Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? June 2017 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? The Distributional Impact of Fiscal Policy in Namibia © 2017 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank in close collaboration with staff of the Namibia Statistics Agency. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. Front cover, top: Basket weavers in Ongula, Northern Namibia @Philip Schuler / World Bank Front cover, bottom: Hilton Windhoek Hotel - Studio One, Photography Markus Weiss Back photo: Ev Thomas/Shutterstock.com The report was designed and typeset by Shereno Printers, Gauteng, South Africa. Contents Preface vi Foreword vii Acknowledgements viii Executive Summary 1 Chapter 1: Poverty and Inequality Context in Namibia 4 A. Why this study? 4 B. Setting the stage: Namibia’s development challenges 4 C. Responding to the challenges of poverty and inequality 9 D. Why now? 11 E. Content of the report 13 Chapter 2: Namibia’s Tax and Public Benefits System 14 A. Structure of taxes 14 B. Public social benefits system 18 C. Non-social spending benefits (indirect subsidies) assessed in this study 21 Chapter 3: Methodology and Data 22 A. What is fiscal incidence analysis? 22 B. Data, incomes, and income components 27 Chapter 4: Incidence and Progressivity of Taxes 30 A. Direct taxes 31 B. Indirect taxes 32 C. Progressivity of total taxes (direct and indirect) 33 Chapter 5: Incidence and Efficiency of Social Spending 35 A. Incidence and efficiency of direct transfers 35 B. Incidence and efficiency of indirect subsidies 44 C. Progressivity of in-kind transfers: education and health 45 D. Tying it all together: Progressivity of the fiscal system 48 Chapter 6: Impact of Fiscal Policy on Poverty and Inequality 50 A. Overall impact of fiscal policy on poverty and inequality 50 B. Marginal contribution of the components to poverty and inequality reduction 54 C. Comparison to other middle-income countries 56 D. Changes in fiscal policy between 2009/10 and 2016 and possible implications 59 Chapter 7: Conclusions and Implications for Policy and Data 61 A. Main findings 61 B. Distributional analysis of taxes 62 C. Transfers and subsidies 62 D. In-kind transfers in health and education 63 E. Impact on poverty and inequality 64 Annex 1: What is Fiscal Incidence Analysis 65 Annex 2: Classifying targeting methods 67 Notes 69 References 75 iii Figures Figure 1. Poverty has declined substantially since 1993 5 Figure 2. Poverty is high in Namibia relative to average income per person 5 Figure 3. Namibia has the second-highest Gini index of inequality in the world 6 Figure 4. GDP growth has been accelerating 7 Figure 5. Unemployment has remained stubbornly high since independence 7 Figure 6. Labor force participation is low in Namibia compared to other developing countries 8 Figure 7. Few poor households benefit from employment income 8 Figure 8. Employment has not shifted into higher productivity activities 9 Figure 9. Employment is falling in industries where productivity is rising 9 Figure 10. Spending on education, health, and other social programs is large and has been growing as a share of the economy 10 Figure 11. Spending on social grants has been generous 11 Figure 12. Spending and deficits are at record levels 12 Figure 13. SACU receipts are declining sharply 12 Figure 14. Composition of Total Government Revenues as a share of GDP 16 Figure 15. Social spending as a share of GDP 19 Figure 16. Definitions of income used in the CEQ fiscal incidence analysis 23 Figure 17. Diagram representing the progressivity of taxes and transfers 24 Figure 18. Income totals by decile (Namibian dollars, trillions) 28 Figure 19. Personal income tax collection total (of total reference income) 29 Figure 20. Tax revenue, cross country comparison (percent of GDP) 30 Figure 21. Concentration curves for PIT (share of total tax paid by reference i ncome deciles) 31 Figure 22. Concentration shares: personal income taxes 32 Figure 23. Progressivity of direct tax system: Kakwani coefficient 32 Figure 24. Concentration curves of indirect taxes (share of total tax paid by reference income deciles) 33 Figure 25. Concentration curves of direct and indirect taxes (share of total tax paid by reference income deciles) 34 Figure 26. Spending on social protection by country (percent of GDP) 36 Figure 27. Distribution of spending on direct transfers by type of program, 2012/13 36 Figure 28. International comparison: coverage of the transfers (poorest decile) 37 Figure 29. International comparison – targeting efficiency of the transfers 37 Figure 30: Progressivity of direct cash transfers by category: concentration curves for transfers and Lorenz curves for market incomes, Namibia and South Africa 38 Figure 31. Targeting accuracy, coverage, and generosity - all direct transfers 39 Figure 32. Progressivity of all direct transfers 39 Figure 33. Targeting accuracy, coverage, and generosity of each direct transfer 41 Figure 34. PMT conditional cash transfers in Indonesia - concentration curves for transfers 43 iv Figure 35. Concentration curves for subsidies 45 Figure 36. Concentration curves, education 46 Figure 37. Percent of households utilizing public education, by reference income decile 47 Figure 38. Concentration curves, health and all in-kind 48 Figure 39. Progressivity of the fiscal instruments (Kakwani Index) 49 Figure 40. Poverty and inequality indicators at each income concept 51 Figure 41. Amounts by which incomes exceed market income (as a share of market income) 53 Figure 42. The impact of fiscal policy on inequality worldwide, circa 2009/2010 (measured in Gini points) 57 Figure 43. Change in poverty headcount ratio (%) from market to consumable income, US$2.50 PPP per day poverty line, circa 2009/2010 59 Tables Table 1. Namibia government revenue collections 15 Table 2. Income tax rates in 2009 17 Table 3. Namibia general government expenditure, percent of GDP 18 Table 4. Coverage (targeting efficiency) of direct transfers 39 Table 5. Poverty and inequality indicators for each income concept 52 Table 6. Fiscal impoverishment at a $1.25 PPP per day poverty line 53 Table 7. Marginal contribution of indirect subsidies and direct transfers to poverty 54 Table 8. Marginal contribution of indirect subsidies and direct transfers to inequality 55 Table 9. Gini coefficient at each income concept 56 Table 10. Poverty headcount ratio for the US$ 2.50 PPP a day for each income concept 58 Table 11. Income tax rates in 2009 and 2013 60 Boxes Box 1. Caveats of the CEQ analyses and data limitations 26 Box 2. Main indicators of performance and targeting efficiency of direct transfers 35 Box 3. Advantages of proxy means tests for targeting social programs: the case of Indonesia 43 v Preface I am pleased to present the report: Does headcounts and the Gini coefficient while Fiscal Policy Benefit the Poor and Reduce drawing comparison from experiences of Inequality in Namibia?, jointly prepared by other countries. This way, the report provides the World Bank and the Namibia Statistics evidence that can shape public debates over Agency. I seize this opportunity to express government spending and overall design of my gratitude to the Namibia Statistics social programs. Agency, particularly John Steytler (former The main conclusion is threefold: firstly, and founding Statistician-General, Namibia Namibia’s generous fiscal policy does reduce Statistics Agency (NSA)) who initiated this poverty and inequality, but its impact is study and to Alex Shimuafeni, Statistician- relatively modest in comparison to other high General (NSA), for his leadership during inequality countries. Secondly, the overall the preparation of this report. This report income tax system in Namibia is mildly goes to the heart of Namibia’s most pressing progressive with the income tax burden challenges of poverty and inequality - as falling on the top income earners and the poor articulated in Vision 2030, the country’s hardly paying income taxes, yet indirect taxes guiding development strategy. The report tend to be rather neutral. Thirdly, generous is also fully aligned with the World Bank and progressive social spending benefits Group’s twin goals to help client countries the low income earners and the poor, but eliminate extreme poverty and boost shared its coverage and efficiency could be further prosperity by 2030. improved. The most progressive programs are Despite progress made in addressing direct transfers, especially the child support development challenges since independence grants and old age pension. in 1990, challenges for poverty and inequality To further reduce poverty and inequality, reduction remain. The economy’s steady this report suggests that Namibia will need growth has not generated enough jobs, further improvements in the efficiency of resulting in sluggish reduction in poverty, social spending through for example better inequality, and unemployment. Poverty targeting efficiency and consolidation of rates are relatively high for an upper middle social programs, and reducing leakages of income country. World Bank estimates show existing programs. Ultimately, higher and that 16.9 percent of the population lived on more inclusive economic growth that creates less than US$1.90 a day in 2015. Inequality more jobs for the poorest members of society is among the highest in the world and is needed. unemployment remains relatively high. The It is my hope that this evidence-based persistence of the triple challenge of poverty, analysis will enhance ongoing public debates inequality and unemployment despite the on fiscal policies that are suitable and high allocation of public resources to address effective to tackle poverty and inequality them calls for an in-depth assessment of while also ensuring that existing policies are whether the government is making the best implemented more efficiently and effectively. possible use of fiscal policy to reduce poverty and inequality. This report contributes to Paul Noumba Um such analyses by assessing the contribution Country Director for Namibia of fiscal instruments to reducing poverty World Bank vi Foreword Poverty and inequality remain two of to those of other countries. Namibia’s pressing challenges. Though falling, Namibia’s fiscal policy is found to poverty and inequality remains relatively reduce both poverty and inequality although high especially considering the amount of inequality is still among the highest in resources the government has invested in the world and is reducing at a slow pace. addressing these challenges. As Namibia takes The report emphasizes the limits of what steps towards accelerating development and redistributive fiscal measures alone can eradicating poverty, new analytical studies accomplish. It stresses the need to promote are vital to generating knowledge, expertise, job creation as a sustainable way to reduce innovations, and new thinking on how poverty and inequality. to address these long standing challenges. This work was initiated by John Steytler Effective partnerships in analytical studies (former and founding Statistician-General, aimed at reducing inequality and poverty Namibia Statistics Agency (NSA)). It was through comprehensive and rigorous tax prepared by a team at the World Bank in close and benefit incidence analysis, and active collaboration with a team in the Household engagement with the policy community and Welfare Statistics division at the NSA. are highly welcome. This report uses this The NSA would like to express appreciation framework to answer two main questions: to the World Bank for its leadership in To what extent do taxes and spending in preparing this report as well as for availing Namibia redistribute income between the funds to support the preparation of the rich and poor households? What is the report. Appreciation also goes to colleagues impact of taxes and spending on poverty and at the Ministry of Finance who generously inequality levels in Namibia? provided administrative data needed to carry In answering these questions, the report out this study. contributes to policy introspection on I am confident that the analysis will whether the Government is making the best deepen policy dialogues and debates on possible use of fiscal policy to reduce poverty policies to tackle the twin challenges of and inequality. It uses an innovative approach poverty and inequality in Namibia. More that combines administrative fiscal data specifically, I hope that this report informs with the Namibia Household Income and policy dialogues on the implementation of Expenditure Survey (NHIES) data to analyze the National Development Plan and the the impact of the tax and benefit system Harambee Prosperity Plan. (fiscal policy) on poverty and inequality in Alex Shimuafeni Namibia. An international perspective is Statistician–General & CEO given by comparing the results from Namibia Namibia Statistics Agency vii Acknowledgements This report was prepared by a core team administrative data needed for the analysis. comprising Victor Sulla, Precious Zikhali, The report benefited from discussions with Philip Schuler, and Jon Jellema (Associate government officials, particularly in two Director, Commitment to Equity Institute). workshops held in Windhoek in October 2015 The report was undertaken under the guidance and October 2016, development partners of Andrew Dabalen (Practice Manager), and stakeholders outside government. Pablo Fajnzylber (Practice Manager), Paul The peer reviewers for the report were Noumba Um (Country Director), Guang Matthew Grant Wai-Poi and Maria Ana Zhe Chen (former Country Director), Lugo. The team also thanks Lucilla Maria Sebastien Dessus (Program Leader), and Ivan Bruni (GSP01) for her continued support Velev (Country Program Coordinator). The throughout the preparation of the report. report was prepared in close collaboration Logistical assistance in the preparation with the Social Statistics division at the of this report was ably provided by Siele Namibia Statistics Agency (NSA), led by Shifferaw Ketema and Mokgabo Molibeli. Daniel Oherein (Manager: Social Statistics). Communications support was provided The team would like to thank John by Zandile Ratshitanga and Maura Leary. Steytler (former and founding Statistician- Last but not least, the team would like to General, NSA) for his fruitful collaboration thank everyone at NSA and the World Bank and specifically for initiating the study. The who contributed to making this a truly support of Alex Shimuafeni (Statistician- collaborative effort. Thank you. General and CEO, NSA) and Liina Kafidi The Commitment to Equity (CEQ) (Executive: Demographic and Social toolkit developed by Tulane University and Statistics, NSA) is greatly appreciated. Special ADePT Software developed by the World thanks to Penda Ithindi, Ndilimeke Iipinge, Bank were used to produce most of the Robert Kaveto, and Kenneth Haludilu at the statistical outputs in the report. Ministry of Finance for generously availing viii Executive Summary Reducing poverty and inequality continues public services. to be an important national priority in Despite the progress, daunting challenges Namibia. Vision 2030 – the country’s guiding for poverty and inequality reduction remain. development strategy – has a subordinate The economy’s steady growth has not vision that points to several goals: “Poverty is generated enough jobs, resulting in sluggish reduced to the minimum, the existing pattern of reductions in poverty, inequality, and income-distribution is equitable and disparity unemployment. Though falling, poverty rates is at the minimum.” Vision 2030 is being are relatively high for an upper middle income implemented via a series of five-year National country. World Bank calculations show that Development Plans, with the current 16.9 percent of the population lived on National Development Plan IV (NDP4) less than $1.90 a day in 2015. Inequality is covering 2012 through to 2017. NDP4 sets among the highest in the world. In 1993/94, specific numerical targets. One is reducing the Gini coefficient stood at 0.646, declining the incidence of extreme poverty to less to 0.60 in 2003/04 and 0.597 in 2009/10. than 10 percent of individuals by the end of Unemployment has remained stubbornly FY2016/17, measured at the national lower high. The Labor Force Survey (LFS) reported bound poverty line of N$277.54 in 2009/10. an unemployment rate of 28.1 percent in Another is reducing the Gini coefficient by 2014, with youth unemployment higher at 3 percent a year on a path toward achieving 39.2 percent (NSA, 2014). Vision 2030’s goal of 0.30. The persistence of the triple challenge Political stability, sound economic of poverty, inequality, and unemployment management, moderate economic growth, despite the high allocation of public resources and a sustained fiscal commitment to social to address them calls for an assessment of programs have helped Namibia confront whether the government is making the developmental challenges since independence best possible use of fiscal policy to reduce in 1990. More than half of government poverty and inequality. This becomes more spending routinely goes to education, health, pertinent now that Namibia has entered a social security, housing, and other social period of fiscal consolidation. This study programs. This has been complemented by seeks to help the country assess poverty and a highly progressive income tax schedule inequality reduction programs during a time and exemptions in the value-added tax of budget cuts. It uses the fiscal-incidence for goods consumed by the poor. This has methodology developed by the Commitment helped achieve notable progress in reducing to Equity (CEQ)1 project to assess the poverty, although variations persist across poverty reduction and redistributive effects the country’s 14 regions. The incidence of of Namibia’s taxes and public benefits. By extreme poverty fell from 58.9 percent of decomposing the contributions of individual individuals in 1993/94 to 15.3 percent in tax and spending measures, the report 2009/10. In addition, the country has made provides a unified framework for measuring strides in upgrading its human development programs’ progressivity, generosity, coverage, record by improving citizens’ access to basic and final impacts, whether they take the 1 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? form of direct cash grants, indirect subsidies, education and outpatient healthcare). or in-kind subsidies. This can contribute Richer households acquire greater shares to the development and monitoring of the of the higher value but not as frequently NDPs and serve as a tool for improving the provided public services (tertiary education targeting efficiency of social spending. and inpatient healthcare). As a result, the Results show that Namibia’s personal distribution of total public expenditures income tax (PIT) is progressive, while on in-kind health and education services indirect taxes (fuel levy and value added is approximately neutral; each population tax), when taken together, are more or less group receives a share of benefits equal to its neutral. Focusing on the tax system as a share of the population. whole reveals a progressive system, driven by Taken separately, spending in public the progressivity of PIT. school education is absolutely progressive An analysis of the incidence and but the system becomes less progressive efficiency of direct transfer spending shows as education levels rise – from primary that: (i) overall spending on direct transfers to tertiary. Although education spending is above the average for Sub-Saharan African matters for everybody, outlays for primary countries, but comparable to the average for education target the poor better. However, developing countries; (ii) it is progressive spending on tertiary education is regressive in terms of targeting, that is, the poor are by enrollment rather than program design more likely to receive the transfers; (iii) the – the small number of students reaching transfers matter for the poor, being generous the tertiary level are disproportionately non- in that they make up a larger share of total poor. As a result, near-neutrality of education income; and (iv) the targeting accuracy of spending is observed. transfers is low and could be improved. Fiscal policy on the whole reduces Taken together, indirect subsidies are poverty in Namibia. Severe poverty, defined progressively targeted, largely because of as the proportion of Namibians living below the water subsidy. This study considered the lower bound poverty line of N$277.54 three indirect subsidies: the rural water per month, falls from 22.2 percent before infrastructure and services program and any fiscal policy measures (i.e., based on two housing subsidies, the Build Together market incomes) to 16.7 percent after Program (BTP) and the National Housing adding direct transfers and indirect subsidies Enterprise (NHE). The progressivity is largely and subtracting taxes. This translates to a driven by the fact that the water subsidy is reduction of 24.7 percent in extreme poverty allocated only to households in rural areas, due to fiscal policy. Not only does fiscal policy where the majority of poor and low income reduce poverty on the whole, the proportion households live. Both the BTP and NHE are of individuals made poor by the application active primarily in urban areas and are less of fiscal policy is low. likely than the water subsidy to reach a poor Direct transfers drive poverty reduction, household. while the role of taxes and indirect subsidies Taken together, expenditures on in- is less significant. The poor barely pay income kind health and education services are taxes and as a result, no change in the poverty approximately neutral. In both education rate is observed when personal income and health, poorer households acquire taxes are deducted from market incomes. greater shares of the lower value but more Introduction of direct transfers, however, 2 frequently provided public services (primary reduces extreme poverty by 30.6 percent to 15.4 percent (6.8 percentage points). The reduction from the Gini coefficient for introduction of indirect subsidies followed market incomes. Approximately 78.2 percent by indirect taxes leads to a 1.3 percentage of the Gini coefficient reduction from market point increase in poverty. to final incomes is attributed to in-kind Internationally, the poverty reduction transfers, 16.4 percent to direct transfers, attributable to fiscal policy in Namibia is and the remaining 5.4 percent to direct and on par with South Africa: from the set of indirect taxes and subsidies. individuals who would be poor without This report demonstrates that Namibia’s fiscal expenditures, approximately 11 percent progressive income tax and generous social are in receipt of fiscal transfers that help spending programs substantially reduce them escape impoverishment (measured as poverty and inequality, but the analysis also expenditure of $PPP 2.50 or less per day). underscores the limits of what redistributive Fiscal policy on the whole reduces fiscal measures alone can accomplish. The inequality in Namibia, largely because of economy must ultimately create more jobs in-kind transfers. The market income Gini for the poorest members of society to change coefficient of 0.635 falls to 0.590, using the underlying distribution of what might be post-fiscal income (which does not include called “pre-fiscal” income; i.e., the income a monetized value of health, education, or before households pay taxes and receive other public in-kind services received). If benefits from social programs. This will taken into consideration, in-kind transfers in require structural transformation through health and education would have the highest greater investment in activities that create redistributive effect. Including monetized employment for unskilled workers and offer value of in-kind transfers would further the potential for continuous productivity reduce inequality to 0.429, a 0.206-point increases. 