Effectiveness and Economic Impact of Tax Incentives in the SADC

Loading...

TECHNICAL REPORT

Effectiveness and Economic Impact of Tax Incentives in the SADC Region

SUBMITTED TO

USAID/RCSA SADC Tax Subcommittee, SADC Trade, Industry, Finance and Investment Directorate SUBMITTED BY

Nathan−MSI Group UNDER CONTRACT NO.

PCE-I-00-98-00016-00 February 2004

TECHNICAL REPORT

Effectiveness and Economic Impact of Tax Incentives in the SADC Region

SUBMITTED TO

USAID/RCSA SADC Tax Subcommittee, SADC Trade, Industry, Finance and Investment Directorate SUBMITTED BY

Nathan−MSI Group PREPARED BY

Dr. Bruce Bolnick Nathan Associates Inc., Arlington, Virginia UNDER CONTRACT NO.

PCE-I-00-98-00016-00 February 2004

Contents Executive Summary 1. Introduction

xi 1-1

1.1 Starting Point—Principles of Taxation for Developing Countries

1-2

1.2 What are Tax Incentives?

1-4

1.3 “Effectiveness” and “Impact” of Tax Incentives

1-5

1.4 Study Organization

1-6

2. Taxation, Investment, and Growth

2-1

2.1 Capital Investment and Economic Growth

2-1

2.2 Theory of the Determinants of Investment

2-3

2.3 Empirical Evidence for Taxation as a Determinant of Investment

2-9

2.4 Lesson for SADC 3. Economics of Tax Incentives: Pros and Cons

2-13 3-1

3.1 Ten Arguments in Favor of Investment Tax Incentives

3-1

3.2 Ten Arguments against Investment Tax Incentives

3-5

3.3 Summary 4. Economic Toolkit for Analyzing Tax Incentives 4.1 Marginal Effective Tax Rate

3-18 4-1 4-1

4.2 Budget Impact—Monitoring “Tax Expenditures”

4-10

4.3 Screening Criteria

4-13

5. Design of Investment Tax Incentives 5.1 Design Considerations

5-1 5-1

II

Contents (continued) 5.2 Assessment of Alternative Tax Instruments 5.3 Summary 6. Economics of Harmful Tax Competition

5-4 5-12 6-1

6.1 What is Harmful Tax Competition?

6-1

6.2 Theory of Tax Competition

6-4

6.3 Can Jurisdictions Cooperate in Mitigating Harmful Tax Competition?

6-8

6.4 Implications for SADC 7. Tax Systems and Tax Incentives in the SADC Region

6-10 7-1

7.1 Revenue Performance

7-1

7.2 Standard Tax Rates

7-2

7.3 Investment Tax Incentives

7-7

7.4 Successes and Failures

7-17

7.5 Institutional Framework for Tax Incentives

7-19

7.6 Summary

7-22

8. Conclusions and Recommendations

8-1

8.1 General Principles for Agreement

8-2

8.2 Structure of Investment Tax Incentives

8-4

8.3 Transparency

8-6

8.4 Capacity Building

8-7

8.5 Harmful Tax Competition

8-8

References Appendix. Investment Tax Incentives in the SADC Region

III

Contents (continued) ILLUSTRATIONS Figures Figure 3-1. Tax Incentives and the User Cost of Capital Figure 7-1. Standard Company and Dividend Tax Rates, SADC Member Countries, 2003 Figure 7-1. Illustrative METR with Standard Tax System and with Tax Incentives

3-11 7-4 7-17

Tables Table 4-1. International Comparisons of the Marginal Effective Tax Rate, circa 1985

4-6

Table 4-2. Relative Cost-effectiveness of Various Tax Incentives

4-7

Table 4-3. Conceptual Basis for Screening Selective Tax Incentives

4-13

Table 5-1. Evaluation of Major Tax Incentive Instruments

5-14

Table 6-1. Prisoner’s Dilemma and Tax Incentives

6-5

Table 7-1. SADC Fiscal Indicators, Three- year Averages, 1999-2001(as % of GDP)

7-3

Table 7-2. Illustrative Marginal Effective Tax Rate Calculations, SADC Member Countries, Standard Tax Systems (non-exporters, without special incentives) Table 7-3. Investment Tax Incentives in the SADC Region (2003)—Who Does What?

7-6 7-8

Table 7-4. Illustrative Marginal Effective Tax Rate Calculations, SADC Member Countries: Case—Tax Incentives for Projects Manufacturing Goods for Export Table 7-5. Elements of Tax Structure Facing Companies in SADC Member Countries

7-16 7-23

Table 7-6. Parameters used for METR Calculation in SADC Countries, with Tax Incentives

7-28

Exhibits Exhibit 2-1. A Productivity Package

2-4

Exhibit 2-2. Policies Affecting Investment Risk and Return

2-5

Exhibit 3-1. Success Stories in Malaysia, Ireland, Costa Rica, and Mauritius

3-3

Exhibit 3-2. Small Companies, Giant Loopholes

3-9

Exhibit 3-3. Import Duties, Effective Rate of Protection, and Economic Efficiency

3-12

Exhibit 3-4. Impact of Tax Differentials on the Allocation of Capital

3-13

Exhibit 3-5. Elimination of Tax Incentives in Uganda, 1997

3-19

Exhibit 3-6. Elimination of Tax Incentives for Indonesia, 1984

3-20

Exhibit 4-1. How Different Tax Incentives Affect the Marginal Effective Tax Rate

4-5

IV

Contents (continued) Exhibit 4-2. METR Model of Dunn and Pellechio Exhibit 4-3. Foregone Revenue from an Investment Tax Incentive

4-8 4-12

Exhibit 5-1. A Cautionary Tale from India

5-7

Exhibit 6-1. Resisting Tax Competition from Next Door, Zambia, 1963

6-7

Exhibit 6-2. Shopping for Tax Incentives: The Case of Ramatex

6-9

Exhibit 7-1. Strategic Industrial Projects Program in South Africa

7-9

Glossary ACT

additional company tax

AETR

average effective tax rate

AGOA

African Growth and Opportunity Act

CFE

controlled foreign entities

CTM

comparative tax model

DAO

Development Approval Order

DBM

Declining Balance Method

DITE

Division on Investment, Technology and Enterprise Development (UNCTAD)

DRC

Democratic Republic of the Congo

EPZ

export processing zone

ERP

effective rate of protection

ERR

economic rate of return

EU

European Union

FDI

foreign direct investment

FRR

financial rate of return

FTZ

free trade zone

GCI

Global Competitiveness Index

GCR

Global Competitiveness Report

GDP

Gross Domestic Product

GNP

Gross National Product

HTC

harmful tax competition

HTS

high-tax state

ICA

initial capital allowance

ICT

information and communications technology

IMF

International Monetary Fund

ITC

investment tax credit

LTS

low-tax state

METR

marginal effective tax rate

MICI

Microeconomic Competitiveness Index

MIDP

Motor Industry Development Program

MOU

Memorandum of Understanding

VI

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

OECD

Organization for Economic Cooperation and Development

R&D

research and development

RCE

relative cost-effectiveness

ROR

rate of return

SADC

Southern African Development Community

SCM

Subsidies and Countervailing Measures (WTO Agreement)

SIP

Strategic Industrial Projects

SL

straight line (depreciation)

SOW

Statement of Work

TFP

total factor productivity

TIFI

Directorate on Trade, Industry, Finance and Investment (SADC)

UCC

user cost of capital

UNCTAD

United Nations Conference on Trade and Development

USAID

United States Agency for International Development

VAT

value added tax

WTO

World Trade Organization

Preface The Scope of Work (SOW) for this study points out that Many countries around the world, including those in the SADC region, use tax incentives to attract investment and promote economic activity. The goal is to increase investment and expand the range of economic activities. These are seen as being critical to these country’s efforts to sustain rapid economic growth and development. Given the scarcity of resources in SADC, and the region as a whole, it is crucial that governments adopt a coherent approach to tax incentives and that the tax incentives are effective.

In this context, the SADC Tax Subcommittee commissioned a study of the effectiveness and impact of tax incentive programs in the region. More specifically, the SOW states that the objective of the study is to provide evidence regarding the extent of coherence of tax incentives across SADC Member States. This evidence will help guide the efforts of Member States in forming a consensus on harmonising tax incentive structures and avoiding harmful tax competition. It will also assist Member States review the tax laws and regulations and suggest the merits of centralizing them with a single ministry (such as the Ministry of Finance). The overall objective is to improve tax administration, enhance revenue collection procedures, and reduce investor confusion regarding tax incentives.

This study approaches the subject from an economic perspective. Our premise is that the impact and effectiveness of any investment tax incentive depends on the context. Also, any evaluation of tax policy involves judgments about tradeoffs and uncertainties. Thus, we do not attempt to offer simple universal rules about how SADC Member States should proceed. Instead, we recognize that every country needs to establish strong capacity for undertaking its own tax policy analysis, including careful analysis of actual or proposed tax incentive programs. In this spirit, our study provides a non-technical primer on the economics of tax incentives for developing countries as a foundation for reviewing the incentive programs in the SADC region. Tax officials and managers, policymakers, and stakeholders who formulate or evaluate tax incentive programs will find the study beneficial, as will students interested in the economics of tax policy as a tool for investment promotion and economic development.

