12.07.2015 Views

PARPA II Review—The Tax System in Mozambique - UNICEF ...

PARPA II Review—The Tax System in Mozambique - UNICEF ...

PARPA II Review—The Tax System in Mozambique - UNICEF ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

<strong>PARPA</strong> <strong>II</strong> Review—The <strong>Tax</strong><strong>System</strong> <strong>in</strong> <strong>Mozambique</strong>Volume ISeptember 2009This publication was produced by Nathan Associates Inc. for review by the United StatesAgency for International Development.


<strong>PARPA</strong> <strong>II</strong> Review—The <strong>Tax</strong><strong>System</strong> <strong>in</strong> <strong>Mozambique</strong>Volume IDISCLAIMERThis document is made possible by the support of the American people through the United States Agency forInternational Development (USAID). Its contents are the sole responsibility of the author or authors and do notnecessarily reflect the views of USAID or the United States government.


ContentsExecutive Summaryvii1. Introduction 12. Background 3The <strong>Tax</strong> <strong>System</strong> at the Start of <strong>PARPA</strong> <strong>II</strong> 3<strong>Tax</strong> Issues <strong>in</strong> <strong>PARPA</strong> <strong>II</strong> 6Major <strong>Tax</strong> Reforms 2006-2009 73. The Current <strong>Tax</strong> <strong>System</strong> 13The <strong>Tax</strong> <strong>System</strong> <strong>in</strong> 2009 13Do<strong>in</strong>g Bus<strong>in</strong>ess Rank<strong>in</strong>gs of <strong>Mozambique</strong>’s <strong>Tax</strong> <strong>System</strong> 23<strong>Tax</strong> Adm<strong>in</strong>istration 25Plans for Moderniz<strong>in</strong>g <strong>Tax</strong> Adm<strong>in</strong>istration 294. Revenue Performance 33Revenue Targets 33Revenue Growth 35International Comparisons 42Revenue Effort 445. Impact of the <strong>Tax</strong> <strong>System</strong> 47Investment 48Employment 53Sav<strong>in</strong>g 55Private Sector Development 57Fairness and Equity 686. Priority Issues 73Medium-Term Revenue Target 73


<strong>II</strong><strong>Tax</strong> Rates 74<strong>Tax</strong> Policy Analysis 75Functional Integration of <strong>Tax</strong> Operations 76E-<strong>Tax</strong>ation 76Risk Management 77Simplification 78<strong>Tax</strong>payer Services 79<strong>Tax</strong> Culture 80<strong>Tax</strong> Tra<strong>in</strong><strong>in</strong>g 80EITI Implementation 81Donor Coord<strong>in</strong>ation 81References 83IllustrationsFiguresFigure 3-1. Pr<strong>in</strong>cipal Sources of Domestic Revenue, 2005 and 2008 14Figure 4-1. Domestic Revenue & Expenditure as a Share of GDP 34Figure 4-2. Pr<strong>in</strong>cipal <strong>Tax</strong> Revenues as a Share of GDP 38Figure 4-3. Revenues from <strong>Tax</strong>es on Income as a Share of Total Domestic Revenues 41Figure 5-1. Lower<strong>in</strong>g <strong>Tax</strong>es on Capital to Stimulate Investment 48Figure 5-2. Company <strong>Tax</strong> Rates <strong>in</strong> the SADC Region, 2009 50Figure 5-3. Company <strong>Tax</strong> + Dividend Withhold<strong>in</strong>g <strong>Tax</strong> <strong>in</strong> the SADC Region, 2009 52Figure 5-4. <strong>Tax</strong> Rates as a Constra<strong>in</strong>t to Bus<strong>in</strong>ess, by Firm size 63Figure 5-5. <strong>Tax</strong> Adm<strong>in</strong>istration as a Constra<strong>in</strong>t to Bus<strong>in</strong>ess, by Firm Size 64Figure 5-6. Responses on Sales Declared for <strong>Tax</strong> Purposes, by Firm Size 64Figure 5-7. Applied <strong>Tax</strong> Rates 69Figure 5-8. Share of Rural and Urban Expenditures Potentially Subject to VAT 72TablesTable 2-1. <strong>Tax</strong> Objectives, Actions, and Strategic Indicators <strong>in</strong> <strong>PARPA</strong> <strong>II</strong> 8Table 2-2. Major Changes <strong>in</strong> the <strong>Tax</strong> Law, 2006 – 2009 10Table 3-1. Approved <strong>Tax</strong> Exemptions, by Major <strong>Tax</strong> (10^6 MT) 22Table 3-2. Do<strong>in</strong>g Bus<strong>in</strong>ess 2009, Pay<strong>in</strong>g <strong>Tax</strong>es Indicator Components 26Table 4-1. Pr<strong>in</strong>cipal Revenues as Share of GDP(%) 35Table 4-2. Revenue Buoyancy with Respect to Nom<strong>in</strong>al GDP 37Table 4-3. <strong>Mozambique</strong> VAT Productivity 39Table 4-4. VAT Productivity for <strong>Mozambique</strong> and Comparator Country Groups 40


<strong>II</strong>ITable 4-5. Fiscal Indicators, Three-year Averages, 2006-2008 (% GDP) 43Table 4-6. <strong>Mozambique</strong> Actual & Predicted Revenue-Ratio & Revenue-Effort 45Table 5-1. Reasons for Be<strong>in</strong>g Informal 59Table 5-2. Fiscal Benefit Code Investments as a Share of Total Approved Investments 61Table 5-3. Number of Firms by Size, Industry, and City 62Table 5-4. Shares of Firms Report<strong>in</strong>g 0% and 100% Sales Declared 65Table 5-5. Distribution of Sales Turnover by Firm Size 66Table 5-6. IRPS Rates 68Table 5-7. Rural and Urban Expenditure Share by Qu<strong>in</strong>tile 71Volume <strong>II</strong>: AppendicesAppendix A. Basic DataAppendix B. Fiscal Benefits <strong>in</strong> <strong>Mozambique</strong>Appendix C. <strong>Tax</strong>es and Fiscal Benefits: International ComparisonAppendix D. <strong>Tax</strong>ation and Bus<strong>in</strong>ess Environment Rat<strong>in</strong>gsAppendix E. Operational Efficiency IndicatorsAppendix F. Revenue EffortAppendix G. <strong>Tax</strong> Incentives for Investment—Pros and ConsAppendix H. Interview Results—Ma<strong>in</strong> Concerns and IssuesAppendix I. Persons Interviewed


AcknowledgementsThis report was written by Bruce Bolnick of Nathan Associates Inc. and Bruce Byiers, aconsultant to Nathan Associates, with contributions from Alexander Greenbaum and Molly Jamesof Nathan Associates and Egildo Massuanganhe, a consultant to the firm. The report reflects<strong>in</strong>formation available as of July 2009.The authors wish to express their deep gratitude for the support provided by everyone listed <strong>in</strong>Appendix I of Volume <strong>II</strong>, which conta<strong>in</strong>s the full list of people <strong>in</strong>terviewed for the study. Specialthanks are reserved for the assistance provided by Ashok Menon, Sab<strong>in</strong>a Sequeira, and SteliaNarotam from USAID’s Trade and Investment project; Herm<strong>in</strong>io Sueia and Ali Algy at theRevenue Authority; Brendan Kelly at the M<strong>in</strong>istry of Plann<strong>in</strong>g and Development; and EmmyBosten and Andrea Lemgruber from the IMF. As always, the authors are fully responsible for thecontent of the report.


Executive SummaryThis study is part of a wider evaluation of government performance <strong>in</strong> achiev<strong>in</strong>g the objectives of<strong>PARPA</strong> <strong>II</strong>, which has the ultimate goal of promot<strong>in</strong>g rapid, broad-based and susta<strong>in</strong>able growthto reduce poverty and improve standards of liv<strong>in</strong>g for the people of <strong>Mozambique</strong>. The studyexam<strong>in</strong>es the impact of the tax system on revenue mobilization, <strong>in</strong>vestment, sav<strong>in</strong>gs, job creation,and private sector development, with special attention to small and medium enterprises, as well asthe issue of tax equity. The study draws on a review of the tax laws and other documents relat<strong>in</strong>gto the tax system; an analysis of data on revenue performance, <strong>in</strong>clud<strong>in</strong>g <strong>in</strong>ternationalcomparisons; and extensive field <strong>in</strong>terviews conducted <strong>in</strong> July 2009. Based on the analysis, thestudy identifies 12 priority issues for the next <strong>PARPA</strong> and recommends correspond<strong>in</strong>g measuresfor consideration by the government.The <strong>Tax</strong> <strong>System</strong>S<strong>in</strong>ce 1998, all of the ma<strong>in</strong> tax <strong>in</strong>struments have undergone deep reform. Pr<strong>in</strong>cipal features of thesystem are now well aligned with best practices for develop<strong>in</strong>g countries, <strong>in</strong>clud<strong>in</strong>g primaryreliance on a broad-based value-added tax and a modern <strong>in</strong>come tax, supplemented by exciseduties on selected products and moderate import duties. The government has made notableprogress, too, <strong>in</strong> improv<strong>in</strong>g tax adm<strong>in</strong>istration s<strong>in</strong>ce establish<strong>in</strong>g the Revenue Authority (AT) <strong>in</strong>2006. As a result, most of the <strong>PARPA</strong> <strong>II</strong> tax objectives have been realized, <strong>in</strong>clud<strong>in</strong>g the centralaim of boost<strong>in</strong>g revenue from 14.0 percent of GDP <strong>in</strong> 2005 to 16.2 percent by 2009. The actualfigure for 2009 will fall short of this mark due to weak economic conditions, but AT data showthat the ratio reached 16.4 percent <strong>in</strong> 2008. Other achievements <strong>in</strong>clude enact<strong>in</strong>g a new General<strong>Tax</strong> Law clarify<strong>in</strong>g rules for tax collection and taxpayer rights; rationaliz<strong>in</strong>g fiscal benefits, <strong>in</strong>particular end<strong>in</strong>g the special regime for large projects; pass<strong>in</strong>g a new Municipal F<strong>in</strong>ance Act;eas<strong>in</strong>g the burden on small bus<strong>in</strong>esses by <strong>in</strong>creas<strong>in</strong>g tax thresholds and enact<strong>in</strong>g a Simplified <strong>Tax</strong>for Small Contributors (ISPC); and strengthen<strong>in</strong>g the tax regime for the m<strong>in</strong><strong>in</strong>g and petroleumsectors. Several objectives, however, are still works <strong>in</strong> progress. These <strong>in</strong>clude <strong>in</strong>tegrat<strong>in</strong>g tax andcustoms <strong>in</strong>formation systems; tax collection via the banks; improv<strong>in</strong>g audit revenues relative tototal revenue; moderniz<strong>in</strong>g tax adm<strong>in</strong>istration; and implement<strong>in</strong>g tax courts.There is little scope for revenue ga<strong>in</strong>s through further tax policy measures, but great potential for<strong>in</strong>creas<strong>in</strong>g revenue by broaden<strong>in</strong>g the effective tax base, allocat<strong>in</strong>g AT resources more efficiently,and facilitat<strong>in</strong>g taxpayer compliance through further measures to modernize tax adm<strong>in</strong>istration.These <strong>in</strong>clude functional <strong>in</strong>tegration of tax and customs operations; automation of procedures andsystems; risk management; human resource management; and development of a service cultureamong tax officials. Most of the required measures are <strong>in</strong>corporated <strong>in</strong> the AT’s Strategic Plan for


V<strong>II</strong>I<strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUE2009–2010. To be effective, it is essential that the reforms be carefully planned, properlysequenced, adequately funded, and well managed.The tax system <strong>in</strong> <strong>Mozambique</strong> is reasonably progressive. The poor bear very little of the taxburden as a result of high thresholds for enter<strong>in</strong>g the tax net (relative to per capita <strong>in</strong>come), thelow tax rate on small contributors via the ISPC, the progressive rate structure under the <strong>in</strong>dividual<strong>in</strong>come tax, and the structure of zero rat<strong>in</strong>gs and exemptions for basic necessities under the VAT.However, import duties and excise taxes on goods consumed by the poor have regressive effects.Tak<strong>in</strong>g <strong>in</strong>to account the strong revenue growth s<strong>in</strong>ce 2000, the likely effects of recent policyreforms, prospective reforms to tax adm<strong>in</strong>istration, and the natural responsiveness of revenues toeconomic growth, it should be possible for the government to achieve a revenue ratio of 18.5percent of GDP by 2015, <strong>in</strong>clud<strong>in</strong>g tax and non-tax revenues. International benchmarks suggest,however, that it is very difficult for low-<strong>in</strong>come countries to boost revenue beyond 18 to 19percent of GDP without special conditions such as oil wealth. In sett<strong>in</strong>g the medium-term revenuetarget, the government should strike a balance between public and private resource requirementsfor national development, job creation, and susta<strong>in</strong>able growth. This consideration suggests that aportion of the revenue ga<strong>in</strong> from reforms and the underly<strong>in</strong>g economic dynamics should be usedto reduce tax rates, rather than aim<strong>in</strong>g solely for a maximum revenue ratio.Effects of the <strong>Tax</strong> <strong>System</strong> on Private Sector DevelopmentStandard tax rates on bus<strong>in</strong>esses <strong>in</strong> <strong>Mozambique</strong> are now slightly above the regional and<strong>in</strong>ternational averages, due to the global trend towards lower company taxes. Nonetheless, theprevail<strong>in</strong>g tax rates should not be a deterrent to most <strong>in</strong>vestments, because other aspects of thebus<strong>in</strong>ess environment are far more important <strong>in</strong> determ<strong>in</strong><strong>in</strong>g the viability of most projects. Lowertax rates would, however, <strong>in</strong>crease the capacity of exist<strong>in</strong>g firms to f<strong>in</strong>ance expansion andimprove productivity by boost<strong>in</strong>g net earn<strong>in</strong>gs and cash flow. One case where tax rates have acritical effect is on <strong>in</strong>ternationally mobile (“footloose”) <strong>in</strong>vestments that could easily go to othercountries; these <strong>in</strong>vestments, however, should be covered by the fiscal benefits apply<strong>in</strong>g to IFZenterprises, which are highly competitive. In contrast, the effective tax rate is very burdensomefor <strong>in</strong>vestments that lack special <strong>in</strong>centives and face double taxation of dividend distributions to<strong>in</strong>dividual shareholders. This feature of the tax code creates a dis<strong>in</strong>centive for <strong>in</strong>corporation.The tax system affects the efficiency of <strong>in</strong>vestment, as well as the level. Special tax breaks underthe Code of Fiscal Benefits tend to have the greatest impact on <strong>in</strong>vestments with relatively lowrates of return, s<strong>in</strong>ce projects with an <strong>in</strong>tr<strong>in</strong>sically high return would be undertaken anyway. Thefiscal benefits also tilt the play<strong>in</strong>g field <strong>in</strong> favor of new <strong>in</strong>vestors who qualify for tax breaks,relative to exist<strong>in</strong>g bus<strong>in</strong>esses and other <strong>in</strong>vestors—especially small bus<strong>in</strong>esses. Similarly, taxbenefits that lower the cost of imported capital tend to encourage capital <strong>in</strong>tensity and importdependence, which <strong>in</strong>hibits the extent of job creation and domestic l<strong>in</strong>kages. Bus<strong>in</strong>ess l<strong>in</strong>kages tosmall enterprises are also h<strong>in</strong>dered, <strong>in</strong> less obvious ways, by the need for registered entities toobta<strong>in</strong> proper tax receipts from suppliers under the VAT and <strong>in</strong>come tax, as well as provisions for20 percent withhold<strong>in</strong>g tax on various transactions.The quality of tax adm<strong>in</strong>istration also has an important effect on <strong>in</strong>vestment and bus<strong>in</strong>essoperations. Indeed, concerns about arbitrary and punitive enforcement practices by tax officials,


E XECUTIVE S UMMARYIXcomplexity of the tax system, lack of taxpayer services, and difficulties <strong>in</strong> recover<strong>in</strong>g refunds arefar more prevalent than compla<strong>in</strong>ts about the level of taxes. These problems particularly affectsmall and medium size enterprises: the primary eng<strong>in</strong>e of job creation for the economy.<strong>Tax</strong> rates, compliance costs and complexity of the tax system also affect decisions by smallbus<strong>in</strong>esses to formalize. The Simplified <strong>Tax</strong> (ISPC) passed <strong>in</strong> 2009 should go a long way towardsreduc<strong>in</strong>g these tax-related barriers, if effectively implemented. The VAT, too, creates <strong>in</strong>centivesfor formalization, for any small enterprises that wish to do bus<strong>in</strong>ess with VAT-registeredcompanies. Incentives to formalize are also <strong>in</strong>fluenced by the cost of non-compliance, which isdeterm<strong>in</strong>ed by the AT’s capacity to enforce registration requirements.The effect of the tax system on sav<strong>in</strong>g is complex. Higher taxes could reduce bus<strong>in</strong>ess andhousehold sav<strong>in</strong>g but <strong>in</strong>crease government sav<strong>in</strong>g, leav<strong>in</strong>g an ambiguous net effect <strong>in</strong> the shortrun. In the medium to long term, however, the most important consideration is how the taxsystem affects the growth of <strong>in</strong>come, which is the primary determ<strong>in</strong>ant of sav<strong>in</strong>g.Priority <strong>Tax</strong> IssuesFrom the analysis of the tax trends and features of the current tax system, 12 issues emerge asmedium-term priorities for the next <strong>PARPA</strong>.• Medium-term revenue target. As noted above, a revenue ratio of 18.5 percent of GDPshould be achievable by 2015, with a strong commitment to further reform. Yet revenuemaximization is not the sole aim of tax policy. Depend<strong>in</strong>g on the desired balance betweenrevenue enhancement and tax relief for the private sector, the study suggests optionsrang<strong>in</strong>g from 16.5 to 18.5 percent of GDP. In any case, the revenue target for each year’sbudget program should be determ<strong>in</strong>ed from an analysis of prevail<strong>in</strong>g conditions.• <strong>Tax</strong> rates. Moderate tax cuts would provide the private sector with additional f<strong>in</strong>ancialresources for expansion, <strong>in</strong>novation, and job creation. Furthermore, <strong>Mozambique</strong> shouldnot lag too far beh<strong>in</strong>d the global and regional trend towards lower tax rates on company<strong>in</strong>come. The study therefore recommends plann<strong>in</strong>g for <strong>in</strong>come tax reductions as revenueprospects improve. Also recommended are a reduction <strong>in</strong> the withhold<strong>in</strong>g tax andelim<strong>in</strong>ation of double taxation on dividends disbursed to <strong>in</strong>dividual shareholders.• Capacity for tax analysis. Decisions on tax policy should be <strong>in</strong>formed by a carefulanalysis of the revenue effects and economic impacts. The study recommendsestablish<strong>in</strong>g a <strong>Tax</strong> Policy Unit at the M<strong>in</strong>istry of F<strong>in</strong>ance and tra<strong>in</strong><strong>in</strong>g a group ofspecialists to conduct quantitative and qualitative analysis of tax policy issues.• Functional <strong>in</strong>tegration with<strong>in</strong> the AT. The Revenue Authority can greatly improveefficiency and effectiveness by further moderniz<strong>in</strong>g tax adm<strong>in</strong>istration. One majorrequirement is functional <strong>in</strong>tegration of common operations between customs and taxservices, <strong>in</strong>clud<strong>in</strong>g audit, risk management, debt management, and customer services, aswell as human resource management and master data files for each taxpayer.• E-taxation. The government and the AT recognize the importance of <strong>in</strong>formationtechnologies (e-Tributação) and have begun plann<strong>in</strong>g and implement<strong>in</strong>g major reforms.Priorities <strong>in</strong>clude e-declarations, electronic payments through the banks, an automated


1. IntroductionThis study exam<strong>in</strong>es the impact of tax policy on susta<strong>in</strong>able revenue mobilization, broad-basedeconomic growth, the development of small and medium enterprises, and other basic objectivesof the government’s second Plano de Acção para a Redução da Pobreza Absoluta, or <strong>PARPA</strong> <strong>II</strong>.The study forms one part of a wider evaluation of government performance <strong>in</strong> achiev<strong>in</strong>g the<strong>PARPA</strong> <strong>II</strong> objectives, with the ultimate goal of reduc<strong>in</strong>g poverty, promot<strong>in</strong>g rapid, yet broadbasedand susta<strong>in</strong>able growth, to improve the standards of liv<strong>in</strong>g and welfare of the people of<strong>Mozambique</strong>.Consistent with the terms of reference, the study provides an overview of the pr<strong>in</strong>cipal issues, thema<strong>in</strong> reforms, and their impacts on <strong>PARPA</strong> <strong>II</strong> objectives, lead<strong>in</strong>g to recommendations onpriorities for further improvement of the tax system <strong>in</strong> the next <strong>PARPA</strong> period. The study drawson three sources of <strong>in</strong>formation: a thorough review of documents and studies relat<strong>in</strong>g to the taxsystem, <strong>in</strong>clud<strong>in</strong>g the ma<strong>in</strong> tax laws; an analysis of data on revenue performance, <strong>in</strong>clud<strong>in</strong>g<strong>in</strong>ternational comparisons; and extensive field <strong>in</strong>terviews conducted <strong>in</strong> July 2009. The analysisfocuses on developments dur<strong>in</strong>g the <strong>PARPA</strong> <strong>II</strong> time frame of 2006 to 2009, but also exam<strong>in</strong>es alonger period of data to put recent trends <strong>in</strong> perspective. 1A number of general po<strong>in</strong>ts that emerge from the study are worth highlight<strong>in</strong>g at the outset. Thefirst is that large strides have clearly been taken to improve the efficiency and capacity of the taxauthorities s<strong>in</strong>ce the establishment of the <strong>Mozambique</strong> Revenue Authority (AT) <strong>in</strong> 2006. This iswidely acknowledged by all parties, and reflects very positively on the recent reforms.A second basic observation, encountered <strong>in</strong> many of the field <strong>in</strong>terviews, is the scale of thechallenge <strong>in</strong> design<strong>in</strong>g and adm<strong>in</strong>ister<strong>in</strong>g tax laws that must deal effectively with firms rang<strong>in</strong>gfrom large mult<strong>in</strong>ational firms with access to highly-skilled account<strong>in</strong>g services to very smallenterprises and even family farmers, many of whom are illiterate and lack basic legal documents.Integrat<strong>in</strong>g all types of economic actors <strong>in</strong>to a s<strong>in</strong>gle tax system is no easy task for any taxadm<strong>in</strong>istration. Many of the issues raised <strong>in</strong> this study relate to this aspect.A third po<strong>in</strong>t merit<strong>in</strong>g emphasis is the divergence of views on the performance of the AT,depend<strong>in</strong>g on the background of the <strong>in</strong>terlocutor. Government officials generally emphasize the1 Although the stated objective of the study is to analyze the “impact” of the tax system on <strong>PARPA</strong> <strong>II</strong>objectives, the analysis does not <strong>in</strong>clude formal model<strong>in</strong>g to quantify cause and effect <strong>in</strong> the complexrelationship between tax variables and <strong>in</strong>dicators represent<strong>in</strong>g <strong>PARPA</strong> objectives. This approach isnecessitated by time and resource constra<strong>in</strong>ts, as well as the limited period of time-series data for<strong>Mozambique</strong>.


2 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEpositives, focus<strong>in</strong>g on recent and planned reforms to broaden the tax base and improve taxadm<strong>in</strong>istration. In contrast, many bus<strong>in</strong>ess representatives, while acknowledg<strong>in</strong>g the progress,emphasize the lack of a service culture on the part of the tax authorities, and the correspond<strong>in</strong>gabsence of a tax-pay<strong>in</strong>g culture on the side of many firms and <strong>in</strong>dividuals. One widely citedgrievance is that many tax officials cont<strong>in</strong>ue to arbitrary and punitive enforcement actions <strong>in</strong>response to m<strong>in</strong>or and un<strong>in</strong>tentional <strong>in</strong>fractions, <strong>in</strong> order to meet revenue targets, rather thanencourag<strong>in</strong>g compliance through education and support for the taxpayers. This approach createsobstacles and costs for bus<strong>in</strong>esses, underm<strong>in</strong>es the <strong>PARPA</strong> objective of us<strong>in</strong>g the tax system topromote the national bus<strong>in</strong>ess community, and leads to an unhelpful perception of the AT as apredatory bureaucracy. These deep-seated problems will not be overcome quickly, but attention isurgently needed.Alongside the divergent perceptions about AT performance are differ<strong>in</strong>g views on the role of thetax system. Government officials typically focus on the need for higher revenue yields, relative toGDP, to f<strong>in</strong>ance public services while reduc<strong>in</strong>g dependency on donor support. In contrast, many<strong>in</strong> the private sector argue that measures to broaden the tax base and improve collectionefficiency should be used as an opportunity to reduce tax rates <strong>in</strong> order to stimulate private<strong>in</strong>vestment and foster more rapid growth. All parties recognize the importance of both objectives,but op<strong>in</strong>ions differ widely on the appropriate balance.F<strong>in</strong>ally, it is clear from the research conducted for this study that few opportunities rema<strong>in</strong> toachieve sizeable revenue ga<strong>in</strong>s through changes <strong>in</strong> tax policy. Yet there is huge potential forimprov<strong>in</strong>g the revenue yield by strengthen<strong>in</strong>g and moderniz<strong>in</strong>g tax adm<strong>in</strong>istration throughmeasures that will broaden the effective tax base, allocate AT resources more productively, andfacilitate taxpayer compliance.The rema<strong>in</strong>der of this study is organized as follows: Chapter 2 provides an overview of the taxsystem, cover<strong>in</strong>g its characteristics at the start of the <strong>PARPA</strong> <strong>II</strong> period, the major issues aris<strong>in</strong>gdur<strong>in</strong>g that period, and recent reforms. Chapter 3 discusses <strong>in</strong> more detail the current tax system,<strong>in</strong>clud<strong>in</strong>g issues relat<strong>in</strong>g to tax adm<strong>in</strong>istration. Chapter 4 provides an analysis of revenueperformance over recent years, look<strong>in</strong>g at revenue targets, revenue performance, and <strong>in</strong>ternationalcomparisons. Chapter 5 assesses the impact of the tax system on broader outcomes for theeconomy <strong>in</strong> light of the goals set out <strong>in</strong> <strong>PARPA</strong> <strong>II</strong>. Chapter 6 then draws on the analysis fromearlier chapters to highlight key issues that warrant attention as priorities for the next <strong>PARPA</strong>period, along with prelim<strong>in</strong>ary recommendations for address<strong>in</strong>g these concerns. In a separatevolume, Appendix A presents basic data tables used <strong>in</strong> the revenue analysis; eight otherappendices provide supplementary <strong>in</strong>formation on aspects of the report, as referenced <strong>in</strong> the text.


2. BackgroundAs a prelude to the analysis of the current tax system, this chapter briefly reviews the system as itstood at the start of the <strong>PARPA</strong> <strong>II</strong> period. This is followed by a summary of <strong>PARPA</strong> <strong>II</strong> prioritiesand targets <strong>in</strong> the area of taxation, progress to date <strong>in</strong> achiev<strong>in</strong>g these goals, and major taxreforms adopted dur<strong>in</strong>g the time frame for <strong>PARPA</strong> <strong>II</strong>.THE TAX SYSTEM AT THE START OF <strong>PARPA</strong> <strong>II</strong>Between 1998 and 2002, all of the ma<strong>in</strong> tax <strong>in</strong>struments were subject to deep reforms.Subsequent studies by the IMF and the World Bank’s Foreign Investment Advisory Service(FIAS) concluded that the tax system <strong>in</strong> <strong>Mozambique</strong> conformed broadly to <strong>in</strong>ternational bestpractices for develop<strong>in</strong>g countries. 2The value added tax (Imposto sobre o Valor Acrescentado, or IVA) was enacted <strong>in</strong> 1998 and<strong>in</strong>troduced <strong>in</strong> 1999 at the rate of 17 percent for the normal regime. Exports and other designatedgoods were subject to zero tax (isenção completa) and qualified for a full refund of VAT paid on<strong>in</strong>puts. In addition, a long list of goods and services were exempt from VAT. The VAT code alsooffered a simplified regime (regime simplificada) to enterprises with a turnover less than 250million MT per year (about $12,500), consist<strong>in</strong>g of a 5 percent tax on gross sales. Enterpriseswith a turnover below 100 million MT per year (about $5,000) were exempt from VAT. Note thatlocal currency figures <strong>in</strong> 2005 referred to the old metical (equal to .001 new MT).In 2002, a new <strong>in</strong>come tax system was <strong>in</strong>troduced, with separate tax codes for corporations(Imposto sobre o Rendimento das Collectivas, or IRPC) and <strong>in</strong>dividuals (Imposto sobre oRendimento das Pessoas S<strong>in</strong>gulares, or IRPS). The standard IRPC tax rate was 32 percent,though <strong>in</strong>come from agricultural activity was subject to a tax rate of 10 percent until 2010. TheIRPS had progressive marg<strong>in</strong>al tax rates rang<strong>in</strong>g from of 10 to 32 percent applied to nearly allsources of personal <strong>in</strong>come, 3 with an adjustment for marital status and deductions for dependents.Income under 24 million MT (about $1,200) per year was not taxable, while the top rate appliedto <strong>in</strong>comes <strong>in</strong> excess of 1,008 million MT (about $50,000). The IRPC and IRPS also provided asimplified regime for small enterprises that lack organized accounts and have a turnover of nomore than 1,500 million MT (about $75,000) per year. For these entities, taxable <strong>in</strong>come could be2 See IMF (2005a); Versano et al. (2006); FIAS (2006).3 Technically, the IRPS def<strong>in</strong>es five categories: <strong>in</strong>come from employment (“dependent” <strong>in</strong>come);professional and bus<strong>in</strong>ess <strong>in</strong>come; capital <strong>in</strong>come; <strong>in</strong>come from build<strong>in</strong>gs; and other <strong>in</strong>comes. Eachcategory covers multiple types a variety of <strong>in</strong>come.


4 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEdeterm<strong>in</strong>ed by tax officials on the basis of “technical-scientific <strong>in</strong>dicators” or deemed to be 20percent of the value of sales of merchandise or products, or 30 percent for other <strong>in</strong>come.After several rounds of import duty reductions, standard duties <strong>in</strong> 2005 ranged from 0 to 25percent, accord<strong>in</strong>g to the import category:• 25 percent of consumer goods (Class C)• 7.5 percent for <strong>in</strong>termediate materials (Class I)• 5 percent for capital goods (Class K) and fuel (Class N)• 2.5 percent for raw materials (Class M)• 0 percent for basic goods (Class E)The simple average tariff was 12 percent, and the trade-weighted average was 9 percent. Bothfigures were among the lowest <strong>in</strong> the region. Accord<strong>in</strong>g to the Diagnostic Trade IntegrationStudy (DTIS) for <strong>Mozambique</strong>, the tariff rate for Class C still provided a fairly high level ofeffective protection to most f<strong>in</strong>al goods produced <strong>in</strong> <strong>Mozambique</strong>. 4 Beyond the standard tariffs,<strong>Mozambique</strong> had begun by 2005 to reduce duty rates on imports from SADC member states,under the SADC trade protocol. Duty free imports of SADC orig<strong>in</strong> covered over 1,500 tariffl<strong>in</strong>es, all <strong>in</strong>volv<strong>in</strong>g products that did not compete aga<strong>in</strong>st local enterprises. Other duty exemptionswere granted under the tax <strong>in</strong>centive regime for <strong>in</strong>vestors (Código de Benefícios Fiscais) and aspecial customs regime for selected manufactur<strong>in</strong>g <strong>in</strong>dustries (Regime Aduaneiro para aIndústria Transformadora).The Specific Consumption <strong>Tax</strong> (Imposto sobre Consumos Éspecificos, or ICE) was restructured<strong>in</strong> 1998 to complement the VAT. The ICE is an ad valorem excise tax on more than 140 products,at rates that ranged from 15 to 65 percent <strong>in</strong> 2005. Items subject to ICE <strong>in</strong>cluded luxuries (such ascars, jewelry, perfume, and sport<strong>in</strong>g goods) and health risks (particularly tobacco and alcohol).The tax is charged on both domestic products and imports. In addition, the Fuel <strong>Tax</strong> (<strong>Tax</strong>a deCombustíveis) was adopted as a specific nom<strong>in</strong>al levy per liter of various types of fuel. The taxwas restructured <strong>in</strong> 2003 with a formula for quarterly <strong>in</strong>flation adjustments to ma<strong>in</strong>ta<strong>in</strong> the realvalue of the tax per unit.Under Decree n o 6/2004 of 1 April, the tax stamp is imposed on a lengthy list of documents,contracts, papers, and economic activities. Many of the charges are fixed amounts per item, butothers are ad valorem assessments at rates rang<strong>in</strong>g from 10 percent on the fee for land useconcessions, to 0.02 percent on certa<strong>in</strong> collateral guarantees.In 2002, the government adopted a major reform of the Code of Fiscal Benefits for <strong>in</strong>vestment(Código de Benefícios Fiscais, or CBF). The reform was designed to rationalize the <strong>in</strong>centivesregime and improve its cost-effectiveness by concentrat<strong>in</strong>g on <strong>in</strong>vestment tax credits andaccelerated depreciation, while scal<strong>in</strong>g back tax holidays. Partial tax holidays were reta<strong>in</strong>ed foragriculture and for enterprises <strong>in</strong> Industrial Free Zones (IFZs). The new code also reta<strong>in</strong>edexemptions from import duty on capital goods, and standard exemptions from <strong>in</strong>direct tax forexporters operat<strong>in</strong>g <strong>in</strong> an IFZ.4 Nathan Associates (2004), p. 7-4.


B ACKGROUND 5By 2006 the government had therefore taken major strides to transform what had been an archaicand <strong>in</strong>efficient tax system <strong>in</strong>to one that was largely <strong>in</strong> l<strong>in</strong>e with <strong>in</strong>ternational best practices.Virtually every observer acknowledged great progress. Yet the post-reform system was stillsubject to sharp criticism, especially from the private sector. 5 As <strong>PARPA</strong> <strong>II</strong> was be<strong>in</strong>g prepared,the most prom<strong>in</strong>ent tax issues <strong>in</strong>cluded:• Low tax yield. In 2005, government revenues totaled 13.8 percent of GDP, and taxrevenues amounted to 11.2 percent of GDP. An IMF analysis found the tax ratio to bewell below the norm for sub-Saharan Africa, and far short of what could be achieved <strong>in</strong><strong>Mozambique</strong>. 6 The government, too, was deeply concerned to improve the revenue yield<strong>in</strong> order to reduce dependency on donor fund<strong>in</strong>g. This was reflected <strong>in</strong> the <strong>PARPA</strong> goalof boost<strong>in</strong>g revenue to 16.2 percent of GDP by 2009.• High tax burden on companies. Two studies found that the tax rate on companies wasbroadly <strong>in</strong> l<strong>in</strong>e with regional averages, and that <strong>in</strong>vestors obta<strong>in</strong><strong>in</strong>g fiscal benefits faced amarg<strong>in</strong>al effective tax rate (METR) that was very competitive compared with othercountries <strong>in</strong> the region. For bus<strong>in</strong>esses without tax <strong>in</strong>centives, however, the METR wasunusually high by regional and <strong>in</strong>ternational standards, due to the comb<strong>in</strong>ed effect of thecompany tax plus the tax on distributed dividends.• Narrow tax base. The comb<strong>in</strong>ation of high effective tax rates and low tax revenuesunderscored the problem of gap<strong>in</strong>g holes <strong>in</strong> the tax base. Both the local private sector andthe IMF recommended stronger efforts to broaden the tax base, improve tax compliance,and reduce the generosity of tax breaks for large scale projects.• Tilted play<strong>in</strong>g field and <strong>in</strong>equitable tax burden. The tax system gave a competitiveadvantage to companies enjoy<strong>in</strong>g tax <strong>in</strong>centives, and also to enterprises that escaped thetax net by virtue of <strong>in</strong>formal operations, bribery, or tax fraud. The system tendedespecially to penalize tax-registered small and medium enterprises.• Excessive complexity. Representatives of the private sector were especially critical of thecomplexity of the <strong>in</strong>come tax and the value added tax. One commentator likened the taxcode to a “Mercedes stuck <strong>in</strong> the sand” – modern, well eng<strong>in</strong>eered and powerful, but<strong>in</strong>appropriate for conditions <strong>in</strong> <strong>Mozambique</strong>.• Lack of public <strong>in</strong>formation and taxpayer services. Even though the <strong>in</strong>come tax and VATservices conducted sem<strong>in</strong>ars and public appearances to present the 2002 tax codes, theeffort fell far short of the need. Equally important was the general absence of a serviceculture among tax officials.• Arbitrary and punitive enforcement. Another major compla<strong>in</strong>t by the private sector wasarbitrary and punitive behavior by tax <strong>in</strong>spectors, especially outside of Maputo. Becausethe tax code was complex, many bus<strong>in</strong>esses (especially SMEs) could not comply fully5 See FIAS (2006); Bolnick (2004); Versano et al. (2006); Arndt, Tarp and Jones (2009); CTA (2006).6 Versano (2006); IMF (2005a). Jones (2009), however, conducted an econometric analysis show<strong>in</strong>g that<strong>Mozambique</strong>’s tax ratio through 2003 was very close to the value predicted, based on the country’sstructural characteristics.


6 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEwith all of its technicalities. Consequently, many entrepreneurs faced arbitrary treatmentand costly penalties, which created <strong>in</strong>centives for unofficial side payments to taxofficials.• VAT refund delays. The s<strong>in</strong>gle most contentious issue <strong>in</strong> 2006 was the problem faced bymany bus<strong>in</strong>esses, especially exporters and contractors, <strong>in</strong> obta<strong>in</strong><strong>in</strong>g timely payment ofVAT refunds. Even large bus<strong>in</strong>esses had problems submitt<strong>in</strong>g acceptable refundpetitions, due to the complex paperwork requirements.S<strong>in</strong>ce 2006 the government has been address<strong>in</strong>g many of these issues. Even so, every po<strong>in</strong>t on thelist rema<strong>in</strong>s today as a major concern.TAX ISSUES IN <strong>PARPA</strong> <strong>II</strong><strong>Tax</strong> issues appear at various po<strong>in</strong>ts <strong>in</strong> <strong>PARPA</strong> <strong>II</strong>. The Macroeconomic and Fiscal Scenario <strong>in</strong>Chapter V (paragraphs 130-131) targets an <strong>in</strong>crease <strong>in</strong> total State revenues from 14.0 percent ofGDP <strong>in</strong> 2005 to 16.2 percent <strong>in</strong> 2009. This is a primary target <strong>in</strong> the Matrix of StrategicIndicators, for the category of Public F<strong>in</strong>ance Management. Accord<strong>in</strong>g to the fiscal scenario, theadditional revenue “must be mobilized <strong>in</strong> a way that does not jeopardize growth of the privatesector or underm<strong>in</strong>e the <strong>in</strong>centives to pursue economic activities <strong>in</strong> the formal sector, byexpand<strong>in</strong>g the tax base curb<strong>in</strong>g tax evasion, and reduc<strong>in</strong>g tax <strong>in</strong>centives.” 7<strong>PARPA</strong> <strong>II</strong> addresses tax policy explicitly under the Economic Development Pillar, <strong>in</strong> the sectionon Macroeconomic Management (Chapter V<strong>II</strong>I, paragraph 489). The objective here is “to reformand <strong>in</strong>crease the efficiency of the tax adm<strong>in</strong>istration with a view to gradually <strong>in</strong>creas<strong>in</strong>g themobilization of domestic funds as a percentage of GDP, with the idea of reduc<strong>in</strong>g externaldependency.” To this end, the <strong>PARPA</strong> lists eight priority actions:• “Domestic revenues will gradually be <strong>in</strong>creased;• “The tax system will be simplified and ref<strong>in</strong>ed, and the tax base broadened;• “Reforms made <strong>in</strong> direct and <strong>in</strong>direct taxes will be consolidated;• “Simplified taxation regimes will be reviewed, the effectiveness of tax and <strong>in</strong>vestment<strong>in</strong>centives will be evaluated, and the process of establish<strong>in</strong>g tax courts will be cont<strong>in</strong>ued;• “Work on moderniz<strong>in</strong>g the tax adm<strong>in</strong>istration will be cont<strong>in</strong>ued, to make it an efficienttax-collection system and to curb fraud and tax evasion;• “Legislation will be approved that simplifies the relationship between the taxadm<strong>in</strong>istration and the taxpayers, mak<strong>in</strong>g it easier for them to exercise their rights andreceive the protection assured them;• “<strong>Tax</strong> and customs courts will be effectively implemented; and7 The statements cited here from <strong>PARPA</strong> <strong>II</strong> are based on the English translation posted by the IMF asCountry Report No. 07/37.


B ACKGROUND 7• “Legislation on local government f<strong>in</strong>ances will be ref<strong>in</strong>ed and the conditions of theresponsible for collection and control of local government taxes will beimproved.”Revenue issues also appear <strong>in</strong> other parts of the <strong>PARPA</strong>. In the section on Promot<strong>in</strong>g PrioritySectors, Broaden<strong>in</strong>g the Bus<strong>in</strong>ess Base, and Creat<strong>in</strong>g Jobs: Industry (Chapter V<strong>II</strong>I, paragraph537) one key action is to “establish tax-free zones, on the condition that an evaluation be made, <strong>in</strong>advance, of the net contribution to the national economy and to tax revenues.” The section onM<strong>in</strong><strong>in</strong>g (Chapter V<strong>II</strong>I, paragraph 551) covers several objectives, one of which is to strengthen thelegal and fiscal framework “as a way of <strong>in</strong>creas<strong>in</strong>g <strong>in</strong>vestment, tax revenues, and exports by them<strong>in</strong><strong>in</strong>g sector.” F<strong>in</strong>ally, the section on Assistance from Cooperation Partners (Chapter IXparagraph 611) states that the government will, among other th<strong>in</strong>gs, exam<strong>in</strong>e “the ideal balancebetween the weight of domestic funds (tax revenues) and external aid as regards Stateexpenditures.”Table 2-1 lists the major <strong>PARPA</strong> <strong>II</strong> objectives, actions and strategic <strong>in</strong>dicators relat<strong>in</strong>g to the taxsystem, along with notes on the status of implementation. Most targets have either been achievedor are now <strong>in</strong> process. In particular, the government succeeded <strong>in</strong> 2008 <strong>in</strong> meet<strong>in</strong>g the targetrevenue ratio that was set for 2009; this year will see a shortfall <strong>in</strong> the revenue ratio, however, dueto the adverse effects of the global economic crisis via slower growth of GDP and trade, and asharp reduction <strong>in</strong> commodity prices. The government also succeeded <strong>in</strong> <strong>in</strong>troduc<strong>in</strong>g a newsimplified tax regime for small bus<strong>in</strong>esses, and revis<strong>in</strong>g the code of fiscal benefits.In some respects, however, achievements fell well short of the <strong>PARPA</strong> <strong>II</strong> targets. The areas ofdeficiency or delay <strong>in</strong>clude the program for implement<strong>in</strong>g new tax <strong>in</strong>formation systems,<strong>in</strong>creas<strong>in</strong>g the collection rate from tax audits, establish<strong>in</strong>g tax courts, and curb<strong>in</strong>g tax fraud andtax evasion.MAJOR TAX REFORMS 2006-2009A series of reforms to the tax code and measures to improve tax adm<strong>in</strong>istration have beenundertaken dur<strong>in</strong>g the period 2006 to 2009. By and large, the reforms have aligned well with the<strong>PARPA</strong> <strong>II</strong> objectives and targets for tax policy.The period began with a new law (Law 1/2006) establish<strong>in</strong>g the <strong>Mozambique</strong> <strong>Tax</strong> Authority(Autoridade Tributária, or AT), comb<strong>in</strong><strong>in</strong>g formerly separate organizations for tax and customs.The purpose of this reform was to create a new organizational foundation for moderniz<strong>in</strong>g taxadm<strong>in</strong>istration and professionaliz<strong>in</strong>g the tax service <strong>in</strong> order to strengthen enforcement, improveservices, combat evasion, and expand the tax-base. S<strong>in</strong>ce the AT began operations, the number ofregistered taxpayers (<strong>in</strong>dividuals plus enterprises) has more than doubled from 374,181 <strong>in</strong> 2006 to864,317 by July 2009. The AT expanded its staff by 18 percent and opened 14 new collectionoffices between 2007 and the first half of 2009; <strong>in</strong>troduced new programs for public educationand customer service; reduced delays <strong>in</strong> the approval of VAT refunds; <strong>in</strong>creased the number ofaudits of large tax payers; <strong>in</strong>troduced post-clearance audits for customs; simplified tax declaration


8 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEforms; and reduced the average time to release imported goods at the border. In addition, the ATopened a new tra<strong>in</strong><strong>in</strong>g <strong>in</strong>stitute <strong>in</strong> 2009 to <strong>in</strong>tensify staff tra<strong>in</strong><strong>in</strong>g. 8Table 2-1<strong>Tax</strong> Objectives, Actions, and Strategic Indicators <strong>in</strong> <strong>PARPA</strong> <strong>II</strong><strong>PARPA</strong> <strong>II</strong> Element Source <strong>in</strong> <strong>PARPA</strong> Status: July 20091. Objective: Progressively <strong>in</strong>creasedomestic revenues and expand tax base.Goal for 2009: Total Revenues as % ofGDP = 16.2%2. Action: Rationalize the grant<strong>in</strong>g andimprov<strong>in</strong>g the management of fiscalbenefits.Goal: <strong>System</strong>atize data on benefits;published statisticsStrategic Indicators Matrix,Management of Public F<strong>in</strong>ances,Objective 5As aboveAT (2009): Total Revenue = 16.4% of GDP for2008IMF (2009): Total Revenue = 16.0% of GDP for2008AT target of 17.3% for 2009 will not be met due toeffects of the global economic crisis; IMF projectsratio decl<strong>in</strong><strong>in</strong>g to 15.7% this year.New Code of Fiscal Benefits approved <strong>in</strong> January2009, end<strong>in</strong>g the special negotiation regime forlarge projects. (For details see Appendix B of thepresent report.)Accord<strong>in</strong>g to CPI, procedures for approval of an<strong>in</strong>vestment rema<strong>in</strong> largely unchanged, but theprocess of obta<strong>in</strong><strong>in</strong>g tax and duty relief has beenstreaml<strong>in</strong>ed s<strong>in</strong>ce the AT was established.Survey of <strong>in</strong>vestors obta<strong>in</strong><strong>in</strong>g CPI approvals <strong>in</strong>2005, 2006 and 2007 shows that 47 of 60 had nosignificant problem with the procedures, but 13 of60 reported major problems <strong>in</strong> terms of costs ordelays (Bolnick, 2009).AT reports data on tax expenditures <strong>in</strong> its AnnualReports s<strong>in</strong>ce 2007; the data also appear <strong>in</strong> theAnnual Report on the Conta Geral do Estado fromthe Tribunal Adm<strong>in</strong>istrativo (without <strong>in</strong>formation onmethodology).3. Action: Integration of <strong>in</strong>formationsystems management <strong>in</strong>to the ATM with<strong>in</strong>the context of the approved PDTI.Goal: By 2009 <strong>in</strong>tegrated systems <strong>in</strong> fulloperation, <strong>in</strong>clud<strong>in</strong>g bank<strong>in</strong>g collectionmanagement moduleAs above Plann<strong>in</strong>g well advanced for e-taxation module of e-SISTAFE, <strong>in</strong>clud<strong>in</strong>g e-declarations, electroniccollection via banks, and customs s<strong>in</strong>gle w<strong>in</strong>dow;pilot system for e-tax module procured and tra<strong>in</strong><strong>in</strong>gbegun via UTRAFE.4. Action: Increase audits.Goal: To be def<strong>in</strong>ed. Basel<strong>in</strong>e annualrevenue recovered by audit = 0.2% ofGDP.5. Objective: Strengthen <strong>in</strong>stitutionalcapacity of local governments.Action: Development of municipalrevenue collection capacity.Goal for 2009: Own revenues, % ofannual budget:Capital cities 60%Other cities 55%Villages 45%As aboveStrategic Indicators Matrix, PublicSector Reform, Objective 9AT Annual Report for 2008 (Table 18) showsrevenue collection from audits = 0.4% of totalreceipts <strong>in</strong> 2007, and 0.2% of total receipts <strong>in</strong> 2008.Municipality of Maputo (calculated from OrcamentoRectificativo 2009 do Municipio de Maputo):Own current revenue/current revenue2008 = 71.5%Budget 2009 = 76.5%Own total revenue/total revenue = Own totalrevenue/budget2008 = 36.8%Budget 2009 = 27.8%8 The data are from IMF (2008, 2008a, 2009b and 2009c), AT Annual Reports for 2007 and 2008, and<strong>in</strong>formation provided by the AT’s Gab<strong>in</strong>ete de Planeamento Estúdios e Cooperação Internacional.


B ACKGROUND 9<strong>PARPA</strong> <strong>II</strong> Element Source <strong>in</strong> <strong>PARPA</strong> Status: July 20096. Domestic revenues will gradually be<strong>in</strong>creasedEconomic Development pillar,Macroeconomic Management and <strong>Tax</strong>Policy - Paragraph 489See item 1 above.7. The tax system will be simplified andref<strong>in</strong>ed, and the tax base broadened8. Reforms made <strong>in</strong> direct and <strong>in</strong>directtaxes will be consolidatedAs above AT Annual Report for 2007 (Table 15) and 2008(Table 30) show new <strong>in</strong>come tax registrations =100227 <strong>in</strong> 2006; 190019 <strong>in</strong> 2007; 191140 <strong>in</strong> 2008.Various IMF reports show total number of taxpayers<strong>in</strong>creas<strong>in</strong>g from 295000 <strong>in</strong> 2005 to 587205 <strong>in</strong> 2007(latest figure).New Simplified <strong>Tax</strong> (ISPC) approved 2009 (seeitem 9)As above New IRPC, IRPS, IVA codes approved <strong>in</strong> 2007consolidat<strong>in</strong>g amendments and other statutorychanges.9. Simplified taxation regimes will bereviewed, the effectiveness of tax and<strong>in</strong>vestment <strong>in</strong>centives will be evaluated,and the process of establish<strong>in</strong>g tax courtswill be cont<strong>in</strong>ued10. Work on moderniz<strong>in</strong>g the taxadm<strong>in</strong>istration will be cont<strong>in</strong>ued, to makeit an efficient tax-collection system and tocurb fraud and tax evasionAs aboveAs aboveNew Simplified <strong>Tax</strong> (ISPC) approved 2009 forsmall bus<strong>in</strong>esses with turnover up to 2,500,000 MT.ISPC replaces simplified regimes for <strong>in</strong>come tax andVAT with s<strong>in</strong>gle tax at 3% of turnover, up to amaximum liability of 75,000 MT.<strong>Mozambique</strong> Revenue Authority (AT) establishedby Law <strong>in</strong> 2006, with operations start<strong>in</strong>g 2007.General agreement among stakeholders <strong>in</strong>terviewedfor this report (government, private sector, anddonors) that AT has achieved major ga<strong>in</strong>s <strong>in</strong>professionalism and efficiency <strong>in</strong> tax collection –but also widespread view that tax fraud and taxevasion are still rampant.11. Legislation will be approved thatsimplifies the relationship between the taxadm<strong>in</strong>istration and the taxpayers, mak<strong>in</strong>git easier for them to exercise their rightsand receive the protection assured themAs above New General <strong>Tax</strong> Code approved <strong>in</strong> 2006establish<strong>in</strong>g clear rules and procedures for exerciseand protection of taxpayer rights.12. <strong>Tax</strong> and customs courts will beeffectively implemented13. Legislation on local governmentf<strong>in</strong>ances will be ref<strong>in</strong>ed and the conditionsof the responsible for collection andcontrol of local government taxes will beimproved.14. Establish tax-free zones, on thecondition that an evaluation be made, <strong>in</strong>advance, of the net contribution to thenational economy and to tax revenues.15. Strengthen the legal and fiscalframework for m<strong>in</strong><strong>in</strong>g “as a way of<strong>in</strong>creas<strong>in</strong>g <strong>in</strong>vestment, tax revenues, andexports16. Exam<strong>in</strong>e the ideal balance between theweight of domestic funds (tax revenues)and external aid as regards Stateexpenditures.As aboveAs aboveEconomic Development pillar,Promot<strong>in</strong>g Priority Sectors,Broaden<strong>in</strong>g the Bus<strong>in</strong>ess Base, andCreat<strong>in</strong>g Jobs: Industry - Paragraph537Economic Development pillar,Promot<strong>in</strong>g Priority Sectors,Broaden<strong>in</strong>g the Bus<strong>in</strong>ess Base, andCreat<strong>in</strong>g Jobs: M<strong>in</strong><strong>in</strong>g - Paragraph551Factors Determ<strong>in</strong><strong>in</strong>g Success,Assistance from Cooperation Partners(Chapter IX paragraph 611)Customs courts <strong>in</strong> operation; few cases on thedocket. Legal framework for tax courts <strong>in</strong> place;judges selected and tra<strong>in</strong>ed; operations expected tostart soon <strong>in</strong> Maputo followed by Beira andNampula. Large backlog of tax disputes(contenciosos) requir<strong>in</strong>g attention.New Local Government F<strong>in</strong>ance Act passed <strong>in</strong>2008, clarify<strong>in</strong>g authority of local governments forexpenditures and own revenue sources.New Fiscal Benefits Code <strong>in</strong>cludes <strong>in</strong>centives forSpecial Economic Zones (ZEEs); plans welladvanced for establishment of ZEE <strong>in</strong> Nampula.Study conducted on revenue effects of the NampulaZEE.New M<strong>in</strong><strong>in</strong>g and Petroleum Act approved <strong>in</strong> 2007,restructur<strong>in</strong>g the regime of tax <strong>in</strong>centives applicableto <strong>in</strong>vestment <strong>in</strong> these sectors to <strong>in</strong>crease revenuesfor the state while rema<strong>in</strong><strong>in</strong>g attractive to <strong>in</strong>vestors.<strong>PARPA</strong> <strong>II</strong> target for <strong>in</strong>creas<strong>in</strong>g the ratio of domesticrevenue to GDP by 0.5% percentage po<strong>in</strong>ts eachyear largely achieved, except for shortfall <strong>in</strong> 2009due to effects of the global economic crisis.


