Tax Policy Measures in Advanced and Emerging Economies: A Novel Database
  • 1 0000000404811396https://isni.org/isni/0000000404811396International Monetary Fund

This paper describes a new, comprehensive database of tax policy measures in 23 advanced and emerging market economies over the last four decades. We extract this information from more than 900 OECD Economic Surveys and 37,000 tax-related news from the International Bureau of Fiscal Documentation using text-mining techniques. The innovation of this dataset lies in its granularity: changes in the rates and bases of personal and corporate income taxes, value added and sale taxes, social security contributions, excise, and property taxes are systematically documented. In addition, the database provides information on the announcement and implementation dates, whether the measures represent major changes, are part of a broader tax package, and phased in over several years. The paper also presents a range of stylized facts suggesting that information from this database is useful to deepen the analysis of tax policy changes for research and policy purposes.

Abstract

This paper describes a new, comprehensive database of tax policy measures in 23 advanced and emerging market economies over the last four decades. We extract this information from more than 900 OECD Economic Surveys and 37,000 tax-related news from the International Bureau of Fiscal Documentation using text-mining techniques. The innovation of this dataset lies in its granularity: changes in the rates and bases of personal and corporate income taxes, value added and sale taxes, social security contributions, excise, and property taxes are systematically documented. In addition, the database provides information on the announcement and implementation dates, whether the measures represent major changes, are part of a broader tax package, and phased in over several years. The paper also presents a range of stylized facts suggesting that information from this database is useful to deepen the analysis of tax policy changes for research and policy purposes.

I. Introduction

Tax policy measures that mobilize additional revenue, enhance competitiveness, and boost productivity remain at the center stage of policy debates in many advanced and emerging market economies (IMF, 2016 and 2017). Tax policy is under constant reform, but it is very cumbersome to obtain detailed information of what transpired in different countries, beyond changes to basic parameters such as tax rates. The study of past reforms and their impacts, however, is crucial to gauge the impact of future reforms and develop informed advice on viable reform directions.

This paper introduces a new, comprehensive database of tax policy measures adopted in 23 advanced and emerging market economies over the last four decades—the Tax Policy Reform Database (TPRD). The TPRD contains more granular information on tax policy actions compared to existing databases. Tax measures are extracted from more than 900 Organization for Economic Cooperation and Development (OECD) Economic Surveys and 37,000 tax-related news from the archives of the International Bureau of Fiscal Documentation (IBFD)1 using text mining techniques (see Gentzkow and others, 2017).2,3

The innovation of the TPRD lies in the systematic documentation of the direction of changes in rates and tax base for six different tax types—personal (PIT) and corporate (CIT) income taxes, value added and sale taxes (VAT), social security contributions (SSC), excises (EXE), and property taxes (PRO). The database also contains information on the exact announcement and implementation dates of tax measures (e.g., day and/or month and year), whether the measures represented major tax changes (e.g., tax reforms), and if they were phased over multiple years. For fiscal consolidation years, as defined in Alesina and others (2017) and Dabla-Norris and Lima (forthcoming), the database reports governments’ estimates of the intended revenue yield, when available. Moreover, the raw data is such that more granular categorizations are possible as needed for policymakers and researchers.

Our database offers several distinct advantages. First, it allows for classifying the precise nature of discretionary tax policy measures across advanced and emerging countries, including in areas for which no time-varying policy indicators currently exist (e.g., changes in exemptions, tax thresholds, or capital gains taxation). In this regard, it is similar to the approaches used by Kawano and Slemrod (2016) to identify CIT rate and base changes, and by Vegh and Vuletin (2015) to document rate changes for PIT, CIT, and VAT. In contrast with these studies, we identify and document both rate and base changes across a range of taxes. Second, we identify the timing (i.e., announcement and implementation dates) of tax policy changes since the early 1970s. This, in turn, enables an assessment of the anticipation effects associated with tax policy changes, which have been found to be empirically relevant (see Mertens and Ravn, 2012, for the United States). Finally, our database complements the information available from “The Taxes in Europe” database of the European Commission (TEDB), which provides detailed information on the nature of tax measures (e.g., type of tax change, timing of tax change, and expected revenue impact) introduced in European Union member countries since 2012.4 In contrast, the TPRD covers a longer time horizon, and contains detailed information for a more diverse group of advanced and emerging market economies.5

Our database lends itself to numerous new applications of relevance to researchers and policymakers alike. For example, our database could help shed light on whether the observed decline in standard tax rates conceals base broadening measures or was accompanied by changes in other tax rates (e.g., whether the historical downward trend in CIT rates across countries has been accompanied by base broadening measures). Similarly, it could help examine the (dynamic) macroeconomic effects of tax policy packages, including their composition and potential synergies across different reforms. The information contained in the TPRD can also help advance knowledge on the economic effects of tax policy measures with significantly different implementation lags—the difference between implementation and announcement dates, and on the political economy of these reforms. Finally, it could help in identifying case studies for specific tax policy changes (e.g., changes in VAT or PIT thresholds) which policymakers may be contemplating.

At the same time, the TPRD should be seen as work in progress since the quality of the information gathered varies across countries, time periods, and types of tax measures. In this regard, PIT, CIT, and VAT measures introduced between 1988–2014 offer the most comprehensive coverage in terms of information available in the database. Moreover, our database does not attempt to measure and compile policy settings across countries. Also, it does not aim at providing an exhaustive accounting of all tax policy measures introduced by a country over the sample period. Going forward, users could further improve and expand the database, including by extending its country and time coverage, adding details on tax measures included in the database or currently excluded from it, and closing information gaps.

The level of detail contained in the TPRD complements the existing literature which has compiled information on fiscal policy actions using different approaches. This paper is closely related to studies that have used the narrative approach to estimate the economic impact of exogenous changes in fiscal policy. This approach consists of gathering information about the size, timing, and motivation of policy interventions from narrative documents, such as legislative documents, news articles, and presidential speeches (see, for example, Ramey and Shapiro, 1998; Romer and Romer, 2010; Ramey and Zubiary, 2015).6 Relatedly, a number of papers have used a “natural experiment” approach for identifying and measuring fiscal policy actions and examining their short-term macroeconomic cyclical impacts (see, for example, Ramey, 2011; Strawcynski, 2013; Auerbach and Gorodnichenko, 2012; Riera-Crichton et al., 2015). For instance, Riera-Crichton et al., (2015) build a value-added tax rate database at a quarterly frequency to estimate tax multipliers for a large sample of advanced economies. However, most studies focus on a single country, typically the US, and do not decompose exogenous tax policy changes between rate and base effects for different taxes. To check for accuracy, we compared our database to other databases constructed using a narrative approach for individual countries.7

This paper presents two sets of stylized facts which could be explored further in future research. The first is relevant for empirical analyses on the effects of tax policy. Namely, we find that the majority of tax policy measures introduced in the sample affect the tax base rather than tax rates, and are part of broader tax policy reform packages. These two aspects of tax policy changes have been neglected in the literature. The second set of stylized facts matters for understanding the drivers of tax policy. Specifically, we examine regularities in the timing of announcements of tax policy changes and find that tax increases occur relatively more frequently in periods of economic recessions and post-election years than in expansions and in the run-up to elections, respectively. We also find that in episodes of fiscal consolidation, tax increases are often offset by tax decreases, suggesting that policymakers attenuate the distortionary and/or distributional effects of higher taxes. These results are subject to significant heterogeneity across countries and tax types, which we likewise document.