3 Chapter 1: Poverty and Inequality Context in Namibia A. Why this study? rates on more highly paid individuals. In the past 25 years, Namibia has devoted Namibia is one of the youngest countries in around half of the government budget to Africa, having gained its independence from social programs. The government provides South Africa in 1990. In recent decades, a number of cash transfers to vulnerable the country has established an enviable segments of the population; these are fairly track record of political stability, prudent generous compared to similar programs in macroeconomic policies, moderate economic other developing countries. growth, and a sustained fiscal commitment This report aims to measure the to social programs. Per capita income has effectiveness of these efforts and draws grown. Financial inclusion is high by regional comparisons to the experiences of other standards. Poverty has declined substantially. countries. It estimates how major taxes and Namibia has achieved these gains while facing social spending programs affect individual constraints imposed by climate, geography, incomes. It then assesses who benefits from or Despite progress and legacies of apartheid and colonialism. bears the burden of each instrument and by made in addressing Daunting challenges remain. Namibia how much. This way, the analysis estimates development suffers from chronic high unemployment, the contribution of each instrument to challenges Human Immune-deficiency Virus (HIV) reducing the poverty headcount and the Gini since attaining and Acquired Immune Deficiency Syndrome coefficient, a standard measure of inequality. independence in (AIDS), and a distribution of income that The analysis provides evidence that 1990, daunting is among the world’s most unequal. Only a can shape public debates over government challenges for poverty minority of the population lives in conditions spending and the design of social programs. and inequality expected in an upper middle income country. This report is particularly timely in light of reduction remain Economic growth has not generated jobs the need for deep budget cuts announced by because the structure of production and the Minister of Finance in November 2015. external trade remains essentially unchanged, It can also contribute to the development and tied closely to metals, minerals, and other monitoring of new anti-poverty programs natural resources. In addition, the country and of the next National Development Plan faces new risks stemming from global climate (NDP5), which had the official launch of its change, a growing number of unemployed formulation process in June 2016. and poorly educated youth, increased debt exposure, and diminished fiscal space. B. Setting the stage: Namibia’s Namibia has for years devoted development challenges considerable fiscal resources to addressing poverty and inequality. Low income workers Poverty is falling, but remains high. Since are exempt from personal income taxes, the early 1990s, the incidence of poverty in which are applied at progressively higher Namibia has declined substantially.2 It fell 4 from 58.9 percent of individuals in 1993/94 depth and severity of poverty show similar to 21.9 percent in 2003/04 and then to 15.3 declines. When compared to other countries, percent in 2009/10, measured at the lower however, Namibia’s poverty remains high bound national poverty line of N$277.54 relative to its level of national income per in 2009/10 (Figure 1). At the upper bound person (Figure 2). According to World Bank poverty line of N$377.96 in 2009/10, poverty calculations, 16.9 percent of the population fell from 69.3 percent in 1993/94 to 37.7 lived on less than $1.90 a day in 2015. At the percent in 2003/04 and then to 28.7 percent $3.10 a day international poverty line, the in 2009/10 (NSA, 2012). Measures of the rate was 42.8 percent in 2015. Figure 1. Poverty has declined Figure 2. Poverty is high in Namibia substantially since 1993 relative to average income per person 90 90 Pov e rty he a dc ount a t the na tiona l Poverty headcount at US$1.90/day 80 80 lower bound poverty line 70 70 60 58.9 60 50 50 PPP 40 40 30 28.1 21.9 30 Namibia 20 20 16.8 15.3 10 6.6 4.2 10 0 2.9 1.7 0 1993/94 2003/04 2009/10 $0 $10,000 $20,000 $30,000 Poverty has been Poverty Incidence falling in Namibia, Poverty gap Per capita Gross National Income (Atlas but remains higher method) Poverty severity than in countries with similar incomes Source: Namibia Statistics Agency. Source: World Development Indicators. Notes: Values are the most recent available over the past six years. In 2009/10, the poor were predominantly in poverty by the main income source shows women, subsistence farmers and pensioners, that although the population share with and Namibians living in rural areas (NSA, subsistence farming as the main source of 2012). Regional variation is also revealed income fell between 2003/04 and 2009/10, with the Kavango region recording the poverty reduction among subsistence farmers highest incidence of poverty. Growth in contributed 2.94 percentage points to the mean consumption drove the reduction 8.34 percentage point reduction in poverty in poverty observed between 2003/04 and when the upper bound national poverty line 2009/10, while the distributional impact is used (NSA, 2012, pp 31). The proportion of inequality put a damper on poverty of pensioners grew during this period reduction. Improvements in education and poverty reduction among pensioners outcomes supported poverty reduction. In contributed to a decline in poverty by 2.56 addition, sectoral decomposition of changes percentage points (NSA, 2012, pp 31). The 5 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? latter underscores the role of fiscal policy on the latest available data, the World Bank – specifically the Old Age Pension – in estimates that Namibia has the second supporting poverty reduction. most unequal distribution of income in the Inequality is among the highest in world after South Africa (Figure 3). Unlike the world. A century of colonial rule and the poverty headcount, the Gini coefficient apartheid concentrated Namibia’s wealth – has not declined substantially over the past including ownership of land, companies, two decades. In 1993/94, it stood at 0.646, and financial assets – in the hands of a small declining to 0.600 in 2003/04 and 0.597 minority. The country has taken great strides in 2009/10. Inequality remains higher in to expand ownership of these assets by those urban compared to rural areas. In terms of who were excluded in the past. Nevertheless, consumption inequality, the Karas region income inequality remains quite high. Based registered the highest inequality. Figure 3. Namibia has the second-highest Gini index of inequality in the world 70 Namibia 60 50 Global average = 38.5 Gini index 40 GDP grew at an 30 average of 5.6 percent a year between 2010 20 and 2014 but the 10 growth did not result in significant job 0 creation Source: World Bank staff calculations in World Development Indicators. Jobs are critical for moving out of behind the sluggish reduction in inequality poverty, but GDP growth has not generated and the relatively high poverty headcount. many jobs. GDP has been growing at an Despite this enviable record of growth, accelerating rate since the 1980s (Figure 4). unemployment has remained stubbornly Between 2010 and 2015, it reached an average high. Since the early 1990s, the strict of 5.6 percent a year. This was supported unemployment rate has remained at around by historically low real interest rates, fiscal 20 percent of the labor force (i.e., those stimulus, rapid private sector credit growth, with jobs or actively seeking work). When and foreign investment in several large discouraged workers are included, which is mining projects. Construction and services the official broad unemployment rate used in industries made the largest contributions to Namibia, the jobless rate rises to 30 percent growth in the economy. Steady growth has or higher (Figure 5). not put Namibians to work – a key factor 6 Figure 4. GDP growth has been Figure 5. Unemployment has remained accelerating stubbornly high since independence 14% 60 12% 50 constant 2000 Namibian dollars Year on year growth of GDP, 5.6% 10% 40 Unemployment rate 4.4% 8% 3.7% 30 6% 1.1% 20 4% 10 2% 0% 0 1991 1998 2005 2012 -2% -4% Strict De nition 1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2011 2014 Broad (incl. discouraged workers) Source: Namibia Statistics Agency. Sources: Namibia Statistics Agency population census (1991, 2001, 2011), household income and expenditure surveys (1993, 2003, 2009), and labor-force surveys (1997, 2000, 2004, 2008, 2012–14). Not only is unemployment high in Few poor households in Namibia benefit Namibia but labor-force participation is from employment income. In the bottom low, relative to other developing countries. four income deciles, only 30 percent of With less than 60 percent of working-age households depend on salary, wages, or a individuals in the labor force when the strict pension from previous employment as their Only 30 percent definition is adopted i.e., either working or primary source of income. Instead, their of the bottom 40 actively seeking employment, Namibia falls incomes come mainly from subsistence percent of households in the bottom third of developing countries farming or the receipt of social grants, depend on ranked by labor-force participation rates drought relief, or private transfers (Figure employment income (Figure 6). 7). These sources provide little potential for boosting households out of poverty. 7 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? Figure 6. Labor force participation Figure 7. Few poor households benefit is low in Namibia compared to other from employment income developing countries 100 100% Share of households in decile 80% 90 Developing 60% country 80 average 40% = 64.4% Namibia 20% 70 0% 2 3 4 5 6 8 9 7 Poorest decile Richest decile Percent 60 50 40 30 Other 20 Business income 10 Employment income 0 Remittances, grants, relief, alimony Subsistence farming Source: International Labor Organization, 2013 data. Source: World Bank calculations from the 2009/10 Namibia Household Income and Expenditure Survey. Employment has The structure of production has changed has grown significantly. Some of these may not been moving little over the past several decades. Despite have potential and could generate jobs for into industries with a movement out of subsistence farming relatively unskilled workers (e.g., tourism). growing productivity (and out of rural areas and into towns), In general, employment has not been manufacturing’s share of employment has moving into industries with growing declined. In the latest Labor Force Survey, productivity. Between 2012 and 2014, more people were employed in construction employment tended to grow mainly in than in manufacturing, mining, and public industries with low productivity growth, utilities combined (Figure 8). Services measured by real value added per worker. that are largely non-tradable—public It fell in industries with higher productivity administration, defense, community services, growth (Figure 9). Construction has been health, education, and household services— an exception with gains in both jobs and account for 31 percent of total employment. productivity, but on average, the sector’s The share of employment in financial, value added per worker is low compared to commercial, transportation, professional, other industries. and other services that are largely tradable 8 Figure 8. Employment has not shifted Figure 9. Employment is falling in into higher productivity activities industries where productivity is rising 50% 50% 45% Mining Public Admin 40% 40% Construction 35% Share of employment 30% 30% Trade 25% 20% Finance and 20% Utilities Growth in employment 15% Insurance 10% Agriculture 10% 5% Social Education 0% 0% 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 -10% Transport Public sector, community, and household services -20% Agriculture, forestry shing Manufacturing -30% Household Commercial and tradable services Hospitality -40% Real estate, Construction Business Manufacturing, mining, and utilities -50% -60% -40% -20% 0% 20% 40% 60% 80% 100% Growth in value added per worker Sources: World Bank calculations from Namibia Statistics Agency Source: World Bank calculations from NFLS and national Labor Force Surveys (1997, 2000, 2004, 2008, 2012–14) and accounts data. population census (1991). Note: The size of bubbles represents 2014 value added per worker (constant 2010 Namibian dollars). C. Responding to the challenges A series of five-year National Development of poverty and inequality Plans is intended to implement Vision 2030. The National The current National Development Plan IV Development Plans have emphasized Reducing poverty and inequality remains an (NDP4) covers 2012 through 2017.3 Unlike inclusive, sustained important national priority in Namibia. The the previous NDPs, it is narrowly focused on and equitable country’s guiding development strategy— a few overall goals, supported by a small set economic growth to Vision 2030—sets forth an overriding of subordinate objectives. The three overall realize Vision 2030 objective: “A prosperous and industrialized goals are high and sustained economic Namibia, developed by her human resources, growth, employment creation, and increased enjoying peace, harmony and political stability” income equality. To reach these goals, NDP4 by the year 2030 (Government of the Republic calls for reducing extreme poverty and for of Namibia, 2004). Beneath is a subordinate improving the institutional environment, vision: “Poverty is reduced to the minimum, education, health, and public infrastructure. the existing pattern of income distribution is The economic priorities are increasing equitable and disparity is at the minimum.” output for targeted sectors: logistics, tourism, Vision 2030 challenges the country to reduce manufacturing, and agriculture. poverty by ensuring that all Namibians enjoy NDP4 sets a specific numerical target of access to safe drinking water, comprehensive reducing the incidence of extreme poverty to health services, housing, sanitation, and below 10 percent of individuals by the end of other basic services. In addition, it calls for FY2016/17.4 Although not setting an explicit social integration of people with disabilities. target for increasing income equality, NDP4 Finally, Vision 2030 sets a numerical target suggests reducing the Gini coefficient by 3 of reducing the Gini coefficient to 0.30 by percent a year on a path toward achieving 2030. Vision 2030’s goal of 0.30. 9 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? Since its independence in 1990, Namibia global economic downturn since the 1990s, has placed a high priority on using public Namibia sharply increased social spending as resources to address poverty, inequality, and a share of GDP, bringing spending to record other social policy objectives (Figure 10). high levels. A highly progressive income tax Education, health, social security, housing, schedule and value-added tax exemptions for and other social programs routinely receive goods the poor consume complement public more than half of government spending. spending on poverty reduction. Despite external shocks caused by the Figure 10. Spending on education, health, and other social programs is large and has been growing as a share of the economy 30% 10% Education 9% 25% 8% All Other Social 7% 20% Share of spending Share of GDP Economic Services 6% 15% 5% Defence 4% 10% Public Order and 3% Safety 2% 5% Health 1% Social spending 0% General Public 0% 1991/92 1994/95 1997/98 2000/01 2003/04 2006/07 2009/10 2012/13 2015/16 1991/92 1994/95 1997/98 2000/01 2003/04 2006/07 2009/10 2012/13 2015/16 routinely accounts Services for more than half of government spending Sources: World Bank calculations using data from the Bank of Namibia (FY1990/91–FY2005/06) and Ministry of Finance (FY2006/07–2014/15). A well-developed program of cash comparison, recent analysis estimates that transfers to vulnerable segments of the all developing countries spend an average of population has been a cornerstone of 1.6 percent of GDP on social safety nets, and social spending in Namibia. Transfers Organization for Economic Co-operation include a non-contributory pension for all and Development (OECD) countries spend Namibians and grants to families caring for 2.9 percent (Honorati et al., 2015). As the orphans and vulnerable children, disabled fiscal incidence analysis presented in this children and adults, and war veterans.5 report will document, these transfers have Collectively, these accounted for 5 percent made important contributions to reducing of government spending and 2.0 percent of poverty and inequality in Namibia. GDP in FY2015/16 (Figure 11). By way of 10 Figure 11. Spending on social grants has been generous 8% 7% 6% 5% 4% 3% 2% 1% 0% 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2016/17 (b) 2017/18 (b) 2018/19 (b) 2015/16 (e) Share of total spending Share of GDP Source: World Bank calculations from Ministry of Finance budget documents. Note: FY2016/17 onwards are budgeted figures. Programs include grants to the elderly, disabled, orphans and vulnerable children, and veterans. D. Why now? GDP, the decline has been 5.4 percentage points of GDP over two years (Figure 13). Namibia has entered a period of fiscal The 2016 Fiscal Policy Strategy forcefully Slower spending consolidation, and this fiscal incidence analysis makes the case for curtailing spending growth is needed for can help assess anti-poverty programs during growth to preserve fiscal sustainability. The fiscal sustainability a time of budget cuts. Namibia increased Medium-term Expenditure Framework government spending and deficits to record (MTEF) presented in 2016 cuts the budget high levels during the past five years (Figure for FY2016/17 and FY2017/18 by N$7.5 12). However, Namibia’s receipts from the billion from the projections in the April Southern Africa Customs Union’s Common 2015 MTEF. During this time of fiscal Revenue Pool – one of the government’s consolidation, it is all the more important to largest revenue sources – are falling sharply evaluate the success of poverty and inequality due to the downturn in the South African reduction measures and to find ways to economy. Measured relative to Namibia’s enhance their effectiveness. 11 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? Figure 12. Spending and deficits Figure 13. SACU receipts are declining are at record levels sharply 50 25 40.5 40 20 Share of GDP 30 34.3 20 15 Share of GDP 10 10 0 -10 5 -6.2 -20 0 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2006/07 2016/17(b) 2017/18(b) 2018/19(b) 2015/16(e) 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2016/17(b) 2017/18(b) 2018/19(b) 2015/16(e) Spending Revenue Domestic taxes SACU receipts Budget balance Non-tax revenue Source: World Bank calculations from Ministry of Finance budget documents. This fiscal incidence analysis can also kind subsidies. help with the design and implementation The creation of the Ministry of Poverty of the National Development Plans and Eradication and Social Welfare in March The creation of the initiatives aimed at reducing poverty and 2015 signaled a renewed commitment to Ministry of Poverty inequality. Reducing the levels of poverty addressing poverty and inequality. The new Eradication and and income inequality are two important ministry has prepared a white paper on Social Welfare in NDP4 priorities. These are expected to poverty reduction and convened the National 2015 signaled a remain priorities in the NDP5, for which Conference on Wealth Redistribution and renewed commitment the formulation process recently began. In Poverty Reduction to stimulate public debate by Government to April 2016, President Geingob launched on existing programs and proposed initiatives, tackle poverty and the “Harambee Prosperity Plan, 2016/17– such as a universal basic income grant for inequality 2019/20: Namibian Government’s Action all Namibians, a “solidarity” super tax on Plan towards Prosperity for All” – a plan to incomes, and development of new social accelerate development via targeted measures protection instruments. These consultations to remove bottlenecks and implementation are expected to culminate in a blueprint challenges. The plan calls for improving the on how to eradicate poverty in Namibia by administration of targeted social safety nets 2025. In his State of the Nation address in (Government of the Republic of Namibia, April 2016, President Geingob indicated 2016). By decomposing the contributions that the blueprint has been finalized, and it of individual tax and spending measures, is set to be tabled in Parliament for debate.6 the report provides a unified framework The President emphasized that aspects of for measuring the progressivity, generosity, the blueprint have also been prioritized and coverage, and final impact of programs under incorporated into the Harambee Prosperity these initiatives, whether they take the form Plan. of direct cash grants, indirect subsidies, or in- 12 E. Content of the report in the analysis. It also includes a discussion of different ways of targeting beneficiaries of The report proceeds as follows: Chapter public benefits, particularly direct transfers. 2 provides an overview of Namibia’s tax This report demonstrates that Namibia’s and public social benefits system. Chapter progressive income tax and generous social 3 presents the fiscal incidence analysis spending programs substantially reduce employed to generate results. It includes poverty and inequality. However, the the methods, data, and the assumptions analysis also underscores the limits to what and choices made as well as the analytical redistributive fiscal measures alone can limitations that result when available data accomplish. The economy must ultimately was not complete. Chapter 4 examines the create more jobs for the poorest members of progressivity or regressivity of selected taxes in society to change the underlying distribution Namibia. Chapter 5 describes the incidence of what might be called “pre-fiscal” income; and efficiency of social spending. In Chapter i.e., the income a household would have before 6, the overall impact of Namibia’s fiscal it pays taxes and receives benefits from social policy on poverty and inequality is estimated. programs. This will require a transformation Chapter 7 concludes and discusses possible of the structure of the economy through policy directions for further reduction in greater investment in activities that create poverty and inequality in Namibia. employment for unskilled workers, with The Annex provides a more detailed a potential for continuous productivity description of the CEQ methodology used increases. 13 Chapter 2: Namibia’s Tax and Public Benefits System This chapter provides an overview of They accounted for around 33.2 percent of Namibia’s tax and public social benefits, total tax collections and 6.7 percent of GDP. focusing on 2009/10. It also gives an Namibia’s tax system shares an important indication of which of these taxes and characteristic with most developing spending programs are included in this study economies – dependence on indirect taxes, to assess the redistributive impact of fiscal and international trade comprised of SACU policy. receipts (Besley and Persson, 2013). Taken together, indirect taxes and receipts from the A. Structure of taxes Southern Africa Customs Union (SACU) revenue pool, made up 18.1 percent of GDP The majority of Namibia’s tax collections in in 2009/10, compared to 11.1 percent for 2009/10 came from direct taxes. Broadly, direct taxes. Excluding the SACU receipts these included taxes on income and profits to capture actual government tax collections as well as property taxes. They made up 54.8 show that the country relied more on direct percent of total tax collections and 11.1 taxes in 2009/10 than it did on indirect percent of GDP (Table 1).7 Indirect taxes domestic taxes on goods and services (Table included value added tax and the fuel levy. 1). 14 Table 1. Namibia government revenue collections Share of revenue Share of total collected by Government the GRN (i.e. Incidence revenue and excluding SACU analysis Percent of grants receipts and (percent GDP (percent) grants)(percent) of GDP) Total Government revenue and grants 32.0 Taxes 29.5 92.2 5.1 Direct taxes 11.1 34.8 54.8 2.3 Personal income tax 6.8 21.1 33.3 2.3 Corporate income tax 3.8 11.9 18.7 Other direct taxes 0.6 1.8 2.8 Indirect taxes* 6.7 21.0 33.2 2.8 Value added tax 6.6 20.7 32.6 2.7 Fuel levy 0.1 0.3 0.5 0.1 SACU receipts 11.4 35.7 Stamp duties 0.2 0.7 1.1 Non-tax revenue 2.5 7.8 12.3 Grants 0.3 0.8 1.3 Current 0.3 0.8 1.3 Capital 0.0 0.0 0.0 Other revenue 2.2 7.0 11.0 Sources: Namibia Ministry of Finance, Communication with Inland Revenue. * Excludes excises as virtually all revenue is collected in South Africa, even when excisable goods are consumed in Namibia. Only 2.3 percent of GDP worth of by households in the NHIES data are far less direct taxes is identified in the analysis than comparable data from budget reporting compared to 11.1 percent of GDP indicated documents and secondary sources. by administrative data. This is largely due to Direct taxes in Namibia are larger than two main reasons. First, the study does not most regional or income-level comparators include corporate income tax which stands at (Figure 14). Relative to other revenue sources, 3.8 percent of GDP. This decision was driven direct taxes in Namibia are also larger than by lack of an appropriate methodology that comparator countries. For example, only makes it possible to comprehensively identify in Georgia and South Africa do direct taxes households that pay corporate income tax. represent greater shares of total revenues than Second, PIT is the only direct tax that is in Namibia and only in Namibia, South analyzed in the study. The challenge, as will Africa and Mexico do direct taxes represent be elaborated later, is that the number of a greater share of total revenues than indirect (observed) taxpayers and tax revenue reported taxes. 