TAX POLICY IN ARAB COUNTRIES

Transcription

1 Distr. LIMITED E/ESCWA/SDD/2014/Technical Paper.3 16 June 2014 ORIGINAL: ENGLISH ECONOMIC AND SOCIAL COMMISSION FOR WESTERN ASIA (ESCWA) TAX POLICY IN ARAB COUNTRIES United Nations New York, 2014 Note: The views expressed in this paper are those of the authors and do not necessarily reflect the views of the United Nations Secretariat

2 Acknowledgments The paper was authored by Mr. Manuel Büsser under the guidance and supervision of Ms. Gisela Nauk, Chief, Social Policy Section, Social Policy Division of the Economic and Social Commission for Western Asia (ESCWA), and the overall guidance of Mr. Frederico Neto, Director, Social Development Division, ESCWA. The paper also benefitted from comments and support by Mr. Kenneth Iversen and Ms. Tanja Sejersen. iii

3 CONTENTS Page Acknowledgements... Abbreviations and acronyms... iii vii Chapter I. INTRODUCTION... 1 II. TAXATION AND FISCAL SPACE... 1 III. CHARACTERISTICS OF ESCWA MEMBER STATES... 2 IV. LEVELS OF TAX REVENUE... 6 V. COMPOSITION OF TAX REVENUE VI. TAX EFFORT ANALYSIS VII. CONCLUSION References LIST OF TABLES 1. Key economic indicators Key development indicators Government budget surplus/deficit Gross public debt Levels of tax revenue, Income versus consumption tax revenue PIT versus CIT revenue LIST OF FIGURES I. Fiscal diamond... 2 II. Tax revenue... 7 III. Levels of tax revenue... 8 IV. Composition of tax revenue in selected ESCWA member States V. Tax capacity VI. Tax effort ratio ANNEXES I. Tax revenue composition in selected ESCWA member States II. Regression output III. Regression dataset for selected Arab countries v

4 vi

5 ABBREVIATIONS AND ACRONYMS CIT ESCWA GCC GDP GST HDI HI ILO IMF LMI OECD PIT UAE UMI UNDP USAID VAT WDI corporate income tax United Nations Economic and Social Commission for Western Asia Gulf Cooperation Council gross domestic product general sales tax Human Development Index high income International Labour Organization International Monetary Fund lower middle income Organisation for Economic Co-operation and Development personal income tax United Arab Emirates upper middle income United Nations Development Programme United States Agency for International Development value-added tax World Development Indicators $ United States dollars vii

6 I. INTRODUCTION Arab countries currently face increasing demand for social protection and improved social services, but adjusting the prevalent welfare mix is difficult. Significant social policy changes will be required in the near future, possibly involving a reallocation of public expenditures and an increase in social spending. In the wake of the Arab uprisings, many Arab countries have increased social protection measures. However, social protection remains limited due to a high rate of informality and unemployment, as well as a low female participation rate in the labour market. 1 As a large share of the expansion of social spending happened on an ad hoc basis, many countries are running substantial deficits. 2 This paper advocates a sustainable long-term approach to ensure a social protection floor. 3 In doing so, it follows the initiative of the International Labour Organization (ILO) and the United Nations, which are promoting a social protection floor in Arab countries. 4 To fund such social spending, Governments have to build fiscal space. While there are different ways to do so, this paper focuses on revenue mobilization through taxation. The objective of the paper is to analyse the role of tax revenues within a fiscal space context in the Arab region. II. TAXATION AND FISCAL SPACE Fiscal space is a relatively new term, which remains vague. Definitions vary slightly: Mr. Heller, former deputy director of the International Monetary Fund (IMF) fiscal affairs department defined the term as the availability of budgetary room that allows a government to provide resources for a desired purpose without any prejudice to the sustainability of a government s financial position. 5 Such a purpose could for instance be the achievement of a specific development goal. A similar definition based on Mr. Heller s was used in a report by the World Bank and IMF in 2006: Fiscal space refers to a government s ability to undertake spending without impairing its solvency, that is without impairing its present and future ability to service its debt. 6 A study by the Economic and Social Commission for Western Asia (ESCWA) 7 refers to the definition used by the United Nations Development Programme (UNDP): Fiscal space is the financing that is available to government as a result of concrete policy actions for enhancing resource mobilization, and the reforms necessary to secure the enabling governance, institutional and economic environment for these policy actions to be effective, for a specified set of development objectives. 8 Following this definition, fiscal space is sustainable if a country does not rely on external financing. This highlights the long-term approach of fiscal space, which encourages growth and poverty reduction in the long run. The fiscal diamond is a stylized illustration of the fiscal space concept (figure I). It is composed of four pillars: domestic revenue mobilization, reprioritization (expenditure shifting), deficit financing and official development assistance. All four indicators are measured relatively to the gross domestic product (GDP) of the country. 1 ILO, 2012, pp ESCWA, 2012b, pp The social protection floor [ ] is an integrated set of social policies designed to guarantee income security and access to social services for all, paying particular attention to vulnerable groups, and protecting and empowering people across the life cycle. ILO, 2011, p ILO, 2009, p. 7; ESCWA, 2013, p Heller, 2005, p Development Committee, 2006, p ESCWA, 2012a, p Roy and others, 2007, p. 2.

