Chapter 2, the Special Feature of this report, examines the proportion of personal income tax (PIT) revenue that comes from different sources of individual income in OECD countries. Based on information provided for the first time by 29 countries for this edition of Revenue Statistics and a methodology developed by the European Commission, the Special Feature analyses the relative importance of PIT revenue levied on income from employed labour, capital, self-employment, and pensions and social transfers, and tracks how their respective shares have changed over time.
Revenue Statistics 2025
Disentangling Personal Income Tax Revenue in OECD Countries
Abstract
Personal income tax (PIT) accounts for almost a quarter of total tax revenue on average across OECD countries. It is also one of the main mechanisms for fiscal redistribution and influences economic growth through a variety of channels. However, analysis of PIT systems within and across countries is constrained by a lack of information about the proportion of PIT revenue that is derived from different sources of individual income, notably labour, capital, self-employment or pensions and social transfers.
For the first time in Revenue Statistics, this Special Feature provides a disaggregation of PIT revenue by income source for a majority of OECD countries and shows how these shares have evolved over time. Based on information provided by 29 countries for this edition of OECD Revenue Statistics and a methodology developed by the European Commission1, the information contained in the Special Feature compares the composition of PIT revenue in OECD countries and can serve as a starting point for more in-depth analysis.
The Special Feature is structured as follows. The first section explains the benefits of knowing more about the specific sources of individual income on which PIT is levied, as well as how this information might enrich international revenue classifications. The second section outlines how countries have estimated the breakdown of PIT revenue by source of income for this Special Feature while the third section provides a high-level analysis of the results of this exercise. A fourth section concludes.
As shown in Chapter 1 of this Report, PIT is an ever-more important source of revenue in OECD countries. In 2023, PIT accounted for 23.7% of total tax revenue on average across OECD countries, an increase of 1.6 percentage points (p.p.) since 2011; over the same period, PIT revenue as a share of GDP rose from 7.3% to 8.2% (Figure 2.1). PIT revenue ranged from 1.4% of GDP in Costa Rica to 25.2% of GDP in Denmark in 2023; between 2011 and 2023, it increased as a share of GDP in 31 out of 38 OECD countries.
In addition to its contribution to revenue raising, PIT is one of the principal direct channels for fiscal redistribution in OECD countries (Gerber et al., 2018[1]) (Jalles and Karras, 2025[2]). It may also affect economic growth through a number of channels, including its potential impact on labour market participation, entrepreneurship, investment in human capital or saving behaviour (OECD, 2017[3]), (Akgun, Cournède and Fournier, 2017[4]), (Zidar, 2019[5]), (Piketty, Saez and Stantcheva, 2014[6]).
The wide range of income sources that constitute the PIT base can make it challenging to analyse PIT revenue trends as well as the economic and redistributive impact of PIT systems. So too can the design of different taxes within PIT systems. This section explains how more-detailed information on the different income sources on which PIT is levied can enrich tax policy and analysis, and it sets out the potential benefits for Revenue Statistics and other international revenue classifications.
The broad base of PIT
Individual income is a broad tax base, encompassing a wide range of sources that includes labour income, income from self-employment, capital income2 (such as dividends, interest, rental income and capital gains), and transfers (including pensions and social benefits). Historically, the income an individual (or household) receives from different sources would be pooled for tax purposes under a ‘comprehensive system’, whereby the tax liability is calculated according to a single schedule. However, in recent decades, there has been a shift across many OECD countries from comprehensive to dual income tax systems, whereby income from labour and capital are taxed separately (Genser, 2006[7]).
Percentage of GDP (left-hand side) and percentage of total taxation (right-hand side)
Source: (OECD, 2025[8]).
Disentangling PIT revenue from these different sources of income is therefore crucial for understanding how a PIT system interacts with the broader economy, including the revenue impact of long-term structural changes or short-term shocks. Among the former category are questions as diverse and important as the future of work, demographic change and labour’s declining share of income across OECD countries; among the latter, the impact of inflation or the COVID-19 pandemic. This information can in turn inform the design of policies to make PIT systems more adaptive to changes in the economy or more robust to shocks.
Two examples of macroeconomic analysis that can be carried out with data on PIT revenue generated from different sources of individual income are the calculation of implicit tax rates (ITRs) and revenue elasticity. ITRs show the overall tax rate on different components of national income while revenue elasticity shows how revenue from a specific tax or taxes responds to changes in the economic base, for which GDP is often used.