The remainder of the paper is organized as follows. Section 2 describes how the database is constructed. Section 3 presents stylized facts on various characteristics of tax policy measures, while Section 4 explores the aspects related to their timing. Section 5 concludes.

II. Database Construction

A. Sources of Information

The information on tax policy measures was obtained by examining documented policy actions reported in 953 OECD Economic Surveys for 23 advanced and emerging economies over the last four decades.8,9 The advantage of using OECD Surveys as main source of information over other types of publications, such as IMF country reports or private company’s country analysis, is that these reports provide the most comprehensive assessment and documentation of a country’s main tax policy changes across all tax types we considered.

Data on tax policy measures was further supplemented with information on announcement and implementation dates. For the period 1988–2014, this information is primarily obtained from tax-related news contained in the archives of the IBFD.10 IBFD news are compiled by tax experts and provide very detailed information on the nature and exact timing of announcement and implementation (i.e., day, month, and year) of tax policy changes in a large number of countries from 1988. For the period preceding 1988, this information is primarily extracted from the so-called OECD “Calendar” or “Chronology” of main economic events, which had been a standard annex of OECD Surveys until 2003–2005. Compared to the IBFD, OECD calendars typically provide less detail on the timing of tax changes (i.e., the day of implementation/announcement is often unavailable). As a result, the precision in dating measures before 1988 is lower than in the case of measures adopted after 1988. When information on the timing of tax changes is not found in IBFD or OECD calendars, it is retrieved by assessing the information available in the textual fragments of the OECD Surveys, which typically allows one to identify announcement and implementation years.

The information on identified tax measures was cross-checked against available external indicators to ensure accuracy and detect possible information gaps. Specifically, identified rate changes for PIT, CIT, and VAT were confirmed by comparing the relevant information from the OECD Surveys with quantitative data available from the IMF tax rate database (1980–2014); the European Commission tax indicator database (1995–2015); the Global KPMG tax rates database (2006–2015), and USAID collecting taxes database (2007–2012). For PIT, CIT, and VAT base measures that were dated with IBFD news, the checks involved comparing the relevant information from the OECD Surveys with the detailed information available from the IBFD archives. Information for some countries was also checked against well-established databases that were constructed using a narrative approach. Cross-checks for SSC, EXE, and PRO taxes are currently ongoing and any potential revision in the coding and/or timing of these measures will be reflected in the next version of the database. It should be noted, however, that while these checks help improving the quality of the database, they do not rule out the risk of finding inconsistencies and/or omissions in the database.

China and India represent two special cases because only a handful number of OECD Surveys is available for these countries.11 To increase the coverage of tax policy changes in these countries, the information from the Surveys was integrated with hand-picked information from alternative sources, such as IMF internal documents, IBFD archives, and national sources. This allowed to us to significantly expand the coverage of tax policy measures for these countries.

B. Steps Involved

The construction of the TPRD involved several steps (Figure 1 and Appendix A). A first step encompassed processing information contained in the 953 OECD Surveys and 37, 943 IBFD news clips with the view to identify excerpts of these documents that describe changes in any of the six taxes considered (i.e., PIT, CIT, VAT, SSC, EXE, and PRO). This was done by defining a system of text-based rules to extract fragments of text from unstructured documents such as the OECD Surveys.12 Such a system of rules represents a fairly flexible apparatus, a “tax vocabulary” of sorts, that codifies how policy changes in any of the six taxes under consideration are typically described in OECD Surveys (i.e., which terms are used to discuss a tax change and how these terms interact with each other).13

Figure 1.
Figure 1.

Steps to Develop the Tax Policy Reform Database

Citation: IMF Working Papers 2018, 110; 10.5089/9781484354865.001.A001

The second step required assessing by hand which of the excerpts of OECD Surveys among the ones identified in the previous step, constituted an actual policy change for any of the six taxes under consideration. The validated policy changes were then classified along several different dimensions, including the type of tax (e.g., CIT, VAT); the type of change (i.e., rules governing the tax base and tax rates), the direction of the change (i.e., increase, decrease); whether the measure represented a major tax change (or “reform”); if the measure was announced as part of a package; whether the measure was phased over several years; and if the measure was introduced in a consolidation year (see Section C for details).

In a third step, announcement and implementation dates of each classified measure were documented using the information available from IBFD archives and OECD’s calendars, or from the excerpts themselves if information from IBFD and OECD was unavailable.14 A fourth and final step consisted of checking the accuracy of the information on PIT, CIT, and VAT measures in the database against quantitative information on rate changes and qualitative information on base changes available from alternative data sources (see Appendix A. and Appendix E). These checks provided confidence on the accuracy of the information contained in the database, and also helped in identifying and documenting information gaps. Such gaps are part of ongoing analysis and will be addressed in the next version of the database.

C. Identifying Tax Policy Measures

The nature of tax measures selected from the OECD Surveys is fairly heterogenous. In particular, identified tax measures span important reforms (e.g., the introduction of VAT in France; the overhaul of PIT taxation in Italy and Poland; and the reform of CIT taxation in Ireland) as well as minor measures related to small changes in tax rates, or the adoption of tax expenditures for specific products or taxpayers. Documented tax measures were, therefore, further differentiated between major tax changes with potentially large macro-fiscal implications (e.g., tax reforms) and those measures with potentially more limited economic effects. This was done by fixing a threshold for rate changes, and relying on informed judgement for base changes using all available information from the OECD Surveys, IBFD archives, as well as analytical studies (see below and Appendix E). For each tax category, we recorded the direction of change (i.e., increase or decrease). Table 1 provides specific examples of different types of base and rate changes contained in the database.