7 Theoretically, new revenues can be raised through any of these four pillars. In certain countries, however, some policy options may not be as feasible as others. Hence, the application of fiscal space in reality is complicated and must be underpinned by sound analysis. 9 Figure I. Fiscal diamond Domestic revenue mobilization Deficit financing Reprioritization and efficiency of expenditure Official development assistance Tax revenues fall under the first pillar (domestic revenue mobilization). They can be increased by raising the current tax rates, introducing new taxes or by expanding the tax base. Within the Arab region, the importance of tax revenues to public finance varies significantly. While in non-oil exporting countries tax revenues account for a large share of total government revenues, oil-exporting countries finance a very low share of their expenditures by taxes. It becomes apparent that Governments mobilize extensive resources through taxation. As a result, tax revenue is vital to the fiscal space discussion. III. CHARACTERISTICS OF ESCWA MEMBER STATES The Arab region is characterized by the economic and social diversity of its countries. Many derive a significant share of income from the export of oil-related products; namely the Gulf Cooperation Council (GCC) countries, Iraq and Libya. In the GCC region, oil rents account for between 20 and 55 per cent of GDP. Income levels are high and overall unemployment rates relatively low. Economically, the non-gcc countries are less developed and fall under the upper or lower middle-income category. In the wake of the global economic crisis and social unrest, many of those countries face severe economic and social challenges. Currently, Egypt, the Sudan and Yemen show low real GDP growth and high inflation, and data suggest low employment. As a result of conflict, the economies of Libya and the Syrian Arab Republic are still in very unstable condition. From a social point of view, there is significant scope for further development. The Sudan and Yemen show low human development as measured by the Human Development Index (HDI), while Egypt, Iraq, Morocco and Palestine show medium human development. The Syrian Arab Republic, currently in a state of conflict, is another lower middle-income country with rather low social development. Income inequalities within countries are not as high as in Latin America, where Gini coefficients can reach 50; 10 however GCC countries fall behind countries from the high income category in other world regions in terms of social development. As the economic situation is unsatisfactory in many Arab countries, tensions within the population are still rising. These circumstances are particularly challenging as Governments fail to mobilize resources to provide much needed social security. 9 Morisse (2013) presents a literature review on fiscal space in the Arab region. 10 UNDP, 2013, pp

8 GDP per capita, in current $ (2012) TABLE 1. KEY ECONOMIC INDICATORS Real GDP growth (2012) 3 Unemployment rate (2012) Oil rents, percentage of GDP (2011) Oil-exporting countries Inflation (2012) Bahrain a/ Iraq Kuwait Libya b/ Oman Qatar Saudi Arabia Sudan Syria a/ UAE Yemen Ecuador Nigeria a/ 32.9 Norway Non-oil-exporting countries Egypt Jordan Lebanon Morocco Palestine c/ 12.2 c/ 1.7 c/ Tunisia France Ghana Peru Turkey Source: IMF, World Economic Outlook database; Palestinian Central Bureau of Statistics; World Bank, World Development Indicators (WDI) database. Note: Two dots (..) indicate that data are not available. a/ Data are for b/ Data are for c/ Data are for TABLE 2. KEY DEVELOPMENT INDICATORS Oil-exporting countries HDI value (2012) HDI rank (2012) Income Gini Index Income level Bahrain HI Iraq a/ UMI Kuwait HI Libya UMI Oman HI Qatar a/ HI Saudi Arabia HI Sudan b/ LMI

9 TABLE 2 (continued) Oil-exporting countries HDI value (2012) HDI rank (2012) Income Gini Index Income level Syria LMI Tunisia c/ UMI UAE HI Yemen c/ LMI Ecuador UMI Nigeria LMI Norway HI Non-oil-exporting countries Egypt d/ LMI Jordan e/ UMI Lebanon UMI Morocco a/ LMI Palestine b/ LMI France HI Ghana LMI Peru UMI Turkey UMI Abbreviations: LMI, lower middle income (1,036 United States dollars ($) - $4,085 per capita); UMI, upper middle income ($4,086 - $12,615 per capita); HI, high income ($12,616 or more per capita). Source: UNDP, 2013, pp , Note: Two dots (..) indicate that data are not available. a/ Data are for b/ Data are for c/ Data are for d/ Data are for e/ Data are for As stated earlier, fiscal space is the room to manoeuvre which a country possesses to mobilize resources for a specific purpose. Many different factors should be analysed to draw an adequate picture of a country s fiscal space, including budget balances. The fiscal constraints of a country or the extent by how much revenue exceeds expenditure determine short-term fiscal space. Before the Arab uprisings, several countries faced challenging fiscal constraints but were able to keep the budget deficit at a sustainable level. In the last three years, however, public finances in the region have been negatively affected. The unrest put a strain on growth rates which impacted government revenues. At the same time, increased social spending as a response to the upheaval worsened budget deficits. Even in countries that were not severely affected by social unrest, discretionary social spending measures were introduced to stem the fear of contagion. The favoured polices included hand-outs increased public work programmes and increased subsidies. Many of the policies adopted in 2011 will be difficult to reverse in the short term. 11 As a result, the fiscal outlook in the region will be characterized by high social spending in the coming years. Currently the actions taken by Governments are based on short-term relief. 12 It is vital that they shift to a solution that is sustainable in the long run and focuses on social security and on implementing a social protection floor rather than on immediate emergency response. Additional resources should be derived without external financing. This could be achieved for instance by increasing tax revenues to ensure long-term fiscal space. 11 ESCWA, 2012a, p ILO, 2012, p