To estimate the ITR on labour, for instance, the proportion of PIT revenue from employment labour is combined with revenue from social security contributions and divided by compensation of employees as shown in national accounts. This figure can be compared with the ITR on capital or other forms of income. Meanwhile, comparing changes in PIT revenue from a specific income source against changes in the corresponding component of national income will improve elasticity analysis, not only because of the closer relationship between the numerator and denominator but also because it will be easier to differentiate between the revenue impact of increases in (different forms of) national income on the one hand and changes to tax policy on the other (OECD, 2023[9]).
Macroeconomic analysis of PIT systems can corroborate microeconomic analysis, such as the calculation of effective tax rates (ETRs) on different income sources (Acciari and Heijmans, 2004[10]). For example, the estimation of ITRs on capital and labour in the economy as a whole can complement the findings of (Hourani, 2023[11]) that ETRs are generally lower for individuals who receive at least some income from capital sources than the ETRs of individuals who receive only wages. A combination of ITRs and ETRs can also help to understand the extent to which PIT systems may incentivise certain forms of tax arbitrage, for instance among owners of closely-held businesses, who tend to have greater flexibility about how they receive their income (Zawisza et al., 2024[12]).
PIT revenue in international classifications
Detailed information on the composition of PIT revenue can enhance comparative analysis of PIT systems across countries using the indicators mentioned above (and others). As this sub-section explains, PIT revenue is currently recorded at an aggregate level in international tax revenue classifications, including the OECD Interpretative Guide, whose classification of taxes forms the basis of Revenue Statistics.
The historic tendency noted above for countries to combine PIT revenue from different sources in a comprehensive system has informed the classification of PIT revenue in the Interpretative Guide. ‘Taxes on income, profits and capital gains’ is one of five main tax categories of the Interpretative Guide3, appearing under heading ‘1000’. As shown below, this category is divided into taxes on individual income (1100) and corporate income (1200) as well as a third category (1300) for instances where it is not possible to allocate income tax revenues either to categories 1100 or 1200.
Within category 1100, which is the focus of this chapter, a disaggregation is possible between taxes on income and profits (1110) and taxes on capital gains (1120). Capital gains are thus the only form that can be identified via the classification contained in the OECD Interpretative Guide, although they are not the only form of capital income individuals may receive.
Percentage of GDP
:Source: (OECD, 2025[8]).
The relatively small number of countries reporting CGT revenue (out of the 33 OECD countries that implement a CGT) attests to the challenges that countries may experience in distinguishing CGT revenue from other taxes on individual income for the purposes of Revenue Statistics4. These challenges in turn reflect differences in the taxation of different forms of capital gains within and across OECD countries as well as differences in how these various forms of capital gains tax are reported to tax administrations (Hourani and Perret, 2025[13]).
The generally low levels of CGT revenue among the countries that report them are likely to reflect exemptions or relief granted to certain asset classes in OECD countries (notably housing). They may also result from the fact that taxes on capital gains are primarily withheld at a flat rate, meaning tax administrations might not have access to detailed information on these revenues. In addition, they may reflect under-reporting for the reasons mentioned above, which impedes cross-country comparison of CGTs of the kind carried out by (Hourani and Perret, 2025[13]).
Revenue Statistics provides a means of distinguishing between the different sources of PIT revenue through the Country Tables contained in Chapter 5 of this Report. These tables show the labels of individual taxes that OECD countries report as part of the annual data submission as well as how much revenue is generated by each (and how each tax is classified within the Interpretative Guide).
While some countries identify a significant number of individual taxes that are classified under 1110, ten countries report two lines or fewer in this sub-category. In any case, the labels for different taxes often do not specify which income source is included and labelling can differ widely across countries, posing further challenges to comparative analysis.
Other international classifications encounter the same challenges concerning PIT. In the IMF’s Government Finance Statistics Manual, the category for taxes on income, profits and capital gains paid by individuals consists of a single category (category 1111) that cannot be disaggregated between capital gains and other forms of income. A single aggregation of income taxes is also used in the 2008 System of National Accounts and the European System of Accounts 2010.
To show the proportion of PIT revenue that comes from different sources of income on a comparable basis requires navigating the complexity of individual countries’ tax systems in a way that is replicable internationally. To this end, the European Commission’s Directorate-General for Taxation and Customs Union (TAXUD) has developed the ‘PIT Split’ exercise, which is described in greater detail in Box 2.1. This section outlines the three different approaches that countries can take to carry out the PIT Split, which have also been adopted for the purposes of this Special Feature.