VIII

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Several limitations bear mention. First, in accordance with the SOW, the study focuses on tax incentives operating through the company income tax (though it touches on interactions between the company and personal tax systems, as well as indirect tax incentives). Second, because the study covers the entire SADC region it is not possible to probe the intricacies of the tax systems in every country. Instead, we address main points of the various systems. The primary resource for this purpose is a preliminary version of the SADC Tax Database (August 2003). Supplementary data were obtained from public documents and websites of member governments, international organizations, and private accounting firms, as well as consulting studies and academic papers. Because tax laws change frequently, some information presented here may be out of date. We welcome corrections. Third, without access to microeconomic data on investment projects or taxpayer records, we built our analysis on principles of tax policy, lessons from international experience, findings of the professional literature, simulations using illustrative cases, and macroeconomic data.1 The author, Dr. Bruce Bolnick of Nathan Associates Inc., would like to express special thanks to Martin Grote and David Hollinrake at the National Treasury in South Africa for their valuable information and comments; to Moeketsi Senaoana, Senior Macroeconomic Policy Advisor at SADC/TIFI for support and information; to Sebastian James, at the Kennedy School of Government at Harvard University, for his excellent research assistance; and to Anthony Pellechio at the IMF for providing an updated version of his Marginal Effective Tax Rate model (with David Dunn) for use in the study and permission to distribute copies to the Tax Subcommittee. Dr. Bolnick also benefited from background studies prepared by Thabo Legwaila at Stellenbosch University, Arbind Modi at the Ministry of Finance in India, and Danies Chisenda at the Ministry of Planning and Finance in Zambia, as well as information provided by Domingos Julio Inacio at the Ministry of Finance in Angola, Duraiswamy Ravindran at the Ministry of Finance and Development Planning in Botswana, Happyson Nkya at the Tanzania Revenue Authority, and Rory Allan at UNCTAD’s Division on Investment, Technology and Enterprise Development. In addition, he benefited from discussions with Louis Wells at the Harvard Business School, Andrea Viol at the Kennedy School of Government, Graham Glenday at the Duke Center for International Development, and Frank Flatters from Queens University. Dr. Bolnick wishes to thank those who kindly provided both time and information during his field visit in March, 2003, including (in Gabarone) Fudzai Pamacheche and Rosalind Thomas at the SADC Secretariat; Stanley Mupanomunda at USAID/RCSA; Ms. Iponeng Sennanyana at the Ministry of Finance and Development Planning; Lameck Nthekela at the Botswana

1 The original SOW called for microeconomic data analysis, but when this proved not to be possible SADC and

USAID approved the revised methodology when the study commenced.

PREFACE

IX

Export Development & Investment Authority; Marcel Belanger at the Ministry of Trade and Industry; Olivia Muzvidziwa at KPMG; Diniar Minwalla at PriceWaterhouseCoopers; and (in Pretoria) Cecil Morden at the National Treasury.

Executive Summary The use of tax incentives in developing countries has been very popular and very controversial for decades. Although such incentives undoubtedly affect investment decisions in some circumstances, it is not at all clear that the overall benefits outweigh the costs. Despite the controversy, every SADC country offers investment tax incentives of one form or another. Many governments in the region face pressure to sweeten these programs, to compete with tax breaks offered elsewhere. At the same time they are pressed to strengthen revenues, to finance essential public goods and services. The challenge is to understand the conditions and the policy design features that determine whether tax incentives in the SADC region are likely to deliver substantial and sustainable net benefits in a particular context. The purpose of this study is to help tax officials, policymakers, and other stakeholders in the region meet the challenge. The fundamental purpose of taxation is to raise revenue effectively, through measures that suit each country’s circumstances and administrative capacity. In fulfilling the revenue function, a well designed tax system should be efficient in minimizing the distortionary impact on resource allocation, and equitable in its impact on different groups in society. According to the SADC Memorandum of Understanding on taxation (2002) tax incentives are “fiscal measures that are used to attract local or foreign investment capital to certain economic activities or particular areas in a country.” This definition excludes “general tax incentives” that apply to all investments. The present study encompasses both general and selective tax incentives, while focusing on the latter. Here, “effectiveness” means the extent to which tax incentives stimulate additional productive investment, whereas “impact” refers to the broader effects on revenue, tax administration, economic efficiency, social equity, and, ultimately, prospects for economic growth.

Taxation, Investment, and Growth Two fundamental premises underpin the case for tax incentive programs in developing countries: first, that additional investment is needed to foster more rapid economic growth; and second, that tax breaks can be effective in stimulating investment. Both propositions may

XII

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

seem self-evident, yet they are subject to important qualifications that are highly pertinent to understanding the effectiveness and impact of investment tax incentives in the SADC region. On the first proposition, the key issue is that investment productivity is at least as important as the quantity of investment in determining growth. Even if tax incentives do stimulate investment, their net impact on growth could be adverse if the incentives reduce productivity. Regarding the second proposition, taxation undoubtedly affects some investments— particularly “footloose” projects that are viable in several competing locations. But non-tax factors are far more important in determining most investment decisions. Moreover, if tax breaks cause fiscal problems that worsen other elements of the investment climate, the net effect of incentives on the volume of investment can be negative rather than positive.

Advantages and Disadvantages of Tax Incentives Arguments in favor of investment tax incentives are widely known. According to proponents, tax incentives clearly enhance returns on investment; they may be justified by positive externalities stemming from investments; they are relatively easy to target and fine tune; they signal openness to private investment; they are useful in a world of capital mobility; they are necessary for responding to tax competition from other jurisdictions; and they compensate for other deficiencies in the investment climate. Another common argument is that incentives can actually enhance revenue by stimulating investments that generate other taxable income via employment and linkage effects. Tax incentives also offer political advantages over direct expenditure programs to stimulate investment. Finally, proponents point out that tax incentives have been successfully used in well known cases like Malaysia, Ireland, and Mauritius. The case against tax incentives is less widely understood, but essential to any careful policy analysis. Those who advise against tax incentives argue that: 1.

The actual revenue cost can be high if the investments would have been viable anyway; the incentives are offset by source-country tax laws; or tax-favored investors take business away from taxable producers.

2.

Abusive tax avoidance schemes, made possible by tax preferences, further erode the revenue base.

3.

Tax incentives also divert administrative resources from revenue collection.

4.

Such revenue losses require painful fiscal adjustments in the form of higher taxes on other entities, cuts in expenditure, or greater dependence on other costly forms of financing.

5.

Tax differentials can introduce serious economic distortions that reduce efficiency and productivity.

EXECUTIVE SUMMARY

6.

XIII

Tax preferences create inequities by favoring some taxpayers over others. This can undermine general compliance.

7.

As a development tool, tax incentives score poorly in terms of transparency and accountability.

8.

The cash value of tax incentives stimulates political manipulation and corrupt practices.

9.

Alternative instruments for promoting investment can have much more favorable and lasting effects on productivity, growth, and development.

10. International experience shows that tax incentives most often do not deliver favorable results! In short, there are arguments both for and against the use of investment tax incentives that must be taken seriously. The applicable balance will vary according to circumstances. The danger is that in the political discourse the benefits are usually exaggerated, while the costs are downplayed or ignored. This creates a strong bias toward the implementation of poor tax incentive policies.

Tools for Analyzing Tax Incentives Three important tools for analyzing tax incentive policies are the marginal effective tax rate (METR) model, tax expenditure budgeting, and the specification of screening criteria for applying selective incentives. The METR model provides a gauge for evaluating the extent to which various tax incentive packages improve the rate of return for representative investment projects, at the margin. Tax expenditure budgeting is a valuable method for monitoring the amount of foregone revenue from tax incentives. Every member country of SADC should be taking steps to adopt these two tools for policy analysis. In screening projects that will benefit from selective incentives, the goal should be avoid forgoing tax revenue for investments that would be undertaken anyway. In general, projects that are efficient and sustainable are likely to materialize even without special tax breaks. The exception is “footloose” investment that can easily locate in other countries. Incentives can also be effective in stimulating investments that are not viable without a tax break. However, these are projects with low productivity, by definition. The criteria used to target investment incentives often fail to pick projects that will deliver large benefits relative to the revenue cost. Furthermore, any selective screening process can be subverted by political maneuvering.

Design of Tax Incentives Common incentives include low overall tax rates, preferential tax rates for certain investments, tax holidays, capital recovery allowances, investment tax credits, the treatment of dividends, excess deductions for designated expenses, special export incentives, reduced

XIV

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

import duties on capital and raw materials, and protective tariffs. The advantages and disadvantages of the various incentives can be analyzed in terms of four criteria: (1) effectiveness in stimulating investment, (2) impact on revenue, (3) economic efficiency, and (4) impact on tax administration. The effectiveness and impact of any tool will depend on local circumstances, the characteristics of each investment, and other details of the tax code. Because each tool has virtues and drawbacks, ranking them requires policy judgments as well as technical analysis. Nevertheless, one may draw some conclusions about the relative merits of the various tax incentive instruments. For example, even a zero-tax regime will not attract much investment if there are serious deficiencies in the investment climate that render projects fundamentally unviable. Hence, concern over the design of tax incentives must not divert attention from other policies and programs that are needed to improve returns and reduce risks for investors.

