International Studies Program Working Paper 07-32 December 2007 Tax Morale and Tax Evasion in Latin America James Alm Jorge Martinez-Vazquez International Studies Program Working Paper 07-32 Tax Morale and Tax Evasion in Latin America James Alm Jorge Martinez-Vazquez December 2007 International Studies Program Andrew Young School of Policy Studies Georgia State University Atlanta, Georgia 30303 United States of America Phone: (404) 651-1144 Fax: (404) 651-4449 Email: [email protected] Internet: http://isp-aysps.gsu.edu Copyright 2006, the Andrew Young School of Policy Studies, Georgia State University. No part of the material protected by this copyright notice may be reproduced or utilized in any form or by any means without prior written permission from the copyright owner. International Studies Program Andrew Young School of Policy Studies The Andrew Young School of Policy Studies was established at Georgia State University with the objective of promoting excellence in the design, implementation, and evaluation of public policy. In addition to two academic departments (economics and public administration), the Andrew Young School houses seven leading research centers and policy programs, including the International Studies Program. The mission of the International Studies Program is to provide academic and professional training, applied research, and technical assistance in support of sound public policy and sustainable economic growth in developing and transitional economies. The International Studies Program at the Andrew Young School of Policy Studies is recognized worldwide for its efforts in support of economic and public policy reforms through technical assistance and training around the world. This reputation has been built serving a diverse client base, including the World Bank, the U.S. Agency for International Development (USAID), the United Nations Development Programme (UNDP), finance ministries, government organizations, legislative bodies and private sector institutions. The success of the International Studies Program reflects the breadth and depth of the in-house technical expertise that the International Studies Program can draw upon. The Andrew Young School's faculty are leading experts in economics and public policy and have authored books, published in major academic and technical journals, and have extensive experience in designing and implementing technical assistance and training programs. Andrew Young School faculty have been active in policy reform in over 40countries around the world. Our technical assistance strategy is not to merely provide technical prescriptions for policy reform, but to engage in a collaborative effort with the host government and donor agency to identify and analyze the issues at hand, arrive at policy solutions and implement reforms. The International Studies Program specializes in four broad policy areas:  Fiscal policy, including tax reforms, public expenditure reviews, tax administration reform  Fiscal decentralization, including fiscal decentralization reforms, design of intergovernmental transfer systems, urban government finance  Budgeting and fiscal management, including local government budgeting, performance- based budgeting, capital budgeting, multi-year budgeting  Economic analysis and revenue forecasting, including micro-simulation, time series forecasting, For more information about our technical assistance activities and training programs, please visit our website at http://isp-aysps.gsu.edu or contact us by email at [email protected]. Tax Morale and Tax Evasion in Latin America James Alm and Jorge Martinez-Vazquez Andrew Young School of Policy Studies Georgia State University Abstract It is hard for tax administrations to levy and collect taxes anywhere and any time. However, taxing certain kinds of activities, sectors, or individuals – the so-called “informal sector” – is an additional challenge for tax administrations in both developing and developed countries, and the “fiscal gap” that arises from the failure to tax this sector can be quite large. This issue is especially pressing in Latin America and Caribbean (LAC) countries, where often over half of the workforce is found in the informal sector. In this paper we examine taxation and tax compliance in LAC countries – and beyond – focusing on several main questions. What is meant by the “informal sector”? What is the size of informal sector in LAC countries? What are some effects from an informal sector, including the size of the “fiscal gap”? What are the reasons for this fiscal gap? What can be done to address these various issues? 