The PIT Split is carried out by all European Union (EU) Member States and Norway. It produces comparable data on the proportion of PIT revenue that can be attributed to income derived from employed labour, capital, self-employment, or social transfers and pensions (European Commission: Directorate-General for Taxation and Customs Union, 2025[14]).5 This Special Feature contains information on the PIT Split not only for Norway and the 22 OECD countries that are also EU Member States but also for Australia, Costa Rica, Israel, Japan, New Zealand and the United States, which have estimated the PIT Split for this edition of Revenue Statistics.
The approach most commonly adopted by EU countries is based on a sample of tax returns submitted by individual taxpayers. Microsimulations are used to calculate the proportion of income that comes from the four different sources, taking into account the characteristics of individual taxpayers that may affect their level of income, how much income they receive from different sources and any tax allowances to which they might be entitled. This information is then used to estimate the proportion of income derived from each source and the tax paid on that income among the wider taxpayer population.
Under the second approach, countries do not use individual taxpayer data but instead base their analysis on tax return data that is aggregated by tax bracket or income class. This is found to produce broadly analogous results to the first approach, being similarly capable of distinguishing between taxpayers in terms of their income level, the proportion of their income that comes from different sources and their associated tax treatment (European Commission: Directorate-General for Taxation and Customs Union, 2025[14]).
For the third approach, the starting point is data on the withholding tax on employee wages. The wage withholding tax approximates tax payments on employed labour income for different groups of taxpayers, although this information needs to be adjusted both in light of corrections that may be made as part of the end-of-year reconciliation and because it may also include taxes levied on pensions or social benefits. Countries that adopt this approach tend to use aggregate tax-return data to calculate PIT receipts from capital and self-employment income.
The European Commission carries out the PIT Split exercise for its Data on Taxation Trends, a set of key taxation indicators for member states (European Commission: Directorate-General for Taxation and Customs Union, 2025[15]). This information in turn feeds into the Annual Report on Taxation, which analyses the design and performance of member states’ tax systems (European Commission: Directorate-General for Taxation and Customs Union, 2025[16]).
Through the PIT Split exercise, EU member states and Norway divide their PIT revenue into four income sources: employed labour, self-employed labour, capital, and social transfers and pensions. Table 2.1 shows the specific items that are included within each. This information is then used to calculate the breakdown of tax revenues by economic function and the implicit tax rates (ITR) on labour and capital.
Information derived from the PIT Split exercise is also shown in the National Tax Lists for each member state, wherein each individual tax is attributed to an economic function.
For this Special Feature, Norway and the 22 OECD countries that belong to the EU used the same methodology as they use for the PIT Split exercise. Australia, New Zealand and the United States6 used micro-data to carry out this exercise while Costa Rica and Israel used aggregate withholding tax and final assessment income tax data with certain adjustments. Japan estimated the proportion of employed labour income in PIT revenue using tax-return data, based on government statistics (sampling survey), aggregated at the level of income classes or tax brackets but did not calculate the proportions for the other three income types.
The following section shows the estimated breakdown of PIT revenue according to the four income sources as a percentage of total PIT revenue. These proportions are then applied to aggregate PIT revenue data reported to Revenue Statistics to estimate revenue from the different sources of income as a share of GDP.7 While this approach may provide or enable insights into the macroeconomic weight of each income source within PIT systems, the results should be interpreted with caution.
In addition to the aforementioned differences within and across the three approaches outlined above, TAXUD identifies a number of specific limitations to the comparability of the PIT estimations between countries (European Commission: Directorate-General for Taxation and Customs Union, 2025[14]). For the most part, these limitations affect the allocation of income to capital and social transfers and pensions.
For example, there are some countries for which not all taxable pensions or social benefits can be identified from tax return data or where the revenue impact of tax allowances cannot be calculated. Joint income declarations also pose a challenge where each spouse receives the bulk of their income from different sources. Further complexity arises from variation within the same country with respect to whether income, especially capital income, is taxed at source or within the framework of the individual tax return.
For this Special Feature, Revenue Statistics data on PIT revenue has been disaggregated according to the source of individual income for 29 OECD countries, although Japan was only able to estimate the share allocable to employed labour income. This section begins by examining the disaggregation of PIT revenue for each country in 2023 before analysing changes in the respective shares between 2011 and 2023. The disaggregation is shown both as a percentage of total PIT revenue and as a share of GDP in each country to assess both the relative importance of each income source within total PIT revenue and its macroeconomic weight.