10 TABLE 3. GOVERNMENT BUDGET SURPLUS/DEFICIT (As a share of GDP) County Average Bahrain Egypt Iraq Jordan Kuwait Lebanon Libya Morocco Oman Qatar Saudi Arabia Sudan Syria Tunisia UAE Yemen Average Source: IMF, World Economic Outlook database. Note: two dots (..) indicate that data are not available. In Egypt and Tunisia, the uprisings directly impacted the government budget. 13 Bahrain, Jordan, Lebanon, Morocco, the Sudan and Yemen also show very high budget deficits which suggests low fiscal space in the short term. Many of the oil-exporting countries find themselves in more comfortable fiscal situations. GCC countries (except for Bahrain) as well as Libya and Iraq have achieved remarkable budget surpluses. Since oil exports tend to fluctuate by a wide margin, non-oil revenues compared to government expenditure show by how far expenditures are offset by more permanent revenues. With many of the Arab countries highly involved in the export of oil, this issue is of particular interest. In the case of Kuwait and the United Arab Emirates, non-oil revenues are not sufficient to sustain a positive balance. 14 The high dependence on oil revenues presents a possible risk and links fiscal space to oil prices. Ultimately, large budget deficits result in high government debt levels. Egypt, Jordan, Lebanon and the Sudan hover around levels above 80 per cent of GDP (table 4). Oil-exporting countries show debts which are much more sustainable and low compared to international levels. TABLE 4. GROSS PUBLIC DEBT (As a share of GDP) County Bahrain Egypt Iraq Jordan Kuwait Lebanon ESCWA, 2012a, p Ibid., p

11 TABLE 4 (continued) County Libya Morocco Oman Qatar Saudi Arabia Sudan Syria Tunisia UAE Yemen Source: IMF, World Economic Outlook database. Note: Two dots (..) indicate that data are not available. IV. LEVELS OF TAX REVENUE To determine a country s fiscal space one must look into several indicators, one of which is tax revenue. For most Governments, revenue from taxes is of highest importance. Besides taxation, income may also be derived from fees or property income, such as revenues from government-owned companies. On a world average, however, 62 per cent of government revenues stem from taxation. 15 While low-income countries around the world depend to a higher extent on tax revenues, high-income countries have a higher share from other sources. On average, middle-income countries derive 73 per cent of revenue from taxation. 16 With only 37 per cent (arithmetic average), the Arab region shows relatively little dependence on tax revenues. It is important to note that this average is a broad generalization. One must go into more detail and look at the data on a country level. The Arab region is very diverse and the variation around the region s average is high. Tax revenues range from 1 per cent in Kuwait to 89 per cent in Palestine (figure II). In GCC oil-exporting countries, taxation plays a minor role in the government budget. These countries rely highly on revenue from oil and gas exports. Yemen shows substantial oil-related revenue as well, which comes along with relatively low tax income. Other oil-exporting countries such as the Sudan (before partition) and the Syrian Arab Republic derive approximately half of their revenue from taxation. State budgets in Jordan, Egypt, Lebanon, Tunisia, Morocco and Palestine are mostly based on tax income. Overall, the reliance on tax revenue in the region (40 per cent on average) seems to be relatively low compared to that of other middle-income countries around the world. In order to assess the relevance of taxation to a country s economy as whole, tax revenues are compared to GDP. According to World Bank, the world average tax revenue was at 14.6 per cent of GDP in While in low income and lower middle-income countries, levels between 10 and 15 per cent can be found, certain developed countries have overall tax revenue around 30 per cent. 18 Upper middle-income countries usually exceed 20 per cent. Empirical evidence indicates a positive correlation between tax levels and GDP World Bank, WDI database. 16 Ibid. 17 Ibid. 18 Ibid. 19 Bahl, 1971, p