Responding to Harmful Tax Competition Since many investors these days can select among alternative tax jurisdictions, tax competition is inevitable. The issue is where to draw the line between legitimate political judgments about offering an attractive tax policy, versus practices that cause serious crossborder fiscal and economic distortions. Competition between jurisdictions can impel tax reform and efficiency in government expenditure. But tax competition can also reduce fiscal resources for the region as a whole and lead to less efficient allocation of investment. Cooperation is called for when competition impairs regional development through a selfdefeating “race to the bottom” or poor policy decisions based on imperfect information. Since most tax incentive programs, worldwide, have not been very successful, it is very likely that cooperation to mitigate harmful tax competition can benefit all SADC Member States.

Tax Systems and Incentives in the SADC Region Every SADC country offers special investment tax incentives in one form or another. The most streamlined program is in Lesotho, while Mauritius has the most extensive and complicated set of programs. The most widely used instruments in the region—appearing in at least 10 Member States—are initial capital allowances, full or partial tax holidays, preferential tax rates, and special export incentives.2 The least common instrument is the investment tax credit (ITC). This is odd because the ITC is regarded by many experts as the most cost-effective, transparent, and simple form of incentive. Only in Mozambique, since 2002, has the ITC been a centerpiece of the tax incentive program. One notable investment tax incentive program in the region is South Africa’s Strategic Industrial Projects, which was introduced in 2001 to encourage selected industrial investments. The SIP has attractive features such as coherent links between policy goals and

2 Special, in the sense of going beyond the international norm of relieving duty and tax on inputs used in the

production of exported goods.

EXECUTIVE SUMMARY

XV

screening criteria; benefits in the form of an initial capital allowance that is valuable to investors while still generating revenue; explicit limits on the budgetary cost; transparent procedures; and “clawback” provisions penalizing beneficiaries who do not deliver results. The institutional framework for tax incentive programs (i.e., policy objectives, procedures, and systems for enhancing transparency) is weak in most of the region. In addition, most Member States face critical fiscal constraints, implying that they should be very cautious about the revenue risks associated with generous incentives. The standard company tax rates in the region range from 25 to 40 percent, clustering around 30 to 35 percent. The combined burden of company tax plus dividend tax is lowest in Mauritius and Botswana and highest in Zambia and DRC. To assess the standard tax system in each SADC country, we evaluate the METR taking into account the tax on company income, dividends and capital gains, as well as depreciation allowances, and loss carry-forward provisions. Under four illustrative project scenarios, the standard tax system in Mauritius offers the lowest METR to investors. Namibia and Botswana also offer standard tax codes with METRs under 35 percent. In contrast, the METR is above 50 percent for the standard tax system in Angola, DRC, Mozambique, Swaziland, and Zimbabwe. We next compare the tax incentive regimes by calculating the METR for the case of a foreignowned investment manufacturing for export. Tax concessions produce a large reduction in the METR in most states, except Angola and Botswana where the differential is relatively small. A large reduction in the METR may have a significant positive impact on the economy if the incentives are well targeted, effectively administered, and complemented by other favorable investment policies. Otherwise, they may simply lead to a misallocation of resources and a large revenue loss.

Conclusions and Recommendations On the basis of our review of how investment tax incentives work in theory and in practice, as well as their economic advantages and disadvantages, we conclude that • Non-tax elements of the investment climate are far more important than tax incentives in determining the level and quality of investment flows; • The effect of incentives on productivity and efficiency is at least as important as the effect on the amount of investment; • Investment tax incentives may work well in some contexts but they work poorly in many others; decisions about tax incentives must be country specific; • The benefits of investment tax incentives are widely exaggerated, while the costs are often underestimated or overlooked altogether; and • Capacity building to strengthen tax policy analysis should be a central priority.

XVI

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Given these conclusions and our review of tax incentive programs and general fiscal health in the SADC region we offer 12 recommendations pertaining to the structure of incentives, transparency, capacity building, and harmful tax competition: 1.

Definition of “tax incentives.” SADC should modify its definition of “tax incentives” to encompass both general and selective incentives. The definition should continue to cover both direct and indirect tax measures.

2.

Policy coherence. SADC members should agree to review the coherence of their tax incentive programs to eliminate inconsistency between goals, criteria, and instruments. For example, the goal of promoting growth may be served by tax measures that are designed to stimulate projects with substantial positive externalities, but not by incentives that support inefficient and uncompetitive activities. Other inconsistencies arise between tax policies and other policies that affect the investment climate, including macroeconomic policies, structural reforms, and institutional reforms.

3.

Tax policy analysis. SADC member states should agree to develop the analytical capacity, organizational arrangements, and institutional procedures needed to conduct a professional review of existing and proposed tax incentive measures and other tax policies. The purpose is to ensure that policy decisions are based on sound information about their likely impact.

4.

Revenue management and the role of the finance ministry. Tax incentive programs must be compatible with prudent revenue management. Countries facing severe revenue constraints should be especially cautious about offering incentives that create substantial revenue risks directly or indirectly. To ensure proper consideration of revenue objectives, the minister responsible for finance must be centrally involved in the formulation of tax incentive policies.

5.

Choice of tax incentive instruments. The advantages and disadvantages of each tax instrument must be judged in light of local conditions, subject to systematic policy analysis to support well informed decisions. Nevertheless, the analysis suggests broad conclusions about the choice among tax incentive instruments.  Overall, the soundest tax incentive is to establish standard tax rates that are fair and moderate.  Relieving exporters of indirect tax on inputs should be a top priority, with due regard to the need for effective procedures to prevent abuse of the remissions.  The most cost-effective tax incentives are the ITC and the initial capital allowance (ICA). These tools yield a large reduction in the METR relative to the revenue cost, with a minimum of administrative complexity. Yet they also create moderate distortions favoring capital-intensity and short-lived assets, and can serve as avenues for tax evasion if not well administered.

EXECUTIVE SUMMARY

XVII

 Tax holidays entailing a full exemption are less likely to be cost-effective. The revenue loss tends to be large relative to the improvement in incentives. Tax holidays strongly favor transitory rather than sustainable investments, and create glaring opportunities for aggressive tax avoidance.  The worst form of tax incentive is the imposition of high protective tariffs on competing imports. This may stimulate investment for the domestic market, but it usually turns out to have low productivity and poor development potential.  In states with serious revenue constraints, selective tax breaks (if used at all) should narrowly focus on activities that are likely to deliver an especially high payoff in terms of policy goals, and which would not be undertaken without special incentives.  Avoid zero tax rates. The vast majority of viable investment projects do not hinge on getting a total exemption from tax.  Eliminating import duties on raw materials and intermediate goods is a poor way to stimulate investment. The measure has a high revenue cost and little effect on investment for the domestic market since indirect taxes are usually passed along to consumers. (As noted above, however, mechanisms are needed to eliminate the burden of duties on inputs used to produce for export.)  Location-dependent investments that are fundamentally viable, especially resourcebased projects that cannot readily move elsewhere, should not receive special tax preferences.  A low-rate alternative minimum tax can ensure that every company contributes at least minimally to the cost of basic public services. 6.

Tax administration. In designing tax incentive programs, SADC members should place a high priority on the administrative implications. Countries with weak tax administration should shun high-value tax breaks that invite aggressive tax planning and abusive tax avoidance.

7.

Automatic versus discretionary incentives. In a country with highly disciplined administrative systems, discretionary programs can have advantages in targeting incentives to enhance their effectiveness and impact. In other circumstances, discretion can create severe problems by increasing administrative costs, politicizing technical decisions, and creating avenues for corrupt practices.

8.

Fiscal transparency. SADC member states should embrace the need for fiscal transparency, using the IMF Code of Good Practices on Fiscal Transparency as a starting point. Specifically, member states should commit to  Transparent tax incentive systems, procedures and criteria;  Public disclosure of discretionary tax incentives granted;  Transparent cost control and fiscal impact; and

XVIII

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

 Effective monitoring mechanisms. 9.

Technical assistance. In implementing the Capacity Building section of the MOU (Article 3), Member States should develop joint technical assistance and training programs to improve the analysis of tax incentive policies, as well as educational programs to help government officials and stakeholders better understand the economic and financial impact of investment tax incentives.

10. Analytical tools. The SADC tax subcommittee should develop a kit of analytical tools that can be used by Member States in analyzing the effectiveness and impact of their tax incentive programs, and by the Community in comparing tax incentives programs within and outside of the region. The tool kit should include a model for evaluating the marginal effective tax rates; information systems for monitoring average effective tax rates; guidelines for monitoring tax expenditures; and guidelines for screening applicants for selective discretionary incentives. 11. Definition of harmful tax competition (HTC). Member states should consider adding additional criteria to the conditions listed as evidence of “harmful tax competition” in Article 4.3(a) of the MOU, based on the OECD (1998) report on harmful tax practices. 12. Areas for cooperation in combating HTC. Article 4.3(b) of the MOU calls on member states to avoid “introducing tax legislation that prejudices” another member state. This broad provision should be supplemented by concrete steps to mitigate the adverse effects of HTC on the Community at large. Some of the recommendations above serve this purpose by improving policy formulation and accountability. In addition, Member states should consider the following steps toward deeper cooperation:  Negotiate bilateral or multilateral protocols or treaties for the effective exchange of information, including coordination on tax audits and investigation of tax crimes.  Establish a periodic joint review of existing tax incentive programs to evaluate compliance with any SADC agreement on tax cooperation.  Participate proactively in international forums dealing with harmful tax practices.  Jointly support the concept of an International Tax Organization with a mandate to evaluate and facilitate international responses to harmful tax practices.  Develop standards to establish clear and effective legal provisions in each country on transfer pricing, thin capitalization, and controlled foreign entities.  Jointly pursue defensive actions to minimize abusive tax avoidance schemes using cross-border tax differentials.  Initiate programs for regional fiscal transfers, similar to the structural funding arrangement in the European Community, to help poorer Member States pursue more

EXECUTIVE SUMMARY

constructive approaches to improving their investment climate, in lieu of aggressive tax concessions.  Finally, any effective regional agreement to mitigate the harmful effects of tax competition requires a mechanism for resolving disputes and enforcing remedial actions in response to violations.