The analysis in this section does not include an ‘OECD average’ indicator for the composition of PIT revenue that would be analogous to the average OECD tax structure shown in Chapter 1 of this Report as not all OECD countries were able to disaggregate their PIT revenue for this exercise.
Disaggregating PIT revenue by income source
As shown in Figure 2.3, employed-labour income was the largest source of PIT revenue in 2023 across the 29 countries, although its share varies between them. At the top of the range is the Slovak Republic, where employed labour contributed more than 100% of total PIT revenue in 20238; in a further eight countries, employed labour accounted for over 80% of total PIT revenue. In 15 countries, revenue from employed-labour income amounted to between 60% and 80% of total PIT revenue, while in five countries revenue summed to between 40% and 60% of the total.9
In 11 of the 28 countries that provided a full disaggregation, PIT on income from social transfers and pensions was the second-largest component of total PIT revenue. This income source amounted to over 20% of PIT revenue in Italy (29.6%), Denmark (25%), Finland (23.7%) and Portugal (22.7%) and was above 15% in an additional six countries. Social transfers and pensions generated less than 5% of PIT revenue in ten countries.
Percentage of PIT
Note: Data for the United States is from 2022 and shows the breakdown of revenue from federal income tax. Due to methodological constraints, Japan was only able to estimate the share of employed-labour income.
Income from self-employment was the second-largest source of PIT revenue in 10 countries, amounting to at least 15% of the total in Poland (37.9%), the Netherlands (28.9%),10 Germany (19.6%), Austria (19.5%), Italy (15.4%) and Greece (15.1%). Self-employed income accounted for less than 5% of PIT revenue in six countries.
Capital income was the second-largest income source for seven countries and contributed at least 10% of total PIT revenue in eight countries. The highest proportions were observed in Greece, Czechia11 (17.9% of total PIT revenue in both cases), Norway (13.6%) and France (13.2%). In ten countries, capital income accounted for less than 5% of total PIT revenue.
In two cases – Belgium and the Netherlands – a negative figure is recorded for capital income’s contribution to PIT revenue. In the case of Belgium, this is because most income from capital does not flow through PIT while some tax expenditures are assigned to the capital income component of the PIT Split. For the Netherlands, the negative figure is a consequence of PIT deductions on mortgage interest payments.
Next, this section shows the decomposition of PIT revenue by the four income sources as a share of GDP. This is calculated by applying the proportions shown above to each country’s overall PIT-to-GDP ratio using data shown in Chapter 1 of the Report (Figure 2.4).
In 2023, PIT revenue from employed-labour income as a share of GDP was highest in Denmark (17.1%) and also exceeded 10% in New Zealand (10.8%) and Australia (10.8%). The high level of PIT revenue from employed labour in these countries corresponds to the high effective PIT rates relative to other OECD countries, as shown in (OECD, 2025[17]). In 2023, the level of PIT in the overall tax wedge on labour income12 was the highest across the OECD in these three countries (excluding Iceland, which did not provide disaggregated data on PIT revenue for this Special Feature). Social security contributions (SSCs) in Denmark amounted to 0.1% of GDP while Australia and New Zealand do not levy SSCs.
Percentage of GDP
Note: Data for the United States is from 2022 and shows the breakdown of revenue from federal income tax as a share of GDP. Due to methodological constraints, Japan was only able to estimate the share of employed-labour income.
PIT revenue from employed-labour income exceeded 5% of GDP in a further 16 countries within the sample. On average across the 29 countries, PIT revenue from employed-labour income amounted to 6.0% of GDP, versus an average level of PIT revenue as a share of GDP of 8.6% for the sample.13
Denmark also observed the highest level of PIT revenue from social transfers and pensions in 2023, at 6.3% of GDP. Social transfers and pensions also exceeded 2% of GDP in Italy (3.3% of GDP), Finland (3.0%), Belgium (2.2%) and Sweden (2.1%). PIT revenue from self-employment exceeded 2% of GDP in the Netherlands (2.6%) and amounted to at least 1% of GDP in a further 11 countries. Meanwhile, PIT revenue from capital income tended to be lowest, amounting to at least 1.0% of GDP in only five countries with the highest level observed in Norway (1.3% of GDP).
On average across the 28 countries, PIT revenue from social transfers and pensions income amounted to 1.2% of GDP, while PIT revenue from self-employed income and from capital income amounted to 0.9% of GDP and 0.5% of GDP respectively.
