Income Tax

An income tax is a tax levied on individuals or entities (taxpayers) on the income or profits they earn. In its most common form, the tax owed equals a rate multiplied by taxable income, which is income after permitted exclusions and deductions. Individual income is often taxed at graduated rates that rise as income rises, while companies are frequently taxed under a separate corporate tax, often at a flat rate. Income taxes are a central instrument of modern government finance, and they also shape household behavior, business investment, and the distribution of resources across a society, which makes them a major topic in economics and public policy.

The history of the income tax runs from the wartime measures of late eighteenth-century Britain and nineteenth-century United States to the broad-based systems of the twentieth century, when mass income taxation and payroll withholding made it the largest revenue source in many countries. Its design raises recurring questions: how to define income, how steeply rates should rise, how to treat savings and capital gains, how to prevent avoidance and evasion, and how to tax income that crosses borders. These questions are answered differently across countries and periods.

This entry describes the concept and its tax base, the mechanics of computing and collecting the tax, the main rate structures and taxpayer categories, the historical development from its origins through the twentieth century, the economic effects and the debates over incidence and design, and the international dimension and comparisons with other tax bases.

Definition and tax base

Income as the base

An income tax applies to a tax base defined as some measure of income received over a period, and the definition varies across jurisdictions. The most comprehensive benchmark is the Haig–Simons concept, which treats income as consumption plus the change in net worth over a period. On that definition, wages, business profits, interest, dividends, rents, realized capital gains, and in-kind benefits all count. No jurisdiction applies it in full. Most systems tax wages and salaries, business profits, and investment returns such as interest and dividends, and most tax rents. Transfers are treated unevenly: pensions, unemployment benefits, and some social insurance payments are taxable in many countries, while child benefits and many means-tested transfers are often excluded. Countries also differ on imputed rent from owner-occupied housing. Switzerland taxes it, whereas most jurisdictions do not. Treatment of asset sales differs as well. Hong Kong and Singapore generally do not tax individuals' capital gains, while many other countries do, sometimes at preferential rates.

Gross income, net income and realization

Gross income is the total of all included receipts before deductions. Net or taxable income is what remains after allowable costs are subtracted, so the definition of deductible expenses shapes the base as much as the definition of receipts does. In business taxation, for example, profit is usually revenue less the expenses of earning it. Realization governs timing. Under the realization principle, a rise in an asset's value is generally not taxed until the owner sells or otherwise disposes of the asset, even if the appreciation built up over many years. Deferral therefore has value to owners, and it can produce a lock-in effect in which holders keep assets to postpone tax. Some regimes depart from strict realization in limited cases, such as mark-to-market rules for certain financial instruments. The United States resets the tax basis of inherited assets at death, so gains accrued during the decedent's lifetime are not taxed as income. This is a further way in which the timing rule shapes who bears tax on appreciation.

Exclusions, exemptions and the narrowed base

Many items that resemble income are excluded by law. Employer-provided health insurance, some employer-funded benefits, gifts received by the recipient, qualifying scholarships, and certain social transfers are commonly left out of the base. Economists often call the revenue forgone through such special treatment a tax expenditure, meaning revenue that could in principle be collected if the base were broader. Each exclusion narrows the base, so a given amount of revenue requires a higher rate on the remaining income. Exclusions also tend to benefit people who have access to the favored benefits or who face higher marginal rates, which makes the distributional effects a central point of dispute. Estimates of these costs depend on assumptions about what taxpayers would have done under the alternative rules, so their magnitude is often contested.

Flow income versus stock wealth

An income tax is levied on a flow, measured over a period such as a year, whereas a wealth tax is levied on a stock, valued at a point in time. The distinction matters because wealth and current income can diverge. Retirees and owners of illiquid assets may hold substantial wealth with modest annual income, and some high-income earners hold little net worth. An income tax therefore reaches the returns that wealth generates, not the stock itself, while wealth, property, and inheritance taxes reach the stock directly. The Haig–Simons framework narrows this gap by counting accrued gains as income, but realization rules still determine when a change in the stock becomes taxable flow. The practical consequence is that the same economic position can be taxed quite differently depending on whether the tax is framed as a levy on income, on wealth, or on transfers at death. Broader connections to taxation and to the measurement of incomes in national accounts are discussed in other sections.

Calculation: rates, brackets, deductions and credits

From gross income to tax payable

The calculation usually starts with gross income. The taxpayer subtracts allowable deductions, and in many systems certain adjustments, to reach taxable income, which is the base that the rate schedule is applied to. In the simplest single-rate case the formula is tax equals rate multiplied by taxable income. Under a progressive schedule, each band of taxable income is multiplied by its own rate and the results are summed. Credits come last: they are subtracted from the tax computed on the schedule, which gives the tax payable or a refund. The order of steps therefore matters, because a deduction changes the base before the rates apply, while a credit changes the result after they apply.

Marginal versus average rates

The marginal rate is the rate on the last unit of taxable income, that is, the rate at which the next dollar would be taxed. The average or effective rate is total tax divided by income. In a progressive system the average is lower than the top marginal rate whenever lower bands exist. A worked example uses the 2026 United States thresholds for a single filer: 10% on the first $12,400, 12% on income from $12,401 to $50,400, and 22% on income from $50,401 to $105,700. For taxable income of $80,000, the tax is $1,240 on the first band, $4,560 on the second, and $6,512 on the $29,600 that lies in the third, for a total of $12,312. That is an effective rate of about 15.4% against a marginal rate of 22%.

The intuition is that a higher bracket reaches only the income above its threshold. If the same filer's taxable income rose from $80,000 to $85,000, the extra $5,000 would be taxed at 22%, adding $1,100, while the earlier income would be taxed as before. Bracket boundaries in the United States are adjusted each year for inflation, so the same mechanism applies to a shifting set of thresholds.

Deductions and credits

A deduction reduces the base, so its value depends on the taxpayer's marginal rate. A $1,000 deduction saves $100 for someone in a 10% bracket and $220 for someone in a 22% bracket. A credit reduces the tax itself, so a $1,000 credit saves $1,000 regardless of bracket, provided the taxpayer owes at least that much. Credits may be nonrefundable, in which case they can reduce liability only to zero, or refundable, in which case they can produce a payment to the taxpayer. Because a deduction's value rises with the marginal rate, a deduction of a given size favors higher-income taxpayers more than a credit of the same size would, which is one reason the two instruments are often compared in policy debates.

