Value-Added Tax

A value-added tax (VAT) is a consumption tax levied on the value each business adds at every stage of production and distribution, with the final consumer ultimately bearing the burden. It is collected in pieces along the supply chain: sellers charge tax on their sales, credit the tax they paid on inputs, and remit the difference to the state. Because tax is assessed on transactions rather than on a single retail sale, the tax base is spread across the economy, which is why VAT is administered through business records and invoices instead of only at the checkout. VAT has become one of the most widespread taxes in the world, used in roughly 175 of the 193 UN member states, and it is a major source of government revenue in many of them. This entry explains the mechanism and its calculation methods, the rules on rates and exemptions, and how VAT differs from older turnover taxes and retail sales taxes.

The entry then traces the tax's intellectual and institutional history, from early twentieth-century proposals and France's 1954 adoption to its spread through the European Community and then across Africa, Asia and Latin America. It goes on to examine the economic effects of VAT, including who bears its burden, how it affects revenue and other taxes, and its distortions and regressivity. Administration and compliance problems, especially invoice-based fraud, are discussed alongside the treatment of exports, cross-border trade and digital services. The final section sets out the main policy debates and open empirical questions, such as the tax's fairness, its place in tax reform, and why some large economies still lack a federal VAT.

Definition and basic principle

A value-added tax (VAT) is a consumption tax levied on the increase in value that occurs at each stage of production and distribution, rather than on the full price of a finished good. Under a retail sales tax, the whole price is taxed once, at the final sale. VAT instead divides the chain into steps. Each seller charges tax on its own sale price and then deducts the tax already paid on its purchases, so the amount remitted at each step corresponds to the margin that seller added. A simple example at a 20 percent rate: a farmer sells wheat for 100 and charges 20 in tax, passing 20 to the government. A baker buys the wheat for 120 and sells bread for 200 plus 40 tax, then deducts the 20 paid to the farmer and remits 20. The government has collected 40 in total, which equals 20 percent of the final price of 200. The chain of deductions is what keeps the tax from compounding on itself, and it is also the reason the base of the tax is conceptually linked to the value added measured in national accounts and to gross domestic product.

VAT is classified as an indirect tax on consumption. Its legal obligation falls on registered businesses, which must collect the tax from their customers and remit it to the state, while the economic burden is intended to fall on the household that buys the final good or service. This separation is the defining feature of an indirect tax, and it is why the concept of tax incidence is central to any analysis of VAT. Businesses act as unpaid tax collectors for the state. Because every firm in the chain can recover the tax it paid on inputs, the tax is neutral from the seller's standpoint in the sense that it does not remain in the firm's cost structure. The only point at which no credit is available is final consumption, so the tax ultimately lands on the buyer. VAT is also distinct from a tariff, which is levied on goods crossing a border; a VAT is charged on domestic and imported goods alike, with adjustments at the border that are discussed in later sections.

The term used for this tax varies by jurisdiction. The expression value-added tax is common in Europe, Latin America and much of Africa. The goods and services tax (GST) label is used in Australia, New Zealand, Singapore, Canada (for its federal component) and India, while Jamaica uses the general consumption tax (GCT) label. Japan's version is called the consumption tax. These names reflect local legislative choices and political history rather than different economic mechanisms: the underlying method of taxing value added through a credit system is largely the same. Differences in naming can, however, obscure important design variations, such as the number of rates, the treatment of exemptions and the rules on registration, which are examined in later parts of this entry. In practice, a consumer who pays a VAT, a GST or a consumption tax at the till faces a very similar outcome, even though the tax is administered through a multi-stage process that is invisible at the point of sale.

Calculation methods and mechanics

The credit-invoice method

Most countries that levy a VAT use the credit-invoice method. Each registered business charges tax on its taxable sales (output tax) and deducts the tax it paid on its business purchases (input tax). The difference is remitted to the tax authority, or refunded if it is negative:

Because each seller must show the tax on its invoice, the buyer's credit can be checked against the seller's records. This is what prevents the tax from cascading as a turnover tax would. A simple chain at a 20% rate shows the mechanics. A manufacturer sells goods for 100 plus 20 tax and remits 20. A wholesaler buys for 120, sells for 200 plus 40 tax, claims the 20 credit and remits 20. A retailer buys for 240, sells for 300 plus 60 tax, claims 40 and remits 20. The three remittances total 60, which is exactly 20% of the final price of 300, even though tax was collected at every stage.

Subtraction and addition methods

The same base can be computed without an invoice trail. Under the subtraction method, a business's base is its taxable sales minus its purchases from other businesses, and the rate is applied to that net figure once per period. Tax does not appear on individual invoices, so the method measures value added directly but makes it harder to verify exemptions and exports. The addition method works from the other side. It builds the base by summing the factors of production, such as wages, rent, interest, depreciation and profit, so the result resembles an income-type measure. In a broad-based system the two produce the same total, but they can be treated differently in law. A 1989 Washington State analysis noted that a subtraction-method tax may look more like a sales tax, and an addition-method tax more like an income tax, even though the two are economically identical. Michigan's Business Activities Tax (1953–1967) was an income-type VAT using the subtraction method.

Accounts-based VAT and Japan

An accounts-based VAT computes tax as a percentage of the difference between sales and purchases recorded in the taxpayer's books, rather than through invoice-level credits. Among countries that collect a VAT, Japan has been the notable user of this approach. Before October 2023, Japanese taxpayers did not rely on a consumption-tax invoice as proof of a credit. Instead, they <!-- -->calculated creditable input tax from company records and prescribed tax books, and any invoice could support the claim. The Qualified Invoice System, in force from 1 October 2023, moved Japan toward the invoice-based model. Under it, input credits are in principle available only for purchases backed by a qualified invoice from a registered issuer, and purchases from unregistered suppliers gradually lose credit eligibility. Japan's consumption tax now carries a 10% standard rate and an 8% reduced rate, and the reform was partly designed to support the multiple-rate system.