15 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? Figure 14. Composition of Total Government Revenues as a share of GDP (ranked by GNI per capita (2011 PPP) – right-hand side) 50% 25 000 45% Revenue as a share of GDP 40% 20 000 GNI per capita 35% 30% 15 000 25% 20% 10 000 15% 10% 5 000 5% 0% 0 Namibia (2009/10) Ghana (2013) Peru (2009) Ethiopia (2011) Jordan (2010) Guatemala (2010) Indonesia (2012) Georgia (2013) Uruguay (2009) Dominican Republic (2013) Sri Lanka (2010) Mexico (2010) Russia (2010) Brazil (2009) Bolivia (2009) South Africa (2010) El Salvador (2011) Direct Taxes Indirect and Other Taxes Social Security Contributions Other Revenues GNI per capita (2011 PPP) Sources: Namibia: World Bank calculations from Ministry of Finance budget documents. Uruguay (Bucheli et al., 2014). Guatemala The majority of (Cabrera et al., 2015), Brazil (Higgins and Pereira, 2014), Peru (Jaramillo, 2014), Bolivia (Paz et al., 2014), Mexico (Scott, 2014), Ghana (Younger et al., 2015), Dominican Republic (Aristy-Escuder et al., 2016), El Salvador (Beneke et al., 2014), Namibia’s tax Ethiopia, Georgia, Sri Lanka, Indonesia, Jordan, South Africa, Russia (Inchauste and Lustig, forthcoming). collections came from direct taxes. Relative Direct taxes to other revenue income is taxed at progressive marginal rates. sources, direct taxes The major component of direct taxes is the It can be argued that the PIT schedules might in Namibia are large personal income tax (PIT), which is levied create disincentives to work in the formal on individuals’ taxable income.8 Namibia sector. While no empirical evidence exists to taxes all receipts and accruals originating support that, the fact that the government from a Namibian source. Foreign residents reduced the marginal tax rate of the bottom are taxed only on income generated within bracket to 18 percent in 2013 suggests Namibia. A self-assessment tax regime is efforts by government to increase labor force employed, and spouses are taxed separately participation and ensure the labor markets on their incomes. No deductions are support poverty and inequality reduction. provided for married persons or children. Reducing poverty and inequality cannot be Table 2 reports the applicable PIT rates at done through fiscal policy alone. different income brackets. It illustrates that 16 Table 2. Income tax rates in 2009 Taxable income (N$) Marginal rate (percent) Up to 40,000 0 40,001 to 80,000 27 80,001 to 200,000 32 200,001 to 750,000 34 Over 750,000 37 Source: Ministry of Finance budget documents. The PIT system includes provisions for (VAT) and the fuel levy. The VAT, introduced withholding tax and certain deductions. in 2000, is levied at a uniform standard rate With regard to salaries or (formal sector) of 15 percent of the value of goods and employment income, employers are services supplied or imported. However, the responsible for registering employees for VAT system includes items that are zero-rated the PIT and withholding tax under the pay- or exempt. Zero-rated food supplies include as-you-earn system. No capital gains tax mahangu (pearl millet), mahangu meal and exists in Namibia. Deductions limited to maize meal; fresh and dried beans (but not N$40,000 a year in 2009 were allowed for canned or frozen beans); sunflower cooking aggregated contributions to approved pension oil; animal fat used for food preparation; funds, retirement annuity funds, travel bread flour, cake flour and bread; fresh milk; expenses, premiums associated with tertiary and white and brown sugar.9 In 2009/10, educational policies, provident funds, and the VAT made up 32.6 percent of total Several exemptions donations to registered welfare organizations tax collections (excluding SACU receipts and zero-ratings exist or approved educational institutions. There and grants), and 6.6 percent of GDP. on goods and services are no deductions for medical expenses or Interestingly, the country’s reliance on the consumed mainly by contributions to medical schemes. PIT and VAT were almost equal in 2009/10 the poor The PIT is the only direct tax used in (6.8 percent of GDP for PIT and 6.6 percent the incidence analysis. In 2009/10, PIT of GDP for VAT). This could compromise made up 33.3 percent of total tax collections the ability of the country’s fiscal system to (excluding SACU receipts and grants) and promote equity10 because personal income 6.8 percent of GDP. It accounted for 60.8 taxes have been shown to be more progressive percent of total direct taxes. We omit from than consumption taxes. Excise tax revenue the analysis the company or corporate is excluded in this analysis. Even though income tax (CIT), which is levied at 35 Namibians consume products on which percent. The CIT accounted for about 34.1 excise taxes are levied (e.g., tobacco, alcohol), percent of total direct taxes, 18.7 percent of most of these products are produced in South total tax collections, and 3.8 percent of GDP. Africa, and therefore the South African rather The omission is largely due to the lack of a than the Namibian government collects excise methodology to comprehensively assign the taxes on these products. Consequently, excise CIT burden to households in the survey. tax revenue does not appear in the Namibian government budget documents, and it is not Indirect taxes possible to estimate excise tax revenue paid Two types of indirect taxes are included in by Namibian households. the incidence analysis – the value-added tax In 2009/10, the fuel levy made up only 17 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? 0.5 percent of total tax collection, and 0.1 survey accounts for 2.8 percent of GDP percent of GDP. Fuel levies are based on the compared to 6.7 percent of GDP indicated importation of petrol, diesel, and paraffin in administrative records. into Namibia. The fuel levy finances the Motor Vehicle Accident Fund, established to B. Public social benefits system “provide assistance and benefits to persons injured in motor vehicle accidents and to dependents of Namibia has several spending programs that persons killed in such accidents; and to provide affect poverty and income distribution. These for incidental matters.” It also contributes include programs that can be classified as to the Road Funds Administration for road social spending (direct and indirect (or in- maintenance. kind) transfers) and non-social spending11 This report accounts for only a portion (Table 3). In the social spending component, of direct and indirect taxes. Table 1 shows the direct spending includes old age pensions, study identifies direct taxes that made up 2.3 veterans’ grants, children’s grants, foster percent of GDP in 2009/10, compared to the parents’ grants, and disability grants for 11.1 percent indicated in the administrative adults and children. Indirect or in-kind account. In terms of indirect taxes, the transfers include education and health. Table 3. Namibia general government expenditure, percent of GDP 2009/10 Incidence analysis Total government spending 14.7 11.5 Social spending made Primary government spending 14.7 11.5 up 12.7 percent of GDP in 2009/10, Social Spending (excludes contributory pensions) 12.7 11.0 higher than other Direct Transfers (Total Cash and Near Cash) 1.6 1.6 middle income Old Age Pension 1.0 1.0 countries for which Veterans grant 0.1 0.1 similar analyses has Children’s grant 0.2 0.2 been done Foster parents grant 0.1 0.1 Disability grant – adults 0.2 0.2 Disability grant – children 0.1 0.1 Total In-kind Transfers 10.8 9.3 Education 7.5 6.7 Health 3.3 2.6 Other Social Spending 0.3 … Non-Social Spending 0.6 0.6 Indirect Subsidies 0.6 0.6 Build Together Program 0.1 0.1 National Housing Enterprise 0.1 0.1 Water Subsidy 0.4 0.4 Source: World Bank calculations from Ministry of Finance budget documents. 18 Social spending in Namibia compares – another form of in-kind transfer – with favorably to other countries for which CEQ spending of 3.3 percent of GDP. Direct cash studies have been undertaken12 (Figure transfers to individuals were only 1.6 percent 15). At 12.7 percent of GDP, Namibia’s of GDP. Despite the different magnitudes in social spending in 2009/10 was higher than overall expenditures, the structure of social other middle-income countries, including spending is similar across all countries in Indonesia (4.9 percent), Peru (8.4 percent), Figure 15 – education constitutes the bulk and Mexico (10.0 percent). At 16.2 percent of social spending, followed by health and of GDP in 2009, Brazil recorded the highest then direct transfers. Unlike many countries, social spending among the countries in Namibia has no contributory national Figure 15. In-kind transfers in the form of pension (or social security) scheme, nor education made the bulk of social spending does it have a national contributory health in Namibia, accounting for 7.5 percent of insurance scheme. GDP in 2009/10. This is followed by health Figure 15. Social spending as a share of GDP 16% 14% 12% Social spending in 10% Namibia is relatively Share of GDP 8% high in comparison to that in other 6% countries for which 4% similar analyses has been done. In-kind 2% transfers towards 0% education and health account for most of Namibia (2009/10) Ethiopia (2011) Peru (2009) Uruguay (2009) Guatemala (2010) Indonesia (2012) Mexico (2010) Armenia (2011) Costa Rica( 2010) Brazil (2009) Sri Lanka (2009) South Africa (2010) Bolivia (2009) El Salvador ( 2011) social spending Direct Transfer Education Health Subsidies Source: World Bank (2014) and Ministry of Finance (Namibia), available online: www.mof.gov.na/documents/57508/107402/ Estimate+book_2.pdf/fe7ade9b-86a9-4638-b68e-fd68978a7476?version=1.0. 19 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? This study’s incidence analysis focuses on parent, a parent receiving an Old Age Pension direct cash transfers and in-kind transfers for grant, or a parent who is absent due to death health and education as components of social or imprisonment. This is a means tested spending. These items are discussed in greater grant, with the threshold for an applying detail below. Collectively, they accounted for parent set at less than a gross income of 84.3 percent of total government spending N$1,000 a month. and 97.7 percent of social spending in The Foster Care Grant targets children 2009/10. Other components of social who the courts have placed in the temporary spending are excluded due to data limitations. care of foster parents. The grant’s value in 2010 was N$200 for the first foster child and Direct transfers (cash transfer N$100 for any additional foster children. programs) There is no upper limit on the number of The Old Age Pension (also referred to as the children per applicant. basic pension) is accessible to all Namibians Namibians, 16 years and older, who over the age of 60, residing in the country.13 are either confirmed by a state doctor to No means test is required. A means test is a be temporarily or permanently disabled method for establishing whether, based on (including the blind) or have full-blown an indicator of means or financial ability, an AIDS are eligible for the disability grant individual or household is eligible for specific (“disability grant A” for adults). A special government assistance (see Annex 2 for a maintenance grant for children under the age discussion of different methods used). The of 16 who are living with disabilities is also number of beneficiaries has been growing, available (“disability grant C” for children). Namibia has a increasing from around 110,000 in 2003 to number of cash In-kind transfers 143,000 in 2013. Close to 92.2 percent of transfers for eligible Namibians received the pension by Education vulnerable groups 2011. including the elderly, war veterans, the War veterans and their dependents are An average of 20 percent of the national disabled, orphaned eligible for the veterans’ subvention. This budget goes to education, a reflection of children, and targets any person who “was a member the high priority which the Namibian other children in of the liberation forces; consistently and government places on education. In need. These are persistently participated or engaged in any 2009/10, spending on education amounted complemented by in- political, diplomatic or underground activity in to 7.5 percent of GDP. The importance of kind transfers largely furtherance of the liberation struggle; or owing education is recognized in the Namibian towards health and to his or her participation in the liberation Constitution. The Education Act of 2001 education as well as struggle was convicted, whether in Namibia further underscores the need to provide indirect subsidies or elsewhere, of any offence closely connected to for an accessible, equitable, qualitative and the struggle and sentenced to imprisonment.” democratic national education service. This is means tested in the sense that veterans The NHIES 2009/10 revealed that 11.6 must either not be employed or, if employed, percent of children between the age of 6 and receive less than a prescribed income. At 13 had never been to school. To deal with N$2,200 a month, the veterans’ subvention this challenge and promote primary school has the highest value of the social grants by enrollment among poor children, Namibia far. introduced a no-fee primary education The Child Maintenance Grant provides (Universal Primary Education) in 2013.14 20 support to children with either a disabled The plan called for the government to cover school fees until students completed primary $20,000, or a maximum joint income of school or reached age 16.15 Free secondary N$30,000 per month. Collateral of 20 school began at the start of 2016. percent or a deposit of 5 percent is required, Financial support is available for higher which means poor households are less likely education. The programs include the to benefit from the initiative. The NHE Namibia Students Financial Assistance Fund obtains its funds from the capital markets. (NSFAF), a loan/grant scheme that supports From 1990 until 2011, the Government gave students in specific priority areas of study. subsidies of N$56.7 million to the NHE. Additional income was generated through Health the development and financing of houses and Public spending on health is relatively large. other loan products. Through the Ministry of Health and Social The Build Together Program targets Services (MHSS), the state is the main households earning less than N$3,000 per provider of health care services in Namibia. month. It consists of four elements: (a) The system is funded via general taxation the urban/rural housing loans scheme; (b) and serves 85 percent of Namibians. MHSS the social housing scheme; (c) the single expenditures were close to N$2.5 billion quarter transformation scheme; and (d) in 2009/10, accounting for around 9.8 the informal settlement upgrading scheme. percent of total public spending. A vibrant Under the urban/rural housing loans scheme, private health sector funded largely through households with low or very low incomes employee and employer contributions exists, receive assistance to build their own houses. parallel to the public sector. The program is administered by the Ministry Although user charges apply to public of Regional, Local Government, Housing health facilities, primary healthcare is and Rural Development (MRLGHRD). subsidized to ensure access by many The rural water infrastructure and households. Despite the public subsidies, services program provides a proxy for water user charges still serve as a barrier to access to subsidies. The program, administered by the health services by the poor. Directorate of Rural Water Supply in the Ministry of Agriculture, Water and Forestry C. Non-social spending benefits (MAWF), benefits only rural residents with (indirect subsidies) assessed access to an improved water source. Namibia in this study is regularly afflicted by droughts, and large rivers are far from the population centers In addition to direct transfers and in-kind and large water users in manufacturing transfers in the form of education and health, and mining establishments. To overcome the incidence analysis considers indirect water shortages, the country has built dams, subsidies. Specifically, it looks at housing and pipelines, potable water re-use systems, water subsidies. The two housing programs and seawater desalination plants. Broadly, in the analysis target low- and medium- Namibia has given control of water supply income households: the National Housing and sanitation to the Department for Water Enterprise (NHE) and the Build Together Resources Management and the Department Program (BTP). for Rural Water and Sanitation Coordination The NHE targets households with in MAWF. monthly incomes between N$5,000 and 21 Chapter 3: Methodology and Data The study’s primary methodology comes from to fiscal incidence analysis.18 It takes place in the Commitment to Equity (CEQ) project an ex-post static setting. The analysis begins (Lustig and Higgins, 2013). The project uses from a “before” or “pre-fiscal” income and a comprehensive fiscal incidence analysis and allocates a tax or transfer to each household a diagnostic framework to assess the poverty or individual. If the fiscal intervention is a reduction and redistributive effects of taxes direct tax (transfer) and the analysis starts and public social benefits. This chapter from pre-tax (pre-transfer) income, the post- provides an overview of this methodology, tax (post-transfer) income is calculated by describes the data used, and discusses the subtracting (adding) the tax paid (transfer assumptions made in the context of data received). The accounting approach does not limitations. take into account behavioral responses that taxes and public spending may trigger in A. What is fiscal incidence individuals or households. analysis? This report makes both partial and comprehensive assessments of Namibia’s Fiscal incidence Fiscal incidence analysis consists of allocating current fiscal system. Partial fiscal incidence analysis examines the burdens of taxes and the benefits of analysis looks at the impact of one or several who ultimately public spending to households or individuals, fiscal policy interventions – for example, bears the burden of making it possible to compare incomes income taxes or the use of public education government taxes and before and after taxes and transfers. The and health services. Comprehensive fiscal who benefits from public spending CEQ methodology makes the allocations incidence analysis assesses the impact of the using household level micro-data. Incomes revenue and spending sides simultaneously, after taxes and transfers may include the measuring the overall impact of direct and monetized value or consumption of free indirect taxes, cash and in-kind transfers, public services. The allocations can be and indirect subsidies. For taxes, this report analyzed to determine how a government’s estimates actual average incidence and revenue-generation and expenditure activities effective average rates (i.e., it uses the average redistribute income among the population.16 rate of tax collection including possible tax Using this approach, it is possible to measure evasion). It values in-kind benefits according the fiscal system’s redistributive impact to the “government cost” approach, assuming through its effects on poverty and inequality. the full cost of a public service is borne by the As long as there is sufficient detail in the government. household survey, the same allocation can We follow Lustig and Higgins (2013) be used to assess the impact of fiscal policy and measure per capita income before and on the welfare of different social groups – for after each set of fiscal interventions (Figure example individuals differentiated by gender, 16). The “before” and “after” measures are ethnicity, or location.17 referred to as income concepts. For example, This report uses the “accounting” approach all earned and unearned income from any 22 source is called “market income,” a measure paid, yielding a measure of the resources of the resources households’ control “before” households control “after” direct taxes but any direct taxes have been applied. “Net “before” any direct transfers. market income” subtracts any direct taxes Figure 16. Definitions of income used in the CEQ fiscal incidence analysis BENEFITS Market Income TAXES Wages and salaries, income from capital, private transfers, contributory pensions - Personal income and payroll taxex Net Market Income + Direct transfers Disposable Income + - Indirect subsidies Indirect taxes Post-fiscal Income In-kind transfers (free government + - services in Co-payment, user education and fees health) Final Income Source: Lustig and Higgins (2013). 23 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? When it comes to poverty and inequality, by the bottom “x” percent of the population the impact of any fiscal component depends (ranked by reference income) is higher than on two factors: its magnitude and its its share in the population, a transfer is progressivity.19 For any measure of household absolutely progressive. Transfer shares are income, this study measures the progressivity higher for the poorest populations, and the of fiscal policy components (taxes and shares decline as income rises.22 transfers) by comparing the cumulative To illustrate, Figure 17 presents a Lorenz distribution and cumulative concentration curve for a reference “market income”. of the component before and after the Along the horizontal axis, the population is component has been received.20 The “before” ranked, poorest to richest, according to this income is called the reference income. A tax reference income. The vertical axis plots the (transfer) is progressive when the cumulative cumulative share of this income.23 Using the share of a tax paid (transfer received) by same ranking of reference income, we can the bottom or poorest “x” percent of the plot on the vertical axis cumulative shares of, population is lower (higher) than that group’s for example, taxes paid or transfers received; share in the pre-tax (pre-transfer) reference these are called concentration curves. income.21 If the share of a transfer received Figure 17. Diagram representing the progressivity of taxes and transfers Transfer progressive in absolute terms The progressivity of 45-degree line; population share Cumulative proportion of benefits, taxes or income taxes can be measured Transfer: neutral in absolute terms by comparing the Tax: upper bound of regressive share of a specific tax collected from each decile of the Transfer: progressive in population relative relative terms Tax: regressive to the share of total income each decile receives Transfer regressive Market Income Lorenz curve: Transfer or tax: neutral in relative terms if on this line Cumulative proportion of population (ordered by market income) Source: Lustig and Higgins (2013). 