Common examples illustrate the distinction. Mortgage interest is generally a deduction, and in the United States it is usually useful only to taxpayers who itemize rather than take the standard deduction. The child and dependent care credit reduces tax for households that pay for care so that they can work; in 2026 it is worth up to 50% of up to $3,000 in qualifying expenses for one dependent, or up to $6,000 for two or more, and it is nonrefundable. Education expenses are typically supported through a mix of credits and deductions, and the rules usually bar claiming two benefits for the same expense.

Filing units

The filing unit is the entity that files a return and is assessed on its combined income. In the United States, the choices are single, head of household, married filing jointly, and married filing separately. Joint filing combines both spouses' income, deductions, and credits on one return, and both spouses are responsible for the result. Separate filing produces two returns, each reporting only that spouse's income, and the spouses generally cannot take credits or deductions that they would not qualify for individually. If one spouse itemizes, the other cannot take the standard deduction. The standard deduction for joint filers in 2025 was $31,500, against $15,750 for each separate filer, and joint brackets were roughly double the single thresholds.

The effect on tax owed depends on the distribution of income between spouses. Joint filing tends to help when one spouse earns much more than the other, because the wider brackets absorb part of the higher earner's income at lower rates. When both spouses earn similar amounts, joint filing can push combined income into higher brackets than either would face alone, which is the so-called marriage penalty. Some tax systems use the household or the family as the unit, and some permit income splitting between spouses, which produces a comparable trade-off through different arithmetic.

Rate structures

Progressive rates and brackets

A progressive or graduated rate structure is one in which the marginal rate, the rate applied to the last unit of income, rises as income rises. Most implementations divide the income scale into brackets, and each bracket carries its own rate. The bracket rate applies only to the slice of income that falls inside that band. Income above a threshold is not retaxed at the higher rate in full.

Suppose a schedule charges 10 percent on the first 10,000 units of taxable income, 20 percent on income from 10,001 to 50,000, and 30 percent above 50,000. A taxpayer with 60,000 pays 1,000 on the first slice, 8,000 on the second (20 percent of 40,000), and 3,000 on the third (30 percent of 10,000), for a total of 12,000. The 30 percent rate touches only the final 10,000. Raising income from 60,000 to 70,000 adds 3,000 in tax, which is 30 percent of the extra 10,000 and nothing else. Because the tax on each band is computed separately, a higher marginal rate never reduces the tax on lower slices of income.

Marginal and average rates

The distinction between marginal and average rates follows from bracket arithmetic. The average rate is total tax divided by taxable income. In the example above, 12,000 divided by 60,000 gives an average rate of 20 percent, while the marginal rate is 30 percent. Because the lower bands still enter the calculation, the average is always below the top marginal rate that the taxpayer faces. In a progressive system the average rate rises with income, although more slowly than the marginal rate.

Proportional and flat-rate systems

A proportional or flat-rate system applies one percentage to all taxable income. With a 15 percent flat rate, a person earning 30,000 pays 4,500 and a person earning 300,000 pays 45,000. Both have an average rate of 15 percent, although the high earner pays far more in absolute terms. A flat rate by itself is therefore proportional, but a flat rate combined with a large personal exemption produces a progressive average rate, because the exempt amount forms a zero-rate band at the bottom. Some analysts describe a flat rate as a special case of the progressive or regressive forms, and the flat-rate case serves as the reference point for defining both.

Supporters of flat taxes argue that a single rate is simple to administer and reduces incentives to shift income or to avoid higher brackets. Critics argue that a single rate shifts the burden toward lower-income households in terms of the share of resources absorbed, and that it lowers redistribution. Actual flat-rate income taxes are used in a number of countries, particularly in Central and Eastern Europe and the Baltic region, which shows that the choice between schedules is a matter of policy design rather than of technical necessity. The debate itself is covered in progressive taxation and flat-tax proposals.

Schedular and global systems

Rate structures also differ in how they treat sources of income. In a schedular system, separate taxes or rate schedules apply to different categories of income, such as wages, business profits, and investment returns. Gross income and deductible expenses are computed separately for each category, and the rates may differ from one category to the next. In a global system, all income is combined into a single base, and one schedule of rates applies to total taxable income. Global systems are generally regarded as more consistent with a taxpayer's overall ability to pay, while schedular systems can be more administratively complicated, and they create incentives to recharacterize income so that it falls under a lower-rated category.

Most income tax systems in practice are mixed. Many countries that use a global base apply separate rates to capital gains or dividends, and some countries that are formally schedular have consolidated the schedules for personal returns. The degree of schedularity is therefore a matter of degree rather than a binary classification.

Indexation and bracket creep

Bracket thresholds are fixed in nominal currency units, so inflation can move taxpayers across them without any gain in real income. This phenomenon is known as bracket creep, and it is sometimes called fiscal drag. Consider a schedule with a 20 percent rate up to 50,000 and 30 percent above that. If a worker's income rises 8 percent from 48,000 to 51,840 during a year of 8 percent inflation, the worker's purchasing power is unchanged. An indexed schedule would move the threshold to 54,000 and leave the worker in the 20 percent band, with tax of about 10,368. The unindexed schedule instead yields 10,000 on the first 50,000 plus 552 on the 1,840 above it, a total of 10,552, an increase of about 184 with no real change in circumstances.

To prevent this, the United States began indexing brackets in the early 1980s. The Economic Recovery Tax Act of 1981 provided for automatic adjustments that took effect in tax year 1985, using the Consumer Price Index. The Tax Cuts and Jobs Act of 2017 changed the index from 2018 onward to the chained CPI, which generally rises more slowly and therefore produces smaller annual threshold increases. Other countries use different indices and methods, and indexation is not uniform. Its relationship to inflation and to the broader fiscal position of governments remains a subject of policy debate.

Taxpayers and taxable entities

Income tax systems tax two broad kinds of taxpayer: individuals, whose liability is usually called personal income tax, and legal entities, mainly companies, which are subject to corporate income tax. Many other organisations sit between these two categories. Whether a business profit is taxed once at the level of the firm, once in the hands of its owners, or in some combination depends on legal form and on how the system is designed.

Personal and corporate income tax

Personal income tax is charged on individuals, and in many countries on households through joint filing. Corporate income tax is charged on a company's taxable profit, normally revenue less allowable expenses, with depreciation and interest among the main deductions. Most jurisdictions apply a single statutory rate to corporate profit, though some set reduced rates for small companies or apply different rates to particular sectors or types of income. Where subnational governments also levy corporate taxes, as in federal countries, the combined rate can be well above the national figure.