Changes in the composition of PIT revenue over time
The composition of PIT revenue by income source has evolved over the period from 2011 to 2023 (Figure 2.5). As mentioned above, PIT’s share in the average OECD tax structure grew over this period, while revenue from PIT also increased as a share of GDP. Across the 29 countries examined in this Special Feature, PIT revenue rose as a share of GDP in all but six: Belgium, Czechia, Ireland, Norway, Slovenia and Sweden. The 2.4 p.p. fall in PIT revenue as a share of GDP in Ireland contributed to an overall decline of 6.3 p.p. in its tax-to-GDP ratio over this period, which was due to an exceptional GDP increase in 2015.
Difference in p.p. of total PIT revenue share
Note: Comparison for Japan is 2014-2023 and for the United States is 2018-2022. The breakdown of revenue from federal income tax is shown for the United States. Due to methodological constraints, Japan was only able to estimate the share of employed-labour income.
In a majority of the 28 OECD countries for which the full PIT disaggregation was available, the proportion of PIT revenue derived from employed-labour income declined over this period while the proportion derived from capital income increased. Revenue from self-employment income also increased as a share of total PIT revenue in a majority of countries while the revenue share of social transfers and pensions declined in slightly more countries than it increased.
Overall, revenue from employed-labour income fell as a share of total PIT revenue in 19 countries. In 17 of these, the share of revenue from capital income increased; the only exceptions were Italy and Poland, where both sources of income became less prominent in the PIT mix.
The largest falls in employed-labour income as a share of total PIT revenue occurred in Czechia and the Netherlands, with declines of 18.6 p.p. and 16.4 p.p. respectively.14 Among the eight countries where employed-labour income rose as a share of total PIT revenue, the largest increases occurred in the Slovak Republic (8.4 p.p.), Lithuania (5.7 p.p.) and France (5.2 p.p.).
Revenue from capital income rose as a share of total PIT revenue in 18 countries and declined in 10 between 2011 and 2023. The largest increases occurred in Czechia (15.7 p.p.), the Netherlands (9.9 p.p.) and Greece (7.8 p.p.) while the largest declines were seen in Portugal (5.9 p.p.) and Lithuania (3.9 p.p.).
Revenue from self-employment income increased as a share of total PIT revenue in 15 countries and declined in 12. The largest increases occurred in Poland (10.0 p.p.) and the Netherlands (9.2 p.p.)15 while the largest declines were in the Slovak Republic (8.9 p.p.) and France (4.4 p.p.).
Meanwhile, slightly more countries observed a decline in the share of social transfers and pensions in PIT revenue (13 countries) than observed an increase (12 countries), with the share unchanged in a further three countries. The largest declines in this category occurred in Greece (6.7 p.p.) and Poland (5.0 p.p.) while largest increases were in Belgium (4.5 p.p.), Spain and Italy (both 4.1 p.p.).
Difference in p.p. of GDP
Note: Comparison for Japan is 2014-2023 and for the United States is 2018-2022. The breakdown of revenue from federal income tax as a share of GDP is shown for the United States. Due to methodological constraints, Japan was only able to estimate the share of employed-labour income.
Figure 2.6 decomposes changes in PIT revenue as a share of GDP between 2011 and 2023 according to the four different income categories. PIT revenue from employed-labour income increased as a share of GDP in 18 of the 29 countries between 2011 and 2023, even though (as noted above) the contribution of employed-labour income to total PIT revenue declined in a majority of countries over this period.
This phenomenon attests to the continued (if diminished) prominence of revenue from employed-labour income in the PIT mix: revenue from other income sources – notably capital – increased by more than employed labour income as a proportion of their 2011 level but their increase was generally smaller than the increase in revenue from employed-labour income over this period when measured as a share of GDP.
The increases in PIT revenue from employed-labour income exceeded 2% of GDP in four countries between 2011 and 2023: Lithuania (3.7 p.p.), Luxembourg (2.6 p.p.), New Zealand (2.2 p.p.) and Australia (2.1 p.p.). Nonetheless, the share of employed-labour income in the PIT mix declined in both New Zealand and Australia.
In the 23 countries where PIT revenue increased as a share of GDP between 2011 and 2023, revenue from employed-labour income was the main driver in all but seven countries. In Denmark and Greece, the increase in revenue from capital income exceeded the increase in revenue from employed-labour income as a share of GDP. In Austria, Finland, Italy, the Netherlands and Poland, PIT revenue from employed-labour income declined as a share of GDP.