10 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEThe adm<strong>in</strong>istrative reforms have been accompanied by a raft of new tax legislation andaccompany<strong>in</strong>g regulations, which are summarized <strong>in</strong> Table 2-2.Table 2-2Major Changes <strong>in</strong> the <strong>Tax</strong> Law, 2006 – 2009Legal InstrumentTitle and DescriptionLaw 1/2006 of 22 March and Decree29/2006 of 30 AprilLaw 2/2006 of 22 March<strong>Tax</strong> Authority of <strong>Mozambique</strong> – Organic statute for establishment of the AT.General Law on <strong>Tax</strong>ation – General pr<strong>in</strong>ciples and provisions regulat<strong>in</strong>g responsibilities andobligations of the tax authorities at all levels and taxpayer rights and responsibilities.Law 3/2007 of 7 February Customs Law – Reduces the duty rate on consumer goods from 25% to 20%.Laws 11/2007, 12/2007, 13/2007 of27 June and Decrees 4/2008 and5/2008 of 9 AprilLaw 28/2007 of 4 DecemberLaw 32/2007 of 31 December andDecree 7/2008 of 16 AprilLaw 33/2007 of 31 December andDecree 8/2008 of 16 AprilLaw 34/2007 of 31 December andDecree 9/2008 of 16 AprilLaw 1/2008 of 16 January andDecree 632008 of 30 DecemberLaw 4/2009 of 12 JanuaryLaw 5/2009 of 12 January andDecree 14/2009 of 14 AprilEITI Board Approval of candidacyfor <strong>Mozambique</strong>, 18 May 2009Law ??/2009 [forthcom<strong>in</strong>g]M<strong>in</strong><strong>in</strong>g and Petroleum Laws – Reforms the tax regime and surface concessions for m<strong>in</strong><strong>in</strong>g andoil production activities, and the <strong>in</strong>centives regime for <strong>in</strong>vestment projects fall<strong>in</strong>g under theM<strong>in</strong><strong>in</strong>g Law and the Petroleum Law.Inheritance and Gift <strong>Tax</strong> CodeValue Added <strong>Tax</strong> Code (IVA) – Consolidates amendments and other legal changes relat<strong>in</strong>g to theVAT code; raises the threshold for registration and expands the scope of the simplified taxscheme; revises the list of exempt goods and transactions.Personal Income <strong>Tax</strong> Code (IRPS) - Consolidates amendments and other legal changes relat<strong>in</strong>gto the VAT code, and adjusts tax bracket thresholds and family allowances; also <strong>in</strong>troduces taxon <strong>in</strong>come from traded securities and term deposits.Corporate Income <strong>Tax</strong> Code (IRPC) - Consolidates amendments and other legal changes relat<strong>in</strong>gto the VAT code, and raises the threshold for pay<strong>in</strong>g tax.Municipal F<strong>in</strong>ance Law – Def<strong>in</strong>es the f<strong>in</strong>ancial, budgetary and asset regimes for localgovernments, and reforms the municipal tax system.Code of Fiscal Benefits (CBF) – Reforms the available package of fiscal benefits for <strong>in</strong>vestmentsapproved under the Investment Act, <strong>in</strong>clud<strong>in</strong>g both generic benefits and specific benefits fordesignated types of projects.Simplified <strong>Tax</strong> for Small <strong>Tax</strong>payers (ISPC) – Introduces a new simplified tax system for smallenterprises, replac<strong>in</strong>g simplified tax regimes for both <strong>in</strong>come tax and VAT.This measure beg<strong>in</strong>s the formal process of qualify<strong>in</strong>g <strong>Mozambique</strong> for participation <strong>in</strong> andadherence to the Extractive Industries Transparency Initiative.Excise <strong>Tax</strong> Code (ICE) – Restructures the excise tax rates on certa<strong>in</strong> goods <strong>in</strong>clud<strong>in</strong>g w<strong>in</strong>es,tobacco products, jewelry, and vehicles; elim<strong>in</strong>ates excise tax on some toys and sport<strong>in</strong>g goods.SOURCE: Deloitte, GTZ APSP and ACIS (2009) and MozLegal, <strong>Mozambique</strong> Investor (2006-2009).In March 2006, a new General Law on <strong>Tax</strong>ation redef<strong>in</strong>ed the basic framework govern<strong>in</strong>g the taxsystem, <strong>in</strong>clud<strong>in</strong>g the duties and responsibilities of the tax authorities, and the obligations andrights of the taxpayers. In 2007, the government adopted a new VAT code that altered the list ofexempt products, and <strong>in</strong>creased the turnover thresholds for VAT exemption and for coverage bythe simplified VAT regime. At the same time, the government revised the tax codes for <strong>in</strong>dividualand corporate <strong>in</strong>come taxes. Both tax codes were shortened considerably, as some provisionswere shifted <strong>in</strong>to the regulations, and others were covered by the new General Law on <strong>Tax</strong>ation.In addition, the new IRPC code updated thresholds for coverage by the simplified regime, and the


B ACKGROUND 11IRPS code updated the threshold for non-taxable <strong>in</strong>come and the various tax brackets. <strong>Tax</strong> rates,however, were unchanged, and the tax structure itself was not noticeably simplified. 9Perhaps the most important reform of 2007, <strong>in</strong> terms of future revenue potential, was the passageof new laws govern<strong>in</strong>g m<strong>in</strong><strong>in</strong>g and petroleum concessions. Among other th<strong>in</strong>gs, the legislationremoved the authority of the Council of M<strong>in</strong>isters to negotiate special tax arrangements form<strong>in</strong><strong>in</strong>g and petroleum <strong>in</strong>vestments and reduced the fiscal benefits for such projects. Morerecently, <strong>in</strong> May 2009, the government followed this up by formally <strong>in</strong>itiat<strong>in</strong>g procedures toqualify for the Extractive Industries Transparency Initiative. Once completed, the EITI processshould <strong>in</strong>stitutionalize improvements <strong>in</strong> transparency and management of revenue from<strong>in</strong>vestments <strong>in</strong> extractive <strong>in</strong>dustries.Other policy reforms <strong>in</strong> 2007 <strong>in</strong>volved a reduction <strong>in</strong> the top rate of customs duty on consumergoods from 25 percent to 20 percent and a new code for <strong>in</strong>heritance and gift tax. In 2008 thegovernment adopted a new Municipal F<strong>in</strong>ance Law, <strong>in</strong>clud<strong>in</strong>g reforms to the municipal taxsystem <strong>in</strong> l<strong>in</strong>e with the <strong>PARPA</strong> objective of decentralization.The SADC free trade area came <strong>in</strong>to effect <strong>in</strong> January 2008, entail<strong>in</strong>g duty-free treatment of mostimports from other SADC member states, though the complex rules of orig<strong>in</strong> have impairedaccess to duty-free trade <strong>in</strong> both directions (IMF 2009a). Additional tariff reductions will be<strong>in</strong>troduced through the Interim Economic Partnership Agreement (EPA) between <strong>Mozambique</strong>,other participat<strong>in</strong>g SADC members, and the EU. Once the EPA is f<strong>in</strong>alized and approved,participat<strong>in</strong>g SADC states will reta<strong>in</strong> quota- and duty-free access to the EU and, on a reciprocalbasis, will reduce customs tariffs on EU imports.In response to civil unrest caused by a steep rise <strong>in</strong> fuel and food prices <strong>in</strong> early 2008 (of 27 and15 percent respectively), the government suspended fuel price <strong>in</strong>creases until June and thendeferred fuel-related taxes until end-2008 (IMF 2008), though this measure was reversed <strong>in</strong> 2009(IMF 2009a).More recently, <strong>in</strong> 2009, the National Assembly revised the Fiscal Benefits Code, passed a newSimplified <strong>Tax</strong> for Small <strong>Tax</strong>payers (ISPC) and revised the excise tax code, <strong>in</strong>clud<strong>in</strong>gadjustments <strong>in</strong> the tax rates on designated alcoholic beverages, tobacco products and luxurygoods. These major reforms are discussed <strong>in</strong> more detail <strong>in</strong> the next chapter.9 Frei, A. The Mozambican Investor, Issues 81 (February 26, 2008) and 82 (March 04, 2008).


3. The Current <strong>Tax</strong> <strong>System</strong>The ma<strong>in</strong> features of the tax code prevail<strong>in</strong>g <strong>in</strong> 2009 are essentially the same as the basel<strong>in</strong>esystem <strong>in</strong> 2005, outl<strong>in</strong>ed <strong>in</strong> the previous chapter. The discussion here provides more details andupdates the <strong>in</strong>formation to 2009. Particular attention is focused on the 2009 Simplified <strong>Tax</strong> forSmall <strong>Tax</strong>payers (ISPC) and the 2009 Code of Fiscal Benefits (CBF). The chapter also exam<strong>in</strong>esthe stand<strong>in</strong>g of <strong>Mozambique</strong>’s tax system <strong>in</strong> the World Bank’s Do<strong>in</strong>g Bus<strong>in</strong>ess rat<strong>in</strong>gs. The f<strong>in</strong>alsection summarizes the Revenue Authority’s plans for further reforms to modernize taxadm<strong>in</strong>istration.THE TAX SYSTEM IN 2009The current tax system <strong>in</strong> <strong>Mozambique</strong> conforms broadly to <strong>in</strong>ternational standards for goodpractice <strong>in</strong> develop<strong>in</strong>g countries, with the ma<strong>in</strong> sources of revenue com<strong>in</strong>g from the value addedtax (IVA) and modern corporate and <strong>in</strong>dividual <strong>in</strong>come taxes (IRPC and IRPS, respectively), andtaxes on trade decl<strong>in</strong><strong>in</strong>g <strong>in</strong> importance. 10Figure 3-1 shows the pr<strong>in</strong>cipal sources of domestic revenue (exclud<strong>in</strong>g grants) <strong>in</strong> 2008 comparedto the basel<strong>in</strong>e conditions <strong>in</strong> 2005. 11 In 2008, total revenue amounted to an estimated 16.0 percentof GDP, with tax revenue (receitas fiscais) at 13.5 percent of GDP. Both numbers represent an<strong>in</strong>crease by more than two percentage po<strong>in</strong>ts relative to GDP between 2005 and 2008, despitereductions <strong>in</strong> the top rate of customs duty, implementation of SADC trade preferences, higherthresholds for enter<strong>in</strong>g the tax net for the VAT and <strong>in</strong>come tax, and temporary deferment of thefuel tax <strong>in</strong> 2008 to cushion the effects of soar<strong>in</strong>g world prices for petroleum. These facts showthat government efforts to improve tax adm<strong>in</strong>istration and enlarge the tax base have <strong>in</strong>deed beenbear<strong>in</strong>g fruit.10 The IMF (2008a, p.13) states that “The legal framework for the tax system is modern, comprehensive,and generally broadly dissem<strong>in</strong>ated.”11 These figures are from the Medium-Term Fiscal Framework of the M<strong>in</strong>istry of Plann<strong>in</strong>g andDevelopment (MPD), as of June 2009. Some figures differ from those <strong>in</strong> the IMF’s Article IV Reviewfrom May 2009 (IMF, 2009b), and from data presented by the AT <strong>in</strong> their Annual Report for 2008 (AT,2009). For example, for 2007 the three sources, respectively, show a ratio of total revenue to GDP of 16.2,15.9, and 16.6 percent. For 2008 the correspond<strong>in</strong>g figures are 16.0, 16.0 and 16.4 percent.


14 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEFigure 3-1Pr<strong>in</strong>cipal Sources of Domestic Revenue, 2005 and 2008Revenue Structure, 2005 and 2008Percent of Total Revenue100%90%80%70%60%50%40%30%20%10%0%2005 2008Source: Calculated from data provided by DNEAP, M<strong>in</strong>istry of Plann<strong>in</strong>g and DevelopmentOther revenueFuel taxCustoms dutiesExciseVATIRPC (company)IRPS (personal)Revenue Structure, 2005 and 2008Percent of GDP18.0%16.0%14.0%12.0%10.0%8.0%6.0%4.0%2.0%0.0%2005 2008Source: Calculated from data provided by DNEAP, M<strong>in</strong>istry of Plann<strong>in</strong>g and DevelopmentOther revenueFuelCustoms dutiesExciseVATIRPC (company)IRPS (personal)


C URRENT T AX S YSTEM 15Look<strong>in</strong>g at the structure of tax receipts, the value added tax (<strong>in</strong>clud<strong>in</strong>g imports) <strong>in</strong> 2008 raised 5.4percent of GDP, up from 4.5 percent <strong>in</strong> 2005. Income taxes amounted to 4.9 percent of GDP, splitalmost evenly between the IRPS (2.5 percent) and IRPC (2.4 percent). This is a hugeimprovement from 2005, when IRPC revenues amounted to just 1.0 percent of GDP, and IRPSrevenues raised 1.9 percent of GDP. In contrast, revenue from customs duties has decl<strong>in</strong>ed <strong>in</strong>relative importance to 1.5 percent of GDP <strong>in</strong> 2008, from 1.9 percent <strong>in</strong> 2005. In absolute terms,however, even customs revenue <strong>in</strong>creased by 29 percent over the period, <strong>in</strong>dicat<strong>in</strong>g that the ris<strong>in</strong>gvolume and value of imports, and a higher nom<strong>in</strong>al exchange rate have more than offset theeffects of SADC trade liberalization.Revenues from the excise tax (ICE) <strong>in</strong> 2008 totaled 1.1 percent of GDP, slightly higher <strong>in</strong> relativeterms than the 1.0 percent yield <strong>in</strong> 2005. Hence, ICE receipts grew slightly faster than nom<strong>in</strong>alGDP, though decl<strong>in</strong><strong>in</strong>g as a share of total revenue. Revenue from the fuel tax also decl<strong>in</strong>ed <strong>in</strong>relative importance from 1.2 percent of GDP <strong>in</strong> 2005 to 1.0 percent <strong>in</strong> 2008, reflect<strong>in</strong>g thetemporary suspension of fuel taxes last year. The stamp tax generated just 0.1 percent of GDP <strong>in</strong>revenues <strong>in</strong> 2008. F<strong>in</strong>ally, the government earned negligible revenue from royalties and surfacetaxes on m<strong>in</strong><strong>in</strong>g and petroleum <strong>in</strong> 2008, amount<strong>in</strong>g to just 0.2 percent of tax receipts–unchangedfrom 2005–and 0.0 percent of GDP.The rema<strong>in</strong>der of this section describes the ma<strong>in</strong> features of the current tax structure. 12Value Added <strong>Tax</strong>The standard value added tax rate of 17 percent applies to imports and domestic supplies of goodsand services from bus<strong>in</strong>esses with turnover greater than 2,500,00 MT per year (approximately$95,000 at the exchange rate <strong>in</strong> July 2009). 13 The 2007 revision of the VAT code reta<strong>in</strong>ed thesimplified regime (regime simplificada) consist<strong>in</strong>g of a 5 percent tax on gross sales, apply<strong>in</strong>g toenterprises with an annual turnover below 2,500,000 MT, exclud<strong>in</strong>g exporters and importers.Enterprises with a turnover below 750,000 MT per year (about $28,000) qualify for VATexemption (isenção). Firms registered under the simplified or exempt regimes cannot claim creditfor VAT paid on <strong>in</strong>puts; they also do not need to ma<strong>in</strong>ta<strong>in</strong> organized accounts conform<strong>in</strong>g toestablished standards. Another provision of the 2007 code (Article 15, section l) reduces the VATliability by 60% for enterprises contracted for construction of public roads, bridges and watersupplies.In l<strong>in</strong>e with <strong>in</strong>ternational norms, exports are subject to zero tax (isenção completa) and qualify fora full refund of VAT paid on <strong>in</strong>puts. Zero rat<strong>in</strong>g also applies to certa<strong>in</strong> basic foods, animal feed,and kerosene for light<strong>in</strong>g. In addition, the system cont<strong>in</strong>ues to exempt a long list of goods andservices from VAT, mean<strong>in</strong>g that they do not charge VAT on sales, and do not recover VAT paidon <strong>in</strong>puts. Exempt items <strong>in</strong>clude medical services and drugs, f<strong>in</strong>ancial services, basic <strong>in</strong>puts toagriculture, and products of agriculture, forestry, animal husbandry, and fish<strong>in</strong>g. Clearly, these12 This short review excludes the vehicle tax, municipal taxes, and other fees and levies.13 For a detailed explanation of the VAT Code, see Deloitte and ACIS, Legal Framework for <strong>Tax</strong>ation <strong>in</strong><strong>Mozambique</strong>: Value Added <strong>Tax</strong> Manual, Edition 1, October 2008, and separate manuals cover<strong>in</strong>g IRPC,IRPS, and VAT. The documents are available for downloads at: http://www.acisofala.com/ .


16 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEprovisions are designed to m<strong>in</strong>imize the burden of VAT on basic goods consumed heavily by thepoor.<strong>Tax</strong>payers <strong>in</strong> a VAT credit position—which arises when a firm owes less VAT on its taxablesales than it the tax paid on its <strong>in</strong>puts—can claim a refund (re-embolso) after twelve months if thecredit balance cont<strong>in</strong>ues to exceed 50,000 MT (approximately $1,900). Exporters can claim arefund without a twelve month wait with a credit balance above 5,000 MT, as can firms withzero-rated supplies constitut<strong>in</strong>g at least 75 percent of total receipts. All parties <strong>in</strong>terviewed forthis study remarked that the VAT refund process has improved s<strong>in</strong>ce 2005, but it is still a majorsource of grievance due to the extremely cumbersome procedures. This issue will be addressedfurther <strong>in</strong> Chapters 5 and 6.Income <strong>Tax</strong>The standard tax rate on corporate <strong>in</strong>come (Imposto sobre o Rendimento das Collectivas, orIRPC) rema<strong>in</strong>s at 32 percent, with a special rate of 10 percent for agriculture and animalhusbandry until December 2010. Withhold<strong>in</strong>g tax at a rate of 20 percent is due on a variety ofpayments such as rentals, services provided <strong>in</strong> Mozambican territory, and royalties for the use of<strong>in</strong>tellectual property. On payments to nonresidents, withhold<strong>in</strong>g tax is a f<strong>in</strong>al payment; <strong>in</strong> othercases it is a pre-payment aga<strong>in</strong>st tax due for the year. There is also a special tax rate of 35 percentapply<strong>in</strong>g to expenses that are not adequately documented, and therefore disallowed asdeductions. 14As with the VAT, the IRPC code <strong>in</strong>cludes a simplified regime apply<strong>in</strong>g to companies with aturnover below 2,500,000 MT (about $95,000), which are not required to ma<strong>in</strong>ta<strong>in</strong> organizedaccounts. Under the simplified regime, taxable <strong>in</strong>come is deemed to equal 20 percent of revenuesfrom the sale of produced goods or the supply of lodg<strong>in</strong>g, restaurant or beverage services, and 30percent on gross revenue for other economic activities. In cases where the coefficient of 0.2 isapplicable, the tax due would amount to 6.4 percent of gross receipts (=.20 x 32%). <strong>Tax</strong>authorities can also use “<strong>in</strong>direct methods” to assess <strong>in</strong>come <strong>in</strong> cases where the taxpayer cannotprovide appropriate or credible and timely accounts.The <strong>in</strong>come tax on <strong>in</strong>dividuals (Imposto sobre o Rendimento das Pessoas S<strong>in</strong>gulares, or IRPS)reta<strong>in</strong>s progressive marg<strong>in</strong>al tax rates of 10, 15, 20, 25, and 32 percent, assessed on totalhousehold <strong>in</strong>come after allowance for personal and family deductions. 15 No tax is due on <strong>in</strong>comesbelow 36 times the maximum m<strong>in</strong>imum wage <strong>in</strong> force at the end of the previous year. For 2008,the highest m<strong>in</strong>imum wage was 2139.50 MT per year, giv<strong>in</strong>g an IRPS threshold of 77,022 MT(about $2,900). In turn, the top marg<strong>in</strong>al tax rate of 32 percent applies to chargeable <strong>in</strong>comeexceed<strong>in</strong>g 1,512,000 MT (about $57,000). Compared to the zero-tax threshold <strong>in</strong> 2005 of 24million (old) MT, or just $1,200, the 2007 IRPS revision frees many more households with very14 For a detailed explanation of the IRPC Code, see Deloitte and ACIS, Legal Framework for <strong>Tax</strong>ation <strong>in</strong><strong>Mozambique</strong>: Manual on Corporate Income <strong>Tax</strong>, April 2009.15 As <strong>in</strong> 2005, the IRPS def<strong>in</strong>es five categories: <strong>in</strong>come from employment (“dependent” <strong>in</strong>come);professional and bus<strong>in</strong>ess <strong>in</strong>come; capital <strong>in</strong>come; <strong>in</strong>come from build<strong>in</strong>gs; and other <strong>in</strong>comes. Eachcategory covers multiple types of <strong>in</strong>come, with restrictions on loss offsets across categories.


C URRENT T AX S YSTEM 17low <strong>in</strong>comes from tax liability. In addition, the zero-rate bracket is now implicitly <strong>in</strong>dexed for<strong>in</strong>flation because it is def<strong>in</strong>ed relative to the national m<strong>in</strong>imum wage. By the same token, thisl<strong>in</strong>kage <strong>in</strong> the tax law could create a problem with revenue loss if the government were ever toapprove a hike <strong>in</strong> the m<strong>in</strong>imum wage far above the <strong>in</strong>flation rate. For employees of registeredbus<strong>in</strong>esses, the IRPS on wage <strong>in</strong>come is withheld at source and remitted monthly to the AT. 16Those earn<strong>in</strong>g employment <strong>in</strong>come of up to MT 100,000 are not obliged to file returns.Under both the IRPC and IRPS, taxpayers can claim refunds if the amount of tax withheld atsource or pre-paid as provisional estimated payments, which are required for corporations and for<strong>in</strong>dividuals receiv<strong>in</strong>g professional or bus<strong>in</strong>ess <strong>in</strong>come. The widely discussed problems with VATrefunds apply as well to the procedures for obta<strong>in</strong><strong>in</strong>g timely refunds of <strong>in</strong>come tax.Simplified <strong>Tax</strong> for Small <strong>Tax</strong>payersThe National Assembly approved the new Simplified <strong>Tax</strong> for Small <strong>Tax</strong>payers (ImpostoSimplificado para Pequenos Contribu<strong>in</strong>tes–ISPC) on 26 December 2008 and the law entered <strong>in</strong>toforce on 1 January 2009. The Council of M<strong>in</strong>isters published the first regulations <strong>in</strong> the Boletimde República on 14 April 2009. Dur<strong>in</strong>g the first half of 2009 the Revenue Authority and otheragencies, <strong>in</strong>clud<strong>in</strong>g private sector organizations, engaged <strong>in</strong> a broad campaign of meet<strong>in</strong>gs andmedia coverage to <strong>in</strong>form the public about the ISPC and their obligations to contribute to nationaldevelopment. The underly<strong>in</strong>g aim is to use this new tax tool to make formalization moreattractive (or at least less daunt<strong>in</strong>g) for small enterprises, and thereby attract large numbers ofnew taxpayers <strong>in</strong>to the system.In June, the Revenue Authority began register<strong>in</strong>g taxpayers under the ISPC, even though formsand procedures were still be<strong>in</strong>g developed. Accord<strong>in</strong>g to <strong>in</strong>formation provided by the AT, 1,800new contributors registered under the ISPC <strong>in</strong> the first month, while 300 taxpayers elected toswitch their status to coverage under the ISPC <strong>in</strong>stead of the simplified regimes for VAT and<strong>in</strong>come tax.The ma<strong>in</strong> features of the ISPC are as follows:• Enterprises with an annual turnover up to 2,500,000 MT (just under $100,000) may electto register under the ISPC. This is the same threshold as provided <strong>in</strong> the 2007 codes forVAT and IRPC to qualify for the respective simplified regimes.• For <strong>in</strong>come covered by the ISPC, taxpayers have an exclusion from VAT, IRPC, andIRPS. Other <strong>in</strong>come is subject to IRPS.• Enterprises with a turnover of up to 36 times the highest m<strong>in</strong>imum wage <strong>in</strong> force as of theprevious year-end are exempt from tax. This provision of the ISPC matches the thresholdfor enter<strong>in</strong>g the IRPS, except that the former relates to turnover while the latter relates to<strong>in</strong>come. Hence, the ISPC will be charg<strong>in</strong>g tax to <strong>in</strong>dividuals operat<strong>in</strong>g micro and smallenterprises who would fall below the tax threshold under the <strong>in</strong>dividual <strong>in</strong>come tax.16 For a detailed explanation of the IRPS Code, see Deloitte and ACIS, Legal Framework for <strong>Tax</strong>ation <strong>in</strong><strong>Mozambique</strong>: Manual on Individual Income <strong>Tax</strong>, January 2009.


18 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUE• The applicable tax rate is 75,000 MT per year (about $2800), or 3 percent of salesvolume, at the option of the taxpayer. 17• For new bus<strong>in</strong>esses, the tax is reduced by half for the first year of operations, as a lure toregister <strong>in</strong> the system.• The tax is paid <strong>in</strong> four quarterly <strong>in</strong>stallments.In essence, the ISPC is designed as a substitute for the simplified schemes under the VAT and<strong>in</strong>come tax laws. Recall that the simplified VAT imposed a tax equal<strong>in</strong>g 5 percent of turnover; <strong>in</strong>addition, <strong>in</strong>come above the zero-bracket threshold is subject to a m<strong>in</strong>imum 10 percent tax rateunder the standard IRPS. Under the IRPC code, the <strong>in</strong>come tax would amount to 2 percent ofturnover for enterprises subject to the coefficient of 0.2 for presumed <strong>in</strong>come. Hence, the ISPCrepresents not only a consolidation of what had been two separate simplified tax regimes, but alsoa reduction <strong>in</strong> the effective tax rate from approximately 7 percent of turnover to just 3 percent. Inaddition, the AT is develop<strong>in</strong>g simpler procedures for registration and simpler formats for officialreceipt books under the ISPC. They further plan to make it easier for small enterprises outsidemajor cities by open<strong>in</strong>g additional offices <strong>in</strong> new localities and launch<strong>in</strong>g mobile tax units. Thereis also discussion of possibly engag<strong>in</strong>g local authorities as tax collection agents <strong>in</strong> more remotelocations, with appropriate <strong>in</strong>centives to make it worth their participation <strong>in</strong> the system.Accord<strong>in</strong>g to sources with<strong>in</strong> the AT, the M<strong>in</strong>istry of F<strong>in</strong>ance, and the M<strong>in</strong>istry of Plann<strong>in</strong>g andDevelopment, the budget program does not anticipate a significant revenue contribution from theISPC <strong>in</strong> the near term. The expectation is that the simplified tax, by draw<strong>in</strong>g large numbers ofnew contributors <strong>in</strong>to the formal tax system, will capture a grow<strong>in</strong>g stream of revenues over timefrom the natural growth of enterprises that enter the system due to the ISPC.Import DutiesAs of 2009, the standard duty rates range from 0 to 20 percent accord<strong>in</strong>g to the import category:• 20 percent for consumer goods (Class C)• 7.5 percent for <strong>in</strong>termediate materials (Class I)• 5 percent for capital goods (Class K) and fuel (Class N)• 2.5 percent for raw materials (Class M)• 0 percent for basic goods (Class E)Under the SADC trade protocol most imports from member states enter <strong>Mozambique</strong> duty-freewhen accompanied by the required (and complicated) documents for compliance with SADCrules of orig<strong>in</strong>. Customs duties will fall to zero by 2012 on all comply<strong>in</strong>g imports from memberstates other than South Africa, and for imports from South Africa by 2015. The SADC InterimEconomic Partnership Agreement (EPA) with the European Union, which <strong>Mozambique</strong> signed <strong>in</strong>June of 2009, will also reduce the tax base for collect<strong>in</strong>g customs duties as duties are graduallyremoved on imports from the EU. In addition, extensive duty exemptions are granted to qualified<strong>in</strong>vestors under the Code of Fiscal Benefits (Código de Benefícios Fiscais), which is discussedbelow.17 The latter condition is clarified <strong>in</strong> the regulations.


C URRENT T AX S YSTEM 19Excise <strong>Tax</strong> and Fuel LevyThe National Assembly recently approved a revision to the tax code for the Specific Consumption<strong>Tax</strong> (Imposto sobre Consumos Éspecificos, or ICE). As of this writ<strong>in</strong>g, the new ICE code was notyet gazetted, but it should come <strong>in</strong>to force at the beg<strong>in</strong>n<strong>in</strong>g of 2010. The ICE cont<strong>in</strong>ues to applyad valorem tax rates to an extensive list of domestic products and imports, ma<strong>in</strong>ly alcoholicbeverages, tobacco products, vehicles, and jewelry, and other luxury goods. In 2005 the ICE ratesranged from 5 to 65 percent. The 2009 ICE code <strong>in</strong>creases the top rate to 75 percent, for tobaccoproducts. It also <strong>in</strong>creases the tax rate on jewelry and several vehicle categories, while reduc<strong>in</strong>gthe ICE rate on some alcoholic beverages, sport<strong>in</strong>g goods and art works, among other th<strong>in</strong>gs.Another feature of the new ICE law is a m<strong>in</strong>imum specific tax (per quantity) on most alcoholicbeverages and tobacco products, as a safeguard aga<strong>in</strong>st undervaluation.The Fuel Levy (<strong>Tax</strong>a de Combustíveis) rema<strong>in</strong>s unchanged as a tax per liter of various types offuel, subject to a formula for quarterly adjustments reflect<strong>in</strong>g changes <strong>in</strong> the cost of petroleumproducts and the exchange rate. Because a large share of the levy is earmarked for use by theRoad Fund, special provisions are <strong>in</strong> place to compensate non-road producers who use largeamounts of fuel, particularly mechanized farmers, m<strong>in</strong>es, and fish<strong>in</strong>g boats. In mid-2008, thegovernment suspended the fuel levy on kerosene until year-end, as well import duties and VATon petroleum products, <strong>in</strong> order to cushion the impact of ris<strong>in</strong>g world fuel prices, which hadsparked civil unrest.Stamp <strong>Tax</strong>The stamp tax has not changed s<strong>in</strong>ce 2005 (see Chapter 2.) The study team understands that theAT is exam<strong>in</strong><strong>in</strong>g the implications of elim<strong>in</strong>at<strong>in</strong>g or narrow<strong>in</strong>g the scope of the stamp tax. Manytax experts view this as an archaic “nuisance tax” with a low revenue yield relative to the cost ofadm<strong>in</strong>istration and compliance. Elim<strong>in</strong>ation of the stamp tax is often recommended <strong>in</strong> the <strong>in</strong>terestof simplification. While the elim<strong>in</strong>ation of any tax reduces revenue, <strong>in</strong> this case the revenue effectis very small and could easily be offset by other measures as part of the overall tax reformprogram.Other Elements of the National <strong>Tax</strong> <strong>System</strong>In addition to the components discussed above, the national tax system also <strong>in</strong>cludes:• Special tax on gambl<strong>in</strong>g <strong>in</strong>come• Inheritance and gift tax• Property transfer tax• Royalties on m<strong>in</strong>eral and petroleum extraction and surface tax on resource concessions• Vehicle tax• National reconstruction taxMunicipal <strong>Tax</strong>esConsistent with the government’s plans for decentralization and devolution, one of the strategic<strong>in</strong>dicators <strong>in</strong> <strong>PARPA</strong> <strong>II</strong> is to <strong>in</strong>crease the percentage of municipal and local government budgetsf<strong>in</strong>anced by their own revenue sources. To support this objective, the Municipal F<strong>in</strong>ance Act of2008 establishes clear guidel<strong>in</strong>es for the types of revenue fall<strong>in</strong>g with<strong>in</strong> the jurisdiction of


20 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEmunicipalities, and limits on the tax rates. Without go<strong>in</strong>g <strong>in</strong>to detail here, 18 municipalities havethe authority to impose the follow<strong>in</strong>g taxes and charges:• Municipal <strong>in</strong>dividual tax• Municipal property tax• Municipal property transfer tax• Municipal tax on vehicles• Contributions for <strong>in</strong>frastructure improvements• Levies for issu<strong>in</strong>g operat<strong>in</strong>g licenses• Tariffs and charges for the provision of municipal services<strong>Tax</strong> Incentives for InvestmentUnder the Investment Act of 1993, <strong>in</strong>vestments approved by the Center for Investment Promotion(CPI) qualify for specified fiscal benefits. Eligibility requires a m<strong>in</strong>imum <strong>in</strong>vestment of $50,000for foreign companies, and $5000 for domestic companies, as well as detailed bus<strong>in</strong>ess plans andevidence of f<strong>in</strong>ancial and management capability. CPI approval is subject to a fee of 0.1% of thetotal <strong>in</strong>vestment, up to a maximum of $50,000 for large projects.In January 2009, the government <strong>in</strong>troduced a new Code of Fiscal Benefits (Código de BenefíciosFiscais, or CBF), to replace the 2002 Code. The stated aim of the 2009 Code is to “rationalize thefiscal benefits for <strong>in</strong>vestments and make them more effective as an <strong>in</strong>strument of the politicaleconomy.” 19 Appendix B provides a detailed po<strong>in</strong>t-by-po<strong>in</strong>t comparison of the 2009 and 2002Codes. The new CBF provides “specific” benefits for ten categories of <strong>in</strong>vestments: basic public<strong>in</strong>frastructure projects; rural commerce and <strong>in</strong>dustry; manufactur<strong>in</strong>g and assembly <strong>in</strong>dustries;agriculture and fisheries; hotels and tourism; science and technology parks; large-scale projects;rapid development zones; <strong>in</strong>dustrial free zones, and special economic zones. The specific benefitsgenerally <strong>in</strong>volve exemption from customs duty and VAT, partial tax holidays <strong>in</strong> the form of<strong>in</strong>come tax reductions for def<strong>in</strong>ed periods of time, and additional provisions such as accelerateddepreciation and deductions for professional tra<strong>in</strong><strong>in</strong>g.For the specific benefits, major changes <strong>in</strong> 2009 are as follows:• Elim<strong>in</strong>ation of “exceptional <strong>in</strong>centives” for large-scale projects, which previously werenegotiated with the Council of M<strong>in</strong>isters. Large projects now bear the standard companytax. The former regime usually applied a 1 percent tax on gross receipts <strong>in</strong> lieu of the<strong>in</strong>come tax, <strong>in</strong> l<strong>in</strong>e with the precedent set for Mozal <strong>in</strong> the 1990s. Large projects stillbenefit from an exemption from customs duty—extended now to <strong>in</strong>clude VAT—onimported capital goods, plus accompany<strong>in</strong>g spare parts and accessories.• Elim<strong>in</strong>ation of a 25 percent tax reduction for eight years for <strong>in</strong>vestments <strong>in</strong> m<strong>in</strong><strong>in</strong>g.• Deep reductions <strong>in</strong> <strong>in</strong>come tax rates for approved <strong>in</strong>vestments <strong>in</strong> basic public<strong>in</strong>frastructure projects, phased over 15 years; for agriculture and fisheries, phased through18 Deloitte and ACIS, General Overview of the <strong>Tax</strong> <strong>System</strong> <strong>in</strong> <strong>Mozambique</strong>, Edition 1, October 2008, pp.30-35, provides a summary of the revenue provisions of Law 1/2008 of 16 January.19 CPI, Code of Fiscal Benefits, Law 4.2009 of 12 January (English version), p. 3.


C URRENT T AX S YSTEM 212025; for science and technology parks, phased over 15 years; and for enterprises <strong>in</strong>Industrial Free Zones (ZFIs) and Special Economic Zones (ZEEs), phased over 15 years.ZFI and ZEE operators (as dist<strong>in</strong>ct from enterprises with<strong>in</strong> a zone) are exempt from<strong>in</strong>come tax for 10 and 5 years, respectively, followed by 5 years with a 50 percentreduction and then a 25 percent tax reduction without time limit. Investments <strong>in</strong>agriculture and IFZs already enjoyed tax rate reductions <strong>in</strong> the 2002 Code, but the timeframes are now longer and the IFZ reduction more generous.• Extension of the customs duty exemption to a broader range of goods <strong>in</strong> addition tocapital goods <strong>in</strong> category “K.” The specification varies by type of <strong>in</strong>vestment, and<strong>in</strong>cludes, for example: accompany<strong>in</strong>g spare parts, accessories and raw materials formanufactur<strong>in</strong>g and assembly <strong>in</strong>dustries; goods required for construction and “outfitt<strong>in</strong>g”of tourism and hotel <strong>in</strong>vestments; and “essential goods” for <strong>in</strong>vestments <strong>in</strong> ruralcommerce.• Extension of the exemption from customs duty to <strong>in</strong>clude exemption from VAT forspecified purchases by most <strong>in</strong>vestments that qualify for specific <strong>in</strong>centives (theexception be<strong>in</strong>g manufactur<strong>in</strong>g and assembly <strong>in</strong>dustries).The 2009 Code also offers “general” fiscal benefits for <strong>in</strong>vestments that do not fall <strong>in</strong>to a categorycovered by “specific” <strong>in</strong>centives. These benefits cover import duty exemptions (but not VATexemptions) on class “K” goods; <strong>in</strong>vestment tax credits; accelerated depreciation; deductions formodernization and new technology; deductions for professional tra<strong>in</strong><strong>in</strong>g; and deductions forcerta<strong>in</strong> expenses on <strong>in</strong>frastructure that will benefit the public. 20 The 2009 Code <strong>in</strong>troducedseveral changes to the general benefits provided <strong>in</strong> the previous Code, <strong>in</strong>clud<strong>in</strong>g:• Scal<strong>in</strong>g back accelerated depreciation from 200 percent to 150 percent of the normal rate.• Scal<strong>in</strong>g back tax deductions for <strong>in</strong>vestments <strong>in</strong> specialized equipment <strong>in</strong>volv<strong>in</strong>g newtechnology, as well as expenditures on public <strong>in</strong>frastructure.• No exemption from the stamp tax and real property transfer tax.One bizarre “benefit” is the deduction for expenses on professional tra<strong>in</strong><strong>in</strong>g of Mozambicanlocals, which is capped at 5 percent of taxable <strong>in</strong>come (or 10 percent for tra<strong>in</strong><strong>in</strong>g on advancedtechnology equipment). With the cap def<strong>in</strong>ed <strong>in</strong> this way, the benefit is worthless for an<strong>in</strong>vestment <strong>in</strong> a tax-loss condition, which is common for start-ups dur<strong>in</strong>g early years of operation.More to the po<strong>in</strong>t, it makes no sense to limit the deduction for a legitimate bus<strong>in</strong>ess expense,especially one that creates positive externalities for the economy by upgrad<strong>in</strong>g labor force skills.Indeed, some countries <strong>in</strong> the region such as Botswana and Swaziland allow a super-deduction(more than 100 percent) for tra<strong>in</strong><strong>in</strong>g expenses, to provide an effective subsidy for these desirableactivities.Both the 2002 and 2009 Codes state that fiscal benefits are to be considered as a form ofgovernment expenditure, requir<strong>in</strong>g a declaration of benefits used each year. Table 3-1summarizes government estimates of the “tax expenditures” for 2006 and 2007, with partial20 The 2009 Code specifies that general benefits may not be cumulated with specific benefits, except asstated <strong>in</strong> the designation of the latter.


22 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEestimates for 2008. 21 It is <strong>in</strong>terest<strong>in</strong>g to see that value of fiscal benefits <strong>in</strong> 2007 under the IRPCjumped to over 90 percent of actual company tax receipts. The Tribunal Adm<strong>in</strong>istrativo (2008)po<strong>in</strong>ts out that this large <strong>in</strong>crease reflects high profits at Mozal, CFM, M-Cel and BancoInternacional de Moçambique. 22 These figures underscore the importance of the 2009 revision tothe CBF, which closes off the special <strong>in</strong>come tax benefit for large projects <strong>in</strong> the future, though itleaves the benefits unchanged for exist<strong>in</strong>g large projects.Table 3-1Approved <strong>Tax</strong> Exemptions, by Major <strong>Tax</strong> (10^6 MT)Import Duty VAT Excise IRPC IRPS Sum<strong>Tax</strong> exemptions2006 822.8 1534.3 271.8 517.4 0.2 3146.52007 923 1857.1 112.8 3967.2 0.9 68612008 710 1291.6 188.1 na na na<strong>Tax</strong> receipts2006 3260.6 9385 1818.6 2535.5 3784.4 20784.12007 3803.8 11314 2076.2 4364.9 4859.1 264182008 3597.4 12969.6 2634.3 5425.9 5957.2 30584.4Exemptions as % receipts2006 25.2% 16.3% 14.9% 20.4% 0.0% 15.1%2007 24.3% 16.4% 5.4% 90.9% 0.0% 26.0%2008 19.7% 10.0% 7.1% na na naExemptions, by type of tax (%)2006 26.1% 48.8% 8.6% 16.4% 0.0% 100.0%2007 13.5% 27.1% 1.6% 57.8% 0.0% 100.0%2008 na na na na na naNotes: IRPC = Company Income <strong>Tax</strong>IRPS = Individual Income <strong>Tax</strong>Sources: 2006 and 2007 exemptions from Tribunal Adm<strong>in</strong>istrativo, Relatório e Parecer sobre a Conta Geral do Estado de 2007, p. V.2008 exemptions from Autoridade Tributária, Relatório de Actividades 2008, p. 50.Receipts, by type of tax from Autoridade Tributária, Execução da Receita do Estado 2008Appendix C provides an <strong>in</strong>ternational perspective on tax rates and tax <strong>in</strong>centive regimes <strong>in</strong> theSADC member states and four other comparator countries <strong>in</strong> Africa (Ghana, Kenya, Senegal, andUganda). It difficult to judge from this <strong>in</strong>formation the extent to which <strong>Mozambique</strong>’s tax systemis regionally competitive because the <strong>in</strong>centive regimes conta<strong>in</strong> many technical details that defysimple comparison, and <strong>in</strong> any case the effects depend on characteristics of particular<strong>in</strong>vestments. The most useful comparisons come from several studies that are discussed <strong>in</strong>Chapter 5 (see Investment), show<strong>in</strong>g that the 2002 Code of Fiscal Benefits <strong>in</strong> <strong>Mozambique</strong> washighly competitive with<strong>in</strong> the region. No comparable studies have been conducted, however, onthe 2009 Code.It is also important to emphasize that the question of tax competition only arises <strong>in</strong> the context of<strong>in</strong>ternationally mobile <strong>in</strong>vestment projects. For <strong>in</strong>vestments that are motivated by access toparticular natural resources, opportunities <strong>in</strong> the domestic market, or proximity to ports and21 The IMF (2008a) notes that government documents fail to expla<strong>in</strong> the methodology beh<strong>in</strong>d theseestimates.22 Us<strong>in</strong>g conservative assumptions, Versano and others (2006) estimate that the <strong>in</strong>come tax benefit forMozal, alone, <strong>in</strong> 2004 was worth approximately $37.5 million, compared to total company <strong>in</strong>come taxreceipts of approximately $45.6 million that year.


C URRENT T AX S YSTEM 23corridors – and are therefore anchored geographically – tax rates <strong>in</strong> other countries do not enter<strong>in</strong>to consideration. 23DOING BUSINESS RANKINGS OF MOZAMBIQUE’S TAX SYSTEMThis section reviews <strong>Mozambique</strong>’s performance on tax system rank<strong>in</strong>gs from the World Bank’sannual Do<strong>in</strong>g Bus<strong>in</strong>ess reports. 24 Although there are technical problems with the <strong>in</strong>dicators (seeExhibit 3-1), the rat<strong>in</strong>gs have proven to be useful <strong>in</strong> highlight<strong>in</strong>g constra<strong>in</strong>ts to the bus<strong>in</strong>essenvironment, provid<strong>in</strong>g both an impetus and political leverage for reforms to improve marketconditions, <strong>in</strong>clud<strong>in</strong>g reforms to establish a more efficient, transparent and effective tax system.The Do<strong>in</strong>g Bus<strong>in</strong>ess reports are designed to provide a set of “standard tools used across a broadrange of jurisdictions to measure the impact of government rule-mak<strong>in</strong>g on bus<strong>in</strong>ess activity…,”particularly with reference to the effect of <strong>in</strong>stitutional conditions on opportunities for domesticsmall and medium-size enterprises. 25 In Do<strong>in</strong>g Bus<strong>in</strong>ess 2009, <strong>Mozambique</strong> ranked 141 out of182 countries for the overall ease of do<strong>in</strong>g bus<strong>in</strong>ess. For the category “pay<strong>in</strong>g taxes,” however,<strong>Mozambique</strong> fares much better, with rank<strong>in</strong>g 88. This is ahead of all but two other SADCcountries with comparable <strong>in</strong>come levels, though Malawi (at 58) and Zambia (at 38) show thatbetter rank<strong>in</strong>gs can be achieved even for low <strong>in</strong>come countries. The overall SADC median rank is58, with the best performers be<strong>in</strong>g Mauritius and South Africa, at 11 and 23, respectively.The Do<strong>in</strong>g Bus<strong>in</strong>ess score for pay<strong>in</strong>g taxes is calculated for a specific case <strong>in</strong>volv<strong>in</strong>g a mediumsizedlimited liability company that started operations one year earlier, undertak<strong>in</strong>g generalcommercial or <strong>in</strong>dustrial activities. 26 The score consists of three components, which have equalweight: the number of tax payments; the hours spent <strong>in</strong> preparation and payment of taxes; and acomputation of the total tax rate as a percentage of profits for the representative bus<strong>in</strong>ess. Thetotal tax rate itself has three components: the tax on profits; taxes and contributions on labor thatare borne by the employer; and other taxes.In Do<strong>in</strong>g Bus<strong>in</strong>ess 2009, the World Bank team estimates that the standardized case-studybus<strong>in</strong>ess would require 230 hours of work to prepare and submit taxes <strong>in</strong> <strong>Mozambique</strong> (rank 89)<strong>in</strong>volv<strong>in</strong>g 37 payments (rank 120), and face a total tax equal<strong>in</strong>g 34.3 percent of the assumed profitmarg<strong>in</strong> (rank 52). Over the past three years these tax scores have not improved for <strong>Mozambique</strong>,while some other countries have leapfrogged ahead by implement<strong>in</strong>g reforms affect<strong>in</strong>g one ormore of the <strong>in</strong>dicators. Consequently, <strong>Mozambique</strong>’s rank on pay<strong>in</strong>g taxes dropped from 77 to 8823 Bolnick (2009) reports the results of a survey of 60 <strong>in</strong>vestments <strong>in</strong> <strong>Mozambique</strong> that were approved byCPI <strong>in</strong> 2005, 2006 and 2007. The most important and frequent motive for <strong>in</strong>vestment was opportunity <strong>in</strong>the domestic market; only 7 of 60 <strong>in</strong>vestors even considered alternative locations, and only 10 of 60 ratedtax <strong>in</strong>centives as be<strong>in</strong>g critical to their <strong>in</strong>vestment decision.24 Appendix C provides a summary of tax <strong>in</strong>dicators from the World Economic Forum’s AfricanCompetitiveness Report for 2009, and executive survey results from the WEF’s Global CompetitivenessReport for 2007-2008.25 http://www.do<strong>in</strong>gbus<strong>in</strong>ess.org/documents/DB09_About.pdf. Accessed June 24, 2009.26 For the full set of assumptions see:http://www.do<strong>in</strong>gbus<strong>in</strong>ess.org/MethodologySurveys/Pay<strong>in</strong>g<strong>Tax</strong>es.aspx.


24 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEExhibit 3-1The Pay<strong>in</strong>g <strong>Tax</strong>es Indicator: How does it measure up?The World Bank’s Independent Evaluation Group (IEG) criticized this the <strong>in</strong>dicator on two grounds <strong>in</strong> its recentreport, Do<strong>in</strong>g Bus<strong>in</strong>ess: An Independent Evaluation. First, whereas most Do<strong>in</strong>g Bus<strong>in</strong>ess <strong>in</strong>dicators gauge theburden of bus<strong>in</strong>ess regulation, the pay<strong>in</strong>g taxes <strong>in</strong>dicator <strong>in</strong>cludes subjective judgments about fiscal efficiency andequity by <strong>in</strong>clud<strong>in</strong>g an estimate of the total tax rate (TTR) as a share of profits for an illustrative bus<strong>in</strong>ess case. This<strong>in</strong>dicator implicitly equates low tax rates with a more efficient and equitable tax system. In reality the situation ismore complex. For <strong>in</strong>stance, <strong>in</strong> a country where revenue mobilization is a paramount concern, lower taxes are notnecessarily advisable. Moreover, <strong>in</strong>clud<strong>in</strong>g the TTR, as calculated, with<strong>in</strong> the score for pay<strong>in</strong>g taxes leads to thecurious result that <strong>Mozambique</strong> ranks ahead of advanced countries like F<strong>in</strong>land, Austria and Japan <strong>in</strong> this category.A second criticism raised by the IEG report is that Do<strong>in</strong>g Bus<strong>in</strong>ess relies on a s<strong>in</strong>gle source and an opaquemethodology for calculat<strong>in</strong>g TTR. The analysis was developed by PricewaterhouseCoopers (PwC) and PwC is thesole <strong>in</strong>formant for the pay<strong>in</strong>g tax scores <strong>in</strong> 142 countries. Collaborat<strong>in</strong>g with a s<strong>in</strong>gle source is <strong>in</strong>consistent with theoverall approach used by Do<strong>in</strong>g Bus<strong>in</strong>ess, and <strong>in</strong>creases the likelihood of anomalous results, especially for acomplex calculation like TTR.As an example, Do<strong>in</strong>g Bus<strong>in</strong>ess regularly accords Zambia a very strong rank<strong>in</strong>g on pay<strong>in</strong>g taxes, largely because ofthe TTR estimate. Do<strong>in</strong>g Bus<strong>in</strong>ess 2009 shows a TTR for Zambia of 16.1 percent, 8 th best <strong>in</strong> the world. Look<strong>in</strong>g atthe TTR elements, the estimated company tax equals 1.7 percent of profits for the standard bus<strong>in</strong>ess case(compared to 27.7 percent for <strong>Mozambique</strong>). Yet Zambia imposes a standard company tax rate of 35 percent,versus 32 percent <strong>in</strong> <strong>Mozambique</strong>. The extremely low TTR for Zambia appears to be an artifact of the particularcost structure assumed for the standard bus<strong>in</strong>ess case. (Note that the methodology is based on the standard taxsystem without reference to any special fiscal benefits.)Another problem with the TTR calculation is that it does not take <strong>in</strong>to account the tax on distributed dividends,which directly affects the net rate of return to shareholders. a Countries that have elim<strong>in</strong>ated double taxation ofdividends therefore get no advantage <strong>in</strong> the TTR score. Furthermore, the methodology adopts a restrictiveassumption about how profit marg<strong>in</strong>s are set for the standardized bus<strong>in</strong>ess case, lead<strong>in</strong>g to the implausible resultthat the bus<strong>in</strong>esses bear a tax exceed<strong>in</strong>g 200 percent of their profit <strong>in</strong> several countries, <strong>in</strong>clud<strong>in</strong>g Burundi andDRC.Estimat<strong>in</strong>g the number of tax payments would appear to be straightforward, but there is an arbitrary element to thecount for electronic payments. Suppose a country moves from a paper system of monthly VAT returns to a systemallow<strong>in</strong>g electronic payments. By assumption, Do<strong>in</strong>g Bus<strong>in</strong>ess would reduce the tax payment count from 12 to 1,produc<strong>in</strong>g a big jump <strong>in</strong> the rank<strong>in</strong>g for pay<strong>in</strong>g taxes. This jump is difficult to justify. If the case-study bus<strong>in</strong>ess hascomputerized accounts, then pr<strong>in</strong>t<strong>in</strong>g out and sign<strong>in</strong>g monthly forms is not that much more onerous than send<strong>in</strong>g anelectronic version. And if the case-study bus<strong>in</strong>ess lacks computerized accounts, it cannot take advantage of theoption for electronic fil<strong>in</strong>g.In short, Do<strong>in</strong>g Bus<strong>in</strong>ess scores for pay<strong>in</strong>g taxes should be viewed with caution, though they may be useful <strong>in</strong>highlight<strong>in</strong>g problems with the tax system that require attention.a World Bank, IFC and PwC (2008), Pay<strong>in</strong>g <strong>Tax</strong>es 2009: The Global Picture, Appendix 2 (Methodology).