Refunds and payment timing

A negative net position arises when input tax exceeds output tax, which is typical for exporters, whose sales are zero-rated, and for firms making heavy capital investments in a period. The IMF's guidance is that such excess credits should be refunded promptly, but practice varies. Immediate refund systems are common in some EU countries, while other regimes carry the credit forward. In the UAE, for example, the excess is offset against future output tax, which allows audits before any cash refund is made. Filing is usually quarterly or monthly, and where tax is payable, the UAE example requires payment within 28 days after the end of the quarter. Refund eligibility also depends on timing: in that regime, the refund is generally claimed in the first period after the taxpayer has both received the supplier's invoice and paid the supplier.

Rates, exemptions and zero-rating

Standard, reduced, zero and exempt treatment

Most VAT systems apply one standard rate to the bulk of supplies and layer reduced or special treatments on top for selected goods and services. The treatments differ mainly in whether the seller can recover the tax on its own purchases. A zero-rated supply is a taxable supply with a 0% rate, so the vendor charges no VAT but can still claim input tax deductions on related expenses. An exempt supply, by contrast, is not subject to VAT, and the vendor cannot claim input tax deductions on related expenses. Because the seller's unrecovered input tax is folded into its prices or margins, exemption is sometimes described as "hidden VAT". This is why an exemption interrupts the chain of credits that the tax is designed to maintain, while zero-rating keeps it intact. Businesses that make both taxable and exempt supplies must apportion their input deductions according to how far the goods or services were used for taxable supplies.

Rate variation across countries

Headline rates differ considerably. EU VAT rates run from 17% in Luxembourg to 27% in Hungary in 2026, and the EU framework itself requires member states to apply a standard rate of at least 15%. Outside Europe the range is wider. The UAE and Oman each levy a VAT rate of 5%, and reduced rates are uncommon in the Gulf states, most of which have a single flat rate. Bahrain's standard rate is 10% with no reduced rate. Japan applies a 10% consumption tax rate with a reduced 8% rate on food and beverages, while Singapore's GST stands at 9%. Some states have not adopted VAT at all: Kuwait and Qatar have yet to implement VAT.

Headline rates can also mislead. In Hungary, for example, meat, eggs, milk and fish are taxed at 5%, so the effective burden on a basket of groceries can be far below 27%. Commentary on the sector notes that almost every country also has reduced rates for essentials such as food, medicine, children's clothing and books, and some items are exempt entirely.

Typical reduced and zero-rated categories

The goods most often given preferential treatment are staples and health-related products. Zero-rating is commonly applied to prescription medications, medical supplies and basic groceries, and some countries extend it to books. Passenger transport is treated differently across jurisdictions, but zero-rating is also applied to certain transactions involving international transportation. Exemptions, which carry no input credit, are more often used for services. Such supplies include health, education, financial and insurance services. Governments choose between the two forms according to whether they want the final price to fall (zero-rating or a reduced rate) or simply want certain activities left outside the tax (exemption), a choice that has distributional and administrative consequences.

The destination principle and export rebates

VAT on international trade is based on the destination principle. Under this principle exports are free of VAT and imports are taxed on the same basis and at the same rate as local production. The OECD describes the consumption tax as one that is ultimately levied only on the final consumption that occurs within the taxing jurisdiction. In practice, the adjustment operates at the frontier: a tax is levied at the border on incoming goods and rebated at the border for outgoing goods. The World Trade Organization's rules sanction this approach, and the GATT explicitly allows, without requiring, the refunding of indirect taxes paid in the production of exported goods. Within the EU, the VAT Directive treats exports as "exempt with credit," which is functionally zero-rating.

Not every country refunds exports fully. China is a notable case: its standard rate of 17 percent applies to goods sold on the domestic and foreign markets, but rebates for exported goods vary by commodity and range from 0 to the 17 percent VAT rate. Zero-rating also depends on proof. Exporters in many systems must keep evidence of export, such as outward permits, bills of lading or air waybills, to support zero-rating.

Comparison with other consumption taxes

VAT is most often compared with three alternatives: the retail sales tax, the turnover tax, and, on the income side, the income tax. Each comparison explains part of why VAT spread after the 1950s, and each also shows where its advantages are limited.

Retail sales tax

A retail sales tax (RST) is collected at a single point, the final sale to a consumer. This is a one-off tax collected only at the point where the final sale to the consumer happens. Because the levy never reaches earlier stages, a firm has no mechanism to offset the tax it paid on its own purchases, so input taxes can stay buried in prices. In practice, the separation is less clean than the textbook version. State RSTs do not exempt all business-to-business sales, leading to double taxation (called cascading). One estimate suggests around 40% of total sales tax revenue comes from tax levied on business-to-business sales.

The main case for VAT over an RST rests on enforcement. The government has no record of transactions with which to verify retailers' tax payments. Under a VAT, by contrast, the chain of crediting creates a natural audit trail, and the seller has more incentive to report the transaction and pay tax. The compliance side is more mixed. An RST may be cheaper to administer at low rates, while a VAT may be less expensive if high rates are needed. Businesses with no retail sales typically do not file under an RST, whereas under a VAT they would still need to file and pay.

Turnover taxes

A turnover tax is levied on gross receipts at each sale, whether the buyer is another firm or a consumer. Turnover taxes are levied on revenues and do not allow for input deductions, resulting in tax cascading where final goods are taxed multiple times throughout the production. The arithmetic is simple: two successive 5 percent levies produce an effective rate of 10.25 percent (1.05 × 1.05 = 1.1025), and the effect grows with each additional stage. Cascading also distorts business structure. In principle, turnover taxes distort business organizations to favor vertical production. The efficiency cost is not limited to the tax on inputs, either. By taxing production, even at a low nominal rate, a turnover tax reduces output in a way that a consumption tax does not.