24 We use the following descriptions to equality (the 45-degree diagonal line), describe how a fiscal policy component the transfer is also progressive in absolute redistributes income: terms. The monetary amount received falls as income rises. • Progressive (regressive): A transfer (tax) with a concentration curve above (below) • Neutral: A transfer (tax) with a the Lorenz curve for the reference income concentration curve that coincides with but below the line of perfect equality (the the Lorenz curve for the reference income 45-degree diagonal line). The transfer or is neutral. tax is progressive only in relative terms.24 • Regressive (progressive): A transfer (tax) • Absolute progressive or “pro-poor”: When with a concentration curve below (above) the concentration curve for a spending the Lorenz curve for the reference income. program is above the line of perfect 25 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? Box 1: Caveats of the CEQ analyses and data limitations The fiscal incidence analysis applied here has some notable limitations: • The analysis does not take into account behavioral, lifecycle, or general equilibrium effects and focuses on average incidence rather than incidence at the margin. Our assumptions about tax shifting and labor supply responses are strong because they imply that both consumer demand and labor supply are perfectly inelastic. In practice, they provide a reasonable approximation and they are commonly used. • The analysis does not take the intra-household distribution of consumption into account. • The analysis does not explicitly take into account the quality of services delivered by the government. However, when fees for access (in health and education) are taken into account, benefits are adjusted to absorb some of those quality differences that are correlated with funding levels. • The analysis excludes corporate income, international trade or property taxes and spending categories such as infrastructure investments (including urban services and rural roads). • The analysis does not capture asset stocks or flows, so the income inequality impacts of savings and investment decisions (and the taxes, subsidies, and fiscal programs that shape those decisions) are not discussed. • The following limitations emerge from the NHIES data used in the analysis: • As suggested by very high levels of household indebtedness, market incomes could be misrepresented. The reliability of income data in the survey depends critically on whether the NHIES module includes an exhaustive possible sources of income (including private transfers). This is not necessarily the case. • As in other countries, the NHIES’ ability to capture comprehensive and reliable income information on households at the top of the distribution is limited. • NHIES coverage limitations are evident in the case of analysis of taxes. Only 2.3 percent of GDP worth of direct taxes are identified compared to 11.1 percent of GDP recorded in administrative records. Even after excluding the corporate taxes, there is a substantial share that is missing. We do not know the distribution of the missing taxes. We assume that under-coverage of the taxes should mostly affect the size of the decline in inequality. 26 B. Data, incomes, and income method is not feasible – for example, for components in-kind health transfers and the water and housing subsidies – we use inference, Incidence studies use both national accounts simulation, imputation, and alternative data and household survey data, or they rely sources to generate a “best guess” allocation on incidence indicators from secondary of benefits or taxes paid.27 sources, usually micro-data sets. This study Two important exceptions are the uses 2009/10 national accounts and budget “market income” concept – we use it as reporting data to estimate the magnitudes the reference income in the analysis – and of the social spending components received personal income taxes paid, which when by and the revenue collected from Namibian applied to market income, arrives at the “net households via different instruments. market income” concept (Figure 16). What NHIES 2009/10 is used as a primary micro- follows is a brief summary of assumptions we dataset to allocate, household by household, relied on to generate the reference “market these transfers and taxes. To determine the income” and the household level personal size of the transfer received or the taxes income tax burden. contributed at each income concept (Figure The NHIES applies an income module 16), the taxes and transfers from Namibia’s to all households selected for enumeration. national accounts and administrative fiscal This module contains a (relatively) complete data are allocated to individual households set of possibilities for market income sources: in the NHIES.25 The survey contains data wages and the value of non-wage benefits on household expenditures, cash transfers, from employers; income from real estate and utilization of educational and health and other assets; income from insurances services, collected from approximately 9,656 and remittances; and income from auto- representative households across the country production (among others). In the survey, over 12 months. Per capita values are obtained market income from these sources totals by dividing the total taxes paid or transfers about N$51 trillion, or approximately 76 received by the total number of household percent of average net national income in members defined as individuals who, during 2009/10.28 the reference period, were spending at least Market income from direct identification four nights in a week or at least two of the records indicates relatively high levels of four weeks of the survey in the household. indebtedness among low income households. For most of the indirect taxes, direct Figure 18 indicates, for example, that cash transfers, and in-kind transfer items total market income in the poorest 10 in this analysis, the NHIES allows for percent of households in the NHIES direct identification to determine whether survey is approximately 16 percent of total a household received a benefit or paid a consumption expenditures. Such an income tax. Simply, the NHIES asks respondents and expenditure profile implies that some to indicate which transfers were received households do not balance their budgets and or which taxable or subsidized items were spend more than what they earn and receive consumed.26 Where the direct identification in transfers. 27 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? Figure 18. Income totals by decile (Namibian dollars, trillions) A. All deciles B. First 6 deciles only, re-scaled 40 1,5 30 1,0 20 10 0,5 0 0,0 2 3 4 5 6 8 9 7 t st es 2 3 4 5 6 st e or ch e or Po Ri Po Observed market income Observed market income Consumption expenditures Consumption expenditures Derived market income Derived market income Source: Authors’ calculations based on NHIES 2009/10. For this reason, the baseline incidence income, the same ratio for is 390. analysis takes as the reference income There is a relatively high discrepancy measure, a derived market income created by between observed market income and working backwards from total consumption consumption expenditures (Figure 18). expenditures.29 In other words, we ignore It is important to mention that “derived what households indicate they are earning market income” includes the value of auto- and begin with what they recall spending. production/auto-consumption. Even so, the From total expenditures we subtract out difference between derived market income the value of direct transfers (social grants) and observed market income is smallest for received – these can be directly identified in the poorest decile, the decile with a relatively the NHIES – to identify net market income. high share of consumption expenditures To this derived net market income, we add coming from auto-consumption/auto- back in personal income taxes paid – also production. Another concern that could directly identifiable in the NHIES – to derive arise with the use of derived market income market income. We add a “balanced budget” is that richer households typically save a condition that does not allow derived market considerable percentage of their income, income (or net market income) to be less than implying their incomes will be considerably zero. Figure 18 indicates that this derived understated by the derived market income market income measure is approximately variable. This suggests a need to adjust for 100 percent greater than the observed savings. A common way to do this is to use market income for the poorest 40 percent the marginal propensity to save (MPS) to of the population and approximately half adjust the derived market incomes higher, the size of observed market income for the based on other survey estimates of the MPS. richest 10 percent of the population. Derived However, in this particular context, Figure market income appears much more equally 18 indicates that the marginal propensity to distributed than observed market income. save is not linear or log-linear particularly in For derived per capita income, the ratio of observed income, which means applying a the tenth to the first decile is 88; for market single MPS to the entire income distribution 28 might not be adequate. Further, the decision expenditure proxy has implications for the to use derived market income rather than measurement of poverty and inequality as well reported incomes was based on discussions as for fiscal components’ impacts on welfare. with officials at NSA that indicated that Take the bottom 20 percent of households consumption expenditures were the better ranked by observed market income: about estimate for NHIES 2009/10 enumerated 40 percent rank higher when derived market households.  The NHIES exhibits several income is used. In fact, approximately 14 weaknesses when it comes to income percent appear in the top half of the derived reporting and these include under- market income distribution.30 represented households (the very rich ones); Fiscal policy instruments exhibit missing households (survey non-response); different profiles within the two different under-represented income sources (private reference income distributions that are not within-family transfers, for example); rank-preserving. Figure 19 summarizes total missing sources of income (item non- personal income taxes paid by decile for both response); recall error or bias, among others. the observed and derived market income Nonetheless, it is important to acknowledge distributions. Personal income taxes (PIT) this methodological limitation as use of paid generally do increase as incomes rise, derived market income could underestimate but the distribution of the PIT burden falls the extent of income inequality. more heavily on low income households in Deriving market income from an the derived market income distribution. Figure 19. Personal income tax collection total (of total reference income) A. All deciles B. First 6 deciles only, re-scaled 10% 1.0% 0.9% 8% 0.8% 0.7% 6% 0.6% 0.5% 4% 0.4% 0.3% 2% 0.2% 0.1% 0% 0.0% 2 3 4 5 6 7 8 9 Poorest Richest Poorest 2 3 4 5 6 Observed market income Derived market income Observed market income Derived market income Source: Authors’ calculations based on NHIES 2009/10. We use the derived market income observed market income components, measure as reference income in the analysis especially for those households with less – summarized here. The expenditure proxy or less reliable – income from the observed is likely measured with less error than the market income components. 29 Chapter 4: Incidence and Progressivity of Taxes This chapter examines the progressivity or indirect taxes is assessed. The government regressivity of selected taxes in Namibia. It imposes direct taxes on individuals and is important to note that the study’s overall organizations – e.g., income taxes, corporate objective is to estimate the poverty and taxes and the like. In the framework of redistributive effects of fiscal policy; it does this study, we analyze personal income not offer a full analysis of whether specific taxes (PIT) as the main direct tax affecting taxes or expenditures are desirable. Good individuals. Analysis of corporate and wealth tax policy will include a range of revenue taxes are beyond the scope of this analysis. collection instruments that produce a desired Indirect taxes are applied to the sale of goods revenue level with minimal distortions and and services. Goods and services providers low administration costs. Public spending collect VAT and fuel levy taxes from the should aim to provide the minimal functions end user to subsequently pass the proceeds of a state (such as security) and to invest in onto the government. Direct and indirect public goods (such as infrastructure) that are taxes affecting end users are first analyzed necessary to ensure prosperity. separately and then looked at collectively. The distributional impact of direct and Figure 20. Tax revenue, cross country comparison (percent of GDP) 70 Tax revenue as a percentage of GDP 60 50 40 30 Namibia Global average = 20 16.3% 10 0 Source: World Development indicators, 168 countries (2010 or last available year if 2010 data are not available). 30 Namibia has relatively high tax-to-GDP A. Direct taxes32 ratio in comparison to other developing countries. Wide national variations exist Incidence analysis shows that the Namibian across countries (Figure 20). At the upper PIT is progressive – the shares of PIT paid end are Lesotho, Algeria and Seychelles, increase with reference income shares.33 which have a tax-to-GDP ratio above 30 Figure 21 demonstrates this, using percent, while at the lower end are Middle concentration curves. Households are ranked Eastern economies with less than 5 percent from poorest to richest (according to the tax-to-GDP ratio. The corresponding average reference income) from left to right along the was 16.3 percent in 2010. Namibia tax-to- x-axis. The curves are generated by plotting GDP ratio (24 percent) was more than 50 the cumulative share of a tax paid (or a transfer percent above the average of the developing received) accounted for by all the households countries.31 The relatively high share of tax to the left of any point on the x-axis (see revenues collected by the government in also Chapter 3). Figure 21 shows that the Namibia reflects the relatively wide range of poorest 80 percent of the population (ranked taxes. As Namibia’s level of the taxation is by derived market income plus pensions) relatively high for an average African country, accounts for approximately 11 percent of this is the time to analyze incidences of the total PIT collections reported in the NHIES. taxes. The poorest 90 percent of the population accounts for approximately 30 percent of all PIT collections, leaving about 70 percent of PIT collections within the richest 10 percent. The poorest 90 percent of the Figure 21. Concentration curves for PIT (share of total tax paid by reference Namibian population income deciles) accounts for about 30 percent of all 100% personal income tax collections Cumulative proportion of PIT 80% 60% 40% 20% 0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Cumulative proportion of the population (ranked by market income + pensions) 45-deg. line PIT Market income + pensions Source: Authors’ estimates based on NHIES 2009/10. 31 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? The progressivity of PIT in Namibia these taxes in total public revenue varies from compares favorably to others in the CEQ country to country. For example, the top country set. Figure 22 summarizes the decile in South Africa pays over 80 percent of incidence of personal income taxes (and total PIT collected, and total PIT collected other direct taxes) relative to reference market represents about 14 percent of GDP. Brazil income for Namibia and four other countries collects two-thirds of PIT from the top in the CEQ country set. Direct taxes are decile, and total PIT receipts represent about progressive everywhere, although the share of 12 percent of GDP. Figure 22. Concentration shares: Figure 23. Progressivity of direct tax personal income taxes system: Kakwani coefficient (Share of taxes by market income deciles) 90% 0.30 80% 0.28 0.27 0.25 PIT relative to market income 70% 0.23 0.23 60% Kakwani index 50% 0.13 40% 30% 20% 10% 0% (2009) (2009) Namibia Ethiopia Uruguai Mexico Armenia (2010) Brazil (2009) (2011) (2010) (2010) Peru Armenia Brazil Ethiopia South Namibia Africa Sources: Armenia (Younger et al., 2014), Brazil (Higgins and Sources: Armenia (Younger et al., 2014), Brazil (Higgins Pereira, 2014), South Africa (Inchauste et al., 2015), and and Pereira, 2014), South Africa (Inchauste et al., 2015), Namibia (own calculations based on NHIES 2009/10). Ethiopia (Woldehanna et al., forthcoming), Indonesia (Jellema et al., forthcoming), Mexico (Scott, 2014), Peru (Jaramillo, Note: Darker bars at left represent poorer deciles; lighter deciles 2014), and Namibia (authors’ calculations based on NHIES at right represent richer deciles. 2009/10). B. Indirect taxes plus direct transfers, net of direct taxes.34 Indirect taxes as a whole35 are more or The indirect taxes covered in the study are the less neutral. At the bottom of the income VAT and fuel levy. Together, their revenues distribution, the cumulative share of total equal 6.7 percent of GDP. VAT accounted indirect taxes paid is approximately equal to for 32.6 percent of total tax collections in each decile’s share of disposable income. For 2009/10 (excluding SACU receipts and deciles two through six, the share of indirect grants), and the general fuel levy contributed taxes paid exceeds disposable income shares, about 0.5 percent. We assess the incidence meaning only the richest decile pays a share of indirect taxes with respect to disposable of indirect taxes that is smaller than its share income, defined as the sum of market income of disposable income (Figure 24). 32 Figure 24. Concentration curves of indirect taxes (share of total tax paid by reference income deciles) 100% 90% Cumulative proportion of indirect taxes 80% 70% 60% 50% 40% 30% 20% 10% 0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Cumulative proportion of the population 45-deg. line VAT + Fuel Levy Disposable income Source: Authors’ estimates based on NHIES 2009/10. The poorest 90 C. Progressivity of total taxes The overall incidence of direct and percent of the (direct and indirect) indirect taxes is progressive; i.e., shares of total Namibian population accounts for about taxes paid increase with reference income 40 percent of all tax Combined direct and indirect taxes shares. The concentration curve presented in collections (excluding SACU receipts) in Namibia are Figure 25 combines information in Figure 21 about 17.8 percent of GDP, making it one of (PIT) and Figure 24 (indirect taxes). It shows the largest total tax burdens of any country that the poorest 75 percent of the population with a CEQ assessment. Most countries with (ranked by derived market income) accounted a similar-sized indirect tax take (e.g. Ethiopia) for approximately 20 percent of total direct have a much smaller direct tax take, and most and indirect taxes. The poorest 90 percent of countries with a similar-sized direct tax take the population accounted for approximately (e.g. South Africa) have a smaller indirect tax 40 percent of total tax collection, leaving take. about 60 percent of the total taxes within the richest 10 percent. 33 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? Figure 25. Concentration curves of direct and indirect taxes (share of total tax paid by reference income deciles) 100% Cumulative proportion of taxes paid or disposable income 80% 60% 40% 20% 0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Cumulative proportion of the population (ranked by disposable income) 45-deg. line Market income + pensions (ranked by market income + pensions) Total (VAT + PIT + Fuel Levy) (ranked by market income + pensions) Source: Authors’ estimates based on NHIES 2009/10. 34 Chapter 5: Incidence and Efficiency of Social Spending A. Incidence and efficiency of effectiveness also depends on the efficiency direct transfers of administration in terms of identifying beneficiaries and delivering benefits. This The effectiveness of social protection section assesses the performance of the direct programs in reducing poverty depends on transfers with respect to: (i) progressivity; whether they cover a significant number of (ii) coverage; (iii) targeting accuracy; (iv) the poor people and whether they are adequate. generosity of benefits; and (v) the impact on The key issues are: Do benefits go mostly to poverty. Box 2 provides definitions of the the poor? Are they adequate to significantly performance indicators.36 reduce the consumption gap? Cost Box 2: Main indicators of performance and targeting efficiency of direct transfers Based on the World Bank’s study “Targeting of Transfers in Developing Countries”37, economic growth is a necessary but insufficient condition for the alleviation of poverty. The asset base of poor households needs to be built up so that they can participate in the growth process. Growth needs to be more intensive in the sectors in which the poor predominate. Short-term public transfers are required to protect and raise the consumption of the poorest households. Implementation of this agenda for reducing poverty requires methods for reaching the poor. In part, this can be accomplished by spending on items such as universal primary education that reach a wide segment of society, including the poor. It also can be accomplished by providing targeted resources directly to the poor. Targeting is a means of increasing program efficiency by increasing the benefit that the poor can get within a fixed program budget. Several methods exist in the social protection literature which ensure resources are directed to the poor. A brief summary of the individual/household assessment, simple means tests, proxy means tests, community based-targeting, and categorical targeting are described in the Annex 2. The main indicators of performance of social assistance cash transfers include: • Coverage: What share of the population receives the transfers, with a focus on the share received by the poorest quintile? • Targeting accuracy: What share of the transfer goes to each quintile, with particular focus on the share of transfers going to the bottom quintile? • Generosity: How much is the transfer as a fraction of post-transfer disposable income or consumption? If this fraction is large, it would imply that the household is getting its income primarily from this transfer. 35 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? • Progressivity index: An alternative method to measure progressivity is the Kakwani Index, calculated as twice the area between the market income Lorenz curve and the tax (transfer) concentration curve. If the tax (transfer) concentration curve is below (above) the Lorenz curve, the Kakwani index will be positive, indicating that taxes (transfers) are progressive. If the tax (transfer) concentration curve is above (below) the Lorenz curve, the Kakwani index will be negative, a signal of regressive taxes (transfers). • Impact of the transfer on poverty and inequality: To what extent does the transfer lift people out of poverty and reduce inequality? To measure this impact, the amount of the transfer is removed from households’ consumption, leading to estimates of how many more individuals would be poor in the absence of the transfers and how the inequality would change. • Under-coverage and leakage: A common approach to evaluating the targeting performance of alternative transfer instruments is to compare under-coverage and leakage rates. Under- coverage is the proportion of poor households that are not included in the program (errors of exclusion). Leakage is the proportion of those who are reached by the program who are classified as non-poor (errors of inclusion). Namibia exceeds the average spending countries (1.6 percent). Namibia’s spending on public transfers for Sub-Saharan African level is comparable to that of Argentina countries in overall spending on direct and Poland, but significantly below that of transfers. Spending on direct transfers is South Africa (3.5 percent) and Mauritius higher than the average for Sub-Sahara (3.3 percent). The Old Age Pension (OAP) African countries (1.46 percent) (Figure 26). constitutes 59 percent of spending on direct It is comparable to the average for developing transfers (Figure 27). Figure 26: Spending on social Figure 27: Distribution of spending protection by country (percent of on direct transfers by type of program, GDP) 2012/13 8 Disability Disability C grant 7 A grant 6% 12% 6 Foster parents' Share of GDP 5 grant 6% 4 3 Children's Old age grant pension Average for 11% 59% 2 developing countries = 1.6% Namibia 1 Veteran's grant 6% 0 Source: Aspire database, the World Bank and Government of Namibia for Namibia. 36 Despite generous spending, the targeting Targeting efficiency of direct transfers in efficiency of direct transfers is relatively Namibia was 16.9 percent compared to the low. International comparison suggests that 26.6 percent world average. In other words, Namibia is below the world average both in despite generally high levels of government terms of coverage of the direct transfers going spending on grants, both coverage and to the poorest population and in terms of targeting efficiency are relatively low in grants’ targeting efficiency. Figure 28 presents Namibia in comparison to other countries. coverage of the programs going to the poorest The lack of targeting efficiency of Namibian quintile of the income distribution for 104 grants is well illustrated in comparison to countries worldwide. The coverage for the South African grants. Both countries spend poorest quintile was 33 percent in Namibia, a significant amount of resources on transfers below the 43.1 percent world average. Figure as a share of GDP, while most of the South 29 illustrates the targeting efficiency of African grants are proxy means tested which direct transfers, expressed as a percentage of significantly improves the targeting efficiency the transfers going to the poorest quintile. of grants. Figure 28. International comparison: coverage of the transfers (poorest decile) 100 90 South Africa = 85.1% Percentage of poorest quintile population 80 70 60 50 Spending on direct Average = 43.1% transfers is quite 40 generous but is 30 Namibia = 33% characterized 20 by relatively low 10 targeting efficiency 0 Figure 29. International comparison – targeting efficiency of the transfers 70 S ha re of be ne fi ts goi ng to the l ow e s t qui nti l e 60 50 40 South Africa = 29% 30 Average = 26.6% 20 Namibia = 16.9% 10 0 Source: Authors’ calculations for Namibia based on NHIES 2009/10. ASPIRE database (The World Bank) for other countries (104 countries). 