12 Figure II. Tax revenue (As a share of government revenue) 100% 80% 71% 72% 76% 80% 88% 89% 88% 73% 60% 51% 52% 60% 61% 40% 43% 40% 23% 20% 0% 1% 2% 3% 6% 6% 6% 7% Source: Bahrain: Ministry of Finance; Egypt: Ministry of Finance; Iraq: Central Bank of Iraq; Jordan: World Bank data; Kuwait: World Bank data; Lebanon: Ministry of Finance; Libya: IMF data; Morocco: Ministry of Finance; Oman: National Centre for Statistics and Information; Palestine: Ministry of Finance; Qatar: World Bank data; Saudi Arabia: IMF data; the Sudan: Central Bank of the Sudan; Syria: Central Bank of the Syrian Arab Republic; Tunisia: Ministry of Finance; United Arab Emirates (UAE): Central Bank of the United Arab Emirates; Yemen: Central Statistical Organisation; low-income countries, middle-income countries, high-income countries and OECD (Organisation for Economic Co-operation and Development members): World Bank data. Note: Value for Bahrain includes taxes and fees. In the Arab region the figure varies by a wide margin across countries (figure III). While the arithmetic average is in the range of 10 per cent, countries are either clearly above or below this value. Even though there is empirical evidence for a positive correlation of GDP and tax level, the region does not necessarily support this hypothesis. With the exception of Qatar, GCC countries, which are the countries with the highest income within the region, have tax burdens below 2 per cent of GDP. In fact, they are among the countries with the lowest tax revenue levels in the world. Other oil-exporting countries seem to follow this trend: Iraq, Libya, Yemen and the Sudan are found far below the 10 per cent average. Yemen and the Sudan, the two poorest countries of the region, show a tax level far lower than that of the region s lower middle-income countries such as the Syrian Arab Republic, Egypt, Palestine and Morocco. With over 20 per cent, Tunisia and Morocco hover around levels of middle-income and certain developed countries such as France (21.3 per cent) and Sweden (22.1). 20 Jordan and Lebanon are around the 15 per cent world average. In Table 5, Qatar stands out as tax revenues are clearly higher than in all other GCC countries. This stems from the fact that the oil sector contributes to a large share to taxation revenue, while in the other GCC countries oil revenue is reported as profits from their state-owned oil companies. Therefore, in Qatar, the total level of tax revenue is much higher. Non-oil taxation revenue in Qatar however was only at 5.1 per cent as a share of GDP in 2011/ World Bank, WDI database. 21 IMF, 2013, p

13 Figure III. Levels of tax revenue (As a share of GDP) 25% 20% 15% 10% 5% 0% 0.8% 1.1% 1.3% 1.5% 1.9% 2.5% 3.8% 6.7% 6.7% 20.8% 21.0% 17.8% 14.7% 15.3% 15.7% 15.9% 11.8% 9.4% Source: Bahrain: Ministry of Finance; Egypt: Ministry of Finance; Iraq: Central Bank of Iraq; Jordan: World Bank data; Kuwait: World Bank data; Lebanon: Ministry of Finance; Libya: IMF data; Morocco: Ministry of Finance; Oman: National Centre for Statistics and Information; Palestine: Ministry of Finance; Qatar: World Bank data; Saudi Arabia: IMF data; the Sudan: Central Bank of the Sudan; Syria: Central Bank of the Syrian Arab Republic; Tunisia: Ministry of Finance; United Arab Emirates (UAE): Central Bank of the United Arab Emirates; Yemen: Central Statistical Organisation. GDP data are derived from World Bank sources, except for Libya (IMF data) and Palestine (Palestine Monetary Authority). Note: Value for Bahrain includes taxes and fees. It seems that there is significant room for additional taxation revenue in the Arab region. According to Heller, a tax revenue level of 15 per cent of GDP should be the minimal goal for any developing country. 22 Eleven of the 17 ESCWA member States are currently below that level. Furthermore, most are below levels of other middle-income countries. In comparison with Latin American and Southeast Asian countries, Arab countries fall behind. 23 From a social development perspective, the question is: are those levels adequate to meet minimal expenditure on social security, i.e. a social protection floor? To achieve the Millennium Development Goals, the United Nations suggests tax levels to be above 20 per cent of GDP. 24 Considering those recommendations, most ESCWA member States should significantly ramp up their tax collection. Nevertheless, it is difficult to evaluate which States do in fact possess the potential to increase tax revenues. Tax systems are designed for long-term well-being and reforms should aim to achieve the same objective. Variations in the level of tax revenues might not be feasible from an efficiency standpoint. However, this does not mean that tax systems should be rigid. On the contrary, the goals should be to build a dynamic system which allows adjustments and reforms over time since tax systems are challenged with the changing needs of a developing country. As the demand for government expenditure and services rises with the overall development of a country, revenue mobilization must follow. Therefore tax revenues should increase vis-à-vis GDP, and the elasticity of income and expenditure should correspond. Ideally the taxrevenue-to-gdp ratio remains constant. A decrease over time is seen as a lack of tax reforms, in which case tax policies should be adjusted. 22 Heller, 2005, pp Abu-Ismail and others, 2012, p UNDP, 2010, p