Statutory corporate rates have fallen sharply over the past quarter century, but the decline has levelled off. From 2020 to 2026, the average STR across all jurisdictions covered has remained stable at an average rate of 21.2% across the years. Figures from other compilers differ mainly because of coverage. One independent tabulation of 181 jurisdictions put the worldwide average at 23.58 percent, and the OECD's own figure for its members is higher than the global average. Rates also vary widely: France maintains the highest combined corporate tax rate in the OECD, at 36.1 percent in 2025, while the lowest marginal corporate tax rate is found in Hungary, at 9 percent, followed by Ireland (12.5 percent). Since 2024, a 15% global minimum effective rate, agreed under the OECD's Pillar Two framework, has applied to large multinational groups, which limits how far headline rates can be pushed down.

Pass-through entities

A pass-through entity does not pay income tax at the level of the business. Its profit or loss is attributed to its owners, who report it on their own returns. Most US businesses are taxed as pass-through (or flow-through) entities that, unlike C-corporations, are not subject to the corporate income tax or any other entity-level tax. Instead, their owners or members include their allocated shares of profits in taxable income under the individual income tax. The category includes sole proprietorships, general and limited partnerships, limited liability companies, and S corporations. Some forms still file an information return. A partnership, for example, files an entity-level tax return (Form 1065), but profits are allocated to owners who report their share of net income on their individual returns.

Pass-through treatment has consequences beyond the rate. A sole proprietor's net business income is generally subject to self-employment payroll taxes, and distributions from a pass-through are usually not taxed again when paid out, because the income has already been taxed once. S corporations are a special case with restrictions: an S-corporation cannot have more than 100 shareholders, so they cannot be publicly traded on a stock exchange. Some US states have also created elective taxes at the entity level, which are deductible for federal purposes and so partly work around the federal cap on state and local tax deductions.

Double taxation and integration

Where a company is taxed on its profit and its shareholders are then taxed on the dividends paid from after-tax profit, the same income bears two layers of tax. Under our current system, corporate income is taxed at two levels: the level of corporate profits and the level of shareholder dividends. The double layer also distorts financing decisions: firms may retain earnings instead of distributing them, or prefer debt, whose interest is deductible, to new equity.

Integration methods aim to reduce this distortion. One approach is an imputation or credit system, in which the corporation serves as a withholding agent, paying taxes on dividend income on behalf of the shareholder. Shareholders then receive a credit for the corporate tax already paid. Australia introduced this approach in 1987 and New Zealand followed in 1989; Canada uses only a partial version. A second approach is a dividend deduction, under which the company may deduct dividends it pays, so the tax is collected at the shareholder level. A third is a shareholder-level exclusion or reduced rate for dividends. Some countries take a different route: Chile has tax integration, and all applicable company profits are taxed only at the shareholder's tax rate, while Singapore instead taxes profits at the company level and does not tax the dividend in the shareholder's hands. The United States has not adopted a full imputation system; it has relied mainly on reduced rates for certain dividends.

Residents and nonresidents

Most systems divide taxpayers into residents and nonresidents, and the two groups are taxed on different bases. Residents are generally taxed on worldwide income, subject to relief for foreign taxes. Nonresidents are generally taxed only on income with a domestic source, such as local employment, property, or business activity, and that income is often collected by withholding at a flat rate rather than through a progressive schedule with personal allowances.

Residence is defined by statute, and the tests differ. Common criteria include a home or domicile, a day-count threshold such as 183 days in a tax year, and the location of a person's centre of economic or family ties. Companies are often resident where they are incorporated or where central management and control is exercised. When two countries claim a person or company as resident, tax treaties usually apply tie-breaker rules. A few countries, notably the United States, tax based on citizenship as well as residence, which is discussed in the international dimension section.

Collection, withholding and compliance

Withholding and pay-as-you-earn

Modern income tax collection depends on taking the tax out of income before the taxpayer receives it. In the United States, the decisive change came during the Second World War. Beardsley Ruml, then chairman of the New York Federal Reserve Bank and treasurer of R.H. Macy and Company, had championed a shift from taxing the previous year's income to taxing the current year's. The main obstacle was transitional: the taxpayers would have to pay two years' taxes in a single year. Ruml therefore proposed to simply cancel 1942 tax liabilities. Congress rejected the most generous version, and the compromise that President Roosevelt signed in June 1943 provided for a 20 per cent withholding tax to start July 1 and forgave from 75 to 100 per cent of a year's tax liability. The arrangement was codified in the Current Tax Payment Act of 1943, which required all taxpayers to stay current on their tax liability through a system of paycheck withholding. Withheld receipts rose from almost $8 billion in 1944 to more than $10 billion in 1945.

Britain followed a similar path. The scheme was devised by Sir Paul Chambers and introduced into the UK in 1944, following trials in 1940–1941. Before it, tax was typically settled in lump sums, when tax was collected annually or twice yearly. Sources differ on the launch date: HMRC states 6 April 1944, while a chronology of 1944 lists 10 February. Wartime expansion of the workforce made the change urgent, and by January 1944 15 million people, those earning more than £100 a year, had received notices telling them their tax code number. Today the system covers wages and salaries, as well as sick pay, maternity pay, directors' fees and pensions (but not the state pension, although it remains taxable). Because PAYE is deducted from each paycheck by the employer and must be remitted promptly to the government. The pay-as-you-earn label is one of many; most countries refer to income tax withholding by other terms, including pay-as-you-go tax.

Annual returns and payments

Withholding is an estimate rather than a final settlement. Amounts withheld are treated as advance payments of income tax due. They are refundable to the extent they exceed tax as determined on tax returns. The annual return therefore reconciles the year's deductions and other payments against the liability actually computed. A surplus produces a refund, while a shortfall leaves a balance due.

Withholding covers only part of the population. Self-employed people and those whose income comes mainly from investments, rents or interest usually have no employer deducting tax at source. In systems such as the United States they make estimated payments during the year, typically in quarterly instalments, and an underpayment can attract a penalty even when the return is filed on time.

Audits, penalties and information reporting

Compliance relies heavily on third-party reporting. Employers report wages to tax authorities and to employees, and banks, brokers and other payers report interest, dividends and similar receipts. Authorities can therefore match the figures on a return against independent records, and discrepancies are a common basis for selecting returns for review. An audit examines the taxpayer's records and can result in additional tax, interest and penalties.