37 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? Each direct transfer covered in this study are much more progressive. is weakly progressive. In the top two panels (A A comparison to South Africa indicates and B) of Figure 30, the concentration curves that the majority of the South African present the distribution of the consumption grants are strongly progressive – in contrast per capita on the far right from the diagonal to Namibian grants that are merely neutral line of perfect equality. The deep bowing or slightly progressive (Figure 30, panels C confirms the very high level of inequality and D). South Africa and Namibia are very in Namibia. However, almost all grants are similar countries in terms of the high level of very close to the 45-degree equality line. This income inequality and availability of a wide means that the transfers, analyzed separately, range of direct transfers. However, South are weakly progressive in relative and in Africa uses proxy means tests for most social- absolute terms (i.e., the transfers represent a assistance programs, and this could be driving larger share of income among lower deciles the observed strong progressivity. If so, proxy in relative terms). All the analyzed grants means tests could be explored in Namibia are generally similar in terms of their weak as a way to improve the progressivity and progressivity. However, South African grants efficiency of social assistance programs. Figure 30: Progressivity of direct cash transfers by category: concentration curves for transfers and Lorenz curves for market incomes, Namibia and South Africa Panel A: Namibia Panel B: Namibia 1 1 0,9 0,9 Cumulative share of bene ts 0,8 0,8 Cumulative share of bene ts 0,7 0,7 0,6 0,6 0,5 0,5 0,4 0,4 0,3 0,3 0,2 0,2 0,1 0,1 0 0 0,2 0,4 0,6 0,8 1 0 0 0,2 0,4 0,6 0,8 1 Cumulative share of population, ranked from poorest to richest Cumulative share of population, ranked from poorest to richest Line of equality Consumption per capita Line of equality Old Age Pension (social grant) Consumption per capita Veteran's grant (social grant) Foster parents' grant (social grant) Disability A grant (social grant) Children's grant (social grant) Disability C grant (social grant) Panel C: South Africa Panel D: South Africa 100% Lorenz for Market Income 100% Lorenz for Market Income Direct Transfers Direct Transfers 90% Old -age pension 90% Child foster care Child support grant 80% 80% Grant-in-aid Disability grant 70% Other grants 70% Population Shares Population Shares 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% 0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10 38 Source: Authors’ calculations based on NHIES 2009/10 for Namibia. For South Africa: World Bank (2014). The coverage of direct transfers is slightly The “all direct grants” category combines progressive in terms of targeting, but they all cash transfers; if a household has access reach a relatively small proportion of the to any of the grants, it will be defined as a population (Table 4). We define “coverage” receiver. Almost one-third of the population as the proportion of people in the respective (27.2 percent) has either direct or indirect population group having access to the access through family members’ access to the specific grant or group of grants. Both direct transfers. The targeting efficiency indicator and indirect beneficiaries are included in is generally progressive: 33 percent of the the analysis – in other words, the whole poorest quintile receive at least one transfer, household is defined as a beneficiary, if and coverage declines to 13.1 percent among any individual in it has access to the grant. the top quintile. Table 4. Coverage (targeting efficiency) of direct transfers     Quintiles of per adult equivalent consumption   Total Q1 Q2 Q3 Q4 Q5 Direct and indirect beneficiaries All direct transfers 27.2 33.0 35.9 31.5 22.4 13.1 Old Age Pension 17.2 22.2 22.6 19.6 14.7 6.8 Children’s grant 6.3 6.6 9.4 7.7 5.3 2.6 Veterans grant 1.3 2.0 1.7 1.6 0.3 1.0 Foster parents grant 2.0 2.3 3.3 1.3 1.2 2.0 Disability grant – adult 4.2 6.3 5.7 4.4 3.3 1.1 Only close to a third Disability grant – child 1.3 0.7 1.3 0.6 1.8 2.2 of the population Source: Authors’ calculations based on NHIES 2009/10 for Namibia. receives direct transfers, directly or Figure 31. Targeting accuracy, coverage, Figure 32. Progressivity of all direct indirectly and the and generosity - all direct transfers transfers poor are more likely to receive direct 100 transfers 1 90 0.9 Cummul a tiv e s ha re of the be ne f its 80 0.8 70 66.1 0.7 Cumulative share of bene ts 60 0.6 50 0.5 40 0.4 30 33.0 0.3 20 20.2 0.2 16.9 13.4 10 13.1 0.1 0 0 Q1 Q2 Q3 Q4 Q5 0 0.2 0.4 0.6 0.8 1 Cumulative proportion of population, ranked from Coverage poorest to richest Targeting Accuracy Line of equality Generosity Consumption per capita All social assistance Source: Authors’ calculations based on NHIES 2009/10. 39 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? Direct transfers matter for the poor. The of transfers going to each income quintile. coverage and generosity of direct transfers The share of all direct transfers going to the as a whole is slightly progressive (Figure poorest quintile, or targeting accuracy, was 31). With respect to coverage, progressivity 16.9 percent in 2009/10 (Figure 31). The means that the poor were more likely to share of benefits going to the richest quintile receive direct transfers in 2009/10. Not only was 20.2 percent. Direct transfers may matter was the coverage progressive, the transfers for the poor in other ways, but their targeting tended to be generous to the poor, making accuracy is low. The finding of regressive up a larger share of their total income. targeting accuracy and slightly progressive Households in the lowest quintiles of the coverage suggests the value of the transfers consumption distribution received up to 66.1 is significantly larger among the rich. This percent of their income from direct transfers is even more evident from the Lorenz curve (generosity); the share was below 20 percent analysis in Figure 32 where the line for direct for the top quintile. On average, generosity of transfers nearly lies on the 45-degree line. direct transfers was 27.2 percent in Namibia This differs from the progressive distribution in 2009/10.38 of direct benefits in South Africa, where The targeting accuracy of direct transfers the poorest quintile received more than 25 is not as progressive despite progressive percent and richest less than 11 percent coverage and generosity as a whole. The (authors’ calculation). targeting accuracy is measured by the share The share of all direct transfers going to the poorest quintile, was 16.9 percent in 2009/10 while it was 20.2 percent for the richest quintile 40 Figure 33: Targeting accuracy, coverage, and generosity of each direct transfer Old age pension Children's grant 100 100 90 90 80 80 70 70.9 70 60 60 Percent Percent 50 50 40 40 30 22.2 30 26.0 20 14.0 20 12.2 16.9 10 19.2 10.2 10 8.4 6.8 6.6 0 0 2.6 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Coverage Coverage Targeting accuracy Targeting accuracy Generosity Generosity Veterans' grant Foster care 100 100 90 90 80 80 70 70 60 60 51.2 Percent Percent 50 50 40 40 30 22.4 30 26.1 28.2 24.0 20 20 14.4 10 11.2 10 6.1 2.0 1.0 2.3 2.0 0 0 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Coverage Coverage Targeting accuracy Targeting accuracy Generosity Generosity Disability grant - adults Disability grant - children 100 100 90 90 80 80 70 70 71.8 60 60 Percent Percent 50 50 40 39.1 40 30 30 27.1 20 19.7 20 22.6 13.7 10 13.2 10 1.9 6.3 1.1 2.2 0 0 0.7 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Coverage Coverage Targeting accuracy Targeting accuracy Generosity Generosity Source: Authors’ calculations based on NHIES 2009/10. 41 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? Analyzing the direct transfers, one- example, veterans’ grants went to 2 percent by-one, reveals a similar picture. All direct of the lowest quintile and 1 percent of the transfers are progressive in terms of coverage top quintile. For adult disability grants, the and generosity, but their targeting accuracy corresponding figures were 6.3 percent for could be improved (Figure 33). the first quintile and 1.1 percent for the top The Old Age Pension (OAP) has the quintile. A similar pattern is observed for highest coverage, and it is progressively foster care grants. targeted. An average of nearly 17.2 percent In Namibia, proxy means tested programs of Namibians receive the OAP. The grant could reduce poverty and inequality. is progressively distributed, going to 22.2 Introducing the Proxy Means Test (PMTs) is percent of the poorest quintile and 6.8 expected to help Namibia make substantial percent of the top quintile. progress toward achieving the NDP goals Children’s grants are given to a relatively as well as offering insights to the strengths small portion of the population, and their and weaknesses of different approaches to targeting efficiency is limited. On average, poverty and inequality reduction. However, 6.3 percent of the population receive PMTs are operationally complex, requiring children’s grants, with coverage of 6.6 percent considerable investment and institutional in the poorest quintile and 9.4 percent in the capacities. Targeting cash transfer and second quintile. However, 2.6 percent of the compliance verification to conditions require richest quintile receive the grant. This pattern highly developed, efficient systems. Box 3 holds in most countries because child related presents the rationale for introducing PMT grants are self-targeting for big families with programs in Namibia, drawing on lessons Direct transfers are a lot of children. These households tend to from PMT in Indonesia. The box also progressive in terms be poorer, making targeting more efficient. indicates administrative and methodological of coverage and With a program coverage size of 6.3 percent, challenges associated with the introduction generosity, but their which is relatively small, one would hope of PMT programs. We have not actually targeting accuracy that the coverage of the poorest quintile is compared the PMT based distribution and could be improved much higher than any other quintile, but at the current distribution. The basis for thinking 6.6 it is barely above the national average, is that PMT could improve the outcomes is quite a bit lower than Q2 (9.4) and Q3 (7.7), based on cross-country experiences. It is also and is not much higher than Q4 (5.3). These important to emphasize that a combination results are not that surprising, given that of measures should be considered as a long there is no PMT-style targeting in Namibia. term solution for Namibia. A whole range The veterans, foster care, and disability of targeting mechanisms (community, grants are generally small in terms of coverage PMT, self-targeting, etc.) should be further and vary in terms of their progressivity. On analyzed. Assuming PMT is probably a average, coverage was 1.3 percent of the preferred method, additional research in population for veterans’ grants, 2 percent this area should be undertaken, suggesting for foster care grants, and 5.5 percent for appropriate methods that could improve disability grants. The poor generally are the targeting efficiency of direct transfers in more likely to receive these grants. For Namibia. 42 Box 3: Advantages of proxy means tests for targeting social programs: the case of Indonesia Many low- and middle-income countries have introduced conditional cash transfers (CCTs) based on the proxy means testing methodology. Among the countries with good examples of well-designed targeted social protection systems are South Africa, Chile, and Brazil. CEQ analysis indicates progressive patterns of CCTs. Indonesia has also developed a well-targeted social assistance system. As Indonesia matures into a middle-income country, the government is trying to improve social assistance as part of its efforts to reduce poverty. The country now offers households a number of social assistance programs, including subsidized rice, health-fee waivers, cash transfers for poor students, a pair of CCTs, and a temporary unconditional cash transfer. Indonesia’s government introduced two complementary CCT programs. The household CCT gives quarterly transfers to poor households identified through a PMT system. The community CCT focuses on communities rather than individually targeted households. The PMT methodology estimates household income by associating indicators with household expenditure or consumption. Proxy means testing uses multivariate regression to correlate certain proxies, such as assets and household characteristics, with poverty and income. The World Bank recently completed a CEQ analysis of the poverty and redistributive impacts of Indonesia’s transfers system. Figure 34 presents the results for the proxy means tested conditional transfers to households. Figure 34: PMT conditional cash transfers in Indonesia - concentration curves for transfers 100% Cumulative Proportion of Market Income or Tax 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Cumulative Proportion of the Population 45 Degree Line Lorenz for Market Income Conditional Cash Transfer Source: Authors’ calculations based on Indonesia National Socio-Economic Survey 2012. The concentration curve for the CCT is significantly above the 45-degree line of equality, indicating the progressivity of the transfers. The program has a very substantial impact on poverty and income inequality in Indonesia. The transfer raises the income of poor households and increases their consumption, reducing poverty and inequality. PMT transfers are considered more efficient and more effective than other types of government transfers. 43 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? Implementing proxy means testing significantly improves targeting of the social protection system, but the method has a number of challenges. Significant administrative costs are associated with introduction and implementation of PMT programs. PMT methodology is based on national household survey data, and the enumerators are not always objective when conducting surveys and do not always have time to verify proxies within households. Some households’ characteristics are difficult to verify. The method does not take into consideration financial shocks and distress when households fall in poverty without observing a change in the household characteristics. To summarize, the foregoing discussion amount is defined as the public expenditure shows that the magnitude of government on this program in 2009/10, divided by spending on direct transfers is adequate in the total number of rural households with Namibia, but targeting efficiency is relatively access. Both the BTP and NHE are allocated low. The transfers are important to the randomly to homeowners within the NHIES beneficiaries, especially to the poor in that dataset who meet the programs’ eligibility they make up a large share of their total rules. The value of benefits is defined as the income. The targeting accuracy of transfers preferential financing subsidy relative to is low and could be improved by introducing expected market or traditional lender finance PMT programs. Targeting accuracy is low rates. mostly because the transfers are categorical Taken together, the three subsidies are (elderly for pensions, child grants), which progressively targeted, driven by the water Indirect subsidies automatically makes richer households subsidy (Figure 35). The three subsidies are progressively eligible. Introduction of PMT schemes could have a relative- to absolute-progressive targeted, driven by improve targeting efficiency in Namibia, incidence in that the poorest and middle- the water subsidy making them more pro-poor. However, it does income households receive subsidy shares take time, money and capacity to introduce that are larger than their shares of disposable PMT schemes.  For the money the Namibian income.39 However, this overall progressivity government spends on social assistance, and is driven by the water subsidy; it accounts given a very ambitious inequality reduction for nearly three-quarters of all subsidy target, the PMT programs would have much expenditures analyzed in the study. The better targeting outcomes and greater cost BTP accounts for nearly one-quarter, and effectiveness. the NHE for approximately 2 percent. The water subsidy’s absolute progressivity (or “pro B. Incidence and efficiency of poor-ness”) are not surprising. The subsidy is indirect subsidies allocated only to rural households, and the majority of poor and low income households This section covers three indirect subsidy reside in rural areas. Both the BTP and programs – the rural water infrastructure NHE are active primarily in urban areas. and services program, the Build Together In this incidence analysis, these programs Program (BTP) housing subsidy, and the are restricted to urban areas and so are less National Housing Enterprise (NHE) housing likely than the water subsidy to be allocated subsidy. Only rural residents with access to to a poor household. In terms of benefits an improved water source are eligible for the delivered, the BTP is about 10 times larger water infrastructure program. The subsidy than the NHE; it is also more progressively 44 distributed than the NHE. BTP targets low one for NHE. The NHE indirectly targets income homeowners. All else equal, they a higher range of incomes by providing the present greater credit risks, so the below- same preferential financing on larger loan market financing BTP benefit for a similarly values (for more expensive houses). valued house is consequently larger than the Figure 35. Concentration curves for subsidies 100% 90% Cumulative proportion of subsidies 80% 70% 60% 50% 40% 30% 20% 10% 0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Cumulative proportion of the population 45-deg. line Water BTP NHE All Disposable income Source: Authors’ calculations based on NHIES 2009/10. The very poorest households (the C. Progressivity of in-kind bottom 10 percent of the disposable income transfers: education and distribution) account for approximately health 1 percent of BTP benefits and none of the NHE benefits. The second through to the Government expenditure data on publicly sixth deciles all have shares of BTP benefits provided education and health services are equal to or greater than their population used to calculate the “unit costs” of services shares (these deciles receive trivial shares of provided – costs per pupil enrolled or cost NHE benefits). Therefore, the larger BTP per outpatient visit, for example. We define program covers more poor households, but it the benefit received by individuals to be does not transfer benefits to the poorest of the equal to the amount spent per pupil or per poor. By virtue of its eligibility criteria and visit.40 While health and education services the population it indirectly targets through are typically viewed as investments in human larger available loans, the NHE program capital, we capture only the input side (or is primarily for middle-income and richer the current public sector purchase price) households. of the provision of these services. In other words, we allocate current public spending on education but not the future returns that such expenditures may generate. 45 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? Progressivity of education of education benefits becomes first relatively Spending on public primary schools is progressive and then regressive at the tertiary absolutely progressive, but the education level, where the bottom 70 percent of the system becomes less progressive as levels population receives just 10 percent of all rise from primary to tertiary. Figure 36 public expenditures. Total public expenditures shows that enrollment rates in public across all education levels are distributed primary schools are high in all income neutrally. Each group of households receives groups, and that the concentration shares of a benefit share approximately equal to its public primary education expenditures are population share. The benefits of public absolutely progressive, or “pro poor.” Low expenditure on education are conditional income households receive shares of primary on attending school, so it is not surprising education spending greater than their that the allocation of education benefits is population shares.41 Moving from secondary correlated with enrollment levels. to tertiary levels, however, the distribution Figure 36. Concentration curves, education (share benefit received by reference income deciles) 100% 90% Cumulative proportion of subsidies 80% 70% Spending on public 60% primary schools 50% is absolutely 40% progressive, but 30% the education 20% system becomes less progressive as levels 10% rise from primary to 0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% tertiary Cumulative proportion of the population 45-deg. line Primary Secondary Tertiary All levels Market income + Pensions Source: Authors’ estimates based on NHIES 2009/10. Education spending matters for level, utilization rates are approximately everybody, but spending on primary equal (and low) in the poorest and richest education has the highest positive impact deciles. While tertiary utilization is low for on utilization rates among the poor. Figure all household groups, the rate in the richest 36 shows that the overall near-neutrality in households is 28 times higher than the education spending obscures very unequal rate in the poorest households.43 Figure 37 access. At the primary level, public education demonstrates that primary enrollment is utilization42 rates are much higher in poorer nearly universal in Namibia. The observed 46 deciles than richer ones. At the secondary high primary enrollment rates among the poor could be suggesting that poorer CEQ approach used in this analysis is based households have more kids. Further, this on government cost of providing education could be reflecting the fact that wealthier and does not take into consideration issues of households opt for private schools given quality or the value that education provides that, with notable exceptions, Namibian to households. It is, therefore, possible that private schools are considered to be of better public spending is pro-poor but not as quality than public schools. The standard beneficial to the recipients. Figure 37. Percent of households utilizing public education, by reference income decile poorest decile richest decile 80% 70% 60% 50% 40% 30% 20% 10% 0% Primary Secondary Tertiary Health spending is Source: Authors’ calculations based on NHIES 2009/10. slightly progressive while middle class Progressivity of health households benefit expectation).  In the bottom 80 percent of Benefitting from public healthcare-related the population ranked by market income, most often from the expenditures requires access to healthcare the expected shares of health spending public healthcare service providers and, as in education, received (through expected utilization) are system utilization rates drive the progressivity of nearly equal to population shares – slightly public health expenditures in Namibia. The smaller than population shares in deciles one NHIES 2009/10 does not include enough through to five and slightly higher in deciles detailed healthcare-related questions to allow six through to eight. Based on administrative an estimate of household- or individual- data on total spending by facilities and total level utilization rates. An alternate survey patients served by facilities,  the per-visit – the Namibia Demographic and Health public cost for inpatient care is approximately survey, or DHS – was used to estimate an seven times higher than an outpatient visit.45  expected household level propensity to utilize Richer households generally have lower publicly provided healthcare services. Total propensities to choose publicly provided verified expenditures on public inpatient and healthcare services.46 outpatient healthcare services were allocated Taken together, expenditures on in- according to these propensities.44 kind education and health services are Health spending is slightly progressive approximately neutral. In both, poorer while middle class households benefit most households acquire greater shares of the often from the public healthcare system  (in low valued but more frequently provided 47 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? public services (primary education and public expenditures on in-kind health and outpatient healthcare). Richer households education services is approximately neutral. acquire greater shares of the high valued but Each population group receives a share of not as frequently provided public services benefits equal to its population share (Figure (tertiary education and inpatient healthcare). 