C URRENT T AX S YSTEM 25between 2008 and 2009. 27 Look<strong>in</strong>g at the components, <strong>Mozambique</strong>’s rank slipped from 112 to120 on the number of payments; from 83 to 89 on the time required; and from 43 to 52 on thetotal tax rate.These scores suggest that complexities affect<strong>in</strong>g tax compliance <strong>in</strong> <strong>Mozambique</strong> are a moreserious barrier to do<strong>in</strong>g bus<strong>in</strong>ess than the tax rates–an observation corroborated <strong>in</strong> all of the<strong>in</strong>terviews conducted for the present study. To ascend <strong>in</strong> the rank<strong>in</strong>gs and significantly improveits score for the ease of pay<strong>in</strong>g taxes–and the overall ease of do<strong>in</strong>g bus<strong>in</strong>ess rank<strong>in</strong>g as well – thefocus of attention should be on reduc<strong>in</strong>g the time and number of payments <strong>in</strong>volved <strong>in</strong> pay<strong>in</strong>gtaxes. The most important action to jump <strong>in</strong> the rank<strong>in</strong>gs would be to <strong>in</strong>troduce electronicdeclarations and payments for the VAT, the corporate <strong>in</strong>come tax, and social <strong>in</strong>surancecontributions for workers. Given the methodology used by the World Bank, automat<strong>in</strong>g these taxpayments would reduce the number of payments for these three taxes from 31 to 3. In Tocalculate the impact of this measure on the Do<strong>in</strong>g Bus<strong>in</strong>ess score, we assume that automation ofthe payments would also halve the estimated time to pay taxes. If other countries ma<strong>in</strong>ta<strong>in</strong> theirpresent policies, these reforms would improve <strong>Mozambique</strong>’s rank on pay<strong>in</strong>g taxes from 88 to 14,and its overall ease of do<strong>in</strong>g bus<strong>in</strong>ess rank from 141 to 130. Elim<strong>in</strong>at<strong>in</strong>g the stamp duty wouldfurther reduce the number of payments by one, and improve the rank<strong>in</strong>g by one more place.The same measures would boost <strong>Mozambique</strong>’s rank from 10 to 3 <strong>in</strong> the SADC region <strong>in</strong> thecategory of pay<strong>in</strong>g taxes, trail<strong>in</strong>g only Mauritius and South Africa (see Table 3-2). In addition,these measures would improve <strong>Mozambique</strong>’s SADC rank<strong>in</strong>g on the overall ease of do<strong>in</strong>gbus<strong>in</strong>ess from 10 to 8. Of course, if other SADC countries are simultaneously pursu<strong>in</strong>g taxreforms there will be less improvement <strong>in</strong> <strong>Mozambique</strong>’s relative position. Nonetheless, thesemeasures would still significantly improve the ease of pay<strong>in</strong>g taxes <strong>in</strong> absolute terms–which is theprimary objective.TAX ADMINISTRATIONChapter 2 outl<strong>in</strong>ed an impressive number of reforms to tax adm<strong>in</strong>istration dur<strong>in</strong>g the <strong>PARPA</strong> <strong>II</strong>period. All of the available <strong>in</strong>formation—from documentary evidence, <strong>in</strong>terview results, and dataanalysis—show that the reforms genu<strong>in</strong>ely improved the quality of tax adm<strong>in</strong>istration andsignificantly <strong>in</strong>creased the revenue yield relative to GDP.Nonetheless, the glass is still less than half full when it comes to modernization of taxadm<strong>in</strong>istration <strong>in</strong> <strong>Mozambique</strong>. This section exam<strong>in</strong>es the current conditions, focus<strong>in</strong>g onshortcom<strong>in</strong>gs relative to best practices for develop<strong>in</strong>g countries. The f<strong>in</strong>al section of the chaptersummarizes the AT’s plans for deal<strong>in</strong>g with many of these challenges. It must be emphasized that<strong>in</strong>stitut<strong>in</strong>g major reforms to tax adm<strong>in</strong>istration is a complex and difficult task that requiresdetailed technical plann<strong>in</strong>g, careful sequenc<strong>in</strong>g, diligent attention to change management,adequate f<strong>in</strong>ancial, technical and personnel resources, and coord<strong>in</strong>ation of adm<strong>in</strong>istrative reforms27 In 2007, <strong>Mozambique</strong> ranked 80 on the ease of pay<strong>in</strong>g taxes. The jump to 72 <strong>in</strong> 2008 occurred despitethe absence of improvement <strong>in</strong> any of the country’s component scores. The 2007 rank<strong>in</strong>gs, however, arenot directly comparable to those for 2008 and 2009. In fact, the time series data set on the World Bank’sDo<strong>in</strong>g Bus<strong>in</strong>ess website shows scores for 2007 and 2008 only <strong>in</strong> the category of pay<strong>in</strong>g taxes.


26 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUETable 3-2Do<strong>in</strong>g Bus<strong>in</strong>ess 2009, Pay<strong>in</strong>g <strong>Tax</strong>es Indicator ComponentsPay<strong>in</strong>g <strong>Tax</strong>esEconomyRankPayments(number)Time(hours)Profit tax(%)Total tax rate(% profit)Angola 130 31 272 24.6 53.2Botswana 17 19 140 17 17.1Congo, Dem. Rep. 153 32 308 0 229.8Lesotho 54 21 324 14.5 18Madagascar 92 25 238 19.6 42.8Malawi 58 19 292 29.6 31.4Mauritius 11 7 161 11.3 22.2<strong>Mozambique</strong> 88 37 230 27.7 34.3Namibia 96 37 375 16.7 25.3Seychelles 40 16 76 23.6 46.6South Africa 23 9 200 24.5 34.2Swaziland 52 33 104 28.1 36.6Tanzania 109 48 172 19.8 45.1Zambia 38 37 132 1.7 16.1Zimbabwe 157 52 256 0 63.7SADC Median 58 31 230 19.6 34.3SOURCE: Do<strong>in</strong>g Bus<strong>in</strong>ess, 2009with support<strong>in</strong>g legal and regulatory changes, all driven by strong champions of reform <strong>in</strong>leadership positions. Otherwise, deep adm<strong>in</strong>istrative reforms can be highly disruptive to the AT’sbus<strong>in</strong>ess processes, to the po<strong>in</strong>t of jeopardiz<strong>in</strong>g the revenue yield and prospects for success of thereforms themselves.OrganizationThe establishment of the Revenue Authority was a giant step towards professionaliz<strong>in</strong>g taxservice. Compared to the previous organizational arrangements under the civil service, the AT isendowed with a high degree of managerial autonomy, <strong>in</strong>clud<strong>in</strong>g flexibility <strong>in</strong> determ<strong>in</strong><strong>in</strong>gremuneration and decisions on hir<strong>in</strong>g, promotion, and retention of staff. Creation of the AT alsoprovides the basis for achiev<strong>in</strong>g major ga<strong>in</strong>s <strong>in</strong> efficiency through the <strong>in</strong>tegration of commonoperations. To date, however, the <strong>in</strong>tegration of customs and domestic tax services has beenlimited, largely <strong>in</strong>volv<strong>in</strong>g back office functions. Greater efficiency can be achieved through the<strong>in</strong>tegration of revenue operations such as audit, debt management, risk management, and refundprocess<strong>in</strong>g, as well as support functions such as human resource management, taxpayer databasepool<strong>in</strong>g, IT systems development, and customer services.


C URRENT T AX S YSTEM 27Procedures and Information <strong>System</strong>sEvery department <strong>in</strong> the AT uses computer applications <strong>in</strong> the daily course of bus<strong>in</strong>ess. Yet manybasic <strong>in</strong>teractions with taxpayers still operate <strong>in</strong> a cumbersome pre-automation mode rely<strong>in</strong>g onpaper documents, error-prone manual <strong>in</strong>puts of data <strong>in</strong>to the IT systems, and physical movementof hard-copy files for process<strong>in</strong>g and approvals. This description applies to tax declarations andpayments, customs clearances, refund claims, and case management for appeals, among otherbus<strong>in</strong>ess processes. The <strong>in</strong>troduction of electronic systems will cut compliance costs for taxpayersand free AT personnel time from push<strong>in</strong>g paper to tasks that generate a higher revenue yield. Thegovernment and the AT are committed to implement<strong>in</strong>g a modern system of e-taxation, <strong>in</strong>clud<strong>in</strong>gelectronic declarations, electronic payments through commercial banks, and the <strong>in</strong>troduction ofan automated s<strong>in</strong>gle w<strong>in</strong>dow (ASW) for customs clearances.Another serious source of adm<strong>in</strong>istrative <strong>in</strong>efficiency and burdensome compliance costs fortaxpayers arises from the cont<strong>in</strong>ued use of out-dated approaches to risk management. Manydevelop<strong>in</strong>g countries have been adopt<strong>in</strong>g systems us<strong>in</strong>g automated statistical analysis of past taxrecords and other third-party data to select high-risk taxpayers as targets for the verification ofreturns, tax <strong>in</strong>vestigations, customs <strong>in</strong>spections, and <strong>in</strong>tegrated audits. This system supportsbroader application of “green channel” procedures to allow simple and rapid process<strong>in</strong>g of mostcustoms and tax transactions with m<strong>in</strong>imal revenue risk.Automated systems are also the hallmark of good practice <strong>in</strong> screen<strong>in</strong>g tax returns and other<strong>in</strong>com<strong>in</strong>g documents to catch arithmetic errors or other <strong>in</strong>consistencies, and for identify<strong>in</strong>g andnotify<strong>in</strong>g non-filers, stop-filers, and late filers. Here, too, more effective use of computerizedsystems can free AT personnel from rout<strong>in</strong>e low-value tasks and allow a reallocation of humanresources to activities that are more productive <strong>in</strong> generat<strong>in</strong>g revenue.The VAT refund process is a prime case <strong>in</strong> po<strong>in</strong>t. Over the past few years the AT has reducedrefund delays and improved its support to taxpayers who have difficulty deal<strong>in</strong>g with theprocedures and documentation requirements. But the system is still paper based, with detailedmanual verification of every refund petition, and approvals centralized <strong>in</strong> Maputo. A moreefficient system would be based on electronic refund petitions, automated error-check<strong>in</strong>g <strong>in</strong> realtime, prompt electronic transmission to a centralized approval center, without requir<strong>in</strong>g theshipment of paper files, selective risk-based verification based on the taxpayers overallcompliance record, and prompt payment of low-risk claims.The VAT refund system also illustrates the critical importance of thoroughly re-eng<strong>in</strong>eer<strong>in</strong>gworkflows <strong>in</strong> conjunction with IT modernization. To realize the potential benefits of electronictax systems, it is essential to restructure basic bus<strong>in</strong>ess practices rather than simply putt<strong>in</strong>gcurrent practices onto a computer. This applies to virtually every area of operation.Accompany<strong>in</strong>g procedural reforms, the AT also needs to expand the management <strong>in</strong>formationsystem (MIS) to <strong>in</strong>clude more <strong>in</strong>dicators of operational efficiency (<strong>in</strong>dicadores de desempenho).The Annual Reports (ARs) of the AT are filled with tables and figures provid<strong>in</strong>g important<strong>in</strong>formation on issues such as revenue results relative to targets; the number of registrations,audits, <strong>in</strong>spections, refunds, and other proceed<strong>in</strong>gs; the size and composition of AT staff; and thebudget for tax adm<strong>in</strong>istration. But management also needs regular reports on operational


28 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEefficiency, such as data on added revenue per audit and per auditor; the ratio of late-filers andnon-filers to registered (non-exempt) taxpayers; the time distribution of refund payments; thepercentage of customs clearances or VAT refunds benefit<strong>in</strong>g from Green Channel treatment; andsurveys of taxpayer knowledge and customer satisfaction. 28Another major challenge is that the software systems currently <strong>in</strong> use by customs (DGA) and thedomestic tax department (DGI) do not share data.. 29 The software platforms need to be changedso that all of data for each taxpayer can be <strong>in</strong>tegrated <strong>in</strong>to a s<strong>in</strong>gle master file, based on his or herunique taxpayer identification number (NUIT). Integration of the data systems is a basicrequirement for efficient risk management and case management.Human ResourcesThe AT has made notable progress <strong>in</strong> <strong>in</strong>creas<strong>in</strong>g the complement of staff, upgrad<strong>in</strong>g educationalcredentials, and expand<strong>in</strong>g tra<strong>in</strong><strong>in</strong>g programs. However, the AT still needs to embed thesedevelopments <strong>in</strong> a strategic approach for <strong>in</strong>tegrated human resource management to ensure thatpersonnel are deployed across the organization to maximum advantage <strong>in</strong> generat<strong>in</strong>g revenue andimprov<strong>in</strong>g taxpayer compliance. As noted above, the <strong>in</strong>troduction of e-taxation systems andmodern approaches to risk management will open new avenues for reallocat<strong>in</strong>g staff and<strong>in</strong>creas<strong>in</strong>g operational efficiency.The on-go<strong>in</strong>g process of adm<strong>in</strong>istrative reform also creates new tra<strong>in</strong><strong>in</strong>g needs. For example, theestablishment of a large taxpayers unit to handle all aspects of tax adm<strong>in</strong>istration for largecorporations with sophisticated f<strong>in</strong>ancial plann<strong>in</strong>g capabilities, accentuates the need for tra<strong>in</strong><strong>in</strong>g<strong>in</strong> advanced audit techniques and comprehensive taxpayer services <strong>in</strong>volv<strong>in</strong>g even arcanetechnicalities <strong>in</strong> the tax code.The Culture of <strong>Tax</strong>ationComments on the tax culture <strong>in</strong> <strong>Mozambique</strong> were among the most frequent observations aris<strong>in</strong>g<strong>in</strong> field <strong>in</strong>terviews. Three related elements stand out. First, nearly every <strong>in</strong>terlocutor from theprivate sector offered vivid anecdotal examples illustrat<strong>in</strong>g the lack of a customer service cultureamong many tax officials, especially outside of Maputo. The <strong>in</strong>cidents generally <strong>in</strong>volveencounters where tax officials pursue a punitive approach to enforcement, such as levy<strong>in</strong>g heavypenalties for <strong>in</strong>advertent and <strong>in</strong>consequential mistakes, or errors stemm<strong>in</strong>g from <strong>in</strong>adequate<strong>in</strong>formation about tax system, rather than tak<strong>in</strong>g these opportunities to educate taxpayers and helpthem avoid errors. The <strong>in</strong>terviews and the documentary evidence suggest that the AT isimprov<strong>in</strong>g taxpayer services and tax education programs. Yet there are still many complexities orprocedural requirements that can trip up even well mean<strong>in</strong>g taxpayers. Also, the <strong>in</strong>centive28 Appendix E contrasts the type of data conta<strong>in</strong>ed <strong>in</strong> the Annual Reports with the type of datarecommended for monitor<strong>in</strong>g operational efficiency, as suggested <strong>in</strong> IMF(2009c). The Appendix alsotabulates data show<strong>in</strong>g trends <strong>in</strong> selected <strong>in</strong>dicators of tax adm<strong>in</strong>istration performance, drawn from theAT’s Annual Reports.29 The customs data system has other significant technical limitations which will not be discussed here.These limitations have been identified <strong>in</strong> restricted reports to the AT from the IMF and a Quality AssuranceGroup (QAG) of tax experts represent<strong>in</strong>g donors contribut<strong>in</strong>g to a Common Fund <strong>in</strong> support of tax reforms.


C URRENT T AX S YSTEM 29structure with<strong>in</strong> the AT encourages harsh enforcement practices whenever a local tax office isfall<strong>in</strong>g short of an assigned revenue target. Revenue targets that are overly ambitious relative toeconomic conditions and <strong>in</strong>stitutional capacity tend to create costly problems for small andmedium bus<strong>in</strong>esses at the local level.Many <strong>in</strong>terviewees emphasized equally the lack of a “taxpay<strong>in</strong>g culture” <strong>in</strong> <strong>Mozambique</strong>. The AThas been work<strong>in</strong>g to address this problem through public <strong>in</strong>formation highlight<strong>in</strong>g the role oftaxation <strong>in</strong> f<strong>in</strong>anc<strong>in</strong>g public services, and the obligation of citizens to contribute to nationaldevelopment. The AT is also <strong>in</strong> the f<strong>in</strong>al stages of develop<strong>in</strong>g its website to provide ready accessto any bus<strong>in</strong>esses or citizens with Internet connections. This <strong>in</strong>formation campaign is animportant step <strong>in</strong> the right direction, though it will take time to alter deeply <strong>in</strong>gra<strong>in</strong>ed attitudesand habits. Chang<strong>in</strong>g the culture of taxpay<strong>in</strong>g is particularly difficult if tax officials are oftenperceived as be<strong>in</strong>g predatory rather than supportive, and if the public does not see correspond<strong>in</strong>gbenefits from government expenditures.A related observation on the tax culture, also widely encountered, is that tax evasion andsmuggl<strong>in</strong>g are still rampant and blatant, along with bribery and corruption. Here, too, the AT hasmade progress <strong>in</strong> adopt<strong>in</strong>g and publiciz<strong>in</strong>g a code of conduct for tax and customs officials, andpromot<strong>in</strong>g greater <strong>in</strong>tegrity and professionalism throughout the organization. By all accounts,however, the revenue loss from unethical practices on the part of taxpayers and tax officials isstill very large.PLANS FOR MODERNIZING TAX ADMINISTRATIONThe discussion of tax system would be <strong>in</strong>complete without <strong>in</strong>clud<strong>in</strong>g a summary of plans forfurther modernization of tax adm<strong>in</strong>istration. The Autoridade Tributária (AT) recently developedand adopted a Strategic Plan for reform <strong>in</strong> 2009 and 2010, along with a Tactical Plan for 2009sett<strong>in</strong>g deadl<strong>in</strong>es and responsibilities for each activity. The Strategic Plan <strong>in</strong>cludes an overallstatement of vision, mission, and values for the AT: 30• Vision: To establish premier quality services <strong>in</strong> collect<strong>in</strong>g revenues and promot<strong>in</strong>g andprotect<strong>in</strong>g the economy and society.• Mission: To collect revenues for f<strong>in</strong>anc<strong>in</strong>g public sector activities by apply<strong>in</strong>g the tax andcustoms laws effectively, efficiently and equitably, with greater convenience fortaxpayers <strong>in</strong> comply<strong>in</strong>g with their obligations, as well as protect<strong>in</strong>g the economy andsociety.• Values: Confidence and mutual respect, fairness, <strong>in</strong>tegrity, transparency, courtesy,dedication and excellence.The Plan def<strong>in</strong>es three strategic objectives and actions:• Strategic objective 1: To <strong>in</strong>crease tax collection <strong>in</strong> a susta<strong>in</strong>able manner.30 Translation by the authors of the present report.


30 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUE⎯ Strategy 1.1: Help taxpayers understand and comply with the tax laws. Actions<strong>in</strong>clude promot<strong>in</strong>g tax education; provid<strong>in</strong>g taxpayer assistance; and simplify<strong>in</strong>gprocedures for pay<strong>in</strong>g taxes.⎯ Strategy 1.2: Strengthen measures to improve compliance with tax obligations.Actions <strong>in</strong>clude the formulation of legislation to combat tax evasion; simplificationand improvement <strong>in</strong> processes for <strong>in</strong>spection, <strong>in</strong>vestigation and audit; more effectivesystems for collection and control of tax debts; and strengthen<strong>in</strong>g regional and<strong>in</strong>ternational cooperation.⎯ Strategy 1.3: Create an <strong>in</strong>tegrated central management for large taxpayers and megaprojects.Actions <strong>in</strong>clude development and implementation of a new organizationalunit to provide more professional treatment of large taxpayers; and strengthen<strong>in</strong>g thespecialized capabilities needed by tax auditors deal<strong>in</strong>g with mega-projects.• Strategic objective 2: To modernize and strengthen tax adm<strong>in</strong>istration.⎯ Strategy 2.1: Develop the human resource management system and improve thequality of life for tax officials. Actions <strong>in</strong>clude recruit<strong>in</strong>g, tra<strong>in</strong><strong>in</strong>g and promot<strong>in</strong>g ATfunctionaries based on qualifications, experience and performance; establish<strong>in</strong>g aunified career framework for AT personnel; and provid<strong>in</strong>g social and professionalassistance to AT staff.⎯ Strategy 2.2: Improve the organizational structure and management system toguarantee well planned and well managed operations. Actions <strong>in</strong>clude develop<strong>in</strong>g anew organizational structure; implement<strong>in</strong>g an effective system of changemanagement; and extend<strong>in</strong>g geographical coverage of tax services throughout thenational territory.⎯ Strategy 2.3: Improve the <strong>in</strong>frastructure to support effective function<strong>in</strong>g of <strong>Tax</strong>Authority. Actions <strong>in</strong>clude new and rehabilitated build<strong>in</strong>gs; new equipment andfurnish<strong>in</strong>gs, <strong>in</strong>clud<strong>in</strong>g vehicles; and effective management <strong>in</strong> provid<strong>in</strong>g essentialgods and services.⎯ Strategy 2.4: Strengthen the organizational culture of the <strong>Tax</strong> Authority. Actions<strong>in</strong>clude programs to promote <strong>in</strong>tegrity and combat corruption; and development of aculture of controls <strong>in</strong> account<strong>in</strong>g, legal procedures and systematic risk management.• Strategic objective 3: To develop <strong>in</strong>formation technology that permits the improvementof management of tax processes⎯ Strategy 3.1: Modernise bus<strong>in</strong>esses processes of tax adm<strong>in</strong>istration through theapplication of <strong>in</strong>formation technology and ma<strong>in</strong>tenance of the function<strong>in</strong>g of thecurrent system. Actions <strong>in</strong>clude design<strong>in</strong>g modern software applications for e-NUITs,e-<strong>Tax</strong>ation, and the automated S<strong>in</strong>gle W<strong>in</strong>dow for customs; provision of modernhardware, <strong>in</strong>tegration of software systems, and implementation of an efficientcommunication environment; effective education and tra<strong>in</strong><strong>in</strong>g for AT personnel; andsystems ma<strong>in</strong>tenance to guarantee smooth function<strong>in</strong>g of operations andadm<strong>in</strong>istration while new <strong>in</strong>tegrated systems are developed.


C URRENT T AX S YSTEM 31F<strong>in</strong>ally, the Strategy establishes four quantitative and qualitative <strong>in</strong>dicators for measur<strong>in</strong>gperformance aga<strong>in</strong>st the objectives:• Annual <strong>in</strong>crease <strong>in</strong> revenues of 0.5 percentage po<strong>in</strong>ts of GDP.• M<strong>in</strong>imum annual <strong>in</strong>crease <strong>in</strong> collections of 2 percent over the previous year.• Survey results on taxpayer satisfaction.• Survey results on AT personnel satisfactionThe Strategy acknowledges major issues fac<strong>in</strong>g the AT <strong>in</strong> order to improve efficiency,effectiveness and fairness <strong>in</strong> tax adm<strong>in</strong>istration. Particularly important are the need for taxpayereducation and taxpayer support services; stronger enforcement and collection, especially relat<strong>in</strong>gto large taxpayers; the <strong>in</strong>troduction of modern and <strong>in</strong>tegrated IT systems; better tra<strong>in</strong><strong>in</strong>g for taxofficials; and creation of a culture of <strong>in</strong>tegrity and control with<strong>in</strong> the organization. The difficulty,of course, comes <strong>in</strong> translat<strong>in</strong>g the strategy <strong>in</strong>to effective action.


4. Revenue PerformanceThis chapter reviews revenue performance over the <strong>PARPA</strong> <strong>II</strong> period and preced<strong>in</strong>g years. Thechapter beg<strong>in</strong>s with a discussion of revenue targets <strong>in</strong> the <strong>PARPA</strong> and the government’s mediumtermfiscal framework (MTFF), and follows this with a descriptive overview of the revenuetrends. For <strong>in</strong>sight <strong>in</strong>to what might be an appropriate revenue target for <strong>Mozambique</strong> the chapteralso exam<strong>in</strong>es <strong>in</strong>ternational benchmarks for the ratio of revenue to GDP, and presents aregression analysis based on cross-section data for develop<strong>in</strong>g countries.While tax revenues for 2009 are projected to fall short of the <strong>PARPA</strong> <strong>II</strong> target, the chapterconcludes that overall revenue growth has been strong over the <strong>PARPA</strong> <strong>II</strong> period, reflect<strong>in</strong>gpositively on AT efficiency and the policy reforms carried out over the last ten years. Inconjunction with evidence presented elsewhere <strong>in</strong> this study, the analysis of revenue buoyancyand appropriate <strong>in</strong>ternational comparisons <strong>in</strong>dicate that there is still scope for <strong>in</strong>creas<strong>in</strong>g therevenue ratio <strong>in</strong> the com<strong>in</strong>g years.REVENUE TARGETSAs shown <strong>in</strong> Chapter 2, <strong>PARPA</strong> <strong>II</strong> set a target for total domestic revenue target to reach 16.2percent of GDP <strong>in</strong> 2009, up from 14.0 percent <strong>in</strong> 2005. For the longer term, the <strong>in</strong>dicative target<strong>in</strong> <strong>PARPA</strong> <strong>II</strong> is to <strong>in</strong>crease revenues to 16.6 percent of GDP <strong>in</strong> 2014 (p.39, Table 11). The<strong>PARPA</strong> also states that these targets are to be achieved without plac<strong>in</strong>g an undue burden on theformal sector, by broaden<strong>in</strong>g the tax-base and reduc<strong>in</strong>g tax exemptions and evasion.The long-term need to <strong>in</strong>crease revenues is highlighted by the persistently large fiscal gap – thedifference between total expenditures and total revenues, exclud<strong>in</strong>g grants and loans.Figure 4-1 illustrates the scale of the issue, 31 show<strong>in</strong>g that external grants and loans f<strong>in</strong>anced 49percent of expenditures <strong>in</strong> 2006 and 43 percent <strong>in</strong> 2008. While a reliance on external f<strong>in</strong>ance isacknowledged as a necessity for the foreseeable future <strong>in</strong> <strong>PARPA</strong> <strong>II</strong>, the long-term goal is toreduce this dependency by <strong>in</strong>creas<strong>in</strong>g government revenues.31 As the latest wave of major tax reforms <strong>in</strong> <strong>Mozambique</strong> goes back to the <strong>in</strong>troduction of VAT <strong>in</strong> 1999,this is taken as the start<strong>in</strong>g po<strong>in</strong>t for analyz<strong>in</strong>g revenue performance, permitt<strong>in</strong>g a longer time horizon foranalysis of trends and patterns although the focus will rema<strong>in</strong> on the <strong>PARPA</strong> <strong>II</strong> period 2006 to 2008.


34 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEFigure 4-1Domestic Revenue & Expenditure as a Share of GDPSOURCE: MPD, "Quadromacro Revisto CFMP Proposta", received June 2009.This long-run goal is reflected <strong>in</strong> government annual budget programs which, <strong>in</strong> agreement withthe IMF, have targeted an <strong>in</strong>crease <strong>in</strong> revenues of 0.5 percent of GDP per year s<strong>in</strong>ce 2005 (IMFreports from 2005 to 2009a). In fact, the latest Medium Term Fiscal Framework (Cenário Fiscalde Médio Prazo- CFMP) projections target average growth of nearer 0.6 percent of GDP from2005 to reach 17.4 percent of GDP <strong>in</strong> 2011, reflect<strong>in</strong>g more ambitious targets than thoseexpressed <strong>in</strong> <strong>PARPA</strong> <strong>II</strong>. 32 Further, a recent IMF Debt Susta<strong>in</strong>ability Analysis assumes that taxrevenue will reach 21 percent of GDP <strong>in</strong> 2028, with total revenues reach<strong>in</strong>g 23.5 percent of GDP(IMF 2009a).This optimistic long-term view is based at least to some degree on two IMF studies. 33 An <strong>in</strong>ternalstudy <strong>in</strong> 2004 used a comb<strong>in</strong>ation of macroeconomic data, household surveys and foreign tradestatistics to estimate the potential tax base for VAT, <strong>in</strong>come tax and import duties, and thenapplies the respective tax rules and rates to the estimated tax base to arrive at an estimate ofpotential revenue. Compar<strong>in</strong>g potential revenue with actual collection, Schenone estimates thatthe maximum revenue potential for these three taxes was roughly double the realized revenue of11 percent of GDP <strong>in</strong> 2001 and 2002, illustrat<strong>in</strong>g a large degree of foregone revenue due toexemptions, evasion, and weak adm<strong>in</strong>istration as of the reference dates. The second IMF studyapplied an econometric “tax frontier” model to estimate the potential tax revenue from crosscountrypanel data for the period 1995 to 2005. This study also estimated the tax capacity for<strong>Mozambique</strong> at around 22 percent of GDP. The authors note that this would be on a par withKenya’s performance.32 From MPD, DNEAP's "Quadromacro Revisto CFMP Proposta", received June 2009.33 Both studies are discussed <strong>in</strong> Varsano et al. (2006).


R EVENUE P ERFORMANCE 35Notwithstand<strong>in</strong>g these conclusions and ambitious revenue targets, growth and revenue forecastshave recently been revised downwards, reflect<strong>in</strong>g effects of the world economic downturn. Themost recent IMF report from June 2009 projects total revenue on the order of 15.7 percent ofGDP <strong>in</strong> 2009, ris<strong>in</strong>g to 16.5 percent <strong>in</strong> 2011 (IMF 2009b).The follow<strong>in</strong>g section discusses revenue performance and the observed trends for the ma<strong>in</strong>contribut<strong>in</strong>g taxes <strong>in</strong> light of the <strong>PARPA</strong> <strong>II</strong> and other targets mentioned above.REVENUE GROWTHRecent revenue growth has been strong. After a dip <strong>in</strong> the ratio of revenue to GDP <strong>in</strong> 2004,growth <strong>in</strong> the ratio resumed <strong>in</strong> 2005 and cont<strong>in</strong>ued for the first two years of <strong>PARPA</strong> <strong>II</strong>, co<strong>in</strong>cid<strong>in</strong>gwith the <strong>in</strong>troduction of the AT. This helped to narrow the gap between domestic revenues andexpenditures (illustrated <strong>in</strong> Figure 4-1). In fact, domestic revenues reached the 2009 <strong>PARPA</strong> <strong>II</strong>target of 16.2 percent of GDP <strong>in</strong> 2007, and narrowly miss<strong>in</strong>g this mark <strong>in</strong> 2008, with 16.0 percentof GDP, as shown <strong>in</strong> Table 4-1.Table 4-1Pr<strong>in</strong>cipal Revenues as Share of GDP (%)1999 2000 2001 2002 2003 2004 2005 2006 2007 2008Total revenues 10.8 11.5 11.2 11.4 13.1 12.4 13.8 15.2 16.2 16.0<strong>Tax</strong> revenues 10.0 10.5 10.0 10.5 12.3 10.8 11.1 12.3 13.4 13.5Income taxes 1.5 1.6 1.8 2.1 2.9 2.7 2.9 3.5 4.5 4.9Personal <strong>in</strong>come tax 0.8 1.0 1.1 1.4 2.0 1.9 1.9 2.1 2.3 2.5Corporate <strong>in</strong>come tax 0.7 0.6 0.6 0.7 0.8 0.8 1.0 1.4 2.1 2.4Expenditure taxes 6.6 7.3 6.9 7.2 7.9 7.2 7.4 8.0 8.3 8.0Value-added tax (VAT) 3.9 4.4 4.2 4.4 4.9 4.5 4.5 5.2 5.4 5.4VAT on domestic transactions 1.0 1.8 1.9 2.0 2.1 1.9 1.8 2.1 3.2 2.4VAT on imports 1.4 2.6 2.4 2.6 2.7 2.6 2.7 3.2 2.2 3.0Spec. consumption tax (SCT) dom. 0.6 0.6 0.6 0.6 0.7 0.6 0.6 0.6 0.6 0.7SCT Beer and soft-dr<strong>in</strong>ks 0.0 0.5 0.4 0.5 0.5 0.4 0.5 0.5 0.5 0.5SCT – tobacco 0.0 0.1 0.2 0.2 0.2 0.2 0.2 0.2 0.1 0.1SCT – Other products 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1Specific consumption tax imports 0.4 0.3 0.3 0.3 0.4 0.4 0.4 0.4 0.4 0.4Customs duties and sugar surtax 1.7 1.9 1.8 1.9 2.0 1.7 1.9 1.8 1.8 1.5Other tax revenues 1.9 1.6 1.3 1.2 2.2 0.9 0.8 0.7 0.7 0.6Other revenues 0.7 1.0 1.2 0.9 0.9 1.8 2.9 3.0 2.7 2.5Mem: Imposto s/ Combustivel 1.3 1.2 1.0 0.9 1.2 1.3 1.2 1.0 1.1 1.0Notes: Years 1999 to 2007 from Conta Geral do Estado, 2008 from Relatorio de Execucao. Data MPD"Quadromacro Revisto CFMP Proposta", received June 2009. Note that here Imposto de Combustivel is presented separately asbut is otherwise <strong>in</strong>cluded partly <strong>in</strong> "Non-tax revenues" and partly <strong>in</strong> "Consigned revenues".With<strong>in</strong> total revenues, tax receipts also <strong>in</strong>creased as a share of GDP from 10.8 percent <strong>in</strong> 2004to13.5 percent <strong>in</strong> 2008. The difference between total revenue growth and tax revenue growth is


36 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEdue to “other revenues”, compris<strong>in</strong>g non-tax revenues, pre-assigned revenues and capitalrevenues, with assigned fuel tax and capital revenues mostly responsible for growth <strong>in</strong> thiscategory.Despite miss<strong>in</strong>g the <strong>PARPA</strong> <strong>II</strong> goal for 2008 and 2009, overall revenue performance suggestssignificant ga<strong>in</strong>s from the <strong>in</strong>troduction of the AT, reforms to tax adm<strong>in</strong>istration, and efforts tobroaden the tax base. This is particularly impressive given that several tax policy measures werecarried out that were likely to reduce revenues, namely lower<strong>in</strong>g the top import tariff rate,implementation of the SADC trade protocol, rais<strong>in</strong>g the VAT threshold, and the temporarysuspension of fuel tax. Part of the <strong>in</strong>crease <strong>in</strong> revenues is also due to the natural elasticity of thetax system.Revenue Elasticity and BuoyancyIn discuss<strong>in</strong>g revenue performance, it is important to understand the concepts of revenueelasticity and buoyancy. Revenue elasticity refers to the <strong>in</strong>crease <strong>in</strong> revenues which occursnaturally as a result of economic growth, exclud<strong>in</strong>g the revenue effects of changes <strong>in</strong> tax policy oradm<strong>in</strong>istrative reforms. Economic growth itself moves <strong>in</strong>dividuals <strong>in</strong>to higher tax brackets,<strong>in</strong>creases the number of households and bus<strong>in</strong>esses above the exemption threshold, pulls moreworkers and firms <strong>in</strong>to the formal sector, <strong>in</strong>creases the <strong>in</strong>comes of successful firms, and expandsexpenditure on taxable goods and services relative to GDP. These basic factors suggest that boththe corporate and <strong>in</strong>dividual <strong>in</strong>come tax, as well as the VAT are <strong>in</strong>herently elastic sources ofrevenue. That is to say that they tend to generate higher revenues relative to GDP as the economygrows, aside from any measures to broaden the tax base, improve collection efficiency, orimprove tax compliance. To a lesser extent the same is true of the excise tax, because sumptuaryspend<strong>in</strong>g tends to <strong>in</strong>crease relative to total consumption expenditure as <strong>in</strong>comes rise. Customsduties, on the other hand, may be less elastic, depend<strong>in</strong>g on trends <strong>in</strong> the ratio of trade to GDPand changes <strong>in</strong> the composition of imports.Revenue elasticity is hard to quantify due to the difficulty of separat<strong>in</strong>g the effect of policymeasures from the underly<strong>in</strong>g dynamics due to growth. Revenue buoyancy is an alternativemeasure that is easier to calculate because it simply looks at the overall change <strong>in</strong> revenuesrelative to <strong>in</strong>come and expenditure, without try<strong>in</strong>g to dist<strong>in</strong>guish the effects of policy oradm<strong>in</strong>istrative measures. Table 4-2 presents revenue buoyancy measures by type of tax, averagedover two periods: 2000 to 2008, and the <strong>PARPA</strong> <strong>II</strong> period of 2006 to 2008. The averagebuoyancy of total revenues over the period 2000-2008 is 1.38, imply<strong>in</strong>g that revenues <strong>in</strong>creasedby 1.38 percent for each 1 percent <strong>in</strong>crease <strong>in</strong> GDP. Over the shorter <strong>PARPA</strong> <strong>II</strong> period, totalrevenue buoyancy has been nearly the same, at 1.34. For both periods the buoyancy rate has beenhigher than one, show<strong>in</strong>g that revenues grew considerably faster than nom<strong>in</strong>al GDP.<strong>Tax</strong> revenues were less buoyant than total revenues over the full period 2000 to 2008, but therelationship reversed <strong>in</strong> the <strong>PARPA</strong> <strong>II</strong> period, with a strong buoyancy factor of 1.48 for taxreceipts. Income tax revenues have been especially buoyant over this period, particularly thecorporate <strong>in</strong>come tax (IRPC) for which revenues <strong>in</strong>creased by 3.5 percent for every 1 percent rise<strong>in</strong> GDP. VAT revenues have also been highly buoyant over the past three years, with domesticris<strong>in</strong>g by 2 percent and the VAT on imports ris<strong>in</strong>g by 1.58 percent for every 1 percent rise <strong>in</strong>GDP.


R EVENUE P ERFORMANCE 37Table 4-2Revenue Buoyancy with Respect to Nom<strong>in</strong>al GDPAve. 00-08 Ave. 06-08Total revenue 1.38 1.34<strong>Tax</strong> revenue 1.32 1.48Income taxes 2.08 2.35Personal <strong>in</strong>come tax 2.13 1.62Corporate <strong>in</strong>come tax 2.10 3.49Expenditure taxes 1.22 1.18VAT on domestic transactions 2.16 2.00VAT on imports 2.00 1.58Specific consumption tax - domestic 1.15 1.15Specific consumption tax - imports 1.17 1.28Customs duties and sugar surtax 0.97 0.52Fuel tax 1.06 0.78Other revenues 2.37 0.64SOURCE: Authors’ calculations us<strong>in</strong>g data from MPD, "Quadromacro Revisto CFMP Proposta",received June 2009.This is important to keep <strong>in</strong> m<strong>in</strong>d <strong>in</strong> discuss<strong>in</strong>g the prospects for further revenue <strong>in</strong>creases as ashare of GDP. The buoyancy evidence suggests that adm<strong>in</strong>istrative and tax policy measures,comb<strong>in</strong>ed with the <strong>in</strong>herent effects of rapid GDP growth, have produced strong growth <strong>in</strong> taxrevenue relative to GDP over the <strong>PARPA</strong> <strong>II</strong> period. There is no reason to expect the favorabledynamics to dissipate over the medium term given the prospective revenue effects of recentchanges <strong>in</strong> tax policy relat<strong>in</strong>g to future large projects, <strong>in</strong>clud<strong>in</strong>g m<strong>in</strong>eral and petroleum ventures,the rapid expansion <strong>in</strong> tax registrations, and the scope for substantial revenue ga<strong>in</strong>s throughfurther adm<strong>in</strong>istrative reforms, as discussed <strong>in</strong> Chapter 3.Revenue SourcesFigure 4-2 disaggregates total revenues <strong>in</strong>to the ma<strong>in</strong> components: <strong>in</strong>come taxes; VAT; customsduties (plus the surcharge on sugar imports); fuel tax; “other taxes;” and “other revenues.” 34 Thisclearly shows that the largest contributors to revenue are <strong>in</strong>come taxes and VAT and that thesehave been <strong>in</strong>creas<strong>in</strong>g as a share of GDP at least s<strong>in</strong>ce 2005, although VAT revenue growth tailsoff <strong>in</strong> 2008. The next largest contributor to revenues is customs duties and the sugar surtax,followed by “other revenues” which have seen a leap <strong>in</strong> their share of GDP s<strong>in</strong>ce 2004. These arediscussed <strong>in</strong> more detail below.34 Revenues from the specific consumption tax (excise tax) on domestic goods and imports are left out to avoidclutter, and given their relatively low share of GDP through time. More detailed data are presented above <strong>in</strong> Table 4-1.


38 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEFigure 4-2Pr<strong>in</strong>cipal <strong>Tax</strong> Revenues as a Share of GDPSOURCE: MPD, "Quadromacro Revisto CFMP Proposta", received June 2009.Expenditure <strong>Tax</strong>esExpenditure taxes are the primary source of domestic revenue, represent<strong>in</strong>g 9.0 percent of GDP <strong>in</strong>2008, with 60 percent of this sum com<strong>in</strong>g from VAT. 35 VAT revenues have <strong>in</strong>creased as a shareof GDP from 3.9 percent of GDP <strong>in</strong> 1999 to 5.2 percent <strong>in</strong> 2006 and 5.4 percent <strong>in</strong> 2007 and2008, mak<strong>in</strong>g VAT by far the most important tax, provid<strong>in</strong>g approximately one third of totalrevenues and 40 percent of overall tax receipts <strong>in</strong> 2008.Given the importance of VAT revenues, it is useful to exam<strong>in</strong>e revenues from domestictransactions and imports separately. As can be seen <strong>in</strong> Table 4-1 above, VAT collections onimports fell from 3.2 percent of GDP <strong>in</strong> 2006 to 2.2 percent of GDP <strong>in</strong> 2007 before rebound<strong>in</strong>g <strong>in</strong>2008 to 3.0 percent of GDP. At the same time, revenues from import duties and the sugar surtaxdecl<strong>in</strong>ed <strong>in</strong> 2008 from 1.8 percent of GDP to 1.5 percent. While this decl<strong>in</strong>e <strong>in</strong> duty revenuesmakes sense follow<strong>in</strong>g the reduction <strong>in</strong> the tariff rate on imports of consumer goods and thephas<strong>in</strong>g <strong>in</strong> of SADC trade preferences, the accompany<strong>in</strong>g VAT surge should reflect the broadertrend <strong>in</strong> imports. In fact, the local currency value of imports <strong>in</strong>creased by just 8.6 percent <strong>in</strong> 2007– below the growth rate for nom<strong>in</strong>al GDP. Imports by value then jumped by 24.7 percent <strong>in</strong> 2008,reflect<strong>in</strong>g <strong>in</strong> part the surge <strong>in</strong> world commodity prices. 3635 Note for 2005 that the breakdown of VAT between domestic and imported goods, and the breakdownof the ICE on national products between beverages, tobacco and other goods is imputed as an average ofthe breakdown <strong>in</strong> 2004 and <strong>in</strong> 2006 due to the absence of the relevant data.36 Figures from the DNEAP Quadromacro from June 2008.


R EVENUE P ERFORMANCE 39In contrast to VAT on imports, domestic VAT receipts <strong>in</strong>creased markedly <strong>in</strong> 2007 and then fell<strong>in</strong> 2008 as a share of total revenue. This may be a consequence of the revised VAT code which<strong>in</strong>troduced a 60 percent reduction <strong>in</strong> VAT on supplies of services to the state for public workssuch as roads, bridges and water supply <strong>in</strong>frastructures. The <strong>in</strong>crease <strong>in</strong> the turnover threshold forVAT exemption may also have had an effect on revenues beg<strong>in</strong>n<strong>in</strong>g <strong>in</strong> 2008. Even though a largenumber of firms may have been affected (no figures are available), this is not likely to have beena major factor given that affected firms would be very small, and the exemption only elim<strong>in</strong>atesVAT on a firm’s f<strong>in</strong>al sales marg<strong>in</strong>, without provid<strong>in</strong>g relief from VAT paid on <strong>in</strong>puts. Furtheranalysis would be needed to expla<strong>in</strong> more fully the recent behavior of VAT receipts. Analysis ofthis sort should be a rout<strong>in</strong>e function of the studies office (GEST) <strong>in</strong> the M<strong>in</strong>istry of F<strong>in</strong>ance.Overall VAT performance can also be exam<strong>in</strong>ed <strong>in</strong> terms of a standard <strong>in</strong>dicator called VATproductivity. This is def<strong>in</strong>ed as the ratio of VAT receipts to GDP, divided by the VAT rate, which<strong>in</strong> this case is 17 percent. The result<strong>in</strong>g value lies between zero and one, where an estimate closeto one implies that actual VAT collections are high relative to the theoretical maximum, or highVAT productivity, while a value close to zero implies low productivity due to a comb<strong>in</strong>ation ofexemptions, zero-rat<strong>in</strong>g, tax evasion, and weak tax adm<strong>in</strong>istration. Table 4-3 shows that VATproductivity <strong>in</strong> <strong>Mozambique</strong> has broadly been ris<strong>in</strong>g s<strong>in</strong>ce 2000, the first full year ofimplementation, with m<strong>in</strong>or decl<strong>in</strong>es <strong>in</strong> productivity 2005 and 2006 only.Table 4-3<strong>Mozambique</strong> VAT Productivity2000 2001 2002 2003 2004 2005 2006 2007 2008VAT Productivity 0.23 0.26 0.25 0.26 0.29 0.26 0.26 0.31 0.32SOURCE: Authors’ calculations us<strong>in</strong>g data from MPD, "Quadromacro Revisto CFMP Proposta", received June 2009.The Mozambican average for the period 2006 to 2008 is 0.30 which, as shown <strong>in</strong> Table 4-4, isabove the average of 0.25 for sub-Saharan Africa, and the average of 0.24 for low-<strong>in</strong>comeeconomies worldwide. Although this still puts <strong>Mozambique</strong> below the levels atta<strong>in</strong>ed <strong>in</strong> the lowmiddle-<strong>in</strong>comeeconomies group of 0.47, it should be noted that VAT productivity is 0.37 forWestern Europe, and 0.32 for the United States and Canada.In relation to other expenditure taxes, the revenue yield from the domestic excise tax and theexcise tax on imports has rema<strong>in</strong>ed roughly constant at around 0.6 percent of GDP and 0.4percent of GDP, respectively. Revenues from fuel tax have also been relatively stable, rang<strong>in</strong>gbetween 1.0 and 1.3 percent of GDP, despite the forego<strong>in</strong>g of fuel tax receipts as of June 2008mentioned above.


40 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUETable 4-4VAT Productivity for <strong>Mozambique</strong> and Comparator Country GroupsCountry/GroupVATProductivityLow-<strong>in</strong>come economies group 0.24Sub-Saharan Africa 0.25South Asia 0.28<strong>Mozambique</strong> 0.30United States and Canada 0.32Western Europe 0.37High-<strong>in</strong>come economies group 0.39Upper-middle-<strong>in</strong>come economies group 0.43Lat<strong>in</strong> America and the Caribbean 0.44East Asia and Pacific 0.45Central Europe and Central Asia 0.45Middle East and North Africa 0.47Low-middle-<strong>in</strong>come economies group 0.47Note: Fiscal Reform data does not refer to one particular year, but employs the most recently available data as of December 2007,the most part com<strong>in</strong>g from 2005, 2006 and 2007 (Fiscal Reform, 2009, accessed athttp://www.fiscalreform.net/<strong>in</strong>dex.php?option=com_wrapper&Itemid=132).SOURCE: Fiscal Reform (2009).Income <strong>Tax</strong>Income tax revenues <strong>in</strong>creased markedly from 1.5 percent of GDP <strong>in</strong> 1999 to 3.5 percent <strong>in</strong> 2006and then to 4.9 percent <strong>in</strong> 2008. This overall growth stemmed from implementation of the revised<strong>in</strong>come tax code (the IRPS and IRPC) <strong>in</strong> 2003 which, from 2004, led to a higher growth path forrevenues, and from more efficient tax enforcement follow<strong>in</strong>g establishment of the AT.Figure 4-3 shows that <strong>in</strong>come tax revenues climbed from under 15 percent of total revenues <strong>in</strong>1999 to over 30 percent <strong>in</strong> 2008. The graph shows that much of the growth has been due to thecorporate <strong>in</strong>come tax revenues (IRPC). These revenues <strong>in</strong>creased from five percent of the total <strong>in</strong>2003 to 15 percent <strong>in</strong> 2008. In addition to the effects of the new tax regime and improvements <strong>in</strong>adm<strong>in</strong>istration, this rapid growth <strong>in</strong> company tax revenue is likely related to strong growth <strong>in</strong> thef<strong>in</strong>ancial sector, which doubled its share <strong>in</strong> GDP from 1.6 percent <strong>in</strong> 1999, to 3.3 percent <strong>in</strong> 2003and cont<strong>in</strong>ued ris<strong>in</strong>g to 5.6 percent of GDP <strong>in</strong> 2008. Expansion of the f<strong>in</strong>ancial sector wasparticularly strong <strong>in</strong> 2004 and 2005, with growth rates of 25.2 and 49.2 percent, respectively (seethe data on GDP <strong>in</strong> Appendix A). 37 The transport and communication sector is also likely to havebeen <strong>in</strong>strument here, given its high average growth rate of 13.1 percent over the period 2006 to2008 and share of around ten percent <strong>in</strong> GDP.37 The manufactur<strong>in</strong>g sector also displayed a higher accumulated growth than most other sectors over theperiod 1999 to 2008, but this is ma<strong>in</strong>ly due to megaprojects, and Mozal <strong>in</strong> particular, which contributesdisproportionately little to government revenue due to the special negotiated tax regime for this project.


R EVENUE P ERFORMANCE 41In contrast, personal <strong>in</strong>come tax receipts received a one-off <strong>in</strong>crease <strong>in</strong> revenues with the<strong>in</strong>troduction of IRPS <strong>in</strong> 2003 and the <strong>in</strong>clusion of public sector salaries <strong>in</strong>to the tax-net, but haves<strong>in</strong>ce rema<strong>in</strong>ed around 15 percent level of total revenue, even decl<strong>in</strong><strong>in</strong>g as a share of totalrevenues <strong>in</strong> 2005 and 2006.Figure 4-2Revenues from <strong>Tax</strong>es on Income as a Share of Total Domestic RevenuesSOURCE: MPD, "Quadromacro Revisto CFMP Proposta", received June 2009.As a result, <strong>in</strong>come tax revenue <strong>in</strong> 2008 was almost equally shared between companies and<strong>in</strong>dividuals (15.0 percent and 15.6 percent of total revenues, respectively). While the currentgrowth slowdown may dampen company profits <strong>in</strong> the short run, the overall trend suggests thatcompany tax revenues should overtake personal <strong>in</strong>come tax revenues <strong>in</strong> com<strong>in</strong>g years. This trendsuggests that efforts to strengthen tax adm<strong>in</strong>istration have had a disproportionate impact oncompany tax collects, as would be expected given that the <strong>in</strong>dividual <strong>in</strong>come tax is largely paidthrough monthly withhold<strong>in</strong>g by employers <strong>in</strong> the formal sector. The trend is likely to be furtherhelped by the new fiscal benefits code, discussed above, which reduces the level of benefits tofuture mega-projects and m<strong>in</strong><strong>in</strong>g and petroleum <strong>in</strong>vestments, br<strong>in</strong>g<strong>in</strong>g these under the IRPC(though the revenue potential from large projects is generally very low <strong>in</strong> the <strong>in</strong>itial years ofoperation due to deductions for capital expenditures and loss carry-forward provisions of the taxcode).Other Revenue SourcesAn important consideration is the dist<strong>in</strong>ction between total domestic revenues and tax revenues.As illustrated <strong>in</strong> Figure 4.1 the gap between the two revenue figures has <strong>in</strong>creased s<strong>in</strong>ce 2004,with total revenues broadly <strong>in</strong> l<strong>in</strong>e with the <strong>PARPA</strong> <strong>II</strong> target even though tax revenues are belowtarget. A part of the <strong>in</strong>crease <strong>in</strong> this gap is due to the reclassification of a portion of fuel tax from“Other <strong>Tax</strong>es” to non-tax “Consigned Revenue” as of 2004. This statistical artifact <strong>in</strong>creased“Other Revenues” by 0.8 percent of GDP. Even exclud<strong>in</strong>g this reclassification effect, however,the gap between total revenues and tax revenues has still <strong>in</strong>creased due to growth <strong>in</strong> “Own


42 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUERevenues” (Receitas Proprias) of central and district government agencies, from zero percent ofGDP <strong>in</strong> 2004 to 0.5 percent s<strong>in</strong>ce 2005, and also an <strong>in</strong>crease <strong>in</strong> capital revenues from 0 percent ofGDP <strong>in</strong> 2004 to 0.7 percent <strong>in</strong> 2005, decl<strong>in</strong><strong>in</strong>g to 0.4 percent of GDP <strong>in</strong> 2008. As privatizationrevenues are now mostly exhausted, capital revenues are ma<strong>in</strong>ly derived from dividends earnedon government <strong>in</strong>vestments state-owned corporations and government shares <strong>in</strong> private ventures.These might be expected to rise further as government stakes <strong>in</strong> Mozal and other mega-projectsbear more fruit.F<strong>in</strong>ally, some commentators contend that tax cuts may <strong>in</strong>crease revenue by stimulat<strong>in</strong>g highercompliance and faster growth. There is no empirical evidence, however, to support theproposition that a moderate reduction <strong>in</strong> the tax rates would have a large enough impact on<strong>in</strong>vestment, productivity, or compliance to avert a revenue loss. One possible exception,mentioned <strong>in</strong> Chapter 4, is the extremely high tax on certa<strong>in</strong> products like vehicles from thecomb<strong>in</strong>ed effect of import duties, VAT, and excise tax. Experience <strong>in</strong> other countries suggeststhat lower<strong>in</strong>g the excise tax or the import duty <strong>in</strong> such cases can actually boost revenue. Thisoption warrants study.INTERNATIONAL COMPARISONSTo ga<strong>in</strong> perspective, it is <strong>in</strong>structive to compare the revenue performance <strong>in</strong> <strong>Mozambique</strong> withthat of other countries. In fact, <strong>in</strong>ternational comparisons were frequently mentioned <strong>in</strong> <strong>in</strong>terviewsconducted for this report, particularly <strong>in</strong> connection with discussions of an appropriate revenuetarget. <strong>Mozambique</strong> currently has a revenue ratio that is one of the lowest of the 15 SADCmember states, and well below the SADC median of 23.3 percent of GDP (averaged over theperiod 2006 to 2008). Closer exam<strong>in</strong>ation, however, shows that this is not an appropriatecomparison. Several member states benefit from hav<strong>in</strong>g abundant natural resources (Angola,Botswana, Namibia, South Africa), while some have access to extraord<strong>in</strong>ary revenues throughallocations from the SACU common revenue pool (Botswana, Lesotho, Namibia, Swaziland), andfive are irrelevant to <strong>Mozambique</strong> due to very different economic and political conditions(Mauritius, Seychelles and South Africa as much wealthier countries, and DRC and Zimbabwe ashighly disrupted economies). This leaves only Malawi, Zambia, Tanzania, and Madagascar asrelevant comparators with<strong>in</strong> the SADC region.Table 4-5 presents fiscal <strong>in</strong>dicators for SADC countries and some additional comparators.Malawi and Zambia have revenue ratios that are higher than <strong>in</strong> <strong>Mozambique</strong>, at 18.7 percent and18.2 percent of GDP, respectively, while Madagascar and Tanzania have revenue sharesconsiderably lower than <strong>in</strong> <strong>Mozambique</strong>, at 12.1 and 13.2 percent of GDP, respectively. Notably,Malawi and Zambia also score better than <strong>Mozambique</strong> <strong>in</strong> the World Bank’s Do<strong>in</strong>g Bus<strong>in</strong>essrank<strong>in</strong>g for “ease of pay<strong>in</strong>g taxes” (as discussed <strong>in</strong> Chapter 3). Although not <strong>in</strong> SADC, anotherstrong reformer <strong>in</strong> the region is Uganda, which also has a revenue ratio well below that of<strong>Mozambique</strong> at 12.7 percent of GDP. This basic comparison might lead to the conclusion thatthat <strong>Mozambique</strong> is well placed, given its level of development and economic structure, and <strong>in</strong>particular given the cont<strong>in</strong>u<strong>in</strong>g level of adm<strong>in</strong>istrative reforms underway.Accord<strong>in</strong>g to World Development Indicators (WDI) data from the World Bank for 2008, onlythree low-<strong>in</strong>come countries managed to generate revenue ratios above 19 percent of GDP <strong>in</strong>2006. These are Ghana (21.9 percent), Kenya (21.1 percent) and Vietnam (27.9 percent). For


R EVENUE P ERFORMANCE 432007, this group also <strong>in</strong>cludes Senegal at 21.1 percent of GDP. However, all of these countriesdiffer considerably from <strong>Mozambique</strong> <strong>in</strong> economic conditions. All four have much higher levelsof per capita <strong>in</strong>come. Measured <strong>in</strong> terms of purchas<strong>in</strong>g power parity, the respective <strong>in</strong>come levels<strong>in</strong> 2007 were $2,602 for Vietnam, $1,424 for Ghana, $1,672 for Kenya, and $1,698 for Senegal,compared to $842 for <strong>Mozambique</strong> (IMF, 2009b). These figures are only illustrative, but<strong>Mozambique</strong> clearly lags far beh<strong>in</strong>d the high-revenue countries of the low <strong>in</strong>come group <strong>in</strong> termsof overall development. In addition, Vietnam benefits from hav<strong>in</strong>g substantial amounts of oilrevenue.Table 4-5Fiscal Indicators, Three-year Averages, 2006-2008 (% GDP)Overall Fiscal BalanceInclud<strong>in</strong>ggrantsExclud<strong>in</strong>ggrantsGovernmentRevenue(exclud<strong>in</strong>ggrants)GDP percapita(currentUSD)SADC Mean 2.5 -1.7 28.0 3,106SADC Median 0.6 -4.85 23.3 1876SADC C OUNTRIESAngola 12.9 12.9 46.9 3,813Botswana 4.6 3.9 35.7 7,011Congo, Dem Rep -0.7 -5.8 15.4 165Lesotho 11.8 10.3 59.8 666Madagascar 10.6 -8.3 12.1 373Malawi -2.4 -5.2 18.7 273Mauritius -4.3 -4.5 20 5,803<strong>Mozambique</strong> -3.7 -13.5 15.8 408Namibia 1 0.8 30.3 4,146Seychelles -4.4 -6.2 37.8 10,943South Africa 0.2 0.2 26.6 5,686Swaziland 5.8 5.3 41 2,750Tanzania -2.8 -8.1 13.2 442Zambia 5.7 -5.8 18.2 1,002COMPARATOR COUNTRIESGhana -9.7 -15 22.5 665Kenya -3.1 -4.3 21.8 766Senegal -4.6 -6.7 20.1 925Uganda -1 -5.2 12.7 391Vietnam n/a -0.9 27.3 866Note: Data on Zimbabwe as a SADC country is excluded as unreliable.SOURCE: Sub-Saharan Africa Regional Economic Outlook, April 2009; World Economic Outlook April 2009


44 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEIn discuss<strong>in</strong>g a suitable revenue target for <strong>Mozambique</strong>, Varsano et al. (2006) draw attention tothe structure of the Mozambican economy. In particular, they highlight the importance of thecommerce, transport and communications sectors which were responsible for 36.9 percent ofGDP <strong>in</strong> 2004, higher than <strong>in</strong> Tanzania, Kenya, Uganda and Malawi. As they po<strong>in</strong>t out, thesesectors should provide relatively good sources of VAT revenue. They also po<strong>in</strong>t out that that<strong>Mozambique</strong> and Kenya have a similar share of manufactur<strong>in</strong>g <strong>in</strong> GDP (13.7 and 13.0 percent ofGDP, respectively) and conclude that their revenue capacity should also be similar, at around 22percent of GDP. This comparison ignores the fact that the <strong>Mozambique</strong> is considerably poorerand less developed than Kenya. Overall, these <strong>in</strong>ternational benchmark comparisons suggest thata revenue target above 19 percent of GDP would be highly ambitious for <strong>Mozambique</strong>.Nonetheless, simply compar<strong>in</strong>g revenue ratios across countries ignores a wide range of factorsrelat<strong>in</strong>g to taxable capacity and demand for tax revenue. The follow<strong>in</strong>g section presents furtheranalysis relat<strong>in</strong>g to the appropriate target which takes economic characteristics moresystematically <strong>in</strong>to account.REVENUE EFFORTRevenue performance can be analyzed <strong>in</strong> terms of tax effort or revenue effort by compar<strong>in</strong>g acountry’s revenue yield with a predicted value of the revenue ratio as estimated from a crosscountryregression analysis relat<strong>in</strong>g revenue ratios to economic and other characteristics. Thisstatistical form of benchmark<strong>in</strong>g takes <strong>in</strong>to account differences <strong>in</strong> economic characteristics likelyto <strong>in</strong>fluence potential revenues.Jones (2009) carried out this exercise for <strong>Mozambique</strong> by compar<strong>in</strong>g tax ratios across low andmiddle-<strong>in</strong>come countries for the period 1990 to 2003 <strong>in</strong> a panel analysis. He controls statisticallyfor differences <strong>in</strong> economic structure and <strong>in</strong>stitutional factors, the implication be<strong>in</strong>g thatrema<strong>in</strong><strong>in</strong>g differences <strong>in</strong> the observed tax ratios relate to differences <strong>in</strong> policy preferences and thequality of tax adm<strong>in</strong>istration. 38 His econometric results suggest that 13 percent of GDP was arealistic tax ratio for <strong>Mozambique</strong>, circa 2003. He also shows that the actual tax ratio over thisperiod closely followed predicted values, given the prevail<strong>in</strong>g economic and <strong>in</strong>stitutionalconditions.For the present paper, a similar analysis was conducted us<strong>in</strong>g more recent data, and with the ratioof total revenues to GDP as the dependent variable. To capture basic characteristics of the countrythat affect revenue performance, the <strong>in</strong>dependent variables <strong>in</strong> the regression analysis <strong>in</strong>clude GDPper capita, measured <strong>in</strong> constant US$2,000; the annual consumer <strong>in</strong>flation rate; the trade share ofGDP (def<strong>in</strong>ed as exports plus imports divided by GDP); the nonagricultural share of GDP; andthree widely cited measures from the World Bank on <strong>in</strong>stitutional quality, cover<strong>in</strong>g government38 Economic structure variables <strong>in</strong>clude GDP per capita, imports as a share of GDP, <strong>in</strong>dustry as a share ofGDP, whether or not the country is resource-rich; while <strong>in</strong>stitutions are controlled for <strong>in</strong>directly by<strong>in</strong>troduc<strong>in</strong>g the share of the country found <strong>in</strong> the tropics, whether or not the country is landlocked, and whothe colonial power, all variables variously associated with <strong>in</strong>stitutions <strong>in</strong> the cross-country empiricalliterature on growth.