Table 1.

Examples of Different Rate and Base Changes by Type of Tax

article image
Source: IMF, OECD, IBFD.Note: The announcement and implementation dates include an “x” when information on the day or month is not available. For example, if only the month and year of a measure is known, the announcement or implementation dates will look like 4/x/1989.

For tax rate changes, a rate change larger than 1 percentage point in absolute value was considered a major reform.15 While arbitrary, this threshold appears to strike a reasonable balance between accuracy (i.e., small rate changes may have significant fiscal implications if applied to a large base) and parsimony (i.e., to keep the number of major tax changes to a manageable level). In the case of excise taxes (per unit), the determination of whether a tax change is major is based on an assessment of the language used in the OECD report.

For base changes, a major tax reform is identified when the change in the tax base (e.g., a broadening or a reduction of the tax net) affects a large group of taxpayers or has the potential to mobilize significant resources. Based on the description of tax policy changes contained in the OECD Surveys and IBFD news stories, we coded a change that broadens the tax base system (defined as increasing tax revenue holding constant the statutory tax rate, other tax base aspects and the behavior of economic agents) as an “increase” while a change that reduces the tax base is denoted as a “decrease”. Detailed information on specific tax base changes and description of the precise legislative and regulatory actions that underpin observed large changes is captured in the database in text format.

A number of conventions were followed in classifying different tax changes. The introduction/removal of a tax was coded as a base measure and so were changes in income brackets (unless specified otherwise). A reduction in the number of tax brackets was coded as a base broadening measure following the assumption that simplification can boost compliance. The extension or postponement of a tax measure (e.g., a temporary surcharge is maintained for an additional year, the reduction in PIT rate is delayed) were coded as an actual tax change aimed at avoiding the effects of the planned tax change. Accordingly, for example, the postponement of a rate reduction was coded as a rate increase because absent such a postponement, the rate would have been lower.

In general, major tax base changes were identified in different ways depending on the type of tax:

  • A CIT base change pertains to any of the following categories: R&D promotion (e.g., tax credit), investment promotion (e.g., depreciation rules), loss-carry rules, thin capitalization, and capital gains. If a base change does not fall into any of these categories (e.g., generic exemptions), it is classified as belonging to “other base changes”. Other base changes are considered as “major” only if the information from the information from the OECD Survey or IBFD archives suggest that such changes constitute a significant reform.16 These categories are broadly consistent with the ones used in Kawano and Slemrod (2016).

  • A PIT base change pertains any of the following categories: standard relief (e.g., single person or family deductions, tax credits); child relief (e.g., tax credit, deductions); relief on capital gains; interest relief; and relief for SSC, insurance premiums, and private pensions. Base change not falling into any of these categories (e.g., deductions for special purposes) are classified as “other base changes”, which may or may not be considered as major depending on the information available from OECD Surveys and IBFD archives. These categories are broadly consistent with the analysis in OECD (2006 and 2016b).

  • A VAT base change pertains any of the following categories: exemptions on food items, exemptions on medical supplies, and exemptions on education. All other VAT base changes (e.g., introduction of VAT, generic exemptions) are classified as “other base changes”. Other base changes are considered as major only if the information available from OECD Surveys and IBFD archives corroborates such a conclusion. These categories are broadly consistent with the analysis in OECD (2016a) and IMF (forthcoming).

  • A base change in SSC, EXE, or PRO is “major” when available information from OECD Surveys and/or IBFD archives suggests that such a change affects large groups of taxpayers or could potentially mobilize significant resources. This criterion is arbitrary and reflects the lack of consensus in the literature on what constitutes a major SSC, EXE, or PRO base change.

The current definition of major tax measure does not imply any loss of information on tax changes, given that the database contains all the underlying tax excerpts. This allows other users and researchers to customize their definition of major change. In this regard, the database represents a unique source of information and lends itself to multiple uses, including by allowing users to generate new databases that better fit their research question.

The information on the type of tax measures was supplemented with additional documentation on the identified measures. A dummy variable (taking the value of 1, and zero otherwise) identifies whether the coded measure is part of a package, if this is explicitly mentioned in the OECD report or if the measures share the same announcement date. Similarly, we identify whether the measure was phased over several years (i.e., multiyear).17 Finally, for tax measures announced during a consolidation period, as defined in Alesina and others (2017), the TPRD provides information on the expected revenue yield of each measure as reported in Dabla-Norris and Lima (forthcoming). This was done by associating, when possible, consolidation measures and related expected revenue yields with the corresponding measures in the tax measures database (see Appendix B for more details).

While the TPRD holds promise to become an important tool for tax policy analysis, important caveats apply. First, the approach does not rely on a common single metric to identify tax base changes, unlike some earlier studies that relied on changes in tax rates. While we use transparent criteria to identify base changes, there is more judgement involved as to whether a given measure constitutes a major base change. Second, the database provides no information regarding the stance of current (or past) tax policy or tax structure. Third, tax changes, other than in consolidation episodes, are identified as a binary dummy as opposed to continuous variables, which does not allow to measure the size of the change. Moreover, in contrast to consolidation episodes, the exact motivation underlying the tax change is not identified. Importantly, the TPRD is preliminary and should be viewed as work in progress. The quality and the accuracy of tax policy information varies across tax types, countries, and over time, with more detail available for recent decades.

III. Stylized Facts18

The database documents 3,285 tax policy measures, equivalent to an average of 5 tax measures per country year. Aggregate figures, however, masks significant cross-country heterogeneity (Figure 2 and Table Appendix E.1). In emerging market economies (Brazil, China, Mexico, and Poland), the database only captures an average of 3 to 4 tax measures per year. For advanced economies, such as France, United Kingdom, Germany, United States, and Italy, the average number of measures exceeds 6 per year.

Figure 2.
Figure 2.

Average Number of Tax Measures by Country

Citation: IMF Working Papers 2018, 110; 10.5089/9781484354865.001.A001

Source: Tax Policy Reform Database, OECD, IBFD.

Table 2 reports different characteristics of the tax measures in the database. Most measures entailed a change in the tax base (about 60 percent of the total identified measures); ⅔ of these changes implied a decrease in tax liabilities. By contrast, the composition of rate changes between increases and decreases appear to be much more balanced, with rate increases accounting for almost half of identified rate changes. Table 2 also indicates that more than 70 percent of all identified tax measures involved “major” tax changes or reforms in a single year, with a majority of these introduced as a package of tax measures. Among major reforms, a decrease in the tax base introduced as part of a policy package in a single year was most common (accounting for a 20 percent of all identified tax measures), followed by a base increase (12 percent of total identified measures). Major rate decreases introduced as part of a package in a single year represent 10 percent of total identified measures.19

Table 2.