The largest declines in PIT revenue from employed-labour income as a share of GDP were in Ireland (1.9 p.p.) and Belgium (1.1 p.p.), although Ireland was one of the relatively small group of countries where revenue from PIT on employed-labour income became a more prominent part of the PIT mix over this period. In six of the 11 countries where PIT from employed-labour income decreased as a share of GDP, overall PIT revenue declined as a share of GDP; in the remaining five countries, revenue from the other sources offset the decline in revenue from employed-labour income, meaning that total PIT revenue increased as a share of GDP.
Meanwhile, PIT revenue from capital income increased as a share of GDP in 22 of the 28 countries between 2011 and 2023, declined in five countries and stayed the same in one. Increases of at least 0.5 p.p. were observed in Greece, Denmark and Czechia.16 The largest falls in PIT revenue from capital income occurred in Portugal (0.3 p.p.) and Ireland (0.2 p.p.).
PIT revenue from self-employed income rose as a share of GDP in 20 of the 28 countries between 2011 and 2023, while it declined in seven countries and stayed the same in one. The largest increases occurred in the Netherlands (1.2 p.p.), Poland, New Zealand (both 0.5 p.p.) and Israel (0.4 p.p.). In the Netherlands,17 Poland and Austria, self-employed income was the main driver of the overall increase in PIT revenue. The largest declines in PIT revenue from self-employed income were in Ireland (0.4 p.p.), the Slovak Republic (0.3 p.p.) and France (0.2 p.p.).
PIT revenue from social transfers and pensions rose as a share of GDP in 19 of the 28 countries between 2011 and 2023, declined in five countries and stayed the same in three. Increases of at least 0.5 p.p. were observed in Finland, Italy (both 0.6 p.p.) and Spain (0.5 p.p.). The largest declines were in Sweden (0.4 p.p.) and Poland (0.2 p.p.).
Revenue from social transfers and pensions was the main driver of the increase in overall PIT revenue as a share of GDP over this period in Finland and Italy. It was the only one of the four income categories for which PIT revenue did not decline as a share of GDP in Ireland.
PIT is at the core of the tax systems in OECD countries, yet analysis of this key source of revenue is constrained by a lack of information about the weight of different forms of individual income within PIT revenue. For this edition of Revenue Statistics, 29 OECD countries have provided for the first time a disaggregation of their PIT revenue according to four income sources that correspond to different economic factors: employed labour income, capital income, income from self-employment, and social transfers and pensions.
Derived from an approach developed by the European Commission, the data presented in this Special Feature can serve as a basis for detailed examination of PIT systems. In particular, it can facilitate analysis of the drivers of revenue trends and the interaction of PIT revenue with different components of the economy, while also allowing for a greater understanding of the redistributive impact of PIT systems.
The high-level analysis carried out in this chapter shows that employed-labour income accounted for the majority of PIT revenue in 2023 in all 29 countries. The prominence of the other income sources varied across the 28 countries that were able to provide a full disaggregation: in 11 countries, PIT on income from social transfers and pensions was the second-largest component of total PIT revenue, while self-employed income was the second-most important source for 10 countries and capital income was the second-most important source for seven countries.
The relative weight of different forms of income has changed between 2011 and 2023 across the countries examined. Revenue from employed-labour income fell as a share of total PIT revenue in 19 countries over this period, while the share of capital income and self-employed income in total PIT revenue increased in 18 countries and 15 countries respectively. Meanwhile, the share of social transfers and pensions in PIT revenue declined in 13 countries and increased in 12 over this period.
Revenue from employed-labour income nonetheless increased as a share of GDP in 18 of the 29 countries between 2011 and 2023 and was the main driver of increases in overall PIT revenue for all but seven of the 23 countries where PIT revenue rose as a share of GDP over this period. PIT revenue from capital income increased as a share of GDP in 22 of the 28 countries, while revenue from self-employed income and from social transfers and pensions increased in 20 countries and 19 countries respectively.
A deeper analysis of the changes in PIT revenue for these different sources of income is beyond the scope of this Special Feature. However, the trends identified here may be of great significance to OECD countries as they confront major structural changes while also dealing with short- and long-term spending pressures and trying to manage the impact of their tax systems on the distribution of incomes and economic growth. The value of information about the composition of PIT revenue is clear, especially if this is generated through an approach that can be replicated across countries, thereby permitting more-detailed comparative analysis via international revenue classifications.
References
[10] Acciari, P. and P. Heijmans (2004), Examination of the macroeconomic implicit tax rate on labour derived by the European Commission, European Commission, https://taxation-customs.ec.europa.eu/document/download/0c3d49dc-fb97-4d0c-9f57-9072c1c44a4c_en?filename=taxrate_labour.pdf.