R EVENUE P ERFORMANCE 45effectiveness, regulatory control and control of corruption, all of which are <strong>in</strong>dices rang<strong>in</strong>gbetween -2.5 to 2.5. 39The regression equation is estimated for low and middle-<strong>in</strong>come countries. To verify robustnessof the estimates, the regression is run for three periods: 1999-2007, 2003-2006, and 2005-2007.Table 4-6 presents the results. In particular, it shows the actual ratio of revenues to GDP and thepredicted ratio, as estimated from the regression equations. Assum<strong>in</strong>g that the predicted revenueshare <strong>in</strong>dicates the expected revenue yield for a country with <strong>Mozambique</strong>’s level of developmentand structural characteristics, the revenue effort can be def<strong>in</strong>ed as the ratio of the actual topredicted revenue ratios. This is reported <strong>in</strong> the f<strong>in</strong>al column of the Table 4-6.Table 4-6<strong>Mozambique</strong> Actual & Predicted Revenue-Ratio & Revenue-EffortPeriodActualR/YPredictedR/YS.E. Normal Range Revenue-Effort99-07 14.28 15.17 1.30 14.30 16.05 0.9403-06 13.85 16.49 1.45 15.51 17.47 0.8405-07 15.03 17.04 1.67 15.91 18.16 0.88Note: Normal range is def<strong>in</strong>ed as the (predicted T/Y)+/-0.67*S.E., which provides the 50 percentile range around the predictedvalue. The S.E. is the standard error of the predicted tax ratio.SOURCE: Authors’ estimation us<strong>in</strong>g WDI and WEO data as described <strong>in</strong> the text.Table 4-6 also shows the standard error of estimate for the predicted value of the revenue ratio,and a correspond<strong>in</strong>g “normal range” around the predicted value, tak<strong>in</strong>g <strong>in</strong>to account the fact thatthe expected value is estimate with a marg<strong>in</strong> of error. Specifically, the “normal range” is def<strong>in</strong>edhere such that 50 percent of the observations should fall <strong>in</strong>side this <strong>in</strong>terval, with 25 percentabove and 25 percent below. 40The most important po<strong>in</strong>t to note <strong>in</strong> Table 4-6 is that the actual ratio of revenue to GDP is belowthe predicted value, imply<strong>in</strong>g that <strong>Mozambique</strong> is consistently collect<strong>in</strong>g less revenue than wouldbe expected given its economic structure and other characteristics, based on evidence from othercountries. This is reflected <strong>in</strong> the revenue-effort measures, which are all below one (where arevenue effort of one would imply that <strong>Mozambique</strong> was achiev<strong>in</strong>g its predicted ratio). For themost recent period, the revenue ratio is even outside the normal range on the low side. This resultadds weight to the view that there is considerable scope for improvements <strong>in</strong> the revenue effort,and that a target ratio of slightly more than 18 percent should be achievable.In contrast to earlier econometric studies which provided evidence for a very low revenue targetof around 13 percent of GDP (Jones 2009) or a very high tax ratio of around 22 percent of GDP(Varsano et al., 2006), the message to emerge from the present analysis is that an appropriate39 See Appendix F for a complete discussion of the methodology and other studies of this sort.40 The “Normal Range” is calculated as Y+/-(0.67428*S.E.), where Y is the predicted tax ratio, 0.67428is the Z-score from the normal distribution associated with the middle 50 percentiles, and S.E. is thestandard error of the predicted tax ratio.


46 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUErevenue ratio for <strong>Mozambique</strong> lies between those extremes, most likely between 16 and 18percent of GDP.


5. Impact of the <strong>Tax</strong> <strong>System</strong>Previous chapters presented an overview of the structure and recent performance of the taxsystem. This chapter provides a more detailed analysis of the impact the tax system on majorobjectives of <strong>PARPA</strong> <strong>II</strong>. The chapter assesses <strong>in</strong> broad terms the impact of the tax system on<strong>in</strong>vestment, sav<strong>in</strong>g and employment, on private sector development, and on fairness and equity.This analysis leads <strong>in</strong>to to a discussion of major tax issues for consideration <strong>in</strong> the preparation of<strong>PARPA</strong> <strong>II</strong>I, which is the subject of Chapter 6.The discussion of tax impacts on the <strong>PARPA</strong> objectives parallels <strong>in</strong> many ways a review of basicpr<strong>in</strong>ciples of taxation. The fundamental purpose of taxation, of course, is to raise revenue tof<strong>in</strong>ance the provision of public goods and services. The first pr<strong>in</strong>ciple is therefore that the taxsystem should be effective <strong>in</strong> mobiliz<strong>in</strong>g and susta<strong>in</strong><strong>in</strong>g revenue. But taxes have a pervasive<strong>in</strong>fluence on economic decisions of <strong>in</strong>dividuals and bus<strong>in</strong>esses, and on social equity. In view ofthese effects, the tax system should achieve the appropriate level of revenue as efficiently andfairly as possible. Hence, a well designed tax system should also conform to three otherpr<strong>in</strong>ciples:• Economic efficiency. An efficient tax system m<strong>in</strong>imizes tax-driven distortions ofeconomic behavior <strong>in</strong> order to foster productivity and economic growth. Economicefficiency considerations are especially important for low-<strong>in</strong>come countries, which canleast afford the cost of avoidable resource misallocation. Efficiency effects are alsorelated to predictability. A tax regime that is subject to unexpected changes or arbitraryenforcement is a major risk factor for <strong>in</strong>vestors.• Equity. A fair tax system is characterized by vertical equity (collect<strong>in</strong>g proportionatelymore from those with higher <strong>in</strong>comes) and horizontal equity (provides relatively uniformand non-discrim<strong>in</strong>atory treatment of taxpayers with similar economic circumstances). Italso m<strong>in</strong>imizes the tax burden on the poor, and avoids excessive tax burdens or arbitraryimpositions all around. Fairness is a fundamental objective <strong>in</strong> its own right. In addition,perceptions of unfairness can erode compliance and underm<strong>in</strong>e the susta<strong>in</strong>ability of thetax system.• Adm<strong>in</strong>istrative efficiency. The tax system should also be adm<strong>in</strong>istered efficiently, withdue regard to both the direct costs of collect<strong>in</strong>g taxes and compliance costs imposed ontaxpayers. Even the best tax code produces poor results when it is not well adm<strong>in</strong>istered.In countries with limited <strong>in</strong>stitutional capacity and weak capacity on the part of taxpayersto deal with complexities of f<strong>in</strong>ancial management, simplicity is a virtue.


48 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEThese pr<strong>in</strong>ciples are implicitly related to the <strong>PARPA</strong> <strong>II</strong> objective of rais<strong>in</strong>g governmentrevenues without <strong>in</strong>creas<strong>in</strong>g the tax burden on the formal sector, by expand<strong>in</strong>g the tax base,curb<strong>in</strong>g tax evasion, and reduc<strong>in</strong>g tax <strong>in</strong>centives.INVESTMENTThis section exam<strong>in</strong>es the effects of the tax system on two central build<strong>in</strong>g blocks for susta<strong>in</strong>ablegrowth: the level and quality of <strong>in</strong>vestment.How <strong>Tax</strong>es Affect Investment: The BasicsFigure 5-1 illustrates the general relationship between taxes and <strong>in</strong>vestment decisions. L<strong>in</strong>e AArepresents the returns to capital. The downward slope shows that there is a cont<strong>in</strong>uum of<strong>in</strong>vestments rang<strong>in</strong>g from high to low returns. L<strong>in</strong>e MC 0 represents the “user cost of capital”(UCC), which <strong>in</strong>cludes cost of funds, the effective tax rate on <strong>in</strong>come from capital, and theimpact of <strong>in</strong>direct taxes on the acquisition cost of capital goods. A reduction <strong>in</strong> any of the taxcomponents shifts the UCC downward from MC 0 to MC 1 , and stimulates an <strong>in</strong>crease <strong>in</strong><strong>in</strong>vestment from INV 0 to INV 1 .Figure 5-1Lower<strong>in</strong>g <strong>Tax</strong>es on Capital to Stimulate InvestmentMarg<strong>in</strong>alproduct &marg<strong>in</strong>alcost ofcapitalBAValue of the marg<strong>in</strong>alproduct of capitalMC 0MC 1<strong>Tax</strong> ReductionABINV 0INV 1InvestmentAA = Strong <strong>in</strong>vestment climate strong effectBB = Weak <strong>in</strong>vestment climate weak effectThis picture illustrates two important po<strong>in</strong>ts. First, tax cuts have little effect on <strong>in</strong>vestmentprojects with <strong>in</strong>herently high rates of return (along the upper segment of AA) or <strong>in</strong>herently lowrates of return (along the lower segment of AA). <strong>Tax</strong> considerations ma<strong>in</strong>ly affect projects that


I MPACT OF T AX S YSTEM 49are marg<strong>in</strong>ally viable (<strong>in</strong> the mid-range of AA). This selective impact matters because economicgrowth depends on quality of <strong>in</strong>vestment as well as the quantity. 41Second, attract<strong>in</strong>g <strong>in</strong>vestment with fundamentally high rates of return requires policies to improvethe bus<strong>in</strong>ess environment and hence shift the entire l<strong>in</strong>e AA to the right. The <strong>in</strong>vestment climatealso has a strong effect on the amount of new <strong>in</strong>vestment that will occur as a result of lower<strong>in</strong>gthe cost of capital through tax breaks. With favorable bus<strong>in</strong>ess conditions a reduction <strong>in</strong> the costof capital may <strong>in</strong>duce a large <strong>in</strong>vestment response (as along AA), but the same tax <strong>in</strong>centives may<strong>in</strong>duce very little new <strong>in</strong>vestment if bus<strong>in</strong>ess conditions are poor (as along BB). Overall, non-taxconsiderations are likely to be of paramount importance for most <strong>in</strong>vestment decisions.Various tax tools can be used to stimulate <strong>in</strong>vestment by reduc<strong>in</strong>g the cost of capital: cutt<strong>in</strong>g thestandard company tax rate; reduc<strong>in</strong>g the double taxation of dividend <strong>in</strong>come; offer<strong>in</strong>g special tax<strong>in</strong>centives to selected <strong>in</strong>vestments; provid<strong>in</strong>g more generous <strong>in</strong>vestment deductions or credits <strong>in</strong>the calculation of taxable <strong>in</strong>come; or reduc<strong>in</strong>g <strong>in</strong>direct taxes on the acquisition of capital goods.Many tax experts regard the latter two approaches as most cost-effective <strong>in</strong> terms of balanc<strong>in</strong>g theimpact on <strong>in</strong>vestment aga<strong>in</strong>st the revenue loss from special <strong>in</strong>centives. 42 The government’sdecision <strong>in</strong> 2002 to reform the Code of Fiscal Benefits was consistent with this view, as itelim<strong>in</strong>ated most tax holidays <strong>in</strong> favor of <strong>in</strong>vestment allowances, accelerated depreciation, and<strong>in</strong>vestment tax credits, on top of the exemption from import duty on capital goods.<strong>Tax</strong>es and Investment <strong>in</strong> <strong>Mozambique</strong>Dur<strong>in</strong>g the <strong>PARPA</strong> <strong>II</strong> period, Gross Domestic Investment ranged between 18 and 19 percent ofGDP, accord<strong>in</strong>g to figures reported by the IMF. The latest actual data po<strong>in</strong>t (as dist<strong>in</strong>ct fromestimates) is 18.0 percent for 2007, of which two-thirds comes from government <strong>in</strong>vestment andjust one-third, or 6 percent, from non-government <strong>in</strong>vestment. Private <strong>in</strong>vestment has fluctuatedbetween 6 and 13 percent of GDP s<strong>in</strong>ce 2004, with variations driven by the tim<strong>in</strong>g of megaprojects.These <strong>in</strong>vestment rates are exceed<strong>in</strong>gly low, to the po<strong>in</strong>t of be<strong>in</strong>g seriously <strong>in</strong>consistentwith the objective of susta<strong>in</strong><strong>in</strong>g rapid economic growth. 43Is the low level of <strong>in</strong>vestment a sign that tax rates <strong>in</strong> <strong>Mozambique</strong> are too high? Several studiesundertaken prior to <strong>PARPA</strong> <strong>II</strong> found that company tax rates <strong>in</strong> <strong>Mozambique</strong> have been broadly <strong>in</strong>l<strong>in</strong>e with regional and <strong>in</strong>ternational norms (see Bolnick 2004a; FIAS 2006; and IMF 2006). 44 Asseen <strong>in</strong> Chapter 3, <strong>Mozambique</strong> also scores well on the World Bank’s Do<strong>in</strong>g Bus<strong>in</strong>ess estimate of41 A 10 percent <strong>in</strong>crease <strong>in</strong> productivity delivers the same boost to growth as a 10 percent <strong>in</strong>crease <strong>in</strong> theamount <strong>in</strong>vested Hence, a policy that adds 10 percent to the amount <strong>in</strong>vested while reduc<strong>in</strong>g <strong>in</strong>vestmentefficiency by 10 percent adds noth<strong>in</strong>g to the growth rate.42 See Zee, Stotsky and Ley (2002) and Bolnick (2004, chapter 5)43 The National Statistics Institute (INE) may systematically underestimate <strong>in</strong>vestment <strong>in</strong> farm works,rural build<strong>in</strong>gs, and mach<strong>in</strong>ery and equipment used by small enterprises, though it does obta<strong>in</strong> reasonablygood data on public sector capital expenditures, FDI <strong>in</strong>flows, imported mach<strong>in</strong>ery and equipment, andmajor construction projects.44 <strong>Mozambique</strong> also scores well on the World Bank’s Do<strong>in</strong>g Bus<strong>in</strong>ess estimate for the Total <strong>Tax</strong> Rate oncompany profits, but the methodology beh<strong>in</strong>d this measure is seriously flawed, as seen <strong>in</strong> Chapter 3.


50 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEthe Total <strong>Tax</strong> Rate on company profits (though the methodology beh<strong>in</strong>d this measure is seriouslyflawed, as discussed earlier).Figure 5-2 compares company tax rates <strong>in</strong> the SADC region, based on the latest available<strong>in</strong>formation for each country. The SADC average of 30.3 percent is now slightly lower than the32 percent rate <strong>in</strong> <strong>Mozambique</strong>. Five years ago, the regional average was slightly higher than therate <strong>in</strong> <strong>Mozambique</strong>, at 32.5 percent (Bolnick 2004). In the <strong>in</strong>terim, the SADC average decl<strong>in</strong>eddue to the entry of Madagascar, with a 24 percent tax rate, as well as company tax reductions <strong>in</strong>South Africa (from 30 to 28 percent), Lesotho (from 35 to 25 percent) and Mauritius (from 25 to15 percent). In Mauritius, the reduction <strong>in</strong> tax rates was part of a move to a “flat tax,” which alsoelim<strong>in</strong>ated special tax <strong>in</strong>centives and discretionary tax breaks. 45 These tax cuts reflect a globaltrend towards lower taxes on capital. With the advent of the SADC free trade area, taxdifferentials among member states will become more prom<strong>in</strong>ent <strong>in</strong> affect<strong>in</strong>g location decisions, asmore <strong>in</strong>vestments will be driven by a regional perspective.Figure 5-2Company <strong>Tax</strong> Rates <strong>in</strong> the SADC Region, 2009Standard Company <strong>Tax</strong> RatesSADC AverageMAURITIUSMADAGASCARLESOTHOBOTSWANASOUTH AFRICAZIMBABWETANZANIASWAZILANDMALAWIMOZAMBIQUEZAMBIANAMIBIAANGOLASEYCHELLESDR CONGO0% 10% 20% 30% 40% 50%SOURCE: Internet <strong>in</strong>formation accessed between July 27 and August 6, 2009 from 2009 <strong>Tax</strong> Highlights produced by Deloitte offices<strong>in</strong> most SADC countries, plus: for Lesotho, Southern Africa Regional Poverty Network and US Embassy; for Mauritius, PKFInternational Ltd. Mauritius <strong>Tax</strong> guide 2009; for Seychelles, Revenue Commission.The studies cited above by Bolnick and by FIAS also showed that <strong>in</strong>vestments subject to thestandard tax regime faced an extremely high marg<strong>in</strong>al effective tax rate (METR) <strong>in</strong> both absoluteterms and relative to regional norms, to the po<strong>in</strong>t of be<strong>in</strong>g a severe impediment to <strong>in</strong>vestment.The pr<strong>in</strong>cipal cause of the elevated METR was the double taxation of dividends paid to <strong>in</strong>dividualshareholders. (Investments with<strong>in</strong> a corporate group would not face this high METR, s<strong>in</strong>cedividend payments <strong>in</strong> this case are exempt from double taxation, subject to conditions of the45 Bolnick (2009) expla<strong>in</strong>s the flat tax concept. He concludes that while <strong>Mozambique</strong> can benefit frommeasures to broaden the tax base and reduce tax rates, the flat tax as such is but one possible means to thisend.


I MPACT OF T AX S YSTEM 51IRPC.) Figure 5-3 shows, for SADC member states, the comb<strong>in</strong>ed effect of the company <strong>in</strong>cometax and withhold<strong>in</strong>g tax on dividend payments to <strong>in</strong>dividual shareholders. 46 For both residentsand non-residents as shareholders, the comb<strong>in</strong>ed tax <strong>in</strong> <strong>Mozambique</strong> is far higher than theregional average. Reduc<strong>in</strong>g the double taxation of dividend <strong>in</strong>come earned by <strong>in</strong>dividualshareholders would therefore remove a significant barrier to <strong>in</strong>vestments that do not enjoy fiscal<strong>in</strong>centives and are not part of a larger corporate group. As a by-product, this measure would alsoelim<strong>in</strong>ate a serious dis<strong>in</strong>centive for domestic enterprises to <strong>in</strong>corporate.For <strong>in</strong>vestment covered by the Code of Fiscal Benefits (CFB), the same METR studies showedthat the effective tax rates <strong>in</strong> <strong>Mozambique</strong> were very competitive under the 2002 Code. The 2009Code has not been subject to detailed analysis, but it clearly offers even more attractive taxbenefits for <strong>in</strong>vestments <strong>in</strong> agriculture and fisheries, public <strong>in</strong>frastructure, and science andtechnology parks, as well as projects <strong>in</strong> Industry Free Zones (IFZs) and the Special EconomicZone (SEE) planned for Nacala. As discussed <strong>in</strong> Chapter 3, the 2009 Code also curtails benefitsfor large projects. This change is warranted because the tax breaks for large projects proved to bevery costly, and <strong>in</strong> any case <strong>in</strong>ternationally mobile <strong>in</strong>vestments, for which tax competition wouldbe a major factor, can still qualify for special treatment under the 2009 CFB provisions forIndustrial Free Zones or the planned Special Economic Zone <strong>in</strong> Nacala. 47In conclusion, the tax burden is very high for <strong>in</strong>vestments subject to double taxation. For other<strong>in</strong>vestments the level of tax rates should not be a serious problem – though it a concern to see thatthe standard tax rate <strong>in</strong> <strong>Mozambique</strong> is now above the regional average. Unfortunately, there isno empirical basis for estimat<strong>in</strong>g the extent to which lower tax rates would stimulate additional<strong>in</strong>vestment <strong>in</strong> a country like <strong>Mozambique</strong>. Given the prom<strong>in</strong>ence of other barriers to <strong>in</strong>vestment,one would not expect a large impact from a moderate tax reduction. Lower tax rates would,however, <strong>in</strong>crease the capacity of exist<strong>in</strong>g firms to f<strong>in</strong>ance expansion plans or <strong>in</strong>vestments toimprove productivity, by boost<strong>in</strong>g net earn<strong>in</strong>gs and cash flow. Empirical studies for othercountries show that lower taxes do have a positive overall effect on <strong>in</strong>vestment (Klemm 2009).The quality of tax adm<strong>in</strong>istration, too, has an important effect on the <strong>in</strong>vestment climate,<strong>in</strong>dependent of the level of tax rates. In the <strong>in</strong>terviews for this study, grievances about arbitraryand punitive enforcement practices by tax officials were far more prevalent than compla<strong>in</strong>ts aboutthe level of taxes. These compla<strong>in</strong>ts undoubtedly reach the ears of many potential <strong>in</strong>vestors.Hence, problems with tax adm<strong>in</strong>istration take precedence over concerns about the level of taxrates. Besides, improvements <strong>in</strong> tax adm<strong>in</strong>istration can provide the revenue leverage needed tosupport reductions <strong>in</strong> the tax rates.46 The calculations here assume that after-tax earn<strong>in</strong>gs at the company level are distributed <strong>in</strong> full toshareholders. Us<strong>in</strong>g some other illustrative assumption would alter the absolute values of the comb<strong>in</strong>ed tax,but would not change the relative comparisons.47 Klemm (2009) f<strong>in</strong>ds that <strong>in</strong>ternational tax competition is the most prom<strong>in</strong>ent reason for countries tooffer tax <strong>in</strong>centives. Appendix 7 summarizes the on-go<strong>in</strong>g debate about tax <strong>in</strong>centives as a tool for<strong>in</strong>vestment promotion.


52 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEFigure 5-3Company <strong>Tax</strong> + Dividend Withhold<strong>in</strong>g <strong>Tax</strong> <strong>in</strong> the SADC Region, 2009(a) Resident ShareholdersSADC AverageMAURITIUSMADAGASCARSOUTH AFRICANAMIBIALESOTHOBOTSWANATANZANIASWAZILANDMALAWISEYCHELLESANGOLAZIMBABWEZAMBIAMOZAMBIQUEDR CONGO0% 10% 20% 30% 40% 50% 60%(b) Non-Resident ShareholdersSADC AverageMAURITIUSMADAGASCARSOUTH AFRICABOTSWANATANZANIAMALAWISWAZILANDNAMIBIAANGOLALESOTHOZIMBABWEZAMBIAMOZAMBIQUESEYCHELLESDR CONGO0% 10% 20% 30% 40% 50% 60%SOURCE: See note to Figure 5.2.The tax system also affects <strong>in</strong>vestment efficiency. While virtually all taxes create efficiencylosses by distort<strong>in</strong>g economic behavior, sharp differences <strong>in</strong> the METR by sector or by type of<strong>in</strong>vestment, as found by FIAS (2006), accentuate the misallocation of resources, to the detrimentof economic growth. To illustrate, an <strong>in</strong>vestment <strong>in</strong> agriculture that has a 15 percent pre-tax rateof return produces a better after-tax yield under the current tax system than an alternative<strong>in</strong>vestment <strong>in</strong> manufactur<strong>in</strong>g that would generate a 20 percent pre-tax rate of return. Yet the latter<strong>in</strong>vestment is more productive and contributes more to growth. The greater the tax differentials,the greater the efficiency loss for the economy through distortions of this k<strong>in</strong>d.


I MPACT OF T AX S YSTEM 53Another source of <strong>in</strong>efficiency arises when the provision of special fiscal benefits creates arevenue loss that necessitates higher tax rates on other bus<strong>in</strong>ess activities. 48 Also, tax concessionsthat reduce the price of imported goods, such as the exemption from import duty and VAT oncategory “K” imports for <strong>in</strong>vestments approved by CPI, create a strong <strong>in</strong>centive favor<strong>in</strong>g the useof imports rather than the establishment of l<strong>in</strong>kages to domestic suppliers. For these reason, andalso on grounds of equity, most economists favor more uniform application of broad-based taxesat moderate rates, rather than offer<strong>in</strong>g special tax breaks to some <strong>in</strong>vestors while impos<strong>in</strong>g highertax rates on others.EMPLOYMENTThe tax system affects both the supply and demand side of the labor market, as well as thedevelopment of labor force skills. Start<strong>in</strong>g on the supply side, taxes on labor <strong>in</strong>come – both thepersonal <strong>in</strong>come tax and payroll tax for social <strong>in</strong>surance (INSS) – create a “tax wedge” betweenthe cost of labor to employers and the <strong>in</strong>come received by the workers, at least <strong>in</strong> the formalsector. Economic models <strong>in</strong>dicate that market <strong>in</strong>teractions tend to shift the effective burden ofthese taxes onto the workers, <strong>in</strong> the form of lower take-home pay. In <strong>Mozambique</strong>, the tax wedgeis very low for most unskilled workers, but may be high enough to affect labor supply decisionsfor some workers at the higher end of the <strong>in</strong>come scale.One method for compar<strong>in</strong>g the tax wedge across countries is to exam<strong>in</strong>e tax thresholds <strong>in</strong> relationto per capita GDP. S<strong>in</strong>ce 2007 a worker <strong>in</strong> <strong>Mozambique</strong> enters the ISPC tax net at an <strong>in</strong>come of36 times the highest m<strong>in</strong>imum wage, which was approximately 7.6 times per capita GDP <strong>in</strong> 2009.Hence, most unskilled workers <strong>in</strong> <strong>Mozambique</strong> suffer no ISPC liability. By comparison, themedian value for this ratio among low-<strong>in</strong>come countries globally is 1.3, and for sub-SaharanAfrica, 0.9. 49 Lower numbers <strong>in</strong>dicate a lower threshold for enter<strong>in</strong>g the tax net, relative toaverage <strong>in</strong>come levels <strong>in</strong> each country.The INSS imposition <strong>in</strong> <strong>Mozambique</strong>, at 7 percent (with 3 percent paid by the employee and 4percent by the employer) is also on the low side, compared to the median of 10 percent for low<strong>in</strong>come countries, and 10.3 percent for sub-Saharan Africa. Both the high tax threshold and lowsocial <strong>in</strong>surance charge <strong>in</strong> <strong>Mozambique</strong> produce a low tax wedge for most workers. Given theabundance of job seekers for any unskilled labor position <strong>in</strong> the formal sector, it is safe to say thatthe tax system has no serious effect on the labor supply at this level. High <strong>in</strong>come workers,however, face a 32 percent marg<strong>in</strong>al tax rate plus INSS charges. This tax wedge is large enoughto impair <strong>in</strong>centives to work <strong>in</strong> the formal sector. While there is no empirical evidence to estimatethe size of this effect, stories of rampant <strong>in</strong>come tax evasion and resistance to formalizationsuggest that the tax system probably does affect<strong>in</strong>g labor market decisions for some members ofthe economically active population.48 A basic proposition <strong>in</strong> tax analysis is that the efficiency loss from taxation rises with the square of thetax rate. Hence, if tax breaks for A necessitate higher tax rates on B and C, the result is a net loss ofeconomic efficiency.49 Calculations by the study team, us<strong>in</strong>g the “Collect<strong>in</strong>g <strong>Tax</strong>es” data base from the USAID Fiscal Reformwebsite, at www.fiscalreform.net, accessed on July 2, 2009.


54 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUETurn<strong>in</strong>g to the demand side of the labor market, the impact of the tax system is determ<strong>in</strong>ed bythree elements: the effect on <strong>in</strong>vestment; on enterprise development and formalization; and on thecapital <strong>in</strong>tensity of bus<strong>in</strong>ess operations. The analysis presented <strong>in</strong> the Investment section suggeststhat the present tax system is not a significant deterrent to most types of <strong>in</strong>vestment. Nonetheless,lower tax rates would likely stimulate additional <strong>in</strong>vestment demand for labor by enhanc<strong>in</strong>greturns and <strong>in</strong>creas<strong>in</strong>g the flow of earn<strong>in</strong>gs as a source of f<strong>in</strong>anc<strong>in</strong>g for bus<strong>in</strong>ess expansion. At thesame time, other deficiencies <strong>in</strong> the bus<strong>in</strong>ess environment are much more important than taxes asbarriers to <strong>in</strong>vestment and job growth.A more serious concern is the impact of the tax system on bus<strong>in</strong>ess development, and <strong>in</strong>particular the growth of small and medium sized enterprises, which are the ma<strong>in</strong> eng<strong>in</strong>e ofemployment creation <strong>in</strong> most countries. As discussed below (see Private Sector Development),the complexity of the tax system, the lack of clear <strong>in</strong>formation on the tax code, and punitiveenforcement practices by tax officials serve to deter formalization. Furthermore, several taxtechnicalities of the tax system, such as withhold<strong>in</strong>g requirements and the denial of expenses dueto m<strong>in</strong>or errors <strong>in</strong> documentation have <strong>in</strong>hibited the development of l<strong>in</strong>kages between large andsmall enterprises, to the detriment of both. In addition, the difficult procedures and long lags<strong>in</strong>volved <strong>in</strong> obta<strong>in</strong><strong>in</strong>g VAT refunds greatly dim<strong>in</strong>ish the scope for small bus<strong>in</strong>ess entry <strong>in</strong>to exportmarkets. The implication is that tax simplification, improved customer services by the AT, andbetter tax <strong>in</strong>formation can, <strong>in</strong> pr<strong>in</strong>ciple, help to spur job growth.A less obvious factor is that the tax system also effects decisions about the capital <strong>in</strong>tensity orlabor <strong>in</strong>tensity of <strong>in</strong>vestment, and hence job creation. As an example, a survey of 60 recent<strong>in</strong>vestors <strong>in</strong> <strong>Mozambique</strong> found that projects depend<strong>in</strong>g critically on fiscal benefits were far morecapital <strong>in</strong>tensive than project that were viable <strong>in</strong> their own right without tax breaks (Bolnick2009a). Bus<strong>in</strong>ess plans for latter group envisioned five times as much job creation, even thoughtotal planned <strong>in</strong>vestment was far higher <strong>in</strong> the former group. More generally, tax breaks such asexemption from import duty and VAT on capital goods, <strong>in</strong>vestment tax credits, and accelerateddepreciation favor capital-<strong>in</strong>tensive projects. These tools may have some positive <strong>in</strong>fluence on thevolume of <strong>in</strong>vestment, but they have less pronounced effects on expand<strong>in</strong>g employmentopportunities for Mozambican workers.A f<strong>in</strong>al aspect of the employment issue is the effect of the tax system on workforce development.In most countries, legitimate tra<strong>in</strong><strong>in</strong>g expenses are recognized as a cost of do<strong>in</strong>g bus<strong>in</strong>ess. Indeed,some countries <strong>in</strong> the region, such as Botswana and Swaziland, allow a super-deduction of morethan 100 percent for tra<strong>in</strong><strong>in</strong>g expenses, to provide an effective subsidy for these activities. In<strong>Mozambique</strong>, Article 18 of the Code of Fiscal Benefits allows a deduction <strong>in</strong> calculat<strong>in</strong>g taxable<strong>in</strong>come for “<strong>in</strong>vestments <strong>in</strong> professional tra<strong>in</strong><strong>in</strong>g” only dur<strong>in</strong>g the first five years of operation,only for qualify<strong>in</strong>g <strong>in</strong>vestment projects, and only up to a maximum of 5 percent of taxable<strong>in</strong>come (10 percent for new technology). Consequently, many bus<strong>in</strong>esses are not allowed todeduct tra<strong>in</strong><strong>in</strong>g expenses, or can do so only to a limited extent. The effect is to discourage greater<strong>in</strong>vestment <strong>in</strong> tra<strong>in</strong><strong>in</strong>g, even though tra<strong>in</strong><strong>in</strong>g activities generate positive externalities for theeconomy through employee turnover, worker <strong>in</strong>teractions. This provision of the tax systemrequires reconsideration <strong>in</strong> the <strong>in</strong>terest of upgrad<strong>in</strong>g worker skills and improv<strong>in</strong>g laborproductivity. Higher productivity, <strong>in</strong> turn, contributes to an <strong>in</strong>crease <strong>in</strong> the demand for labor, andhigher <strong>in</strong>comes for the workers.


I MPACT OF T AX S YSTEM 55SAVINGThere are three components of Gross Domestic Sav<strong>in</strong>g (GDS): sav<strong>in</strong>g by households (Sh); bybus<strong>in</strong>esses (Sb); and by the government (Sg). The gap between gross domestic <strong>in</strong>vestment andgross domestic sav<strong>in</strong>g is filled by net foreign sav<strong>in</strong>g (Sf), which <strong>in</strong>cludes foreign aid and net<strong>in</strong>flows private capital. In <strong>Mozambique</strong>, the IMF estimates that GDS has been extremely low overthe <strong>PARPA</strong> <strong>II</strong> period, at around 3 percent of GDP. 50 Most of the meager supply of domesticsav<strong>in</strong>g orig<strong>in</strong>ates <strong>in</strong> the private sector, as Sg has been under 1 percent of GDP. Foreign sav<strong>in</strong>g hastherefore f<strong>in</strong>anced virtually all of the capital formation <strong>in</strong> the government sector, and most private<strong>in</strong>vestment as well.Household Sav<strong>in</strong>gThe ma<strong>in</strong> determ<strong>in</strong>ant of household sav<strong>in</strong>g is the level of personal <strong>in</strong>come. S<strong>in</strong>ce lower taxeswould boost personal <strong>in</strong>come, they would encourage more sav<strong>in</strong>g. Over the long term, however,the critical consideration is whether the tax system fosters rapid and susta<strong>in</strong>able growth, to boosthousehold <strong>in</strong>come and, with it, household sav<strong>in</strong>g.Household sav<strong>in</strong>g is also strongly affected by the distribution of <strong>in</strong>come, which <strong>in</strong> turn is heavily<strong>in</strong>fluenced by the tax system. A progressive tax system, as <strong>in</strong> <strong>Mozambique</strong> (see Fairness), taxeshigh-<strong>in</strong>come families more heavily, and m<strong>in</strong>imizes the tax burden on the poor. This tax structureis justified on equity grounds, and is fully consistent with the central <strong>PARPA</strong> goal of povertyreduction. But it does tend to reduce household sav<strong>in</strong>g because affluent families tend to have ahigher propensity to save.<strong>Tax</strong>es also reduce the rate of return on assets, and hence the <strong>in</strong>centive to save. Concern has beenexpressed that the decision <strong>in</strong> 2007 to tax <strong>in</strong>come from tradable securities will have an adverseeffect on sav<strong>in</strong>g. For develop<strong>in</strong>g countries, however, there is little evidence of a strong effect ontotal sav<strong>in</strong>g from moderate changes <strong>in</strong> the yield, as long as the returns to sav<strong>in</strong>g are significantlypositive <strong>in</strong> real terms. The ma<strong>in</strong> effect of this measure is likely to be a more level play<strong>in</strong>g field forallocat<strong>in</strong>g sav<strong>in</strong>g across different f<strong>in</strong>ancial assets. Hence the 2007 measure was justified not onlyon grounds of equity, but also to elim<strong>in</strong>ate a tax distortion that <strong>in</strong>creased the cost of credit to theprivate sector by mak<strong>in</strong>g (tax-free) government securities more attractive to the banks.Several tax measures could improve the <strong>in</strong>centive to sav<strong>in</strong>g. In theory, the best approach wouldbe to elim<strong>in</strong>ate the tax on <strong>in</strong>come from capital, but this would be extremely regressive. Morepractical measures <strong>in</strong>clude a reduction <strong>in</strong> the double taxation of dividend <strong>in</strong>come earned by<strong>in</strong>dividual shareholders, as discussed earlier; 51 implementation of the monetary adjustment forthe calculat<strong>in</strong>g capital ga<strong>in</strong>s, as called for <strong>in</strong> the tax laws (and long overdue); and elim<strong>in</strong>at<strong>in</strong>g theStamp <strong>Tax</strong> on f<strong>in</strong>ancial transactions. No estimates are available on the likely impact of these50 See the Selected Economic Indicators tables <strong>in</strong> IMF (2007) and IMF (2009b).51 Hassan (2009) shows that the tax system <strong>in</strong> <strong>Mozambique</strong> strongly favors debt f<strong>in</strong>anc<strong>in</strong>g over equityf<strong>in</strong>anc<strong>in</strong>g, due to the deductibility of <strong>in</strong>terest expenses and the double taxation of dividend <strong>in</strong>come.


56 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEmeasures on domestic sav<strong>in</strong>g, but they all have the basic advantage of be<strong>in</strong>g consistent with goodtax practices. 52Bus<strong>in</strong>ess Sav<strong>in</strong>gGross sav<strong>in</strong>g by the bus<strong>in</strong>ess sector consists of reta<strong>in</strong>ed earn<strong>in</strong>gs plus the accumulation ofdepreciation allowances (capital amortization). These <strong>in</strong>ternal sources of sav<strong>in</strong>g are a major formof f<strong>in</strong>ance for private <strong>in</strong>vestment everywhere <strong>in</strong> the world, and especially <strong>in</strong> countries like<strong>Mozambique</strong> where credit access is limited and borrow<strong>in</strong>g costs are high. A reduction <strong>in</strong> thecompany tax rate would clearly enhance net earn<strong>in</strong>gs, and some portion of the extra cash flowwill be reta<strong>in</strong>ed to f<strong>in</strong>ance bus<strong>in</strong>ess expansion. This is one of the strongest arguments <strong>in</strong> favor ofus<strong>in</strong>g base-broaden<strong>in</strong>g measures to reduce the tax rate on bus<strong>in</strong>ess <strong>in</strong>come, rather thanchannel<strong>in</strong>g the benefits fully <strong>in</strong>to boost<strong>in</strong>g government revenue. The extent to which additionalearn<strong>in</strong>gs are channeled to <strong>in</strong>vestment depends, however, on the availability of profitopportunities, which <strong>in</strong> turn depend on the quality of the bus<strong>in</strong>ess environment and growthprospects for the economy.Another important tax factor affect<strong>in</strong>g bus<strong>in</strong>ess sav<strong>in</strong>g is the familiar problem of delays anddenials <strong>in</strong> obta<strong>in</strong><strong>in</strong>g refunds of VAT and <strong>in</strong>come tax. These refund problems impair bus<strong>in</strong>esscash flow and reduce the availability of <strong>in</strong>ternal earn<strong>in</strong>gs that could be used for bus<strong>in</strong>essexpansion.Government Sav<strong>in</strong>gGovernment sav<strong>in</strong>g is the difference between current revenues and current expenditures – <strong>in</strong>other words, the funds available for capital formation from current government <strong>in</strong>come.(Government borrow<strong>in</strong>g is a claim on the sav<strong>in</strong>g of some other sector, rather than a source ofsav<strong>in</strong>g by government.) An <strong>in</strong>crease <strong>in</strong> Sg can be achieved by reduc<strong>in</strong>g current expenditures or<strong>in</strong>creas<strong>in</strong>g current revenue – or any comb<strong>in</strong>ation of the two that would provide more resourcesfor capital formation.Sum of the PartsThe tax effect on any one component of GDS is less important than the sum of the parts. Forexample, a tax cut might reduce Sg while <strong>in</strong>creas<strong>in</strong>g Sh and Sb. The net impact on GDS could bepositive or negative, depend<strong>in</strong>g on the conditions. In addition, the short run effects on sav<strong>in</strong>gfrom a change <strong>in</strong> the tax yield could be very different from the effects <strong>in</strong> the medium to long run,depend<strong>in</strong>g on how the government spends the revenue. A more concrete conclusion is that taxpolicy should avoid creat<strong>in</strong>g undue dis<strong>in</strong>centives to sav<strong>in</strong>g or unnecessary distortions favor<strong>in</strong>gcerta<strong>in</strong> forms of sav<strong>in</strong>g over others. This pr<strong>in</strong>cipal lends support to the measures noted above <strong>in</strong>the sections on household and bus<strong>in</strong>ess sav<strong>in</strong>g.52 Versano et al. (2006) recommended reta<strong>in</strong><strong>in</strong>g the stamp tax on f<strong>in</strong>ancial transactions on the argumentthat <strong>in</strong>come from f<strong>in</strong>ancial assets was not taxed. This condition changed <strong>in</strong> 2007.


I MPACT OF T AX S YSTEM 57PRIVATE SECTOR DEVELOPMENTThis section exam<strong>in</strong>es how the tax system affects private sector development. Given the smallsize of most bus<strong>in</strong>esses <strong>in</strong> <strong>Mozambique</strong>, this objective is <strong>in</strong>separable from the promotion of smalland medium sized enterprises (SMEs), and efforts to <strong>in</strong>crease formalization. Thus, <strong>PARPA</strong> <strong>II</strong>prioritizes (on page 1) “greater <strong>in</strong>tegration of the national economy and an <strong>in</strong>crease <strong>in</strong>productivity” via “measures to help small and medium-size companies to flourish <strong>in</strong> the formalsector.” To this end, the Council of M<strong>in</strong>isters <strong>in</strong> 2007 adopted an “SME Development Strategy,”which identified an excessive tax burden and the high cost of pay<strong>in</strong>g taxes as a major obstacle toSME development. <strong>Tax</strong> considerations also significantly affect a second issue highlighted <strong>in</strong> thestrategy document: the lack of vertical and horizontal l<strong>in</strong>kages between firms. 53<strong>Tax</strong> Policy and Enterprise PerformanceThe most direct (and obvious) impact of the tax system on enterprise performance results fromthe fact that <strong>in</strong>come tax payments reduce net earn<strong>in</strong>gs, which, <strong>in</strong> a credit-constra<strong>in</strong>ed economy,are a primary source of f<strong>in</strong>ance for <strong>in</strong>vestment and expansion. Import duties and export taxessimilarly affect bus<strong>in</strong>ess conditions by <strong>in</strong>creas<strong>in</strong>g the cost of compet<strong>in</strong>g imports and thus prices <strong>in</strong>the domestic market; the benefits are greatest for firms that are otherwise uncompetitive.Inefficient process<strong>in</strong>g of tax refunds, for both VAT and <strong>in</strong>come tax, further constra<strong>in</strong> work<strong>in</strong>gcapital and funds for expansion. Ebrill et al. (2002, p.6) highlight the “tension between theimportance of assur<strong>in</strong>g prompt refunds—without which the VAT loses many of its economicmerits—and the desire of governments to guard their revenues aga<strong>in</strong>st fraud and the temptationthey face to strengthen revenues by simply delay<strong>in</strong>g refund payments.” 54 The adoption ofmodern <strong>in</strong>formation technology and risk-based procedures for verify<strong>in</strong>g and approv<strong>in</strong>g refundrequests can go a long way towards reduc<strong>in</strong>g this tension (see Chapter 3: <strong>Tax</strong> Adm<strong>in</strong>istration).The tax code and the Code of Fiscal Benefits also affect <strong>in</strong>vestment efficiency (see Investment)and bus<strong>in</strong>ess decisions on labor <strong>in</strong>tensity and job creation (see Employment). Furthermore, theprovision of fiscal benefits tilts the play<strong>in</strong>g field <strong>in</strong> favor of new <strong>in</strong>vestors who qualify for CPIapproval, relative to exist<strong>in</strong>g bus<strong>in</strong>esses and <strong>in</strong>vestors who do not enjoy the same tax advantages.Enterprise development is also strongly affected by the quality of tax adm<strong>in</strong>istration. Highcompliance costs, particularly for SMEs, stem from the complexity of the tax code, the use ofoutdated procedures by the AT for many tax operations, and <strong>in</strong>adequate public <strong>in</strong>formation andtaxpayer services (see Chapter 3: <strong>Tax</strong> Adm<strong>in</strong>istration). 55 Faced with these complexities, even53 Other major obstacles identified <strong>in</strong> the SME Development Strategy are excessive regulatory barriers;high f<strong>in</strong>ance costs and limited access to f<strong>in</strong>ance; poorly qualified labor; poor market access; and a lack ofentrepreneurial spirit.54 Given the apparently widespread under-declaration of sales (discussed below), some firms may bereport<strong>in</strong>g a credit position with regards to VAT only because they are fully declar<strong>in</strong>g VAT paid on <strong>in</strong>putswhile under-declar<strong>in</strong>g that collected through sales.55 Accord<strong>in</strong>g to private sector tax specialists, many technical details of the tax code have not yet beenclarified through support<strong>in</strong>g regulations or formal <strong>in</strong>terpretations from the AT. Faced with ambiguities ofthis sort, even companies who try to comply fully with the tax law us<strong>in</strong>g professional account<strong>in</strong>g servicescan be vulnerable to costly penalties.


58 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEwell <strong>in</strong>tentioned firms are vulnerable to arbitrary and unpredictable decisions by AT officialsregard<strong>in</strong>g <strong>in</strong>terpretation of the law, and costly penalties for even <strong>in</strong>consequential <strong>in</strong>fractions (suchas misspell<strong>in</strong>g a name on a receipt). Cases of this sort were mentioned by nearly every privatesector <strong>in</strong>terlocutor <strong>in</strong>terviewed for this study, with particular reference to situations <strong>in</strong> which localtax offices need additional revenue quickly to meet revenue targets.As discussed <strong>in</strong> Chapter 3, plans by the AT to modernize tax adm<strong>in</strong>istration will eventuallyaddress many of these issues. In the meantime, adm<strong>in</strong>istrative problems impose extra costs formany firms and create unequal competitive conditions <strong>in</strong> the market place, as some firms beardisproportionate tax costs while others benefit from generous tax breaks or elude the tax netaltogether.<strong>Tax</strong>es and FormalizationEncourag<strong>in</strong>g formalization of micro and small enterprises is an important objective ofgovernment policy. Informal sector activity is undoubtedly a large part of the economy.Widespread <strong>in</strong>formality deprives the government of revenue, and limits the scope for bus<strong>in</strong>essexpansion and thus growth potential for the economy. Informal bus<strong>in</strong>ess activities even alsoimpair the growth potential of firms <strong>in</strong> the formal sector. This can be seen <strong>in</strong> the World Bank’smost recent Bus<strong>in</strong>ess Climate Assessment for <strong>Mozambique</strong>, which found that <strong>in</strong>formal sectorcompetition ranked first on the list of “major or severe” obstacles to bus<strong>in</strong>ess operations, <strong>in</strong> asurvey of almost 600 firms <strong>in</strong> 2008 (World Bank 2009). In the World Bank study, “formality”was synonymous with tax registration. An earlier Informal Sector Survey (INFOR) by INE <strong>in</strong>2005 def<strong>in</strong>ed “<strong>in</strong>formality” more broadly to <strong>in</strong>clude firms that are not fully registered taxes orother purposes as obliged by law. From this survey INE estimated that 75 percent of theeconomically active population earns their livelihoods <strong>in</strong> the <strong>in</strong>formal sector (GoM 2006b).The extent of <strong>in</strong>formality <strong>in</strong> any economy depends on the benefits and costs of formaliz<strong>in</strong>g, whichare strongly <strong>in</strong>fluenced by the tax system. In particular, the costs <strong>in</strong>clude both tax payments andexpenses related to tax compliance, as well as direct and <strong>in</strong>direct expenses for registration andlicens<strong>in</strong>g, and identity cards. Even travel costs to government offices can be a significant barrierfor very poor household enterprises. Informality avoids these costs, though it often requires thepayment of bribes, and restricts the scope for bus<strong>in</strong>ess expansion. On the other side of theformula, the benefits of formalization <strong>in</strong>clude access to the bank<strong>in</strong>g sector, access to the legalsystem for contract enforcement, greater flexibility to expand operations, and protection fromexposure to penalties or various forms of official harassment. If the costs of formalizationoutweigh the benefits, as perceived by each firm, or if the benefits are not sufficiently clear toallow an <strong>in</strong>formed judgment, then <strong>in</strong>formality is the default option for micro and smallenterprises.<strong>Tax</strong> evasion is widely considered to be one of the primary motives for <strong>in</strong>formality <strong>in</strong><strong>Mozambique</strong>. Data from INE’s most recent survey of the <strong>in</strong>formal sector (INFOR), conducted <strong>in</strong>2004, cast doubt on this hypothesis (INE 2006). Table 5-1 summarizes the INFOR f<strong>in</strong>d<strong>in</strong>gs forself-employed non-agricultural <strong>in</strong>formal operators.