Penalties are graded by culpability. An honest mistake, such as miscalculating a deduction or forgetting to report a small amount of interest, is treated as negligence, not fraud. The IRS addresses negligence with civil penalties (typically a 20% accuracy-related penalty), not criminal charges. Deliberate misstatement is treated far more severely, and criminal prosecution in the United States requires proof of wilfulness.

Tax evasion and tax avoidance

The legal line separates the two concepts. Avoidance is the lawful use of tax law to reduce a tax bill, while evasion is the illegal act of not paying taxes owed through deceit or concealment. Avoidance typically relies on taking advantage of tax credits, deductions, or retirement account contributions. Evasion includes practices such as creating a complex scheme to hide income offshore, which indicates intent to evade.

Enforcement against evasion requires proof of intent. In U.S. federal cases, the government must prove you owed funds, knew about the obligation and willfully attempted to evade payment. Penalties include repayment of the original liability with interest. Reported maximums vary by source: one describes a fine of up to $100,000 for individuals, while another cites fines of up to $250,000, and imprisonment of up to five years is commonly reported.

Avoidance is not automatically safe. Tax avoidance doesn't attract penalties, though authorities may challenge some avoidance schemes. Many countries therefore maintain general anti-avoidance rules that allow authorities to disregard arrangements whose main purpose is tax reduction, and the boundary between aggressive planning and evasion remains a matter of contested interpretation in individual cases.

Historical origins and early development

Before the modern era, the states of Mesopotamia, Egypt, Greece and the Roman Empire financed themselves mainly through tribute, tithes on harvests, customs duties, poll levies, land and property charges, and compulsory labour or service obligations. These were generally not income taxes in the modern sense. They were assessed on people, land, goods or trade rather than on a periodic measure of net income from all sources, and they lacked the record-keeping, employer withholding and annual filing that a comprehensive income tax later required.

Britain: from Pitt to Peel

The modern income tax is usually traced to Britain. Income tax was first implemented in Great Britain by William Pitt the Younger in his budget of December 1798, as preparation for the Napoleonic Wars. Accounts differ on its structure. One popular account describes a flat 10 percent rate on incomes above £60 (applied to incomes above £60 per year), while the more detailed histories describe a graduated scale, starting at about 1 percent and reaching 10 percent on incomes above £200 (increased up to a maximum of 2 shillings (10 per cent) on annual incomes of over £200). Pitt had hoped to raise £10 million a year, but receipts for 1799 totalled just over £6 million.

The tax lapsed during the short peace of 1802, and the income tax was reintroduced by Addington in 1803 when hostilities recommenced. Addington's version, presented as a "Property Tax", was built on schedules: taxpayers had to report income from particular sources rather than a single total, a structure that later became the basis of British schedular taxation (he required returns of income from particular sources (under Schedules A to E) instead of the total demanded in the 1799 Act). The tax was finally abolished in 1816, a year after the end of the Napoleonic War, by popular demand. Opponents had argued that such a tax should be used only to finance wars, and the official records were publicly burned (Records were publicly burned by the Chancellor of the Exchequer, but copies were retained).

The tax returned in 1842. Sir Robert Peel reintroduced income tax in 1842 at 3% (sevenpence in the pound) to plug a budget deficit. Peel presented it as a temporary, three-year measure that would allow free-trade reforms to be financed (He assured Parliament it was a three-year temporary measure to allow free trade reforms to raise revenue in the longer term.). It was never repealed after that, which is why 1842 is treated as the start of the permanent British tax.

The United States: Civil War experiments

The first federal income tax in the United States came with the Civil War. The Revenue Act of 1861 imposed a flat tax with a rate of 3% on incomes above $800. Its enforcement machinery was weak, and it was replaced within months. The Revenue Act of 1862 introduced a graduated structure: incomes below $600 were exempt, a 3 percent rate applied to incomes from $600 up to $10,000, and a percentage of 5% of total income was demanded in tax for those above $10,000. The Internal Revenue Act of 1864 then raised the schedule to 5% on income starting at $600; 7.5% on income at $5,000; and 10% on income over $10,000. The tax was repealed in 1872, but it had provided nearly a quarter of the war revenue.

The 1894 tax and the path to the Sixteenth Amendment

Renewed pressure for an income tax followed the war. Congress introduced more than sixty bills between 1871 and 1894 to restore the income tax, and the measure finally passed as part of the Wilson–Gorman Tariff Act of 1894, which lowered tariffs and made up for some of the lost revenue by introducing taxes on income, corporate profits, gifts, and inheritances. The income tax applied a 2 percent rate to incomes above $4,000.

In Pollock v. Farmers' Loan & Trust Co. (1895), the Supreme Court ruled 5–4 that the tax on income from property was a direct tax that had to be apportioned among the states by population, which was impossible in practice for a progressive levy. The income tax, on the other hand, taxed wages and earnings derived from property. The Court's reasoning was that taxes on land and personal property were direct, so the rents, interest and dividends taxed under the act were also direct. The decision was highly unpopular, but it stood until a constitutional amendment overrode it. Congress proposed the Sixteenth Amendment in 1909, and it was ratified in 1913, giving Congress power to tax incomes from any source without apportionment among the states. That amendment is the constitutional basis for the modern federal income tax.

Expansion in the twentieth century

The First World War turned the income tax from an elite levy into a tool of wartime finance. In Britain, the rate rose from 17.5 percent in 1915 to 25 percent in 1916 and 30 percent in 1918, while individual exemptions were lowered. The government also introduced an excess profits tax of 50 percent on profits above prewar norms, which was raised to 80 percent in 1917. Even so, taxes covered at most about 30 percent of national spending, and borrowing financed the rest. In the United States, the War Revenue Act of 1917 cut the exemption levels sharply and raised the top bracket from 15 to 67 percent on income above $2 million. The Revenue Act of 1918 consolidated the system into a normal tax and a surtax, lifting the top rate to 77 percent on income above $1 million. American excess profits taxation worked differently. The 1917 law applied graduated rates of 20 to 60 percent to profits above prewar earnings, and the 1918 legislation added an alternative tax of up to 80 percent for that year. Congress repealed the excess profits tax in 1921. Even at its wartime peak, the federal income tax reached only a small minority: about 5 percent of Americans paid it in 1918, up from roughly 1 percent in 1913, yet it still financed about a third of the war's cost.

