OECD Taxation Working Papers, N. 36- Domestic resource mobilisation is critical to fund government services and to support development. Taxes are a critical domestic revenue source that can also impact other social or economic outcomes. Understanding differences in the level and structure of tax revenues is therefore foundational to discussions of domestic resource mobilisation and of tax reform. This paper presents evidence on the level and structure of tax revenues in 80 countries, drawing on the new Global Revenue Statistics Database. It compares tax-to-GDP ratios and tax structures across countries, regions and over time. Links between tax-to-GDP ratios, GDP per capita and tax structures are assessed in a correlation analysis. The new database provides invaluable insights for researchers and fiscal policy analysts and offers a high level of comparability and reliability. Introduction. Taxes are a critical component of government resources in almost all economies. Together with non-tax revenues, they support the role of the government in providing public services, re-distributing income and implementing other fiscal policy concerns, such as securing sustained growth and encouraging desirable socio-economic behaviour (Wahrig and Gancedo Vallina, 2011[1]). Relative to non-tax revenue sources, tax revenues are typically larger, more stable and less vulnerable to external shocks (OECD/AfDB/UNECA, 2010[2]; Te Velde, 2014[3]; European Parliament, 2014[4]; OECD, 2014[5]). As such, taxes are a critical input to governance and development. The way in which the government chooses to raise tax revenues (i.e. through the choice of taxes and their levels) itself also has social and economic implications. For these reasons, consideration of the level and structure of taxes in an economy is a critical first step in tax policy reform and in domestic revenue mobilisation. The Addis Tax Initiative (International Tax Compact, 2015[6]) recognises the importance of measuring domestic resource mobilisation. The Global Revenue Statistics Database contributes to this goal by providing detailed comparable tax revenue data for 80 countries over the period 1990 to 2016, as well as unweighted averages in the African and Latin America and the Caribbean (LAC) regions, and across the OECD.2 Together, these countries total nearly 60% of global GDP and provide a comprehensive basis for analysis across the four regions included in the analysis. 3 The database includes two key comparative indicators for all countries: the tax-to-GDP ratio and the tax structure (the share of a tax category in total tax revenue). The Global Revenue Statistics Database also provides tax revenue data in national currency and in USD. These data can be used to inform tax policy or administrative reforms within a country, and to conduct cross-country comparative analysis. Examining the policies and practices and historical trends in the data, and comparisons across similar economies, can provide an initial insight into potential avenues for tax policy or administration reforms. There is a great deal of heterogeneity across the countries included in this database and there has been a large amount of change in many countries in recent decades. A comparison of tax-to-GDP ratios for the most recently available year shows the diversity of tax revenue levels across countries and regions, ranging from 10.8% of GDP to 30.3% in Africa, 11.8% to 30.7% in Asia, 12.4% to 38.6% in Latin America and the Caribbean and 16.2% to 45.9% in the OECD. However, the range has narrowed considerably in recent decades, particularly as a result of movement at the lower end of the distribution: since 2000, nearly threequarters of these countries have increased their tax-to-GDP ratios and countries with lower levels of taxation have increased the most. Countries with larger increases tended to be in Africa, Latin America and the Caribbean, while OECD countries, whose ratios started generally at a higher level, experienced limited increases or decreases over this period. There is also a great deal of variation across regions and countries in their tax structures, i.e. the mix of taxes used to generate revenue. Across all countries, the three major sources of revenue are income and profit taxes, social security contributions (SSCs) and taxes on goods and services. Property taxes, payroll taxes and other taxes4 represent a more modest source of revenue; although each are used to widely different degrees in individual countries. The underlying Global Revenue Statistics Database is constructed using data from four Revenue Statistics publications: Revenue Statistics in Africa (OECD/ATAF/AUC, 2017[7]), Revenue Statistics in Asian Countries (OECD, 2017[8]), Revenue Statistics in Latin America & the Caribbean (OECD et al., 2018[9]) and Revenue Statistics in OECD countries (OECD, 2017[10]). These publications are prepared by the OECD in conjunction with a number of regional partners and with the financial support of the European Union (see Box 1 for further information). The database draws on the OECD classification of tax, which is common to all publications and harmonised with other major reporting standards including the System of National Accounts (European Commission et al., 2009[11]), the European System of Accounts (European Commission, 2010[12]), and the Government Finance Statistics Manual (International Monetary Fund, 2014[13]). All data are reported at the general level of government, which includes central, and sub-national and social security funds data5 .The database is updated four times a year as new data and country information become available. This paper provides an initial scan of the levels and structure of taxes for the 80 countries included in the database. It provides a comparative analysis for the most recent year and across time, taking into account the different regions covered by the database. The level and structure of tax revenues in these countries is determined by a multitude of underlying factors, including government expenditure commitments, income levels, international engagement, economic structure, tax morale, administrative capability as well as resource endowments: for example, oil-rich countries tend to have low tax-to-GDP ratios due to narrow tax bases and a high reliance on revenue from the oil sector. Understanding the impact of these factors on the revenue levels and structures of each country would provide a fruitful avenue for further research, building on key studies in the literature (Koenig and Wagener, 2012[14]; Tosun and Abizadeh, 2005[15]; Xing, 2011[16]), but is not covered in this paper. However, a preliminary analysis of correlations provides some initial observations on the relationship between GDP-per-capita, tax-to-GDP ratios and tax structures. This paper is structured as follows. Section 2 provides a brief overview of the methodology used to construct the database and its key indicators. Section 3 presents information on tax levels across the 80 countries included in the database, examining the levels of taxes in different regions and the changes across time. Section 4 examines tax structures across the 80 countries, drawing conclusions about the types of tax systems observed in the different countries. Section 5 undertakes preliminary analysis on the links between tax structure and tax-to-GDP levels, using a correlation analysis to provide initial observations and to suggest avenues for further research. Section 6 concludes. Modica, E., S. Laudage and M. Harding (2018).