XIX

1. Introduction Why should schemes whose impact is either slight or unknown be so widespread and popular in developing countries?-- Shah and Toye (1978) The use of tax incentives is widespread even though the available empirical evidence on the cost effectiveness of such incentives in stimulating investment is highly inconclusive.-- Zee, Stotsky and Ley (2002)

These passages come from two major papers on tax incentives in developing countries, written a quarter century apart. They show that the use of tax incentives to stimulate investment has long been very popular and very controversial. The passages, however, differ in one important respect. The first paper found no evidence that tax incentives actually stimulate investment, whereas the second questions the cost-effectiveness of such incentives. This difference reflects the fact that research conducted during the intervening 25 years has demonstrated that tax incentives can indeed stimulate investment under some circumstances. Yet the balance between benefits and costs remains very much a matter for debate. This debate echoes throughout the SADC region. Even among well-informed officials, opinions cover the spectrum from outright opposition to strong support. Some believe that tax incentives rarely matter in attracting viable long-term investments, and that the associated distortions and abuses swamp whatever positive response may occur. Others contend with equal conviction that tax incentives can work and are essential in attracting investment, promoting growth, and creating jobs. In fact, every SADC country offers tax incentives. Indeed, many governments in the region face pressure to sweeten their incentive programs to compete with tax breaks offered elsewhere in the region and in other parts of the world. Yet they are also under great pressure to strengthen the domestic revenue systems in order to provide adequate and sustainable funding for essential public goods and services. In response to these influences, some SADC governments have recently introduced new tax incentives while others have scaled back their incentive programs. For example, South Africa and Lesotho eliminated tax holidays while Zambia recently introduced new tax exemptions for companies designated as export processing zones. In short, the appropriate use of incentive policies remains elusive.

1-2

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

The lack of consensus can be attributed in part to the technical difficulties involved in measuring the benefits and costs of tax incentives, especially without hard data on the actual performance of investments receiving fiscal benefits. Equally important, the effect of tax incentives depends on the circumstances. The challenge is to understand the conditions and the policy design features that determine whether tax incentives are likely to deliver substantial and sustainable net benefits in a particular country. The purpose of the present study, then, is to help tax officials, policymakers, and other stakeholders in the SADC region meet this challenge by explaining the lessons learned from the latest theories and empirical studies about tax incentives and their implications for the SADC region.3

1.1 Starting Point—Principles of Taxation for Developing Countries 4 The fundamental purpose of taxation is to mobilize revenue to finance the provision of public goods and services through the government budget. Therefore, the core principle of taxation is that the tax system should be an effective instrument for raising revenue. While fulfilling the revenue function, taxes also have a pervasive influence on economic decisions of individuals and businesses, and on social equity. Hence, the tax system should be structured to achieve the appropriate level of revenue as efficiently and fairly as possible. In short, a well designed tax system should be • Effective in raising revenue; • Efficient in its effect on resource allocation decisions of households and businesses; and • Equitable in its impact on different groups in society.5

EFFECTIVENESS An effective tax system is one that satisfies revenue requirements, given the desired scope and size of government and the availability of non-tax financing. In dynamic terms, an effective tax system should be elastic, in the sense that revenue rises naturally with GDP without requiring frequent ad hoc measures. An effective tax system must be consistent with a country’s administrative capacity. Even the best tax code can produce poor results if it is not

3 Especially recommended resources include Zee, Stotsky and Ley (2002); OECD (2001); Shome (1995); Shah

(1995), Bird & Oldman (1990).

4 World Bank (1988), chapter 4; World Bank (1991); Bird (1992). For a more recent review of basic tax policy

issues, see Tanzi and Zee (2000).

5 Another basic principle, though less often cited, is predictability. A tax system that is prone to unexpected

changes or arbitrary administration can be a major risk factor for investors. However, tax incentive programs frequently include a guarantee that prior recipients will not suffer a loss of tax benefits when the programs are changed—as they often are.

INTRODUCTION

1-3

well administered.6 The introduction of special investment incentives inherently complicates tax administration and creates loopholes through which companies and wealthy taxpayers avoid or evade other tax obligations. These are serious problems worldwide, but the costs are especially high in countries with weak tax administration and critical revenue constraints. Under these conditions, simplicity is a cardinal principal. Many SADC countries face a critical need to enhance tax collections (relative to GDP) in order to finance urgent demands for public services, including those essential to economic growth and social welfare, such as education, health, public security, legal and judicial systems, and economic infrastructure. Therefore, the revenue effect of tax incentives is a central concern. In countries with adequate fiscal resources, revenue risks may not be a major issue; but in countries where revenue performance is precarious, such policies may be highly imprudent.

EFFICIENCY An efficient tax system is one that minimizes the loss of economic welfare and growth due to tax-induced distortions in the incentives that guide private decisions on investment, production, technology, consumption, saving, work effort, financing, and even the legality of activities. Efficiency is especially important for poor countries that can least afford economic losses due to avoidable resource misallocation. To minimize efficiency losses, most tax reform programs in developing countries aim to apply a moderate tax rate to a broad tax base. To the extent that special incentives shrink the tax base, revenue targets can only be achieved with higher tax rates on other activities or persons that remain chargeable. This may magnify the efficiency costs that inevitably accompany taxation. 7 Yet the intent of any tax incentive policy is precisely to alter economic incentives in a direction that enhances growth potential and improves national welfare. Depending on the prevailing economic and political conditions and the design of the incentives, the actual impact can go either way -- as discussed in depth in the rest of the report.

EQUITY The concept of tax equity is endlessly debated. Yet there is widespread agreement that an equitable tax system

6 As famously claimed by Casanegra de Jantscher (1990, p. 179): “Tax administration is tax policy.” 7 As a rule, the welfare cost of any tax rises with the square of the tax rate. Thus, the welfare loss from a 40

percent tax rate is four times higher than that of a 20 percent tax rate. An exception is the case of an excise duty on goods with highly inelastic supply or demand. Rate differentials can also be rigorously justified by optimal tax theory (Newbury and Stern, 1987; Burgess and Stern, 1993). But the theoretical advantages are vastly outweighed in practice by administrative considerations. The efficiency cost of distortionary tax policy can be astonishingly high. For example in the World Development Report 1988 (85) the World Bank cites a general equilibrium study of the Philippines which found that the marginal economic cost of imposing a 25 percent import tariff exceeds 2 pesos per peso of revenue collected.

1-4

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

• Minimizes the tax burden on the poor; • Collects more from the rich than from those with lower incomes (vertical equity); • Avoids excessive tax rates and arbitrary impositions all around; and • Provides relatively uniform and non-discriminatory treatment of taxpayers with similar economic circumstances in terms of ability to pay (horizontal equity). Equity issues are often neglected in deliberations about tax incentives, but they surely bear consideration as a matter of principle, and also because perceptions of unfairness can undermine the political sustainability of an incentive program. Investment incentives directly reduce the tax burden on income earned by relatively wealthy investors. As a result other taxpayers may bear a greater tax burden. For example, if investment incentives reduce company tax revenue, then governments may depend more heavily on indirect taxes which impose a greater burden on poorer segments of society. In addition, most programs are designed to favor certain taxpayers over others in similar economic conditions. To compensate for these inequities there must be a clear expectation that the tax incentives will truly and substantially foster equitable growth and job creation. From these three basic principles, one can see why many public finance specialists ask hard questions about the advisability of tax incentive programs. Yet most governments deviate from the pure principles of taxation to pursue other public policy objectives. There is no a priori reason not to do so because the tax code is an obvious tool for social and economic engineering. The primary question, then, is whether tax incentives stimulate the intended behavior and whether their overall impact is beneficial or detrimental to ultimate policy objectives. The secondary question is how best to design investment tax incentives to achieve the objectives at minimum cost.

1.2 What are Tax Incentives? The SADC MOU on taxation (2002) defines “tax incentives” as “fiscal measures that are used to attract local or foreign investment capital to certain economic activities or particular areas in a country.”8 Zee, Stotsky and Ley (2002) adopt a similar definition in a recent review of this topic. They claim that “any tax provision that is applicable to all investment projects does not constitute a tax incentive.” This definition excludes “general tax incentives,” such as accelerated depreciation that applies to all investments. Such general tax provisions deserve to be called incentives for three reasons. First, they are designed as such and function as such. Second, it makes sense for a government to broadcast that it is offering attractive tax incentives for investment even if they take the form of general rather than selective provisions of the tax code. And third, a number of countries, such as Indonesia and Uganda, have shifted

8 Essentially the same definition is used by the International Bureau for Fiscal Documentation. While the

definition covers both direct and indirect taxes, the MOU focuses largely on the former category.