The Great Depression reversed the postwar retreat. Federal revenues collapsed after 1929, and Herbert Hoover's Revenue Act of 1932 raised the top marginal rate from 25 to 63 percent, roughly doubled rates across most income levels, quadrupled the lowest rate from about 1 to 4 percent, and increased corporate and estate taxes. The act also introduced a range of new excise levies on goods such as gasoline and consumer products. Historians disagree about its macroeconomic effects. Some argue that it deepened the contraction by discouraging spending, while others note that falling revenues make the "tax increase" label ambiguous. Franklin D. Roosevelt's New Deal kept and extended the framework. Between 1934 and 1936 Congress raised taxes on capital gains and dividends and increased marginal rates on net incomes above $50,000. Congress also passed a series of Revenue Acts between 1932 and 1939 and two mild excess profits taxes in 1933 and 1935. Despite the higher rates, the broad base remained narrow, and most American families still paid no federal income tax. Some historians link the later tax policies to the 1937–38 recession, though this view is contested.

World War II produced the shift from a class tax to a mass tax. Congress passed four excess profits statutes between 1940 and 1943, and the Revenue Act of 1942 added the Victory Tax, a temporary levy designed to reach a much larger share of wage earners. The decisive step was the Current Tax Payment Act of 1943, signed by Roosevelt on June 9. Before it, taxpayers filed a return each spring and paid the previous year's liability. The act required employers to withhold tax from wages and remit it directly to the Treasury, which turned the income tax into a pay-as-you-go system. To avoid double payments during the transition, the law cancelled 75 percent of one year's liability, the lower of the 1942 or 1943 amount. Withholding had appeared briefly in the Tariff Act of 1913 and had been dropped in 1916, so the 1943 law restored a practice that had not been firmly established. Treasury officials argued that without withholding the income tax could not function as a mass tax. The change was dramatic: income taxes had averaged about 1 percent of personal income in 1939, and they rose above 11 percent after the new system took effect.

In the decades after 1945, personal income tax became the largest or one of the largest revenue sources in many advanced economies. Its persistence has several explanations. Withholding made collection routine and largely invisible to taxpayers, and revenue rose automatically as nominal incomes grew, which allowed governments to finance expanding social programmes and defence without frequent visible rate increases. The mass base also made the tax politically hard to repeal, because a large share of voters now paid it. Reforms since then have mostly reduced top rates and altered brackets and deductions, but they have rarely removed the wage-withholding framework that the wartime changes established. The degree to which these features explain cross-country differences remains a matter of scholarly debate.

Economic effects and incidence

Who ultimately bears an income tax

The person the law names as the payer is not always the person who loses economically. Economists separate statutory incidence from economic incidence, the latter measured as the change in real income or wealth a tax produces (It distinguishes between statutory incidence (the legal requirement to remit a tax) and economic incidence (the change in real income or wealth resulting from a tax).). The gap between the two is what the literature calls tax shifting. A simple case shows the mechanism: if a 10% tax is imposed on sellers of butter, but the market price rises 8% as a result, then 80% of the tax incidence falls on buyers, not sellers.

Income taxes raise the same question. An employer may withhold tax from wages and remit it, but the burden can be spread between workers and employers. Under standard supply-and-demand reasoning, the side of the market that is less able to change its behaviour bears the larger share. The side of the market that is more inelastic bears the larger share of the tax, regardless of who legally pays it. If labour supply is relatively inelastic, workers may absorb more of the cost through slower wage growth. If labour demand is less responsive, employers may absorb more. Empirically, the answer depends on the elasticities of the specific labour market, which is why statutory assignments tell us little on their own.

Effects on labour supply

Taxes on earnings work through two channels. A lower after-tax wage makes leisure relatively cheaper, which is the substitution effect. A lower after-tax income means workers can afford less leisure, which pushes them to work more; this is the income effect. The substitution effect measures the decline in effort when the return to work is lowered. The income effect measures the increase in effort when income declines. The two effects pull in opposite directions, so the net outcome is an empirical question.

Despite disagreement about size, the literature has reached a broad consensus on direction. Reviewers note that while authors may differ on the magnitude of labour-supply elasticities, they largely agree on the sign. The consensus is that increasing tax rates usually reduces work effort. Exceptions arise when income effects dominate, especially at high hours of work. Timing matters too: a permanent tax tends to trigger larger income effects, while a temporary one mainly works through substitution. The more permanent is a tax, the more workers will feel the income effect while if a tax is transitory, the substitution effect is likely more important.

The elasticity debate concerns who responds most. A U.S. review found that for men and single women, estimates of substitution elasticities have increased, and income elasticities still appear to be smaller in absolute value than substitution elasticities. Estimates for married women have moved in the other direction. The question of high earners is contested. One review found little compelling evidence that high-income taxpayers have substantially higher elasticities with respect to their labor input than lower-income taxpayers. Higher estimates of the elasticity of broad income among high-income taxpayers appear to reflect their greater ability to time their income rather than greater changes in their labor supply. By contrast, a study of a Norwegian reform found that high-income individuals are more responsive to tax changes than middle-income individuals.

Saving, investment and deadweight loss

Income taxes also reduce the return to saving and investment, because interest, dividends and capital gains are taxed alongside wages. This can shift the timing of consumption and the form in which wealth is held. The net effect on total saving is ambiguous: higher after-tax returns can increase saving for some households while discouraging it for others.

Distortion is a central cost. Because taxes change relative prices, people alter choices such as hours, timing, compensation form or reporting, and these changes can generate deadweight loss beyond the revenue raised. Using a standard approximation, Feldstein's framework gives a loss that rises with the square of the tax rate and is scaled by the elasticity of taxable income. Goolsbee's presentation of this result is written as $$DWL \approx \frac{1}{2}\,\frac{\tau^{2}\,\varepsilon\,TI}{1-\tau}$$ where $\tau$ is the marginal rate, $\varepsilon$ the taxable-income elasticity and $TI$ taxable income. The Norwegian study cited above concluded that these findings imply a substantial excess burden of taxation, and that reducing top-income tax rates would increase tax revenue. Such conclusions depend heavily on the elasticity estimate used.