Frequency Distribution of Tax Policy Changes in the Database

(in percent of total identified measures)

article image
Source: Tax Policy Reform Database, OECD, IBFD

Because multiple tax measures can occur in the same year, it is important to also examine the distribution of tax policy changes in terms of country-year occurrences. This allows to gain a better understanding of how frequent tax changes are in terms of sample years.20 Table 3 describes the basic feature of the database in country year occurrences and average number of measures per country years. Base measures occurred more often than rate measures in the whole sample (575 versus 520 country years out of 672 country years). In addition, base decreases outnumbered base increases, while rate increases were more frequent in the case of minor measures (132 versus 101 country years) and measures that were not announced as part of a tax package (183 versus 167 country years).

Table 3.

Distribution of Tax Policy Changes in the Database

(count of country years)

article image
Source: Tax Policy Reform Database, OECD, IBFD

Table 4 presents the co-occurrence of tax base and rate changes as well as increases and decreases in tax liabilities expressed in country years. In the sample, changes to the tax rate are more frequent when the tax base also changes (423 country years out of 520 country years in which there a rate change)21 than in cases where the tax base does not change (97 country years out of 520).22 Furthermore, Table 4 shows that decreases in tax liabilities occur more frequently when there are also increases in tax liabilities than in cases where tax liabilities are not increased. These results are consistent with the conclusions reached by Kawano and Slemrod (2016) for the relationship between CIT base and rate changes. By extending the analysis to more types of taxes, our results confirm that ignoring the effects of base changes could potentially bias the estimated economic effect of rate changes and provide a narrower perspective on the various dimensions involved with tax policy reforms. This fact also illustrates well the potential advantages of using our more comprehensive database of occurrences of changes in multiple aspects of tax bases for different tax types.

Table 4.

Tax Policy Measures by Type and Direction of Change

(count of country years)

article image
Source: Tax Policy Reform Database, OECD, IBFD.

In the reminder of this section, we present additional stylized fact that focus on major tax measures (representing over 80 percent of observed tax policy changes or 95 percent of country years occurrences). Figure 3 presents the breakdown of major tax measures by tax type. Changes in the PIT, CIT, and VAT account for around 80 percent of total major tax changes for the entire sample. Changes in SSC are also quite frequent (7 percent of total major measures), while EXE and PRO measures occur less frequently in our database. This hierarchy across different tax types holds irrespective of whether the sample of major measures is restricted based on whether the tax changes are part of a tax package or multiyear in nature.

Figure 3.
Figure 3.

Frequency of Major Tax Reforms by Tax Type

(number of observed tax policy changes)

Citation: IMF Working Papers 2018, 110; 10.5089/9781484354865.001.A001

Source: Tax Policy Reform Database, OECD, IBFD.

Country-level information confirms the relative frequency of reforms tax type (Figure 4). Namely, the bulk of tax reforms is represented by changes to PIT, CIT, and VAT in all countries, except in France, Italy, and Brazil, where SSC measures were more common than changes in the VAT. PRO reforms represent a share of total tax changes above 6 percent in Japan, South Korea, France, China, and Italy; while, reforms to excises occurred more than in 8 percent of tax changes in Turkey, United Kingdom, Canada, Germany, and Denmark.

Figure 4.
Figure 4.

Frequency of Major Tax Reforms by Country and Tax Type

(number of observed tax policy changes)

Citation: IMF Working Papers 2018, 110; 10.5089/9781484354865.001.A001

Source: Tax Policy Reform Database, OECD, IBFD.

Changes in the different taxes can be further decomposed into reforms related to changes in the rate or base for each tax instrument (Figure 5). Most PIT, CIT, and PRO measures comprise base changes, while rate changes were more salient in the case of VAT, SSC, and EXE reforms. In particular, about 2/3 of all PIT, CIT, and PRO consisted of base changes. This suggests that any analysis assessing the economic impact of PIT and CIT rate changes without considering base changes is likely to suffer from significant biases.

Figure 5.
Figure 5.

Composition of Major Tax Reforms by Tax Type and Type of Change

Citation: IMF Working Papers 2018, 110; 10.5089/9781484354865.001.A001

Source: Tax Policy Reform Database, OECD, IBFD.

Aggregate figures, however, mask significant cross-country heterogeneity. Figure 6 shows that in most countries major PIT, CIT, and PRO measures were dominated by base changes, with the exception of Japan and China for PIT; Ireland and Luxemburg for CIT; and China, Portugal, Denmark, and Czech Republic for PRO. At the same time, major VAT base changes outnumbered rate changes in approximately 1/3 of the sample. By contrast, major SSC and EXE measures were typically dominated by rate change, except for Spain, Italy, France, and Turkey for SSC; and Portugal for excises.

Figure 6.
Figure 6.

Composition of Major Tax Reforms by Tax Type, Type of Change, and Country

Citation: IMF Working Papers 2018, 110; 10.5089/9781484354865.001.A001

Source: Tax Policy Reform Database, OECD, IBFD.

Figure 7 provides the breakdown of major tax policy changes by tax and direction of change (i.e., increase or reduction in tax liability). A reduction in the tax liability was more common for major CIT and PIT reforms, while increases in tax liabilities were more pronounced for SSC, VAT, EXE, and PRO reforms. This result is broadly consistent with the gradual shift from direct taxation to indirect taxation observed in many advanced economies over the last decades (OECD, 2010). This does not necessarily imply that documented tax changes led to an increase or reduction in the tax burden expressed as the ratio of tax revenue-to-GDP.23 Information on the size of rate changes from available databases suggests that, on average, CIT and PIT rates have declined, while VAT rates have increased (Box 1).

Figure 7:
Figure 7:

Composition of Major Tax Reforms by Tax Type and Direction of Change

Citation: IMF Working Papers 2018, 110; 10.5089/9781484354865.001.A001

Source: Tax Policy Reform Database, OECD, IBFD.

Our results do not seem to be driven by outliers (Figure 8). Namely, the frequency of major reductions in tax liabilities of the PIT and CIT is common across all countries in the sample, except for the case of Poland and Portugal for the PIT, and Austria and Greece for the CIT. SSC and PRO measures appear to be more evenly distributed between increases or decreases in tax liabilities. In these cases, however, some countries only saw increases or decreases in tax liabilities, and not both. For example, Spain, Japan, South Korea, and Greece only experienced increases in the tax liabilities related to SSC, while Portugal experienced only decreases. In the case of major PRO measures, the number of countries that introduced only one type of change (i.e., increase or decrease in tax liability) is significantly higher (e.g., in United States, Australia Turkey, Ireland, India, Portugal, Luxembourg, and Czech Republic).