I MPACT OF T AX S YSTEM 59Table 5-1Reasons for Be<strong>in</strong>g InformalFrequencyPercentToo complex 113 11.7Too expensive 144 14.9In process 71 7.3Not obligatory 198 20.5Don't know if need 398 41.2Anti-state 8 0.8Other 35 3.6Total 967 100SOURCE: Sub-sample of self-employed, nonagricultural <strong>in</strong>formal firms from INE's 2004INFOR Survey (INE 2006).These results <strong>in</strong>dicate that a lack of <strong>in</strong>formation is far more important as a reason for <strong>in</strong>formalitythan the cost of formalization or complexity of the process. Yet better <strong>in</strong>formation would notnecessarily lead to a surge <strong>in</strong> formalization; it could <strong>in</strong>stead conv<strong>in</strong>ce many <strong>in</strong>formal firms thatthe costs of formalization <strong>in</strong>deed outweigh the benefits.In particular, tax rates, tax compliance costs and the complexity of the tax system, as applied tovery small bus<strong>in</strong>esses, strongly <strong>in</strong>fluence firm-level decisions on formalization. The newSimplified <strong>Tax</strong> (ISPC) passed <strong>in</strong> 2009 should go a long way to reduce these tax-related costs, andtherefore shift the balance towards formalization for many more enterprises. For some smallbus<strong>in</strong>esses, however, even the ISPC will be too heavy a tax to bear because the 3 percent rateapplies to sales volume rather than <strong>in</strong>come or profit. For many small enterprises, the tax liabilityunder the ISPC will exceed the standard IRPC tax rate as well as the maximum effective tax onhigh-<strong>in</strong>come households under the IRPS. Take the case of an enterprise operat<strong>in</strong>g with a narrowprofit marg<strong>in</strong> of 5 percent, for example: the 3 tax on sales represents a 60 percent tax on <strong>in</strong>come.F<strong>in</strong>ally, registered entities start pay<strong>in</strong>g tax at a much lower level of <strong>in</strong>come under the ISPC thanunder the <strong>in</strong>come tax, because the exemption threshold–equal to 36 times the m<strong>in</strong>imum wage <strong>in</strong>each case – relates to sales rather than <strong>in</strong>come. The risk is that many <strong>in</strong>formal enterprises,especially those with highly variable <strong>in</strong>come from one year to the next, will still have an <strong>in</strong>centiveto avoid registration under the ISPC.In practice, the outcome from <strong>in</strong>troduction of the ISPC will depend on how the program isadm<strong>in</strong>istered. Prospects will be favorable if the AT succeeds accord<strong>in</strong>g to plan <strong>in</strong> establish<strong>in</strong>gsimple procedures, simple forms and receipt documents, and extensive geographic coverage oftax services. Early evidence is very positive: the AT reported that 1800 new taxpayers registeredunder the ISPC <strong>in</strong> June, 2009, the first month of operation for this new tax. In addition, 350 smallenterprises converted their registration from the simplified VAT and <strong>in</strong>come tax regimes, to takeadvantage of the lower tax rate under the ISPC (as discussed <strong>in</strong> Chapter 3).The VAT, too, is designed to create <strong>in</strong>centives for formalization. Registered firms are allowed todeduct the VAT component of <strong>in</strong>put costs only if the supplier is VAT-registered and provides aproper VAT receipt. This feature of the tax encourages all participants <strong>in</strong> the value-cha<strong>in</strong> to


60 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEregister, for any product or service that passes through a formal sector entity on its way to thef<strong>in</strong>al consumer. The same <strong>in</strong>centive does not arise for value cha<strong>in</strong>s operat<strong>in</strong>g entirely with<strong>in</strong> the<strong>in</strong>formal sector, or among VAT-exempt registered bus<strong>in</strong>esses. Hence, the <strong>in</strong>centive forformalization due to VAT depends <strong>in</strong> part on the extent to which the economy is alreadyformalized.Another important feature of the VAT, highlighted by Keen (2008), is that this tax effectivelygenerates revenue from <strong>in</strong>formal activities, to the extent that unregistered enterprises purchase<strong>in</strong>puts or supplies though channels that have been subject to VAT, <strong>in</strong>clud<strong>in</strong>g either registereddomestic enterprises or imports (other than smuggled goods). S<strong>in</strong>ce the unregistered enterprisecannot claim credit for VAT paid on these <strong>in</strong>puts, the tax (to this extent) is <strong>in</strong>corporated <strong>in</strong>to thecost of do<strong>in</strong>g bus<strong>in</strong>ess.The <strong>in</strong>centives for formalization are also <strong>in</strong>fluenced by the capacity of the AT to monitor taxcompliance. Effective enforcement of registration, fil<strong>in</strong>g, and payment requirements, along withbroader geographical coverage of tax <strong>in</strong>spections, <strong>in</strong>crease the likelihood of detection and the costof noncompliance. This can shift the benefit-cost balance <strong>in</strong> the direction of encourag<strong>in</strong>gregistration (and also pay<strong>in</strong>g an appropriate amount of tax). The danger here is that punitive andarbitrary enforcement practices applied to registered enterprises could underm<strong>in</strong>e the advantagesof compliance, and foster bribery and corrupt practices rather than formalization. Improvements<strong>in</strong> taxpayer services must therefore go hand <strong>in</strong> hand with efforts to strengthen enforcement.As noted above, one problem with <strong>in</strong>formality is that it deprives the government of revenue.Many people <strong>in</strong>terviewed for this study expressed the view that gett<strong>in</strong>g large numbers of <strong>in</strong>formalsector enterprises to register will significantly enhance revenue, and provide fiscal space thatcould be used to reduce the tax burden on firms that are already comply<strong>in</strong>g with the tax laws.This is probably a false hope, for two related reasons. First, the amount of revenue that can beobta<strong>in</strong>ed from very small bus<strong>in</strong>esses is likely to be very small, even <strong>in</strong> the aggregate.International experience suggests that large taxpayers account for 70 to 80 percent of totalrevenues, and small taxpayers just 5 to 10 percent. 56 The potential revenue yield from register<strong>in</strong>gsmall, <strong>in</strong>formal enterprises is further reduced by virtue (and it is <strong>in</strong>deed a virtue) of the relativelyhigh exemption threshold under the <strong>in</strong>come tax and the ISPC, and the low tax rate that nowapplies to firms registered under the ISPC. Neither the M<strong>in</strong>istry of F<strong>in</strong>ance nor the AT areexpect<strong>in</strong>g substantial revenue ga<strong>in</strong>s <strong>in</strong> the near term from the ISPC.The second reason not to expect a large revenue ga<strong>in</strong> from the ISPC is that the AT should notdevote substantial resources to chas<strong>in</strong>g large numbers of very small enterprises. It is important, ofcourse, to enforce the tax law and promote tax compliance across the board. But resources for taxadm<strong>in</strong>istration are limited, and there is a high opportunity cost from try<strong>in</strong>g to cover very smallenterprises thoroughly, <strong>in</strong> the form of revenue foregone from allocat<strong>in</strong>g AT personnel and budgetresources to other operations with a higher yield.56 IMF (May, 2009c), , pp. 26-27.


I MPACT OF T AX S YSTEM 61These remarks about the limited revenue potential from br<strong>in</strong>g<strong>in</strong>g <strong>in</strong>formal enterprises <strong>in</strong>to the taxnet do not apply to <strong>in</strong>formal or unreported operations of larger enterprises or high-<strong>in</strong>come<strong>in</strong>dividuals. In the <strong>in</strong>terest of both revenue generation and tax equity, this set of “<strong>in</strong>formal”activities should be a primary target for tax <strong>in</strong>vestigations and audits, us<strong>in</strong>g all available legaltools.Differential <strong>Tax</strong> Effects by Firm SizeWhat is the differential effect of the tax system on small, medium, and large firms? Byiers (2009)exam<strong>in</strong>ed this question us<strong>in</strong>g data from a 2006 enterprise survey. Although the survey <strong>in</strong>volved arelatively small sample, and limited <strong>in</strong>formation on f<strong>in</strong>ancial issues, the analysis suggests that theeffective tax burden does vary considerably by firm size, across firms <strong>in</strong> similar sectors andlocations. The results show an <strong>in</strong>verted-U relationship: small firms tend to bear a low tax burdenthrough evasion or exemption, and large firms tend to m<strong>in</strong>imize taxes through sophisticated taxplann<strong>in</strong>g and exemptions provided under the Code of Fiscal Benefits. Middle-size, <strong>in</strong> contrast,cannot easily escape the tax net or afford professional services to engage <strong>in</strong> sophisticated taxplann<strong>in</strong>g. These f<strong>in</strong>d<strong>in</strong>gs provide only a crude <strong>in</strong>dication of <strong>in</strong>equities, because the observed taxdifferentials could reflect legitimate and systematic differences <strong>in</strong> firm structure, such asdeductions due to capital costs, <strong>in</strong>terest expenses, and loss carry forward among large firms, andthe policy of exempt<strong>in</strong>g very small <strong>in</strong>comes from tax through the threshold level. Nonetheless,the evidence suggests that the tax system tends to be a greater h<strong>in</strong>drance to bus<strong>in</strong>ess operations asfirms reach a certa<strong>in</strong> “visible” size.Table 5-2 provides some <strong>in</strong>sight on access to fiscal benefits, by size of firm, us<strong>in</strong>g CPI data for2006 and 2007. Out of 192 <strong>in</strong>vestments approved <strong>in</strong> 2007, more than half were fully foreigndirect <strong>in</strong>vestments (FDI), ris<strong>in</strong>g to 79 percent <strong>in</strong>clud<strong>in</strong>g jo<strong>in</strong>t ventures with a majority foreignequity share. Just 17 percent were fully national enterprises, with 3 percent be<strong>in</strong>g jo<strong>in</strong>t ventureshav<strong>in</strong>g a national majority and 3 percent with equal shares. The story is the same for 2006.Table 5-2Fiscal Benefit Code Investments as a Share of Total Approved InvestmentsCategory 2006 2007Total approved <strong>in</strong>vestments 157 192FDI only 46% 51%Jo<strong>in</strong>t ventures 36% 33%FDI majority share 31% 28%National majority share 2% 3%Equal shares 3% 3%National only 18% 17%Equity < $50,000 6% 5%$50,001 to $100,000 2% 2%$100,001 to $250,000 3% 4%> $250,000 8% 6%SOURCE: Calculations based on data provided by CPI.


62 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUENational enterprises can qualify for CPI approval and fiscal benefits with an <strong>in</strong>vestment as low as$5,000. Yet only 10 national projects were approved <strong>in</strong> 2007 with equity capital less than$50,000, and only 13 with equity capital under $100,000. The fact that very few small<strong>in</strong>vestments by national companies qualify for fiscal benefits suggests that the procedural costsexceed the prospective benefits. Moreover, anecdotal evidence from the <strong>in</strong>terviews suggests thateven with project approval, the stipulated fiscal benefits are not always obta<strong>in</strong>ed due toprocedural complications at the implementation stage.The World Bank’s recent Investment Climate Assessment (ICA) provides further <strong>in</strong>formation onthe relationship between taxation and firm size, based on a survey of bus<strong>in</strong>ess managers cover<strong>in</strong>gthe ma<strong>in</strong> economic sectors and cities, as summarized <strong>in</strong> Table 5-3 (World Bank 2009). 57Table 5-3Number of Firms by Size, Industry, and CityFirm SizeTotalMicro (1-4) Small (5-19) Medium (20-99) Large (100+)I NDUSTRYFood 7 37 42 10 96Garments 5 41 8 3 57Other manufactur<strong>in</strong>g 5 56 15 0 76Retail 86 74 26 6 192Other services 17 100 51 10 178C ITYMaputo 71 239 97 18 425Matola 15 19 25 3 62Beira 24 17 12 2 55Nampula 9 33 8 6 56Unknown 1 0 0 0 1Total 120 308 142 29 599SOURCE: Authors' calculation us<strong>in</strong>g World Bank (2009).Figure 5-4 shows survey responses on the extent to which tax rates are an obstacle to firmoperations, on a scale from 0 (no obstacle) to 4 (serious obstacle). For micro-enterprises themedian response is 1 (m<strong>in</strong>or obstacle) while small, medium and large firms all have medianresponses of 2 (moderate obstacle). Also, the spread of responses is more concentrated around themedian for micro and small firms than for medium and large firms. (The shaded box representsthe middle fifty percent of firms, those <strong>in</strong> the 25 th to 75 th percentiles). This evidence suggests that57 The World Bank survey data discussed here is from the Investment Climate Survey carried out <strong>in</strong>2007, which surveyed approximately 600 manufactur<strong>in</strong>g and service <strong>in</strong>dustry firms <strong>in</strong> Maputo, Matola(Maputo prov<strong>in</strong>ce), Beira and Nampula, stratified by location and firm size.


I MPACT OF T AX S YSTEM 63micro-firms are not much affected by tax rates, but as firms <strong>in</strong>crease <strong>in</strong> size, tax rates become amore significant burden. The higher variation among medium to large size firms may <strong>in</strong>dicate<strong>in</strong>consistent implementation of the tax laws. Nonetheless, very few of the firms <strong>in</strong> the surveyregarded tax rates as a serious h<strong>in</strong>drance, and fully 160 of the 599 firms <strong>in</strong>terviewed did notregard tax rates as an obstacle at all.Figure 5-4<strong>Tax</strong> Rates as a Constra<strong>in</strong>t to Bus<strong>in</strong>ess, by Firm sizetax rates0 1 2 3 4Micro (1-4) Small (5-19) Medium (20-99) Large (100 and above)excludes outside valuesNote: This graph displays the distribution of responses. The middle l<strong>in</strong>e is the median response, and the shaded box shows therange between the 25 th and 75 th percentiles. Source: Authors’ calculations us<strong>in</strong>g data from World Bank (2009).The number of zero-obstacle scores rises to 276 for tax adm<strong>in</strong>istration, suggest<strong>in</strong>g (surpris<strong>in</strong>gly)that adm<strong>in</strong>istrative issues are viewed as an obstacle less frequently than tax rates. Here, too, theresponses differed by firm size, as shown <strong>in</strong> Figure 5-5. For micro-firms the median score wasactually zero (no obstacle), while for large firms the median was 2 (moderate obstacle). This maybe due to a greater concentration by the AT on large firms, to ensure greatest revenue ga<strong>in</strong>s, alongwith the fact that many micro-firms are outside the tax net through avoidance or exemption.F<strong>in</strong>ally, Figure 5-6 shows the response by firm managers when asked to estimate the share ofsales declared for tax purposes by firms similar to theirs. This question is a proxy for the extent oftax evasion. The results <strong>in</strong>dicate that large firms are, on average, perceived as declar<strong>in</strong>g a veryhigh share of sales for tax purposes, with small firms report<strong>in</strong>g the lowest share.


64 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEFigure 5-5<strong>Tax</strong> Adm<strong>in</strong>istration as a Constra<strong>in</strong>t to Bus<strong>in</strong>ess, by Firm Sizetax adm<strong>in</strong>istration0 1 2 3 4Micro (1-4) Small (5-19) Medium (20-99) Large (100 and above)excludes outside valuesNote: This graph displays the distribution of responses. The middle l<strong>in</strong>e is the median, and the shaded box shows the rangebetween the 25 th and 75 th percentiles. For micro-firms, the median score co<strong>in</strong>cides with the bottom of the shaded box.SOURCE: Authors’ calculations us<strong>in</strong>g data from World Bank (2009).Figure 5-6Responses on Sales Declared for <strong>Tax</strong> Purposes, by Firm Sizesales tax0 20 40 60 80 100Micro (1-4) Small (5-19) Medium (20-99) Large (100 and above)excludes outside valuesNote: This displays the distribution of responses as follows: the middle l<strong>in</strong>e is the median, the surround<strong>in</strong>g shaded box the middle50 percentiles (between the 25 th and 75 th percentiles).SOURCE: Authors’ calculations us<strong>in</strong>g data from World Bank (2009).To highlight an important detail from this question, Table 5-4 reports the share of respondents, byfirm size, who estimate that firms declare either zero or 100 percent of sales. Among large firms<strong>in</strong> the survey, 45 percent estimated that a typical firm reports 100 percent of sales for taxpurposes, compared to just 22 percent for small firms and 20 percent for micro firms. And <strong>in</strong> allsize categories, the percentage of zero estimates is very small, with the highest fraction (just 4.2percent) com<strong>in</strong>g from micro-firms. While highly subjective, these results suggest that tax evasion


I MPACT OF T AX S YSTEM 65varies substantially by firm size, most likely due to the vary<strong>in</strong>g ability of firms at different scalesto avoid the scrut<strong>in</strong>y of the tax authorities. The results also support the contention, mentioned <strong>in</strong>many <strong>in</strong>terviews, that the tax system creates unfair competition for firms that are compliant, tothe degree that other firms evade taxes or hide sales.Table 5-4Shares of Firms Report<strong>in</strong>g 0% and 100% Sales DeclaredFirm Size 0% Declared 100% Declared NMicro (1-4) 4.2 20.0 120Small (5-19) 3.2 22.1 308Medium (20-99) 2.8 32.4 142Large (100 and above) 3.4 44.8 29Total 3.3 25.2 599SOURCE: Authors’ calculations us<strong>in</strong>g data from World Bank (2009).The World Bank survey results <strong>in</strong>dicat<strong>in</strong>g that under-report<strong>in</strong>g is greatest among small and microfirms does not <strong>in</strong>validate the po<strong>in</strong>t made earlier (see Formalization) that the AT should not devotesubstantial resources to pursu<strong>in</strong>g small taxpayers. The argument is that the revenue yield fromdeal<strong>in</strong>g with a large number of small payments will be low relative to the opportunity costs ofdeploy<strong>in</strong>g personnel to other uses. Information on this po<strong>in</strong>t can be seen <strong>in</strong> INE’s most recentAnnual Enterprise Survey of over 24,000 firms (INE, 2009) that are registered <strong>in</strong> one form oranother. Table 5-5 presents the distribution of number of firms and sales turnover, by size of firm.Micro-enterprises constitute 82.4 percent of the firms, which account for 36.2 percent of theaggregate sales turnover. In contrast, the top 9.3 percent of firms account for 50 percent of salesturnover, and more than half of the total sales come from the largest 1.1 percent of firms. Bycompar<strong>in</strong>g turnover shares with the number of firms <strong>in</strong> each size category, one can see that theaverage large firm <strong>in</strong> this sample generates more than 50 times as much bus<strong>in</strong>ess volume as theaverage micro firm on INE’s list of enterprises. These figures imply that the tax authorities needto deal with more than 50 times as many micro firms to cover the same volume of bus<strong>in</strong>ess aswould be covered, on average, with one large firm. 58 If anyth<strong>in</strong>g, the data appear to underestimatethe sales share of large firms. This could be the case if very small firms registered with INE tendto be those with higher revenues, such as professional service providers.58 A similar calculation from Enterprise Survey can be made us<strong>in</strong>g data by sector, rather than firm size.This calculation shows that the volume of bus<strong>in</strong>ess per firm is greatest by far <strong>in</strong> the electricity, water andgas sector. Other sectors with relatively high sales per firm are, <strong>in</strong> descend<strong>in</strong>g order: transport, storage andcommunication; extractive <strong>in</strong>dustries; construction; and manufactur<strong>in</strong>g.


66 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUETable 5-5Distribution of Sales Turnover by Firm SizeWorkers Share of firms Share of SalesTurnoverRatio: sales share tofirms share=100 1.1 25.5 24.1Total 100.0 100.0SOURCE: INE (2009) and authors’ calculations.The problems aris<strong>in</strong>g from VAT and <strong>in</strong>come tax refunds also have a differential effect by firmsize, because large firms are much more capable of deal<strong>in</strong>g with fil<strong>in</strong>g and documentationrequirements. They are also more able to cope with work<strong>in</strong>g capital costs of long delays <strong>in</strong>obta<strong>in</strong><strong>in</strong>g a refund, by hav<strong>in</strong>g a stronger f<strong>in</strong>ancial position and better access to credit. VAT refundproblems are especially pernicious for small bus<strong>in</strong>esses that may have an eye on export markets.Exports are zero-rated under the VAT to ensure that Mozambican products do not bear the burdenof home-country <strong>in</strong>direct taxes when compet<strong>in</strong>g <strong>in</strong> external markets. If small bus<strong>in</strong>esses areeffectively blocked from obta<strong>in</strong><strong>in</strong>g VAT refunds due to complex procedures and long delays, thenthey are also largely blocked from penetrat<strong>in</strong>g export markets. See Bolnick (2007) for moredetails.<strong>Tax</strong>es and Bus<strong>in</strong>ess L<strong>in</strong>kagesAs stated above, one of the <strong>PARPA</strong> priorities is to strengthen l<strong>in</strong>kages with<strong>in</strong> the domesticeconomy. The tax system actually discourages the achievement of this objective <strong>in</strong> severalun<strong>in</strong>tended ways. Import duties have an obvious <strong>in</strong>fluence on bus<strong>in</strong>ess decisions to use importedversus domestic <strong>in</strong>puts, and thus the development of l<strong>in</strong>kages with<strong>in</strong> the domestic economy. Inparticular, the full remission of duty and tax on imported capital goods and other <strong>in</strong>puts forcerta<strong>in</strong> types of <strong>in</strong>vestments under the Code of Fiscal Benefits creates an <strong>in</strong>centive favor<strong>in</strong>g theprocurement of imported <strong>in</strong>puts. An example is the purchase of imported versus locally producedfurniture by a new tourism facility.In less obvious ways, the VAT and <strong>in</strong>come tax also affect the development of bus<strong>in</strong>ess l<strong>in</strong>kages.Look<strong>in</strong>g first at the VAT, the same feature that encourages formalization–namely, the need forsuppliers to registered bus<strong>in</strong>esses to provide VAT receipt–serves to impede l<strong>in</strong>kages betweenformal bus<strong>in</strong>esses and small suppliers who are not registered for VAT and have reasons toma<strong>in</strong>ta<strong>in</strong> this status (see Formalization). In an economy with a large <strong>in</strong>formal sector, smallbus<strong>in</strong>esses may decide aga<strong>in</strong>st register<strong>in</strong>g if they purchase their own supplies of goods andservices from unregistered enterprises, and would face substantially higher costs <strong>in</strong> deal<strong>in</strong>g withregistered vendors. The complexity of the VAT refund process is another factor h<strong>in</strong>der<strong>in</strong>g thedevelopment of l<strong>in</strong>kages. This occurs <strong>in</strong> cases where a small domestic supplier would be eligibleto claim a refund of VAT paid on <strong>in</strong>puts to goods or services that are sold to potentially important


I MPACT OF T AX S YSTEM 67clients operat<strong>in</strong>g <strong>in</strong> IFZs, such as Mozal. Without timely refunds, small suppliers may be unableto susta<strong>in</strong> these transactions, even if they would otherwise be a very favorable bus<strong>in</strong>essopportunity.The <strong>in</strong>come tax, too, creates un<strong>in</strong>tended barriers to the development of l<strong>in</strong>kages between largeand small bus<strong>in</strong>esses. Under the <strong>in</strong>come tax code, “expenses not duly documented” aredisallowed as bus<strong>in</strong>ess deductions and subject to a tax of 35 percent. This provision may be<strong>in</strong>voked by tax officials wherever a tax-registered firm has purchased goods or services from asmall-scale supplier who could not issue a receipt comply<strong>in</strong>g with legal requirements <strong>in</strong> everydetail, <strong>in</strong>clud<strong>in</strong>g proper spell<strong>in</strong>g and abbreviations that match tax registration records. In theory,this provision creates <strong>in</strong>centives for formalization on the part of small-scale suppliers, but <strong>in</strong>practice it raises serious problems similar to those noted for VAT. Examples encountered dur<strong>in</strong>g<strong>in</strong>terviews for this study <strong>in</strong>clude large food processors purchas<strong>in</strong>g crops from large numbers ofsmall family farmers <strong>in</strong> outgrower schemes; tourist facilities <strong>in</strong> remote locations purchas<strong>in</strong>g foodsor crafts from local households; and two <strong>in</strong>stances of medium sized firms (one <strong>in</strong> manufactur<strong>in</strong>gand one <strong>in</strong> services) be<strong>in</strong>g constra<strong>in</strong>ed from procur<strong>in</strong>g transportation services from small bus<strong>in</strong>esspartners.Under the <strong>in</strong>come tax code, larger enterprises can avoid disallowance of a deduction and thepenalty tax rate on improperly documented expenses by withhold<strong>in</strong>g 20 percent at source onpayments to suppliers who cannot issue formal and accurate tax receipts. Even with this course ofaction, the tax system still impedes l<strong>in</strong>kages, because 20 percent withhold<strong>in</strong>g on gross receiptscreates an onerous burden on small suppliers. If, for example, the small supplier operates on a 20percent gross marg<strong>in</strong>, then withhold<strong>in</strong>g tax would absorb 100 percent of the <strong>in</strong>come – from abus<strong>in</strong>ess that ought to be pay<strong>in</strong>g a very low effective tax rate, or even no tax at all if net <strong>in</strong>comefalls below the tax threshold. In theory, prompt payment of <strong>in</strong>come tax refunds could mitigate theproblem, but, as discussed earlier, the AT is very <strong>in</strong>efficient <strong>in</strong> pay<strong>in</strong>g such claims, and <strong>in</strong> anycase small bus<strong>in</strong>esses are least able to deal with the required forms and procedures. Of course, thesmall supplier could mark up its <strong>in</strong>voice by 25 percent <strong>in</strong> order to obta<strong>in</strong> full payment after theclient withholds 20 percent on the transaction, 59 but this alternative makes the small suppliermuch less price-competitive.The new ISPC should go a long way towards mitigat<strong>in</strong>g tax impediments to the development ofbus<strong>in</strong>ess l<strong>in</strong>kages. But it is unlikely to overcome the problems altogether, given the practicaldifficulties of register<strong>in</strong>g every family farmer, local trader, and handicraft producer – many ofwhom are illiterate, and often lack even a proper identity card. Also, as noted above, the actualimpact of the ISPC rema<strong>in</strong>s to be seen.On balance, these features of the tax system work at cross purposes to the stated <strong>PARPA</strong>objectives of <strong>in</strong>creas<strong>in</strong>g the <strong>in</strong>tegration of the national economy and promot<strong>in</strong>g SMEdevelopment. The fundamental problem is the difficulty of design<strong>in</strong>g a tax system suitable for thefull range of economic actors <strong>in</strong> <strong>Mozambique</strong>. Nonetheless, the discussion above po<strong>in</strong>ts to several59 To see this, suppose the small supplier seeks a payment of $100. This can be obta<strong>in</strong>ed by <strong>in</strong>voic<strong>in</strong>g for$125, from which 20 percent of the gross amount (<strong>in</strong> this case $25) will be withhold at source as tax.


68 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEmeasures that can reduce the disconnect. In addition to effective implementation of the ISPC,these <strong>in</strong>clude streaml<strong>in</strong><strong>in</strong>g the payment of refunds (with risk-based management of the revenueeffects), reduc<strong>in</strong>g the withhold<strong>in</strong>g tax rate, allow<strong>in</strong>g bus<strong>in</strong>ess to deduct legitimate expenses evenwhen there are m<strong>in</strong>or and <strong>in</strong>consequential mistakes on receipts or <strong>in</strong>voices, and reduc<strong>in</strong>g the<strong>in</strong>centive under the CFB for us<strong>in</strong>g imported <strong>in</strong>puts rather than domestic products.FAIRNESS AND EQUITYThe Enterprise section above discussed effects of the tax system on equity and fairness acrossenterprises, particularly <strong>in</strong> connection with wide variances <strong>in</strong> tax compliance. This sectionfocuses equity across households, through tax effects on <strong>in</strong>comes and prices, as well as taxdifferentials result<strong>in</strong>g from tax evasion and corruption.Income <strong>Tax</strong>esAs noted earlier, the IRPS has a progressive structure. Annual <strong>in</strong>comes below 36 times them<strong>in</strong>imum monthly wage are free of tax, 60 and marg<strong>in</strong>al tax rates above that threshold range from10 to 32 percent (Table 5-6). This structure conforms to the pr<strong>in</strong>ciple of vertical equity, byelim<strong>in</strong>at<strong>in</strong>g the tax on very poor households, and subject<strong>in</strong>g households with higher <strong>in</strong>comes tohigher marg<strong>in</strong>al and effective tax rates. The IRPS code also broadly satisfies the pr<strong>in</strong>ciple ofhorizontal equity, because the tax applies to nearly all sources of personal <strong>in</strong>come, 61 adjusted formarital status and deductions for dependents.Table 5-6IRPS Rates<strong>Tax</strong>able Annual Income (MT) Rate (%) Deducted (MT)Up to 42000 (US$1.570) 10 0From 42.001 to 168.000 (US$1.570-6.279) 15 2,100From 168.001 to 504.000 (US$6.279-18.839) 20 10,500From 504.001 to 1.512.000 (US$18.839-56.519) 25 35,700More than 1.512.000 (US$56.519) 32 141,540SOURCE: IRPS Law 33/2007.To illustrate the progressivity of IRPS, Figure 5-7 shows the effective tax rate <strong>in</strong> 2009 at differentlevels monthly <strong>in</strong>come, for the case of a married couple, assum<strong>in</strong>g for simplicity no otherdependents or deductions. The tax structure results <strong>in</strong> an effective tax rate that gradually <strong>in</strong>creasesfrom zero percent at the exemption threshold of MT 8,235 per month ($308), ris<strong>in</strong>g to 8.1 percentfor a monthly <strong>in</strong>come of MT23.000 ($860), 13.2 percent for monthly <strong>in</strong>come of MT50.00060 There are n<strong>in</strong>e m<strong>in</strong>imum wages, by sector, for agriculture, fisheries, m<strong>in</strong><strong>in</strong>g, manufactur<strong>in</strong>g, electricity,construction, non-f<strong>in</strong>ancial activities, f<strong>in</strong>ancial activities, and public service:http://allafrica.com/stories/200904280769.html .61 The IRPS def<strong>in</strong>es five categories: <strong>in</strong>come from employment (“dependent” <strong>in</strong>come); professional andbus<strong>in</strong>ess <strong>in</strong>come; capital <strong>in</strong>come; <strong>in</strong>come from build<strong>in</strong>gs; and other <strong>in</strong>comes. Each category covers multipletypes a variety of <strong>in</strong>come.


I MPACT OF T AX S YSTEM 69($1,869), 17.0 percent at an <strong>in</strong>come of MT 100,000 ($3,738) and so on up to 32 percent as anasymptotic ceil<strong>in</strong>g. Detailed data on the distribution of <strong>in</strong>come <strong>in</strong> 2009 would be needed todeterm<strong>in</strong>e the population share subject to each tax rate, but it is clear enough that the effective taxrate is very low for most households, and that very few are fac<strong>in</strong>g an effective tax rate above 20percent.Figure 5-7Applied <strong>Tax</strong> Rates25%Effective <strong>Tax</strong> Rate20%15%10%5%0%010,00020,00030,00040,00050,00060,00070,00080,00090,000100,000110,000120,000130,000Monthly Income (MT)140,000150,000160,000170,000180,000190,000200,000SOURCE: Authors’ calculations based on IRPS Law33/2007.Even though the statutory structure of the <strong>in</strong>dividual <strong>in</strong>come tax is suitably progressive, many<strong>in</strong>terviewees suggested that there are major problems of horizontal equity due to widespread taxevasion. While formal sector employees cannot easily avoid pay<strong>in</strong>g IRPS on wage <strong>in</strong>come, due towithhold<strong>in</strong>g by employers, it is easy for evaders to under-report <strong>in</strong>comes from professionalservices, hous<strong>in</strong>g rentals, and bus<strong>in</strong>ess activities lack<strong>in</strong>g modern accounts. It is difficult andcostly <strong>in</strong> such cases for the AT to assess directly the unreported <strong>in</strong>comes. For <strong>in</strong>dividuals withhigh unreported <strong>in</strong>comes, however, the benefits of tougher enforcement–<strong>in</strong> terms of both revenueand equity–greatly outweigh the adm<strong>in</strong>istrative costs. The AT has the authority to use third partydata to determ<strong>in</strong>e gross under-declarations of <strong>in</strong>come and to use “<strong>in</strong>direct methods” to assess tax<strong>in</strong> cases there is clear evidence of improper declarations. In comb<strong>in</strong>ation with tougherenforcement, efforts by the AT to provide better <strong>in</strong>formation to the public about taxresponsibilities and obligations, and to <strong>in</strong>still pride <strong>in</strong> contribut<strong>in</strong>g to national development, mayalso have a positive effect <strong>in</strong> improv<strong>in</strong>g compliance and reduc<strong>in</strong>g <strong>in</strong>equities from tax evasion.Another equity aspect of the personal <strong>in</strong>come tax <strong>in</strong>volves adm<strong>in</strong>istrative problems similar tothose discussed above regard<strong>in</strong>g VAT refunds. Under the IRPS, taxpayers should be reimbursedby the AT when the amount of tax withheld at source plus estimated payments dur<strong>in</strong>g the tax yearexceeds the f<strong>in</strong>al amount payable on reconciled total <strong>in</strong>come at the end of the tax year. In reality,the AT handles refund very slowly and <strong>in</strong>efficiently, result<strong>in</strong>g <strong>in</strong> long delays <strong>in</strong> payment or nopayment at all. This problem creates serious <strong>in</strong>equities, and contributes to a perception ofunfairness that underm<strong>in</strong>es <strong>in</strong>centives for tax compliance. Apart from devot<strong>in</strong>g more personnel


70 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEand resources to handl<strong>in</strong>g claims for <strong>in</strong>come tax refunds, the AT could facilitate these paymentsthrough two measures that are warranted <strong>in</strong> their own right (for reasons discussed elsewhere <strong>in</strong>this report): reduc<strong>in</strong>g the withhold<strong>in</strong>g tax rate; and develop<strong>in</strong>g risk-based management of refundclaims with immediate settlement of low-risk declarations, subject to selective audit. Anothermeasure that might be considered, though it would entail m<strong>in</strong>or <strong>in</strong>equities of its own, would be tosimplify the system by disallow<strong>in</strong>g refund claims below a certa<strong>in</strong> threshold (as done <strong>in</strong> the case ofVAT refunds). This would probably require an amendment to the law.Consumption <strong>Tax</strong>esTurn<strong>in</strong>g to taxes on consumption, the most important equity effects of the tax system <strong>in</strong>volve theVAT, because it is the largest source of revenue for the government. There is a widespread viewthat the VAT, as a broad-based tax on consumption, is <strong>in</strong>herently regressive. This view is notsupported by the data on household consumption patterns.Table 5-7 presents a breakdown of rural and urban expenditure patterns for rural and urbanhouseholds by <strong>in</strong>come qu<strong>in</strong>tile, based on 2002/03 data. For rural households, between 52 and 61percent of total expenditure is auto-consumption across all qu<strong>in</strong>tiles, and therefore not subject totax. In addition, there is little or no VAT on important items like farm products, wheat flour,health and education expenditures, and kerosene for light<strong>in</strong>g. Thus, for the bottom two ruralqu<strong>in</strong>tiles, only about 12 percent of household expenditures are potentially subject to VAT, ris<strong>in</strong>gto 23 percent for the top rural qu<strong>in</strong>tile. This implies a maximum VAT burden of around 2 percentof <strong>in</strong>come for the poorest qu<strong>in</strong>tiles. Even this figure is too high to the extent that poor householdspurchase taxable goods from small enterprises that qualify for VAT exemption. Figure 5-8converts the data to visual form, show<strong>in</strong>g the share of expenditures potentially subject to VAT forboth rural and urban households, by <strong>in</strong>come qu<strong>in</strong>tile. The graph demonstrates that the VATstructure <strong>in</strong> <strong>Mozambique</strong> is progressive, because higher shares of expenditures are subject toVAT for households <strong>in</strong> the top qu<strong>in</strong>tiles.A similar analysis applies to the equity effects of trade taxes on consumer goods. While importduties affect the prices and availability of many basic goods, the l<strong>in</strong>k is very weak for ruralhouseholds, which devote a large share of expenditure to auto-consumption or the purchase oflocally produced goods. The effect of trade taxes on household welfare is stronger <strong>in</strong> urban areas,where market purchases of imported goods are far more important. 62Sonne-Schmidt (2009) provides further <strong>in</strong>formation on the equity effects of <strong>in</strong>direct taxes. Thisstudy estimates for selected goods the comb<strong>in</strong>ed effect of import duties, VAT and excise taxes onhouseholds at different <strong>in</strong>come levels, us<strong>in</strong>g household expenditure data. Sonne-Schmidt f<strong>in</strong>dsthat the import duty on flour and tomatoes is highly regressive, despite a VAT exemption, andthat taxes on kerosene, sugar, and tobacco have a significant adverse impact on the poor, due to62 This issue is discussed <strong>in</strong> detail <strong>in</strong> Chapter 2 of Nathan Associates (2004), Diagnostic TradeIntegration Study for <strong>Mozambique</strong>.


I MPACT OF T AX S YSTEM 71the relative importance of these items <strong>in</strong> their consumption basket. 63 In contrast, he f<strong>in</strong>ds thattaxes on automobile fuel, bottled gas, w<strong>in</strong>e and beer have a progressive effect, due to the lowshare of spend<strong>in</strong>g on these products by poor household. 64 More studies of this sort are needed toprovide better <strong>in</strong>formation on the poverty impact of the tax system.Table 5-7Rural and Urban Expenditure Share by Qu<strong>in</strong>tileProduct Group Rural UrbanQI Q<strong>II</strong> Q<strong>II</strong>I QIV QV QI Q<strong>II</strong> Q<strong>II</strong>I QIV QVFood: Autoconsumption46.0 52.1 52.8 54.0 47.4 11.7 16.6 16.2 12.4 4.6Food: purchased 15.8 15.2 14.9 13.2 15.5 39.1 36.3 35.2 31.9 24.3Hous<strong>in</strong>g & fuel: AutoconsumptionHous<strong>in</strong>g and fuel:Purchased11.4 8.3 6.8 6.8 4.8 7.1 4.3 3.5 1.7 0.514.1 11.1 10.2 8.5 8.4 24.6 23.3 21.7 25.1 28.4Beverages 0.8 1.0 1.0 1.7 1.7 0.6 1.0 1.1 1.8 2.5Cloth<strong>in</strong>g 2.3 3.7 4.6 6.3 7.8 2.8 3.7 6.0 6.5 8.7Furniture anddecoration6.1 5.1 5.4 5.5 7.3 6.6 6.7 7.5 9.7 14.6Health 0.9 0.7 0.6 0.7 0.7 1.0 1.1 1.2 1.5 1.5Transport 1.0 1.3 1.8 1.7 3.8 2.7 3.1 3.1 4.0 6.1Communication 0.0 0.0 0.0 0.1 0.1 0.1 0.3 0.5 0.8 2.0Recreation and leisure 0.5 0.3 0.5 0.3 0.4 0.8 0.7 0.6 0.9 1.8Education 0.3 0.1 0.1 0.1 0.1 0.8 0.7 0.6 0.6 1.1Restaurants, hotels 0.4 0.6 0.7 0.7 0.9 0.9 0.7 1.0 1.8 1.6Other goods andservices0.3 0.5 0.4 0.5 1.1 1.1 1.3 1.8 1.5 2.3Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0Note: Qu<strong>in</strong>tile breakdowns computed by MPF differ from those reported by INE because the former use regional deflators tonormalize expenditure levels.SOURCE: MPF calculations based on IAF 2002/03 household survey data.63 Sugar is subject to import duties of 7.5 percent (putt<strong>in</strong>g it <strong>in</strong> the category of <strong>in</strong>termediate good), VATand a sugar surtax rang<strong>in</strong>g from 90 percent to 100 percent <strong>in</strong> order to provide protection to domestic sugarproducers.64 This is <strong>in</strong> broad agreement with the Fuel <strong>Tax</strong> Poverty and Social Impact Analysis (PSIA) carried out<strong>in</strong> 2003 which found that “the aggregate short-term impact of a rise <strong>in</strong> fuel tax on poverty is modest”(Nicholson et al., 2003), while there was some qualitative evidence of pockets of longer-term impact on“vulnerable activities” such as rural transport and market<strong>in</strong>g, gra<strong>in</strong> mill<strong>in</strong>g, and fish<strong>in</strong>g.


72 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEFigure 5-8Share of Rural and Urban Expenditures Potentially Subject to VATSOURCE: Authors’ calculations based on IAF 2002/03 household survey data.


6. Priority IssuesThis chapter identifies twelve issues that emerge from the analysis <strong>in</strong> previous chapters asmedium-term priorities for the next <strong>PARPA</strong> period, and offers recommendations for address<strong>in</strong>geach. 65 The priorities <strong>in</strong>clude the overall revenue target, several policy and process issues, and aset of <strong>in</strong>ter-related requirements for moderniz<strong>in</strong>g tax adm<strong>in</strong>istration, which overlap <strong>in</strong> manyrespects with the AT’s agenda for reform.MEDIUM-TERM REVENUE TARGETThe strategic objective <strong>in</strong> <strong>PARPA</strong> <strong>II</strong> for tax policy is “to reform and <strong>in</strong>crease the efficiency of taxadm<strong>in</strong>istration with a view to gradually <strong>in</strong>creas<strong>in</strong>g the mobilization of domestic funds as apercentage of GDP, with the idea of reduc<strong>in</strong>g external dependency” (paragraph 489). Morespecifically, <strong>PARPA</strong> <strong>II</strong> targeted a revenue ratio of 16.2 percent of GDP by 2009. Look<strong>in</strong>g aheadto the next <strong>PARPA</strong> period, determ<strong>in</strong>ation of a medium-term revenue target is aga<strong>in</strong> a centralstrategic issue for tax policy. This is <strong>in</strong>herently a political decision, but one that should be<strong>in</strong>formed by an analysis of economic conditions and the trade-offs <strong>in</strong>volved <strong>in</strong> levy<strong>in</strong>g taxes. In<strong>in</strong>terviews for this study, some government officials contended that the government shouldcont<strong>in</strong>ue aim<strong>in</strong>g at an <strong>in</strong>crease <strong>in</strong> revenue by 0.5 percentage po<strong>in</strong>ts of GDP per year up to theSADC average, now 23 percent (us<strong>in</strong>g the median). This target is consistent with estimates of themaximum potential revenue yield from two IMF technical studies <strong>in</strong> 2004 and 2006. Otherofficials, however, expressed concerns about the economic effects of squeez<strong>in</strong>g the private sectorto achieve such a large <strong>in</strong>crease <strong>in</strong> the revenue ratio.As discussed <strong>in</strong> Chapter 4, the SADC average is largely irrelevant for <strong>Mozambique</strong> due tostructural differences <strong>in</strong> the economies of member states. International benchmarks suggest that amedium-term target <strong>in</strong> the range of 18 to 19 percent of GDP is the maximum that should beexpected. This target should be atta<strong>in</strong>able <strong>in</strong> view of recent reforms to broaden the tax base,cont<strong>in</strong>u<strong>in</strong>g efforts to modernize tax adm<strong>in</strong>istration, and the underly<strong>in</strong>g elasticity of the tax system<strong>in</strong> response to economic growth. And yet one must also recognize that the optimal tax ratio maybe lower than the maximum potential ratio (which was the object of the IMF studies). Just asgovernment borrow<strong>in</strong>g crowds out f<strong>in</strong>anc<strong>in</strong>g for the private sector, boost<strong>in</strong>g the revenue ratio alsodraws resources away from both enterprises and households. The objective of tax policy shouldbe to balance public and private requirements <strong>in</strong> the <strong>in</strong>terest of national development, jobcreation, and susta<strong>in</strong>able growth. The government should therefore use a portion of the65 This report is not <strong>in</strong>tended to provide a comprehensive analysis of the agenda for tax reforms.Accord<strong>in</strong>g to several sources, there are plans for a comprehensive review of the tax system <strong>in</strong> 2010 throughtechnical assistance from the IMF and the World Bank.


74 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEprospective ga<strong>in</strong>s <strong>in</strong> revenue capacity to reduce tax rates, rather than focus<strong>in</strong>g exclusively on<strong>in</strong>creas<strong>in</strong>g the revenue ratio.The extent and tim<strong>in</strong>g of ga<strong>in</strong>s <strong>in</strong> revenue capacity will depend on the implementation of taxreforms, the pace of global economic recovery, and the flow of new <strong>in</strong>vestment to <strong>Mozambique</strong>.Consider<strong>in</strong>g the effects of the depressed global economy, the latest IMF estimates (from May2009) project that the revenue ratio will not <strong>in</strong>crease significantly for the next few years.Recommendation. Modify the strategic objective for tax policy from <strong>PARPA</strong> <strong>II</strong> as follows: “toreform and <strong>in</strong>crease the efficiency of the tax adm<strong>in</strong>istration with a view to gradually <strong>in</strong>creas<strong>in</strong>gdomestic revenue as a percentage of GDP to reduce external dependency, while also reduc<strong>in</strong>g taxrates affect<strong>in</strong>g private sector development.”Recommendation. Consider three options for the medium-term revenue target, depend<strong>in</strong>g onpolitical decisions on the balance between higher revenues and lower tax rates:• If the central objective is to <strong>in</strong>crease revenue, an ambitious upper-bound target for 2015would be approximately 18.5 percent of GDP. This option cont<strong>in</strong>ues the recent practiceof target<strong>in</strong>g a revenue <strong>in</strong>crease of around 0.5 percentage po<strong>in</strong>ts of GDP per year.• If the objective is to balance revenue ga<strong>in</strong>s with tax relief for the private sector, aplausible mid-range target for 2015 would be 17.5 percent of GDP. 66• If the objective is to use base-broaden<strong>in</strong>g measures and revenue enhancements to lowertax rates, a reasonable lower-bound target would be to stabilize the revenue ratio at 16.5percent of GDP.Recommendation. Medium-term targets are valuable guidel<strong>in</strong>es, but revenue targets for eachbudget year should be determ<strong>in</strong>ed from an analysis of prevail<strong>in</strong>g economic conditions and therevenue effects of concurrent and recent measures to reform tax policy and tax adm<strong>in</strong>istration.TAX RATESThe call for tax cuts is a common theme of policy discussions with the private sector, who viewthe prospect of lower tax rates as a major objective of measures to broaden the tax base andstrengthen tax adm<strong>in</strong>istration. As the headl<strong>in</strong>e issue, the private sector would like to see areduction <strong>in</strong> the standard rates for <strong>in</strong>come tax. Empirical evidence on the effect of a lower tax rateis actually <strong>in</strong>conclusive. The World Bank’s ICA survey <strong>in</strong> 2008 found that the majority ofrespondents did not view tax rates as a major or severe problem. International benchmarks showthat the tax on company <strong>in</strong>come is not especially high. In contrast, studies of the marg<strong>in</strong>aleffective tax rate (METR) on <strong>in</strong>vestment show it is very high for <strong>in</strong>vestors who do not enjoyspecial tax breaks under the code of fiscal benefits (though very competitive for those who doqualify). In any case, the quality of the bus<strong>in</strong>ess environment is far more important than tax ratesas a determ<strong>in</strong>ant of most <strong>in</strong>vestment decisions. Consequently, moderate tax cuts are unlikely tospur a wave of new <strong>in</strong>vestment. And yet a lower <strong>in</strong>come tax rate would directly improve net66 This option is consistent with the revenue assumption used by the IMF <strong>in</strong> their most recent DebtSusta<strong>in</strong>ability Analysis for <strong>Mozambique</strong>. See IMF (2009b), p. 60


P RIORITY I SSUES 75earn<strong>in</strong>gs for most enterprises and provide an important resource for expansion, <strong>in</strong>novation and jobcreation. Furthermore, to compete for mobile capital, <strong>Mozambique</strong> should not lag too far beh<strong>in</strong>dthe global and regional trend towards lower taxes.The analysis <strong>in</strong> Chapter 5 (Private Sector Development) suggests that two technical tax cuts takepriority over general reductions <strong>in</strong> the tax rate: reduc<strong>in</strong>g the withhold<strong>in</strong>g tax rate, to facilitate thedevelopment of l<strong>in</strong>kages between large and small enterprises; and reduc<strong>in</strong>g or elim<strong>in</strong>at<strong>in</strong>g thedouble taxation of dividend payments to <strong>in</strong>dividual shareholders, to lower the METR on new<strong>in</strong>vestments subject to the standard tax regime. These two measures target significant problemsfor the private sector, at a relatively low revenue cost.Because the VAT is the primary source of domestic revenue, the cost of reduc<strong>in</strong>g this tax ratewould be high. Furthermore, the VAT (as structured) falls primarily on non-poor households, noton bus<strong>in</strong>esses.Recommendation. To avoid revenue risks, package any tax cuts with measures to enhancerevenues, or calibrate the tax cuts so that the revenue costs are covered by expected ga<strong>in</strong>s fromearlier reforms. 67Recommendation. Reduce or elim<strong>in</strong>ate the double taxation of dividend <strong>in</strong>come paid to<strong>in</strong>dividual shareholders, tak<strong>in</strong>g <strong>in</strong>to account the effects on revenue.Recommendation. Reduce the withhold<strong>in</strong>g tax rate to 10 percent on all transactions currentlysubject to a 20 percent rate, tak<strong>in</strong>g <strong>in</strong>to account the effects on revenue.Recommendation. Target a reduction <strong>in</strong> the company <strong>in</strong>come tax (IRPC) and the <strong>in</strong>dividual<strong>in</strong>come tax (IRPS) as revenue prospects permit, to improve the earn<strong>in</strong>gs and growth prospects forboth corporate and non-corporate enterprises.TAX POLICY ANALYSISAll decisions on tax policy should be <strong>in</strong>formed by a careful quantitative analysis of the revenueeffects and at least a qualitative analysis of the economic impacts. This <strong>in</strong>cludes the decisions onrevenue targets for the budget program each year. The MOF therefore requires the capability toundertake serious tax policy analysis. The standard approach is to establish a <strong>Tax</strong> Policy Unit(under various names and organizational arrangements) staffed by a small group of economistsand tax adm<strong>in</strong>istration specialists. The mandate for the TPU can <strong>in</strong>clude: produc<strong>in</strong>g revenueforecasts for the annual budget cycle and revenue projections for the medium-term fiscalframework; analyz<strong>in</strong>g the fiscal, economic and distributional effects of tax policy issues underconsideration; review<strong>in</strong>g and analyz<strong>in</strong>g the current tax system to make recommendations forimprov<strong>in</strong>g its effectiveness <strong>in</strong> generat<strong>in</strong>g revenues and facilitat<strong>in</strong>g an improved bus<strong>in</strong>ess climate;67 <strong>Tax</strong> reform packages are often designed to be revenue neutral, <strong>in</strong> which case the ga<strong>in</strong>s from revenueenhancement measures are fully passed along to the private sector <strong>in</strong> the form of tax cuts. One commonmethod for f<strong>in</strong>anc<strong>in</strong>g tax cuts is by elim<strong>in</strong>at<strong>in</strong>g or reduc<strong>in</strong>g the scope of special tax <strong>in</strong>centives or taxpreferences. Given that <strong>Mozambique</strong> recently adopted a new Code of Fiscal Benefits, however, majorchanges <strong>in</strong> the near term are best avoided <strong>in</strong> the <strong>in</strong>terest of stability and predictability <strong>in</strong> the tax system.