14 TABLE 5. LEVELS OF TAX REVENUE, (As a share of GDP) Country Average Bahrain Egypt Iraq Jordan Kuwait Lebanon Libya Morocco Oman Palestine Qatar Saudi Arabia Sudan Syria Tunisia UAE Yemen Source: Bahrain: World Bank data ( ) and Ministry of Finance ( ); Egypt: Ministry of Finance; Iraq: Central Bank of Iraq; Jordan: World Bank data; Kuwait: World Bank data; Lebanon: Ministry of Finance; Libya: IMF data; Morocco: Ministry of Finance; Oman: Central Bank of Oman; Palestine: Palestine Monetary Authority ( ) and Ministry of Finance ( ); Qatar: World Bank data; Saudi Arabia: IMF data; the Sudan: Central Bank of the Sudan; Syria: Central Bank of the Syrian Arab Republic; Tunisia: World Bank data ( ) and Ministry of Finance ( ); United Arab Emirates (UAE): Central Bank of the United Arab Emirates; Yemen: Central Statistical Organisation. GDP data are derived from World Bank sources, except for Libya (IMF data) and Palestine (Palestine Monetary Authority). Note: Two dots (..) indicate that data are not available. Bahrain data ( ) include taxes and fees. As made visible in table 5, a number of countries show heavy fluctuations. In Jordan, for instance, despite a number of reforms, the tax level has decreased from well above 20 per cent in 2007 to 15 per cent in Other countries with high variation are Lebanon, Morocco, Yemen, Palestine and Qatar. In Palestine a significant share of tax revenues is collected by Israeli authorities. Therefore the fluctuations of reported tax revenues can partly be attributed to the heavy dependence on tax transfer from Israel. 26 In the past such transfers have been withheld on several occasions. 27 Fairly stable tax-to-gdp ratios are found for GCC countries with very low tax burden. Tunisia seems to successfully keep tax revenues in line with GDP. 25 Iversen, 2012, p See Sabri and Jaber, 2006 for more details about the Palestinian tax system. 27 Reuters,

15 V. COMPOSITION OF TAX REVENUE As taxes can be derived from many different sources, it is inevitable to disaggregate total tax revenue. Taxes can broadly be divided into direct, indirect and international taxes. Direct taxes are paid directly to the tax authorities. Taxes on income, capital gains, profits and property fall under this category. Indirect taxes are collected through an intermediary. They consist of taxes on goods and services, such as value-added tax (VAT), general sales tax (GST) and excises. In other words, direct and indirect taxes stand for income and consumption taxes respectively. Taxes on international trade stem from the taxation of imports and exports (tariffs). It is of particular interest to analyse the ratio of income to consumption taxes, as this takes into account efficiency as well as equity issues. In academic literature, taxation of income is seen as the more inefficient way to levy taxes. The reduction of savings leads to a higher economic deadweight loss than taxation on consumption. 28 From an equity point of view, however, income taxes are favourable because a progressive tax structure can balance unequal income distributions. Addressing inequality is a key aspect to sustainable economic and social development. Policymakers should trade off efficiency and equality aspects. Tax revenues are exposed to shocks and variation in economic activity. It is therefore advisable not to depend on just one kind of tax income but to diversify revenue sources. 29 Many developing countries however show reliance on consumption taxes such as VAT. The income-consumption tax ratio impacts equity. To a certain degree, taxation can contribute to a more equitable society. This aspect should be given particular relevance, as income is unequally distributed in developing countries. Traditional consumption taxes such as VAT impose an equal burden on buyers of a good. Nevertheless, empirical studies have proven that consumption taxes are slightly regressive. 30 While the poor consume most of their income, the rich spend a lower share for consumption goods, meaning that they are less affected by consumption taxes. Income taxes are often progressive in nature. Still, this does not mean that substantial transfers from the rich to the poor are the norm. In the Andean countries for instance, the distributional impact is fairly limited due to the poor collection of income taxes. 31 This result is supported by findings from Central America where the redistributive effect of progressive and regressive taxation was found to be weak. 32 To assess if a tax system is pro-poor, one also must take into account the redistributive capacity of public social expenditure. One should not only consider how tax money is collected but also how those revenues are distributed among the population. Targeted social spending might balance a disproportionately large tax burden on the poor. In countries with very low social security standards, however, there is a lot of potential for improving social protection. Nevertheless, when isolating the composition of tax revenue, it seems appropriate to state that regressive taxation is unfavourable to the vulnerable. A large share of indirect taxes is therefore a burden to the ones in need of social protection. In the composition of tax revenue as a share of GDP in ESCWA member States (table 6), there is a typical pattern found in many developing countries: The tax revenue is to a large extent dependent on consumption taxes. On average, countries that levy consumption taxes derive around 50 per cent of their tax revenue from this type of taxation. In the Arab region, countries with a consumption tax-based system tend to also be the ones with higher overall tax levels. They might struggle to increase the tax burden on income, not least of all because of political reasons. Furthermore, informality is an issue. Indirect taxation is more efficient to widen the tax base. This leads to even more concentration on consumption taxes. From an equity stand point, this is worrisome and limits the redistributive effect. Over the past 10 years, several 28 Tanzi and Zee, 2000, p Bird and Zolt, 2003, p Bird and Gendron, 2005, pp Barreix and others, 2007, pp Cubero and Vladkova Hollar, 2010, p