Turnover taxes were once common. In 1967, 19 Organisation for Economic Co-operation and Development countries had some form of wholesale, retail, or "turnover" tax, and by 1995, all had converted to value-added taxes (VATs) that collect revenue at each stage of production. Today they survive mainly as a simple form of taxation for small taxpayers, or as a minimum tax on businesses.

Empirical evidence from switching regimes

Several reform episodes have been studied as natural experiments. Canadian provinces that replaced sales taxes with VATs saw a sharp rise in capital spending: a study cited by the IMF found that the transition from a retail sales tax (largely applying to business inputs) to a VAT led to a 17 percent increase in investment in machinery and equipment. In India, Indian firms' sales increased by 57% in the median run after the move from sales taxes to a VAT. In China, a reform that replaced turnover taxes with VAT for service industries effectively removed tax cascading. The authors find that sales of reformed service firms rose by around 22 percent, with the gains mainly driven by increased outsourcing from manufacturing firms.

Why the invoice trail matters

VAT's enforcement advantage over single-stage taxes comes from the purchase invoice. A purchase invoice, which is generated by businesses and not the government, automatically entitles the purchasing firm to a tax credit from the government. Because the credit depends on the seller's invoice, the VAT generates third-party information on business-to-business (B2B) transactions through the input tax credit mechanism, raising firms' cost of underreporting their true revenue especially from B2B sales. A Chilean study using two randomized experiments among over 400,000 Chilean firms found that announced monitoring had less effect on transactions that left a paper trail, which the authors read as a preventive deterrence effect.

The limits are equally important. The self-enforcing logic works well in B2B (business-to-business) transactions but fails at the retail stage (business-to-consumers, B2C). This is the so-called last-mile problem. Brazil's single-stage manufacturers' tax was designed to combine the two: it combines the relatively low compliance burden of an RST by requiring remittance only at a single stage, with the improved enforcement under a VAT by requiring upstream firms to remit.

Comparison with income taxes

The contrast with income taxes concerns the tax base, the treatment of investment, and incidence. An income base includes all income from both labor and capital (e.g., wages and interest on investments). A consumption base instead covers all money spent on goods and services. On capital spending, a consumption-type VAT treats capital purchases the same way as the purchase of any other input, which excludes all capital goods purchases from the tax base (i.e., it is equivalent to expensing under an income tax). An income-type VAT takes a different route, since the VAT paid on the purchases of capital inputs is amortized (credited against the firm's VAT liability) over the expected life of the capital inputs. Because the consumption base is narrower, revenue neutrality implies a higher rate. A consumption tax is likely to have a higher tax rate than an income tax if the taxes are to be revenue neutral, because the tax base under a consumption tax is necessarily smaller than the tax base under an income tax. On incidence, a VAT does not tax the returns (such as dividends and capital gains) from new capital investment, and income from capital makes up a larger portion of the total income of high-income households.

Origins and early history

The idea of taxing only the value added at each production stage is usually traced to Germany in 1918. The industrialist Georg Wilhelm von Siemens... (no cite; body only)

The idea of taxing only the value added at each production stage is usually traced to Germany in 1918, when the industrialist Georg Wilhelm von Siemens proposed replacing the German turnover tax with a tax on value added. The turnover tax was levied on gross sales at every stage, so the same goods were taxed repeatedly as they moved from supplier to manufacturer to wholesaler to retailer. Exports carried the accumulated charge with them, which made the levy a cascading tax that distorted production choices. Siemens was motivated by these effects, and he saw a tax that counted only the added value at each step as a neutral alternative. His proposal went nowhere at the time. Germany kept a turnover tax in some form until 1968, when it was replaced. Historians also disagree about the origin of the idea. Some credit the American economist Thomas S. Adams, who wrote between about 1910 and 1921 and envisaged a value-added levy as a substitute for business income taxation, while the attribution to Siemens was later promoted by his younger brother, Carl Friedrich von Siemens, in a 1921 book.

The first working modern system came from France. Maurice Lauré, a senior official in the tax administration, designed the tax after the Direction générale des impôts was created in 1948 and he helped establish it. The name taxe sur la valeur ajoutée is credited to him. The tax took effect on 10 April 1954. Accounts differ on where it first applied: one source says it was first introduced in the French colony of Ivory Coast and extended to metropolitan France in 1958, while others describe a national introduction in 1954. In either case, the initial coverage was limited. The tax was aimed at large businesses first, and it was then extended in stages until it covered every business sector. It replaced an earlier system of turnover taxation that had been collected in installments, so the change simplified a complicated levy rather than adding a new burden on top of it.

Lauré's central design choice was to make every taxpayer in the production chain responsible for accounting for the tax on its own sales and purchases. Under the older arrangements, collection relied mainly on retailers or on assessments made by the tax authorities. Under the new system, a registered firm could recover the tax it had paid on inputs, provided it held documentation, and the tax liability of one firm was checked against the claims of its customers. Supporters argued that this made the tax more difficult to evade and less dependent on discretionary assessment. They also presented it as fairer than earlier taxes, which were often criticized for loopholes and for favoring well-connected firms. Early debate was not uniformly favorable, however. Critics questioned whether a tax on consumption fell too heavily on households regardless of their incomes, and the administrative burden on smaller firms was a recurring concern.

The design spread quickly after 1954 for several reasons. It produced revenue reliably, it reduced evasion through the invoice trail, and its apparent success in France encouraged other governments to examine it. Within the decade, the European Economic Community began to consider a common system, and the adoption of VAT by its members is discussed in the section on international spread.

International spread and regional variants

From the 1960s proposals to the Sixth Directive

Harmonising national turnover-tax laws for the common market was on the agenda from the start of the European Economic Community. The founding members of the EU agreed to "consider how the legislation of the various Member States concerning turnover taxes … can be harmonised in the interest of the Common Market". Progress was slow because member states were almost entirely free to provide for exemptions in their national VAT systems. The Commission's 1973 draft sought much closer alignment, proposing rules on a limited number of uniform and compulsory tax exemptions.