INTRODUCTION

1-5

from selective to general incentives, all with the intention of stimulating investments. Therefore, this study will encompass both general and selective tax incentives, while focusing primarily on selective incentives operating through the company income tax.

1.3 “Effectiveness” and “Impact” of Tax Incentives The theme of the study is the effectiveness and impact of tax incentives in the SADC region. Here, effectiveness is usually taken to mean the extent to which investment tax incentives stimulate investment. This definition focuses on the amount of investment, yet the quality of investment is at least as important as the quantity. Incentives that foster unsound and unsustainable investments impede economic development, just as driving faster in the wrong direction only leads you farther from your destination. Such investments should not be counted as a sign that incentives are effective. Hence, in this study effectiveness means the extent to which the investment tax incentive stimulates additional productive investment.9 If tax incentives are effective in this sense, a sound policy analysis must also take into account the associated costs. For example, an incentive that stimulates $10 of investment at a cost of $15 to the economy is a losing formula. Focusing on the gain of $10 yields poor decisions and ultimately adverse outcomes. For this reason, Zee, Stotsky and Ley (2002) emphasize the criterion of cost-effectiveness. In their view, the central issue is whether “the benefits to the economy that can be expected from an increase (if any) in the incentive-favored activities would actually outweigh the total costs of the tax incentives granted.”10 The concept of cost-effectiveness addresses consequences beyond the additional investment as such. In this study, impact refers to the broader fiscal, economic, and social implications of investment tax incentives. Thus, the analysis must examine both the effectiveness of various tax incentives in stimulating productive investment, and their impact on government revenue, tax administration, economic efficiency, social equity, and ultimately, a country’s prospects for economic growth.

9 In the case of infant industries, the expectation is that they will not initially be productive and competitive,

but will become so within a reasonable time frame so that the present value of future net benefits is expected to exceed the short-term efficiency costs. 10 Shah (1995) uses the term cost-effectiveness more narrowly to refer to the amount of additional investment per dollar of revenue loss. A ratio of less than one indicates poor cost-effectiveness. This is not a logical gauge of cost-effectiveness. The numerator should measure the present value of benefits to the domestic economy, not the gross amount of investment; and the denominator should include the present value of future revenue losses, not just the annual cost. Even when redefined this way, the ratio is still oversimplified because it neglects indirect costs and benefits.

1-6

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

1.4 Study Organization The remainder of study is organized as follows. Chapter 2 summarizes major insights from the literature on determinants of investment and growth, with special attention to the role of taxes and tax incentives. Chapter 3 focuses on the economics of tax incentives by reviewing the advantages and disadvantages of these tools for investment promotion. Chapter 4 discusses some of the main analytical tools for evaluating the economic and financial impact of tax incentives. Chapter 5 examines the advantages and disadvantages of widely used tax incentive instruments, and Chapter 6 reviews the economics of harmful tax competition. Chapter 7, which reviews tax systems and tax incentive programs in the SADC region, completes

our

analysis.

Chapter

8

summarizes

our

main

points

and

presents

recommendations for consideration by the SADC Directorate on Trade, Industry and Finance (TIFI) and SADC member countries. The appendix describes the tax incentive regimes in each SADC member country, drawing on information from the SADC tax database (as of September 2003) and supplementary sources as noted therein.

2. Taxation, Investment, and Growth Although many development practitioners and researchers continue to target capital accumulation as the driving force in economic growth, ‘something else’ besides capital accumulation is critical for understanding differences in economic growth and income across countries. -- Easterly and Levine (2001) … tax incentives do affect the decisions of some investors some of the time. —Morisset and Pirnia (2001)

In this chapter we examine two fundamental premises that underlie the justification for offering investment tax incentives in developing countries: first, that additional investment leads to faster economic growth and higher standards of living; and second, that tax breaks stimulate additional investment. These statements may seem self-evident, but both of them are subject to important qualifications that are directly relevant to understanding the effectiveness and impact of investment tax incentives.

2.1 Capital Investment and Economic Growth A standard tool used by economists to study economic growth is an empirical framework called “growth accounting.” Given the rate of population growth, the growth of per capita income is driven by • Investment in physical capital • Investment in human capital, and • Productivity growth. If productivity and the quality of labor were constant then investment in physical capital would directly determine the growth of per capita income. Empirically, however, changes in productivity and investment in human capital are central to the growth process. Virtually every analysis of economic growth highlights the importance of investing in human capital—through education, technical training, preventive and primary health care, and

2-2

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

improved nutrition—to augment labor productivity. Broad-based human capital investment is also essential for poverty reduction because it directly enhances capabilities and opportunities for the poor. The productivity effect is often referred to as technical change. In fact it refers to any source of overall efficiency enhancement. A better label is total factor productivity (TFP) growth, which measures the difference between actual output growth and that which would be explained by increases in factor inputs (capital and labor) alone. In addition to technology, TFP growth also stems from structural changes that shift resources to more productive sectors, the elimination of impediments to competitive markets, stronger financial markets, and improved management techniques. Recent models of economic growth focus on the sources of productivity growth. These theories emphasize determinants such as knowledge externalities, economies of scale in production, and a “catch-up” effect. Knowledge externalities arise when knowledge-building activities not only enhance productivity for the immediate production unit, but also create diffuse, lasting, and cumulative efficiency benefits for the economy. When these spill-over effects are significant, private decision makers will systematically under invest in knowledge-building activities because investors have no incentive to take into account the gains that accrue to others. In effect, advances in knowledge are a type of public good. This provides a justification for costeffective interventions to strengthen the incentives for knowledge-generating activities such as investments in science and technology, education and training, capital investment to upgrade production facilities, or foreign direct investment involving technology transfer. The importance of scale economies has been understood for centuries, but only recently has it been incorporated into formal growth theory. For many (though not all) production processes, a larger scale yields lower unit costs, higher productivity, and greater competitiveness. This is an important reason for pursuing outward-looking strategies, particularly in low-income countries where the domestic market is very small compared to vast opportunities in regional and global market. Finally, the catch-up effect suggests that poor countries ought to have the best prospects for rapid growth. This is ironic because most poor countries have not achieved rapid growth. (Otherwise they would no longer be poor!). The basis for the idea is that poor countries can benefit from a deep pool of existing knowledge. They can also supplement domestic resources with foreign capital and know-how, and take advantage of large global markets. Statistical evidence strongly confirms the validity of the idea, and reveals that poor growth performance in poor countries stems largely from policies and institutions that inhibit investment and undermine productivity. With appropriate and consistent policies and institutions, poor

TAXATION, INVESTMENT, AND GROWTH

2-3

countries should be in a position to achieve high rates of growth— as evidenced by more than a few remarkable success stories over the past 40 years.11 The immediate implication is that a successful growth strategy must be seriously concerned with the impact of policies on productivity, innovation, and knowledge acquisition. Investment in physical capital is one determinant of economic growth, but not the most important one. This is why Easterly and Levine, in the passage quoted at the beginning of this chapter, emphasize that “something else” beyond capital accumulation accounts for most of the differences in economic growth across countries. Similarly, Michael Porter’s influential theory of competitiveness highlights the importance of productivity. According to Porter, policies fostering investments that depend on continued subsidies or protection are “simply inappropriate… no matter what the stage of development of a nation’s industry.” (Porter 1998, 674.) Even more sharply, Devarajan, Easterly and Pack (2002) provide evidence showing that “investment is not the constraint on African development.” Their empirical analysis indicates that low growth rates in Africa are primarily due to low and declining productivity resulting from policies that foster and sustain inefficient production activities. In short, neglect of the productivity component in the growth equation can totally undermine the benefit of higher investment and impair the quest for prosperity.12 Exhibit 2-1 outlines policies that help foster productivity. One of them is the use of tax incentives to encourage efficient investment, training, and research and development. The impact of investment tax incentives on efficiency will be examined in more detail in the following chapter. But first let us examine in general terms the influence of tax policy on investment decisions.

2.2 Theory of the Determinants of Investment RISK AND RETURN The driving force behind private investment can be conceptualized in terms of risk and return. Investment projects are viable when the expected rate of return on capital exceeds a threshold (hurdle) rate that reflects the cost of capital plus a premium to compensate for perceived risk. If two mutually exclusive alternatives are both viable—for example, two

11 Examples include Korea, Taiwan, Malaysia, Indonesia, China, Botswana, Mauritius, Chile, and more

recently, Vietnam and India. All of these countries have achieved sustained periods of growth and poverty reduction at rates far higher than today’s rich countries experienced historically. 12 Moran (1998) found that nearly half of the FDI projects that he examined in 30 countries had net negative effects on host-country welfare. Pigato (2000) reviews several such studies and concludes that negative effects of FDI arise mainly from “lack of competitiveness of input and output markets in the host country, and from distortions in the domestic incentive framework.”