16 countries have shifted from a more direct to an increasingly indirect taxation system. The Sudan, Lebanon, Jordan and Yemen are among the countries where this trend has been the most significant. Jordan and Palestine show levels of indirect taxation which hover around 65 per cent of total tax revenue. This excessive focus on indirect taxes does not favour the poor. GCC countries, Libya and the Syrian Arab Republic however mainly focus on income and trade taxes. Most of those States have not introduced VAT. Since tax levels in those countries are very low, the potential of redistributing resources is marginal. The high income tax level in Qatar is to a large extent contributed by the oil sector. Revenue from international trade taxes in the region hovers around 1 per cent of GDP. In the long run, increasing trade liberalization will likely put downward pressure on the tax rates levied on imports and exports. TABLE 6. INCOME VERSUS CONSUMPTION TAX REVENUE (As a share of GDP) Country Year Total tax revenue Income tax Consumption tax International tax Income/consumption tax Bahrain Egypt Iraq Jordan Kuwait Lebanon Libya Morocco Oman Palestine Qatar Saudi Arabia Sudan Syria Tunisia UAE Yemen France Peru Thailand Turkey Source: Bahrain, Egypt, Jordan, Kuwait, Lebanon, Morocco, Oman, Qatar, France, Peru, Thailand and Turkey: World Bank data; Iraq: Central Bank of Iraq; Libya: IMF data; Palestine: Ministry of Finance; Saudi Arabia: IMF data; the Sudan: Central Bank of the Sudan; Syria: Central Bank of the Syrian Arab Republic; Tunisia: Ministry of Finance; United Arab Emirates (UAE): Central Bank of the United Arab Emirates; Yemen: Central Statistical Organization. GDP data are derived from World Bank sources, except for Libya (IMF data) and Palestine (Palestine Monetary Authority). Note: Two dots (..) indicate that data are not available. Total tax revenue data for Bahrain include taxes and fees. Income taxes can be divided into personal income taxes (PIT) and corporate income taxes (CIT), i.e. taxation on companies. The ratio of CIT to PIT is crucial to any taxation system. In developed countries, about four times as many revenues are derived from PIT than from CIT. 33 Income from salaries as a share of national income is rather low in many developing countries and therefore taxation of personal income might not have the same potential as in developed countries. However, there are other reasons such as an inefficient tax administration hindering efficient taxation of the highest incomes. 33 Tanzi and Zee, 2000, p

17 In many ESCWA member States there is not sufficient tax data that divides income tax revenues into PIT and CIT. Hence a comparison among all countries is not feasible. In table 7, data are presented for seven selected countries. TABLE 7. PIT VERSUS CIT REVENUE (As a share of GDP) Country Year Income tax PIT CIT CIT/PIT Egypt 2012/ Jordan Lebanon Morocco Saudi Arabia >10 Sudan Tunisia Source: Egypt, Jordan, Lebanon, Morocco and Tunisia: ministries of Finance; Saudi Arabia: IMF data; the Sudan: Central Bank of the Sudan. GDP data are derived from World Bank sources. The ratio of CIT to PIT in Middle Eastern countries 34 was 4.3 on average in Ten years later this average had decreased to The latest available data show that there are significant differences among countries. There are broadly three different country groups: countries with an equal share of PIT and CIT; countries with a low share of PIT; and countries with a very low (or inexistent) share of PIT. In Qatar, non-hydrocarbon CIT revenue amounted to 5.1 per cent as a share of GDP in 2011/ Compared to other ESCWA countries this is relatively high and exceptional to other GCC countries. Saudi Arabia is a representative example for the other GCC countries, which have either none or very low taxes on personal income and therefore show a very high ratio of CIT to PIT. Even though significant revenue could be derived from that source one cannot blame this on inefficiency. In the case of those countries the low burden on personal income is desired by Governments and hence implemented in the taxation system. Morocco and Tunisia show very high levels of income taxation and, compared to other countries in the region, also high levels of taxation on personal income. As a share of GDP, revenues from PIT alone are higher than total income taxes in most other countries of the region. Consequently, an equal share is derived from PIT and CIT. While there is still further potential, those two countries seem to derive a relatively large share of income taxation from wages. Egypt and Jordan represent another group. Even though their tax systems aim to tax income and namely personal income, the share of PIT is low. This suggests that tax rates on wages could be increased significantly. 34 The term Middle East refers to the definition of IMF of the Middle East region, which might include non-arab countries. 35 Ibid., p IMF, 2013, p

18 Figure IV. Composition of tax revenue in selected ESCWA member States (As a share of total tax revenue) (A) 100% 90% 80% Other taxes Fiscal stamp Customs 70% Excises 60% GST 50% VAT 40% 30% 20% 10% Taxes on goods & services Property Tax Income tax: capital gains Income tax: corporate Income tax: personal 0% Egypt (2012/13) Jordan (2011) Lebanon (2012) Morocco (2012) Palestine (2013) Sudan (2010) Tunisia (2012) Yemen (2012) Income tax: total Source: Egypt, Jordan, Lebanon, Morocco, Palestine and Tunisia: ministries of finance; the Sudan: Central Bank of the Sudan; Yemen: Central Statistical Organization. (B) 100% 90% Other taxes 80% 70% Customs 60% 50% Taxes on goods & services 40% 30% 20% Property Tax 10% 0% Iraq (2006) Kuwait (2010/11) Libya (2008) Oman (2012) Qatar (2010) Saudi Arabia (2012) Syria (2010)UAE (2007) Income tax Source: Iraq: Central Bank of Iraq; Kuwait: Central Bank of Kuwait; Libya: IMF data; Oman: National Centre for Statistics and Information; Qatar: World Bank data; Saudi Arabia: IMF data; Syria: Central Bank of the Syrian Arab Republic; United Arab Emirates (UAE): Central Bank of the United Arab Emirates. VI. TAX EFFORT ANALYSIS The following section will examine the country-specific potential to increase revenue from taxation. The question is how much taxes should be levied or, in other words, what is the optimal level of taxation. The term optimal may be interpreted in different ways: if one only considers macroeconomic matters, the optimal tax level might be lower than if one also considers social welfare issues. 13