38). The result is that the distribution of total Figure 38. Concentration curves, health and all in-kind (share benefit received by reference income deciles) 100% Cumulative proportion of in-kind transfer 80% 60% 40% 20% 0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Cumulative proportion of the population (ranked by market income + pensions) 45-deg. line Health Market income + pensions All in-kind Source: Authors’ calculations based on NHIES 2009/10. D. Tying it all together: Progres- only two regressive items among the fiscal sivity of the fiscal system instruments included in the analysis – the VAT system and tertiary education spending. This section summarizes progressivity of Tertiary education’s regressivity is driven by the fiscal system (taxes and expenditures) enrollment rather than program design – the based on the Kakwani index, a measure of very small number of students reaching the progressivity of programs. A negative index tertiary level are disproportionately non- suggests regressive programs. The higher the poor. Cash transfers (the social grants), in- index, the more progressive the programs. kind transfers (except for tertiary education), Figure 39 shows the Kakwani progressivity and water and housing subsidies (except index for taxes and transfers.47 for the NHE program) are all progressively The majority of the fiscal components are distributed. Therefore, the bulk of fiscal progressive in Namibia. Consistent with what policy is progressively distributed. has already been discussed, Figure 39 identifies 48 Figure 39. Progressivity of the fiscal instruments (Kakwani Index) Disability grant - children 1,09 Disability grant - adults 1,04 Old age pension 1,02 Direct transfers 1,01 Veteran's grant 0,99 Children's grant 0,93 Foster care grant 0,91 Primary school 0,75 Water subsidy 0,69 All transfers and subsidies (excl. pensions) 0,68 Health spending 0,68 Education, total spending 0,66 BTP housing subsidy 0,63 Secondary school 0,54 Personal income taxes (imputed) 0,20 Fuel Levy 0,18 NHE housing subsidy 0,03 Tertiary education -0,04 VAT -0,07 -0,2 0,0 0,2 0,4 0,6 0,8 1,0 1,2 Kakwani index Source: Authors’ calculations based on NHIES 2009/10. 49 Chapter 6: Impact of Fiscal Policy on Poverty and Inequality A. Overall impact of fiscal and indirect subsidies and subtracting taxes policy on poverty and (post-fiscal incomes). This translates to a inequality reduction of 24.7 percent in extreme poverty due to fiscal policy. In-kind transfers are not This chapter estimates the overall impact generally included in poverty calculations of Namibia’s fiscal policy on poverty and because households do not see a monetary inequality, using the accounting approach contribution. described in Chapter 3. This entails first Direct transfers drive poverty reduction, comparing a household’s market income – while the role of taxes and indirect subsidies the level before payment of taxes and receipt is less significant. The headcount ratio at the of benefits – to its income after all direct taxes national extreme poverty line shows that the have been paid (net market income). Market percentage of people below the poverty line On the whole, income is then compared to the income decreases by 6.8 percentage points by moving fiscal policy after direct taxes have been paid and direct from market income (22.2 percent) to reduces poverty transfers have been received (disposable disposable income (15.4 percent).50 The poor and inequality in income). Including indirect taxes and indirect pay very little in income taxes, which explains Namibia with direct subsidies gives post-fiscal income. Adding in- stagnation in poverty due to introduction of transfers driving the reduction in poverty kind benefits to households for education direct taxes to market incomes. Indirect taxes while the reduction and health yields final income, which takes and indirect subsidies together do not impact in inequality is driven into account all direct and indirect taxes paid poverty for disposable incomes. by in-kind transfers as well as all benefits received. Fiscal policy on the whole reduces On the whole, fiscal policy reduces inequality in Namibia, largely because of in- poverty in Namibia.48 Fiscal policy’s impact kind transfers. The Gini coefficient falls from on poverty and inequality is analyzed at 0.635 for market incomes to 0.590 for post- each income concept. Figure 40 summarizes fiscal incomes. If taken into consideration, the changes using national extreme poverty in-kind transfers in health and education levels49 (estimated at the national lower would have the highest redistributive effect. bound poverty line of N$277.54 per Including the monetized value of in-kind month in 2009/10 prices) and the Gini transfers would further reduce the Gini coefficient for inequality. Table 5 expands coefficient to 0.429. Approximately 78.2 the assessment to alternative indicators of percent of the Gini coefficient reduction poverty and inequality. Extreme poverty falls from market to final incomes is attributed from 22.2 percent before the introduction of to in-kind transfers, 16.4 percent to direct any fiscal policy measure (market incomes) transfers, and the remaining 5.4 percent to to 16.7 percent after adding direct transfers direct and indirect taxes and subsidies. 50 Figure 40: Poverty and inequality indicators at each income concept Inequality Market Income Poverty (Gini index) BENEFITS TAXES (%) Wages and salaries, income from capital, private transfers, contributory pensions 0.635 22.2 - Personal income and payroll taxex Net Market Income 0.628 22.2 Direct + transfers 0.594 Disposable Income 15.4 Indirect + - subsidies Indirect taxes 0.590 Post-fiscal Income 16.7 In-kind transfers (free + government - Co-payment, services in user fees 0.429 education Final Income and health) Source: Authors’ calculations based on NHIES 2009/10. 51 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? Table 5. Poverty and inequality indicators for each income concept Market   Net Dis- Post- Income Final Market posable fiscal + Income Income Income Income Pensions Inequality indicators         Gini coefficient 0.635 0.628 0.594 0.590 0.429 Theil Index 0.830 0.819 0.760 0.750 0.385 90/101 16.8 15.4 10.8 10.6 5.9 Headcount poverty indicators         National extreme poverty 22.2 22.2 15.4 16.7   (lower bound poverty) National poverty 34.7 34.9 28.8 31.6   (upper bound poverty)2 $1.90 PPP 2011 22.0 22.1 15.3 16.6   Source: All data points based on own estimates based on NHIES 2009/10. Final incomes include in-kind transfers in health and education that are not included in the poverty calculations. 1 The ratio of the average adult equivalent consumption of the 90th percentile and the average of the 10th percentile. 2 The official national poverty rate estimated at the upper bound poverty line of N$377.96 is 28.7 percent. However, the present CEQ calculations find a rate of 28.8 percent at the disposable income concept. The difference - which is marginal - is statistical and due to households that were not included in the disposable income calculation because they did not record answers to all the other NHIES questions necessary to create the other income concepts.   60 percent of the population are net recipients of public Most Namibians are net recipients of for all but the very richest households. The transfers government transfers. Figure 41 shows that share of net contributors is generally much fiscal policy makes a positive contribution to higher in other countries. In Armenia market incomes up to the sixth percentile. In (2011), Bolivia (2009), Uruguay (2009), and other words, 60 percent of the population are Ethiopia (2011), for example, third-decile net recipients of public transfers. The ratio households are already net contributors to of net recipients to net payers/contributors public revenues (and all wealthier deciles is much larger in many other countries. At remain net contributors). For Sri Lanka final income concept, Namibia’s fiscal policy (2009) and Peru (2009), second-decile provides net positive transfers (on average) households are already net contributors. 52 Figure 41: Amounts by which incomes exceed market income (as a share of market income) 1,4 1,2 1,0 Share of market income 0,8 0,6 0,4 0,2 0,0 Poorest 2 3 4 5 6 7 8 9 Richest decile decile Final income Post- scal income Disposable income Source: Authors’ calculations based on NHIES 2009/10. The proportion of individuals made US$1.25 PPP per day). No one with market poor by the application of fiscal policy is income above US$2.50 PPP per day was low. Table 6 summarizes “fiscal mobility,” or impoverished. Table 6 also indicates that the movement of individuals across income there are more net “losers” from fiscal policy brackets, between market income and post- than net “winners” in every market-income fiscal income. The standard is the international group other than the poorest. However, the poverty rate of US$1.25 PPP per day. After amount “winners” gain is greater than the applying all the fiscal instruments in this amount “losers” lose for all but the very analysis, less than 4 percent of individuals richest 14 percent of the population (those with market incomes between US$1.25 and who start with market incomes above US$10 US$2.50 PPP per day slipped into poverty PPP per day). (i.e., had post-fiscal incomes of less than Table 6. Fiscal impoverishment at a $1.25 PPP per day poverty line Post-fiscal < market Post-fiscal > market Average Popula- Made Average gain Market Income loss (of tion (%) poor (%) % % (of market Group ($ PPP per market income) day) income) y < 1.25 21 … 29 -5.2 71 293 1.25 <= y < 2.50 28 3.9 62 -6.5 38 29.9 2.50 <= y < 4.00 16 0.0 79 -8.1 21 22.3 4.00 <= y < 10.00 22 0.0 90 -10.7 10 15.7 10.00 <= y < 50.00 12 0.0 97 -15.4 3 10.9 50.00 <= y 2 0.0 99 -17.4 1 2.9 Source: Authors’ calculations based on NHIES 2009/10. 53 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? B. Marginal contribution of the marginal contribution of direct transfers components to poverty and is the difference between the poverty inequality reduction headcounts for disposable income and net market income.51 The comparison is carried This section analyzes the marginal out on the basis of the poverty rates and contribution of each direct transfer and coefficients for disposable incomes. This is indirect subsidy to poverty and inequality in done to compare the increase and decrease Namibia. A transfer’s marginal contribution in observed poverty rates. A similar analysis is calculated by taking the difference could be done taking into consideration between the inequality indicator without indicators based on market incomes. the transfer and with it. For example, the Table 7. Marginal contribution of indirect subsidies and direct transfers to poverty   Poverty measure Increase in poverty (%) Poverty gap Poverty gap headcount headcount Poverty Poverty Severity Severity Poverty Poverty   FGT0 FGT1 FGT2 FGT0 FGT1 FGT2 Poverty indicator (based on disposable income) 0.154 0.042 0.017       Indicator without listed transfer Direct and indirect transfers 0.245 0.107 0.069 58.4 156.6 312.2 Indirect Subsidies 0.173 0.052 0.023 12 23.8 39 Build Together Program 0.158 0.045 0.019 2.2 7.9 17 National Housing Enterprise 0.154 0.042 0.017 0 0 0 Water Subsidy (rural) 0.169 0.048 0.02 9.6 15.8 21.4 Direct transfers 0.225 0.094 0.059 45.8 125.6 251.6 Old Age Pension 0.205 0.078 0.045 32.8 87 169.8 Veterans grant 0.156 0.043 0.018 1.2 4.2 8.8 Children’s grant 0.165 0.045 0.019 6.7 8.6 12 Foster parents grant 0.157 0.043 0.018 1.8 3.7 6.2 Disability grant - adults 0.162 0.047 0.021 5.2 12.8 25.2 Disability grant - children 0.156 0.043 0.017 1 2.2 4.8 Source: Authors’ calculations based on NHIES 2009/10. 54 Severe poverty would be higher in the on the depth and severity of poverty than on absence of direct and indirect transfers. the poverty headcount. Table 7 presents the increases in the poverty Direct transfers reduce poverty more headcount, poverty gap, and squared poverty than indirect transfers (subsidies), with the gaps (severity) in the absence of direct and Old Age Pension being the most important indirect transfers, together and separately. contributor. In the absence of the OAP, the The poverty headcount would rise by 58.4 rest of the transfers and subsidies each play percent higher without the transfers, the a relatively small role in poverty reduction. poverty gap would be 156.6 percent higher, This could suggest poor targeting of most and the squared poverty gap would increase transfers in terms of poverty reduction. by 312.2 percent. The impact is thus greater Table 8. Marginal contribution of indirect subsidies and direct transfers to inequality Change in Gini   Gini coefficient (percent) Gini (based on disposable income) 0.594 Indicator without listed transfer Direct and indirect transfers 0.634 6.7 Indirect Subsidies 0.602 1.4 Build Together Program 0.596 0.3 National Housing Enterprise program 0.594 0 The Old Age Pension Water Subsidy (rural) 0.6 1 drives the poverty- Direct transfers 0.625 5.3 reducing impact of direct transfers Old Age Pension 0.616 3.7 Veterans grant 0.595 0.2 Children’s grant 0.597 0.6 Foster parents grant 0.595 0.2 Disability grant - adult 0.597 0.6 Disability grant - child 0.594 0.1 Sources: Authors’ calculations. 55 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? The impact of transfers on income well when compared with other middle- inequality is less than the impact on income countries in terms of the impact inequality. Table 8 illustrates the increases of fiscal policy on income inequality. As in the Gini coefficient in the absence of the shown in Table 9, the reduction in the Gini transfers. Taken together, the transfers reduce coefficient in moving from market income to inequality only 6.7 percent. As in the poverty disposable income is among the highest of analysis, the OAP is the most effective the countries included in the CEQ sample. program in reducing poverty. To a large extent, this is associated with direct transfers. In absolute terms, Namibia’s C. Comparison to other Gini coefficient reduction trails only South middle-income countries Africa’s. Considering the percentage change, Namibia’s reduction remains among the Although income inequality is very high top performers, but the country’s level of in Namibia, the country performs quite inequality is still very high. Table 9. Gini coefficient at each income concept Ratio Post- Post- Ratio Market Disposable fiscal Final fiscal/ Final/ Income Income Income Income Market Market A B C D C/A D/A Although income Georgia (2013) 0.507 0.395 0.411 0.383 81 76 inequality is very South Africa high in Namibia, the 0.771 0.694 0.695 0.595 90 77 (2010) country performs Namibia (2010) 0.635 0.594 0.590 0.429 93 68 quite well when Armenia (2011) 0.403 0.373 0.374 0.357 93 89 compared with other middle-income Russia (2010) 0.394 0.362 0.366 0.331 93 84 countries in terms of Ethiopia (2011) 0.322 0.305 0.302 0.302 94 94 the impact of fiscal Jordan (2010) 0.342 0.328 0.325 0.319 95 93 policy on income Sri Lanka (2010) 0.371 0.365 0.36 0.344 97 93 inequality Indonesia (2012) 0.394 0.39 0.391 0.37 99 94 Source: Authors’ calculations. 56 Figure 42: The impact of fiscal policy on inequality worldwide, circa 2009/2010 (measured in Gini points) Dominican Rep. (2013) South Africa (2010) Namibia (2009/10) Guatemala (2010) Indonesia (2012) Sri Lanka (2010) Uruguay (2009) Armenia (2011) Ethiopia (2011) Georgia (2013) Mexico (2010) Bolivia (2009) Jordan (2010) Ghana (2013) Russia (2010) Brazil (2009) Peru (2009) 0,00 -0,02 -0,04 -0,06 -0,08 -0,10 -0,12 -0,14 -0,16 -0,18 -0,20 Taxes and direct transfers In-kind transfers Sources: Namibia: World Bank calculations from Ministry of Finance budget documents. Uruguay (Bucheli et al., 2014), Guatemala (Cabrera et al., 2015), Brazil (Higgins and Pereira, 2014), Peru (Jaramillo, 2014), Bolivia (Paz et al., 2014), Mexico (Scott, 2014), Ghana (Younger et al., 2015), Dominican Republic (Aristy-Escuder et al., 2016), El Salvador (Beneke et al., 2014), Ethiopia, Georgia, Sri Lanka, Indonesia, Jordan, South Africa, Russia: (Inchauste and Lustig, forthcoming). Using the international poverty line of (Table 10). The impact is lower than in $2.50 a day, the performance of Namibia’s South Africa and Mexico but higher than in fiscal policy in terms of poverty reduction is Armenia, Brazil, and Costa Rica. As in most generally similar to other CEQ countries. In countries, net indirect consumption taxes terms of percentage point poverty reduction, increase the poverty headcount in Namibia. Namibia is in the middle of the distribution 57 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? Table 10. Poverty headcount ratio for the US$ 2.50 PPP a day for each income concept Net Market Disposable Post-fiscal   Market Income Income Income Income (1) (2) (3) (4) (4)= (3)--Indi- (2) =(1) -- (3)=(2)+Cash     rect Taxes + In- Direct Taxes Transfers direct Subsidies Armenia (2011) 31.3 32.0 28.9 34.9 Bolivia (2009) 19.6 19.6 17.6 20.2 Brazil (2009) 15.1 15.7 11.2 16.3 Costa Rica (2010) 5.4 5.7 3.9 4.2 El Salvador (2011) 14.7 15.1 12.9 14.4 Ethiopia (2011) 81.7 82.7 82.4 84.2 Guatemala (2010) 35.9 36.2 34.6 36.5 Namibia (2012) 48.5 49.3 44.6 46.5 Mexico (2010) 12.6 12.6 10.7 10.7 Peru (2009) 15.2 15.2 14.0 14.5 South Africa (2010) 46.2 46.4 33.4 39.0 Uruguay (2009) 5.1 5.1 1.5 2.3 Sources: Armenia (Younger et al., 2014), Bolivia (Paz et al., 2014), Brazil (Higgins and Pereira, 2014), Ethiopia (Woldehanna et al., 2014), Mexico (Scott, 2014), Peru (Jaramillo, 2014), Uruguay (Bucheli et al., 2014), adapted from Lustig (2014) for Costa Rica (Sauma and Trejos, 2014), El Salvador (Beneke et al., 2014), and Guatemala (Cabrera et al., 2015), South Africa (Inchauste et al., 2014) and authors’ estimates for Namibia based on NHIES 2009/10. Notes: Year of the survey in parenthesis. Bolivia and Indonesia include indirect taxes only. Internationally, the poverty reduction them escape impoverishment (measured attributable to fiscal policy in Namibia is as expenditure of US$2.50 PPP or less per on par with South Africa: from the set of day). Across the low- and middle-income individuals who would be poor without countries included in Figure 43, this result is fiscal expenditures, approximately 11 percent approximately average. are in receipt of fiscal transfers that help 58 Figure 43: Change in poverty headcount ratio (%) from market to consumable income, US$2.50 PPP per day poverty line, circa 2009/2010 3) 01 (2 ic ) ) bl 10 10 ) 0) 11 pu 2) 1) 20 9/ 9) 01 1) 20 3) 01 01 0) Re 9) 0) ) 00 00 01 ) a( (2 13 01 ) 10 r( 01 (2 09 00 (2 01 9) (2 (2 (2 n ri c ala 0 (2 do 20 (2 i ca 00 0 (2 si a (2 ia* (2 Af ia ay ia (2 tem lva ia a( i co (2 in ne ib ia en op an na gu rg h il ssi liv om Sa am do ut ru m az ex rd ha hi ua eo ru Ru Bo Ar So Br Pe M Et El Jo In N D G U G G 20% 13.3%11.4% 8.1% 4.2% 3.3% 1.6% 0% -0.2% -0.8% -2.9% -4.4% -10.6%-11.6% -20% -15.1% -23.3% -29.1% -40% -34.8% -60% -54.9% Pensions as Deferred Income Pensions as Transfer -80% -100% Sources: Namibia: World Bank calculations from Ministry of Finance budget documents. Uruguay (Bucheli et al., 2014), Guatemala (Cabrera et al., 2015), Brazil (Higgins and Pereira, 2014), Peru (Jaramillo, 2014), Bolivia (Paz et al., 2014), Mexico (Scott, 2014), Ghana (Younger et al., 2015), Dominican Republic (Aristy-Escuder et al., 2016), El Salvador (Beneke et al., 2014), Ethiopia, Georgia, Sri Lanka, Indonesia, Jordan, South Africa, Russia: (Inchauste and Lustig, forthcoming). D. Changes in fiscal policy modernization of IT systems, and increasing between 2009/10 and 2016 the number of staff working on tax collection. and possible implications The government has also been working to close loopholes in tax policies and to reduce Given that this analysis uses data from tax avoidance by international companies. 2009/10, it is important to highlight changes These have helped to improve filing rates, in fiscal policy that took place between then improve compliance rates, reduce arrears, and now and their possible implication on etc. This has partly contributed to an increase the incidence of fiscal policy. On the tax in domestic taxes. In addition, the personal side, since 2010 the Ministry of Finance has income tax schedule was changed in 2013 been reforming inland tax administration by reducing rates applied at lower levels of through institutional reorganization into income (Table 11). Among other effects, this functional units (e.g., large taxpayers’ should encourage Namibians to work in the office), reengineering of business processes, formal sector. 59 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? Table 11. Income tax rates in 2009 and 2013 2009 Taxable income (N$) Marginal rate (percent) Up to 40,000 0 40,001 to 80,000 27 80,001 to 200,000 32 200,001 to 750,000 34 Over 750,000 37 2013 Taxable income (N$) Marginal rate (percent) Up to 50,000 0 50,001 to 100,000 18 100,001 to 300,000 25 300,001 to 500,000 28 500,001 to 800,000 30 800,001 to 1,500,000 32 Over 1,500,000 37 Source: Ministry of Finance budget documents. Spending on social sectors such as of recipients of the OAP and the disability education and health has been growing as grant were introduced in 2015. Furthermore, a share of GDP (Figure 10). Notably, the new programs targeted at reducing poverty government has consistently increased the and inequality have been introduced since coverage (number of beneficiaries) and 2009/10. In June 2016, for example, the first the level of support given to social welfare Namibian food bank initiative was launched grant recipients. For example, the OAP and targets households whose income falls grant increased from N$500 in 2010 to below N$400 per month. N$600 in 2013 and then to N$1,000 in These abovementioned developments 2015. The 2016/17-2018/19 Medium- suggest that fiscal policy is likely to have term Expenditure Framework commits to assumed an increased role in poverty and increasing this to N$1,200 per month in the inequality reduction, particularly in light of next two years. Funds for the funeral benefits high unemployment. 