76 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEand serv<strong>in</strong>g as the government’s focal po<strong>in</strong>t for dialogue with the public on tax policy issues,<strong>in</strong>clud<strong>in</strong>g the dissem<strong>in</strong>ation of data reports and policy briefs. The TPU should collaborate closelywith the Gab<strong>in</strong>ete de Planeamento e Estúdos at the AT, with the latter focus<strong>in</strong>g on monitor<strong>in</strong>gand assess<strong>in</strong>g revenue performance, ma<strong>in</strong>ta<strong>in</strong><strong>in</strong>g and develop<strong>in</strong>g the tax data base, and provid<strong>in</strong>ganalytical support to AT management.Recommendation. Establish a <strong>Tax</strong> Policy Unit (TPU) at the M<strong>in</strong>istry of F<strong>in</strong>ance, possibly with<strong>in</strong>the Gab<strong>in</strong>ete de Estudos, and tra<strong>in</strong> a group of specialists to conduct quantitative and qualitativeanalysis of tax policy issues. The TPU should have primary responsibility for revenueforecast<strong>in</strong>g, technical analysis of tax policy issues, and dialogue with the public on tax policyissues.FUNCTIONAL INTEGRATION OF TAX OPERATIONSThe modernization of tax adm<strong>in</strong>istration is a work <strong>in</strong> progress. The AT has achieved majorimprovements <strong>in</strong> tax adm<strong>in</strong>istration, with highly favorable effects on revenue mobilization. Yetlarge additional ga<strong>in</strong>s <strong>in</strong> efficiency and effectiveness can be obta<strong>in</strong>ed through further steps tomodernize tax adm<strong>in</strong>istration, improve compliance, and reduce the scope for tax evasion andunethical practices. These adm<strong>in</strong>istrative reforms will also facilitate private sector developmentand improve the bus<strong>in</strong>ess environment for small and medium enterprises by simplify<strong>in</strong>gprocedures, lower<strong>in</strong>g compliance costs, and reduc<strong>in</strong>g glar<strong>in</strong>g <strong>in</strong>equities caused by uneven andarbitrary enforcement of the tax laws.One major requirement – the first of five to be cited here – is the functional <strong>in</strong>tegration ofcommon operations between customs and domestic tax services, <strong>in</strong>clud<strong>in</strong>g (<strong>in</strong> rough order ofpriority) audit, risk management, debt management, and customer services, as well as supportfunctions such as human resource management and an master files for each taxpayer cover<strong>in</strong>g alltax and customs transactions. The purpose of consolidat<strong>in</strong>g services is to reduce duplication ofeffort with<strong>in</strong> the AT, strengthen synergies across different tax l<strong>in</strong>es, and provide more seamlessservices to taxpayers. As discussed <strong>in</strong> chapter 4, progress on functional <strong>in</strong>tegration has beenlimited to date. Further action <strong>in</strong> this direction is part of the AT’s strategic plan.Recommendation. Formulate and implement specific plans for <strong>in</strong>tegrat<strong>in</strong>g basic operationalfunctions of the customs and tax departments of the AT, focus<strong>in</strong>g first on audit, risk managementand debt management, and customer services.Recommendation. Formulate and implement specific plans for further <strong>in</strong>tegration of supportoperations <strong>in</strong> the AT, <strong>in</strong>clud<strong>in</strong>g compatible IT systems, a unified master data base for eachtaxpayer, and comprehensive systems for human resource management.E-TAXATIONThe government and the AT recognize the importance of <strong>in</strong>troduc<strong>in</strong>g modern <strong>in</strong>formationtechnologies <strong>in</strong>to the tax system (e-Tributação), and have already begun the process of plann<strong>in</strong>gand implement<strong>in</strong>g these reforms. The ma<strong>in</strong> focal po<strong>in</strong>ts <strong>in</strong>clude <strong>in</strong>tegration of revenue accounts<strong>in</strong>to the government’s automated public f<strong>in</strong>ance management system (e-SISTAFE), <strong>in</strong>troduction


P RIORITY I SSUES 77of electronic tax declarations and electronic payments through the bank<strong>in</strong>g system, andimplementation of electronic s<strong>in</strong>gle w<strong>in</strong>dow systems to facilitate border clearances.Compared to exist<strong>in</strong>g processes, IT-based systems offer enormous potential for enhanc<strong>in</strong>grevenue by improv<strong>in</strong>g the efficiency of AT operations, while reduc<strong>in</strong>g compliance costs for theprivate sector. But caution is very much <strong>in</strong> order. The reforms must be carefully planned andsequenced, well managed, properly funded, and effectively coord<strong>in</strong>ated with technical support, toavoid potentially serious pitfalls that could impede ongo<strong>in</strong>g revenue operations, or lock <strong>in</strong>systems that will not meet future needs (as <strong>in</strong> the case of the TIMS system now be<strong>in</strong>g used byCustoms). It is extremely important, too, for the AT to use e-taxation as a catalyst forfundamental re-eng<strong>in</strong>eer<strong>in</strong>g of basic bus<strong>in</strong>ess and work-flow processes, <strong>in</strong>stead of just automat<strong>in</strong>gexist<strong>in</strong>g systems. F<strong>in</strong>ally, the transition to e-taxation may require supportive legal or regulatorychanges. 68Recommendation. Proceed expeditiously with the <strong>in</strong>troduction of e-taxation systems, but withdiligent attention to the design, plann<strong>in</strong>g, sequenc<strong>in</strong>g, and management of the change process. Toimprove efficiency and effectiveness of the AT’s revenue operations, priorities <strong>in</strong>clude e-declarations, e-payments through the banks, the automated s<strong>in</strong>gle w<strong>in</strong>dow for customs, as well asautomat<strong>in</strong>g the systems for deal<strong>in</strong>g with stop-filers and late-filers, and systems for riskmanagement.Recommendation. Use the <strong>in</strong>troduction of e-taxation as a catalyst for fundamental reeng<strong>in</strong>eer<strong>in</strong>gof bus<strong>in</strong>ess and work-flow processes <strong>in</strong> each of the areas covered.Recommendation. Carefully review the legal and regulatory environment to ensure that allrequired revisions are <strong>in</strong> to support e-taxation, <strong>in</strong>clud<strong>in</strong>g not only tax laws but also laws relat<strong>in</strong>gto electronic <strong>in</strong>formation and transactions.RISK MANAGEMENTIn conjunction with the <strong>in</strong>tegration of tax and customs operations and the <strong>in</strong>troduction of e-taxsystems, there is also enormous room for improvement <strong>in</strong> collection efficiency through the<strong>in</strong>troduction of modern risk management practices – which will simultaneously reduce thecompliance burden and facilitate tax transactions for most taxpayers. The concept is to usecomputerized data records and automated statistical systems to screen for cases with high revenuerisk, as well as those with low revenue risk, enabl<strong>in</strong>g the AT to focus its resources moreproductively on cases where the potential revenue ga<strong>in</strong>s are highest. For cases with low revenuerisk, most tax and customs procedures can and should be handled quickly and at m<strong>in</strong>imum cost,effectively support<strong>in</strong>g broader application of automatic “green channel” approvals and “goldcard” treatment of taxpayers with excellent compliance records—<strong>in</strong> the <strong>in</strong>terest of collectionefficiency. By comparison, exist<strong>in</strong>g risk management systems are rudimentary, subjective, and68 See, for example, Nathan Associates (2008), ASEAN S<strong>in</strong>gle W<strong>in</strong>dow: The Intersection of Law &Technology, for a detailed discussion of legal technicalities associates with the <strong>in</strong>troduction of AutomatedS<strong>in</strong>gle W<strong>in</strong>dow processes. The issues <strong>in</strong>clude not only enabl<strong>in</strong>g legislation for electronic transactions, butalso laws relat<strong>in</strong>g to the protection and storage of personal data, cross-border authentication of electronicsignatures, and electronic transfer of rights <strong>in</strong> goods.


78 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEhighly <strong>in</strong>efficient. Even low-risk cases must face the chance of audit, and serious penalties ifmajor <strong>in</strong>fractions are detected; but the audit rate for this group should be low.Recommendation. Once taxpayer data systems are <strong>in</strong> a condition to support statistical riskanalysis, seek technical assistance for the development of automated risk-management systems.Priorities for modern risk management <strong>in</strong>clude customs clearances, audit selection, and thepayment of VAT and <strong>in</strong>come tax refunds.Recommendation. Expand the application of “green channel” and “gold card” treatment forclearances and approvals with low revenue risk as identified by the risk management systemsSIMPLIFICATIONTwo of eight “priority actions” for tax policy <strong>in</strong> <strong>PARPA</strong> <strong>II</strong> dealt with simplification. One calledfor a review and evaluation of the simplified tax regimes under the <strong>in</strong>come tax and VAT, whilethe second called for approval of legislation “that simplifies the relationship between the taxadm<strong>in</strong>istration and the taxpayers, mak<strong>in</strong>g it easier for them to exercise their rights and receive theprotection assured them” (paragraph 489). These aims have been realized on paper at leastthrough adoption of the 2009 ISPC code for small taxpayers and the 2006 General Law on<strong>Tax</strong>ation – though the implementation of both laws is still work <strong>in</strong> progress. In addition, the AThas simplified tax forms and greatly improved public <strong>in</strong>formation campaigns. The Authority isalso <strong>in</strong>vest<strong>in</strong>g heavily <strong>in</strong> expansion of the physical <strong>in</strong>frastructure <strong>in</strong> order to make it easier fortaxpayers outside major cities to comply with tax obligations and seek assistance.Despite this genu<strong>in</strong>e progress, simplification rema<strong>in</strong>s high on the agenda of key issues for taxreform. Complexities <strong>in</strong> the tax code, <strong>in</strong> tax forms, and <strong>in</strong> adm<strong>in</strong>istrative procedures cont<strong>in</strong>ue tocreate serious problems, especially for small and medium size enterprises with weak f<strong>in</strong>ancialmanagement and lack of access to professional assistance. <strong>Tax</strong>payers who, as a result, make<strong>in</strong>advertent errors are vulnerable to arbitrary reassessments and burdensome penalties – arecurrent theme <strong>in</strong> <strong>in</strong>terviews for this study. Complicated documentation requirements andapproval procedures also cont<strong>in</strong>ue to bog down the process of obta<strong>in</strong><strong>in</strong>g VAT refunds, despiteimprovements <strong>in</strong> recent years. The refund process also functions poorly for the IRPC and IRPS.Similar problems arise from the complex rules and cumbersome procedures <strong>in</strong>volved <strong>in</strong> obta<strong>in</strong><strong>in</strong>gcertification for preferential tariff treatment under the SADC trade protocol and the emerg<strong>in</strong>gSADC free trade area. These problems affect both importers and exporters.The <strong>in</strong>troduction of e-taxation and modern risk management systems should simplify taxcompliance and reduce payment problems for many taxpayers. Additional measures can also beconsidered to simplify the system, such as elim<strong>in</strong>at<strong>in</strong>g end-year reconciliations under the IRPSfor households whose sole source of <strong>in</strong>come is from formal sector employment, even if more thanone member of the family is work<strong>in</strong>g.Recommendation. Ensure effective implementation of the simplified tax for small contributors(ISPC) and taxpayer protection provisions of the General Law on <strong>Tax</strong>ation.Recommendation. Seek and pursue further opportunities to simplify tax forms and taxprocedures, with special attention to the refund process for both VAT and the <strong>in</strong>come tax, through


P RIORITY I SSUES 79the application of <strong>in</strong>formation technology and efficient risk management procedures. A review ofthe legal and regulatory environment may also be needed to support simplification reforms.Recommendation. Simplify the procedures for certify<strong>in</strong>g rules of orig<strong>in</strong> (ROO) on exports toother SADC states. For example, ROO documents should be available <strong>in</strong> the prov<strong>in</strong>ces, not only<strong>in</strong> Maputo. The government should also negotiate with SADC partners to simplify ROOcertification on exports to <strong>Mozambique</strong>, so that importers <strong>in</strong> this country can benefit from thepreferential tariff rates.TAXPAYER SERVICESThe AT has been made impressive progress <strong>in</strong> improv<strong>in</strong>g taxpayer services through <strong>in</strong>novativepublic <strong>in</strong>formation and education programs, new tax service and call centers, geographicalexpansion of facilities, greater delegation of authority to regional offices (such as <strong>in</strong> check<strong>in</strong>gVAT refund documents for accuracy before transmission to Maputo), development of thewebsite, and upgrad<strong>in</strong>g human resources to provide taxpayers with better support and moreobjective treatment. These are important steps <strong>in</strong> the direction of facilitat<strong>in</strong>g compliance, help<strong>in</strong>gtaxpayers avoid costly errors, and improv<strong>in</strong>g taxpayer attitudes about the AT and their taxobligations.But the process of improv<strong>in</strong>g taxpayer services is far from complete. Better provision of<strong>in</strong>formation is a central part of the problem. This <strong>in</strong>cludes not only cont<strong>in</strong>ued media coverage ofbasic facts about the tax system, but also <strong>in</strong>creased capacity with<strong>in</strong> the AT to provide clear,consistent and correct answers to questions from taxpayers at all levels of sophistication about thetax code or procedural requirements. Many technical ambiguities <strong>in</strong> the tax code have not yetbeen clarified through subsidiary regulation, and several tax specialists told the study team that itis very difficult (or impossible) to obta<strong>in</strong> written rul<strong>in</strong>gs or <strong>in</strong>terpretations from the AT.Another cont<strong>in</strong>u<strong>in</strong>g problem is the long-stand<strong>in</strong>g concern of the private sector aboutunpredictable and punitive enforcement practices by tax officers, especially <strong>in</strong> reaction to<strong>in</strong>significant errors or un<strong>in</strong>tended <strong>in</strong>fractions caused by lack of <strong>in</strong>formation about details of thetax code. This behavior creates an adversarial relationship, <strong>in</strong>centives for corrupt practices, andsources of dispute, where educational support from tax officials would be more productive allaround.Recommendation. Develop and dissem<strong>in</strong>ate specific plans for improv<strong>in</strong>g taxpayer services <strong>in</strong> themedium term, apply<strong>in</strong>g computer technology where possible to improve efficiency. The planmust <strong>in</strong>clude implementation benchmarks and results targets. It should cover public <strong>in</strong>formationprograms as well as direct services to taxpayers who require support.Recommendation. Ma<strong>in</strong>ta<strong>in</strong> an active file of frequently asked questions, with authoritativeanswers presented <strong>in</strong> language that can readily be understood by most taxpayers. Wherenecessary, formulate and issue support<strong>in</strong>g regulations. These <strong>in</strong>formation products should bemade available to taxpayers through the Internet, through easily operated computer kiosks <strong>in</strong> taxservice centers nationwide, and through cooperat<strong>in</strong>g non-government organizations.


80 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUERecommendation. Expand tra<strong>in</strong><strong>in</strong>g on taxpayer services for all tax officials who deal with thepublic, not just those assigned specifically to taxpayer service functions.Recommendation. Conduct regular surveys of taxpayer knowledge and satisfaction with AToperations, to provide management <strong>in</strong>formation for monitor<strong>in</strong>g the results of the taxpayer service<strong>in</strong>itiatives.TAX CULTUREMany of those <strong>in</strong>terviewed for the study emphasized the deep-seated weakness <strong>in</strong> the “taxculture” as a primary obstacle to improv<strong>in</strong>g compliance and curb<strong>in</strong>g fraud and evasion. Asidefrom the <strong>in</strong>herent tendency for <strong>in</strong>dividuals and bus<strong>in</strong>esses to favor “free-rid<strong>in</strong>g” behavior if theycan get away with it, the weak tax culture also reflects a widespread lack of understand<strong>in</strong>g of howtaxes contribute to national development. It is also <strong>in</strong>fluenced by the complexity of the taxsystem, and the limited geographical coverage of revenue services, which makes it hard to paytaxes <strong>in</strong> many regions of the country. Further fuel<strong>in</strong>g popular resistance to pay<strong>in</strong>g taxes is aperception of blatant tax evasion due to bribery and political favoritism.Strengthen<strong>in</strong>g the tax culture requires a comb<strong>in</strong>ation of carrots and sticks to strengthen <strong>in</strong>centivesfor voluntary compliance and <strong>in</strong>crease the costs (or reduce the benefits) of cheat<strong>in</strong>g. Actionsaddress<strong>in</strong>g the priority issues discussed above will shift the balance of considerations <strong>in</strong> thedirection of compliance. But a more focused effort can hasten the cultural transformation. Thiscan be done most effectively through measures that resonate with the public and underscore theimportance of pay<strong>in</strong>g taxes, such as the two that follow. To be sure, these approaches are easiersaid than done, and require high-level political back<strong>in</strong>g.Recommendation. Crack down on affluent taxpayers who have been escap<strong>in</strong>g their taxobligations with impunity. The AT has legal authority to pursue and catch blatant tax evadersthrough <strong>in</strong>direct assessments based on external signs of wealth and third party data such asautomobile registrations. Where possible, consistent with tax privacy laws, make an example ofpenaliz<strong>in</strong>g or prosecut<strong>in</strong>g these cases.Recommendation. Require senior members of the government and members of the nationalassembly to submit <strong>in</strong>come and wealth declarations to an <strong>in</strong>dependent audit authority and to theAT.TAX TRAININGThere is an enormous need for expand<strong>in</strong>g and improv<strong>in</strong>g tax tra<strong>in</strong><strong>in</strong>g for both AT cadres and taxprofessionals serv<strong>in</strong>g the private sector. The AT has already embarked on an ambitious programto strengthen staff tra<strong>in</strong><strong>in</strong>g, <strong>in</strong>clud<strong>in</strong>g the establishment <strong>in</strong> 2009 of a new Tra<strong>in</strong><strong>in</strong>g Institute <strong>in</strong>Matola. The tra<strong>in</strong><strong>in</strong>g program should be developed on the basis of a thorough assessment oftra<strong>in</strong><strong>in</strong>g needs and a clear plan for prioritiz<strong>in</strong>g and sequenc<strong>in</strong>g tra<strong>in</strong><strong>in</strong>g activities, tak<strong>in</strong>g <strong>in</strong>toaccount the benefits <strong>in</strong> revenue collection and customer services, as well as skill requirements forimplementation of the modernization program (as discussed above). From a revenue perspective,tra<strong>in</strong><strong>in</strong>g <strong>in</strong> advance audit skills for deal<strong>in</strong>g with large taxpayers is especially important.


P RIORITY I SSUES 81The private sector likewise suffers from an acute shortage of well tra<strong>in</strong>ed tax professionals. Largecompanies, of course, have few problems <strong>in</strong> this respect. In addition, the new ISPC should allowvery small enterprises to manage their tax affairs without engag<strong>in</strong>g accountants or tra<strong>in</strong>edbookkeepers. For many other SMEs, especially outside of Maputo, professional services for taxsupport and advice are either unavailable or unaffordable. In large parts of the country, it isliterally impossible for a medium-sized bus<strong>in</strong>ess to submit properly certified accounts for taxpurposes, due to the absence of account<strong>in</strong>g services. This situation <strong>in</strong>creases the costs and risks ofdo<strong>in</strong>g bus<strong>in</strong>ess, because enterprises without properly signed accountant are vulnerable to heavypenalties. The problem <strong>in</strong>volves the quality of tax tra<strong>in</strong><strong>in</strong>g as well as the quantity.Recommendation. Conduct a thorough tra<strong>in</strong><strong>in</strong>g needs assessment for the AT. On that basisestablish a tra<strong>in</strong><strong>in</strong>g plan with priorities and targets for the short term and medium term, <strong>in</strong>clud<strong>in</strong>gformal tra<strong>in</strong><strong>in</strong>g courses, structured on-the-job tra<strong>in</strong><strong>in</strong>g – and tra<strong>in</strong><strong>in</strong>g of tra<strong>in</strong>ers. Priorities arelikely to <strong>in</strong>clude advanced audit for deal<strong>in</strong>g with large taxpayers; taxpayer services; <strong>in</strong>tegrity <strong>in</strong>tax adm<strong>in</strong>istration; <strong>in</strong>ternal <strong>in</strong>vestigation; and new skills requirements relat<strong>in</strong>g to themodernization of systems and procedures.Recommendation. Devote far more attention and resources to expand<strong>in</strong>g and improv<strong>in</strong>g thequality of tra<strong>in</strong><strong>in</strong>g for accountants, <strong>in</strong>clud<strong>in</strong>g specialized courses on tax account<strong>in</strong>g.EITI IMPLEMENTATIONAlong with the adoption <strong>in</strong> 2007 of a new legal framework for <strong>in</strong>vestments <strong>in</strong> the m<strong>in</strong><strong>in</strong>g andpetroleum sectors, and elim<strong>in</strong>ation of special negotiations on the tax treatment of future megaprojects<strong>in</strong> the 2009 Code of Fiscal Benefits, the government’s formal application <strong>in</strong> May 2009 tojo<strong>in</strong> the Extractive Industries Transparency Initiative (EITI) is a landmark commitment torevenue mobilization from the exploitation of m<strong>in</strong>eral resources, as well as fiscal transparency. Inbrief, the EITI is a mechanism requir<strong>in</strong>g, first, that companies <strong>in</strong>volved <strong>in</strong> oil, gas and m<strong>in</strong>eralproduction disclose their tax and royalty payments to a participat<strong>in</strong>g government; second, that thegovernment discloses the receipt of these payments; and third, that these payments are subject to<strong>in</strong>dependent verification.The formal application is one step <strong>in</strong> a multi-year validation process for full compliance with theEITI (see http://eitransparency.org). Ensur<strong>in</strong>g follow-through over the next few years is one ofthe most important revenue issues for the medium term. As an objective, it also has the advantageof be<strong>in</strong>g well def<strong>in</strong>ed and easy to monitor.Recommendation. Adhere to the required schedule for full compliance with the EITI.DONOR COORDINATIONMost of the priority reforms outl<strong>in</strong>ed above require f<strong>in</strong>ancial and technical support from<strong>in</strong>ternational partner agencies. Several donors are now channel<strong>in</strong>g f<strong>in</strong>ancial support through aCommon Fund that is controlled by the AT but monitored periodically by a Quality AssuranceGroup of <strong>in</strong>ternational tax and customs experts, who have been work<strong>in</strong>g with specialists from theIMF and the U.S. Treasury. The Common Fund streaml<strong>in</strong>es the relationship between the AT andthe donors by establish<strong>in</strong>g a s<strong>in</strong>gle set of procedures and controls. It also helps to strengthen theAT’s capacity for plann<strong>in</strong>g and manag<strong>in</strong>g reforms. Another coord<strong>in</strong>ation mechanism is the <strong>Tax</strong>


82 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEWork<strong>in</strong>g Group, which br<strong>in</strong>gs donors together to exchange <strong>in</strong>formation and discuss commonconcerns relat<strong>in</strong>g to tax and customs reforms. Recently, however, this group has not been veryactive.Look<strong>in</strong>g ahead, it appears that other donors will be provid<strong>in</strong>g assistance to the AT throughchannels outside the Common Fund framework. It is important for these donors to develop anddeliver assistance <strong>in</strong> a manner that m<strong>in</strong>imizes the burden on senior managers of the AT of cop<strong>in</strong>gwith a multiplicity of donor agency procedures and requirements.Recommendation. Revitalize the <strong>Tax</strong> Work<strong>in</strong>g Group as a primary mechanism for coord<strong>in</strong>at<strong>in</strong>gdonor support on tax and customs issues, with particular attention to ensur<strong>in</strong>g liaison betweenCommon Fund activities and other support programs. As far as possible, donors who chose tooperate outside the Common Fund should seek to develop mechanisms that m<strong>in</strong>imize theadm<strong>in</strong>istrative burden placed on the AT.


ReferencesArndt, C., F. Tarp and S. Jones, eds. 2009. <strong>Tax</strong>ation <strong>in</strong> a Low-Income Economy: The Case of<strong>Mozambique</strong>. New York. Routledge.Autoridade Tributária de Moçambique (ATM). 2008. Relatório Anual da Autoridade Tributáriade Moçambique–2007.____. 2009. Relatório de Actividades Desenvolvidas em 2008 e Perspectivas de Acções para2009.____. 2009a. Plano Estratégico 2009-2010.____. 2009b. Plano Táctico 2009.Bolnick, B. 2004. The Effectiveness and Economic Impact of <strong>Tax</strong> Incentives <strong>in</strong> the SADCRegion. Prepared by Nathan Associates for USAID under the SADC-TIFI Project._____. 2004a. <strong>Tax</strong> Reform and the Bus<strong>in</strong>ess Environment <strong>in</strong> <strong>Mozambique</strong>: A Review of Private-Sector Concerns. Prepared by Nathan Associates for USAID/<strong>Mozambique</strong> TechnicalAssistance to CTA and MIC Project.____. 2007. The Economic Costs of VAT Refund Delays <strong>in</strong> <strong>Mozambique</strong>. Prepared forUSAID/<strong>Mozambique</strong> Trade and Investment Project by Nathan Associates, Inc.____. 2009. Is the Flat <strong>Tax</strong> Right for <strong>Mozambique</strong>? Prepared for USAID/<strong>Mozambique</strong> Trade andInvestment Project by Nathan Associates, Inc.____. 2009a. The Motivation for Investments <strong>in</strong> <strong>Mozambique</strong>: The Role of Fiscal Incentives.Prepared for USAID/<strong>Mozambique</strong> Trade and Investment Project by Nathan Associates, Inc.Byiers, B. 2009. Enterprise Survey Evidence. In Arndt, Tarp and Jones.Collect<strong>in</strong>g <strong>Tax</strong>es Database. USAID/EGAT Fiscal Reform and Economic Governance Project.www.fiscalreform.netConfederação das Associações Económicas de Moçcambique (CTA). 2006. Opportunities of aGlobal Market: 9th Private Sector Annual Conference. Maputo.Centro de Promoção de Investimentos (CPI). 2009. Code of Fiscal Benefits, Law 4.2009 of 12January.


84 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUEDeloitte and ACIS. 2008. General Overview of the <strong>Tax</strong> <strong>System</strong> <strong>in</strong> <strong>Mozambique</strong>, Edition 1.____. 2008a. Legal Framework for <strong>Tax</strong>ation <strong>in</strong> <strong>Mozambique</strong>: Value Added <strong>Tax</strong> Manual, Edition1.____. 2009. Legal Framework for <strong>Tax</strong>ation <strong>in</strong> <strong>Mozambique</strong>: Manual on Corporate Income <strong>Tax</strong>.____. 2009. Legal Framework for <strong>Tax</strong>ation <strong>in</strong> <strong>Mozambique</strong>: Manual on Individual Income <strong>Tax</strong>.Ebrill, L., Keen, M., et al. 2002. The Allure of the Value-Added <strong>Tax</strong>. F<strong>in</strong>ance and Development,Vol. 39, No.2. Wash<strong>in</strong>gton DC: International Monetary Fund.Foreign Investment Advisory Service. 2006. Study on the Impact of <strong>Tax</strong>es, Customs, Licensesand Other Fees on the Investment Climate.Frei, A. Various dates. The Mozambican Investor. Maputo: MozLegal.Hassan, S. 2009. <strong>Tax</strong>ation and the Cost of Capital <strong>in</strong> <strong>Mozambique</strong>. Chapter 14 <strong>in</strong> Arndt, Tarp andJones.Instituto Nacional de Estatística (INE). 2004. CEMPRE: Censo das Empresas, Revisado 2004,[National Enterprise Census Data, Revised 2004].____. 2006. O Sector Informal em Moçambique: Resultados do Primeiro Inquérito Nacional(INFOR–2004). Direcção de Estatísticas Sectoriais e de Empresas.⎯⎯. 2009. Inquérito Annual ás Empresas – 2006: Pr<strong>in</strong>cipais Resultados. Presentation to PrivateSector Work<strong>in</strong>g Group, July 8. Maputo, <strong>Mozambique</strong>.International Monetary Fund. 2005. Republic of <strong>Mozambique</strong>: Article IV Consultation SecondReview. IMF Country Report No.05/318.____. 2005a. <strong>Tax</strong> and Customs Reforms <strong>in</strong> <strong>Mozambique</strong>: an Overview. In Republic of<strong>Mozambique</strong>: Selected Issues and Statistical Appendix. IMF Country Report No.05/311.⎯⎯. 2006. Republic of <strong>Mozambique</strong>: Fourth Review Under the Three-Year Arrangement Underthe Poverty Reduction and Growth Facility. IMF Country Report No. 06/254.____. 2007. Republic of <strong>Mozambique</strong>: Poverty Reduction Strategy Paper. IMF Country ReportNo. 07/37.⎯⎯. 2007. Republic of <strong>Mozambique</strong>: Fifth Review Under the Three-Year Arrangement Underthe Poverty Reduction and Growth Facility. IMF Country Report No.07/262.⎯⎯. 2007a. Republic of <strong>Mozambique</strong>: Selected Issues. IMF Country Report No.07/258.⎯⎯. 2008. Republic of <strong>Mozambique</strong>: Second Review Under the Policy Support Instrument.IMF Country Report No.08/220.


R EFERENCES 85⎯⎯. 2008a. Republic of <strong>Mozambique</strong>: First Review Under the Policy Support Instrument. IMFCountry Report No. 08/15.____. 2008b. Republic of <strong>Mozambique</strong>: Report on Observance of Standards and Codes—FiscalTransparency Module. IMF Country Report No.08/152.⎯⎯. 2009. Republic of <strong>Mozambique</strong>: Poverty Reduction Strategy Paper Annual ProgressReport. IMF Country Report No. 09/117.⎯⎯. 2009a. Republic of <strong>Mozambique</strong>: Third Review under the Policy Support Instrument. IMFCountry Report No.09/49.⎯⎯. 2009b. Staff Report for the 2009 Article IV Consultation, Fourth Review Under the PolicySupport Instrument. EBS/09/98. June 19 2009.___. 2009c. Moçambique: Acompanhamento do Processo de Consolidação da AutoridadeTributária de Moçambique, by A. Lembruber, A. Teixeira and J. Rachid. (For official useonly.)Jones, E.S., Arndt, C. 2007. Aid, Revenue and Fiscal Space: Considerations for Least DevelopedCountries with Application to <strong>Mozambique</strong>. Paper for the 2007 Nordic Conference <strong>in</strong>Development Economics, June 18-19, Copenhagen, Denmark.Jones, E.S. 2009. Comparative Perspective. In Arndt, Tarp and Jones.Kaufmann, D., Kraay, A., Mastruzzi, M. 2009. Governance Matters V<strong>II</strong>I: Aggregate andIndividual Governance Indicators 1996-2008. World Bank Policy Research Work<strong>in</strong>g PaperNo.4978. World Bank.Keene, M. 2008. VAT Tariffs and Withhold<strong>in</strong>g: Border <strong>Tax</strong>es and Informality <strong>in</strong> Develop<strong>in</strong>gCountries. Journal of Public Economics, Vol. 92, Issues 10-11.Klemm, A. 2009. Causes, Benefits and Risks of Bus<strong>in</strong>ess <strong>Tax</strong> Incentives. IMF Staff Work<strong>in</strong>gPaper WP/09/21.Nathan Associates Inc. 2004. Remov<strong>in</strong>g Obstacles to Economic Growth <strong>in</strong> <strong>Mozambique</strong>: ADiagnostic Trade Integration Study. Volume 2, Ma<strong>in</strong> Report.Nicholson, K., O’Laughl<strong>in</strong>, B., Francisco, A., Nhate, V. 2003. Fuel <strong>Tax</strong> <strong>in</strong> <strong>Mozambique</strong>: Povertyand Social Impact Analysis. UK Department for International Development.Republic of <strong>Mozambique</strong>. 2006. Action Plan for the Reduction of Absolute Poverty 2006-2009(<strong>PARPA</strong> <strong>II</strong>).⎯⎯. 2007. Estratégia para o Desenvolvimento das Pequenas e Médias Empresas emMoçambique, Aprovada pela 22ª Sessão Ord<strong>in</strong>ária do Conselho de M<strong>in</strong>istros em 21 deAgosto.


86 <strong>PARPA</strong> <strong>II</strong> R EVIEW—TAX S YSTEM IN M OZAMBIQUESonne-Schmidt, C. 2009. Household Survey Evidence. In Arndt , Tarp and Jones.Tribunal Adm<strong>in</strong>istrativo. 2008. Relatório e Parecer sobre a Conta Geral do Estado de 2007.United Nations Development Programme. 2007. Human Development Report 2007/2008.USAID. 2004. Diagnostic Trade Integration Study. Remov<strong>in</strong>g Obstacle to Economic Growth <strong>in</strong><strong>Mozambique</strong>.Varsano, R., et al. 2006. Post-Reform Evaluation of the <strong>Tax</strong> <strong>System</strong>. IMF Fiscal AffairsDepartment.World Bank. 2009a. <strong>Mozambique</strong> Investment Climate Assessment (Draft Report).World Bank and International F<strong>in</strong>ance Corporation. 2008. Do<strong>in</strong>g Bus<strong>in</strong>ess, 2009.World Bank, International F<strong>in</strong>ance Corporation, and PriceWaterhouseCoopers. 2008. Pay<strong>in</strong>g<strong>Tax</strong>es 2009: The Global Picture.Zee, H., J. Stotsky and E. Ley. 2002. <strong>Tax</strong> Incentives for Bus<strong>in</strong>ess Investment: A Primer for PolicyMakers <strong>in</strong> Develop<strong>in</strong>g Countries. World Development 30:9. 1497-1516.


<strong>PARPA</strong> <strong>II</strong> Review—The <strong>Tax</strong><strong>System</strong> <strong>in</strong> <strong>Mozambique</strong>Volume <strong>II</strong>: AppendicesSeptember 2009This publication was produced by Nathan Associates Inc. for review by the United StatesAgency for International Development.


<strong>PARPA</strong> <strong>II</strong> Review—<strong>Tax</strong>PolicyVolume <strong>II</strong>DISCLAIMERThis document is made possible by the support of the American people through the United States Agency forInternational Development (USAID). Its contents are the sole responsibility of the author or authors and do notnecessarily reflect the views of USAID or the United States government.


ContentsAppendix A. Basic Data 1Appendix B. Fiscal Benefits <strong>in</strong> <strong>Mozambique</strong> 9Appendix C. <strong>Tax</strong>es and Fiscal Benefits: International Comparison 19Appendix D. <strong>Tax</strong>ation and Bus<strong>in</strong>ess Environment Rat<strong>in</strong>gs 27Appendix E. Operational Efficiency Indicators 33Appendix F. Revenue Effort 39Appendix G. <strong>Tax</strong> Incentives for Investment—Pros and Cons 45Appendix H. Interview Results—Ma<strong>in</strong> Concerns and Issues 49Appendix I. Persons Interviewed 53


Appendix A. Basic Data


Table A-1Revenues by Type of <strong>Tax</strong>, % GDP1999 2000 2001 2002 2003 2004 2005 2006 2007 2008Receita Total 10.8 11.5 11.2 11.4 13.1 12.4 13.8 15.2 16.2 16.0Receitas fiscais 10.0 10.5 10.0 10.5 12.3 10.8 11.1 12.3 13.4 13.5Impostos sobre Rendimentos 1.5 1.6 1.8 2.1 2.9 2.7 2.9 3.5 4.5 4.9Imposto s/ rendimentos de pessoal 0.8 1.0 1.1 1.4 2.0 1.9 1.9 2.1 2.3 2.5Imposto s/ rendimentos de empresas 0.7 0.6 0.6 0.7 0.8 0.8 1.0 1.4 2.1 2.4Imposto Especial s/ o Jogo 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0Impostos sobre Bens e Serviços 6.6 7.3 6.9 7.2 7.9 7.2 7.4 8.0 8.3 8.0Imposto s/ Valor Acrescentado (IVA) 3.9 4.4 4.2 4.4 4.9 4.5 4.5 5.2 5.4 5.4IVA Operações <strong>in</strong>ternas 1.0 1.8 1.9 2.0 2.1 1.9 1.8 2.1 3.2 2.4IVA Importação 1.4 2.6 2.4 2.6 2.7 2.6 2.7 3.2 2.2 3.0Imposto s/ Consumo Específico - Prod.s Nacionais 0.6 0.6 0.6 0.6 0.7 0.6 0.6 0.6 0.6 0.7ICE - Cerveja e Refrigerantes 0.0 0.5 0.4 0.5 0.5 0.4 0.5 0.5 0.5 0.5ICE - Tabaco 0.0 0.1 0.2 0.2 0.2 0.2 0.2 0.2 0.1 0.1ICE - Outros Produtos 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1Imposto s/ Consumo Específico - Prod.s Importados 0.4 0.3 0.3 0.3 0.4 0.4 0.4 0.4 0.4 0.4Direitos Aduaneiros + sobretaxa de acucar 1.7 1.9 1.8 1.9 2.0 1.7 1.9 1.8 1.8 1.5Outros Impostos 1.9 1.6 1.3 1.2 2.2 0.9 0.8 0.7 0.7 0.6Imposto do Selo 0.2 0.2 0.1 0.1 0.2 0.2 0.2 0.1 0.2 0.1Imposto sobre Veículos 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0Imposto Reconstrução Nacional 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0Licenças de Pesca 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0Imposto s/ Combustíveis 1.3 1.2 1.0 0.9 1.2 0.5 0.4 0.3 0.3 0.3Royalties e Impostos de Superfície 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0Juros de Mora e <strong>Tax</strong>a de 3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0SISA 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0Diversos Outros Impostos 0.3 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.1 0.1Receitas não fiscais 0.3 0.6 0.5 0.7 0.7 0.7 0.7 0.5 0.6 0.5


1999 2000 2001 2002 2003 2004 2005 2006 2007 2008<strong>Tax</strong>as Diversas de Serviços 0.1 0.1 0.1 0.1 0.2 0.1 0.1 0.1 0.2 0.1Compensação de Apos. e Pensão de Sobrevivência 0.2 0.3 0.2 0.3 0.3 0.3 0.3 0.3 0.3 0.3Recuperação Crédito B. Austral 0.0 0.0 0.0 0.1 0.1 0.0 0.0 0.0 0.0 0.0Rendas de Casa 0.0 0.1 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.0Outras Receitas Não Fiscais 0.1 0.1 0.0 0.0 0.1 0.1 0.2 0.1 0.1 0.1Receitas Consignadas 0.4 0.2 0.3 0.2 0.2 1.0 1.0 1.0 1.1 1.1Imposto s/ Combustíveis (consig.) 0.0 0.0 0.0 0.0 0.0 0.8 0.8 0.7 0.7 0.7<strong>Tax</strong>a de Sobrevalor. da Castanha de Caju 0.0 0.0 0.0 0.1 0.0 0.1 0.0 0.0 0.0 0.0Assist. Médica e Medicamentosa 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0<strong>Tax</strong>as de Serviços Aduaneiros 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0Outras Receitas Consignadas 0.0 0.2 0.3 0.2 0.1 0.2 0.2 0.3 0.2 0.3Receitas Próprias 0.0 0.0 0.0 0.0 0.0 0.0 0.5 0.9 0.5 0.5Receitas de Capital 0.0 0.2 0.4 0.0 0.0 0.0 0.7 0.6 0.5 0.4Dividendos 0.0 0.0 0.0 0.0 0.0 0.0 0.5 0.4 0.0 0.0Privatizações líquidas 0.0 0.2 0.4 0.0 0.0 0.0 0.1 0.1 0.0 0.0Mem: Imposto s/ Combustiveis 1.3 1.2 1.0 0.9 1.2 1.3 1.2 1.0 1.1 1.0Note: Years 1999 to 2007 from Conta Geral do Estado, 2008 from Relatorio de Execucao. All from DNEAP's "Quadromacro Revisto CFMP Proposta,"received June 2009. Note that here Imposto de Combustivel is presented separately as but is otherwise <strong>in</strong>cluded partly <strong>in</strong> "Non-tax revenues" and partly <strong>in</strong>"Consigned revenues."


Table A-2GDP Sectoral Growth RatesAverage1999 2000 2001 2002 2003 2004 2005 2006 2007 20081999-08 2006-08PIB Real 8.4 1.5 12.3 9.2 6.5 7.9 8.4 8.7 7.4 6.8 7.7 7.6Agricultura, produção animal, caça e silvicultura 6.5 -13.1 10.6 12.1 5.2 5.1 6.9 10.4 9.8 9.4 6.3 9.9Pesca, aquacultura, e actividades relacionadas -2.1 4.8 0.6 1.1 8.6 0.2 0.9 7.7 6.1 4.5 3.2 6.1Industria M<strong>in</strong>eira -6.5 59.6 10.8 28.7 16.1 71.6 0.7 27.8 34.6 13.1 25.7 25.2Industria Manufactureira 14.7 15.1 34.7 8.7 17.0 13.2 2.1 3.0 1.3 2.9 11.3 2.4Electricidade e Agua 78.3 -8.3 9.9 10.4 10.1 16.3 17.2 13.1 8.7 -2.6 15.3 6.4Construção 3.4 13.0 6.7 10.8 9.7 -7.0 13.2 10.4 7.3 13.0 8.0 10.2Comercio e Serviços de Reparação 2.5 3.2 17.4 4.6 6.6 7.1 12.1 21.3 7.2 6.9 8.9 11.8Alojamento, restaurantes e similares 5.4 6.8 4.0 5.1 6.0 3.0 13.6 10.1 15.2 6.1 7.5 10.4Transportes, armazenagem e comunicações 9.0 2.6 6.9 8.4 2.9 9.6 7.9 10.4 10.6 18.3 8.7 13.1Serviços F<strong>in</strong>anceiros -26.9 80.8 21.3 15.8 10.5 25.2 49.3 3.7 10.9 12.9 20.3 9.1Activ. imobiliarias, alugueres e serviços às empresas 3.0 1.3 5.0 0.7 1.1 5.9 1.3 0.8 0.6 0.1 2.0 0.5Adm<strong>in</strong>istração pública, defesa e segurança social 18.1 6.1 22.3 7.4 4.8 4.6 6.9 11.2 5.1 7.4 9.4 7.9Educação 9.5 9.7 19.0 4.7 8.3 11.7 11.9 8.3 12.8 10.6 10.7 10.6Saúde e acção social 17.1 11.7 9.0 5.7 5.8 7.5 7.1 14.5 16.3 8.2 10.3 13.0Outras activ. de serviços colectivos, sociais e pessoais 10.0 18.3 6.8 3.3 2.4 2.4 2.4 2.4 2.4 0.5 5.1 1.8Dir.s de Importação - SIFIM (residual) -35.2 2.6 -3.9 26.9 -1.8 0.2 3.8 0.3 4.3 -8.2 -1.1 -1.2SOURCE: DNEAP Quadromacro, from June 2009.


Table A-3GDP Sectoral Shares(%)1999 2000 2001 2002 2003 2004 2005 2006 2007 2008Agricultura, produção animal, caça e silvicultura 27.4 23.5 23.1 23.7 23.4 22.8 22.5 22.9 23.4 24.0Pesca, aquacultura, e actividades relacionadas 2.2 2.3 2.0 1.9 1.9 1.8 1.7 1.6 1.6 1.6Industria M<strong>in</strong>eira 0.3 0.5 0.5 0.5 0.6 0.9 0.9 1.0 1.3 1.4Industria Manufactureira 10.1 11.4 13.7 13.6 15.0 15.7 14.8 14.0 13.2 12.8Electricidade e Agua 5.0 4.5 4.4 4.4 4.6 4.9 5.3 5.5 5.6 5.1Construção 3.1 3.5 3.3 3.3 3.4 3.0 3.1 3.2 3.2 3.3Comercio e Serviços de Reparação 9.7 9.9 10.4 9.9 9.9 9.9 10.2 11.4 11.4 11.4Alojamento, restaurantes e similares 1.6 1.6 1.5 1.5 1.5 1.4 1.5 1.5 1.6 1.6Transportes, armazenagem e comunicações 10.2 10.3 9.8 9.7 9.4 9.5 9.5 9.7 9.9 11.0Serviços F<strong>in</strong>anceiros 1.6 2.8 3.0 3.2 3.3 3.9 5.3 5.1 5.3 5.6Activ. imobiliarias, alugueres e serviços às empresas 11.7 11.7 10.9 10.1 9.6 9.4 8.8 8.1 7.6 7.1Adm<strong>in</strong>istração pública, defesa e segurança social 3.4 3.5 3.9 3.8 3.7 3.6 3.6 3.6 3.6 3.6Educação 3.0 3.2 3.4 3.2 3.3 3.4 3.5 3.5 3.7 3.8Saúde e acção social 1.2 1.3 1.2 1.2 1.2 1.2 1.2 1.2 1.3 1.4Outras activ. de serviços colectivos, sociais e pessoais 2.1 2.5 2.4 2.2 2.1 2.0 1.9 1.8 1.7 1.6Dir.s de Importação - SIFIM (residual) 7.6 7.6 6.5 7.6 7.0 6.5 6.2 5.7 5.6 4.8SOURCE: DNEAP Quadromacro from June 2009.


Figure A-1Sectoral GDP Cumulative Growth


Appendix B. Fiscal Benefits <strong>in</strong><strong>Mozambique</strong>This appendix provides a short history of fiscal benefits <strong>in</strong> <strong>Mozambique</strong> s<strong>in</strong>ce 1993, and adetailed comparison of the 2009 Code of Fiscal Benefits with the 2002 Code. (See Tables B-1 forthe comparison of specific benefits, and Table B-2 for general benefits.)FISCAL BENEFITS IN THE 1990SThe peace accord <strong>in</strong> 1992 signaled the beg<strong>in</strong>n<strong>in</strong>g of an era of stability and development <strong>in</strong><strong>Mozambique</strong>. One hallmark of this transformation was the adoption of a new Investment Law(Law No. 3/93 of 24 June) and a Code of Fiscal Benefits (Decree 12/93 of 21 July). The regime<strong>in</strong>cluded a guarantee of property rights, access to foreign exchange for remittance of capital andprofits, and generous tax breaks for a wide range of economic activities. Later <strong>in</strong> the 1990s thegovernment added special <strong>in</strong>centives for Special Economic Zones (SEZs) <strong>in</strong> 1998 and IndustrialFree Zones (IFZs)—another name for export process<strong>in</strong>g zones—<strong>in</strong> 1999.Dur<strong>in</strong>g most of that period the standard tax on corporate profit (Contribuição Industrial) was 35percent for agriculture, 40 percent for <strong>in</strong>dustry (<strong>in</strong>clud<strong>in</strong>g construction and m<strong>in</strong><strong>in</strong>g) and 45percent for the trade and service sectors. Customs duties ranged up to 35 percent, <strong>in</strong>clud<strong>in</strong>g a 10percent duty on capital goods. After 1998, the company tax rate was reduced to 35 percent for allsectors but agriculture and fisheries, which enjoyed a special 10 percent tax rate. Other specialbenefits applied to m<strong>in</strong><strong>in</strong>g, oil and gas, hotels and tourism, and sugar. 1The ma<strong>in</strong> fiscal <strong>in</strong>centives were reductions <strong>in</strong> the tax rate for def<strong>in</strong>ed periods of time (partial taxholidays) together with exemptions on customs duties for certa<strong>in</strong> capital goods. The “general”<strong>in</strong>centives <strong>in</strong>cluded a 50 percent reduction <strong>in</strong> the company tax and complementary tax for up to10 years. In three northern prov<strong>in</strong>ces, the reduction was 80 percent for 10 years, and 50 percentfor another 6 years. In four other prov<strong>in</strong>ces, <strong>in</strong>vestments outside the capital obta<strong>in</strong>ed a 65 percenttax reduction for 10 years followed by a 40 percent tax break for 3 more years.The IFZ package <strong>in</strong>cluded exemption from customs duty and <strong>in</strong>direct tax on <strong>in</strong>puts used forexport production, and an unusual “royalty fee” <strong>in</strong> lieu of <strong>in</strong>come tax, set at 1 percent of gross1 Adrien Goorman, Randa Sab and Paulo Ramos, Moçambique: Raçionalização dos Incentivos Fiscais,IMF, August 2000, p. 13. This study was not released to the public.


10 <strong>PARPA</strong> <strong>II</strong> R EVIEW—THE T AX S YSTEM IN M OZAMBIQUErevenue or a flat rate per square meter of occupied area. For <strong>in</strong>vestments <strong>in</strong> the Zambezi RiverValley SEZ (from 1998), the <strong>in</strong>centives <strong>in</strong>cluded an exemption from customs duty and <strong>in</strong>directtax for both equipment and <strong>in</strong>termediate goods and a full tax holiday for 5 years followed by an80 percent reduction thereafter; for undertak<strong>in</strong>gs <strong>in</strong> agriculture and fisheries the full tax holidayapplies until 2025.Under this regime private <strong>in</strong>vestment <strong>in</strong>creased substantially <strong>in</strong> the late 1990s (the trends areexam<strong>in</strong>ed more fully <strong>in</strong> the next section). There is no doubt that fiscal <strong>in</strong>centives were essential tothe negotiations on two early showcase projects of enormous strategic value to the country—theMozal alum<strong>in</strong>um smelter and the Maputo-Witbank toll road project. More generally, though, onecannot easily disentangle the impact of <strong>in</strong>centives from the effect of political stability,macroeconomic stabilization after 1996, and the privatization of hundreds of formerlynationalized enterprises.One piece of evidence came from a study conducted by Jose Macamo for the M<strong>in</strong>istry ofPlann<strong>in</strong>g and F<strong>in</strong>ance <strong>in</strong> 2000, to determ<strong>in</strong>e the importance of adm<strong>in</strong>istrative barriers to<strong>in</strong>vestment as viewed through the eyes of a randomized sample of 30 recent <strong>in</strong>vestors. The studyfound that 76 percent of the respondents would have undertaken the same <strong>in</strong>vestment without taxand customs <strong>in</strong>centives. 2 Also <strong>in</strong> 2000 the IMF conducted a detailed study for the government onthe fiscal <strong>in</strong>centives regime. Based on calculations from a sample of tax files, the study concludedthat the exist<strong>in</strong>g benefits were “not cost effective.” The IMF team was especially critical of costlytax holidays, recommend<strong>in</strong>g <strong>in</strong>stead <strong>in</strong>vestment tax credits and accelerated depreciation. 3THE 2002 REFORMIn the wake of these f<strong>in</strong>d<strong>in</strong>gs, the government <strong>in</strong> 2002 adopted a comprehensive reform of the<strong>in</strong>come tax, <strong>in</strong>clud<strong>in</strong>g a new Code of Fiscal Benefits. Major provisions of the tax reform program<strong>in</strong>cluded a reduction <strong>in</strong> the standard corporate tax rate to 32 percent (with agriculture still at 10percent) and elim<strong>in</strong>ation of the complementary surtax. At the same time, the new Code of FiscalBenefits was designed “to rationalize the concession of fiscal <strong>in</strong>centives so that this regime can bemore efficient and efficacious as an <strong>in</strong>strument of economic policy” and to consolidate <strong>in</strong> onelegal <strong>in</strong>strument what had been “a very scattered system of fiscal benefits.” 4A central feature of the 2002 Code was the elim<strong>in</strong>ation of most tax holidays <strong>in</strong> favor of<strong>in</strong>vestment tax credits and accelerated depreciation, along the l<strong>in</strong>es recommended by the IMF.These general <strong>in</strong>centives applied to a wide range of activities, exclud<strong>in</strong>g most wholesale andretail undertak<strong>in</strong>gs other than rural commerce and <strong>in</strong>vestments <strong>in</strong>volv<strong>in</strong>g new commercial<strong>in</strong>frastructure. In addition, partial holidays rema<strong>in</strong>ed <strong>in</strong> place for agriculture (80 percent reduction2 Jose Macamo, Adm<strong>in</strong>istrative barriers to Investment <strong>in</strong> Moçambique: Lessons learned from theExperience of Recent Investors, M<strong>in</strong>istry of Plann<strong>in</strong>g and F<strong>in</strong>ance, Gab<strong>in</strong>ete de Estudos Discussion Paper#17, December 2000.3 Goorman et al, (2000), op. cit., pp. 5 and 13. This recommendation echoed the predom<strong>in</strong>ant IMF viewof tax <strong>in</strong>centives, as subsequently expressed <strong>in</strong> Zee, Stotsky and Ley (2002).4 Preamble to the Code of Fiscal Benefits, Decree n o 16/2002 of 27 th June.


A PPENDIX B 11<strong>in</strong> the company tax rate until 2012), m<strong>in</strong><strong>in</strong>g (25 percent reduction for 8 years), and IFZ entities(60 percent reduction for 10 years). The thresholds to qualify for CPI authorization under theInvestment Act rema<strong>in</strong>ed unchanged at $5,000 for domestic entities and $50,000 for foreignentities.FDI <strong>in</strong>flows decl<strong>in</strong>ed when the 2002 Code took effect, but this was largely due to the tim<strong>in</strong>g ofmega-projects and uncerta<strong>in</strong>ties created by <strong>in</strong>flation, budget deficits, and exchange rate volatility.Still, average FDI <strong>in</strong> the four years after 2002 matched the average for the correspond<strong>in</strong>g periodbefore 2002, at US$259 million and US$253 million, respectively. FDI then reached new heights<strong>in</strong> 2007 once macroeconomic conditions stabilized. Similar trends are seen <strong>in</strong> the data on CPIapprovals, which <strong>in</strong>dicate the level of <strong>in</strong>terest by <strong>in</strong>vestors plann<strong>in</strong>g to commit resources to<strong>Mozambique</strong>. Exclud<strong>in</strong>g large-scale projects (because they were covered by the special regimethat did not change <strong>in</strong> 2002), the average level of approvals was higher <strong>in</strong> the four years after2002 than <strong>in</strong> the prior four years—at US$552 million and US$ 508 million respectively.Build<strong>in</strong>g on Macamo (2000), Bolnick (2009) conducted a study of <strong>in</strong>vestment motives at the endof 2008, cover<strong>in</strong>g a stratified random sample of projects approved by CPI <strong>in</strong> 2005, 2006 and2007. The results were very similar. Of 60 companies covered <strong>in</strong> the survey, 83 percent <strong>in</strong>dicatedthat tax <strong>in</strong>centives were not a critical factor <strong>in</strong> their <strong>in</strong>vestment decision; the correspond<strong>in</strong>g figurefor import duty relief was 73 percent. While these f<strong>in</strong>d<strong>in</strong>gs confirm a very high redundancy ratefor the fiscal benefits, the study also found that the tax and duty benefits were critical factors forthe largest <strong>in</strong>vestments that were approved dur<strong>in</strong>g this time frame. However, the bus<strong>in</strong>ess plansfiled with CPI showed that <strong>in</strong>vestments that h<strong>in</strong>ged on the availability of <strong>in</strong>centives were verycapital <strong>in</strong>tensive and were designed to create far fewer jobs than the projects that would havebeen undertaken with or without the tax breaks. 5THE 2009 CODE OF FISCAL BENEFITSA new Code of Fiscal Benefits came <strong>in</strong>to force <strong>in</strong> January 2009 (under Law 4/2009 of 12 thJanuary). Major changes <strong>in</strong> the new Code are discussed <strong>in</strong> Chapter 4 of the text. Here we providea more complete description of the new benefits package <strong>in</strong> the form of a tabular comparison ofthe 2009 and 2002 Codes. Table B-1 shows the ma<strong>in</strong> features of the “specific benefits” under thetwo codes, while Table B-2 provides a direct comparison of the “general benefits.” Note thatspecific benefits apply to designated sectors or activities, whereas general benefits apply toqualify<strong>in</strong>g <strong>in</strong>vestments not covered by designated specific benefits. Under Article 4 of the 2009Code, specific benefits may not be aggregated with other specific or general benefits, unlessotherwise <strong>in</strong>dicated <strong>in</strong> the law.5 Bruce Bolnick, Invest<strong>in</strong>g <strong>in</strong> <strong>Mozambique</strong>: The Role of Fiscal Incentives, Nathan Associates, February2009.


Table B-1Specific Fiscal Benefits 2002 and 2009, Ma<strong>in</strong> FeaturesType of Investment Specific Benefits 2002 2009Public <strong>in</strong>frastructure <strong>in</strong>vestmentby private sector or by publicprivatepartnershipsCustoms duty and VAT exemption 2002 General Benefits only Exemption from payment of import duties and VAT on class “K”imports <strong>in</strong>clud<strong>in</strong>g spare parts and accessories.Income tax reductionSee Table of General Benefits, Item 6 Basicfor 200280% reduction <strong>in</strong> the IRPC tax rate for the first five (5) tax years;60% reduction <strong>in</strong> the IRPC tax rate for tax years 6 to 10;25% reduction <strong>in</strong> the IRPC tax rate for tax years 11 – 15.Rural commerce and <strong>in</strong>dustry Customs duty and VAT exemption 2002 General Benefits only For rural commerce, exemption from payment of import dutiesand VAT on class “K” imports as well as other essential goods asenumerated, such as freezers and scales.For rural <strong>in</strong>dustry, exemption from payment of import duties andVAT on class “K” imports <strong>in</strong>clud<strong>in</strong>g spare parts and accessories.Manufactur<strong>in</strong>g and assembly<strong>in</strong>dustriesCustoms duty exemption 2002 General Benefits only Exemption from payment of import duties on imported rawmaterials for the production process.For assembly of motor vehicles, electronic equipment, computerand communications technology, exemption from payment ofimport duties on materials.Above exemptions require annual <strong>in</strong>voic<strong>in</strong>g above 3 million MTand value added of at least 20%.Agriculture and fisheries(Agriculture only <strong>in</strong> 2002)Customs duty and VAT exemptionExemption from payment of customs dutieson class “K” equipment, for required goodsnot produced <strong>in</strong> <strong>Mozambique</strong>.Exemption from payment of customs duties and VAT on theimport of class “K” equipment and accompany<strong>in</strong>g spare parts andaccessories.Income tax reduction80% reduction of the <strong>in</strong>come tax rate until2012, on profits from agricultural ventures80% reduction of the <strong>in</strong>come tax rate until 31 December 201550% reduction <strong>in</strong> the <strong>in</strong>come tax rate between 2016 and 2025.Additional benefitsGeneral benefits for professional tra<strong>in</strong><strong>in</strong>g,public <strong>in</strong>frastructure expenditures, stamptax and property transfer taxGeneral benefits for professional tra<strong>in</strong><strong>in</strong>g and public <strong>in</strong>frastructureexpendituresHotel and tourism Customs duty and VAT exemption Exemption from payment of import dutieson class “K” equipment, for requiredgoods not produced <strong>in</strong> <strong>Mozambique</strong>.Exemption from payment of import duties and VAT on class “K”equipment and other <strong>in</strong>dispensable goods for the construction andoutfitt<strong>in</strong>g of tourism and hotel activities (see list).