Directive 77/388/EEC, dated 17 May 1977 and usually called the Sixth Directive, aimed to give every member state a broadly identical VAT base, so that the same transactions were taxed everywhere. It did not attempt any general harmonisation of VAT rates. It did set limits, requiring a minimum standard rate of 15%, and reduced rates of VAT must be above 5%. The directive was later amended many times, and a consolidated text, the 2006 recast, took effect on 1 January 2007. The recast consolidates and updates the EU's common VAT rules, covering scope, rate structures, place of taxation and obligations of taxable persons. Internal border controls were removed at the end of 1992, after which a transitional regime applied from January 1993 (a transitional VAT system was adopted on 1 January 1993, owing to a lack of political will and for technical reasons).

The United Kingdom

Britain's adoption was tied to its accession to the EEC. Edward Heath's Conservative government took office in 1970 with the intention of joining the European Economic Community, with a value added tax a requirement for entry. The Heath government therefore abolished the SET and the purchase tax, and introduced the UK's first general sales tax in the form of VAT in 1973. VAT began on 1 April 1973, with essential goods such as food, fuel and housing exempted, while all other goods and services were charged at a single rate of 10%. The Labour government later lowered the standard rate to 8% in 1974 (the standard rate of VAT was reduced from 10% to 8%). The standard rate now stands at 20%, the default rate since 4 January 2011.

Adoption elsewhere and national variants

Outside Europe, adoption accelerated in the 1980s and 1990s, often alongside tariff reductions that made domestic indirect taxes more important. Japan introduced a 3% consumption tax in 1989, which has since risen to a 10% standard rate with a reduced rate for food and certain other items. New Zealand adopted its GST in 1986, Canada introduced a federal GST in 1991, and Singapore followed in 1994 at 3%. By the end of the twentieth century, VAT-type taxes had spread across much of Africa, Asia and South America.

Several systems depart from the classic model. Canada's structure is layered: 5% federal GST plus provincial taxes, with some provinces merging their levies into a combined harmonised tax. India's GST, introduced in 2017, replaced a range of central and state indirect taxes. Before the September 2025 reform it used four slabs (5%, 12%, 18% & 28%). The reform took effect on September 22, 2025, and sources differ on how to describe the result. One counts only three main slabs (0%, 5%, 18%) and one premium slab (40%), while another describes just two major GST rates (5% and 18%), along with 0% for essentials and a 40% sin/luxury tax.

Current global coverage

VAT is now the norm in most of the world, but there are notable gaps. The United States has no federal VAT. Its federal system relies on income taxes, while individual states levy retail sales taxes. Hong Kong also has no VAT. Rates vary widely: Hungary's standard rate of 27% is among the highest in the world, while the United Arab Emirates and Saudi Arabia introduced 5% rates in 2018. The EU framework itself leaves rate levels to member states; the VAT Directive does not set out specific VAT rates.

Economic effects, incidence and revenue

Revenue role

VAT is one of the largest sources of government revenue in the world. Across the OECD, VAT accounts for more than one-fifth of total tax revenues (20.8%) on average, representing 20% or more of total taxes in 21 of the 37 OECD countries that operate a VAT. The 2025 OECD tables put the 2023 figure slightly lower, at 20.5% of total tax revenue. Elsewhere the share is higher. In Asia and the Pacific, VAT is the largest contributor to tax revenue among major tax types, accounting for 25.8% of total tax revenue, and the regional averages for Africa and Latin America and the Caribbean were higher still in the most recent data.

Yield per point of rate is a different question from the share of revenue. The OECD's VAT Revenue Ratio compares actual collections with what a single standard rate on the entire potential base would raise. It combines the effect of loss of revenues due to exemptions and reduced rates, fraud and non-compliance. The unweighted OECD average was 0.58 in 2022, meaning that actual collections were on average a little over half of the theoretical benchmark.

Neutrality and distortions

In principle a broad-based VAT is close to neutral with respect to production. The IMF argues that because the tax does not change the prices firms ultimately pay for their inputs, it does not distort production choices and avoids the cascading that arises when tax is levied on inputs and again on the output that embodies them. Because the VAT does not affect the prices firms ultimately pay for inputs, it does not distort production decisions and does not create "cascading". This neutrality depends on the tax applying to nearly all transactions at one rate.

Exemptions and reduced rates break that condition. Theoretical work by Keen, cited in the IMF's 2025 working paper, shows that the welfare cost of a turnover-type or VAT-type distortion rises as the share of taxed inputs rises, and that it also grows with the elasticity of substitution between inputs. The same paper notes that the share of taxed inputs tends to be higher under a turnover tax than under a VAT. Reduced rates also shift household consumption toward favored goods, creating a deadweight loss that a uniform rate would avoid.

Incidence and regressivity

Whether VAT is regressive depends heavily on the denominator. Consumption taxes are regressive when measured as a percentage of household income, but are generally either proportional or slightly progressive when measured as a percentage of household expenditure. Thomas's 2022 study of 27 OECD countries, which adopts the expenditure basis, found the VAT either roughly proportional or slightly progressive in most of the 27 OECD countries examined. Redistribution through the VAT is nonetheless small. One review found that the redistributive effect of VAT is very low under both methods used.

Reduced rates are a central part of this debate. OECD work distinguishes between rates designed to help the poor and rates motivated by other goals. Reduced rates on food and energy are found to have a small progressive effect, but to be poorly targeted. Reduced rates on books, restaurant meals and hotel accommodation are often found to be so poorly targeted that they have a regressive effect. A 2025 study of EU countries found that in 10 out of 26 countries the redistributive effect of reduced VAT rates is negative. Evidence from developing countries is more mixed and more country-specific; a study of Indonesian food spending found a regressive pattern.