Figure 8.
Figure 8.

Composition of Major Tax Reforms by Tax Type, Direction of Change, and Country

Citation: IMF Working Papers 2018, 110; 10.5089/9781484354865.001.A001

Source: Tax Policy Reform Database, OECD, IBFD.

Changes in direct taxation featured more prominently for measures that were announced as part of broader tax reform packages (Figure 9). Specifically, major PIT, CIT, and PRO changes accounted for about ¾ of all tax changes occurred within a package, while this proportion dropped to [60] percent in the case of tax measures outside a package. The relative importance of SSC measures increased when such measures were not announced in conjunction with a broad tax package.

Figure 9.
Figure 9.

Composition of Major Tax Reforms by Tax Type and Package

Citation: IMF Working Papers 2018, 110; 10.5089/9781484354865.001.A001

Source: Tax Policy Reform Database, OECD, IBFD.

Developments in Tax Rates

On average, CIT and PIT rates have been on a declining path in advanced and emerging market economies. Based on the recently developed database (Vegh and others, 2015), CIT rates in advanced economies (16 countries included in our database for which data was available) have on average declined from about 44 percent in 1970s to around 26 percent in 2017. In emerging market economies (6 countries included in our database for which data was available), CIT rates on average have declined from about 33 percent in 1980s to 22 percent in 2017. Similarly, PIT rates for advanced and emerging market economies for the same period have on average declined from nearly 70 percent and 60 percent to 44 and 36 percent, respectively. In contrast, VAT rates in both groups of countries increased. In particular, the average VAT rate increased for advanced and emerging market economies for the same period from 14 to 18 percent and from 13 to 16 percent, respectively.

uA01fig01

Tax rate changes in advanced and emerging market economies (in percent)

Citation: IMF Working Papers 2018, 110; 10.5089/9781484354865.001.A001

Source: Vegh and others, 2015

The tax package announced in a typical year comprised about 4 tax measures, included both rate and base changes in approximately 60 percent of total country years (Table 5). Only in 11 percent of the cases were rate changes announced alone. This result confirms the fact that changes in the tax rates are typically announced in conjunction with base changes, often times of a different nature. Namely, 24 percent of total country years combined at least one measure to increase a tax rate with at least one base narrowing measure. Similarly, approximately 26 percent of total country years combined at least one measure to decrease a tax rate with at least a base broadening tax change.

Table 5.

Major Tax Reforms in a Tax Package by Type and Direction of Change

article image
Source: Tax Policy Reform Database, OECD, IBFD

Countries, on average, announced less than a tax package per year or an average of about 25 tax packages over the entire sample period. However, this masks significant differences across countries. Namely, G7 countries with the exception of Japan announced significantly more tax packages (approximately between 35 and 45 packages over the sample period) which appeared to be broadly balanced in terms of number of tax rate and base changes (Figure 10).24

Figure 10.
Figure 10.

Frequency of Major Tax Reform by Type of Tax Change, Country, and Number of Tax Packages

(The size of the bubble is given by the number of tax packages announced)

Citation: IMF Working Papers 2018, 110; 10.5089/9781484354865.001.A001

Source: Tax Policy Reform Database, OECD, IBFD.

The database also provides information on the implementation lags for most tax measures included in the sample, particularly for the period 1988 to 2014.25 Figure 11 shows that most major CIT, PIT, and VAT measures were implemented with sizable delays. The average and median implementation lag for these three taxes was 153 and 78 days, respectively. While the average implementation lag points to a rather long lead time for economic agents to adjust their behavior, there is significant variation across countries, tax types, and years. In particular, CIT measures featured a higher variation in implementation lags than PIT measures across countries, although both type of tax measures showed a similar median implementation lag. By contrast, the median implementation lags and cross-country variation are much lower for VAT measures.

Figure 11:
Figure 11:

Aggregate Implementation Lags

Citation: IMF Working Papers 2018, 110; 10.5089/9781484354865.001.A001

Source: Tax Policy Reform Database, OECD, IBFD.

It is important to note that the average implementation lag and variation also reflect the retroactive introduction of tax measures, which is captured by negative implementation lags (Figure 11). The median implementation lag and variation for PIT and CIT measures appears to have declined since the onset of the global financial crisis, with a less pronounced reduction for CIT than for PIT measures. This likely reflects the urgency of implementing various tax policy measures in response to the crisis. At the same time, the median implementation lag and variation of VAT measures across countries appears to have increased after the global financial crisis.

Figure 12 presents information on implementation lags by tax type and country. Some countries (e.g., China and Turkey), were particularly effective in implementing new PIT and CIT reforms with an average implementation lag below 40 days, while others (e.g., Czech Republic, Germany) faced longer implementation delays, possibly reflecting the introduction of broader tax reforms that required time for discussion among various stakeholders (e.g., public consultations for changes in the tax code), and/or faced a complex legislative process. Figure 12 also suggests that the dispersion of implementation lags varies significantly across countries. Indeed, Australia, Germany, and Japan showed higher dispersion consistently across different tax types, while Japan, India, and Australia showed a significantly high variation in the implementation lags for the VAT.

Figure 12.
Figure 12.

Implementation Lags by Country and Tax Type

Citation: IMF Working Papers 2018, 110; 10.5089/9781484354865.001.A001

Source: Tax Policy Reform Database, OECD, IBFD.

The implementation lag can be used to differentiate between tax policy changes that are likely to be anticipated or unanticipated by economic agents. Following Mertens and Ravn (2012), any measure with an implementation lag equal to or of less than 90 days is considered as “unanticipated”, while any measure with a lag of more than 90 days is considered as “anticipated”. Based on this definition, Figure 13 shows that over 50 percent of CIT and PIT changes in the sample have been unanticipated. The corresponding number is 60 percent in the case of VAT changes. The prevalence of anticipated relative to unanticipated measures does not change significantly for tax rate or base measures, particularly for the CIT and PIT (Figure 14). By contrast, only ⅓ of all VAT rate changes were anticipated. In terms of the direction of change (i.e., increase/decrease in tax liability), Figure 15 indicates that almost 60 percent of tax decreases were unanticipated, while about 40 percent of tax increases were anticipated. Looking at the breakdown by tax changes, one in every two tax measures that increases tax liabilities was anticipated, while more than ⅔ of VAT decreases were unanticipated.