1 1. Introduction It is well accepted that most people do not like to pay taxes, and, because of this fundamental reason, it is hard for tax administrations to levy and collect taxes anywhere and any time. However, taxing certain kinds of activities, sectors, or individuals – the so-called “informal sector” – is an additional challenge for tax administrations in both developing and developed countries, and the “fiscal gap” that arises from the failure to tax this sector can be quite large. Even aside from the collection of additional tax revenues from taxing those in the informal sector, there are other important tangible effects that arise from taxing the informal sector. One benefit is an improvement in horizontal and vertical equity. Another is an increase in economic efficiency. There are also significant intangible effects, including higher overall “tax morale” – or citizens’ intrinsic motivation to pay taxes – in the country. If there is a growing sense of the inability or unwillingness of tax authorities to catch tax evaders, the resulting unfairness of relative tax burdens could potentially lead over time to much lower tax yields than the lower tax yields due directly to the failure to tax this sector. This issue is especially pressing in Latin America and Caribbean (LAC) countries, where often over half of the workforce is found in the informal sector. In this paper we examine taxation and tax compliance in LAC countries and beyond, focusing on several main questions. What is meant by the “informal sector”? What is the size of informal sector in LAC countries? What are some effects from an informal sector, including the size of the “fiscal gap”? What are the reasons for this fiscal gap? What can be done to address these various issues? We begin with an overview of the tax systems of LAC countries. 2 2. Basic structural features of LAC tax systems The basic structural features of LAC tax systems, as represented by the structure of personal income taxes (PIT), corporate income taxes (CIT), and value-added taxes (VAT) or other general consumption taxes, are presented in Tables A-1 to A-3 in the Appendix. These tables show a great variety of approaches as to what is taxed, what is exempted, what are the rate structures, and so on. In the case of the PIT there are many quite varied approaches. For example, Bolivia has a flat rate tax of 13 percent , while Colombia has a multiplicity of progressive rates (132 in all) ranging from 0.26 percent to 35 percent and Chile similarly has a very progressive schedule with 8 brackets and minimum and maximum rates of 5 percent and 40 percent, respectively. In the area of exemptions, most forms of income from capital are exempt in Argentina (although Argentina imposes a wealth tax on gross asset values with rates ranging from 0.5 percent to 0.75 percent), while most other LAC countries tax capital income, even though realized capital gains are often either fully or partially exempt. Other forms of diversity in personal income taxation are provided by Mexico’s important low income tax credit, which no other country in the LAC region has, or the fact that several countries provide partial or total credit for VAT paid. There is more uniformity in the structure of the CIT, in income determination and allowable deductions, but even here the rates imposed range considerably, from 10 percent to 38.5 percent. In the case of the VAT, most LAC countries operate with a single rate, but again there are significant differences in rates, ranging from 5 percent in Panama to 23 percent in Uruguay. Although most countries zero-rate exports, there are wide differences in the scope of exempted commodities. Diversity in tax structure is accompanied also by diversity in typical tax processes (e.g., time spent preparing taxes, number of payments for tax purposes, and so on), the overall level of 3 taxation as percent of GDP, and the composition of tax revenues (e.g., the direct/indirect tax revenue shares). There is also diversity across LAC countries by size of GDP, its composition, and the level of GDP per capita. Information on these variables is presented in Appendix Table A-4. For example, in the number of payments that the average taxpayers has to make, the average for the region is 41, but it is as high as 68 in the case of Colombia and as low as 8 in the case of Ecuador. The average business spends 430 hours in filing and paying taxes, but again here there are wide variations, with 2600 hours in Brazil to 224 hours in El Salvador and 198 in Suriname. The average ratio of tax revenue to GDP in LAC countries is relatively low (as we discuss in more detail later), at 11.85 percent as an average in the period 1995-2005. This figure is fairly representative of the countries in the region, but there are some outliers. For example, the tax revenue to GDP ratio was as high as 17.80 percent in Uruguay and as low as 9.54 in Guatemala.1 On average, taxes on domestic good and services are twice more important, as a percent of total revenues, than taxes on income, profits, and capital gains. This relatively heavier reliance on indirect domestic taxes is likely to make the tax systems more regressive. Of course, there are again significant variations across countries. The share of indirect domestic taxation in total revenues is 58 percent in Mexico and 56 percent in Guatemala but as low as 9 percent in Panama. Many countries in the LAC region still rely quite significantly on taxes on international trade, which are likely associated with significant distortions in the allocation of productive resources. For example, Argentina received on average in the period 1995-2005 over 15 percent of its revenues from taxes on international trade. 