The policy implication most often drawn from this evidence is that transfers matter more than rate structure. An OECD study concluded that removing reduced rates could raise revenue with smaller distributional effects than previously assumed, but that such reform should be accompanied by additional measures – for example, targeted tax credits or benefits – to compensate poorer households.

Replacing other taxes

Proponents argue that a VAT could finance cuts in income and corporate taxes, which are often seen as more harmful to saving and investment. Hufbauer's proposal to replace the US corporate income tax with a small VAT is a frequently cited example. Henry Aaron responded that the "effect of the tax swap on saving is many fold too high." More recent modelling is also contested. A Tax Foundation estimate suggests that a 5% VAT in place of the existing corporate tax would increase GDP by 5.5% and raise $3.1 trillion in federal revenue on a dynamic basis. A Urban Institute analysis of a US broad-based VAT found that the new tax would largely be offset by reductions in other taxes, with the revenue partially offset by $56 billion in reduced individual income tax liability, $30 billion in reduced payroll tax liability and $11 in billion reduced corporate income tax liability.

Empirical evidence is more limited. The IMF's 2025 review reports that, for middle- and high-income countries, increasing reliance upon consumption taxation while reducing that on the personal (and, even more, than corporate) income tax has been associated with faster growth. The relationship is less clear for low-income countries. Earlier IMF work on past swaps in the United Kingdom, Sweden and Italy found that exchange-rate and world-demand fluctuations could make the effects of such swaps appear insignificant. Pass-through is another open question: some evidence suggests that businesses pass on tax increases but do not pass on tax reductions, which would weaken the expected gains from a corporate tax cut. Finally, VATs have generally come to replace other forms of consumption tax and exist simultaneously with a corporate income tax, so the pure substitution scenario is rarely observed in practice.

Compliance, fraud and administration

Registration, invoicing and filing

A VAT system relies on businesses that register with the tax authority, show the tax on their invoices, keep transaction records and file returns at regular intervals. In the UK, businesses charge VAT on their sales and include it on invoices, submit online returns to HMRC, usually every quarter, and keep digital records that meet the Making Tax Digital rules. Registration is compulsory once taxable turnover passes a threshold, which has applied since 1 April 2024 and is measured over any rolling 12-month period. Firms below the line may register voluntarily, which lets them reclaim input tax on purchases. Compliance costs have two parts: a one-off cost to initially register, and then a periodic cost of completing a VAT return. A UK study notes that these costs are relatively low by international standards, but are not negligible.

Carousel and missing-trader fraud

Missing-trader intra-community (MTIC) fraud exploits the zero-rating of goods moving between EU member states. Traders charge VAT on a sale but abscond with the tax rather than paying it to the tax authority. The fraudulent company is the "missing trader", and the buffer traders in the chain distance the missing trader from the end consumer, which makes detection harder for tax authorities. In a carousel variant, a group of traders in several member states cycles products, real or imaginary, through a series of bogus sales. Each link can generate input credit claims, so downstream buyers claim input VAT deductions on invoices from the missing trader, while the tax authority loses both the uncollected output tax and the refunds it pays out. The scale can be large. Operation Admiral is described as the largest VAT carousel fraud ever investigated in the EU, involving over 400 companies across at least 16 EU countries. Fraud is frequently concentrated in telecoms, mobile phones and other high-value electronic goods, precious metals and gas. Estimates of the EU-wide VAT gap also vary by source: one European Parliament-related summary puts it at roughly €90 billion, while another reports that it reached €128 billion in 2023. The figures likely reflect different methodologies.

Detection by tax authorities

Because invoice trails leave a record of every transaction, authorities can analyse the trading network itself. One study on Bulgarian data proposes methods that detect around 50 percent of the VAT fraud, outperforming well-known techniques that ignore the information provided by the network of VAT transactions. Such machine learning approaches draw on network analysis and anomaly detection. Their design is constrained by data: a study of tax fraud notes that such limitations, coupled with a low rate of tax return labels, restrict the usefulness of supervised techniques, so unsupervised methods that flag outliers are also used. Practice is moving quickly. Poland's system analyses data provided by banks and credit unions on taxpayers' accounts and transactions on a daily basis. In Italy, electronic invoicing lets authorities flag fraudulent supply chains immediately when combined with automated analysis. Authorities also assign risk scores to taxpayers based on a range of criteria, which guides audit selection. Several jurisdictions, among them Italy, Spain, France and China, have already begun deploying AI and data-mining tools to identify VAT anomalies.

Compliance costs and small-business thresholds

Most jurisdictions exempt the smallest firms from registration to limit these fixed costs, but the threshold itself creates distortions. In the UK, the threshold has no phase-in period or tapering, so the moment turnover crosses it, registration is required, whether by £5,000 or just £5. This "cliff edge" can influence behaviour: many businesses deliberately limit their turnover to avoid crossing the threshold and the administrative burden, increased costs and competitive disadvantages that follow. An IMF study of UK firms finds evidence that annual growth in turnover starts to slow when the firm turns over close to the threshold. Business-to-consumer firms are especially exposed, since they cannot easily pass the tax on: one accountant notes that once a sole trader exceeds the £90,000 threshold, they're suddenly 20% more expensive than their competitors. The Federation of Small Businesses has estimated that the extra bureaucracy of being VAT-registered adds £4,100 on average to the running costs of a business. Policy proposals in the UK have therefore called for a higher threshold or simpler rules for small firms.

Cross-border trade and digital services

Because VAT is a consumption tax, international transactions raise a basic question: which jurisdiction may tax a given sale. Most systems follow the destination principle, under which a transaction is taxed where the good is consumed. Exports are relieved of tax and imports are charged on arrival, which keeps the tax from distorting trade.

Border adjustments for goods

Goods crossing a customs frontier are normally charged import VAT at the importing country's rate, collected at entry. A registered importer can usually reclaim that charge as input tax, so for businesses it is neutral and the final consumer bears it. Exports are generally zero-rated: the seller charges no VAT on the sale abroad and recovers the input tax on its own costs, so the goods leave tax-free.