Figure 13.
Figure 13.

Anticipated and Unanticipated Measures by Tax Type

Citation: IMF Working Papers 2018, 110; 10.5089/9781484354865.001.A001

Source: Tax Policy Reform Database, OECD, IBFD.Note: Following Mertens and Ravn (2012), (un)anticipated measures occur when the difference between the announcement and implementation date is more (less) than 90 days.
Figure 14.
Figure 14.

Anticipated and Unanticipated Measures by Reform Type

Citation: IMF Working Papers 2018, 110; 10.5089/9781484354865.001.A001

Source: Tax Policy Reform Database, OECD, IBFD.Note: Following Mertens and Ravn (2012), (un)anticipated measures occur when the difference between the announcement and implementation date is more (less) than 90 days.
Figure 15.
Figure 15.

Anticipated and Unanticipated Measures by Direction of Change

Citation: IMF Working Papers 2018, 110; 10.5089/9781484354865.001.A001

Source: Tax Policy Reform Database, OECD, IBFD.Note: Following Mertens and Ravn (2012), (un)anticipated measures occur when the difference between the announcement and implementation date is more (less) than 90 days.

IV. Timing of Tax Measures

This section provides preliminary evidence on empirical regularities associated with the announcement of major tax reforms. We document the frequency and composition of tax policy reforms announced over specific periods that could influence policy decisions. These include periods of economic expansion versus recessions (where certain tax policy changes may become easier to implement)26; fiscal consolidation episodes where governments may be forced to implement less popular tax policy changes as opposed to “normal” fiscal times 27; and proximity to elections where political considerations may induce governments to announce popular tax policy measures28. The focus of this section is on major PIT, CIT, and VAT reforms, given greater availability of precise announcement dates for these measures.

Major tax policy changes appear to be clustered around specific periods, possibly reflecting waves of tax reforms. Most CIT and PIT measures were announced between the late 80s and early 2000s (Figure 16). For many European countries in the sample, this period corresponded with increasing economic integration (i.e., European Single Market, euro adoption). The frequency of major CIT and PIT reforms also increased prior to the onset of the global financial crisis. At the same time, major VAT policy changes are more evenly distributed across years (with the exception of the early 1990s).

Figure 16.
Figure 16.

Heat Map of Major Tax Measures by Tax and Reform Type

(count of tax changes)

Citation: IMF Working Papers 2018, 110; 10.5089/9781484354865.001.A001

Source: Tax Policy Reform Database, OECD, IBFD.

On average, countries in the sample announced 3.8 major CIT, PIT, and VAT changes per country year, half of these changes typically consisted of PIT measures (Table 6).

Table 6.

Characteristics of Major CIT, PIT, and VAT Measures

(average values unless otherwise indicated)

article image
Source: Tax Policy Reform Database, OECD, IBFD.

Decrease (increase) in tax liabilities refer to years in which there were only decreases (increases) in taxes.

Tax reforms were more frequent when economies were in a recession as opposed to an expansion (Table 6). However, the average number of tax measures announced in periods of economic expansion was slightly higher than in periods of recession. Moreover, Table 7 shows that decreases in tax liability occurred more often than increase in liabilities during expansions (41 versus 15 percent of country years29). Interestingly, during expansions, the average number of measures that increase tax liabilities was significantly lower (1.7 measures per country year) than the average number of measures that decrease tax liabilities (2.7 measures per country year), suggesting that tax policy could have had a pro-cyclical bias in the sample.

Table 7.

Major CIT, PIT, and VAT Measures in Recessions and Expansions

(count of country years)

article image
Source: Tax Policy Reform Database, OECD, IBFD.Note: Recessions are defined as years of negative growth. The sample comprises 22 countries and spans from 1970 to 2014 (see Table Appendix E.5 for the list of countries and years in the sample).

Major CIT, PIT, and VAT measures appear to be less frequent in consolidation periods than in normal times (Table 6). Consolidation periods show a more pronounced tax policy activism as the average number of major tax changes announced was 4.3 measures per country year, as opposed to 3.9 measures per country year during normal fiscal times. Not surprisingly, the average number of measures that increase tax liability is higher in consolidation periods than in normal times (respectively, 2.0 and 1.6 measures per country year). Table 8 further shows that during consolidation episodes decreases in tax liabilities happen in conjunction with increases in tax liability (55 percent of total country years). This suggests that governments may use policy measures that decrease tax liabilities as ‘sweeteners’ to buy political support for fiscal consolidation measures.

Table 8.

Major CIT, PIT, and VAT Measures in Consolidation and Normal Times

(count of country years)

article image
Source: Tax Policy Reform Database, OECD, IBFD.Note: Fiscal consolidation years are borrowed from Alesina and others (2015). The sample comprises 13 countries and spans from 1978 to 2014 (see Table Appendix E.5 for the list of countries and years in the sample).

Major tax measures were more common in the twelve months following an election (137 country years) than in the twelve months preceding an election 100 country years). In terms of the average number of measures, post-election years were characterized by 1.9 measures per country year as opposed to 2.7 measures per country year in pre-election years (Table 9). Interestingly, tax measures that increase the tax liability are more likely to occur in post-electoral periods (43 percent of 137 country years) than in pre-election years (15 percent of 100 country years). Moreover, decreases in tax liability (63 percent of country years) far outnumber increases in tax liabilities (15 percent of country years) in pre-electoral periods. These results suggest that incumbent governments want to avoid announcing unpopular tax measures before elections.

Table 9.

Major CIT, PIT, and VAT Measures in Pre- and Post-Electoral Years

(count of country year)

article image
Source: Tax Policy Reform Database, OECD, IBFD.Note: The proximity to elections is measured as twelve months before or after election. Data on election years are taken from the “Comparative Political Data Set” (CPDS) available at http://www.cpds-data.org/. The sample comprises 17 countries and spans from 1969 to 2014 (see Table Appendix E.5 for the list of countries and years in the sample).

V. Conclusions

This paper presents a novel database of tax policy measures (TPRD) that is unique in terms of its coverage, comprehensiveness, and granularity. The TPRD identifies tax policy changes in six types of taxes (i.e., CIT, PIT, VAT, SSC, EXE, and PRO) for 23 advanced and emerging countries. It classifies these changes according to several dimensions, including whether the tax measure resulted in a rate or base change, in an increase or decrease in tax liabilities, and whether it represented a major tax change. In addition, the database provides information on whether the measure was part of a broader tax package, phased over several years, and was announced in the context of a fiscal consolidation. Given the way it is constructed, future research can further improve and expand the database.