19 From a theoretical point of view, little guidance is found about optimal taxation levels. This study follows the framework applied by several researchers using international comparisons. 37 Analyses solely based on tax revenue as a share of GDP might be too simplistic. Country-specific characteristics should be considered. For instance, it may not be possible to expect the same level of taxation for Cambodia (10.0 per cent of GDP) and Norway (28.1 per cent of GDP) 38 as the two countries are at completely different stages of development. Each country s tax capacity is estimated taking into account certain country-specific characteristics. The tax capacity is subsequently compared with the effective taxation level, which gives an indication of the so-called tax effort a measure for a country s effort to collect taxes. As stated by Pessino and Fenochietto, it is better to speak of effort rather than efficiency, as certain countries may choose to have a low level of taxation but still be efficient in collecting taxes. 39 The data set, featured in annex III, covers ESCWA member States between 2000 and The main source was the WDI database. Certain data series were complemented with data from national sources. The following regression was estimated: 40 = The four explanatory variables used to estimate the tax levels are the natural logarithm of GDP per capita (Y); the value added by the agricultural sector (Ag); primary exports (fuel metal and other raw materials) as a share of merchandising exports (Rent); and money and quasi money as share of GDP (M). Empirical studies indicate that economic development and tax levels are positively correlated. As the demand for government spending increases with economic development, tax levels should be increased to cover the increased expenses. 41 When it comes to taxation, the informal sector falls through the cracks and lowers a country s tax capacity. It is usually hard to levy taxes on farmers which is why Ag can be seen as a proxy for informality. In GCC countries, agriculture plays a minor role unlike in other ESCWA member States. In the Arab region in general, informality is a major factor that lowers the tax base. The variable Rent takes into account the sector composition of an economy. M indicates the degree of monetization. M2 (money and quasi money) increases with the development of the financial sector and therefore widens the tax base. Y and M are expected to be positively correlated with the tax level, while Ag and Rent are anticipated to have a negative influence. A detailed output of the regression is found in Annex All described indicators are statistically significant and the respective influence on the regression is as expected. To tackle the problem of informality, other indicators were included. The variable urban population as a share of total population takes into account the urbanization of countries. It was expected that countries with a higher share of rural population would show a lower tax revenue level, as rural areas are generally subject to more informality. This indicator however was not significant in the regression. The share of selfemployed is another indicator testing for informality that turned out insignificant. To account for the openness of an economy, the variable trade as a share of GDP was added to the model. Countries with a high share of imports and exports are expected to levy more taxes through international taxation such as customs. The 37 See Bahl, 1971; Eltony, 2002; and Gupta, World Bank, WDI database. 39 Pessino and Fenochietto, 2010, p The model applied here is the same as the one used in Iversen, 2012, p Tanzi and Zee, 2000, p The same regression model was applied to estimate levels of income taxation and consumption taxation, respectively. However, neither of the regressions came out significant as the composition of tax revenue could not be explained with the underlying variables. 14

20 influence of the trade variable on the model was however insignificant. On the basis of the values derived from the regression, the tax capacity for each country is calculated. Figure V shows the results. Figure V. Tax capacity (As a share of GDP) 25% 20% 15% 12.7% 12.7% 13.9% 14.3% 14.4% 14.7% 15.0% 16.0% 16.0% 16.4% 16.9% 18.3% 20.1% 10% 10.3% 10.6% 7.0% 5% 0% Sudan Yemen Iraq Libya Syria Saudi Arabia Egypt Oman Kuwait Bahrain Morocco Qatar UAE Tunisia Jordan Lebanon Figure VI. Tax effort ratio (Actual/potential) Kuwait Saudi Arabia UAE Bahrain Oman Iraq Libya Yemen Jordan Lebanon Qatar Egypt Sudan Syria Tunisia Morocco Note: Data are for different years, as follows: Bahrain: 2011; Egypt: 2011; Iraq: 2009; Jordan: 2011; Kuwait: 2009; Lebanon: 2011; Libya: 2008; Morocco: 2010; Oman: 2011; Qatar: 2010; Saudi Arabia: 2011; the Sudan: 2009 (before separation of South Sudan); Syria: 2009 (pre-conflict figures); Tunisia: 2011; United Arab Emirates: 2007; Yemen: Values for Palestine are not available. Subsequently the tax capacity is divided by the effective tax revenue-per-gdp ratio. If the effective tax revenues meet the capacity, the tax effort will be 1. A value below 1 suggests that there is potential to increase tax revenue, while a ratio above 1 indicates limited potential. Figure VI shows the tax effort data for ESCWA member States. Immediately standing out are the low values for certain oil-exporting countries: Kuwait, the United Arab Emirates, Saudi Arabia, Oman, Bahrain, Iraq and Libya. These are the same countries that place very low reliance on tax revenues, as shown earlier. It should be noted, as stated earlier, 15