Small consignments were long a gap in this system. The EU removed its import VAT exemption for small consignments, which was up to EUR 22, so that all goods imported in the EU are subject to VAT. It also created the Import One-Stop Shop (IOSS), which simplifies the declaration and payment of VAT for distance sales of low value goods not exceeding 150€.

Place of supply for services

Services do not pass through customs, so their VAT treatment depends on statutory rules. Cross-border services do not pass through physical customs borders. Instead, their VAT treatment is governed by statutory "place of supply" rules. These place-of-supply rules decide which state holds the taxing right.

For business customers, the general rule places the supply where the customer is established. For business-to-business (B2B) supplies, the place of supply is normally where the customer belongs for VAT purposes. In such cases the recipient usually accounts for the tax through a reverse charge. For consumers, the starting point is the supplier's location, For business-to-consumer (B2C) supplies, the place of supply is generally where the supplier belongs for VAT purposes. Many exceptions override this default. The EU list, for example, covers B2C supplies of telecommunications, broadcasting and e-services, passenger transport, restaurant and catering services and services in respect of or related to admission to cultural, artistic, sporting, scientific, educational, entertainment or similar events. Land-related services are taxed where the land lies. For digital services sold to consumers, the place of supply for consumer sales is where the consumer resides.

Platforms and the EU One-Stop Shop

Online marketplaces have been drawn into the chain of collection. Since 1 July 2021, the EU has allowed marketplaces to become the deemed supplier in some cross-border B2C transactions, making them responsible for collecting, reporting and remitting the VAT due from the consumer. A platform qualifies when it controls the terms of sale, authorises the charge to the customer, and orders or delivers the goods. The deemed-supplier rule covers goods valued below €150 imported by EU or non-EU sellers and sold to EU customers, as well as goods of any value sold by a non-EU seller to an EU customer. Platforms that do not meet this test still face record-keeping duties.

To simplify compliance, the One-Stop Shop (OSS) lets a business usually register once (e.g., in Poland), apply the VAT rate of the buyer's country, and file one quarterly OSS return covering all Member States of consumption. Non-EU businesses supplying services to EU consumers use a separate non-Union scheme. A threshold also applies: the EUR 10,000 EU-wide threshold applies to combined B2C distance sales of goods and TBE services, and below that level a seller may generally keep charging its home-country rate. Other jurisdictions have adopted similar regimes; the UK, for instance, introduced marketplace VAT rules on 1 January 2021.

Coordination problems

Mismatches between national rules are the main source of error. Misapplying them creates double taxation or unexpected liabilities. Even under shared principles, different countries have different interpretations of the same provisions, so a service may be taxed twice or not at all. Thresholds, definitions of "electronic" services, and the treatment of goods shipped through intermediaries all vary. International bodies such as the OECD have issued guidelines for cross-border VAT/GST, but these are not binding and depend on national implementation. Coordination thus remains a central challenge as trade and digital services continue to grow.

Policy debates and open questions

Fairness and regressivity

The most persistent dispute concerns whether VAT is fair. Lower-income households save a smaller share of their income, so a tax on consumption takes a larger slice of their resources when burdens are measured against income. VAT has often been deemed regressive because the propensity to save rises with income, and low-income households therefore spend a greater share of their income on consumption. Measured against expenditure instead, the picture changes. An OECD study of 27 member countries concluded that the VAT is found to be either roughly proportional or slightly progressive in most of the 27 OECD countries examined. The same study cautions that broad-based VAT systems that have few reduced VAT rates or exemptions can produce a small degree of regressivity.

Reduced rates are the usual tool for softening the burden, but evidence on their targeting is poor. An earlier OECD review found that reduced rates on food and energy do favour poorer households proportionately, yet rich households receive as much aggregate benefit from a reduced VAT rate as do poor households. Reduced rates on books, hotel accommodation and restaurant food may even be regressive. The World Bank concludes that the optimal indirect tax literature finds no redistributive role for reduced VAT rates when other more direct instruments are available. The mainstream answer is therefore to compensate poor households through the transfer system, since ensuring the progressivity of the tax-benefit system as a whole is the more reliable way to offset lost purchasing power.

Rate changes as fiscal and behavioural tools

Governments adjust VAT rates for revenue, cost-of-living and sectoral reasons, and recent examples show the range. Ireland has confirmed a VAT rate reduction from 13.5% to 9% for food, catering services and hairdressing, effective 1 July 2026. Austria's National Council has approved a proposed cut in the VAT rate for certain staple foodstuffs from 10% to 4.9%, intended to apply from July 1, 2026, subject to completion of the remaining legislative process. In the United Kingdom, the government has cut VAT on domestic electricity from 5% to 0% for six months, from 1 October 2026 to 31 March 2027, while gas remains at the reduced 5% rate.

Some changes run in the other direction. Belgium's 2026 federal budget maintains the standard VAT rate at 21% while increasing several reduced VAT rates from 6% to 12%, and in Germany the reduced rate on restaurant meals was made permanent at 7%. The stated aim there was to support the hospitality sector's recovery while keeping menu prices down. Whether such cuts are passed through to shelf prices, and how far they stimulate demand, are empirical questions that vary by sector and market structure.

Proposals to adopt VAT

Countries without a VAT face a recurring debate about whether to adopt one. The clearest case is the United States, which is the only major developed economy without a federal VAT or GST. Instead, state sales taxes are levied at point of sale, varying from 0% to 10.25% (combined state and city in some California cities). Proponents point to the invoice trail and to the revenue a broad base could raise, while opponents stress regressivity and the difficulty of coexisting with state sales tax systems. Other jurisdictions have moved in the opposite direction of the debate: Bhutan finally introduces a GST (5%) in place of its sales taxes. The design trade-offs in such reforms, including the choice between a single rate and multiple rates, remain contested.