Empirical evidence and methodological problems

Much of the evidence comes from natural experiments, especially the U.S. Tax Reform Act of 1986. Martin Feldstein used a Treasury panel of more than 4,000 taxpayers to estimate the sensitivity of taxable income to changes in tax rates based on a comparison of the tax returns of the same individual taxpayers before and after the 1986 tax reform. His early findings were striking, with an elasticity at least one and that could be substantially higher.

Later work challenged the magnitude. Gruber and Saez reported that the overall elasticity of taxable income is 0.4, well below the original estimates of Feldstein but roughly at the mid-point of the subsequent literature. Several methodological problems have been identified. One is that the income distribution is not static, so a fixed real income cutoff can produce apparent responses unrelated to tax changes; Feldstein's panel design was partly meant to address this (the use of a constant real income cutoff will naturally lead to a finding that tax cuts for the wealthy are leading to higher taxable incomes in that group.). Another is control-group bias: Saez and coauthors show that the estimated elasticity is twice as large as the true elasticity in the treatment group. Economists also note that the taxable-income concept captures more than labour supply. Feldstein himself emphasizes that the response of taxable income involves much more than a change in the traditional measures of labor supply.

These difficulties explain why the field relies on difference-in-differences designs, reform comparisons and panel data, and why estimates from the same reform can differ substantially. Careful interpretation requires attention to control groups, the timing of reported income, and the distinction between compensated and uncompensated responses.

International dimension

Residence and source as competing claims

A cross-border income stream can plausibly be taxed in two places: the country where the income is generated (the source state) and the country where the recipient lives (the residence state). Source-based taxation gives the state where a factory, mine, or bank branch sits the first claim on profits and payments made from its territory. Residence-based taxation lets the state of the taxpayer's home reach that same income, usually on a worldwide basis. Most countries use both principles at once. Residents are generally taxed on their worldwide income, while non-residents are taxed only on income tied to the country, a distinction often described as unlimited versus limited tax liability.

The two principles pull in different directions for capital. Capital-exporting countries tend to prefer residence taxation, because their investors' foreign returns would otherwise be taxed at source. Capital-importing countries, including many developing economies, tend to favour source taxation, because the income is earned on their soil and their tax bases would otherwise be thin. The OECD Model Tax Convention generally tilts toward residence, for example by giving the residence state exclusive taxing rights over many royalties and most gains from shares, and by suggesting low withholding rates on dividends and interest. The UN Model Convention leaves more taxing rights with the source state, particularly for royalties, equipment rentals, and business activity that is short of a full permanent establishment.

Relieving double taxation

When two states tax the same income, the usual remedies are bilateral tax treaties and unilateral domestic rules. Treaties allocate taxing rights category by category and set ceilings on source-state withholding. Since the League of Nations began developing model conventions in the 1920s, and after the OECD and UN continued that work, the global treaty network has grown into several thousand bilateral agreements. Treaties also include mutual agreement procedures, which allow tax authorities to resolve disputes over how a case should be treated.

Where a treaty does not apply, or the residence state retains some taxing power, the residence state usually relieves double taxation in one of two ways. Under the credit method it taxes the foreign income but deducts the foreign tax already paid, up to the domestic tax on that income. Under the exemption method it excludes the foreign-source income from its base altogether. Many countries combine the two, and the choice shapes the incentives facing investors: an exemption system can make a foreign tax rate less important to the home-country taxpayer, while a credit system can leave a taxpayer owing the home state the difference whenever the foreign rate is lower.

Base erosion, profit shifting, and transfer pricing

Multinational groups can move reported profits to low-tax jurisdictions by exploiting gaps and mismatches between national rules, a practice the OECD calls base erosion and profit shifting (BEPS). Typical techniques include charging high intra-group interest or royalties, locating intangible property in favourable regimes, and routing income through entities with little real activity. The OECD has argued that such practices cost governments roughly $100–240 billion a year in lost revenue, though estimates of this kind depend heavily on method.

The main domestic tool against profit shifting is transfer pricing law. Most countries apply the arm's-length principle, which requires related companies to price their transactions as unrelated parties would, and they require documentation to support those prices. Enforcement is contentious because the arm's-length benchmark is often hard to apply to unique intangibles and integrated business models. In response, the OECD and G20 launched the BEPS project in 2013, which produced 15 action items published in 2015. In 2016 the Inclusive Framework on BEPS was set up to monitor implementation, and it has since grown to more than 145 member jurisdictions. Its later "Pillar Two" design introduces a global minimum effective tax rate of 15 percent for large multinational groups, implemented through domestic rules such as the income inclusion rule and the undertaxed profits rule, though not all members have adopted these measures in the same way.

Citizenship-based and residence-based systems

Most countries tax individuals according to where they live, so that a person who moves abroad and ceases to be a resident usually leaves the domestic tax net, subject to rules on departure and source income. The United States is a major exception. It taxes citizens and green-card holders on worldwide income regardless of where they reside, so a US citizen living in another country must generally file a US return every year. The US relies on the foreign earned income exclusion and the foreign tax credit under Internal Revenue Code section 901 to prevent double taxation, but these reduce tax owed rather than removing the obligation to file. Treaty "saving clauses" typically preserve the US right to tax its citizens as though the treaty did not exist. Eritrea is usually cited as the other country that taxes its citizens on this basis, and the practice has drawn criticism from expatriates and has been defended on grounds of revenue and tax-compliance enforcement. For multinationals and wealthy individuals, the difference between citizenship and residence rules can be significant, which is why both principles remain central to treaty negotiation.

Policy debates and alternative tax bases

Degree of progressivity

Debate over how steeply rates should rise with income rests on three arguments that often appear together but rely on different premises. The equity argument, grounded in ability to pay, holds that a given amount of tax imposes less sacrifice on a high-income household than on a low-income one, so progressive rates are fairer. Critics respond that the fairness standard is a value judgment and that progressivity can be justified without any particular top rate. The work-incentive argument draws on James Mirrlees's 1971 model of optimal income taxation, in which high marginal rates reduce effort and the best top rate depends on how sensitive taxable income is to rates. Empirical estimates of that sensitivity vary widely, so conclusions differ. The revenue argument asks whether high rates raise enough money once behavioral responses are counted. The United States shows the range of positions: the top marginal rate reached 94% in 1944 and about 91% in much of the 1950s, was cut to 50% in 1981, and fell to 28% under the Tax Reform Act of 1986. It has since stayed roughly between 35% and 40%.