21 that a low tax effort does not coincide with low tax efficiency. It merely implies that there is room to derive more revenues from taxation. This argument is particularly important in the case of GCC countries (with the exception of Qatar), as well as Iraq and Libya. These countries have the lowest tax effort values. As oil companies are usually state-owned, revenues from the oil sector count as investment income and are therefore free of taxes. Taxation revenue in Jordan 28% 26% 24% 22% 20% 18% 16% 14% 12% 18.1% 19.0% 18.7% 17.5% 18.2% 21.0% 24.4% 24.6% 24.7% 17.7% 17.0% 15.9% 15.0% 15.3% Tax revenue (as a share of GDP) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Tax on income and profits Property tax General sales tax Taxes on foreign trade Other taxes Source: World Bank data and Ministry of Finance. In the last ten years, a gradual increase of tax revenues has been reported in Jordan. Between 2002 and 2007, tax revenues as a share of GDP increased substantially from 17.5 per cent to 24.7 per cent. Thereafter a constant decline in relative tax revenues to 15.3 per cent in 2012 was witnessed. Tax revenues have not been able to keep up with GDP growth, which negatively affected the country s budget balance (-12.7 per cent in 2011) and increased gross public debt. The results of the tax effort regression suggest that Jordan s taxation revenue is currently 18 per cent under potential. The tax revenue is lower than one would expect given Jordan s countryspecific characteristics. The Jordanian tax system relies heavily on consumption taxes. Revenues from GST accounted for 64.5 per cent of total tax revenue in The relative importance of GST has accelerated since 1999 and has increasingly replaced revenues from taxes on foreign trade, which have become a minor source of revenue. Income taxes are limited and dominated by corporate income tax revenue. Source: Iversen, Another category of countries consists of Yemen (0.65), Jordan (0.82) and Lebanon (0.85), where tax effort levels are significantly higher than in GCC countries. Nevertheless they remain below potential. Yemen has a high share of oil-related revenue, which may partially explain the low tax effort. Jordan s and especially Lebanon s government budget is however predominantly financed by taxation revenue, as there is no involvement in the oil and gas industry. The tax effort values indicate respectable upside potential. Even though Jordan has undergone tax reforms in the recent years, its tax effort has further decreased. In previous years the tax effort ratio in Qatar was well above the expected value (1.23 in 2009). In 2010, tax revenue remained under potential (0.91). The result should be analysed within the context of the oil sector, accounting for the vast majority of taxes paid. If taxation on oil-revenues were excluded tax effort in the non-oil sector would be as low as in other GCC countries. Egypt (0.98), the Sudan (1.01) and the Syrian Arab Republic (1.08) are the countries in the sample with tax revenues that are more or less in line with tax capacity. With a tax-to-gdp ratio of only 6.7 per cent, the Sudan registers among the lowest figures in the world. With levels around 13 per cent, the Syrian Arab Republic is under the world average as well. When applying Heller s 15 per cent rule, the Sudan and the Syrian Arab Republic possess certain scope to expand tax revenue. 16

22 The remaining two countries, Tunisia (1.24) and Morocco (1.45), are clearly above the expected value and potential is therefore rather limited. Tunisia and especially Morocco show a high tax effort. Both of their budgets are heavily dependent on taxation revenues, the main source of income (over 80 per cent of government revenues). VII. CONCLUSION Countries in the Arab region are currently challenged both socially and economically. While GCC countries have ample fiscal space, many other Arab countries are limited by their budget constraints. On average, ESCWA member States show relatively low reliance on taxation revenues. Both as a share of total government revenue and of GDP, tax revenues of the region s countries are lower than those of other comparable countries. These low levels present significant upside potential for almost all ESCWA member States. In addition, many countries show fluctuation in tax revenue levels, suggesting inefficient policy. An assessment of the taxation revenue structure in the region revealed that ESCWA member States are largely dependent on consumption taxes. This reliance has increased in recent years in a number of countries. Income taxation generally accounts for a rather small share. Personal income taxation in particular generates very low revenues in the region. A tax effort analysis showed that GCC countries (with the exception of Qatar), Iraq, Libya, Yemen, Jordan and Lebanon are below their respective expected levels. They therefore possess the potential to ramp up their revenue mobilization from taxation and create more fiscal space. This potential suggests that additional tax revenue could contribute to the financing of a social protection floor. Morocco and Tunisia, which rely heavily on tax revenues, may have to find other ways to finance a social protection package, as their taxation revenues are already higher than expected by the tax effort analysis. In general, tax revenues should be increased in ESCWA member States. While a focus on consumption taxes such as VAT broadens the tax base, extended income taxation is favourable. Given the redistributive effects, this form of revenue mobilization should be increased. Ideally, policy changes should aim to increase the tax base while taking into account equity concerns. Taxation revenue is crucial to the fiscal space discussion; however other measures should be taken into consideration. While deficit financing is not sustainable in the long run, expenditure-shifting policies present another source of fiscal space, which should be the subject of further research. 17