Open empirical questions

Three empirical questions remain unsettled. The first is the size of the compliance gap. In the EU, estimates of the VAT compliance tax gap alone amount to 128 billion EUR in 2023. The gap is far from uniform, with Romania (30%) and Malta (24.2%) recording the highest VAT compliance gaps, while Austria (1%) and Finland (3%) performed best. Estimates also depend on methodology. The Commission's measures are top-down, comparing theoretical liability derived from national accounts with actual receipts, so their precision depends on the underlying macroeconomic data.

The second question concerns whether digital reporting closes the gap. A 2026 study of EU member states reports that the baseline staggered difference-in-differences estimate indicates a reduction in the compliance gap of 4.23 percentage points, although the authors acknowledge that the design of the reporting system matters. The result is a working paper, and its identification strategy is still being tested.

The third question concerns the revenue effects of rate changes and the long-run effects on growth. Estimates of how much revenue a rate change will raise or lose depend on assumptions about consumer response and avoidance, and the empirical record is thin. Claims about long-run growth effects, whether favourable or adverse, rely heavily on modelling choices and remain contested.

References

1. Value-added taxen.wikipedia.org
2. Value Added Tax (VAT) Definitiontaxfoundation.org
3. What Is a Value-Added Tax and Should the United States Have One?pgpf.org
4. What is value-added tax (VAT)avalara.com
5. value added taxfool.com
6. How to Manage Value-Added Tax Refunds in: IMF How To Notes Volume 2021 Issue 004 (2021)elibrary.imf.org
7. The Tax Credit Mechanism (Chapter 6) - Value Added Taxcambridge.org
8. Advantages of the Credit-Invoice Method for a Partial ...scholarship.law.georgetown.edu
9. Japan JP PINT tax qualified invoices updated - vatcalc.comvatcalc.com
10. Japan - Corporate - Other taxes - Worldwide Tax Summariestaxsummaries.pwc.com
11. Japan Qualified Invoice System (Invoice Seido): What Foreign Companies Must Knowaqpartners.jp
12. Calculate consumption tax under Japan’s Invoice Systemstripe.com
13. Alternatives Subcommittee Paper on Value Added Taxesdor.wa.gov
14. VALUE ADDED TAXES AND TRANSACTION TAXEScga.ct.gov
15. GAO Report to the Joint Committee on Taxation, U.S. Congress June 1989gao.gov
16. VAT Regulations - Part 2: Recoverability of Input Tax and Practical Accounting Issues - Lexologylexology.com
17. How would a VAT be collected?taxpolicycenter.org
18. Concept clarification: Zero-rated vs exempt supplies - Cliffe Dekker Hofmeyr (CDH)cliffedekkerhofmeyr.com
19. VAT Concepts Explained: Exemptions vs zero‑rating (and the hidden cost of exemption)vatupdate.com
20. Distinguish Between Zero-Rated VAT, Exempt VAT & Non-Taxable Supplies - Acuaccountsacuaccounts.com
21. EU VAT Rates by Country (2026): Table, Average, Highesteurofiscalis.com
22. Global VAT Rates by Country (July 2026)vatupdate.com
23. VAT Rates by Country 2026: Global VAT Rate Comparisoncleartax.com
24. Global VAT Rates by Country (2026)vatupdate.com
25. VAT Rates by Country 2026calcbase.io
26. VAT and Sales Tax Rates by Country (2026), Mapped - Mapprmappr.co
27. What is the Difference Between Exempt and Zero Rated?lytaxadvisors.com
28. OECD INTERNATIONAL VAT/GST GUIDELINES ...accountancyeurope.eu
29. A VAT Is Not a Disguised Trade Barriercato.org
30. Stop Saying a Value-Added Tax Is an Export Subsidy - Truth on the Markettruthonthemarket.com
31. 17 Interjurisdictional Issues in: The Modern VAT - IMF eLibraryelibrary.imf.org
32. Estimating the repercussions from China's export value‐added tax rebate policy* - Gourdon - 2022 - The Scandinavian Journal of Economics - Wiley Online Libraryonlinelibrary.wiley.com
33. Exempt vs Zero-Rated VAT: Key Differences Explainedvatit.com
34. VAT Rates by Country 2026: Full Global Comparison Tablecalculatevat.co.uk
35. Understanding VAT Risks: The Pitfalls of Export Exemptions Across the EU, UK, Norway, and Switzerlandvatupdate.com
36. VAT vs. Sales Tax: The Key Differencesvatai.com
37. Annual Report to Congress - Taxpayer Advocate Service - IRStaxpayeradvocate.irs.gov
38. VAT vs Sales Tax: Key Differences Explainedfonoa.com
39. Why is the VAT administratively superior to a retail sales tax?taxpolicycenter.org
40. NBER WORKING PAPER SERIES HOW DISTORTIVE ARE TURNOVER TAXES?nber.org
41. Efficiency Aspects of the Value Added Taximf.org
42. What is the experience of other countries with national retail sales taxes?taxpolicycenter.org
43. The VAT at 100: A Retrospective Survey and Agenda for Future Research - Joel Slemrod, Tejaswi Velayudhan, 2022journals.sagepub.com
44. Deterrence and Self-Enforcement in the Value Added Taxnber.org
45. Reaching the last mile: Digitalization of tax administration and VAT compliance at the retail stage - ScienceDirectsciencedirect.com
46. Consumption Taxes: An Overview - EveryCRSReport.comeverycrsreport.com
47. Who would bear the burden of a VAT?taxpolicycenter.org
48. Income tax credits for consumer services: A tool for tackling VAT evasion? - ScienceDirectsciencedirect.com
49. A Brief History of VATtaxually.com
50. European Union value added taxen.wikipedia.org