Flat taxes and consumption-based alternatives

A flat tax applies one marginal rate to income above a personal allowance, so that $T = t(Y - A)$, where $t$ is the rate, $Y$ is income, and $A$ is the allowance. Average liability rises toward $t$ as income grows, which makes the system progressive at low incomes and close to proportional above the allowance. Proponents argue that a single rate simplifies compliance, lowers administrative costs, and reduces distortions in choices between work and leisure and between forms of income. Estonia adopted a single rate in 1994, and Robert Hall and Alvin Rabushka's 1995 proposal popularized the model in the United States. Critics argue that a single rate gives up redistribution at the top, that revenue losses force either higher rates or a narrower base, and that the exemption supplies whatever progressivity the system has.

Consumption-based alternatives tax spending rather than earnings. An expenditure tax, analyzed in the twentieth century by economists including Irving Fisher and in the 1978 Meade Report prepared for the UK's Institute for Fiscal Studies, exempts saving and taxes withdrawals or consumption, removing the tax on the return to saving that an income tax imposes. Hall and Rabushka's design also works as a cash-flow tax on businesses, which deduct investment immediately. Supporters argue that this neutrality encourages saving and investment. Opponents point to transition costs for existing asset holders, the difficulty of achieving progressivity through a consumption base, and the record-keeping needed to track saving.

Capital gains, dividends and retirement saving

Capital income is among the most contested areas of design. Gains are usually taxed only on realization, which permits deferral and, in the United States, a step-up in basis at death that removes tax on gains accrued during life. Many countries tax long-term gains at lower rates than wages. Supporters argue that such rates offset inflation-driven gains, relieve double taxation of corporate profit, and encourage risk-taking. Critics argue that they mainly benefit high-income households, that the realization requirement causes lock-in that reduces asset mobility, and that inflation indexing would be a more precise remedy. Dividends are taxed at the corporate level first. The 2003 US reduction of the dividend rate to match long-term gains rates was defended as a partial integration measure.

Retirement saving is usually taxed in one of two main patterns. Under an EET model, contributions and returns are exempt and withdrawals are taxed. Under a TEE model, as with Roth-style accounts, contributions are taxed and later returns are exempt. The choice shifts tax timing and affects distribution, because tax-favored accounts are used disproportionately by higher-income savers. Proposals range from contribution caps to moving toward Roth-style treatment.

Comparison with other tax bases

The value-added tax, first introduced in France in 1954, is levied on value added at each stage of production and is used in well over 150 countries. As a tax on consumption it is generally regarded as regressive relative to income, since lower-income households spend a larger share of their earnings. Reduced rates on necessities and rebates can offset this. Payroll and social insurance contributions fund pensions, health care and unemployment benefits in many countries. Where they apply only up to an earnings ceiling, as with the US Social Security portion, their burden falls more heavily on lower and middle earners. Economists debate how much of employer contributions workers ultimately bear, which is a question of tax-incidence.

Wealth taxes, levied on net worth or selected assets, exist in few countries. Several European states, including Sweden in 2007, abolished them, citing valuation difficulties, administrative costs and capital mobility. Property taxes, levied on land and buildings, rest on an immobile base and are a major source of local revenue. They can become regressive when assessments lag market values. California's Proposition 13 (1978) limited annual assessment increases to 2% unless a property changed hands. Each base reaches different groups. Income taxes fall on those with taxable labor and capital income. Consumption taxes fall on spenders in proportion to what they buy. Payroll taxes fall on wage earners below the ceiling. Wealth and property taxes fall on asset holders, and property taxes weigh most on homeowners.