1 Although the ratio of tax revenue to GDP changes over time for some countries, it is possible to divide the LAC countries into three categories of relatively high, intermediate, and relatively low ratios. Gomez Sabaini (2005) categorizes as relatively high ratio countries Brazil, Uruguay and Argentina; as relatively low, Paraguay, Mexico, Ecuador, Venezuela, Guatemala, and Haiti; and as intermediate all other countries in the region. 4 It is important also to emphasize the diversity in income levels across LAC countries. GDP per capita in Argentina on average over 1995-2005 was close to $12,000, while in Bolivia per capita GDP was $2,400 and was $3,210 in Nicaragua. However, both Bolivia and Nicaragua had during that period a significantly higher tax revenue to GDP ratio than Argentina. The composition of GDP, which likely affects the ability to raise taxes and the overall elasticity of tax revenues, also differed markedly across LAC countries; for example, value added from agriculture was only 6.18 percent in Chile and 7.02 percent in Argentina while it reached 23.16 in Guatemala and 24.20 percent in Paraguay. Given the existing tax structures as stated in the tax laws and economic environments of each of the countries in the LAC region, the actual level of tax revenues raised in each country depends on the ability and willingness to administer the existing taxes. There is some evidence that the overall effectiveness in using the existing tax structure differs significantly across LAC countries. For example, as shown in Table 1, the VAT productivity, defined as the yield of each percent point of the VAT rate as a share of GDP, ranges form a high of 0.64 in Chile to a low of 0.17 in Guatemala. Although some of the differences in productivity are probably linked to different structures of the VAT, especially the structure of exemptions, Chilean tax administration practices would appear to be over three times more effective than those in Guatemala in raising VAT revenues. Table 1. VAT Productivity for 17 Latin-American Countries VAT or GST Ratea VAT Productivity b Last Period of Available Country (2006, as percent) (as percent of GDP) Informationc Argentina 21% 35% 2004 Bolivia 15% 21% 2004 Brazil 15% 43% 2004 Chile 18% 64% 2004 Colombia 16% 26% 2003 Costa Rica 13% 38% 2004 Dominican Republic 12% 18% 2003 Ecuador 12% 54% 2003 5 Guatemala 12% 17% 2004 Honduras 12% 45% 2002 Mexico 15% 24% 2004 Nicaragua 15% 20% 2003 Panama 5% 30% 2003 Paraguay 10% 49% 2004 Peru 19% 35% 2003 Uruguay 23% 43% 2004 Venezuela 14% 47% 2004 a The rate corresponds to the highest rate defined in the law. Only Brazil has defined different VAT rates, between 10 percent and 15 percent. b “VAT Productivity” is defined as the yield of each percent point of the VAT rate, expresses as percent of GDP. c The most recent available information on VAT collections is limited to 2004 and, in some cases, 2003 and 2002. The GDP used to compute the VAT Productivity coefficient corresponds to the same period, but the VAT rate is the one in place on 2006. Some inaccuracies can be observed for those countries where the VAT rate has changed over the last years. Sources: For general consumption tax collections in 2002-2004 (current US$ millions), Centro Interamericano de Administraciones Tributarias (CIAT); for GDP in 2002-2004 (current US$ millions), World Marketing Data & Statistics; for the VAT or GST rate, PricewaterhouseCoopers’ Worldwide Tax Summaries online (December 2006). A conventional way to look at the performance of tax systems is to ask whether the country’s tax effort is “in line” with other countries of the same level of development and general economic characteristics. Although it is clear that there is no definitive way to establish how high taxes should be in a country, the comparison with international practice allows us to know how far a particular may be below or above the “international norm”. If the level of such “tax effort” is low relative to the international norm, this would be an indication that less than the adequate level of public services and infrastructure may be being provided and that tax effort could increase without appearing to be a “high tax” country to potential foreign direct investors.2 However, as we have seen repeatedly in the discussion above, different countries may differ in their ability to collect taxes because of different economic structures. To control for these difference, regression analysis is typically used to estimate the average capacity to collect taxes for a sample of countries, controlling for GDP per capita and other proxies for the ability to 2 Of course, there are many factors other than taxes that have been shown to affect foreign direct investment. The quality of a country’s governance institutions, low levels of corruption, the quality and skill levels of the labor force and infrastructure, and other factors have been shown to be as important, if not more important, determinants of foreign direct investment flows than taxes. Even so, many of these other determinants depend heavily on the ability of a country to generate adequate revenues. 6