To demonstrate the usefulness of the database, the paper presents a range of novel stylized facts about tax policy that can motivate future research. We examine the anatomy of tax policy across countries and time. Our findings suggest that changes to the tax base are frequent and typically accompany rate changes, an aspect that is are often ignored in the literature. Second, tax policy measures are often taken as part of a broader reform packages, which is another dimension of tax policy for which little information has existed up to now. In most advanced and emerging economies, changes to PIT are most frequent, followed by changes to CIT and the VAT. The average implementation lag of PIT, CIT, and VAT measures is around 2–5 months, providing lead time for economic agents to adjust their behavior, but this differs across tax types. Finally, while these findings hold broadly across countries, there is significant cross-country heterogeneity on the nature and timing of tax reforms.

The paper also investigates whether the basic characteristics of tax policy decisions vary depending on the timing of such measures (e.g., in recessions, during fiscal consolidations, proximity to elections). We find that the number of announced tax policy changes differs significantly, tending to be markedly lower before elections. Moreover, the average number of measures that decrease tax liabilities during expansions are typically higher than during recessions, potentially suggesting a pro-cyclical bias to policymaking. During fiscal consolidations, governments tend to adopt measures that both increase and decrease tax liabilities. This suggests that governments may try to offset the adverse effects of implementing politically difficult tax measures. Finally, decreases in tax liability far outnumber increases in pre-electoral periods.

References

  • Alesina A., O. Barbiero, C. Favero, F. Giavazzi, and M. Paradisi, 2017, “The Effects of Fiscal Consolidations: Theory and Evidence,” NBER Working Papers 23385 (Cambridge, Massachusetts: National Bureau of Economic Research, Inc.).

    • Search Google Scholar
    • Export Citation
  • Alesina, A., C. Favero, and F. Giavazzi, 2015, “The Output Effect of Fiscal Consolidation Plans”, Journal of International Economics, Vol. 96, pp. S19S42.

    • Search Google Scholar
    • Export Citation
  • Auerbach, A. J. and Y. Gorodnichenko, 2012, “Measuring the Output Responses to Fiscal Policy,” American Economic Journal: Economic Policy, Vol. 4, pp. 127.

    • Search Google Scholar
    • Export Citation
  • Cloyne, J. S., 2013, “Discretionary Tax Changes and the Macroeconomy: New Narrative Evidence from the United Kingdom,” American Economic Review, 103, 4, 150728.

    • Search Google Scholar
    • Export Citation
  • Dabla-Norris, E., and F. Lima, Forthcoming, “The Macroeconomic Effects of Tax Changes: Evidence from Fiscal Consolidations, IMF Working Paper (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Duval, R., D. Furceri, J. Jalles, B. Hu and H. Nguyen, 2018, “A Narrative Database of Major Labor and Product Market Reforms in Advanced Economies,” IMF Working Paper No. 18/19 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Gentzkow, M., B. T. Kelly, and M. Taddy, 2017, “Text as Data,” NBER Working Papers 23276 (Cambridge, Massachusetts: National Bureau of Economic Research, Inc.).

    • Search Google Scholar
    • Export Citation
  • Gil, P., F. Martí, R. Morris, J. J. Pérez, and R. Rarnos, 2017, “The Output Effects of Tax Changes: Narrative Evidence from Spain,” Documentos de Trabajo No. 1721 (Madrid: Banco de España).

    • Search Google Scholar
    • Export Citation
  • IMF, Forthcoming, “Tax Policy Assessment Framework (TPAF)” (Washington: International Monetary Fund).

  • IMF, 2017, “IMF Fiscal Monitor: Achieving More with Less,” (Washington: International Monetary Fund).

  • IMF, 2016, “IMF Fiscal Monitor: Acting Now, Acting Together,” (Washington: International Monetary Fund).

  • Leeper, E. M., T. B. Walker, and S. S. Yang, 2013, “Fiscal Foresight and Information Flows,” Econometrica, Vol. 81, pp. 11151145.

    • Search Google Scholar
    • Export Citation
  • Kawano, L. and J. Slemrod, 2016, “How do Corporate Tax Bases Change when Corporate Tax Rates Change? With Implications for the Tax Rate Elasticity of Corporate Tax Revenues,” International Tax and Public Finance, Vol. 23, pp. 401433.

    • Search Google Scholar
    • Export Citation
  • Kremer, J., Braz, C. R., Brosens, T., Langenus, G., Momigliano, S. and Spolander, M., 2006, “A disaggregated framework for the analysis of structural developments in public finances,” ECB Working Paper Series , No. 579.

    • Search Google Scholar
    • Export Citation
  • Mertens, K. and M. O. Ravn, 2012, “Empirical Evidence on the Aggregate Effects of Anticipated and Unanticipated US Tax Policy Shocks,” American Economic Journal: Economic Policy, American Economic Association, Vol. 4, pp. 145181.

    • Search Google Scholar
    • Export Citation
  • OECD, 2006, “Fundamental Reform of Personal Income Tax”, OECD Tax Policy Studies No. 13, (Paris: Organization for Economic Co-operation and Development).

    • Search Google Scholar
    • Export Citation
  • OECD, 2010, “Tax Policy Reform and Economic Growth,” OECD Tax Policy Studies No. 20, (Paris: Organization for Economic Co-operation and Development).

    • Search Google Scholar
    • Export Citation
  • OECD, 2016a, “Consumption Tax Trends 2016: VAT/GST and Excise Rates, Trends and Policy Issues,” (Paris: Organization for Economic Co-operation and Development).

    • Search Google Scholar
    • Export Citation
  • OECD, 2016b, “Taxing Wages,” (Paris: Organization for Economic Co-operation and Development).

  • Pereira, M. C. and L. Wemans, 2013, “Output Effects of a Measure of Tax Shocks based on Changes in Legislation for Portugal,” Working Papers 15 (Lisbon: Banco de Portugal).

    • Search Google Scholar
    • Export Citation
  • Ramey, V. A., 2011, “Identifying Government Spending Shocks: It’s all in the Timing,” The Quarterly Journal of Economics, Oxford University Press, Vol. 126, pp. 150.

    • Search Google Scholar
    • Export Citation
  • Ramey, V. A. and M. D. Shapiro, 1998, “Costly Capital Reallocation and the Effects of Government Spending,” Carnegie-Rochester Conference Series on Public Policy, Vol. 48, pp. 145194.