References

1. Income Tax Brackets: How Marginal Tax Rates Workbritannica.com
2. 2026 Tax Brackets and Marginal Federal Income Tax Rates: What to Know for Your Next Returnkiplinger.com
3. Tax Brackets Explained: How Marginal Tax Rates Work in 2026taxrate.org
4. Marginal tax rate: What it is and how to calculate itfidelity.com
5. Tax Credit vs. Tax Deduction: What’s the Difference?experian.com
6. What's the Difference Between a Tax Credit and a Tax Deduction?money.usnews.com
7. What is the difference between a tax deduction and a tax credit?jacksonhewitt.com
8. Married Filing Jointly vs Separately: Which Should I Choose?blog.turbotax.intuit.com
9. Married Filing Separately or Jointly: Which Is Better in 2026?cnbc.com
10. Married Filing Jointly vs Separately: Which to Choose?mutualofomaha.com
11. Married Filing Jointly vs. Separately: What's Right For You?taxslayer.com
12. Inflation Indexingtaxfoundation.org
13. Federal Individual Income Tax Brackets, Standard Deductions, and Personal Exemption: 1988 to 2026congress.gov
14. What is Bracket Creep? - TurboTax Tax Tips & Videos - Intuitturbotax.intuit.com
15. www.bls.govbls.gov
16. 14 Individual Income Tax in: Tax Law Design and Drafting, Volume 2elibrary.imf.org
17. Income Tax - an overviewsciencedirect.com
18. Proportional taxen.wikipedia.org
19. Progressive vs Regressive Taxes: Examples Explainedeconlearn.org
20. What Is Progressive Tax?raisin.com
21. inflation and personal tax bracket creep a bigger picturedeloitte.com
22. Statutory corporate income tax rates: Corporate Tax Statistics 2026oecd.org
23. 2025 Corporate Tax Rates by Countrytaxfoundation.org
24. Chart: Global Corporation Tax Levels In Perspectivestatista.com
25. What are pass-through businesses?taxpolicycenter.org
26. How Do States Tax the Profits of Pass-through Entities?itep.org
27. FAQs: Corporate Integration Tax Reform and Philanthropycof.org
28. The Taxation of Dividends: Background and Overviewcongress.gov
29. Dividend imputationen.wikipedia.org
30. New OECD data highlight stabilisation in statutory corporate tax rates worldwideoecd.org
31. Corporations and Partnershipsnationalparalegal.edu
32. Australian dividend imputation systemen.wikipedia.org
33. Imputation tax system: Meaning, Criticisms & Real-World Usesdiversification.com
34. Corporate double taxation: Meaning, Criticisms & Real-World Usesdiversification.com
35. The Plan That Slogans Built: The Revenue Act of 1943taxnotes.com
36. Wartime Origins of Modern Income-Tax Withholdingfee.org
37. Tax History: Compromising on the Current Tax Payment Act of 1943taxnotes.com
38. FDR signs pay as go tax - UPI Archivesupi.com
39. Tax History Museumtaxnotes.com
40. Pay-as-you-earn taxen.wikipedia.org
41. The cost of war - UK Parliamentparliament.uk
42. Walking the halls of history with History Alice: PAYE at 80mynewsdesk.com
43. 1944 in the United Kingdomen.wikipedia.org
44. Tax Fraud vs. Tax Evasion for U.S. Taxpayers Explainedgreenbacktaxservices.com
45. Tax Evasion vs. Tax Avoidance: What's the Difference? - FindLawfindlaw.com
46. The Difference Between Tax Evasion and Tax Avoidance in Criminal Casesthelinderfirm.com
47. Tax evasion vs. tax avoidance: Key differences under federal lawstipp-law.com
48. Understanding the Fine Line Between Tax Evasion and Tax Avoidance for Expatsgreenbacktaxservices.com
49. Tax History: The Business of Withholding in 1943: Ambivalence and Internal Divisiontaxnotes.com
50. Tax History: Beardsley Ruml: The Man Who Invented Withholding — Sort Oftaxnotes.com
51. History of taxation in the United Kingdomen.wikipedia.org
52. UK Income Tax History: How Rates Have Changed Since 1799calculatetax.uk
53. Taxation in the United Kingdomen.wikipedia.org
54. William Pitt and his Taxes By John Jeffrey-Cooktaxadvisers.org.uk
55. Key dates - UK Parliamentparliament.uk
56. Income taxen.wikipedia.org
57. Revenue Act of 1861en.wikipedia.org
58. Revenue Act of 1862en.wikipedia.org
59. REVENUE ACTS OF 1861 & 1862 (1st Federal Income Tax)statutesandstories.com
60. Understanding Taxes - Theme 2: Taxes in U.S. History - Lesson 3: Income Tax Issuesapps.irs.gov
61. 1894 Income Tax and the Wilson-Gorman Tariff Actencyclopedia.com
62. Pollock v. Farmers' Loan & Trust Co.en.wikipedia.org
63. Worksheet Solutions: A Brief History of the Income Taxapps.irs.gov
64. Info Sheet 1 Timeline of Events: 1861–1913 Theme 2: Taxes in U.S. Historyapps.irs.gov
65. Sixteenth Amendment to the United States Constitutionen.wikipedia.org
66. Today in History: Income Tax Ruled Unconstitutional in Pollock v. Farmers Loan Trust Co.taxfoundation.org
67. Pollock v. Farmers' Loan & Trust Cosimple.wikipedia.org
68. Excess profits taxen.wikipedia.org
69. The History of Excess Profits Taxes Not as Effective or Harmless as Today’s Advocates Portraytaxfoundation.org
70. War Revenue Act of 1917en.wikipedia.org
71. Revenue Act of 1918en.wikipedia.org
72. Revenue Act of 1921en.wikipedia.org
73. Revenue Actsebsco.com
74. Revenue Act of 1932en.wikipedia.org
75. The Economic Impact of Tax Changes, 1920–1939cato.org
76. Revenue Actbritannica.com
77. The Republican Roots of New Deal Tax Policytaxnotes.com
78. Income and Wealth Taxes (1934-1941) - Living New Deallivingnewdeal.org
79. Current Tax Payment Act of 1943en.wikipedia.org
80. U.S. Tax History Timeline: Class to Mass Tax During World War IItaxnotes.com
81. Tax Incidencelink.springer.com
82. Tax incidenceen.wikipedia.org
83. Tax Incidence: Who Really Pays the Tax?econlearn.org
84. A Review of Recent Research on Labor Supply Elasticitiescbo.gov
85. Income tax and labour supply: Let’s acknowledge what we don’t knowcepr.org
86. Labor Supply — Documentation Micro Notespcmichaud.github.io
87. Substitution and Income Effects of Labor Income Taxationnber.org
88. The Effect of Marginal Tax Rates on Taxable Income: A Panel Study of the 1986 Tax Reform Actideas.repec.org
89. Journal of Public Economics 84 (2002) 1–32 www.elsevier.com/locate/econbasepiketty.pse.ens.fr
90. The Elasticity of Taxable Income with Respect to Marginal ...eml.berkeley.edu
91. Tax incidencebritannica.com
92. The effect of marginal tax rates on income: a panel study of ...eml.berkeley.edu
93. Evidence on the High-Income Laffer Curve from Six Decades of Tax Reformbrookings.edu
94. DO TAXPAYERS REALLY RESPOND TO CHANGES IN TAX RATES?home.treasury.gov
95. The effect of marginal tax rates on income: a panel study of ‘bracket creep’ - ScienceDirectsciencedirect.com
96. Base erosion and profit shifting (BEPS)oecd.org
97. OECD/G20 Base Erosion and Profit Shifting Project A Decade of the BEPSoecd.org
98. OECD/G20 Inclusive Framework on BEPS shows progress in making dispute resolution more effective and in improving tax transparency through Country-by-Country reportingoecd.org
99. Base erosion and profit shifting (OECD project)en.wikipedia.org
100. BEPS 2.0: Pillar One and Pillar Two - KPMG Internationalkpmg.com
101. A guide to OECDelementalcosec.com
102. beps actionsdeloitte.com
103. Tax Justice Network: Source and residence: the UN and the OECDtaxjustice.blogspot.com
104. TAXATIONunctad.org
105. Source and residence taxation - SEP-2005taxjustice.net
106. Chapter 8 Are Tax Treaties Worth It for Developing Economies? in: Corporate Income Taxes under Pressureelibrary.imf.org
107. Model Tax Convention on Income and on Capital FULL VERSIONoecd.org
108. Tax residenceen.wikipedia.org
109. 1 An introduction to tax treaties BRIAN J. ARNOLD* I. Introduction 1.financing.desa.un.org
110. Clash of the Titans: OECD vs. UN Model Treaty - Archipel Tax Advicearchipeltaxadvice.nl
111. UN Model vs OECD Model Treaty: Key Differences in Taxation for Developing Countries - Academy of Tax Lawacademyoftaxlaw.com
112. US expat citizenship-based taxation: 2026 filing guidetaxesforexpats.com
113. Citizenship-Based Taxation: Why the US Taxes Expatstscpatax.com
114. Citizenship-Based Taxation vs. Residency-Based Taxation: What's the Difference?greenbacktaxservices.com
115. Foreign tax credit and Form 1116 explainedtax.thomsonreuters.com