Type of Investment Specific Benefits 2002 2009Investment credit and accelerateddepreciationInvestment tax credit as per GeneralBenefits plus 3 percentage po<strong>in</strong>ts.Accelerated depreciation up to 3 times thenormal rate on new immovable assets,automotive vehicles and other fixed assets.These benefits apply only until 31December 2007.Accelerated depreciation <strong>in</strong>creased by 50% on new immovableassets, vehicles and other fixed assets.Additional benefits All general benefits. General benefits for the <strong>in</strong>vestment tax credit and accelerateddepreciation.Science and technology parks Customs duty and VAT exemption N/A Exemption from payment of import duties and VAT on scientific,teach<strong>in</strong>g and laboratory equipment, <strong>in</strong>clud<strong>in</strong>g software and itssupport materials, <strong>in</strong>clud<strong>in</strong>g accessories and spare parts.Income tax reduction N/A Exemption from the <strong>in</strong>come tax <strong>in</strong> the first 5 tax years.50% reduction <strong>in</strong> the <strong>in</strong>come tax <strong>in</strong> years 6 to 10.25% reduction <strong>in</strong> the <strong>in</strong>come tax <strong>in</strong> years 11 to 15.Large scale projects2002 Code: Investmentsexceed<strong>in</strong>g US$500 million2009 Code: Investmentsexceed<strong>in</strong>g 12.5 billion MTExemption of import duties andVATExceptional fiscal <strong>in</strong>centivesExemption from payment of import dutieson class “K” equipment, for requiredgoods not produced <strong>in</strong> <strong>Mozambique</strong>.To be granted by the M<strong>in</strong>istry of Plann<strong>in</strong>gand F<strong>in</strong>ance under a contractual regimeapproved by the Council of M<strong>in</strong>isterscover<strong>in</strong>g import duties, <strong>in</strong>come tax,property transfer tax and stamp taxExemption from payment of import duties and VAT on importedconstruction materials, mach<strong>in</strong>ery, equipment, and accompany<strong>in</strong>gspare parts.Exceptional fiscal <strong>in</strong>centives elim<strong>in</strong>atedInvestment tax creditInvestment tax credit rang<strong>in</strong>g from 5% to10%, subject to 5-year carry forward.For projects <strong>in</strong> Gaza, Sofala, Manica, Tete,Zambezia and Nampula prov<strong>in</strong>ces, ITC of10% to 20%.For projects <strong>in</strong> Cabo Delgado, Inhambaneand Niassa, ITC of 15% to 30%.General benefits for professional tra<strong>in</strong><strong>in</strong>g,public <strong>in</strong>frastructure expenditures, stamptax and property transfer tax.General benefits for <strong>in</strong>vestment tax credit, accelerateddepreciation, deductions for modern technology, professionaltra<strong>in</strong><strong>in</strong>g, and public <strong>in</strong>frastructure <strong>in</strong>vestments.Rapid Development Zones –designated eligible activities <strong>in</strong>Zambezi Valley, Niassa Prov<strong>in</strong>ce,Nacala District, MocambiqueInsland and Ibo IslandExemption of import duties andVATExemption from payment of import dutieson class “K” and “I” imports, dur<strong>in</strong>g first 3years of implementation, for goods notproduced <strong>in</strong> <strong>Mozambique</strong>Until 31 December 2015.Exemption from payment of import duties and VAT on the importof class “K” equipment, <strong>in</strong>clud<strong>in</strong>g accessories and spare parts.


Type of Investment Specific Benefits 2002 2009Income tax benefitsIncome tax credit equal to 20% of the totalrealized <strong>in</strong>vestment.Until 31 December 2015.Same, but without the “sunset” date.Additional benefitsExemption from property transfer tax andgeneral benefits for professional tra<strong>in</strong><strong>in</strong>g,public <strong>in</strong>frastructure expenditures, stamptax.General benefits for professional tra<strong>in</strong><strong>in</strong>g and public <strong>in</strong>frastructureexpenditures.Industrial Free Zones (ZFIs)Exemption of import duties andVATExemption from customs duty and VAT onthe importation of construction materials,mach<strong>in</strong>ery, accessories, spare parts andgoods and merchandise to be used <strong>in</strong> theimplementation of projects and operation ofapproved activities.VAT exemption <strong>in</strong>cludes <strong>in</strong>ternalacquisitions as well as importsSameIncome tax reduction60% reduction <strong>in</strong> the corporate <strong>in</strong>come taxrate for 10 years.For operators and enterprises <strong>in</strong> ZFIsExemption from <strong>in</strong>come tax for first 10 years.50% reduction for years 11 to 15.25% reduction for the rema<strong>in</strong><strong>in</strong>g life of the project.For enterprises <strong>in</strong> isolated free zones:Exemption from <strong>in</strong>come tax for first 5 years.50% reduction for years 6 to 10.25% reduction for the rema<strong>in</strong><strong>in</strong>g life of the projectAdditional benefits Exemption from property transfer tax. N/ASpecial Economic Zones (ZEEs)Exemption of import duties andVATN/AExemption from customs duty on the importation of constructionmaterials, mach<strong>in</strong>ery, accessories, spare parts and other goodsused <strong>in</strong> carry<strong>in</strong>g out licensed ZEE activity.VAT exemption <strong>in</strong>cludes <strong>in</strong>ternal acquisitions as well as imports.


Type of Investment Specific Benefits 2002 2009Income tax reduction N/A For ZEE operators:Exemption from company <strong>in</strong>come tax for 5 years;50% reduction for years 6 to 10;25% reduction for the rema<strong>in</strong><strong>in</strong>g life of the project;For ZEE enterprises:Exemption of company <strong>in</strong>come tax for 3 years;50% reduction for years 4 to-10;25% reduction for years 11 to 15.For ZEE service enterprises:50% reduction <strong>in</strong> company <strong>in</strong>come tax for 5 years.Investments under the M<strong>in</strong>es ActExemption of import duties andVATExemption from customs duty, VAT andexcise duty on the importation of all articlesrelat<strong>in</strong>g to prospect<strong>in</strong>g, exploration andexploitation of m<strong>in</strong>eral resources.Exemption from customs duty, VAT and excise duty on theimportation of class “K” equipment for prospect<strong>in</strong>g, explorationand exploitation of m<strong>in</strong>eral resources, for a period of 5 years fromthe date of commencement.Income tax reductionUntil 2010, a 25% reduction <strong>in</strong> thecorporate <strong>in</strong>come tax rate for the first 5years of production, on <strong>in</strong>vestments above$500,000.N/AInvestments under the PetroleumActExemption of import duties andVATExemption from customs duty, VAT, andexcise duty on the importation of goods<strong>in</strong>tended for use <strong>in</strong> petroleum operations.Exemption from customs duty, VAT and excise duty on theimportation of class “K” equipment and other designated goodsfor use <strong>in</strong> oil operations, for a period of 5 years from the date ofapproval.Income tax reductionUntil 2010, a 25% reduction <strong>in</strong> thecorporate <strong>in</strong>come tax rate dur<strong>in</strong>g the first 8years follow<strong>in</strong>g the start of production.N/A


Table B-2General Fiscal Benefits 2002 and 2009, Ma<strong>in</strong> FeaturesBenefit 2002 2009Benefits on the import of goodsFiscal benefits <strong>in</strong> respect of <strong>in</strong>comeAccelerated depreciationModernization and <strong>in</strong>troduction ofnew technologyExemption from payment of import dutieson equipment <strong>in</strong>cluded <strong>in</strong> class “K” of thecustoms Tariff ScheduleInvestment carried out under the<strong>in</strong>vestment law shall benefit for theperiod of five (5) years from an<strong>in</strong>vestment tax credit equal to 5% of thetotal <strong>in</strong>vestment realised.In Maputo prov<strong>in</strong>ce the percentage of theITC shall be of 5%; <strong>in</strong> other prov<strong>in</strong>ces theITC shall range from 10% to 15%.Permitted for new immovable assets.Accelerated depreciation at twice thenormal rate set by law for the purpose ofdeterm<strong>in</strong>ation of taxable <strong>in</strong>come subjectto Corporate Income <strong>Tax</strong> ( IRPC) andPersonal Income <strong>Tax</strong>The amount <strong>in</strong>vested <strong>in</strong> specialisedequipment shall dur<strong>in</strong>g the first five yearsfrom the date of commencement ofactivity, benefit from a deduction fromtaxable <strong>in</strong>come for the purpose of theIRPC up to a maximum amount of 15%of taxable <strong>in</strong>come.Exemption from payment of import dutieson equipment and accessories <strong>in</strong>cluded <strong>in</strong>class “K” of the customs Tariff Schedule.SameIn Maputo prov<strong>in</strong>ce the percentage of theITC shall be of 5%; <strong>in</strong> other prov<strong>in</strong>ces theITC is 10%.Same.Accelerated depreciation at 50% above thenormal rate set by law for the purpose ofdeterm<strong>in</strong>ation of taxable <strong>in</strong>come subject toCorporate Income <strong>Tax</strong> ( IRPC) andPersonal Income <strong>Tax</strong>The same conditions also apply torehabilitated immovable assets andequipment and mach<strong>in</strong>ery for <strong>in</strong>dustrialand/or agro <strong>in</strong>dustrial activities.The amount <strong>in</strong>vested <strong>in</strong> specialisedequipment shall dur<strong>in</strong>g the first five yearscount<strong>in</strong>g from the date of commencementof activity, benefit from a deduction fromtaxable <strong>in</strong>come for the purpose of theIRPC up to a maximum amount of 10% oftaxable <strong>in</strong>comeProfessional tra<strong>in</strong><strong>in</strong>gInvestment expenditure for professionaltra<strong>in</strong><strong>in</strong>g of Mozambican locals shall up toa maximum amount of 5% of taxable<strong>in</strong>come be deductible from taxable<strong>in</strong>come for the purpose of calculat<strong>in</strong>gcorporate <strong>in</strong>come tax.When the professional tra<strong>in</strong><strong>in</strong>g is for theuse of technologically advancedequipment, the allowable <strong>in</strong>come taxdeduction for the purpose of thecalculation of the Corporate Income <strong>Tax</strong>shall be a maximum amount equal to 10%of taxable <strong>in</strong>come.Same.Same.<strong>Tax</strong> Deductible Expenditure – Dur<strong>in</strong>ga period of 10 years count<strong>in</strong>g from thedate of production, enterprises certa<strong>in</strong>expenditures may be treated asdeductible expenditure for the purposeof calculation of Corporate Income<strong>Tax</strong> (IRPC)In the case of undertak<strong>in</strong>gs carried out <strong>in</strong>the city of Maputo, 120% of the value ofexpenditure <strong>in</strong> the construction andrehabilitation of roads, railways, airports,mail delivery, telecommunications, watersupply, electric energy, schools, hospitalsand other works that are considered to beof public utility by the competentauthority and documented by the <strong>Tax</strong>Adm<strong>in</strong>istration.In the case of other prov<strong>in</strong>ces, an amountequal to 150% of the expenditureIn the case of undertak<strong>in</strong>gs carried out <strong>in</strong>the city of Maputo, 110% of the value ofexpenditure <strong>in</strong> the construction andrehabilitation of roads, railways, airports,mail delivery, telecommunications, watersupply, electric energy, schools, hospitalsand other works that are considered to beof public utility by the competent authorityand documented by the <strong>Tax</strong>Adm<strong>in</strong>istration.In the case of other prov<strong>in</strong>ces, an amountequal to 120% percent of the expenditure


Benefit 2002 2009Exemption from Stamp <strong>Tax</strong>Reduction <strong>in</strong> the rate of the realproperty transfer taxThe acts for the <strong>in</strong>corporation ofcompanies <strong>in</strong>clud<strong>in</strong>g the alteration of theshare capital and article of association areexempt from stamp duty dur<strong>in</strong>g the firstfive (5) years.Undertak<strong>in</strong>gs shall benefit from a 50%percent reduction <strong>in</strong> the rate of the realproperty transfer tax (SISA) with regardto the acquisition of immovable propertyused <strong>in</strong> <strong>in</strong>dustry, agro <strong>in</strong>dustry, and hotelas long as the property is acquired with<strong>in</strong>the first three (3) years.N/AN/ANote: Under Article 13 of the 2009 Code, “general” fiscal benefits apply to <strong>in</strong>vestments that are not covered by any of the specificbenefits provided <strong>in</strong> the Code, and may not be cumulated with specific fiscal benefits unless otherwise specified <strong>in</strong> the latterprovision.


Appendix C. <strong>Tax</strong>es and FiscalBenefits: InternationalComparisonTable C-1 (<strong>in</strong> two parts) provides selected details about the tax structures and tax <strong>in</strong>centiveregimes <strong>in</strong> the SADC member states, as well as four other comparator countries <strong>in</strong> Africa (Ghana,Kenya, Senegal, and Uganda). It difficult to extract clear judgments from the comparative taxtables on the extent to which <strong>Mozambique</strong>’s tax system is regionally competitive, because the<strong>in</strong>centive regimes conta<strong>in</strong> many technical details that defy simple comparison, and <strong>in</strong> any case theeffects depend on characteristics of particular <strong>in</strong>vestments. The most useful comparisons comefrom two studies that are discussed <strong>in</strong> Chapter 5 of the ma<strong>in</strong> text (see Investment) show<strong>in</strong>g thatthe 2002 Code of Fiscal Benefits <strong>in</strong> <strong>Mozambique</strong> was highly competitive <strong>in</strong> the region. Nocomparable studies have been conducted on the 2009 Code.


Table C-1Overview of Incentives (Part 1)CountryStatutoryCorporateIncome <strong>Tax</strong>Rate aSummary ofInvestment IncentivesTreatment ofDividends ofBus<strong>in</strong>ess AssetsPersonalIncome <strong>Tax</strong>Marg<strong>in</strong>alRatesFavorable <strong>Tax</strong> Rates or ExemptionsAngola 35 % Incentive code grants tax holidays forproject of national <strong>in</strong>terest and projectlocated <strong>in</strong> special development zones.Favorable tax rates and <strong>in</strong>centive aregranted <strong>in</strong> agriculture, forestry, m<strong>in</strong><strong>in</strong>g,and oil.10% withhold<strong>in</strong>g ondividends (residentsand non-residents)Progressive to15%20% for agriculture, forestry, 30% on rent for urbanproperty; 40% legislation govern<strong>in</strong>g m<strong>in</strong><strong>in</strong>g activities;50% for <strong>in</strong>come from oil is taxed; 65.75% for foreignproduction shar<strong>in</strong>g agreement partners and jo<strong>in</strong>t ventures.Exporters exempt from excise tax.Botswana 25 % b Botswana’s <strong>in</strong>vestment promotion ischaracterized by a low statutory corporate<strong>in</strong>come tax rate and a simple <strong>in</strong>centiveregime.15 % withhold<strong>in</strong>gwith set off aga<strong>in</strong>st 10% ACT liability.5-10-15-20-25 % Certa<strong>in</strong> manufactur<strong>in</strong>g and companies operat<strong>in</strong>g underthe jurisdiction of the International F<strong>in</strong>ancial ServicesCenter receive a lower 5% (+10%) tax rate.DR Congo 40 % Incentive code provide 3 to 5 year taxholiday for new companies and 60% ICAfor manufactur<strong>in</strong>g export<strong>in</strong>g greater 20%of output. Higher corporate and personal<strong>in</strong>come tax rates.20 % withhold<strong>in</strong>g(residents and nonresidents)3-5-10-15-20-25-30-35-40-45-50 %Approved companies exempt from export duties andtaxes.Lesotho 25% Lesotho’s current code does not grant taxholidays. 10% tax rate available formanufactur<strong>in</strong>g and farm<strong>in</strong>g and nowithhold<strong>in</strong>g on dividends distributed bymanufactur<strong>in</strong>g companies.15 % withhold<strong>in</strong>gresidents; 25 %withhold<strong>in</strong>g nonresidents;Nowithhold<strong>in</strong>g tax ondividends distributedby manufactur<strong>in</strong>gcompanies toshareholders25-35 % 10% for manufactur<strong>in</strong>g and farm<strong>in</strong>g0% on <strong>in</strong>come generated from manufactured goodsoutside SACUMadagascar 24 % Limited <strong>in</strong>formation collected.Madagascar offers multiple export<strong>in</strong>centives for EPA firms and exportcompanies.No withhold<strong>in</strong>g 24 % Multiple export <strong>in</strong>centives (eg. Refund of VAT forExport Process<strong>in</strong>g Zone firms and professional exportcompanies )Malawi 30 % Malawi’s <strong>in</strong>centive code provides priority<strong>in</strong>dustries with options of 5-10 year taxholidays (depend<strong>in</strong>g on size of <strong>in</strong>vestment)or fixed 15% tax rate. Favorable tax ratesavailable to EPZ firms, manufacturers andfarmers.10 % withhold<strong>in</strong>g(residents and nonresidents)10-20-30-40 % Companies <strong>in</strong> EPZs exempt; 21% <strong>in</strong>surance bus<strong>in</strong>esses;35% branch for foreign companies, 25 % forecclesiastical, charitable or educational <strong>in</strong>stitutions ortrustsMauritius 15 % Flat tax rate of 15%. <strong>Tax</strong> holidaysofficered to small enterprises whichregister for the first time.No withhold<strong>in</strong>g 15 % Double deduction of export market<strong>in</strong>g costs; tax credit15%-40% on export volume such that tax not less than15%


CountryStatutoryCorporateIncome <strong>Tax</strong>Rate aSummary ofInvestment IncentivesTreatment ofDividends ofBus<strong>in</strong>ess AssetsPersonalIncome <strong>Tax</strong>Marg<strong>in</strong>alRatesFavorable <strong>Tax</strong> Rates or Exemptions<strong>Mozambique</strong> 32 % <strong>Mozambique</strong>’s <strong>in</strong>centive code does notgrant ax holidays, but all sectors andregions are granted tax credits withpreferable rates for certa<strong>in</strong> sectors andregions of <strong>in</strong>terest. Favorable tax rates andspecial <strong>in</strong>centives are granted toagriculture, tourism, and other <strong>in</strong>vestmentsconsidered a high priority.20% withhold<strong>in</strong>g(residents and nonresidents)10% for shares listedon the Maputo stockexchangeProgressive to32 %10% rate on agricultural activities until 12/31/2010.; 15%for 5 yrs for <strong>in</strong>vest<strong>in</strong>g <strong>in</strong> specialized equipment.10 yrs: 120% deductible expenditure for public utilityprojects <strong>in</strong> Maputo, 150% for rest of country.80% on <strong>in</strong>come tax for agriculture.“Exceptional <strong>in</strong>centives” for projects over $500m or<strong>in</strong>frastructure or creat<strong>in</strong>g 500-1000 jobs <strong>in</strong> 3 years.M<strong>in</strong><strong>in</strong>g: 25% reduction <strong>in</strong> IRPC for 5 yrs. if $500,000+.Oil: 25% for 8 yrsIFZs 60%, exempt from real property transfer tax.Export <strong>in</strong>centive 60% tax reduction for 10 years.Namibia 35 % Malawi’s <strong>in</strong>centive code offers 50%abatement for 5 years for registeredmanufacturers, a 20% ICA on build<strong>in</strong>gsAnd a 3-year write-off for developmentexpenditure <strong>in</strong> m<strong>in</strong><strong>in</strong>g and petroleum.0% withhold<strong>in</strong>g forresidents; 10%withhold<strong>in</strong>g nonresidentsProgressive to35 %35% m<strong>in</strong><strong>in</strong>g companies; 55% diamond m<strong>in</strong><strong>in</strong>g; 35%petroleum m<strong>in</strong><strong>in</strong>g companies; 37.5 other m<strong>in</strong><strong>in</strong>gcompanies.EPZ companies exempt from all taxes and duties;additional deduction of 25% to 75% on costs for exportpromotion and market<strong>in</strong>g80% allowance on taxable <strong>in</strong>come from export ofmanufactured goods (excl. fish and meat)Seychelles 40 % Seychelles <strong>in</strong>centive code does not offertax holidays but provides 150%-200%special deductions for tra<strong>in</strong><strong>in</strong>g. Multipledeductions and allowance granted tomanufactur<strong>in</strong>g, tourism, agriculture,mar<strong>in</strong>e resources, professional services.0% for residents; 15%withhold<strong>in</strong>g for nonresidentsNone.15% rate for export companies under IPA and companies<strong>in</strong> special growth areas; 25% and 35% bracket for smallbus<strong>in</strong>esses; 15% fisheries, tourism and manufactur<strong>in</strong>gSouth Africa 28 % 6 year tax holiday for IDZs ended <strong>in</strong> 1999but 0 duty and VAT on <strong>in</strong>puts still <strong>in</strong>effect. Favorable rates granted to smallmanufacturers, and <strong>in</strong>centives offered forfarm development, m<strong>in</strong><strong>in</strong>g and qualify<strong>in</strong>gstrategic projects.0% withhold<strong>in</strong>g forresidents and nonresidents(expected tochange to 10% <strong>in</strong>2009); 12.5%secondary tax chargedon declared dividends;18-25-30-35-38-40 %15% rate for small manufacturers ( taxable<strong>in</strong>come


CountryStatutoryCorporateIncome <strong>Tax</strong>Rate aSummary ofInvestment IncentivesTreatment ofDividends ofBus<strong>in</strong>ess AssetsPersonalIncome <strong>Tax</strong>Marg<strong>in</strong>alRatesFavorable <strong>Tax</strong> Rates or ExemptionsSwaziland 30 % Extensive tax holiday of 5 years for newexport manufactur<strong>in</strong>g <strong>in</strong>dustry and 10 yearholiday at 10% tax rate + exemption fromdividend w/hold<strong>in</strong>g available underDevelopmental Approval Order.Tanzania 30 % In recent years the government hasattempted to simplify/harmonize its<strong>in</strong>centive system. It is noteworthy thatTanzania does not offer tax holidaysoutside of the EPZs.Zambia 35 % Zambia’s tax regime provides favorabletax rates and allowances for agriculture,manufactur<strong>in</strong>g, m<strong>in</strong><strong>in</strong>g, and tourism withspecial <strong>in</strong>centives granted through theZambian Development Act to <strong>in</strong>vestmentsthat are considered a high priority.10% withhold<strong>in</strong>gresidents; 15%withhold<strong>in</strong>g onpayments to nonresidents; 12.5% forSACU basedcompanies10% for residents andnon-residents; 5%withhold<strong>in</strong>g forcompanies on stockexchange; 10% foragriculture orcertificate of<strong>in</strong>vestment; 0% form<strong>in</strong>eral sector & EPZ15% withhold<strong>in</strong>g(residents and nonresidents;exceptionfor m<strong>in</strong><strong>in</strong>g sector,EPZ5 year exemption forfarmers20-25-30-33 % 10% Development Approval Order;15% tax rate for export companies under IPA; zero tax onoffshore companies operat<strong>in</strong>g <strong>in</strong> <strong>in</strong>ternational trade zonesFTZ; duty credit certificate scheme for textile andcloth<strong>in</strong>g exporters.17.5-20-25-30 % Same corporate tax rate across all sectors.25-30 % 15% farm<strong>in</strong>g, fertilizer companies, and non-traditionalexports; 30% for m<strong>in</strong><strong>in</strong>g companies; 40% bank amountsover ZMK 250 millionEPZ Act implemented 2003 provides for stand aloneEPZ sites; <strong>in</strong> addition to standard EPZ tax benefits, alsoprovides full exemption from corporate tax, withhold<strong>in</strong>gtax, capital ga<strong>in</strong>s and excise dutyZimbabwe 30 % Zimbabwe’s <strong>in</strong>centive code offers a 5-yearholiday for qualify<strong>in</strong>g <strong>in</strong>vestor, EPZcompanies. ICAs also available undercurrent tax code.20% withhold<strong>in</strong>g(resident and nonresident);15% forquoted companies20-25-30-35-40-45 %15% for licensed <strong>in</strong>vestor (after 5 yr holiday) or new<strong>in</strong>frastructure project <strong>in</strong> growth po<strong>in</strong>t area; 25% form<strong>in</strong><strong>in</strong>g; 20% for export manufactur<strong>in</strong>g or process<strong>in</strong>g andsome tourist facilities; 10% for new manufactur<strong>in</strong>g <strong>in</strong>ngrowth po<strong>in</strong>t area.O THER S UB-SAHARAN C OUNTRIESGhana 25 % Ghana offers corporate <strong>in</strong>come tax<strong>in</strong>centives allocated on a sectoral andlocational basis. Sectors of particular<strong>in</strong>terest are agriculture/agribus<strong>in</strong>ess,f<strong>in</strong>ancial services, and hotels. Exports arealso of <strong>in</strong>terest. Industries located outsideof the regional capitals of Accra and Temareceive favorable <strong>in</strong>centives, and firmslocated <strong>in</strong> Free Trade Zones receiveextensive <strong>in</strong>centives.8% withhold<strong>in</strong>g(residents and nonresidents)Residentsprogressive to 25%;non-residents 15%Favorable tax rates for Free Zones (8%); favorable ratesfor agribus<strong>in</strong>ess’ that source locally and are located <strong>in</strong>Accra (20%) and other regional capitals (10%); andcomplete exemptions for agribus<strong>in</strong>ess located outside ofcapitals.


CountryStatutoryCorporateIncome <strong>Tax</strong>Rate aSummary ofInvestment IncentivesTreatment ofDividends ofBus<strong>in</strong>ess AssetsPersonalIncome <strong>Tax</strong>Marg<strong>in</strong>alRatesFavorable <strong>Tax</strong> Rates or ExemptionsKenya30 % (Resident)37.5% (Non-Resident)Kenya has a stream l<strong>in</strong>ed system where<strong>in</strong>centives are primarily provided to theEPZ, and any additional <strong>in</strong>centives arereserved for targeted priorities. Under thenew (more restrictive) FDI regimeimplemented <strong>in</strong> 2004, one of the threecriteria for receiv<strong>in</strong>g an InvestmentCertificate is that the company contributesto tax revenues or other governmentrevenues. c5% withhold<strong>in</strong>gresidents; 10%withhold<strong>in</strong>g nonresidents;0% forcompanies thatcontrol 12.5% ormore of capital10-15-20-25 % 25% tax rate for ten years follow<strong>in</strong>g the tax holiday forEPZ companies only. Companies that float a m<strong>in</strong>imum of20 – 30% of their capital on the Kenya stock exchangereceive a preferable tax rate of 27 – 25% respectively forfive years from the year of list<strong>in</strong>g.Senegal 33% Senegal’s <strong>in</strong>centive code does not provideof tax holidays or favorable <strong>in</strong>come taxrates, but rather grants tax credits for newenterprises and extension projects. The<strong>in</strong>centive system is relatively simple andstraight forward.16% withhold<strong>in</strong>g(residents and nonresidents)Progressive up to 50%CFCE exemption for 5 or 8 yrs. if 200 jobs created or90% of jobs outside Dakar or if 25% <strong>in</strong>crease <strong>in</strong>production capacity or <strong>in</strong>vestments over $100m FCFAfor extensions; custom duties & VAT cancelled;exemption on taxes on salaries, property tax, taxes on<strong>in</strong>come for stocks & sharesUganda 30 % Uganda has undergone significant reformsto its <strong>in</strong>vestment <strong>in</strong>centive policy <strong>in</strong> recentyears. The new code has replaced taxholidays, with a system of tax allowanceswith the exception of EPZs.Notes15% withhold<strong>in</strong>g(residents and nonresidents)30 % None <strong>in</strong>dicated.a <strong>Tax</strong> rates for Botswana, Ghana, <strong>Mozambique</strong>, Tanzania, and Uganda are from World Development Indicators 2006 figures for highest marg<strong>in</strong>al corporate tax rate. Figures for Gambiaand Zambia are taken from their <strong>in</strong>vestment promotion agencies. Kenya’s rate is from the East African <strong>Tax</strong> Guide 2008 of Price Waterhouse Coopers; Rwanda (2006) and Senegal’s(2007) rates are from the Federation of International Trade Associations (FITA).b Botswana has a special case of a 15% corporate <strong>in</strong>come tax plus a 10% additional tax.c UNCTAD, Investment Policy Review Kenya.


Table C-1Overview of Incentives (Part 2 )Country<strong>Tax</strong> Holidays(Full/partial, duration)Special Deductions forEmployment or Tra<strong>in</strong><strong>in</strong>gInvestment <strong>Tax</strong>Credit Or <strong>Tax</strong>RebateInitial Capital Allowance orAccelerated DepreciationAngolaProjects of national <strong>in</strong>terest or projects located <strong>in</strong> specialdevelopment zones – total exemption from corporate<strong>in</strong>come tax from three up to five years. Also reduction of50% of corporate <strong>in</strong>come tax for up to ten years. The<strong>in</strong>centives are granted by the M<strong>in</strong>ister of F<strong>in</strong>ance.None <strong>in</strong>dicated.None (employmentsubsidy <strong>in</strong>stead).None Indicated.Investments <strong>in</strong> agriculture, farm<strong>in</strong>g, transformative<strong>in</strong>dustries, transportation, education and health, canbenefit from an eight to 15 year corporate <strong>in</strong>come taxholiday, depend<strong>in</strong>g on the <strong>in</strong>vestment's geographicallocation.Botswana 5 years typically with development approval order (DAO). Deduction of 200% of the cost oftra<strong>in</strong><strong>in</strong>g if approved by theCommissioner. In the event thatit’s a manufactur<strong>in</strong>g bus<strong>in</strong>ess,and is approved by the MOFthere is a reduction <strong>in</strong> corporatetax to 15%, i.e. 5% company tax& 10% additional company tax.None <strong>in</strong>dicated.Annual allowance of 10% - 25% of plant and mach<strong>in</strong>ery canbe claimed. An <strong>in</strong>itial allowance of 25% for new or improvedbuild<strong>in</strong>gs used for <strong>in</strong>dustrial purposes.There is immediate depreciation of m<strong>in</strong><strong>in</strong>g capitalexpenditure.DR Congo3-5 years exemption for new companies;None <strong>in</strong>dicated. None <strong>in</strong>dicated. 60% ICA for manufactur<strong>in</strong>g export<strong>in</strong>g > 20% of outputDiscretionary exemptions under contractual regimeLesotho None <strong>in</strong>dicated. 125% for tra<strong>in</strong><strong>in</strong>g or tertiaryeducation costs formanufactur<strong>in</strong>g companies.None Indicated.None Indicated.Madagascar None Indicated. None Indicated. None Indicated. None Indicated.MalawiPriority <strong>in</strong>dustries have options of 5-10 year tax holiday(depend<strong>in</strong>g on size of <strong>in</strong>vestment) or fixed 15% tax rate.Additional 50% of tra<strong>in</strong><strong>in</strong>g costfor employee to earn degree,diploma or certificateNone Indicated.Full expens<strong>in</strong>g of farm works, <strong>in</strong>dustrial build<strong>in</strong>gs, railwaysl<strong>in</strong>es, 40% ICA for manufacturers ( additional 15% <strong>in</strong>designated areas)Mauritius4 years for small enterprises converted <strong>in</strong>to companies andwhich register for the first time with Income <strong>Tax</strong>10 years for foreign <strong>in</strong>come of certified regionalheadquarters; tax holiday (or 15% tax rate) for <strong>in</strong>vestmentunder ICT schemeNone Indicated.10% on <strong>in</strong>vestment bycompanies <strong>in</strong> certa<strong>in</strong>categories other than tax<strong>in</strong>centive companies, suchthat tax payable is not lessthan 15%.10-25% additional ICA on <strong>in</strong>dustrial premises, plant andmach<strong>in</strong>ery, computer software, and state of art technology <strong>in</strong>manufactur<strong>in</strong>g; for ICT equipment 50% ICA plus 3 year writeoff at 33.3% per year.


Country<strong>Tax</strong> Holidays(Full/partial, duration)Special Deductions forEmployment or Tra<strong>in</strong><strong>in</strong>gInvestment <strong>Tax</strong>Credit Or <strong>Tax</strong>RebateInitial Capital Allowance orAccelerated Depreciation<strong>Mozambique</strong> None <strong>in</strong>dicated. 5% for 5 yrs. 10% for tra<strong>in</strong><strong>in</strong>gsfor technologically advancedequipment.5% of total <strong>in</strong>vestment,duration 5 years.10% for projects <strong>in</strong> Gaza,Sofala, Tete & ZambeziaProv<strong>in</strong>ces. 15% <strong>in</strong> CaboDelgado, Inhambane andNiassa Prov<strong>in</strong>ces. +3%for projects <strong>in</strong> hotel andtourism. RDZs developersget 20% decrease <strong>in</strong> CFI.Full exemption of special equipment for advancedtechnology, up to max of 15% of taxable <strong>in</strong>come;120-150%(depend<strong>in</strong>g on location) for <strong>in</strong>vestment <strong>in</strong> public utility<strong>in</strong>frastructure.2 times normal rate for new or rehabilitated immovableassets, mach<strong>in</strong>ery & equipment used <strong>in</strong> <strong>in</strong>dustrial & agro<strong>in</strong>dustrialactivities.Namibia 50% abatement for 5 years, phas<strong>in</strong>g out over follow<strong>in</strong>g 10years for registered manufacturersManufacturers qualify foradditional deduction on tra<strong>in</strong><strong>in</strong>gexpenses and 25% additionaldeduction for production l<strong>in</strong>ewages.None Indicated.20% ICA on build<strong>in</strong>gs, with 8% per year write-off of balance<strong>in</strong> manufactur<strong>in</strong>g; full expens<strong>in</strong>g of farm works; 3 year writeofffor dev. Expenditure <strong>in</strong> m<strong>in</strong><strong>in</strong>g and petroleum.Seychelles None Indicated. 150%-200% deductions None Indicated. 20% additional ICA on manufactur<strong>in</strong>g plant; Under IPA:45-40-30-25-10% ( total 150%) for capital assets <strong>in</strong> mfg, tourismand small <strong>in</strong>dustry;45-40-20-15-5 ( total 120%) <strong>in</strong> agriculture,mar<strong>in</strong>e resources and professional servicesSouth Africa6 year for export companies <strong>in</strong> Industrial DevelopmentZones ended <strong>in</strong> 1999Additional deduction up toR50,000 per employee underapproved leadership programs.None Indicated.Full expens<strong>in</strong>g of capital for farm development andm<strong>in</strong><strong>in</strong>g;(2) 50-40-30-20% for farm mach<strong>in</strong>ery and eqpt;(3)50% to 100% additional allowance for <strong>in</strong>dustrial asset <strong>in</strong>qualify<strong>in</strong>g for strategic projects ( subject to cap)Swaziland5-year holiday new export manufactur<strong>in</strong>g <strong>in</strong>dustryDevelopmental Approval Order gives 10 year holiday at10% tax rate + exemption from dividend w/hold<strong>in</strong>g. Note:<strong>Tax</strong> applies to excess <strong>in</strong>come, as per formula; additionalholidays granted at MoF discretion.200% deduction for cost ofapproved tra<strong>in</strong><strong>in</strong>g expensesNone <strong>in</strong>dicated.50% additional ICA for plant and mach<strong>in</strong>ery <strong>in</strong>manufactur<strong>in</strong>g, for <strong>in</strong>frastructure assets, for hotels;50% ICA for farm build<strong>in</strong>gs and employee hous<strong>in</strong>g;Full write off of capital for m<strong>in</strong><strong>in</strong>g and farm developmentTanzaniaZambia10 -20 year full tax holiday for EPZs followed by 24% taxrate (details depend on location).Special agreements for tourism <strong>in</strong> Liv<strong>in</strong>gstone; 5 yearexemption for some small scale <strong>in</strong>dustry ; one –seventhreduction for rural enterprise for 5 years.None Indicated. None Indicated. 50% for <strong>in</strong>vestment <strong>in</strong> lead and priority sectors (reduced from100% <strong>in</strong> 2002); 20% for <strong>in</strong>dustrial build<strong>in</strong>g, mach<strong>in</strong>ery andfarm works; 100% for agriculture and m<strong>in</strong>eral sectors; 50%for tourism;50 – 12.5% for m<strong>in</strong><strong>in</strong>g.None Indicated. None Indicated. 10% allowance on low cost hous<strong>in</strong>g and 5% for otherbuild<strong>in</strong>gs <strong>in</strong> the manufactur<strong>in</strong>g sector.100% deduction on capital expenditures on build<strong>in</strong>gs,railways, equipment, and shaft s<strong>in</strong>k<strong>in</strong>g for m<strong>in</strong><strong>in</strong>g; 50% ofplant and mach<strong>in</strong>ery for tourism.10% <strong>in</strong>vestment allowance plus 10% ICA for <strong>in</strong>dustrial


Country<strong>Tax</strong> Holidays(Full/partial, duration)Special Deductions forEmployment or Tra<strong>in</strong><strong>in</strong>gInvestment <strong>Tax</strong>Credit Or <strong>Tax</strong>RebateInitial Capital Allowance orAccelerated Depreciationbuild<strong>in</strong>gs; expens<strong>in</strong>g of farm works; 10% ICA for <strong>in</strong>vestment<strong>in</strong> certa<strong>in</strong> tree and bush cropsZimbabwe5-yr holiday + 5 yrs at 15% for qualify<strong>in</strong>g <strong>in</strong>vestors; 5 yrsfor EPZ companies, followed by 15% rate; 5 yrs each at0%,15% and 20% for BOOT arrangement and tourismfacility <strong>in</strong> tourist zone; l-yr discount of 2% granted to anewly listed company of the stock exchangeDouble tax deduction on wagesand salaries for additionalemployees <strong>in</strong> manufactur<strong>in</strong>gNone Indicated.Expens<strong>in</strong>g of certa<strong>in</strong> farm works and m<strong>in</strong><strong>in</strong>g <strong>in</strong>vestment;15% ICA for companies <strong>in</strong> growth po<strong>in</strong>t area; 25% ICA peryear for 4 years on <strong>in</strong>dustrial and commercial build<strong>in</strong>gs andmach<strong>in</strong>ery; 50% Special Initial Allow on most capital assetsGhana5-10 year full tax holiday for: Real estate, rural banks,farm<strong>in</strong>g and agro-<strong>in</strong>dustry, waste process<strong>in</strong>g and freezones.None <strong>in</strong>dicated.25 – 50 percent rebate formanufactur<strong>in</strong>g <strong>in</strong>dustrieslocated <strong>in</strong> alternativeregional capitals.40% for computers and data handl<strong>in</strong>g equipment; 30% ofautomobile, trailers, plant and mach<strong>in</strong>ery used <strong>in</strong>manufactur<strong>in</strong>g and plantation equipment; 80% cost <strong>in</strong> year ofpurchase with 50% annually there after for transportationequipment, build<strong>in</strong>gs and plant and mach<strong>in</strong>ery for m<strong>in</strong><strong>in</strong>g andpetroleum.5% on mach<strong>in</strong>ery and equipment <strong>in</strong> all sectors exceptbank<strong>in</strong>g, f<strong>in</strong>anc<strong>in</strong>g, <strong>in</strong>surance, m<strong>in</strong><strong>in</strong>g and petroleum.Kenya 10 year full tax holiday for EPZ companies only. None <strong>in</strong>dicated. None <strong>in</strong>dicated. Across the sectors there is straight-l<strong>in</strong>e depreciation ofbuild<strong>in</strong>gs (2.5%), mach<strong>in</strong>ery (37.5%), IT equipment (30%),vehicles (25%), office equipment (12.5%). M<strong>in</strong><strong>in</strong>g, andfarm<strong>in</strong>g, hotels and manufactur<strong>in</strong>g receive exceptions foraccelerated depreciation.Senegal None Indicated. None Indicated. 40% tax credit on theeligible <strong>in</strong>vestment for 5yrs.; new enterprises: 50%taxable profit for newcompanies; extensions:25% taxable profit forextension projectsNone Indicated.Uganda 3-6 year general tax holidays were repealed <strong>in</strong> the 1997F<strong>in</strong>ance Statue and were replaced a new <strong>in</strong>vestmentregime.10 year full tax holiday for companies <strong>in</strong> EPZs.Investment capital allowance of100% for tra<strong>in</strong><strong>in</strong>g.None <strong>in</strong>dicated.40% for computer and data handl<strong>in</strong>g equipment. 35% forlight vehicles.30% for heavy vehicles.20% for other depreciable assets and for farm equipment. 5%<strong>in</strong>dustrial build<strong>in</strong>g allowance, 20% for horticulture, variablerate for <strong>in</strong>tangible assets.SOURCE: Bolnick (2004), updated for most countries with data accessed from the Internet between March 15 and April 14, 2008 from Investment Promotion Centers <strong>in</strong> Botswana, Ghana,Kenya, Senegal, Tanzania, Uganda, and Zambia; and data accessed from the Internet between July 27 and August 6, 2009 from 2009 <strong>Tax</strong> Highlights guides produced by Deloitte (forAngola, Botswana, Madagascar, <strong>Mozambique</strong>, Namibia, South Africa, Tanzania and Zambia) and <strong>Tax</strong> Guides produced by PKF International (for Angola, Mauritius, South Africa, andUganda); plus for Lesotho, Southern Africa, US Embassy; for Seychelles, Revenue Commission; for Namibia, M<strong>in</strong>istry of Trade and Industry; for Senegal, Federation of InternationalTrade Associations.


Appendix D. <strong>Tax</strong>ation andBus<strong>in</strong>ess Environment Rat<strong>in</strong>gsChapter 3 reviews the scores for <strong>Mozambique</strong>’s tax system from the World Bank’s annual Do<strong>in</strong>gBus<strong>in</strong>ess reports. This appendix exam<strong>in</strong>es tax system scores <strong>in</strong> two other sources of <strong>in</strong>ternationalcomparisons: the World Bank’s Investment Climate Assessments, as tabulated <strong>in</strong> the AfricanCompetitiveness Report for 2009, and the World Economic Forum’s Global CompetitivenessReport.The World Bank has conducted face-to-face enterprise surveys <strong>in</strong> more than 110 countries tocapture bus<strong>in</strong>ess perceptions of the major obstacles to enterprise growth, the relative importanceof various constra<strong>in</strong>ts to <strong>in</strong>creas<strong>in</strong>g employment and productivity, and the effects of eachcountry’s bus<strong>in</strong>ess environment on its <strong>in</strong>ternational competitiveness. The Bank publishes theresults <strong>in</strong> Investment Climate Assessments, cover<strong>in</strong>g a broad range of bus<strong>in</strong>ess-environmenttopics for each country, <strong>in</strong>clud<strong>in</strong>g access to f<strong>in</strong>ance, corruption, <strong>in</strong>frastructure, crime,competition, and performance measures. Selected results for Africa have been compiled <strong>in</strong> theAfrica Competitiveness Report 2009 (ACR 2009), which is a jo<strong>in</strong>t publication of the WorldEconomic Forum, the World Bank, and the African Development Bank.The ACR 2009 reports five <strong>in</strong>dicators relat<strong>in</strong>g to the tax and customs environment:1. Average number of visits or required meet<strong>in</strong>gs with tax officials.2. Percentage of firms stat<strong>in</strong>g that they are expected to give a gift <strong>in</strong> meet<strong>in</strong>gs with taxofficials.3. Percentage of firms express<strong>in</strong>g that a typical firm reports less than 100% of sales for taxpurposes.4. Average time to clear direct exports through customs (days)5. Average time to claim imports from customs (days)Table D-1 presents the data on these five <strong>in</strong>dicators for the SADC region. The results for<strong>Mozambique</strong> po<strong>in</strong>t to the follow<strong>in</strong>g problems:<strong>Tax</strong> evasion: 73 percent of firms expressed the view that a typical firm reports less than 100percent of sales for tax purposes. This <strong>in</strong>cluded 78 percent of the small firms, and 63 percent ofthe large firms. For this <strong>in</strong>dicator, <strong>Mozambique</strong> has the second worst score <strong>in</strong> the SADC region.


28 <strong>PARPA</strong> <strong>II</strong> R EVIEW—THE T AX S YSTEM IN M OZAMBIQUECustoms: Time to clear direct exports through customs and time to claim imports from customsboth averaged just over 10 days. Medium-sized firms appear to be most affected by customsdelays. They reported that time to clear direct exports averaged 14 days and time to claim importsaveraged 12 days. Here, too, <strong>Mozambique</strong>’s scores are amongst the poorest <strong>in</strong> the region.Corruption: 10 percent of the firms <strong>in</strong> the survey expected to give gifts <strong>in</strong> meet<strong>in</strong>gs with taxofficials, <strong>in</strong>clud<strong>in</strong>g 12 percent of small firms, 7.5 percent of medium sized firms and 4 percent oflarge firms. Although 90 percent of the respondents <strong>in</strong> <strong>Mozambique</strong> do not report hav<strong>in</strong>g to givegifts to tax officials, the score on this <strong>in</strong>dicator is still above the median for the region.Table D-1Investment Climate Profile <strong>Tax</strong> Relevant IndicatorsAvg. No. ofVisits orRequiredMeet<strong>in</strong>gs with<strong>Tax</strong> OfficialsFirms Expectedto Give Gifts <strong>in</strong>Meet<strong>in</strong>gs With<strong>Tax</strong> Officials(%)Firms Express<strong>in</strong>gThat a TypicalFirm ReportsLess than 100%of Sale for <strong>Tax</strong>Purposes, %Avg. Time toClear DirectExportsThroughCustoms(days)Avg. Time toClaimImports fromCustoms,(days)Angola 5.2 14.8 67.8 16.5 28.2Botswana 2.4 4.5 65.3 1.3 3.1DRC 10 64.4 65.4 3.6 13Lesotho 3.2 10.6 . 8 .Madagascar 1.7 6.8 35.6 14.2 19.3Malawi 8.9 15.3 55.3 3.5 6.4Mauritius 3.1 0.3 36.2 10.3 11.7<strong>Mozambique</strong> 2.7 9.8 73.1 10.1 10.4Namibia 1.6 2.6 45.5 1.5 3.3Seychelles . . . . .South Africa 1.8 3.1 40.3 4.6 5.9Swaziland 1.9 3.3 74.6 4 2.2Tanzania 3.3 14.7 71 5.7 14.3Zambia 2.9 5.4 . 3.1 6.6Zimbabwe . . . . .SADC Median 2.9 6.8 65.3 4.6 8.5SOURCE: World Economic Forum.Accord<strong>in</strong>g to the most recent Investment Climate Profile for <strong>Mozambique</strong> (2007), high tax ratesare among the top ten most serious constra<strong>in</strong>ts to <strong>in</strong>vestment as perceived by entrepreneurs. Thiscompla<strong>in</strong>t is found throughout the SADC region, with the exception of South Africa (see TableD-2). <strong>Tax</strong> rates are cited as a problem even <strong>in</strong> Zambia, which has one of the lowest scores <strong>in</strong> theworld on the Do<strong>in</strong>g Bus<strong>in</strong>ess estimate of Total <strong>Tax</strong> Rate (but see Exhibit 3.1 <strong>in</strong> the text for acritique of the methodology for this <strong>in</strong>dicator.) Tanzania, Zambia, and Madagascar also cite taxadm<strong>in</strong>istration <strong>in</strong> their top ten lists.


A PPENDIX D 29Table D-2<strong>Tax</strong>es <strong>in</strong> Top 10 Most Serious Constra<strong>in</strong>ts to Investment, asPerceived by Local Entrepreneurs<strong>Tax</strong>Rates<strong>Tax</strong>Adm<strong>in</strong>istrationAngolaBotswanaXXDRC . .LesothoXMadagascar X XMalawiMauritius<strong>Mozambique</strong>NamibiaXXXXSeychelles . .South AfricaSwazilandXTanzania X XZambia X XZimbabwe . .SOURCE: World Economic Forum.Similar results were found <strong>in</strong> the two enterprise surveys conducted <strong>in</strong> <strong>Mozambique</strong>, one <strong>in</strong> 2003and one <strong>in</strong> 2008, where tax adm<strong>in</strong>istration was perceived by entrepreneurs to be a majorconstra<strong>in</strong>t to bus<strong>in</strong>ess (see Table D-3). In 2003, it was the 10th most severe obstacle to bus<strong>in</strong>ess;<strong>in</strong> 2008, it was the 8th most severe obstacle.Another widely cited source of bus<strong>in</strong>ess environment rat<strong>in</strong>gs is the World Economic Forum’sGlobal Competitiveness Report. This annual report assesses “the ability of countries to providehigh levels of prosperity to their citizens” 6 based on scores for more than 113 <strong>in</strong>dicators. Most ofthe data come from an annual Executive Op<strong>in</strong>ion Survey, which obta<strong>in</strong>s subjective rat<strong>in</strong>gs on abroad range of issues that are difficult to measure us<strong>in</strong>g hard data. Tally<strong>in</strong>g the scores, the reportranks country performance <strong>in</strong> the follow<strong>in</strong>g areas: <strong>in</strong>stitutions; <strong>in</strong>frastructure; macroeconomicstability; health and primary education; higher education and tra<strong>in</strong><strong>in</strong>g; goods market efficiency;labor market efficiency; f<strong>in</strong>ancial market sophistication; technological read<strong>in</strong>ess; market size;bus<strong>in</strong>ess sophistication; and <strong>in</strong>novation.On taxation, the Global Competitiveness Report for 2007-2008 conta<strong>in</strong>s only two <strong>in</strong>dicators otherthan those taken from the World Bank’s Do<strong>in</strong>g Bus<strong>in</strong>ess reports (discussed <strong>in</strong> the ma<strong>in</strong> text): the“extent and effect of taxation” and the trade-weighted average tariff.6 http://www.weforum.org/en/<strong>in</strong>itiatives/gcp/FAQs/<strong>in</strong>dex.htm#network3. Accessed July 6, 2009.


30 <strong>PARPA</strong> <strong>II</strong> R EVIEW—THE T AX S YSTEM IN M OZAMBIQUETable D-3Major or Severe Obstacles to Bus<strong>in</strong>ess2008 2003Practices of <strong>in</strong>formal competition 1 5Access to f<strong>in</strong>ance 2 1Crime 3 8Corruption 5 3Electricity 6 2Transport 7 14<strong>Tax</strong> adm<strong>in</strong>istration 8 10Workforce education 9 12Licens<strong>in</strong>g & permits 10 13Customs & trade regulations 11 9Access to land 12 15Telecommunications 13 16Political <strong>in</strong>stability 14 .Labor regulation 15 11Courts 16 .SOURCE: <strong>Mozambique</strong> Enterprise Surveys 2003, 2008.For the first <strong>in</strong>dicator, respondents are asked to rate “the level of taxes <strong>in</strong> your country” on a scaleof 1 to 7, where 1 = significantly limits the <strong>in</strong>centives to work or <strong>in</strong>vest, and 7 = has little impacton the <strong>in</strong>centives to work or <strong>in</strong>vest. Table D-4 shows that the perceived “extent and effect oftaxation” <strong>in</strong> <strong>Mozambique</strong> has not exhibited major improvement over the <strong>PARPA</strong> <strong>II</strong> period,hover<strong>in</strong>g around a score of 3. Though the validity of cross-country comparisons us<strong>in</strong>g perceptionsdata between countries is questionable, the result for <strong>Mozambique</strong> is <strong>in</strong> l<strong>in</strong>e with median amongSADC countries, though well below the results for the lead<strong>in</strong>g countries of South Africa andMauritius.Table D-5 shows that the trade weighted average tariff for <strong>Mozambique</strong> is relatively low <strong>in</strong>absolute terms <strong>in</strong> l<strong>in</strong>e with the SADC medians, though higher than the average tariff <strong>in</strong> SouthAfrica and Mauritius.Table D-4Extent and Effect of <strong>Tax</strong>ation2005/2006 2006/2007 2007/2008 2008/2009Angola . . . .Botswana 4.9 4.7 4.8 4.9DRC . . . .Lesotho . 3.1 2.9 3.0Madagascar 3.0 3.1 3.3 3.5Malawi 2.1 2.9 . 3.0Mauritius 3.8 4.7 4.8 5.4


A PPENDIX D 312005/2006 2006/2007 2007/2008 2008/2009<strong>Mozambique</strong> 3.3 2.7 3.0 3.1Namibia 3.7 3.6 3.6 3.7Seychelles . . . .South Africa 3.7 4.2 4.3 4.5Swaziland . . . .Tanzania 3.5 3.4 3.5 3.4Zambia . 2.1 2.5 2.8Zimbabwe 2.6 2.5 2.4 2.2SADC median 3.5 3.1 3.2 3.1SOURCE: Global Competitiveness Report.Table D-5Trade-weighted Tariff2008/2009Angola .Botswana 4.6DRC .Lesotho 4.2Madagascar 8.4Malawi 12.7Mauritius 3.6<strong>Mozambique</strong> 7.7Namibia 8.5Seychelles .South Africa 6.2Swaziland .Tanzania 7.7Zambia 11.6Zimbabwe 13SADC median 7.7NOTE: The rate is the average rate of duty per imported value unit <strong>in</strong> 2007, weighted by 2006 import values. Time series data arenot available, as the methodology for calculat<strong>in</strong>g the <strong>in</strong>dicator was changed <strong>in</strong> the 2008/09 report.SOURCE: Global Competitiveness Report.