    • Search Google Scholar
    • Export Citation
  • Ramey, V. A. and S. Zubairy, 2014, “Government Spending Multipliers in Good Times and in Bad: Evidence from U.S. Historical Data,” NBER Working Papers 20719 (Cambridge, Massachusetts: National Bureau of Economic Research, Inc.).

    • Search Google Scholar
    • Export Citation
  • Riera-Crichton, D., C. A. Vegh, and V. Guillermo, 2015, “Procyclical and Countercyclical Fiscal Multipliers: Evidence from OECD Countries,” Journal of International Money and Finance, Vol. 52, pp. 1531.

    • Search Google Scholar
    • Export Citation
  • Romer, C. D. and D. H. Romer, 2010, “The Macroeconomic Effects of Tax Changes: Estimates Based on a New Measure of Fiscal Shocks,” American Economic Review, Vol. 100, pp. 763801.

    • Search Google Scholar
    • Export Citation
  • Romer, C. D. and D. H. Romer, 2009, “A Narrative Analysis of Postwar Tax Changes,” (Berkeley: University of California, Berkeley).

  • Vegh, C. A. and G. Vuletin, 2015, “How is Tax Policy Conducted over the Business Cycle?American Economic Journal: Economic Policy, Vol. 7, pp. 32770.

    • Search Google Scholar
    • Export Citation

Appendix A. Data Source and Definitions

This appendix describes in detail the data sources used in the compilation of the TPRD, and the definitions of the variables.

Data Sources

Organization for Economic Co-operation and Development (OECD) and International Bureau of Fiscal Documentation (IBFD)

The database covers tax policy reforms in 23 countries over the last four decades. The countries covered in the database are Australia, Austria, Brazil, Canada, China, Czech Republic, Denmark, France, Greece, Germany, India, Ireland, Italy, Japan, Korea, Luxembourg, Mexico, Poland, Portugal, Spain, Turkey, United Kingdom, and the United States. The primary source of information was the OECD Economic Surveys. These surveys are publicly available for download from the OECD website. 30 The other primary source of information was the IBFD archives, in the form of news clips. The archives contain detailed tax information from 1988 onwards.31

Tax Rate Databases

To check the accuracy and coverage of our database, we performed various checks against external rate databases. We compared instances where our database recorded a major rate change to changes in tax rate levels from various internal and external sources. These are listed below.

  • IMF Tax Rate Database. The IMF Tax Rate Database is compiled internally, containing the statutory top rates for CIT, PIT, and VAT. The database covers the period 1980 to 2015.

  • European Commission Tax Indicator Database. This database contains information on top statutory CIT and PIT rates, standard and reduced VAT rates, labor-implicit tax rates (used for checks of SSC rates), and consumption-implicit tax rates (used for checks of EXE rates), and served as a secondary source for CIT, PIT, VAT, SSC, and EXE rates. The database covers the period 1995 to 2015.

  • Global KPMG Tax Rates. The KPMG database served as a secondary source of information for CIT, PIT, and VAT rates, and also contains information on the social security rates for employees and indirect rates. The database covers the period 2006 to 2015.

  • USAID Collecting Taxes Database. The USAID database was used as another supplementary source for CIT, PIT, and VAT rates. The database covers the period 2007 to 2012.

Other Tax Narratives

As another check for accuracy, we compared our database to other databases constructed using a narrative approach. The other narrative databases typically only covered tax changes in one country. The countries for which narrative databases were considered are United States (Romer and Romer, 2009), the United Kingdom (Cloyne, 2013), Spain (Gil and others, 2017), and Portugal (Pereira and Wemans, 2013). In a systematic manner, we verified whether all the major tax changes included in the narrative databases were also recorded in our database. Figure Appendix E.3 lists the countries and years for which possible information gaps have been identified.

Definitions

The database contains the following variables:

Sentence_OECD is a textual variable containing an excerpt from the OECD Surveys on OECD and non-OECD countries that mention changes in one or more of the six tax types covered in the database (i.e., PIT, CIT, VAT, SSC, EXE, and PRO). The quality of textual information extracted from OECD country report varies significantly: some textual fragments are very generic (e.g., the government reduced personal income taxation) while others provide details (e.g., the authorities increased the VAT standard rate from 15% to 20%). In addition, each excerpt can contain one or multiple tax changes.

Paragraph_OECDs is a textual variable containing the paragraph from which the excerpts included in the ‘Sentence_OECD’ variable were taken.

Tax_type is a categorical variable identifying which type of tax is discussed in the Sentence_OECD variable. The database covers six types of taxes that are defined following the Government Finance Statistics Manual (GFSM) 2014 (Table Appendix A.1). 32

Specifically:

  • PIT includes GFSM items 1111 (tax payable by individuals) and the portion of item 1113 (other taxes on income, profits, and capital gains) that refers to personal taxation.

  • CIT includes the GFSM items 1112 (taxes payable by corporations and other enterprises) and the portion of item 1113 (other taxes on income, profits, and capital gains) that refers to corporate taxation.

  • VAT includes the GFSM items 11411 (value-added taxes), 11412 (sales taxes), 11413 (turnover and other general taxes on goods and services).

  • EXE includes the GFSM item 1142 (excise)

  • SSC includes the GFSM items 112 (taxes on payroll and workforce) and 12 (social contributions).

  • PRO includes the GFSM item 113 (taxes on property).

Importantly, the above mention classification excludes the following GFSM tax items: 11414, 1143, 1144, 1145, 1146, 115, 116, 13, and 14. This reflects two considerations: namely that OECD Surveys do not consistently cover these items, as well as the decision to focus the analysis on major domestic taxes.

Tax_reform type is a categorical variable indicating whether the tax measure reported in Sentence_OECD refers to a rate (RATE) or base (BASE) change. Note that changes in per unit (i.e., specific) taxes are classified as rate changes.

Tax_change is a categorical variable reporting whether the tax measure described in Sentence_OECD entails an increase (INC) or decrease (DEC) in the rate or base.

Tax_major is a dummy variable taking on value 1 if the change is major. The following definitions of major rate and base changes are used in the database:

Major rate changes are identified when the rate changes by at least 1 percentage point in absolute terms (i.e., ΔFIATE ≥ 1pp) or when, in absence of quantitative information in the Sentence_OECD or Paragraph_OECD variables, the text describing the measures says that the change is major. In the case of per unit taxes, the determination on whether a change is major is based on an assessment of the language used in the Sentence_OECD or Paragraph_OECD variables.

Table Appendix A.1:

Summary Classification of Revenue According to GFSM 2014

article image
Source: GFSM 2014.