51. Maurice Laur%C3%A9en.wikipedia.org
52. A brief history of VAT in Europe and the UKicaew.com
53. 1 The rise of the value-added taxassets.cambridge.org
54. Ad valorem taxen.wikipedia.org
55. THE RATES OF VAT - European Parliamenteuroparl.europa.eu
56. National Tax Day: Remembering 3 VAT heroes - The Himalayan Times - Nepal's No.1 English Daily Newspaperthehimalayantimes.com
57. History of the EU VAT system · summitto blogblog.summitto.com
58. EU PERSPECTIVE ON VAT EXEMPTIONS Joachim Englisch OXFORD UNIVERSITY CENTRE FORoxfordtax.sbs.ox.ac.uk
59. VAT Directive: definitionitaxa.it
60. REPORT on the implementation of the Sixth VAT Directive: what is the missing part to reduce the EU VAT gap?europarl.europa.eu
61. Value added tax (VAT) at 50 - House of Lords Librarylordslibrary.parliament.uk
62. The Complete Guide To UK VATvatcalculatoruk.co.uk
63. Value-added tax in the United Kingdomen.wikipedia.org
64. GST 2.0 Reform 2025: Simplified Tax Slabs, Key Changes, and Industry Impactsageuniversity.edu.in
65. New GST Rates 2025: GST 2.0 Reform Bill Explainedujjivansfb.bank.in
66. New GST Tax Slabs in India 2025religareonline.com
67. GST Reforms in India: New Rates, Slabs and Key Changes in 2025cleartax.in
68. Consumption Tax Trends 2024oecd.org
69. Revenue Statistics in Asia and the Pacific 2025oecd.org
70. Finance & Development, June 2002 - The Allure of the Value-Added Taximf.org
71. No. 22 OECD Tax Policy Studies The Distributional Effectsoecd.org
72. Reassessing the regressivity of the VAT - Thomas - 2022 - Fiscal Studies - Wiley Online Libraryonlinelibrary.wiley.com
73. Contrary to Popular Belief, Value-Added Taxes Found to Be Slightly Progressivetaxfoundation.org
74. Policy Research Working Paper 10677 VAT Rate Structures in Theory and Practicedocuments1.worldbank.org
75. Taxing fairly or failing badly? Reduced VAT rates and redistribution - ScienceDirectsciencedirect.com
76. OECD Taxation Working Papers No. 49 Reassessing the regressivity of the VAToecd.org
77. chapter 11 Effects of VAT on the Economy in: Value Added Taxelibrary.imf.org
78. A Value-Added Tax: a Revenue-Neutral Alternative to the Corporate Income Taxtaxfoundation.org
79. Effects of Imposing a Value-Added Tax To Replace Payroll ...urban.org
80. Value-Added Tax: Administrative and Policy Issueselibrary.imf.org
81. The VAT at 100: A Retrospective Survey and Agenda for Future Research - Joel Slemrod, Tejaswi Velayudhan, 2022journals.sagepub.com
82. REVENUE STATISTICS 2025 Disentangling Personal Income Tax Revenueoecd.org
83. (PDF) The Distributional Effects of VAT Rate Changes: Evidence from Indonesian Household Food Expenditureresearchgate.net
84. Why the VAT registration threshold is a problem for small businesses - ByteStartbytestart.co.uk
85. The VAT Threshold Trap: Should Small Businesses Limit Growth to Avoid VAT? - U&Wuwaccountancy.com
86. Small Firm Growth and the VAT Threshold: Evidence for the UKimf.org
87. Missing trader frauden.wikipedia.org
88. Carousel fraudacfe.com
89. What is VAT carousel fraud?autofact-solutions.com
90. EUR 500 billion crime:- How carousel VAT fraud works and how you can stop itcomsuregroup.com
91. Carousel VAT Fraudstpaulschambers.com
92. Combating VAT Fraud Through Machine Learning and Predictive Artificial Intelligence - VAT and GST Tax Trendsglobalindirecttaxmanagement.com
93. How tax authorities are using AI to transform VAT compliancefintua.com
94. [2106.14005v3] A network and machine learning approach to detect Value Added Tax fraudarxiv.org
95. CSSEtechscience.com
96. How can artificial intelligence be used to help reduce the VAT gap?blog.groupseres.com
97. VAT Evasion & AI: Closing the Global Tax Gapmlogica.com
98. VAT Registration Threshold: SMEs - Hansard - UK Parliamenthansard.parliament.uk
99. The VAT threshold limit is a big issue for small business - AAT Commentaatcomment.org.uk
100. index envat-one-stop-shop.ec.europa.eu
101. VAT on overseas invoices: rules for EU and non-EU customersavalara.com
102. Cross-Border services and VAT: Start with place of supplythevatteam.co.uk
103. place of supply goods vs services vatdeloitte.com
104. Deemed supplier rules for EU and non EU companies - hellotaxhellotax.com
105. New EU ecommerce VAT and marketplace rules go into effect July 1avalara.com
106. One Stop Shop (OSS) In The EUpolishtax.com
107. VAT Reverse Charge and Place of Supply Rules Explainedkurums.com
108. EU VAT One Stop Shop (OSS) - Your Europe - European Unioneuropa.eu
109. WP/21/40 Encouraging Formal Invoicing and Reducing the VATimf.org
110. Reassessing the regressivity of the VAToecd.org
111. VAT rate changes coming in 2026: What you need to knowfintua.com
112. May 2026 global indirect tax and sales & use tax updatesavalara.com
113. UK VAT Rates 2026: Standard, Reduced & Zero Rate Explainedfincsolaccountancy.com
114. Global upcoming VAT rate changesvatupdate.com
115. Global VAT legislative Changes as of January 1, 2026vatupdate.com
116. VAT Rates by Country 2026: Global VAT & GST Comparison Tablecountrytaxcalc.com
117. European Commission releases report on tax gaps to support competitiveness and fairer tax systems - Taxation and Customs Uniontaxation-customs.ec.europa.eu
118. Mandatory Transaction-Based Reporting and the VAT Compliance Gap: Evidence from EU Member Statesdoi.org