2-4

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Exhibit 2-1 A Productivity Package Steady gains in aggregate productivity emerge from

barriers, eliminating legal and administrative

a complicated process of evolutionary changes at

impediments to entry and exit of new

the firm and industry level, as some firms grow and

businesses, and reforming or privatizing

others shrink or fold. Joseph Schumpeter long ago

industries

identified this "creative destruction" as a central

inefficient state monopolies.

feature of growth dynamics. The following policies can make an important contribution: •

operating

as

Establish a welcoming environment to attract

of technical and managerial innovation, along

determined primarily by market mechanisms

with domestic investment and innovation.

reflecting resource costs and effective demand.



Develop policies to facilitate the introduction and adaptation and development of more productive

technologies,

Invest in human capital, including technical,

protection

scientific,

and

enforcement of contracts.

particularly

where

managerial linked

education,

to clusters

of

national competitiveness.



been

and facilitate productive FDI as a major source

and self-interest in response to price signals



have

Allow resource allocation decisions to be

that harness the power of personal initiative





that

of

property

including rights

and

the the

Tax incentives may be a useful tool for stimulating productive investment, research and development,

Expand the scope for specialization and scale

and training. But they must be carefully designed

economies by improving transportation and

and well administered so as to avoid side-effects

communications infrastructure (which also

that diminish productivity by distorting resource

reduces costs throughout the supply chain)

allocation, sustaining inefficient or unsustainable

and facilitating production for the export

activities, and losing revenue needed for other

market.

components of the productivity package.

Strengthen

competition

throughout

the

economy by reducing highly protective trade

possible locations for a given investment project—then the investor will normally favor the location offering the highest risk-adjusted rate of return. Exhibit 2-2 shows that a variety of government policies affect risks and returns for investors. The tax system, in particular, influences the viability of investment through several channels. The rate of return depends clearly on the effective tax rates. In addition, the perceived risk is affected by unpredictability of the tax code, the degree of transparency or arbitrariness in tax administration, and technical factors such as investment allowances and loss carry-forward provisions. The tax system, however, is just one of many policy variables entering the risk/return analysis. Furthermore, a host of non-policy variables have a strong effect on investment decisions. These include the distance to major markets, proximity to raw materials, climate,

TAXATION, INVESTMENT, AND GROWTH

2-5

Exhibit 2-2 Policies Affecting Investment Risk and Return How can public policies address the underlying determinants of investment—risk and return? Perceived risk can be reduced by policies aimed at •

Maintaining a commitment to macroeconomic stability, including low inflation, a sustainable budget deficit, a sustainable debt profile, and a reasonably stable real exchange rate.



Strengthening institutions to protect property rights, enforce contracts, and control crime.



Reducing the risk of a financial crisis through strict prudential regulations and effective banking supervision.



Eliminating, as far as possible, bureaucratic discretion in the implementation of laws and regulations affecting investments and business operations.



Ensuring repatriation of capital and profits from foreign investment, without restrictions.



Improving transparency in macroeconomic and financial management by adopting international standards for data dissemination.



Reducing balance of payments risk by maintaining a market-based exchange rate, pursuing policies to diversify exports, and reducing dependency on foreign aid.



Improving the dependability of basic infrastructure services through mechanisms to introduce effective management incentives, appropriate pricing for cost recovery, and timely maintenance systems.



Reducing political risks by respecting basic human rights, establishing participatory approaches to governance, and developing effective procedures for dispute resolution.



Enhancing the predictability of the tax system by establishing a well defined strategy for tax reforms and minimizing the scope for arbitrary assessments.

Investment returns can be enhanced by government policies aimed at •

Improving the quality of infrastructure, particularly in transportation, telecommunications (where moving slowly guarantees falling behind), energy, and water.



Lowering duties and barriers on international trade, to reduce costs and provide better access to wider regional and global markets.



Firmly committing to macroeconomic stability, to reduce borrowing costs and minimize risk premiums up and down the supply chain.



Investing in education and health services, including programs to roll back pandemic diseases such as HIV/AIDS and malaria.



Eliminating red-tape through deregulation, simplification of procedures, and civil service reform.



Establishing effective laws and institutions to control corruption (which is a heavy implicit tax on business).



Establishing an attractive tax system with moderate effective tax rates.



Providing special incentives that may take the form of tax breaks, direct grants, or subsidies, targeted training programs, or improved infrastructure services.

2-6

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

and the size of the local market. With so many factors at work, how important are the tax considerations? To answer this question, it is helpful to identify the principal determinants of investment by reviewing the economic theory of investment behavior.

TAXES IN THE THEORY OF INVESTMENT BEHAVIOR A simple theory of investment postulates that the desired stock of capital for any firm is proportional to the target level of output. Hence, the desired change in the capital stock each year—that is, net investment13—is proportional to the expected change in output. It follows that the ratio of net investment to GDP depends on the expected rate of GDP growth.14 This is called the “accelerator” model because it shows that investment rises when output growth accelerates and falls when output growth decelerates (even if the growth rate is still positive). In other words, an economy on a rapid growth path attracts a high rate of investment, while a stagnant or shrinking economy offers no inducement for net investment aimed at the domestic market. Of course, capital investment is itself a determinant of growth. Hence, we have an interactive system that can create either a virtuous circle of high growth and high investment, or a vicious circle of low growth and low investment. A more realistic version of the model, called the “flexible accelerator,” recognizes that the desired capital stock depends not only on output, but also on the user cost of capital (UCC),15 defined below. In this model investment takes place as long as the value of the added output from an investment exceeds the UCC—in other words, when the benefits exceed the cost. From this basic condition one can readily incorporate tax considerations into the analysis. In particular, tax elements heavily influence the UCC, which is the cost per year of deploying capital in an investment project. Neglecting taxes for a moment, the UCC consists of three elements: • The financial cost of funds tied up in capital goods. This cost depends on the price of the capital good relative to output prices (PK), and the real (inflation-adjusted) cost of funds. • The cost of depreciation or wear-and-tear of the capital asset. This cost depends on PK and the economic rate of depreciation.

13 Net investment (In) = total investment (It) – replacement investment (Ir), where Ir is the investment needed to

maintain the existing capital stock as it depreciates through wear-and-tear. In is therefore the net addition to the capital stock. For simplicity, we focus on net investment in fixed business capital, and neglect replacement investment as well as investment in inventory stocks and residential housing. 14 Let α represent the ratio of capital to output for the whole economy. Then: K* = α Q* Æ In = ∆K* = α ∆Q*, where In is net investment Æ In/Q = α ∆Q*/Q = α x gQ, where Q is real GDP. 15 This presentation draws on Chirinko (1993), Mankiw (2002), Shome (1995) and OECD (2001a).

TAXATION, INVESTMENT, AND GROWTH

2-7

• Any change in the relative price of capital goods, due to changing market conditions. This enters the picture because gains and losses on holding physical assets clearly affect the returns. Taxation enters the UCC through three channels:

Company tax rate (u). From the point of view of the investor, the effective return on capital is diminished to the extent of tax due on company income. Tax incentives. The cost of paying company tax is offset by any benefit which may accrue to the investor from tax incentives such as tax holidays, preferential tax rates, investment credits, or capital allowances in excess of economic depreciation. These benefits arise at different points in time and vary year to year. To handle this complexity, the standard approach is to take the present discounted value of the tax benefits, per unit of the investment outlay. Tax treatment of the cost of funds. Investment can be financed by equity or debt.16 Hence, the overall cost of funds depends on both the real interest rate on debt financing and the riskadjusted real rate of return required by shareholders who provide equity financing. The real cost of debt financing (rd) is the nominal interest rate (i) less the rate of inflation (π). Since nominal interest payments are deductible in most tax systems, the after-tax real cost of debt financing for the firm is rd = (1 – u ) i – π. Now let re represent the risk-adjusted real rate of return required by shareholders. If the company and personal tax codes are integrated to avoid double taxation of dividends, as in Botswana, then re is the effective cost of equity financing facing the firm. However, if a separate tax is imposed on dividends at the rate td, then shareholders require a higher tax-inclusive return on equity, namely (1+ td) re. The overall cost of funds is then a weighted average of these two elements. In this framework, investment takes place as long as the gross return on additional investment exceeds the tax-adjusted user cost of capital. In effect, the hurdle value of VMPK rises with the company tax rate and the tax on dividends, and falls with the value of the tax incentive package. A higher user cost of capital reduces the set of viable investment projects. It also provides an incentive for companies to pursue more labor-intensive projects. Conversely, a lower UCC expands the set of viable investment projects, and favors capital-intensive projects. Note that the net impact of tax breaks on job creation is ambiguous, since the changes in investment and labor intensity work in opposite directions. The theoretical effect of taxation on investment is mediated by three other considerations. First, the gestation period for many investments may span several years, particularly for large projects. So there can be substantial lags before tax policies to stimulate investment have an

16 A third major source of finance, retained earnings, is omitted at this point for simplicity.

2-8

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

actual impact. (Still, policy changes that worsen profitability may provoke an immediate cessation of planned investments.) Second, recent models that highlight the effect of uncertainty show that investors may defer projects even if they are fundamentally viable. Faced with substantial uncertainty about economic stability or the sustainability of pro-investment policies, along with irreversible start-up costs, investors may choose to wait and see how events unfold before committing funds. Implicitly, they demand a higher hurdle rate consisting of the standard UCC plus the value of the “option to wait.”