[4] Akgun, O., B. Cournède and J. Fournier (2017), “The effects of the tax mix on inequality and growth”, OECD Economics Department Working Papers, No. 1447, OECD Publishing, Paris, https://doi.org/10.1787/c57eaa14-en.
[16] European Commission: Directorate-General for Taxation and Customs Union (2025), Annual report on taxation 2025 – Review of taxation policies in the EU Member States, ublications Office of the European Union, https://doi.org/10.2778/6367826.
[15] European Commission: Directorate-General for Taxation and Customs Union (2025), Data on Taxation Trends, https://taxation-customs.ec.europa.eu/taxation/economic-analysis/data-taxation-trends_en (accessed on 3 October 2025).
[14] European Commission: Directorate-General for Taxation and Customs Union (2025), Data on Taxation Trends - Methodological and explanatory notes, Publications Office of the European Union, https://taxation-customs.ec.europa.eu/document/download/9392a0fe-fc17-4cb6-b9c9-56e71f21caec_en.
[7] Genser, B. (2006), “The Dual Income Tax: Implementation and Experience in European Countries”, Ekonomski pregled, Vol. 57/3-4, pp. 271-288, https://hrcak.srce.hr/clanak/12398.
[1] Gerber, C. et al. (2018), “Personal Income Tax Progressivity: Trends and Implications”, IMF Working Papers, Vol. 246 (2018), https://doi.org/10.5089/9781484383087.001.
[11] Hourani, D. (2023), The taxation of labour vs. capital income: A focus on high earners, OECD Publishing, Paris, https://doi.org/10.1787/04f8d936-en.
[13] Hourani, D. and S. Perret (2025), Taxing capital gains: Country experiences and challenges, OECD Publishing, Paris, https://doi.org/10.1787/9e33bd2b-en.
[2] Jalles, J. and G. Karras (2025), “Rethinking tax progressivity: a new look at its role in OECD economic growth”, Int Tax Public Finance, https://doi.org/10.1007/s10797-025-09902-y.
[8] OECD (2025), Revenue Statistics 2025: Disentangling Personal Income Tax Revenue in OECD Countries, OECD Publishing, Paris, https://doi.org/10.1787/3a264267-en.
[17] OECD (2025), Taxing Wages 2025: Decomposition of Personal Income Taxes and the Role of Tax Reliefs, OECD Publishing, Paris, https://doi.org/10.1787/b3a95829-en.
[9] OECD (2023), Revenue Statistics 2023: Tax Revenue Buoyancy in OECD Countries, OECD Publishing, Paris, https://doi.org/10.1787/9d0453d5-en.
[3] OECD (2017), Taxation and Skills, OECD Tax Policy Studies, No. 24, OECD Publishing, Paris, https://doi.org/10.1787/9789264269385-en.
[6] Piketty, T., E. Saez and S. Stantcheva (2014), “Optimal Taxation of Top Labor Incomes: A Tale of Three Elasticities.”, American Economic Journal: Economic Policy, Vol. 6(1), pp. 230–7, https://doi.org/10.1257/pol.6.1.230.
[12] Zawisza, T. et al. (2024), “Tax arbitrage through closely held businesses: Implications for OECD tax systems”, OECD Taxation Working Papers, No. 70, OECD Publishing, Paris, https://doi.org/10.1787/24b4ed4d-en.
[5] Zidar, O. (2019), “Tax Cuts for Whom? Heterogeneous Effects of Income Tax Changes on Growth and Employment”, Journal of Political Economy, Vol. 127:3, pp. 1437-1472, https://doi.org/10.1086/701424.
← 1. The authors are very grateful for the support of colleagues from the European Commission: Directorate-General for Taxation and Customs Union (TAXUD) throughout the production of this Special Feature. This support includes a webinar that was jointly organised by TAXUD and the Secretariat for delegates of Working Party No. 2 (WP2) of the Committee on Fiscal Affairs on 20 June 2025 about the disaggregation of PIT revenue by income source. Nevertheless, analysis contained in this chapter and the views expressed herein can in no way be taken to reflect the official opinion of the European Union. The Secretariat is also very grateful to the WP2 delegates who presented their approach to the PIT Split at the webinar (Ladislav Jelínek from Czechia; Paolo Acciari from Italy; and Remco Mocking from the Netherlands) as well as all those who attended the event.
← 2. This Special Feature focuses on capital income taxes specifically; it does not examine taxes on capital, such as taxes on wealth, property or wealth transfer.
← 3. The other main tax headings are social security contributions (category 2000), taxes on payroll and workforce (category 3000), taxes on property (category 4000), and taxes on goods and services category 5000).