60 Chapter 7: Conclusions and Implications for Policy and Data A. Main findings individuals and households as well as a large pool of non-poor households. Namibia’s social spending is generous, Direct transfers and subsidies are producing significant impacts on poverty important for the poor but not well targeted. and inequality, but its efficiency can be Direct transfers are progressive in absolute significantly improved. Namibia dedicates terms, meaning they more effectively target a substantial part of its GDP to direct the poor than in-kind transfers or subsidies. transfers, and it is one of the few African However, they are loosely targeted, and countries that fully funds these programs out households in all deciles receive some of these of its own resources. Overall spending on transfers. As a result, less than half of all direct direct transfers is above the average for Sub- transfer expenditures are actually working to Saharan African countries, but comparable reduce existing poverty. Predictably, direct to the average for developing countries. The transfers have only a modest impact on distribution of social assistance spending is measured inequality. The targeting efficiency biased toward programs that are not meant of selected programs should be improved. to target only the poor. Namibia’s 2009/10 Direct transfers and subsidies reduce poverty fiscal instruments reduced market income by 58 percent, with most of the impact poverty through a system of progressive – but coming from the Old Age Pension (OAP) only loosely targeted – social spending. and water programs. The OAP’s impact on The tax system is merely neutral – taxes inequality is minor. Namibia lacks efficiently are generally paid by the relatively better- designed means tested social protection off population and the redistribution effect programs. is low. On the tax side, replenishing fiscal In-kind transfers are generally larger resources (with household contributions) than direct transfers, but they are received is accomplished directly by taxing incomes only upon access. While poorer households and indirectly by taxing certain consumption benefit from both public education activities. Consumption taxes are neutral to and health services; nonetheless, richer slightly regressive. Personal income taxes are households consume enough of the high progressive, but few of these revenues are value services that the overall distribution of collected from households that appear in the in-kind transfers is approximately neutral. A NHIES. Everyone in Namibia contributes to neutral distribution of benefits still produces tax revenues – some through consumption large inequality reductions because benefits activities at the least, others through income received by poorer households are so large taxes. The government uses these (and other) relative to their incomes. resources to provide transfers and in-kind Introduction of proxy means tested spending that raises incomes for the poorest (PMT) programs and consolidation of 61 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? programs could improve efficiency of the taxes. However, only about one-quarter of social protection system. Consolidation households pay any fuel taxes. Indirect taxes could be helpful if the consolidated benefit is are about 6.7 percent of GDP when using given to all current beneficiaries who receive administrative records – small relative to the at least one of the six transfers analyzed in other low and middle-income countries that this study. To eradicate poverty efficiently, have done CEQ assessments. Only Indonesia Namibia would need to put in place a good (at 6 percent of GDP) and Mexico (at 4.3 targeting system of identifying the poor and percent of GDP) collect noticeably less than granting them benefits adequate to close Namibia in indirect taxes. their consumption deficits. The overall incidence of direct and indirect taxes is progressive, but the marginal B. Distributional analysis of impact of taxes on poverty and inequality is taxes minimal. In Namibia’s actual fiscal system, taxes are progressive with respect to market A relatively large tax burden falls on the income. Direct taxes do not impact poverty working population. Administrative data headcount as very few poor or near-poor shows that taxes (PIT plus indirect taxes, households pay taxes. Direct taxes, being excluding SACU receipts) are about 17.8 progressive, slightly reduce inequality. percent of GDP. This is a relatively large Indirect taxes and subsides, however, are tax burden among countries that have done distribution neutral and have almost no CEQ assessments. Most countries with a effect on inequality, but they increase poverty similar-sized indirect tax take (e.g. Ethiopia) slightly. have a much smaller direct tax take, and most countries with a similar-sized direct tax take C. Transfers and subsidies (e.g. South Africa) have a smaller indirect tax take. Namibia has also a relatively high total Namibia’s expenditure on transfers is tax revenue as a percentage of GDP. significant, but its targeting efficiency Direct taxes included in the analysis are could be improved. The grants are generally large but have low coverage. At about 6.8 progressive in terms of their coverage and percent of GDP according to administrative almost one-third of the population benefits records, direct taxes (PIT) are large relative to from them. However, proxy means testing the other low- and middle-income countries could significantly improve efficiency. The that have done CEQ assessments. Only OAP tops other grants in coverage and it Brazil and South Africa have larger direct is progressively targeted. Children’s grants tax burdens as a percentage of GDP. Direct are given to a relatively small portion of taxes have low coverage: the household data the population and they are generally well indicates that 16 percent of households targeted. Veterans, foster care, and disability pay PIT. The International Monetary Fund grants are generally small in terms of coverage, (IMF) sources indicate that the top 0.1 but they also show a progressive pattern. The percent of households pay over 45 percent of targeting efficiency of the many transfers in all PIT, and the top 4.5 percent pay over two- Namibia could be significantly improved thirds of all PIT. by the introduction of proxy means tested Indirect taxes have high coverage, but programs. they are rather small in comparison to the Although direct transfers are low in 62 direct taxes. All households pay some indirect coverage and poor in efficiency, they matter for the poor. The OAP, the largest grant, D. In-kind transfers in health is the clear “winner” in terms of impact, and education although this is probably because of its large coverage rather than its targeting. In a distant In-kind transfers in health and education second place is the children’s grant. With the are weakly progressive. Their impact is exception of the OAP, which benefits about driven almost entirely by primary education 17 percent of the population, coverage is low, spending. For example, 78 percent of the benefits are small, and targeting is loose. Even 0.206 Gini point reduction from market to with the OAP, the richest decile still captures final income is due to the marginal impact of 5 percent of the benefits. in-kind transfers. Two-thirds of that is due to In comparison to other countries, primary education spending. The remaining transfers in Namibia do not do much for third comes from secondary and tertiary inequality even though they are absolutely education and inpatient and outpatient progressive in their allocation. Perhaps healthcare spending. because of the many transfers in Namibia, Primary education is the most only a few countries (Indonesia, Guatemala, progressively distributed in-kind transfer. This Peru) spend noticeably less on direct transfers. is largely because enrollment is nearly universal Subsidies are more progressively targeted and poorer households have more children. in Namibia than other transfers, largely Secondary education is approximately neutral because of the water subsidy. In addition because enrollment at this level starts to skew to water, the subsidies section analyzed the towards richer households; tertiary education two housing programs, the BTP and the is regressive. This pattern can be seen clearly NHE. The subsidies are slightly progressive. in both the concentration curves and the However, this progressivity is largely driven marginal contributions to inequality. by the water subsidy; it accounts for nearly Primary education’s marginal three-quarters of all subsidy expenditures contribution is 10 Gini points, which is the analyzed in the study, while BTP accounts largest single factor in inequality reduction for nearly one-quarter and NHE for about of the expenditures and revenues in the 2 percent. study. The single contribution from primary Proxy means testing is recommended. By education accounts for nearly 70 percent of itself, PMT could provide better coverage at the overall contribution of in-kind transfers existing spending levels, providing a greater and about 50 percent of the total Gini poverty and inequality impact. However, reduction from market to final income. PMT requires large sunk costs and high Secondary education’s marginal maintenance/operational expenses (e.g. contribution – at 1.6 Gini points – is similar continuous updates of beneficiary registers). to the marginal contribution from the Old Consolidation could be helpful if it captures Age Pension. The old-age grant covers about all current beneficiaries who receive at least 17 percent of the population, and public one of the six transfers; in practice, though, secondary education covers about 34 percent. that would mean higher expenditures because Hence, secondary education is slightly pro- a larger benefit package has to be transferred poor relative to market income but not as to the same (cumulative) beneficiary pro-poor as the OAP. Tertiary education population. is regressively allocated – its concentration coefficient is larger than the market income Gini. 63 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? E. Impact on poverty and Gini points. The inequality-reducing impact inequality of in-kind social spending can be attributed to access (enrollment) as well as demographics. Public social spending reduces poverty and The poverty-reducing impact is inequality, but the impacts of the components associated primarily with the direct transfers differ. Social spending is weighted heavily – low spending leading to low coverage toward in-kind spending, which in turn is and/or small transfer magnitudes. Social weighted heavily toward primary education, spending is progressively distributed and which turns out to be the most inequality- reduces inequality, but its impact would be reducing benefit. If there were no in-kind much more significant with better targeting spending, for example, the reduction in efficiency of direct transfers. Poverty and inequality would only be 0.045 Gini points, inequality could be much lower if targeting while actual social spending, including in- of direct transfers improved. kind transfers, yields a reduction of 0.206 64 Annex 1: What is Fiscal Incidence Analysis This appendix presents additional details individual or a household) as Ih, and net taxes on the methodology and data as well as the of type i as Ti. Let us define the “allocator” revenue and expenditure components that of tax i to unit h as Sih (or the share of net make up CEQ income concepts. The CEQ tax i borne by unit h). Then post-tax income framework and the results discussed in this of unit h can be defined as: Yh= Ih- ∑iTiSih. report measure any income redistribution Complications in the accounting approach achieved by the application of fiscal policy’s arise when actual incidence in the micro-data revenue and expenditure components.54 differs from statutory incidence (or incidence Fiscal incidence analysis consists of according to policies and their implementing allocating taxes and public spending to regulations). households or individuals so that incomes Partial fiscal incidence analysis assesses before taxes and transfers can be compared the impact of one or several fiscal policy with incomes after taxes and transfers, where interventions – for example, income taxes the latter may include the monetized value or the use of public education and health or consumption of free public services. services. Comprehensive fiscal incidence CEQ methodology makes these allocations analysis assesses the impact of the revenue in household level micro-data, so the CEQ and spending sides simultaneously; namely, analysis can (in principle) generate the the impact of direct and indirect taxes, cash impact of taxes and transfers disaggregated by and in-kind transfers, and indirect subsidies. any characteristics or profiles observed in the In addition, there is point-in-time versus micro-data – for example, gender, ethnicity, lifetime fiscal incidence analysis, which or location. can assess a current system or estimate The most common type of fiscal incidence the potential or actual effects of particular analysis, the one used in this report, is the reforms. It can make different assumptions accounting approach, which takes place about tax shifting and the value of in-kind in a post-hoc, static setting. This approach benefits. The analysis can assess the average begins from a “before” or “pre-fiscal” income incidence of a tax or benefit or it can assess and allocates the proper amount of a tax or the incidence on the margin; e.g., the a transfer to each household or individual. distribution of an increase in the spending of If the fiscal intervention is a direct tax public education. (transfer), the post-tax (post-transfer) income In this study, we assess the current fiscal is calculated by subtracting (adding) the tax system on partial and comprehensive bases. paid (transfer received). In the accounting We estimate actual average incidence and approach, behavioral responses that taxes and effective average rates (i.e., we use average public spending may trigger in individuals or tax collection rates including possible tax households are not taken into account. evasion) and we value in-kind benefits More formally, define a “before taxes according to the government cost approach. and transfers” income of unit h (typically an Incidence studies use micro-data from 65 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? household surveys or rely on incidence incidence analysis include measures of indicators from secondary sources (usually progressivity, such as incidence – i.e., the micro-data sets themselves). In practice, share of taxes (transfers) paid (received) as surveys will not include information on a proportion of the pre-tax (pre-transfer) every tax paid or transfer received (or the income – and concentration coefficients, information if it exists may be inaccurate), which measure the share (by decile or so that information must be generated in a quintile) of specific or overall taxes and consistent and methodologically solid way. transfers. In addition, fiscal incidence studies Frequently, the information will have to report inequality and poverty indicators – be generated using more than one method such as headcount ratios and Gini coefficients to check the sensitivity of the results to – before and after taxes and transfers. Some assumptions that cannot be externally studies include indicators of horizontal equity validated. to capture how policies impact individuals Indicators that can be generated by who are equal before fiscal intervention. 66 Annex 2: Classifying Targeting Methods55 Individual/household assessment is a method simple means tests are used for both direct in which an official (usually a government transfer programs and for fee-wavering employee) directly assesses, household by programs, with or without the visit to the household or individual by individual, household. whether the applicant is eligible for the Proxy means tests, while relatively rare, program. It is the most laborious of targeting are being instituted in a growing number methods. The gold standard of targeting is of countries. We use the term to denote a a verified means test that collects (nearly) system that generates a score for applicant complete information on a household’s households based on fairly easy to observe income and/or wealth and verifies the characteristics of the household such as information collected against independent the location and quality of the dwelling, sources such as pay stubs or income and ownership of durable goods, demographic property tax records. This requires the structure of the household, and the education existence of such verifiable records in the and possibly, occupations of adult members. target population, as well as the administrative The indicators used in calculating this score capacity to process this information and to and their weights are derived from statistical continually update it in a timely fashion. analysis (usually regression analysis or For these reasons, verified means tests are principal components) of data from detailed extremely rare in developing countries where household surveys of a sort, too costly to the poorest households receive income from be carried out for all applicants to large a myriad of diverse sources and formal record programs. The information provided by the keeping is nonexistent. Other individual applicant is usually partially verified by either assessment mechanisms are used in the collecting the information on a visit to the absence of the capacity for a verified means home by a program official. test. Three common ones are simple means Community based-targeting uses a group tests, proxy means tests, and community- of community members or a community based targeting. leader whose principal functions in the Simple means tests, with no independent community are not related to the transfer verification of income, are not uncommon. program to decide who in the community A visit to the household by a program social should benefit. School officials or the parent- worker may help to verify in a qualitative way teacher association may determine entry that visible standards of living (which reflect to a school-linked program. A group of income or wealth) are more or less consistent village elders may determine who receives with the figures reported. Alternatively, the grain provided for drought relief, or social workers’ assessment may be wholly special committees composed of common qualitative, taking into account many factors community members or a mix of community about the household’s needs and means but members and local officials may be specially not having to quantify them. These types of formed to determine eligibility for a program. 67 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? The idea is that local knowledge of families’ Age, gender, ethnicity, land ownership, living conditions may be more accurate than demographic composition, or geographical the results of a means test conducted by a location are common examples that are fairly government social worker or a proxy means easy to verify. Age is a commonly used category, test. with cash child allowances predominant in Categorical targeting refers to a method transition countries, supplemental feeding in which all individuals in a specified category programs for children under five common —for example, a particular age group or in poor countries, and non-contributory region — are eligible to receive benefits. pensions for the elderly common in many This method is also referred to as statistical places. Geographic targeting is even more targeting, tagging, or group targeting. It common, often used in combination with involves defining eligibility in terms of other methods. Unemployment or disability individual or household characteristics that status is somewhat harder to verify, but cash are fairly easy to observe, hard to falsely assistance to these groups may be categorically manipulate, and correlated with poverty. targeted as well. 68 Notes 1 Source: http://www.commitmentoequity. 3 The formulation process for National org/.The Commitment to Equity (CEQ) Development Plan V (NDP5) was was designed to analyze the impact of launched in June 2016. taxation and social spending on inequality 4 The poverty headcount in this target is and poverty in individual countries and measured at the national poverty line of provide a roadmap for governments, N$277.54 in 2009/10. multilateral institutions, and non- governmental organizations in their 5 See ILO (2014) for complete descriptions efforts to build more equitable societies. of these and other social programs. As of June 2016, the CEQ analysis has 6 See http://www.op.gov.na/ been completed in 20 countries and documents/84084/264431/ carried out in 15 countries. SONA+2016,+final.pdf/c57aa559-673f- 2 The Namibia Statistics Agency (NSA) 4618-8a49-37aeddc3f919 uses three national poverty lines (NSA, 7 Total tax collections are taken to exclude 2012). These include the food, lower Southern Africa Customs Union (SACU) bound and upper bound poverty lines receipts, which are not paid by Namibian which were set at N$204.05, N$277.54, households. SACU receipts are revenue and N$377.96 per month, respectively, for the Government but not part of its in 2009/10 prices. The food poverty line “tax collections.” recognizes that all human beings have a 8 Taxable income refers to gross income less basic minimum nutritional requirement. exemptions and allowable deductions. The lower bound poverty line is based on Individual incomes could be salary/ households that sacrifice some of their wages, pension/annuity payments basic food requirements in order to meet and investment income (interest and their non-food needs. Therefore, the dividends). minimum amount set on non-food basic needs is added to the food line. The upper 9 Other zero-rate items include: petrol, bound poverty line is based on households diesel and paraffin; direct export of goods whose food expenditure is very close to and services; international transport; the food line. For these households, in sale of a going concern; sale of land and addition to the basic food requirements buildings for residential purposes and that are measured by the food poverty erection of residential buildings; supply line, there are certain basic non-food items of municipal services to residential that they need. The terminology used in accounts; supply of agricultural land to be this report is consistent with that used by used for resettlement purposes; supplies NSA which considers the proportion of made in respect of guarantees; supply of the population below the upper bound undertaking services for funerals; supply poverty as poor while those below the of services physically rendered outside lower bound poverty line are considered Namibia; postage stamps (but not postage to be severely poor. Severe poverty and stamps for collectors); telecommunication extreme poverty are used interchangeably services to residential accounts; supplies in this report. by charitable organizations and similar 69 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? institutions; supply of livestock (on the education) is typically for children up hoof ); supply and repair of goods to be to age 6; this is followed by seven years used as aids by physically handicapped of primary school, made up of lower persons who are blind, deaf, crippled primary (grades 1-4) and upper primary or a chronic invalid; services for any (grades 5-7); secondary school constitutes adjustment or modification in respect of junior secondary (grades 8-10) and senior a vehicle used for these purposes. secondary (grades 11-12); tertiary/higher education is then offered at teacher Exemptions are applicable on the following education colleges, agricultural colleges, supplies: financial services; medical the University of Namibia, and the services and services provided by hospitals; Polytechnic of Namibia. According to group finance/management companies Fischer (2010), the private sector runs and inter-company loans; residential pre-schools and kindergartens, while the leases and fringe-benefit accommodations; state runs other schools. public transport services; educational services; management of group housing 16 For more details on the approaches to and commercial premises; employee fiscal incidence analysis see, for example, organizations; local authorities; fringe Adema and Ladaique (2005), Alleyne et benefits; supplies to foreign heads of state. al. (2004), Atkinson (1983), Barr (2004), Bergh (2005), Birdsall et al. (2008), It is important to acknowledge other 10 Bourguignon and Pereira da Silva (2003), criterion besides equity for evaluating Breceda et al. (2008), Coady (2006), a tax system, including the impact on Ferreira and Robalino (2010), Fiszbein et economic growth. al. (2009), Fullerton and Metcalf (2002), The classification into these categories 11 Goñi et al. (2011), Grosh et al. (2008), is purely for the purposes of this study. Kakwani (1977), Lambert (2002), Lora The study acknowledges that most of (2006), Lustig and Higgins (2013), the government’s budget is allocated to Morra et al. (2009), O’Donnell et al. sectors other than “social”. (2008), Shah (2003), Suits (1977), van de 12 At 14.9 percent of GDP in 2009/10, Walle and Nead (1995), and World Bank total general government spending in (2000/2001, 2006, 2009). Namibia was relatively low compared to 17 We do not provide an assessment of the the middle-income country average. The drivers or causes of poverty or inequality, corresponding values for comparable nor do we estimate how welfare might middle-income countries are: South change if fiscal policy were changed; the Africa at 34.8 percent; Indonesia at 17.5 analysis only describes the level of poverty percent, and Peru at 25.5 percent. or inequality that exists before and after 13 See ILO (2014) for details on the coverage the government undertakes fiscal policy. of different types of social assistance 18 Complications in the accounting approach programs. arise when actual incidence (in the micro- 14 This meant that parents were no longer data) differs from the statutory incidence required to pay fees toward the School (or incidence according to policies and Development Fund. their implementing regulations). 