows a clear upward jump in government spending at the adoption threshold, with an average increase on the order of 7.2 percentage points of GDP after a VAT is introduced. However one interprets the precise causal channel, the pattern is hard to reconcile with the notion that a VAT tames government. The more plausible reading is that a VAT enables higher spending by providing a stable, easily expanded base.
The money‑machine dynamic also shows up in the composition of taxes. The first figure below shows that since 1966, taxes on income and profits in the VAT adopters increased by about 4.7% of GDP, compared with 3.6% in the United States. The second figure below shows that corporate‑income‑tax receipts rose by 1% of GDP in VAT countries while falling by 1.4% of GDP in the United States. If VAT revenue were consistently used to reduce direct taxes, we would expect the opposite. In the world as it is, VATs appear to add fiscal capacity rather than replace other sources.
The Federalism Problem: Layering, Not Replacing
In the European Union, the VAT replaced a patchwork of turnover taxes that created cascading burdens as goods moved through the supply chain. In the United States, by contrast, the proposal is to layer a federal VAT on top of state and local retail sales taxes. That is not a cosmetic difference.
The basic economics of welfare loss implies that layering is costly. The excess burden of a tax—the deadweight loss—rises faster than proportionally with the marginal rate. Reaching a given revenue target with two stacked taxes rather than one raises the efficiency cost, other things equal. In addition, because all taxes reduce the activities they fall on, a federal VAT would shrink the state sales‑tax base even if state rates were unchanged. The most likely outcome is a revenue squeeze for states, coupled with administrative duplication and friction for firms that must navigate two conceptually different consumption taxes.
The administrative comparison is also unfavorable to a federal VAT. A retail sales tax taxes the final transaction. A VAT requires tax to be collected and credited at each stage. Florida State University economist Randall Holcombe illustrates the practical difference in his work on VAT. Under a 10% sales tax, a retailer charges $2.50 on a $25 sale. Under a 10% VAT, the retailer computes tax on the sale and nets out input credits for VAT already paid upstream, say $0.80, $0.30 and $0.40, leaving $1.00 to remit. That arithmetic is hardly prohibitive for a single firm, but every firm in the chain must keep detailed records, file returns and face audit risk. The number of taxpayers increases markedly, and the government must operate a large crediting system that functions smoothly both in good times and under stress.
European evidence describes the burden from the firm’s point of view. A 2023 European Parliament study (using 2022 data) reports that businesses in many member states spend upwards of €3,000 per year just on VAT compliance. For micro‑enterprises—typically those with fewer than 10 employees—about 21% of all tax‑compliance costs are VAT‑related, a larger share than is spent on corporate‑income‑tax compliance. VAT audits are not incidental: Estimated out‑of‑pocket costs range from a little over €1,000 for micro‑firms to more than €5,000 for medium‑sized firms. These are fixed burdens, and they fall most heavily on small and young firms.
Thresholds, Bunching and the Suppression of Small‑Firm Growth
The strongest micro‑evidence on how VATs distort economic activity comes from registration thresholds. Many countries exempt very small firms, requiring VAT registration only above a turnover threshold. That design creates a sharp notch in tax liability. Firms that would otherwise grow across the threshold have a powerful incentive to bunch just below it. Recent work using U.K. administrative data demonstrates exactly this pattern. As firms approach the registration threshold, turnover growth slows by roughly one to two percentage points, a decline of about a quarter relative to average firm growth. A meaningful share of “non‑crossers” dynamically limit their size to remain under the threshold. The evidence suggests genuine reductions in activity, not merely reporting games.
For the United States, where business dynamism and the rapid scaling of young firms are core strengths, importing a tax that embeds growth cliffs into the business landscape would be costly. When one adds the fixed compliance burden described above, the net effect is a two‑sided squeeze on entrepreneurship: a cliff at the threshold and nontrivial fixed costs from the moment a firm enters. That is a poor fit for a country that depends on experimentation and scale‑ups for productivity growth.
Prices, Real Wages and Asymmetric Incidence
The VAT’s effect on prices is neither neutral nor easily reversed. In an influential study exploiting quasi‑experimental rate changes, economists Youssef Benzarti, Dorian Carloni, Jarkko Harju and Tuomas Kosonen document asymmetric pass‑through: When VAT rates rise, consumer prices increase substantially; when VAT rates fall, prices do not fall by a symmetric amount. In their estimates, approximately 34% of VAT increases pass through to prices, compared with only 6% for decreases, and the asymmetry persists for at least 18 months. The implication is a ratchet: The cost of living climbs more readily than it falls.
The labor‑market channel adds a further layer of complexity to the VAT system. As the OECD has observed, consumption taxes reduce real after‑tax wages and can curb labor supply in much the same way as proportional income taxes; in the short run they can also depress labor demand if firms attempt to preserve real wages while consumption‑tax-driven prices rise. In combination with the price ratchet, this means workers face a squeeze that is difficult to unwind politically. Attempts to shelter households through reduced rates and exemptions erode the base, reduce C‑efficiency and ultimately encourage rate increases, perpetuating the cycle.
The Myth of Replacement
In American debates the VAT is sometimes sold as a path to replacing the federal income tax. That promise has no precedent in the modern world. Of the roughly 170‑plus countries that have adopted a VAT, none has repealed its income tax. The actual pattern is layering: Governments introduce a VAT and keep their direct taxes, often increasing them later. The cross‑country trends in income‑ and corporate‑tax receipts summarized earlier are consistent with this pattern, and so is the upward drift of VAT rates over time.