← 4. In instances where it is not possible to distinguish between sub-categories, the reporting guidelines for Revenue Statistics allow countries to report all revenue data under a single sub-category even if a portion of the revenue is derived from another sub-category. Where it is not possible to distinguish between sub-categories 1110 and 1120, countries may report all their revenues under 1100, as has been done by 25 countries for this edition of Revenue Statistics.
← 5. There is variation between how countries calculate the PIT Split, even within the same approach. For detailed information about how the EU countries and Norway, please consult pages 44-59 of the Methodological guide (European Commission: Directorate-General for Taxation and Customs Union, 2025[14]).
← 6. The United States calculated the proportion of individual federal income tax allocable to each of the four sources of income for the period from 2018 to 2022. The analysis for the United States in this Special Feature therefore does not include total PIT revenue reported to Revenue Statistics, which also comprises revenue from income tax at the state and local level.
← 7. For countries such as Belgium that apply the time-shifted cash method to record PIT revenue in national accounts (and in Revenue Statistics), data for the PIT Split does not fully match the revenue data reported in national accounts and, hence, in Revenue Statistics. The PIT Split is compiled when the standard return and assessment procedure for a given income year is completed. This permits the PIT Split to remain unaffected by issues regarding the timing and method of tax payments (earned income withholding, advance payments or assessment payments). As such, the PIT Split approximates full accrual but it takes more than 1.5 years before the split is available (e.g. for Belgium the split for 2023 only became available in September 2025). The time lags in national accounts reporting are much shorter due to the use of time-shifted cash (instead of full accrual). With the time-shifted cash method, tax receipts from withholdings and advance payments are aggregated with assessment balances which to a large extent do not relate to the same income year. Moreover, it may take several years to settle disputes related to PIT payments. With the time-shifted cash method, the receipts resulting from disputes and regular annual assessments are recorded when the assessments are determined, which may be months or even years after the year in which the underlying income was earned.
← 8. The Slovak Republic explained its methodology for calculating the PIT Split as follows: The Ministry of Finance split total PIT revenues into employed labour, self-employment, and capital shares by combining aggregate tax figures with individual-receipt data. Three aggregates were observed separately: (1) withholding tax, (2) PIT on employed-labour income, and (3) PIT on self-employment and other sources. The first two aggregates directly feed into the capital and employed-labour categories, respectively. The third aggregate was apportioned between capital and self-employment income using microdata: individual tax bases were summed by income type, and the capital-income share of that aggregate was derived from the ratio of capital to total income bases. The capital share of PIT is then calculated as withholding tax plus the capital share of the self-employment/other aggregate; the self-employment share is the remainder. Social transfers and pensions are tax-exempt, so their PIT share is zero.
← 9. The share of employed-labour income in the Netherlands fell from 64.2% in total PIT revenue in 2022 to 56.0% of total PIT revenue in 2023, while the share of PIT revenue from self-employment rose from 21.9% to 28.9% over the same period. This shift in the structure of PIT revenue was a result of closely held businesses distributing profits to themselves in anticipation of a tax reform in 2024. The once-off nature of the change to the proportions of employed-labour income and self-employed income within overall PIT revenue should be borne in mind when considering the PIT Split for 2023 and the analysis of changes in the structure of PIT revenue between 2011 and 2023 elsewhere in this Special Feature.
← 10. See Note 9.
← 11. The results shown for Czechia and other EU countries in this Special Feature should be treated with caution due to a slight difference in the reporting of revenue data between Revenue Statistics and the European Commission’s Directorate-General for Taxation and Customs Union. In Revenue Statistics, only the transfer component of tax credits is added to tax income, whereas the European Commission includes both the transfer and expenditure components. Further information regarding the treatment of tax credits for income taxation within Revenue Statistics for OECD Member countries can be found in the section “Non-wastable Tax Credits” in Chapter 1 of this report.
← 12. The tax wedge is an indicator of the effective tax burden levied on the total labour costs for a hypothetical worker in full-time employment. The indicator includes PIT, SSCs paid by employers and employees and the cash transfers received by the government. The results shown here refer to the tax wedge of a single worker without children earning the average wage in each country.
← 13. These averages include only federal income tax for the United States, not state or local PIT revenue.
← 14. See Note 9.
← 15. See Note 9.
← 16. In addition, revenue from capital income in the Netherlands changed from -0.8% of GDP to -0.2% of GDP between 2011 and 2023.
← 17. See Note 8.