Appendix E. OperationalEfficiency IndicatorsAs noted <strong>in</strong> the text, the Annual Reports (ARs) of the AT are filled with tables and figuresprovid<strong>in</strong>g valuable data on many facets of revenue collection and tax adm<strong>in</strong>istration. Yet thereports also lack many types of <strong>in</strong>dicators that would provide management with better<strong>in</strong>formation for monitor<strong>in</strong>g operational efficiency (<strong>in</strong>dicadores de desempenho). The purpose ofthis appendix is to clarify this statement.In its Annual Report for 2008 (AT, 2009) the AT presents 39 tables and 43 graphs. Without go<strong>in</strong>g<strong>in</strong>to full detail, the ma<strong>in</strong> types of <strong>in</strong>formation are as follows:• Revenues collected, relative to targets, by type of revenue• Revenue foregone from the temporary suspension of tax on petroleum products <strong>in</strong> 2008• Revenue contribution from mega-projects and f<strong>in</strong>ancial <strong>in</strong>stitutions• Number of audits and <strong>in</strong>spections, and revenue results, <strong>in</strong>clud<strong>in</strong>g post-clearance audits bycustoms• <strong>Tax</strong> disputes by status and region (number and tax amounts <strong>in</strong>volved)• Refund requests received, paid , outstand<strong>in</strong>g, by type of tax• Revenue foregone due to fiscal benefits, by type of tax• New registrations by companies and <strong>in</strong>dividuals, relative to targets• Staff<strong>in</strong>g levels, <strong>in</strong>clud<strong>in</strong>g education levels• Costs of tax adm<strong>in</strong>istration, budget and actual• Rat<strong>in</strong>g of tax units, by realization of revenue targets• New <strong>in</strong>vestments <strong>in</strong> AT <strong>in</strong>frastructure• Revenue recovered as a result of <strong>in</strong>ternal audits• Efficiency of tax service offices, operational cost as % of collectionsTable E-1 summarizes selected <strong>in</strong>dicators of operational performance from the AT AnnualReports for 2007 and 2008, show<strong>in</strong>g time series comparisons where possible.All of this <strong>in</strong>formation is important, but only the last-mentioned item on the forego<strong>in</strong>g list (fromTable 39, p. 81 of the 2008 Annual Report) provides direct <strong>in</strong>formation about cost-effectiveness


34 <strong>PARPA</strong> <strong>II</strong> R EVIEW—THE T AX S YSTEM IN M OZAMBIQUEor operational efficiency, as dist<strong>in</strong>ct from descriptive data on the level of activity, the outcomes,and characteristics of the organization.What is miss<strong>in</strong>g? The latest IMF tax mission report, which is not available to the public (IMF,2009c) provides a full page of suggested performance <strong>in</strong>dicators that are standard elements of awell developed management <strong>in</strong>formation system for tax adm<strong>in</strong>istration. Only few of thesuggested <strong>in</strong>dicators are found <strong>in</strong> the AT Annual Report, and, judg<strong>in</strong>g from <strong>in</strong>terview results, theAT does not yet collect most of the miss<strong>in</strong>g <strong>in</strong>formation. Examples <strong>in</strong>clude:• Survey f<strong>in</strong>d<strong>in</strong>gs on taxpayer satisfaction• Survey f<strong>in</strong>d<strong>in</strong>gs on AT personnel satisfaction• Number of taxpayer service requests handled, by function and location• Quality of <strong>in</strong>formation provided to taxpayers• Number of declarations processed, by type of tax• Process<strong>in</strong>g time per declaration• Process<strong>in</strong>g time for refund requests• Number of audits per audit staff and type of contributor• Average time per audit• Percent of audits f<strong>in</strong>alized and accepted by taxpayer without contest• Percent of challenged audits settled <strong>in</strong> favor of the government• Value and number of debts collected, and outstand<strong>in</strong>g• Average time, and stratification by value of outstand<strong>in</strong>g tax debts• Average time for contact<strong>in</strong>g taxpayers about overdue payments, and <strong>in</strong>itiat<strong>in</strong>g action• Average time for resolv<strong>in</strong>g disputes• Average time for customs <strong>in</strong>spections• Average amounts collected per post-clearance audit• Percent of customs declarations, by channelAs the tax data base develops, the AT should also be collect<strong>in</strong>g data that will allow managementto target audits, <strong>in</strong>spections, and verifications more efficiently by allocat<strong>in</strong>g AT resources totransactions <strong>in</strong>volv<strong>in</strong>g high revenue risk.


Table E-1AT Operational IndicatorsTopic Item Source NotesAuditAuditorias e fiscalizaçõesNumberOther useful audit <strong>in</strong>dicatorsDAFI DAFI/UGC Total Amount due per audit/process2006 500 AR2007, p14 Amount collected per audit/process2007 498 460 958 AR2007, Table 9 Amount collected/due2008 769 AR2008, Table 18 Number of audit staffNumber of audits per audit staffAmount due (10^6 MT)Amount due per audit staffDAFI DAFI/UGC Total Amount collected per audit staff2006 660.3 AR2007, p14 Audit revenue, % GDP2007 780.3 AR2007, Table 102008 352.8 AR2008, Table 14, 18Amounts collected (milhoes MT, excl multas)DAFI DAFI/UGC Total2006 na na2007 124.1 14 138.1 AR2007 Table 12 = 0.4% total receipts: AR2008, table 182008 89.4 AR2008, Table 15, 18 = 0.2% total receipts: AR2008, table 18Processos de contas das empresas analisadas (a/c)No.Add’l tax collected (10^6 MT, excl multas))2006 1637 AR2007 p152007 1336 17.8 AR2007 Table 122008 1478 93.1 AR2008 Table 17Operações de prevenção e combate a fuga ao fiscoNo,Collected2006 na na2007 40 na AR2007 p142008 na naAuditorias Pós-Desembaraço AduaneiroNo. Value expected Collected (10^6 MT)2007 59 12.8 5.04 AR2007 Table 112008 63 89.3 12.0 AR2008 Table 16Registration Registo de contribu<strong>in</strong>tes Other useful registration <strong>in</strong>dicatorsIRPS IRPC Total Total number registered, by type of tax2006 96448 3779 100227 AR2007 Table 15 No. monthly returns / No. registered, by type of tax2007 186368 3651 190019 AR2007 Table 152008 186671 4469 191140 AR2008 Table 30 Regional breakdown availableNumber of <strong>Tax</strong>payers2005 295000 2005: IMF (2006) PRGF 4th review2006 391719 2006: IMF (2007) PRGF 6th Review2007 587205 2007: IMF (2008) PSI 2nd Review2008 na


Disputes (contensiosos) Evolução de processos de contencioso fiscal (year-end)No.Value (10^6 MT)2006 8333 1169.6 AR2007 Table 162007 8289 1315.3 AR2007 Table 162008 7606 916.24 AR2008 Table 19of which No. 2007 No. 2008 Value 2007 Value 2008 AR2008 Table 19Instaurados 6.476 4489 484.4 491.4Cobrados 3.638 5998 223.4 171.7Anulados 27 9 7.9 1.9Relaxados 2.855 1224 107.4 218.0Contestados 812 1229 721.3 69.2Recorr. a TA 134 230 45.1 321.5Evolução de processos executivos (??)No. Value (10^6 MT) No ValueSaldo 2005 224916 1127.5 AR2007 Table 17Instaurados 7408 390.5 AR2007 Table 17Cobr + Anul 12887 124 AR2007 Table 17Saldo 2006 219432 1380.0 AR2007 Table 17Instaurados 4062 748.6 AR2007 Table 17Cobr + Anul 3014 63.3 AR2007 Table 176656 26.8Saldo 2007 213824 2036.6 AR2007 Table 17Instaurados 4512 498.9 AR2008 Table 23Cobrados 3390 85.5 AR2008 Table 23Anulados 20833 990.7 AR2008 Table 23Saldo 2008??? 189605 1748.7 AR2008 Table 23Evolução de processos de contencioso aduaneiroNumbers small -- omit AR2007 Table 18Refunds IVA - Pedidos de reembolso tratados e pagos (excl diplomatas) Other useful refund <strong>in</strong>dicatorsNo. Value (10^6 MT) % petitions closed w/<strong>in</strong> 30 days2006 na na % petitions overdue at year-end2007Received 389 941.4 AR2007 Table 19Paid 385 562.2 AR2007 Table 19 Incl pyts on petitions from prior yrs2008Received 430 1071.9 AR2008 Table 25Paid 487 863.1 AR2008 Table 25 Incl pyts on petitions from prior yrsIRPS - Pedidos de reembolsoNo.Value (10^6 MT)2006 na na2007 3473 21.6 AR2007 Table 20Autorizados 1069 5.6Indeferido 73 1.7Pendentes 2331 14.32008 4142 28.5 AR2008 Table 27Autorizados 2075 12.7Indeferido 189 0.9Pendentes 1872 14.8IRPC - Pedidos de reembolsoNo.Value (10^6 MT)2006 na na2007 35 59 AR2007 Table 20Autorizados 12 0.5Indeferido 1 0.3Pendentes 22 58.22008 31 94.6 AR2008 Table 28Autorizados 10 11.8Indeferido 1 10.4Pendentes 18 72.3


Staff<strong>in</strong>gDistribution of personnelTributaria Aduaneira Regime Geral Total2006 na na2007 1017 1718 2735 AR2007 Table 212008 1324 1757 10 3091 AR2008 Anexo 2Education level Maestrado Lic/Bach Med. Prof Med.Ger BasicGeral InfBasic2006 na2007 na 15.3% 17.9% 40.0% 22.8% 4.0% AR2007 ???2008 0.7% 16.0% 20.6% 38.7% 13.6% 8.3% AR2008 Anexo 2.1Cost of tax adm<strong>in</strong>.Distribuição pelos Centros de custos - Orc. disponivel (10^6 Mt)2006 2007 2008 Other useful cost <strong>in</strong>dicatorsPessoal na 543.7 940.7 AR2007 Table 23 Operation cost as % receipts - DGI, DGABens e Servicos na 208.4 393.9 AR2008 Table 33 Total cost as % receipts - DGA, DGATransf corr. na 2.0 1Other corr. na 103.2 79.3Desp capital na 129.7 143.8Total na 987.0 1558.8Total tax revenue (10^6 MT) 22142.1 27965.1 32315.2 ATM-GPECI - Execução da Receita do EstadoAT total cost % tax revenue 3.53% 4.82% CalculatedAT operat<strong>in</strong>g cost % tax rev 3.07% 4.38% CalculatedFiscal benefitsIsenções concedidosDireitos IVA ICE IRPC IRPS Total2006 822.8 1534.3 271.8 517.4 0.2 3146.6 TA Rel CGE2007 p.V-332007 923.0 1857.1 112.8 3967.2 0.9 6861.0 TA Rel CGE2007 p.V-332008 710.0 1291.6 188.1 na na 2191.8 AR2008 Table 29% total tax revenue2006 14.2%2007 24.5%2008 6.8% Excludes IRPC ??Megaprojects <strong>Tax</strong>-registered megaprojects, end 2008 AR2008 p28-9ActiveNot yet activeMozal, SARL;Projecto de Ferro e Aço de Maputo,Sasol Petroleum Temane;Projecto da Zona Franca Industrial da Beira e de Ferro e Aço da Beira,Sasol Petroleum Moçambique;Complexo Petroquímico da Beira,Projecto Areias Pesadas de Moma;Riversdale <strong>Mozambique</strong>, Lda,Areias Pesadas de Chibuto;Procana,Hidroeléctrica de Cahora BassaAyrs Logispics (ref<strong>in</strong>aria de petróleo de Nacala).Companhia do Vale do Rio Doce.Contribution of megaproject and f<strong>in</strong>ancial <strong>in</strong>stitution (%) AR2008 Table 112008 % Rec DGI % Rec Tot % PIBMegaprojects 8.4% 5.5% 0.9%Inst F<strong>in</strong> 5.1% 3.3% 0.5%Total 13.5% 8.8% 1.4%Contribution of megaproject, by type of tax (10^6 MT) AR2008 Table 122008 IRPS IRPC Imp s/ Prod <strong>Tax</strong>a Concess Dividendos TotalEnergia 83.5 33.4 0 338.6 0 455.4Petróleo 0 0 0 0 0 0Recursos M<strong>in</strong>erais 103.4 101.2 96.2 841.1 0 1141.7Outros 132.5 284.9 0 0 129.8 547.6Total 319.3 419.5 96.2 1179.6 129.8 2144.3


Appendix F. Revenue EffortJones (2009) carries out a cross-country analysis for <strong>Mozambique</strong> by compar<strong>in</strong>g tax-ratios acrosslow and middle-<strong>in</strong>come countries for the period 1990 to 2003 <strong>in</strong> a panel analysis. He controls fordifferences <strong>in</strong> economic structure and <strong>in</strong>stitutional factors, the implication be<strong>in</strong>g that rema<strong>in</strong><strong>in</strong>gdifferences <strong>in</strong> the tax-ratio between countries relate to differences <strong>in</strong> tax-policy andadm<strong>in</strong>istration. 7 With average tax-ratio (tax revenue over GDP) as the dependent variable, thefollow<strong>in</strong>g are <strong>in</strong>cluded as <strong>in</strong>dependent variables: real GDP per capita; the share of <strong>in</strong>dustry <strong>in</strong>GDP; exports as a percentage of GDP; and imports as a share of GDP. In addition, Jones (2009)attempts to account for long-run <strong>in</strong>stitutional differences by <strong>in</strong>clud<strong>in</strong>g the share of territory <strong>in</strong> thetropics, the colonial power, whether or not the country is landlocked, the cont<strong>in</strong>ent to which thecountry belongs, and whether or not the country is resource-rich. His results lead him to concludethat 13 percent of GDP is a realistic tax-ratio for <strong>Mozambique</strong> given that the actual tax-ratioappears to faithfully follow the predicted tax-ratio accord<strong>in</strong>g to the economic structure and<strong>in</strong>stitutions <strong>in</strong> place.Davoodi and Grigorian (2006) also carry out a cross-country regression analysis of tax-effort onup to 141 countries over the period 1990 to 2004. 8 Similarly to Jones (2009) they use taxrevenues as dependent variable and <strong>in</strong>clude GDP per capita, trade (imports plus exports percapita), and natural resources (oil) as explanatory variables, but also <strong>in</strong>clude the share of urbanpopulation, and the share of agriculture <strong>in</strong> the economy rather than <strong>in</strong>dustry, which may providemore <strong>in</strong>formation regard<strong>in</strong>g the ease or difficulty of collect<strong>in</strong>g taxes. Other variables <strong>in</strong>clude<strong>in</strong>flation and an estimated <strong>in</strong>formal sector share us<strong>in</strong>g <strong>in</strong>strumental variables given that the<strong>in</strong>formal sector share itself is affected by the tax-ratio. 9 Davoodi and Grigorian (2006) alsoattempt to account for <strong>in</strong>stitutional quality by employ<strong>in</strong>g variables from the International CountryRisk Guide. Their results suggest that many low and low-middle <strong>in</strong>come countries collect revenueabove their predicted revenue shares, accord<strong>in</strong>g to their economic characteristics. However, this7 Economic structure variables <strong>in</strong>clude: GDP per capita, imports as a share of GDP, <strong>in</strong>dustry as a share ofGDP, whether or not the country is resource-rich; while <strong>in</strong>stitutions are controlled for <strong>in</strong>directly by<strong>in</strong>troduc<strong>in</strong>g the share of the country found <strong>in</strong> the tropics, whether or not the country is land-locked, andwho the colonial power, all variables variously associated with <strong>in</strong>stitutions <strong>in</strong> the cross-country empiricalliterature on growth.8 Davoodi, H.R., Grigorian, G.A., (2007), “<strong>Tax</strong> Potential vs <strong>Tax</strong> Effort: A Cross-Country Analysis ofArmenia’s Stubbornly Low <strong>Tax</strong> Collection”, IMF Work<strong>in</strong>g Paper WP/07/106.9 To be an <strong>in</strong>strument, a variable must be highly correlated with the <strong>in</strong>dependent variable to be<strong>in</strong>strumented (relevance), but only impact the outcome through that <strong>in</strong>dependent variable (exogeneity).


40 <strong>PARPA</strong> <strong>II</strong> R EVIEW—THE T AX S YSTEM IN M OZAMBIQUErelates to the regression methodology which implies that some countries <strong>in</strong> the sample will havetax-ratios above the estimated regression l<strong>in</strong>e, and some below.Bird et al. (2007) carry out a similar exercise, <strong>in</strong>clud<strong>in</strong>g GDP per capita, the rate of populationgrowth, trade as a share of GDP, the non-agricultural share of GDP, and <strong>in</strong>dex figures forvoice/accountability and corruption from the World Bank’s governance <strong>in</strong>dicators. 10 They use across-section of mean tax-ratio values for the period 1990 to 1999 and f<strong>in</strong>d that <strong>in</strong>deed corruptionand voice/accountability play a strong role <strong>in</strong> determ<strong>in</strong><strong>in</strong>g the tax effort of develop<strong>in</strong>g andtransition countries. S<strong>in</strong>ce these <strong>in</strong>stitutional dummy variables may be endogenously determ<strong>in</strong>ed,someth<strong>in</strong>g overlooked by Davoodi and Grigorian (2006), they check for their importance byemploy<strong>in</strong>g two-stage least squares, where these are modeled as be<strong>in</strong>g determ<strong>in</strong>ed by legal orig<strong>in</strong>,and an ethnic fractionalization <strong>in</strong>dex, lead<strong>in</strong>g them to conclude that “a more legitimate andresponsive state is likely an essential precondition for a more adequate level of tax effort <strong>in</strong>develop<strong>in</strong>g countries.”, and therefore that <strong>in</strong>stitutional factors relat<strong>in</strong>g to corruption andgovernment accountability matter. 11 The variables employed <strong>in</strong> the various studies aresummarized <strong>in</strong> Table F-1.Table F-1Variables Employed <strong>in</strong> <strong>Tax</strong>-Effort StudiesDavoodi &Grigorian(2006)Bird et al.(2007)Jones(2009)GDP per capita x x xAgricultural sharexIndustry shareXNonagricultural sharexTrade shares x x XInstitutions: x x XBureaucratic riskComposite <strong>in</strong>dexxxShare of land <strong>in</strong> tropicsColonial regimeLandlockedXXx10 Bird, R.M., Mart<strong>in</strong>ez-Vasquez, J., Torgler, B., (2007), “<strong>Tax</strong> Effort: The Impact of Corruption, Voiceand Accountability”, CREMA Work<strong>in</strong>g Paper 2007-13, Center for Research <strong>in</strong> Economics, Managementand the Arts, Basel, Switzerland.11 A considerably earlier study on tax-effort focus<strong>in</strong>g on 43 Sub-Saharan African countries is carried outby Stotsky and WoldeMariam (1997) us<strong>in</strong>g panel analysis for the period 1990 to 1995. Their explanatoryvariables <strong>in</strong>clude the share of agriculture <strong>in</strong> GDP, the share of m<strong>in</strong><strong>in</strong>g, the share of manufactur<strong>in</strong>g, percapita <strong>in</strong>come, and the trade shares of GDP. The focus here is on the determ<strong>in</strong>ants of a country’s tax-ratio,their results suggest<strong>in</strong>g that much relies on the share of agriculture <strong>in</strong> GDP, the share of m<strong>in</strong><strong>in</strong>g and theshare of exports. They also test for the impact of IMF programs and f<strong>in</strong>d no strong effect. See J. Stotskyand A. WoldeMariam., (1997), “<strong>Tax</strong> Effort <strong>in</strong> Sub-Saharan Africa”, IMF Work<strong>in</strong>g Paper WP/97/107.


A PPENDIX F 41Davoodi &Grigorian(2006)Bird et al.(2007)Jones(2009)Accountability/VoiceCorruptionxxMETHODOLOGY FOR REVENUE-EFFORT ESTIMATIONA choice must be made between analyz<strong>in</strong>g tax-ratios or revenue ratios. While tax-ratios are thesubject of the above-cited studies, develop<strong>in</strong>g country data on tax-ratios is less readily availablethan that on total revenues. This need not be a problem and <strong>in</strong>deed there is some ambiguity abouthow different countries classify particular revenues, with differences aris<strong>in</strong>g between what is<strong>in</strong>cluded as tax revenue and what is considered non-tax revenue. As such, total revenue is likelyto be more comparable across countries as long as economic characteristics are also taken <strong>in</strong>toaccount.In terms of explanatory variables, similar to Bird et al. (2007) this study employs World Bankgovernance <strong>in</strong>dicators on i) government effectiveness, ii) regulatory quality and iii) control ofcorruption. These are def<strong>in</strong>ed respectively as i) “perceptions of the quality of public services, thequality of the civil service and the degree of its <strong>in</strong>dependence from political pressures, the qualityof policy formulation and implementation, and the credibility of the government's commitment tosuch policies”; ii) “perceptions of the ability of the government to formulate and implementsound policies and regulations that permit and promote private sector development”, <strong>in</strong>clud<strong>in</strong>gtaxes; and iii) “perceptions of the extent to which public power is exercised for private ga<strong>in</strong>,<strong>in</strong>clud<strong>in</strong>g both petty and grand forms of corruption, as well as "capture" of the state by elites andprivate <strong>in</strong>terests” (Kaufman et al., 2009). For all of these, higher scores correspond to betteroutcomes. 12The specification employed here is therefore as follows, with total revenues over GDP as thedependent variable: 13• GDP per capita – measured <strong>in</strong> constant US$2,000• Inflation rate – consumer prices, annual percentage rate12 For details on how these <strong>in</strong>dices are calculated see D. Kaufmann, A. Kraay and M. Mastruzzi (2009),Governance Matters V<strong>II</strong>I: Aggregate and Individual Governance Indicators, 1996-2008, World BankPolicy Research Work<strong>in</strong>g Paper 4978.13 GDP per capita, <strong>in</strong>flation, the trade share of GDP and non-agricultural share of GDP are all based ondata from the World Bank’s World Development Indicators (WDI) (WB, 2008). Revenue-ratio data alsocome from WDI except for the follow<strong>in</strong>g develop<strong>in</strong>g economies for which data come from the IMF’sWorld Economic Outlook dataset (IMF, 2009b): Angola, Botswana, Congo, Dem Rep, Ghana, Kenya,Lesotho, Madagascar, Malawi, Mauritius, <strong>Mozambique</strong>, Namibia, Senegal, Seychelles, South Africa,Swaziland, Tanzania, Ugandan , Zambia, Zimbabwe, Vietnam.


42 <strong>PARPA</strong> <strong>II</strong> R EVIEW—THE T AX S YSTEM IN M OZAMBIQUE• Trade share of GDP – def<strong>in</strong>ed as (exports of goods and services as a share ofGDP)+(imports of goods and services as a share of GDP)• Nonagricultural share of GDP – def<strong>in</strong>ed as 100-(Agricultural value-added share of GDP)• Government effectiveness – aggregated <strong>in</strong>dex with range of approximately -2.5 to 2.5.• Regulatory control - aggregated <strong>in</strong>dex with range of approximately -2.5 to 2.5.• Control of corruption - aggregated <strong>in</strong>dex with range of approximately -2.5 to 2.5.The regression equation is for estimated low and middle-<strong>in</strong>come countries over three periods. Theperiods chosen are 1999-2007, 2003-2006, and 2005-2007 given their relevance to the periodunder exam<strong>in</strong>ation <strong>in</strong> this report, data availability and <strong>in</strong> the <strong>in</strong>terests of verify<strong>in</strong>g robustness.As Bolnick (1978) 14 po<strong>in</strong>ts out, tax effort calculations are a better measure for <strong>in</strong>ternationalcomparisons than simple tax-ratios, but there are clear limitations, the pr<strong>in</strong>cipal one be<strong>in</strong>gsimultaneity. Factors considered to impact on tax capacity, such as demand factors, arethemselves to some degree dependent on tax policy and therefore related to tax-ratios: revenuesupply and demand are potentially determ<strong>in</strong>ed simultaneously. Estimation <strong>in</strong> l<strong>in</strong>ear form is afurther limitation given that capacity constra<strong>in</strong>ts are found to ease at an accelerat<strong>in</strong>g pace as<strong>in</strong>come grows. This is recognized <strong>in</strong> a number of studies and dealt with through the use of<strong>in</strong>strumental variables (e.g. Bird et al. (2007) and Davoodi and Grigorian (2006)).Explanatory variables employed <strong>in</strong> this study, such as those relat<strong>in</strong>g to <strong>in</strong>stitutions, may also besimultaneously determ<strong>in</strong>ed along with the tax-ratio, lead<strong>in</strong>g to endogeneity and bias <strong>in</strong> thecoefficient estimates. However, the pr<strong>in</strong>cipal objective of this analysis is to arrive at an estimateof <strong>Mozambique</strong>’s predicted tax capacity, rather than present po<strong>in</strong>t estimates of the impact of these<strong>in</strong>stitutional variables so that bias <strong>in</strong> the coefficient estimates is not considered an importantconstra<strong>in</strong>t.As is common practice <strong>in</strong> these analyses, to avoid the <strong>in</strong>terference of dynamic effects, averagesare taken of all time-vary<strong>in</strong>g variables. This also allows various periods to be used, allow<strong>in</strong>g forrobustness checks for the results obta<strong>in</strong>ed. Nonetheless, it is important to note that this analysis is<strong>in</strong>tended to provide further <strong>in</strong>dicative <strong>in</strong>formation regard<strong>in</strong>g revenue targets based on a simpleanalysis given the time available. In addition to the caveats above, it should therefore also benoted that us<strong>in</strong>g World Development Indicators data, there are a considerable number of countrieswith miss<strong>in</strong>g data mak<strong>in</strong>g the sample under analysis relatively small at only 43 countries.RESULTSThe results of the revenue-effort regressions are shown <strong>in</strong> Table F-2. The regression coefficients<strong>in</strong> the table <strong>in</strong>dicate that the most statistically robust determ<strong>in</strong>ant of revenue-ratios <strong>in</strong> thesesamples is the trade share of GDP, which is positively related to tax-ratios and highly statisticallysignificant across all three periods. This is <strong>in</strong>dicative of the importance of revenues related to14 B. Bolnick, "<strong>Tax</strong> Effort <strong>in</strong> Develop<strong>in</strong>g Countries: What Do Regression Measures Really Measure?" <strong>in</strong>John Toye (ed.), <strong>Tax</strong>ation and Development, London: Frank Cass, 1978.


A PPENDIX F 43trade, <strong>in</strong>clud<strong>in</strong>g duties, consumption taxes and VAT, all seen to be important for <strong>Mozambique</strong> <strong>in</strong>the discussion above. Of the <strong>in</strong>stitutional variables <strong>in</strong>cluded, control of corruption emerges asbe<strong>in</strong>g positively related to the revenue-ratio and statistically significant at the ten percent level atleast, with a large effect <strong>in</strong> <strong>in</strong>creas<strong>in</strong>g the tax-ratio. In contrast, regulatory control is negativelyassociated with revenue-ratios, by a similarly sized coefficient to that on corruption control. Thatwould suggest that the better government is perceived as implement<strong>in</strong>g sound policies and taxesto promote private sector development, the lower the revenue share. These variables are likely tobe coll<strong>in</strong>ear, although given the focus on the predicted values rather than estimated coefficients,this not of major concern.Table F-2<strong>Tax</strong> Effort Regression CoefficientsPeriod 1999-2007(1)2003-2006(2)2005-2007(3)GDPPCINFLTRADENONAGGOVCORRUPTREGULConstant0 0 0[0.002] [0.002] [0.002]0.071 0.224 0.406[0.063] [0.147] [0.308]0.115*** 0.119*** 0.125***[0.032] [0.032] [0.036]0.213 0.207 0.340*[0.160] [0.166] [0.183]0.11 0.095 0.059[0.170] [0.173] [0.144]5.784* 7.128** 8.025**[3.084] [3.436] [3.858]-5.551* -7.570** -8.653***[2.901] [2.943] [3.063]-5.888 -7.323 -17.98[10.817] [11.134] [12.608]Observations 43 43 43R-squared 0.62 0.6 0.59F-test 8.23 7.47 7.15Notes: Standard errors <strong>in</strong> brackets*significant at 10%** significant at 5%*** significant at 1%


Appendix G. <strong>Tax</strong> Incentives forInvestment—Pros and ConsThis appendix summarizes the controversy about offer<strong>in</strong>g tax <strong>in</strong>centives to attract or stimulateforeign and domestic <strong>in</strong>vestment <strong>in</strong> develop<strong>in</strong>g countries. 15 The case <strong>in</strong> favor of fiscal <strong>in</strong>centivesis usually advocated with great vigor by officials responsible for <strong>in</strong>vestment promotion andm<strong>in</strong>istries responsible for <strong>in</strong>dustrial development, often with support from their advisors andconsultants. Not surpris<strong>in</strong>gly, potential <strong>in</strong>vestors are also vocal proponents of tax <strong>in</strong>centives, andfrequently demand them as a condition for committ<strong>in</strong>g funds to a country with a weak bus<strong>in</strong>essenvironment (even if they would <strong>in</strong>vest anyway). The case aga<strong>in</strong>st special tax breaks is usuallyput forth by m<strong>in</strong>istries responsible for tax policy and public f<strong>in</strong>ance management, with supportfrom their advisors and consultants. Divergent views co-exist, too, with<strong>in</strong> the donor community.Many <strong>in</strong>ternational experts advise host governments on the importance of fiscal <strong>in</strong>centives whileothers advise governments to scale back or elim<strong>in</strong>ate them.THE CASE IN FAVOR OF TAX INCENTIVESThe case <strong>in</strong> favor of tax <strong>in</strong>centives rests on familiar and plausible arguments:Incentives enhance the return on <strong>in</strong>vestment. Investment decisions are driven by expectationsabout prospective risks and returns – specifically the after tax returns. Lower<strong>in</strong>g the tax burdenfor designated <strong>in</strong>vestors boosts the expected returns and shifts the balance <strong>in</strong> favor ofimplement<strong>in</strong>g <strong>in</strong>vestments.Incentives send a signal to <strong>in</strong>vestors. <strong>Tax</strong> <strong>in</strong>centives are a market<strong>in</strong>g device that can get <strong>in</strong>vestorsto look seriously at bus<strong>in</strong>ess opportunities <strong>in</strong> the host country. Gett<strong>in</strong>g them <strong>in</strong> the door is animportant first step <strong>in</strong> <strong>in</strong>vestment promotion.Incentives are essential due to tax competition. Foreign direct <strong>in</strong>vestment operates on a globalstage. Each develop<strong>in</strong>g country has to compete with other dest<strong>in</strong>ation countries for <strong>in</strong>vestmentsthat will stimulate growth and job creation, <strong>in</strong>clud<strong>in</strong>g countries that offer attractive <strong>in</strong>centives. 1615 The discussion here draws heavily on Bolnick (2009a).16 Klemm (2009) contends that tax competition is likely to be a major driv<strong>in</strong>g force.


46 <strong>PARPA</strong> <strong>II</strong> R EVIEW—THE T AX S YSTEM IN M OZAMBIQUEIncentives correct for externalities. Economic theory justifies <strong>in</strong>terventions that alter marketsignals <strong>in</strong> order to stimulate <strong>in</strong>vestments that generate significant positive externalities. Theseexternalities would otherwise lead to under-<strong>in</strong>vestment <strong>in</strong> activities such technology transfer andtra<strong>in</strong><strong>in</strong>g, where returns to the economy generally exceed the returns faced by private <strong>in</strong>vestorsthemselves. <strong>Tax</strong> <strong>in</strong>centives are a practical tool for <strong>in</strong>fluenc<strong>in</strong>g decisions <strong>in</strong> favor of such<strong>in</strong>vestments.Incentives are needed to compensate for other deficiencies <strong>in</strong> the <strong>in</strong>vestment climate. <strong>Tax</strong><strong>in</strong>centives help to offset the costs and risks faced by <strong>in</strong>vestors <strong>in</strong> low-<strong>in</strong>come countries, which arecharacterized by serious structural and <strong>in</strong>stitutional constra<strong>in</strong>ts such as poor <strong>in</strong>frastructure, a weaklegal/judicial system, and low labor productivity due to poor education, health, and nutrition.Incentives can work! <strong>Tax</strong> <strong>in</strong>centives have been an effective <strong>in</strong>strument for attract<strong>in</strong>g foreign<strong>in</strong>vestment to many countries. Malaysia is a widely cited prototype, where generous <strong>in</strong>centiveshelped trigger the transition from a poor resource-based economy to a rapidly <strong>in</strong>dustrializ<strong>in</strong>g“tiger.” Countries as diverse as Mauritius, Ireland and Ch<strong>in</strong>a followed similar strategies withsimilar results.THE CASE AGAINST TAX INCENTIVESSome arguments on the other side of the debate <strong>in</strong>volve technicalities that are less familiar tononspecialists, and therefore merit a bit more explanation for the sake of clarity.Incentives have limited effects. <strong>Tax</strong> <strong>in</strong>centives will not affect <strong>in</strong>vestments for which the expectedprofit rate, before tax, falls short of the <strong>in</strong>vestor’s risk-adjusted target rate of return– <strong>in</strong> otherwords, <strong>in</strong>vestments that are not fundamentally viable due to the underly<strong>in</strong>g risks and costs ofdo<strong>in</strong>g bus<strong>in</strong>ess <strong>in</strong> low-<strong>in</strong>come countries. Many possible <strong>in</strong>vestments will be immune on this countto the <strong>in</strong>fluence of tax <strong>in</strong>centives.Among projects that are fundamentally viable, given the local bus<strong>in</strong>ess environment, <strong>in</strong>centiveswill only affect the <strong>in</strong>vestment decision when tax considerations flip the expected return frombelow to above the target rate. For <strong>in</strong>vestments with marg<strong>in</strong>ally <strong>in</strong>adequate returns, <strong>in</strong>centives can<strong>in</strong>deed be a critical factor. Projects with solid fundamentals, however, would achieve the targetrate of return with or without <strong>in</strong>centives. In such cases <strong>in</strong>centives are “redundant,” <strong>in</strong> that they arenot a determ<strong>in</strong><strong>in</strong>g factor driv<strong>in</strong>g the <strong>in</strong>vestment decision. 17 To be sure, the <strong>in</strong>centives stillimprove the recipient’s cash flow and bottom l<strong>in</strong>e (at the expense of the Treasury). Hence,potential <strong>in</strong>vestors will always seek <strong>in</strong>centives and argue for their importance. This is no<strong>in</strong>dication, however, that the tax breaks are required.An important case where <strong>in</strong>centives can be decisive is <strong>in</strong> attract<strong>in</strong>g <strong>in</strong>ternationally mobile(“footloose”) export-oriented <strong>in</strong>vestments, which could just as well locate <strong>in</strong> another country. If17 <strong>Tax</strong> <strong>in</strong>centives are also redundant when granted to a U.S.-based company that repatriates profits; this isbecause the host-country tax break is offset by loss of a tax credit at home. <strong>Tax</strong> <strong>in</strong>centives may also be<strong>in</strong>effective if the procedures entail costs or delays that outweigh the benefits.


A PPENDIX G 47alternative venues offer dist<strong>in</strong>ctly more favorable conditions, then even attractive <strong>in</strong>centiveswould be <strong>in</strong>effectual. Also, offer<strong>in</strong>g generous tax holidays to lure this type of <strong>in</strong>vestment maylead only to short-term ga<strong>in</strong>s, as there is a tendency for beneficiaries to leave when the tax breaksend, or restructure to qualify aga<strong>in</strong> for <strong>in</strong>centives as a new entity.Incentives are costly. Some contend that the revenue effect of fiscal <strong>in</strong>centives may be nil or evenpositive because <strong>in</strong> the absence of <strong>in</strong>centives there would be no <strong>in</strong>come to tax from thebeneficiary projects. As a generalization, this is totally <strong>in</strong>valid. Revenue losses are likely, andlikely to be large. To the extent that <strong>in</strong>centives are redundant (which is difficult to measure), thereis a direct revenue loss to the full extent of the tax break. Even where <strong>in</strong>centives do stimulate new<strong>in</strong>vestment, there is still a revenue loss to the extent that a smaller benefit would have done thetrick (which is also hard to measure).Another revenue cost arises from the fact that preferences create loopholes. <strong>Tax</strong> lawyers andaccountants are <strong>in</strong> the bus<strong>in</strong>ess of structur<strong>in</strong>g corporate transactions to shift <strong>in</strong>come from fullytaxed affiliates to tax-favored affiliates, at potentially great expense to the Treasury. A thirdsource of revenue loss occurs <strong>in</strong>directly, when tax-favored <strong>in</strong>vestors pull bus<strong>in</strong>ess away fromfully taxed competitors.Incentives create economic distortions. <strong>Tax</strong>es distort economic decisions and reduce theefficiency of resource allocation. Fiscal <strong>in</strong>centives (when they work) add to the distortions bydraw<strong>in</strong>g resources <strong>in</strong>to tax-favored activities at the expense of others. For example, a generoustax benefit for agriculture may stimulate <strong>in</strong>vestment <strong>in</strong> an agriculture project with a 15 percentrate of return, <strong>in</strong>stead of an alternative <strong>in</strong>vestment with a 25 percent rate of return. Promot<strong>in</strong>g<strong>in</strong>vestment at the expense of productivity is not good formula for susta<strong>in</strong>ed growth. In addition,<strong>in</strong>centives l<strong>in</strong>ked to capital tend to favor capital <strong>in</strong>tensity over job creation. Insofar as fiscal<strong>in</strong>centives cause a revenue loss, additional efficiency costs arise from the need to either ma<strong>in</strong>ta<strong>in</strong>higher taxes on non-beneficiaries, or reduce government expenditures that could be used fordevelopment purposes.Where tax <strong>in</strong>centives are well designed to foster activities generat<strong>in</strong>g positive externalities, thenthe “distortions” can be efficiency enhanc<strong>in</strong>g. The problem is that <strong>in</strong>centives are often driven bypolitical considerations or special <strong>in</strong>terest pressures rather than careful economic analysis. Inaddition, where <strong>in</strong>centives are ad hoc and discretionary, they <strong>in</strong>vite rent-seek<strong>in</strong>g and corruption,which can derail even the best-<strong>in</strong>tentions programs (and add to the revenue risk).Incentives create costly precedents. When <strong>in</strong>centives are given to certa<strong>in</strong> <strong>in</strong>vestors, the precedentleads other groups to press for similar treatment. International experience suggests that it isdifficult to avoid this political “slippery slope,” which leads to a proliferation of <strong>in</strong>centives,shr<strong>in</strong>kage of the tax base, and an <strong>in</strong>tensification of other adverse effects.Incentives often don’t work! Just as proponents cite selective cases where <strong>in</strong>centives have clearlyworked, one can f<strong>in</strong>d many cases demonstrat<strong>in</strong>g the opposite. Thus, a study of the issue <strong>in</strong> 2004by McK<strong>in</strong>sey Global Institute concluded that: “Governments around the world woo foreign direct


48 <strong>PARPA</strong> <strong>II</strong> R EVIEW—THE T AX S YSTEM IN M OZAMBIQUE<strong>in</strong>vestment by offer<strong>in</strong>g costly tax breaks, import duty exemptions, land and power subsidies, andother enticements. Yet our evidence suggests that they are largely <strong>in</strong>effective.” 18 A NathanAssociates study of tax <strong>in</strong>centives <strong>in</strong> the SADC region similarly found that “generous tax<strong>in</strong>centives rarely stimulate a substantial <strong>in</strong>vestment response where the basic climate for do<strong>in</strong>gbus<strong>in</strong>ess is seriously deficient.” 19 This study also cited <strong>in</strong>stances <strong>in</strong> Uganda and Indonesia wheretax <strong>in</strong>centives were elim<strong>in</strong>ated <strong>in</strong> favor of broader tax reforms, with no adverse effect on<strong>in</strong>vestment trends.In general, the effectiveness of <strong>in</strong>centives varies by country and by type of <strong>in</strong>vestment. As notedabove, “footloose” export <strong>in</strong>dustries are highly responsive to tax competition among potentialhost countries offer<strong>in</strong>g otherwise adequate bus<strong>in</strong>ess conditions. In contrast, tax competition is farless important for <strong>in</strong>vestments anchored by geography, <strong>in</strong>clud<strong>in</strong>g activities target<strong>in</strong>g the domesticmarket or projects <strong>in</strong>volv<strong>in</strong>g access to natural resources. For resource-based <strong>in</strong>vestments, hostcountries should seek to maximize their share of value added, while still allow<strong>in</strong>g the <strong>in</strong>vestor toearn a reasonable rate of return.18 McK<strong>in</strong>sey Global Institute, New Horizons: Mult<strong>in</strong>ational Company Investment <strong>in</strong> Develop<strong>in</strong>gCountries, San Francisco, October 2003, p. 29.19 B. Bolnick, <strong>Tax</strong> Reform and the Bus<strong>in</strong>ess Environment <strong>in</strong> <strong>Mozambique</strong>, Nathan Associates, 2004. p.7-1.


Appendix H. Interview Results—Ma<strong>in</strong> Concerns and IssuesGOVERNMENT1. Appropriate target tax ratio: 20% target? Converge to regional average of 22-24%?Unrealistic revenue targets lead AT to punitive treatment of taxpayers.2. Trade-offs from lower tax rates: revenue versus growth? Offer scenarios, long-run vs. shortrunimpacts; impact on private sector and competitiveness.3. E-taxation, <strong>in</strong>clud<strong>in</strong>g e-declarations, payment through banks, s<strong>in</strong>gle w<strong>in</strong>dow for customs, andunified IT systems with<strong>in</strong> the AT.4. Contention that large projects are not contribut<strong>in</strong>g enough. What is the <strong>in</strong>ternational bestpractice for countries like <strong>Mozambique</strong>?5. Need simpler, more practical system, less complicated tax forms (especially M10 for IRPS).Adm<strong>in</strong>istrative complications from handl<strong>in</strong>g reconciliation of returns and refund claims. <strong>Tax</strong>adm<strong>in</strong>istration too arbitrary.6. Country needs more tra<strong>in</strong><strong>in</strong>g of accountants.DONORS1. Need for pro-growth tax system, not just pro-revenue. Short-run/long-run trade offs, need forcompetitive tax environment to stimulate <strong>in</strong>vestment. Government focus<strong>in</strong>g narrowly onrevenue targets (0.5% per year <strong>in</strong>crease <strong>in</strong> Rev/Y, up to 20% or 22%), to the detriment ofcustomer service. Heavy pressure to <strong>in</strong>crease receipts to reduce dependence on donors. Whatare the options?2. Coherence between revenue system and objectives to promote <strong>in</strong>vestment and SMEdevelopment. Trade off between <strong>in</strong>dustrial policy and revenue needs. How does tax systemaffect competitiveness?3. Major issue should be improv<strong>in</strong>g taxpayer service and enforcement. Both needed to expandrevenue base.4. <strong>Tax</strong> system favors large and small bus<strong>in</strong>esses, unfair to middle-size local enterprises. <strong>System</strong>of fiscal benefits skewed to large bus<strong>in</strong>ess, creat<strong>in</strong>g serious equity problems.5. Integration of AT so that customs and <strong>in</strong>come tax departments provide a s<strong>in</strong>gle face totaxpayers, with unified taxpayer database and systems for IT, audit, debt collection, riskassessment, dispute settlement, etc.6. New IT systems vital for improv<strong>in</strong>g tax adm<strong>in</strong>istration, but must be a catalyst forrestructur<strong>in</strong>g AT bus<strong>in</strong>ess processes (“leap forward”), not just automation of traditional


50 <strong>PARPA</strong> <strong>II</strong> R EVIEW—THE T AX S YSTEM IN M OZAMBIQUEsystems. But also serious risks <strong>in</strong>volved <strong>in</strong> IT modernization, if the process is not wellplanned, well managed and well funded, with adequate technical support.7. AT staff qualifications: tra<strong>in</strong><strong>in</strong>g needs; capacity build<strong>in</strong>g. Includ<strong>in</strong>g, especially, audit of largetaxpayers, customer service, especially <strong>in</strong> the prov<strong>in</strong>ces.8. Expansion of physical presence of AT to make it easier for small bus<strong>in</strong>esses to pay tax: new<strong>in</strong>frastructure; mobile brigades, or <strong>in</strong>termediary tax collectors (such as water utility).Decentralization of AT operations.9. Need simplification to make people’s lives easier, improve compliance.10. Important to develop risk-selection system to allocate AT staff efforts more productively: forverification, <strong>in</strong>spection, audit. Green channel, gold card – reward the good taxpayers.Exist<strong>in</strong>g systems rudimentary. “Enormous room for improvement” and revenue ga<strong>in</strong>s throughmodern risk-management systems.11. AT needs to develop <strong>in</strong>ternal MIS <strong>in</strong>dicators to monitor and assess operational efficiency.12. VAT refund process still far too complicated for compliance, especially by small bus<strong>in</strong>esses,mak<strong>in</strong>g it impossible for them to export.13. Donor coord<strong>in</strong>ation. The Common Fund is an important approach, but several donors will beprovid<strong>in</strong>g support on tax and customs issues outside that framework. Need one uniformapproach.14. Need to broaden tax base by br<strong>in</strong>g<strong>in</strong>g <strong>in</strong> the <strong>in</strong>formal sector via ISPC, to lower tax burden onthe formal sector. The big challenge is implementation.15. Trade facilitation. Green channel, post-clearance audit <strong>in</strong>tegrated with DGI audit functions.16. Need better MIS system for monitor<strong>in</strong>g operational <strong>in</strong>dicators, <strong>in</strong>clud<strong>in</strong>g prov<strong>in</strong>cial offices.PRIVATE SECTOR1. Complexity of the system. Moz does not have competency <strong>in</strong> the private sector or the AT toimplement the present laws effectively. Need simplification: fewer types of tax, <strong>in</strong>clud<strong>in</strong>gmultiple municipal taxes. Problem: simplification of VAT, <strong>in</strong>come tax difficult withoutopen<strong>in</strong>g the door to widespread abuse.2. E-payment and s<strong>in</strong>gle w<strong>in</strong>dow to reduce compliance costs for taxpayers, control evasion, andreduce opportunities for corruption. As much as 30-35% <strong>in</strong>crease <strong>in</strong> customs collectionsexpected from catch<strong>in</strong>g unrecorded imports via s<strong>in</strong>gle w<strong>in</strong>dow system.3. Tra<strong>in</strong><strong>in</strong>g and education – for both the ATM and the private sector4. Tra<strong>in</strong> more accountants! Need for better quality a/c education. Current programs turn outaccountants who are <strong>in</strong>capable.5. <strong>Tax</strong> rates too high for VAT and IRPC/IRPS, <strong>in</strong>hibit<strong>in</strong>g bus<strong>in</strong>ess development. Effective<strong>in</strong>come tax can be even higher due to disallowance of expenses on the basis of technicalerrors <strong>in</strong> the <strong>in</strong>voices or receipts. Need for exact completion of AT-approved receipts makesit difficult to develop l<strong>in</strong>kages to small bus<strong>in</strong>esses.6. ISPC – implementation is the key question. Will the simplified tax succeed <strong>in</strong> broaden<strong>in</strong>g thetax base by br<strong>in</strong>g<strong>in</strong>g <strong>in</strong>formal enterprises <strong>in</strong>to the tax net? Concern that the tax is still too highfor many micro and small enterprises that operate on low marg<strong>in</strong>s; anyone without adequatesales records faces full 75.000 MT.7. Rampant and blatant tax evasion. Need to catch tax evaders with assessments based on signsof wealth and third party data such as automobile registrations. Equally, need to change the


A PPENDIX H 51culture of pay<strong>in</strong>g tax. Noth<strong>in</strong>g will work well without a change <strong>in</strong> attitudes – pride <strong>in</strong> pay<strong>in</strong>gtaxes. <strong>Tax</strong>payers must see the advantage.8. Need modern system of risk management, <strong>in</strong>clud<strong>in</strong>g green channel for taxpayers with provenrecord of compliance (subject to audit).9. Customs scanners: 100% scann<strong>in</strong>g at $100 per conta<strong>in</strong>er (plus 17% VAT). Private sector notaga<strong>in</strong>st the concept, but it should be applied <strong>in</strong> l<strong>in</strong>e with <strong>in</strong>ternational standards, imply<strong>in</strong>gselective approach, low fee.10. VAT refunds much improved but still far too complicated. Payments still centralized <strong>in</strong>Maputo, and require paper documentation rather than electronic fil<strong>in</strong>g. Local officials whonow do <strong>in</strong>itial verification not well qualified. Refunds for IRPC and IRPS also function<strong>in</strong>gpoorly.11. 20% withhold<strong>in</strong>g punitive for many types of transactions, <strong>in</strong>clud<strong>in</strong>g payments to smallfarmers! Withhold<strong>in</strong>g tax on payments to foreign contractors simply causes them to <strong>in</strong>crease<strong>in</strong>voice price by 25% to break even net of tax.12. Both importers and exporters fail<strong>in</strong>g to benefit from SADC free trade area due to complexity<strong>in</strong> comply<strong>in</strong>g with documentation requirements for rules of orig<strong>in</strong>. In <strong>Mozambique</strong>, only twoofficials – <strong>in</strong> Maputo – are authorized to sign ROO documents.13. Multiplicity of taxes, tak<strong>in</strong>g <strong>in</strong>to account the many municipal levies. Has anyone added upthe tax cost and compliance cost?SMALL BUSINESSES1. <strong>Tax</strong> system too complex; it requires a tremendous amount of time to comply with alldocumentation, and reconcile at year-end tax paid with tax due; requirements not simpleenough for an ord<strong>in</strong>ary person to understand; further simplification needed to encouragepeople to comply with tax laws.2. The bureaucracy deal<strong>in</strong>g with tax documentation gives rise to corruption and bribery.3. There is skepticism over the success of the ISPC.4. More tra<strong>in</strong><strong>in</strong>g and education needed on tax issues; public education on tax obligations <strong>in</strong>steadof heavy tax penalties.5. The culture of comply<strong>in</strong>g with tax laws is weak. People <strong>in</strong> general don’t see the real benefitsaris<strong>in</strong>g from pay<strong>in</strong>g taxes promptly.6. The VAT reimbursement system is cumbersome and subject to long delays that affect thecash flow of SMEs; VAT reimbursement is centralized <strong>in</strong> Maputo, which adds to the delays;at times documentation returned without proper explanation.7. Many tax officials and accountants work<strong>in</strong>g <strong>in</strong> private sector are not well tra<strong>in</strong>ed to deal withcomplicated tax issues.8. Very few small companies can afford to hire professional tax advisory or account<strong>in</strong>g firms soas to comply with the tax laws; there are very few accountants <strong>in</strong> remote areas or prov<strong>in</strong>cessuch as Niassa; this situation <strong>in</strong>creases the cost of do<strong>in</strong>g bus<strong>in</strong>ess s<strong>in</strong>ce small enterprise areheavily penalized for not present<strong>in</strong>g accounts signed by designated professional accountants.9. The tax rates are relatively high compared to other countries <strong>in</strong> the region; <strong>Mozambique</strong>should learn from South Africa’s system <strong>in</strong> terms of provid<strong>in</strong>g tax relief for emergent SMEs.10. There is a tremendous tax evasion and a lot of bribery; <strong>in</strong>fluential people have ways to alwaysavoid the taxes; some tax official facilitate people to evade taxes.


52 <strong>PARPA</strong> <strong>II</strong> R EVIEW—THE T AX S YSTEM IN M OZAMBIQUE11. <strong>Tax</strong> payments should be done through private banks or onl<strong>in</strong>e payment services to reduce theamount time needed to wait <strong>in</strong> queues to deal with tax documentations.


Appendix I. Persons InterviewedIN MOZAMBIQUEGovernmentAntonio Cruz MPD/DNEAP DirectorBrendan Kelly MPD/DNEAP AdvisorFausto Mafambissa MPD/DNEAP Head of CFMM unitAugusto Sumburana MOF/GEST DirectorHerm<strong>in</strong>io Sueia AT/DNEP National DirectorAli Algy AT/DNEP Department HeadTapu Mamane AT/DSPE DirectorArl<strong>in</strong>do Jose Antonio da Graca AT/GPECIRevenue AdvisorMaria Otilia Santos AT/DGI General DirectorIlidio Rafael Guibalo AT/DGI Deputy General DirectorMoises Patricio Marrime AT/DGI General Coord<strong>in</strong>atorDanielo Nala MPD/GANEZA General DirectorJulieta Dom<strong>in</strong>gas Much<strong>in</strong>e MIC/GASP DirectorHoracio Dombo CPI Chief, Special ProjectsDonors and Donor ProgramsNelson Guilaze USAID Senior Policy AnalystEric Johnson USAID Agriculture and Bus<strong>in</strong>ess AdviserEmmy Bosten IMF Technical Assistance Coord<strong>in</strong>atorAntonio Nucifora World Bank Senior EconomistTelma LoForte SDC Senior EconomistRalf Orlik KfW DirectorAndrew Clark DFID Economic AdvisorCarlos Mate Norwegian Embassy Program OfficerRosario Marapusse Italian Cooperation EconomistAlberto Musatti Italian Cooperation EconomistBus<strong>in</strong>ess RepresentativesKekobad Patel CTA Executive DirectorCarie Davies ACIS Executive DirectorGraeme White ACIS President, Management CommitteeJoao Mart<strong>in</strong>s PWC Partner, <strong>Tax</strong> and Legal ServicesPaula Ferreira Deloitte Manag<strong>in</strong>g PartnerMaria Basto Deloitte Senior Manager, Consult<strong>in</strong>gAvelar da Silva Intertek General ManagerFilipe Franco AFIM Executive DirectorMatola CargoManag<strong>in</strong>g Director


54 <strong>PARPA</strong> <strong>II</strong> R EVIEW—THE T AX S YSTEM IN M OZAMBIQUESevi George AFIM <strong>Mozambique</strong> Director:Hold<strong>in</strong>gsGerry MarketosCIMPOGESTKenneth GunnCIMPOGESTNolifer Lakhani SGL Executive DirectorZenalda Carlota Mats<strong>in</strong>he SGL Director, Market<strong>in</strong>gNatividade Bule Ass. dos Empreendedores PresidentSimeão Sevene Machava Ass. dos Jovens PresidentAgricultoresSudecar Novela Ass. dos Mukherista PresidentAlb<strong>in</strong>o MacieFederação Moçambicanados Empreiteiros Executive DirectorMário FerroAss. Moçambicana de Empresas de Market<strong>in</strong>gPublicidade e Relações PúblicasIN WASHINGTON D.C.Andrea Lemgruber IMF Fiscal Affairs DepartmentJose Sulemane IMF Advisor to Exec Director for Moz.Larry Westfall US Treasury Team Leader, UST supportto the ATSebastian James FIAS/IFC EconomistJim LaFleur CTA Economic Advisor

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!