If a U.S. VAT did finance reductions in income and payroll taxes dollar for dollar, one could stage a debate about the relative efficiency of different tax wedges—the ways in which various taxes drive a wedge between what employers pay and what workers take home, or between pre-tax and post-tax returns on investment. But the empirical record suggests the opposite is far more probable: A VAT would join the existing revenue instruments, not replace them, and would expand total taxation over time.
VAT and Fiscal Consolidation: What the Literature Actually Shows
The strongest political claim for a U.S. VAT is not microeconomic at all; it is macro‑fiscal. The United States has a debt problem. A VAT, we are told, is a responsible way to address it with minimal harm. That claim misreads the consolidation literature and underestimates the way VATs change the long‑run size of government.
On consolidation, the evidence is decisive: Expenditure‑based adjustments succeed where tax‑based adjustments fail. The Alesina–Favero–Giavazzi research program, covering dozens of countries and hundreds of fiscal episodes, shows that consolidations relying on spending cuts are associated with better debt dynamics and smaller output losses, while those relying on tax hikes produce deeper, more persistent recessions and tend to be reversed. The point is not that revenue can never rise; it is that raising the overall tax wedge is macroeconomically costly and politically brittle.
VAT advocates sometimes reply that the Federal Reserve’s dual mandate makes the United States special: A contractionary tax hike would be neutralized by monetary policy. This monetary‑offset argument is misplaced. Central banks rarely offset fiscal shocks one‑for‑one, and—more important—they cannot undo the structural effects of broad‑based consumption taxes: lower real wages, reduced labor supply and diminished incentives to invest. Studies that find low short‑run multipliers for marginal VAT changes from low initial rates do not speak to the introduction of a new national VAT in a large federal economy where the tax would likely expand government and complicate federal‑state finances. Implementing a VAT is not a quarter‑point tweak; it is a permanent institutional change. Furthermore, the multiplier estimates of this single working paper are well outside the range of credible estimates in the empirical literature.
Holcombe’s back‑of‑the‑envelope exercise, though simple, conveys the growth stakes. In the period from 1999 to 2004, U.S. real GDP growth averaged 3%. Had a 5% VAT been in place, growth would have been roughly 2.73%; with a 10% VAT, 2.46%. Even if one treats these numbers as indicative rather than definitive, they illustrate how the VAT’s layered burdens (higher overall taxation, compliance costs, threshold distortions and price effects) translate into a lower growth path.
Innovation and Capital Markets
VATs create special problems in financial services, where it is technically difficult to apply the tax to implicit‑fee intermediation. Many jurisdictions therefore exempt financial services from VAT, which has the unintended effect of blocking input credits for banks and insurers. The hidden VAT then becomes a cost that is passed on to business users of finance, raising intermediation costs and dulling investment. Proposals to fix this administratively exist on paper but are complex and largely untested at scale. For an economy whose comparative strength lies in deep, innovative capital markets, this is not a trivial footnote.
A Road Not Taken: Better Paths to a Sound Budget
Rejecting a VAT is not a plan to ignore the debt. It is a recognition that how a nation consolidates matters as much as whether it consolidates. If the evidence indicates that a VAT will expand government while imposing growth costs, the responsible course is to pursue alternative tools that work.
The first is expenditure‑based reform. In the U.S. context, this means entitlement changes that alter the long‑run trajectory of mandatory spending, healthcare reforms that improve procurement and price signals, and procedural rules that bind appropriations to real caps rather than aspirational targets. Consolidations built around spending cuts have better track records because they improve expectations about the future tax burden and the return to investment.
The second is productivity‑oriented tax simplification within the existing framework: broader bases and lower rates where possible; fewer targeted tax expenditures; stable treatment of saving and investment. Simplicity is not merely a cosmetic virtue. It reduces compliance costs, strengthens the rule of law and improves the allocation of capital by reducing the return to lobbying.
The VAT’s virtues on paper—broad base, neutrality, easy administration—bear only a partial resemblance to its performance in the world. Cross‑country evidence indicates that VAT adoption is associated with a durable increase in the total tax take. After adoption, rates rise, and government grows. Micro‑evidence from the U.K. shows that the VAT’s registration threshold discourages small firms from expanding, while European data documents substantial fixed compliance costs, especially for the smallest firms. Prices ratchet upward more readily than they fall when VAT rates change, and the OECD cautions that consumption taxes can depress labor supply and demand much like proportional income taxes. In a federal system that already relies heavily on retail sales taxes at the state level, a federal VAT would compound welfare losses and erode state revenues rather than replace existing distortions.
Nor is the VAT a path to fiscal responsibility. The consolidation literature is clear: Spending‑based adjustments succeed; tax‑based adjustments don’t. Monetary policy cannot neutralize the structural drags of a broad new consumption tax. In the real world, countries that adopt a VAT do not repeal their income taxes; they fund higher levels of spending with a larger mix of taxes. There is little reason to think the United States would be the exception.
A federal VAT would thus be a poor fit for the United States: It would increase the size of government, complicate federalism and dampen entrepreneurship and growth. If the aim is to stabilize the debt while preserving a dynamic economy, the United States should pursue the consolidation strategies that work—spending restraint and productivity‑enhancing reform—rather than importing a tax that works mainly as a money machine.
[1] These are the 15 European countries that adopted a VAT between 1967 and 1995, excluding resource-dependent Norway and former Soviet states.