17

The result may be a very sluggish investment response. The

antidote is to reduce uncertainty by establishing a track record of dependable policy management and political stability. Third, liquidity constraints and imperfections in the financial markets can enhance the effectiveness of tax cuts. The neoclassical model assumes that investors have access to debt and equity financing at a market-determined cost of funds (adjusted for risk). This is a reasonable assumption for large multinational companies. But for many companies the main source of funds for investment is retained earnings. In this case, tax cuts can foster investment by augmenting the company’s net cash flow, providing the means to take advantage of viable investment opportunities that otherwise would be missed for lack of finance.

ROLE OF INDIRECT TAXES How do indirect taxes such as import duties and VAT enter the theoretical analysis of the investment decision? The usual premise is that a uniform tax on production or sales does not substantially affect investment incentives because it is largely passed along to final consumers in the form of higher product prices. However, investment behavior can be strongly influenced by differential indirect taxes which fall on some activities more than others. For example, the common policy of imposing lower import duty on inputs than on outputs enhances the profitability of import-substitution activities, stimulating investment that would otherwise not be profitable or sustainable. (Chapter 3 discusses the economic efficiency cost of this investment promotion policy.) Another important indirect tax issue is the imposition of import duty or tax on the purchase of machinery and equipment. If this tax (tK) were fully passed along to consumers, then it would not significantly affect the profit maximization decision of investors. But it might not be passed through fully if the duty on competing imports of final products remains unchanged, or if competing producers use more labor-intensive production processes and therefore pay less of this tax. In these cases, competition constrains the output price, so the producer would bear the cost of the indirect tax. The price of capital goods in the model rises from PK to (1 + tK) PK. By increasing the price of capital goods, the indirect tax on machinery

17 See Hubbard (1994) for an excellent review of these “real option” models.

TAXATION, INVESTMENT, AND GROWTH

2-9

and equipment increases the user cost of capital and renders some investments non-viable, especially for capital-intensive projects. Conversely, reducing tK lowers the user cost of capital and encourages additional investment and higher capital intensity.

SUMMARY In summary, the standard economic theory indicates that lower tax rates or higher tax benefits can be a significant factor in stimulating investment, even though the fundamental risks and returns are determined largely by non-tax considerations. For some investments, tax considerations can tip the balance in deciding where to locate, if non-tax fundamentals are similar in several alternative host countries. Still, for many projects, the tax system has no effect on the go/no-go decision. Tax breaks can even have a negative effect on investment if they lead to a revenue loss that adversely affects macroeconomic stability or the provision of public services which are critical to the investment climate. Thus, one cannot conclude from the theory whether tax effects on investment are large, small, or even negative. The effect depends on the context, and the outcome can only be determined from empirical analyses, to which we turn.

2.3 Empirical Evidence for Taxation as a Determinant of Investment Three types of empirical studies provide information on the importance of taxation as a determinant of investment: econometric studies, surveys of investors and businesses, and case studies. This section summarizes some of the main findings.18

ECONOMETRIC STUDIES Most econometric analyses of investment behavior have been based on data for the United States and other OECD countries. One must be very cautious in applying these results to the SADC region. Moreover, research in this field has inherent problems due to the use of simplified measures of investment flows, financing costs, and tax effects, in addition to technical problems in controlling fully for other factors that influence investment decisions. Empirical research using data for developing and emerging economies is less extensive and suffers from even more serious data problems.19

18 This section draws on reviews of the empirical literature provided in Chirinko (1993), OECD (2001), Morisset

and Pirnia (2001), and Zee, Stotsky and Ley (2002), and other studies listed in the bibliography.

19 For example, the study by Wilhems (1998) on determinants of FDI in 67 emerging economies (1978-1995)

found a highly significant negative effect of taxes on FDI. But the tax variable is simply the ratio of tax revenue to GDP. One cannot say whether the negative effect of a high tax ratio is due to the tax burden on

2-10

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

A major review of the literature on investment behavior a decade ago, by Chirinko (1993), reported that variables reflecting the UCC, which includes tax effects, had only a small effect on investment. In contrast, output variables had a large effect, in line with the accelerator model discussed above. But Chirinko also concluded that investment models in general produced weak results. Most of the studies examined in this survey used macroeconomic data for estimating investment functions. Perhaps the most important lesson from this early work is that one does not get strong and consistent results about investment behavior from macroeconomic models. To overcome this problem many researchers have begun to use microeconomic data, which produces a far larger sample size and more discerning statistical tests. These papers generally find that variables reflecting the UCC, including tax effects, are statistically significant and quantitatively important as determinants of investment. Zee et al. (2002) cite several estimates of the elasticity of investment with respect to the UCC in the range of -0.5 to -1.0. This means that a 10 per cent reduction in the UCC, due to lower financing costs or tax benefits, tends to increase investment by 5-10 per cent. This is a large impact.20 Looking more specifically at evidence for developing countries, however, the same paper reaches the following conclusion: The main messages of this research are that tax incentives can stimulate investment, but that a country’s overall economic characteristics may be more important for the success or the failure of industries than any tax incentives package; and even if tax incentives stimulate investment, they are not generally cost effective.21

Here the statement that incentives are not “cost effective” means that the amount of additional investment is less than the estimated revenue loss. In this sense, cost-effectiveness appears to vary widely by the type of investment and by the tax instrument used (Shah 1995). General tax reductions appear to be less cost-effective than incentives tied directly to new investment because a large part of the revenue loss in the former case benefits existing capital. Also, R&D investment appears to be more sensitive to tax incentives than other types of investment. Numerous studies focus on the determinants of FDI, in particular, using firm-level data or international cross-section and time-series data. Much of this research finds significant effects of both host-country and home-country tax parameters. Also, FDI seems to be getting more tax-sensitive over time as capital has become more mobile and production more global. Bear in mind, though, that the data and the results are dominated by FDI flows between industrial countries. As OECD countries have become more homogenous in infrastructure,

returns from investment, or high trade taxes that distort the trade regime, or simply the presence of a large and intrusive government sector. 20 In a widely cited study of FDI determinants, which focuses on corruption, Wei (2000) finds highly significant effects of both the statutory tax rate and the average effective tax rate. His famous conclusion: “An increase in corruption from the Singapore level to the Mexico level would have the same negative effect on inward FDI as raising the corporate income tax rate by eighteen percentage points.” 21 Zee et al., 2002, p.1508

TAXATION, INVESTMENT, AND GROWTH

2-11

labor force skills, macroeconomic performance, regulations, and so forth, it is no surprise to find that tax differences have gained in significance. It is plausible to think that tax factors have also become increasingly important in developing countries for investment projects that can readily operate in other locations with similar non-tax characteristics, but there is no clear evidence to this effect. A review of econometric studies of FDI into Africa, by Basu and Srinivasan (2002), does not even mention tax effects.22

SURVEY EVIDENCE Surveys of business decision makers are often used as a source of information on determinants of investment in developing countries. The quality of such information is often difficult to judge since the samples may not be representative, survey responses involve subjective judgments, and respondents may interpret the questions in different ways. In addition, surveys often include a bewildering list of pertinent factors. As one example, McMillan, Pandolfi and Salinger (1999) report on a survey of American electronics companies that invest in labor-intensive operations in developing countries. Asked to list their top five concerns for an investment decision, only 10 percent of the respondents listed tax incentives—ranking well below infrastructure, political stability, skilled labor, and proximity to customers and suppliers. The study also asked respondents to rank 95 “investment factors” in terms of importance for location decisions. On a scale of 1 (low) to 5 (high), tax holidays obtained a score of 4.0. Yet 18 other factors scored higher, including the overall corporate tax rate (with a score of 4.26). Topping the list were power, telecommunications, intellectual property rights, and skilled labor. International reports on competitiveness have become a virtual industry in recent years, based on a combination of business survey responses and statistical data. Perhaps the best known is the annual Global Competitiveness Report (GCR) of the World Economic Forum.23 The GCR for 2002 offers two composite indices. The first, called the Growth Competitiveness Index (GCI), is based on established academic research on the determinants of economic growth. It entails a weighted average of 32 indicators falling into three broad categories: technology, public institutions, and macroeconomic environment. Tax variables do not enter the analysis. The second, called the Microeconomic Competitiveness Index (MICI), is based on Michael Porter’s theory of competitiveness. It consists of 16 indicators of company operations and strategy, and 49 indicators of the National Business Environment. The latter set includes data or scores on physical and administrative infrastructure, human resources, technology infrastructure, capital markets, demand conditions, supporting industries, incentives, and competition. Taxes enter the MICI analysis as an element of the incentive variable labeled

22 Nor does an earlier IMF study of investment in Africa by Hadjimichael et al. (1995). 23 The WEF also published an African Competitiveness Report. This will be discussed in Chapter 7, which reviews

investment conditions in the SADC region.