15 The Namibian education system is For more on the definitions of 19 structured as follows: pre-school (early progressivity, see Lustig and Higgins childhood development, care, and (2013). 70 See, for example, Duclos and Araar 20 28 Net national income is gross national (2006), p. 136. income less depreciation and other payments to fixed capital. Total market 21 A progressive transfer (so defined) will income from all sources in the survey be equalizing in the sense that inequality represents about 67 percent of average measured after receipt of the transfer gross national income in 2009/10. will be lower than was before receipt of the transfer. See, for example, Lambert The main reason for the use of 29 (1985). consumption instead of income for this analysis is the significant underreporting 22 To establish progressivity in the CEQ of incomes among richer households. process, it is not necessary for transfers Consumption data represents the actual (taxes) to be progressive (either relatively welfare of the population much better or absolutely in the case of transfers) at and thus is used as a proxy for market every point (i.e., for every individual) in income. the distribution. Transfers (taxes) can be globally progressive, even if they are not In other words, the derived market 30 everywhere progressive. See, for example income measure not only compresses the Duclos and Araar (2006). distribution of welfare, it also changes the rank position of some households in the The 45-degree line indicates the 23 distribution. cumulative share of the total population that an individual represents. It is the line 31 Despite high level of revenues in Namibia of perfect equality. in comparison to other developing countries, Namibian tax-to-GDP ratio 24 See Lindert, Skoufias and Shapiro (2006), is significantly lower than the OECD Scott (2011) and O’Donnell et al. (2008). average (34 percent in 2009/10, based on 25 NHIES is enumerated once every five the WDI data). years; the 2015/16 survey was being Direct tax payers are imputed. If a 32 enumerated and collected as this report household recalls paying income taxes was being prepared. or having income taxes withheld, that 26 Even when the “event” of paying a tax household is a taxpayer. The amount paid or receiving a benefit can be identified is simulated, with the total amount paid directly, it is not always possible to by NHIES respondents as a percent of determine the amounts paid or received. total disposable income equal to the total In cases where inference or imputation amount collected (reported in budget is needed to estimate a value of a tax or documents) as a percent of national benefit for a directly identified payer or accounts final consumption expenditures. recipient, the CEQ Master Workbook 33 Two caveats should be kept in mind in provides the details on the algorithms interpreting this result. First, the profile used to estimate these values. of PIT payers – including their rank 27 Often, a combination of these methods position in the income distribution – works best. For example, simulating can change significantly depending on distributions of benefits in the primary the reference income used. For example, survey using empirical distributions the poorest 20 percent of the observed estimated from other sources. Lustig and market income distribution contains no Higgins (2013) describe the methods to taxpayers; however, the poorest 20 percent allocate taxes and transfers in detail. of the derived market income contains 71 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? about 0.5 percent of all taxpayers. In the analyze program performance, as well as bottom 60 percent, the market income to provide estimates of coverage, targeting group contains approximately 3 percent and generosity of social spending. Using of all taxpayers, compared with 7 percent household survey data has its limitations. for derived market income. Second, the While household surveys, unlike number of (observed) taxpayers and tax administrative data, allow estimating revenue reported from NHIES households incidence of transfers for different socio- are far less than comparable data from economic groups, some social assistance budget reporting documents and transfers are captured by surveys, while secondary sources. For example, NHIES others are not. Also the survey is not reports just over 71,000 PIT taxpayers in designed to capture some of the small 2009/10; secondary sources indicate that transfers and the results should be the number of actual taxpayers was nearly interpreted with caution. double the NHIES number at 139,000. 37 Coady, D., M. Grosh and J. Hoddinott. Secondary sources also indicate that the 2004. “Targeting of Transfers in top 1 percent of taxpayers (ranked by PIT Developing Countries: Review of Lessons paid) contribute more than 50 percent of and Experience”, World Bank and IFPRI. total PIT revenues; NHIES households with minimum taxable income equal to It is important to acknowledge the 38 the top 1 percent of taxpayers contribute possibility that the relatively high generosity only 28 percent of PIT. of transfers could create disincentives to work in the formal sector. An investigation We view the incidence of indirect 34 of this is beyond the scope of this report. taxes (and subsidies) with respect to disposable income because households 39 We view the incidence of subsidies with make consumption decisions over a respect to disposable income because household budget that includes cash households make consumption decisions transfers received, affording them a over household budgets that include the higher consumption level than market cash transfers they receive, affording them income alone would allow. In turn, this higher consumption levels than what implies households would have paid market incomes alone would allow. In less in indirect taxes (or received less in turn, this implies that households would consumption subsidies), in the absence of have received less in consumption subsidies these transfers. in the absence of these transfers. 35 The fuel levy is a much smaller source of 40 We caution that this approach does not public revenues than the VAT; in addition, address the average, marginal, or relative we cannot determine from NHIES how quality of the services provided, although much fuel was consumed for household public service provision quality does vary use (i.e., not for vehicles/transport). For in Namibia. both reasons, VAT has a 98 percent share 41 Specifically, concentration shares will be in total indirect tax collected from NHIES proportional to shares of enrolled children. households. As a result, we discuss here Poorer households in Namibia have the cumulative impact of both VAT and more school-age children and a higher the fuel levy. propensity to send them to public schools; 36 NHIES was used to assess performance both tendencies produce a larger share of of specific programs in Namibia. children in public schools, if enrollment Disaggregated NHIES data was used to rates are equal across deciles. 72 42 A “utilizing” household is one with at index will be positive, which indicates least one child enrolled in a public school. that taxes (transfers) are progressive. If the Utilization can be higher because of the tax (transfer) concentration curve is above greater frequency of school-age children in (below) the Lorenz curve, the Kakwani some households, the greater propensity index will be negative, which indicates to enroll a school-age child in the public that taxes (transfers) are regressive. system, or both. 48 As Bibi and Duclos (2010) and Lustig 43 For a single student, the tertiary education (2014) explain, the potential impact of benefit is approximately 2.7 times larger any individual intervention should not than the primary benefit, and richer be calculated by taking the difference households account for virtually all public between consecutive pairs of income tertiary utilization. Nonetheless, total concepts. For example, taking the education expenditures are approximately difference between the Gini coefficient neutrally distributed because of the great for post-fiscal and disposable income is numbers of publicly enrolled primary- not equal to the contribution of indirect aged children from poor households. subsidies and indirect taxes to the decline of inequality from market to post-fiscal 44 From administrative and budget records, income. There is an error component we determined the government’s “unit to our estimates when they are arranged cost” of a publicly-provided outpatient sequentially because (a) the contribution or inpatient healthcare visit. From the of each intervention is path dependent; DHS, we determined shares of inpatient (b) we are showing one possible path; and outpatient visits accounted for by and (c) the actual path is unobserved. income-ranked population groups. The We can compare, however, the impact DHS also allowed us to determine a of interventions on any indicator with propensity to visit public providers (as respect to market income. This is what we opposed to private providers). We use the do in this section. Perhaps a more precise latter share and propensity estimates to way of stating our result is the following: allocate verified public healthcare visits Without the redistributive process set in (and attendant benefits) to representative motion (via taxes and transfers) by fiscal households in the NHIES. policy, measured inequality would be 45 This is based on total verified visits to higher in Namibia (in a static setting). outpatient care facilities, total verified 49 The use of ‘extreme poverty’ is consistent inpatient days, and shares of government with the terminology used by NSA which expenditures on health going to facilities considers the proportion of the population providing both types of services. below the lower bound poverty severely 46 Similar to education, private healthcare poor. Severe poverty and extreme poverty facilitates are typically considered to are used interchangeably in this report. be of better quality compared to public The official national extreme/severe 50 facilities. poverty rate estimated at the lower bound The Kakwani index of tax (transfer) 47 poverty line of N$277.54 is 15.3 percent. progressivity is twice the area between the However, the present CEQ calculations market income Lorenz curve and the tax find a rate of 15.4 percent at the disposable (transfer) concentration curve. If the tax income concept (reported in Figure 40, (transfer) concentration curve is below Table 5, and Table 7). The difference - (above) the Lorenz curve, the Kakwani which is marginal - is statistical and due 73 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? to households that were not included in the disposable income calculation because they did not record answers to all the other NHIES questions necessary to create the other income concepts. One drawback of the marginal 51 contribution method is that the sum of all the marginal contributions is not equal to the total redistributive effect because of the interaction between policies, limiting the accuracy of the magnitude of each contribution (Shorrocks, 2013). 52 Based on the World Bank’s WDI databases (http://data.worldbank.org/indicator/ GC.TAX.TOTL.GD.ZS), Namibia is among the list of the top 30 percent of the countries in terms of the total tax revenue as a percentage of GDP. We have not actually compared the 53 PMT based distribution and the current distribution. The basis for thinking that PMT will improve the outcomes is based on cross-country experiences. It is also important to emphasize that a combination of measures should be considered as a long term solution for Namibia. A whole range of targeting mechanisms (community, PMT, self- targeting, etc.) should be considered. An additional research in this area should be undertaken, suggesting appropriate methods that will improve targeting efficiency of direct transfers in Namibia. 54 For more details, see Lustig, Nora and Sean Higgins. (2013). Commitment to Equity Assessment (CEQ): Estimating the Incidence of Social Spending, Subsidies and Taxes. Handbook. CEQ Working Paper No. 1, Center for Inter-American Policy and Research and Department of Economics, Tulane University and Inter- American Dialogue, September. 55 This annex is based on (and generally taken from) Coady et al. (2004). 74 References Adema, Willem, and Maxime Ladaique. Transfers, Taxes and Market Income Across 2005. “Net Social Expenditure, 2005 Five OECD Countries,” Bulletin of Economic Edition: More Comprehensive Measures of Research 62(4): 387-406. Social Support.” OECD Social, Employment Birdsall, Nancy, Augusto de la Torre, and Migration Working Papers 29. and Rachel Menezes. 2008. Fair Growth: Alleyne, Dillon, James Alm, Roy Bahl, Economic Policies for Latin America’s Poor and and Sally Wallace. 2004. “Tax Burden Middle-Income Majority. Washington DC: in Jamaica.” Georgia State University Brookings Institution Press. International Studies Program Working Bourguignon, François, and Luiz A. Paper 04-34. Pereira da Silva, eds. 2003. The Impact of Aristy-Escuder, Jaime, Maynor Cabrera, Economic Poverty and Income Distribution. Blanca Moreno-Dodson, and Miguel Washington DC: World Bank and Oxford Sánchez-Martín. 2016. “Fiscal policy and University Press. redistribution in the Dominican Republic.” Breceda, Karla, Jamele Rigolini, and CEQ Working Paper No. 37, CEQ Institute. Jaime Saavedra. 2008. “Latin America and the Atkinson, Anthony B. 1983. Social Justice Social Contract: Patterns of Social Spending and Public Policy. MIT Press. and Taxation.” Policy Research Working Barr, Nicholas. 2004. Economics of the Paper 4604. World Bank, Latin American Welfare State, Fourth Edition. New York: and Caribbean Region, Poverty Department, Oxford University Press. Poverty Reduction and Economic Beneke, Margarita, Nora Lustig, and Management Division. Washington DC: José Andrés Oliva. 2014. El impacto de los World Bank. impuestos y el gasto social en la desigualdad Bucheli, Marisa, Nora Lustig, Máximo y la pobreza en El Salvador. CEQ Working Rossi, and Florencia Amábile. 2014. “Social Paper No. 26, Center for Inter-American Spending, Taxes and Income Redistribution Policy and Research and Department of in Uruguay,” in Lustig, Nora, Carola Pessino, Economics, Tulane University and Inter- and John Scott, eds., The Redistributive American Dialogue. Impact of Taxes and Social Spending in Latin Bergh, Andreas. 2005. “On the America, special issue. Public Finance Review counterfactual problem of welfare 42(3). state research: How can we measure Cabrera, Maynor, Nora Lustig, and redistribution?” European Sociological Review Hilcias Moran. 2015. “Fiscal Policy, Inequality 21(4): 345-357. and the Ethnic Divide in Guatemala,” World Besley, Timothy, and Torsten Persson. Development 76: 263–279. 2013. “Taxation and Development,” in Coady, David. 2006. “The distributional Auerbach, Alan, Raj Chetty, Martin Feldstein, impacts of indirect tax and public pricing and Emmanuel Saez, eds., Handbook of Public reforms,” in Coudouel, Aline, and Stefano Economics Volume 5, 2013. Paternostro, eds., Analyzing the Distributional Bibi, Sami, and Jean-Yves Duclos. 2010. Impact of Reforms: A Practitioner’s Guide. “A Comparison of the Poverty Impact of World Bank: Washington DC. 75 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? Coady, David, Margaret Grosh, and John Washington DC: World Bank. Hoddinott. 2004. “Targeting of Transfers in Higgins, Sean and Claudiney Pereira. Developing Countries: Review of Lessons 2014. “The Effects of Brazil’s Taxation and Experience.” World Bank and IFPRI. and Social Spending on the Distribution Duclos, Jean-Yves, and Abdelkrim Araar. of Household Income,” in Lustig, Nora, 2006. Poverty and Equity: Measurement, Carola Pessino, and John Scott, eds., The Policy, and Estimation with DAD. New York: Redistributive Impact of Taxes and Social Springer and International Development Spending in Latin America, special issue, Research Centre. Public Finance Review 42(3). Ferreira, Francisco, and David Robalino. Honorati, Maddalena, Ugo Gentilini, 2010. “Social Protection in Latin America: and Yemtsov Ruslan. 2015. The state of social Achievements and Limitations.” Policy safety nets 2015. Washington, D.C: World Research Working Paper 5305. Washington Bank Group. DC: World Bank, Latin America and Inchauste, Gabriela, Nora Lustig, Caribbean Region, Office of the Chief Mashekwa Maboshe, Catriona, Purfield and Economist, and Human Development Ingrid Woolard. 2015. The Distributional Network Social Protection and Labor Unit. Impact of Fiscal Policy in South Africa. Fischer, Gereon. 2010. The Namibian Policy Research Working Paper 7194. World educational system. Windhoek: Friedrich- Bank: Poverty Global Practice Group & Ebert-Stiftung. Macroeconomics and Fiscal Management Fiszbein, Ariel, Norbert Schady, Francisco Global Practice Group. Ferreira, Margaret Grosh, Nial Kelleher, Inchauste, Gabriela, and Nora Lustig. Pedro Olinto, and Emmanuel Skoufias. The Distributional Impact of Fiscal Policy: 2009. Conditional Cash Transfers: Reducing Experience from Developing Countries, Present and Future Poverty. Washington DC: edited by. Washington, D.C: World Bank. World Bank. Forthcoming. Fullerton, Don, and Gilbert Metcalf. International Labor Organization 2002. “Tax Incidence,” NBER Working (ILO). 2014. Namibia Social Protection Floor Paper 8829, March. Assessment. Geneva: ILO, 2014. Goñi, Edwin, Humberto López, and Jaramillo, Miguel. 2014. “The Incidence Luis Servén. 2011. “Fiscal Redistribution of Social Spending and Taxes in Peru,” in and Income Inequality in Latin America,” Lustig, Nora, Carola Pessino, and John World Development 39(9): 1558-1569. Scott, eds., The Redistributive Impact of Taxes Government of the Republic of Namibia. and Social Spending in Latin America, special 2004. Namibia Vision 2030: Policy Framework issue, Public Finance Review 42: (3). for Long-Term National Development. Office Jellema, Jon, Matthew Wai-Poi, and of the President, Windhoek. Rythia Afkar. “The Distributional Impact ________. 2016. Harambee Prosperity of Fiscal Policy in Indonesia.” World Bank. Plan, 2016/17–2019/20: Namibian Forthcoming. Government’s Action Plan towards Prosperity Kakwani, Nanak C. 1977. “Measurement for All. Office of the President, Windhoek. of Tax Progressivity: An International Grosh, Margaret, Carlo del Ninno, Comparison,” The Economic Journal 87(345): Emil Tesliuc, and Azedine Ouerghi. 2008. 71-80. For Protection and Promotion: The Design Lambert, Peter. 1985. “On the 76 and Implementation of Effective Safety Nets. redistributive effect of taxes and benefits,” Scottish Journal of Political Economy 32(1): Household Survey Data: A Guide to Techniques 39-54. and Their Implementation. Washington, D.C: ________. 2002. The Distribution and The World Bank. Redistribution of Income, Third Edition. Paz Arauco, Verónica, George Gray Manchester United Kingdom: Manchester Molina, Wilson Jiménez Pozo, and Ernesto University Press. Yáñez Aguilar. 2014. “Explaining Low Lindert, Kathy, Emmanuel Skoufias, Redistributive Impact in Bolivia,” in Lustig, and Joseph Shapiro. 2006. “Redistributing Nora, Carola Pessino, and John Scott, eds., Income to the Poor and Rich: Public Transfers The Redistributive Impact of Taxes and Social in Latin America and the Caribbean.” Spending in Latin America, special issue, Social Protection Discussion Paper 0605. Public Finance Review 42(3). Washington, D.C: The World Bank. Sauma, Pablo, and Juan Diego Trejos. Lora, Eduardo, ed. 2006. The State of 2014. “Gasto Público, Social Impuestos, State Reforms in Latin America. Washington Redistribución del Ingreso y Pobreza en DC: World Bank. Costa Rica,” CEQ Working Paper No. 18. Lustig, Nora, and Sean Higgins. Scott, John. 2011. “Gasto Público y 2013. Commitment to Equity Assessment Desarrollo Humano en México: Análisis de (CEQ):  Estimating the Incidence of Social Incidencia y Equidad.” Working Paper for Spending, Subsidies and Taxes.  Handbook. Informe sobre Desarrollo Humano México CEQ Working Paper No. 1, Center for 2011. Mexico City: UNDP. Inter-American Policy and Research and ________. 2014. “Redistributive Impact Department of Economics, Tulane University and Efficiency of Mexico’s Fiscal System,” and Inter-American Dialogue, September. in Lustig, Nora, Carola Pessino and John ________. 2014. “Taxes, Transfers, Scott, eds., The Redistributive Impact of Taxes Inequality and the Poor in the Developing and Social Spending in Latin America, special World. Round 1.” CEQ Working Paper No. issue, Public Finance Review 42(3). 23, Center for Inter-American Policy and Shah, Anwar, ed. 2003. Handbook Research and Department of Economics, on Public Sector Performance Reviews. Tulane University and Inter-American Washington DC: The World Bank. Dialogue. Shorrocks, A. 2013. “Decomposition Morra Imas, Linda G., and Ray C. Rist. procedures for distributional analysis: a 2009. The Road to Results: Designing and unified framework based on the Shapley Conducting Effective Development Evaluations. value,” Journal of Economic Inequality 11(1): Washington DC: World Bank. 99-126. Namibia Statistics Agency. 2012. Poverty Suits, Daniel B. 1997. “Measure of Tax Dynamics in Namibia: A comparative study Progressivity,” The American Economic Review using the 1993/94, 2003/04 and the 2009/10 67(4): 747-752. NHIES surveys. Windhoek: Namibia Van de Walle, Dominique, and Kimberly Statistics Agency. Nead, eds. 1995. Public Spending and the ________. 2014. Namibia Labor Force Poor: Theory and Evidence. Baltimore and Survey 2014 Report. Windhoek: Namibia London: Published for the World Bank by Statistics Agency. John Hopkins University Press. O’Donnell, Owen, Eddy van Doorslaer, Woldehanna, Tassew, Tsehaye Eyasu, Adam Wagstaff, and Magnus Lindelow. Ruth Hill, Gabriela Inchauste, and Nora 2008. Analyzing Health Equity Using Lustig. 2014. “A Fiscal Incidence Analysis 77 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? for Ethiopia.” Background Paper for World Bank Poverty Assessment. Forthcoming. Work Bank. 2000/2001. “World Development Report 2000/2001: Attacking Poverty.” Washington, DC and New York: Published for the World Bank by Oxford University Press. ________. 2006. “Country Policy and Institutional Assessments.” Operations Policy and Country Services. Washington, DC: World Bank ________. 2009. “The World Bank’s Country Policy and Institutional Assessment, An Evaluation.” Independent Evaluation Group. Washington, DC: World Bank. ________. 2009. “Levels and Patterns of Safety Net Spending in Developing and Transition Countries.” Safety Net Primer, Washington DC: The World Bank. ________. 2014. South Africa Economic Update: Fiscal Policy and Redistribution in an Unequal Society. November. Washington, DC: World Bank. Younger, Stephen, and Artsvi Khachatryan. 2014. “Fiscal Incidence in Armenia.” Background Paper for World Bank Armenia Public Expenditure Review (forthcoming). Younger, Stephen, Eric Osei-Assibey, and Felix Oppong. 2015. “Fiscal Incidence in Ghana.” CEQ Working Paper No. 35. Center for Inter-American Policy and Research and Department of Economics, Tulane University, Ithaca College, University of Ghana and World Bank. 78 Notes: 79 Does Fiscal Policy Benefit the Poor and Reduce Inequality in Namibia? Notes: 80 This report uses the Commitment to Equity methodology to analyze the progressivity of the main components of Namibia’s tax and social spending programs and quantify their impact on poverty and inequality. The main conclusions are threefold. First, the incidence of taxes (personal income tax, Value-Added Tax, and fuel levy) is progressive: the rich bear relatively more of the burden of taxes than the poor. Second, a generous and progressive social spending system benefits low income earners and the poor, but its coverage and efficiency could be further improved. The most progressive programs are direct transfers, especially the Old Age Pension. Third, Namibia’s fiscal policy does reduce poverty and inequality, but its impact is relatively modest in comparison to other comparator countries for which a similar methodology has been applied. Despite the important role of fiscal policy in reducing poverty and inequality in Namibia, poverty remains relatively high for a middle income country and the country is among the most unequal countries in the world. This shows that there are limits to what fiscal policy alone can achieve in tacking the challenge of poverty and inequality in Namibia. Looking ahead, further reductions in poverty and inequality will require further improvements in the efficiency of social spending through for example better targeting efficiency and consolidation of social programs, and reducing leakages of existing programs. Ultimately, higher and more inclusive economic growth that creates more jobs for the poorest members of society is needed.