2-12

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

“extent of distortive government subsidies,” which has a very low weight in the index. Thus, tax considerations are deemed to be unimportant in the GCR rankings, compared to more fundamental determinants of the investment climate. In 1999−2000, the World Bank conducted a survey of more than 10,000 firms of all sizes in 80 countries, obtaining data or respondent’s scores on hundreds of variables covering firm characteristics, investment, and perceptions of the investment climate.24 The latter category includes taxes and regulations, finance, policy instability, inflation and the exchange rate, corruption, crime, infrastructure, anti-competitive practices, and the judicial system. The most important finding from this data base is that the rankings of business constraints differ greatly by region (and presumably by country as well). In sub-Saharan Africa, the survey covers 15 countries, including 6 SADC members. For this group the category “taxes and regulations”25 is viewed as a major or moderate concern by just 33 percent of the respondents—the lowest for all reported categories. In comparison, more than 60 percent of the respondents view financing, inflation, corruption, and infrastructure as serious constraints. Surprisingly, the importance of “taxes and regulations” in Africa is also far lower than in other regions, with the exception of East Asia/China. On balance, survey studies consistently show that tax considerations rank far behind other factors as a determinant of investment. But how can this evidence of a distinctly minor role for taxation be reconciled with recent econometric findings that indicate significant tax effects on investment? The main explanation is that econometric tests identify the influence of taxes on investment after controlling for the effects of non-tax factors. Another consideration is that many international investment decisions involve a two-step process.26 Step one involves an assessment of the fundamental viability of operating in different locations. At this stage, businesses may indeed pay little attention to the tax system, unless it is unusually bad or exceedingly good. Step two is the selection of a particular host country from the shortlist of viable locations. If these alternatives are similar in terms of their fundamentals, then tax considerations can be decisive at this stage. Thus, survey evidence would show that respondents place a low weight on tax factors in the overall decision process, even though tax variables can have a significant effect on the final decisions.

CASE STUDIES Case studies provide a third source of empirical information on the determinants of investment. This information is less systematic and less representative than econometric

24 Batra, Kaufman and Stone (2002). 25 This category consists of 8 factors: high taxes; tax administration and regulations; customs, and trade

regulations; labor regulations; business registration; environment; foreign exchange regulations; and fire/safety regulations. Among this set, tax issues top the list as potential business constraints, in every region. 26 This observation is from Toft (1996).

TAXATION, INVESTMENT, AND GROWTH

2-13

studies or survey results, but often rich in detail. One of the earliest reviews of fiscal incentives in developing countries (Shah and Toye 1978) set the stage by citing three studies in Mexico, Jamaica, and Pakistan, which showed that most investment decisions were not affected by tax considerations. Another study on Brazil, however, showed that investment in the northeast region did depend critically on tax benefits. Not much has changed since that early review of the case-study literature. Most case studies assign tax considerations an insignificant role (as long as the tax system is reasonably well structured), and emphasize the role of non-tax factors in investment decisions. In some cases, such as Mauritius and Ireland, generous tax incentives have undoubtedly lured investors. And in others, such as Uganda and Indonesia, the elimination of incentives had no effect on the flow of investment—providing real-world “experiments” to prove that selective incentives are not needed if the underlying investment climate is attractive. So case study evidence can be used to support just about any conclusion.

2.4 Lesson for SADC The main lesson for SADC from the empirical literature on determinants of investment is that taxation does affect some investments in some circumstances, but in general it is not a primary consideration for investors. The context matters. In presenting case studies on FDI in Africa, Basu & Srinivasan (2002) clarify this point by classifying investments into four categories: 1.

Investment in natural resources

2.

Investment triggered by locational advantages

3.

Investment triggered by broad-based economic reforms

4.

Investment triggered by the provision of specific incentives.

Implicit in their analysis is that the fourth group—investment attracted by specific incentives —is the exception rather than the rule. To the extent that taxation does affect investment, it is important to bear in mind that the tax burden can be reduced in two ways: by making the general tax system more attractive, or by offering selective incentives. The empirical literature provides no clear guidance about which approach is preferable. On the one hand, some econometric evidence suggests that carefully targeted incentives are more cost-effective than general tax reductions. On the other hand, some survey results and case studies indicate that the overall tax system is much more important to investors than specific incentives. It is time, then, to focus our attention on the economics of tax incentives as such.

3. Economics of Tax Incentives: Pros and Cons Chapter 2 tested the dual premise that tax incentives stimulate investment, and that higher investment leads to more rapid growth and development. On both points, the conclusion is: sometimes so, but it depends on the circumstances. Tax incentives can stimulate investment, particularly for projects that could be viable in several alternative locations where non-tax conditions are reasonably similar. But non-tax factors determine the viability of most projects and dominate most investment decisions. Tax considerations are not decisive for most investments. And if tax breaks cause fiscal problems that worsen other elements of the investment climate, then the net effect on investment may even be negative. On the second premise, capital investment is indeed a major source of growth. But the productivity of investment is at least as important. Incentive policies that encourage investments with low productivity can backfire and have an adverse impact on growth. In this chapter we address not only the effectiveness of tax incentives but also their broader economic impact, in the course of summarizing arguments both for and against incentives.

3.1 Ten Arguments in Favor of Investment Tax Incentives The following propositions are reasonably familiar and well understood by most stakeholders in the policy debates. Hence, they do not require elaborate explanation. 1. Higher Profits. Measures that reduce the tax imposed on income from capital leave investors with a higher net rate of return. This translates directly into greater incentives for investment, and (hopefully) more investment. 2. Positive Externalities. Modern growth theory highlights the importance of positive spillover effects from the accumulation of knowledge and innovation. The benefits are often embodied in new capital investment, investment in research and development, and training associated with new investment, especially for export projects or investments that help to develop networks or clusters as centers of excellence. If spillover effects are significant, then

3-2

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

private decisions based on free-market signals lead to an inefficient undersupply of investment and innovation. There is then a logical justification for government intervention to enhance the market incentive for investment, research and development, and training. Tax breaks can be useful for this purpose. 3. Practicality. Although the fundamental purpose of taxation is to raise revenue, the tax system inherently and invariably affects economic incentives. It is therefore a convenient, practical, and flexible instrument for influencing incentives in a direction that contributes to other policy objectives such as investment promotion, job creation, development of disadvantaged regions, or upgrading of the labor force. 4. Signaling. In conjunction with other measures to create a welcoming investment climate, introducing tax breaks for investors can signal a country’s commitment to facilitating investment. It also provides a headline banner for marketing the country as a desirable investment destination. 5. Capital Mobility. In a global economy with high mobility capital, the effective tax rate on capital has to be low to attract inward flows of foreign investment and keep domestic savings at home to finance productive investment. 6. Tax Competition. When other countries vie for the same investments by offering special tax breaks, the stark choice may be to match the tax breaks or lose the benefits that may come from the investments, including growth-enhancing spillover effects. 7. Compensating for Other Deficiencies in the Investment Climate. A common argument in less developed countries is that attractive fiscal benefits are essential to gain the interest of investors who would otherwise not consider investing because of problems, such as unreliable or high cost infrastructure, macroeconomic instability, or a weak legal and judicial system. 8. Revenue Gains. If it is so that investors would go elsewhere in the absence of getting special tax breaks, then the direct revenue loss from offering such incentives is nil. And the indirect revenue impact can be favorable, because the new investments that materialize through the tax incentive program will create jobs and linkage effects that generate tax revenue. 9. Political Cover. The cost of tax incentives is less visible than that of investment promotion policies that involve explicit budget outlays. This argument is rarely uttered out loud, but it undoubtedly contributes to the political attraction of tax incentives, compared to alternatives that have a direct budgetary impact, such as subsidies or infrastructure development for industrial zones. 10. Experience Shows that Incentives Can Work! Above all, the main argument for investment tax incentives is that they have worked well in a number of countries. Exhibit 3-1 provides a

ECONOMICS OF TAX INCENTIVES

3-3

snapshot of four widely known success stories that have been an inspiration for many developing countries – Malaysia, Ireland, Costa Rica, and Mauritius. Taken together, these ten arguments offer impressive justification for investment tax incentives. But a proper analysis must balance these considerations by examining the other side of the story.

Exhibit 3-1 Success Stories in Malaysia, Ireland, Costa Rica, and Mauritius One strong argument for granting tax incentives is the evidence from countries that have used such policies effectively to stimulate investment and growth. This exhibit outlines the experience in four highly successful small countries: Malaysia, Ireland, Costa Rica, and Mauritius.

MALAYSIA As a charter-member of the Asian Tigers club, Malaysia has been seen as a prototype for countries that wish to make the leap from being poor, resource-dependent economies to prosperous newly-industrializing states. In 1958 Malaysia began offering tax holidays of up to 5 years aimed at import-substitution pioneering industries. By the early 1970s the government recognized that sustained growth would not be achieved through import substitution, and established export processing zones (EPZs) offering subsidized infrastructure, zero import duties, and zero export taxes. The incentive programs were amended several times over the years, including a 1986 Promotion of Investments Act that extended tax holidays of up to 10 years for export-oriented foreign investments with “pioneer” status. Other incentives included an investment tax allowance, export allowance, reinvestment allowance, special deductions for training and R&D, and direct grants for high-tech activities. Aside from financial incentives, the government pursued programs to maintain economic and political stability, improve the infrastructure and upgrade the labor force. These investment promotion programs have been famously successful. In early years they attracted labor-intensive investment in electronic assembly and textiles. As wages and skill levels rose, the investment pattern shifted towards high-tech strategic projects. According to the World Investment Report 2002 (207), “The actual impact of the incentives offered is hard to assess, although it appears that incentives have been an important element in attracting TNCs to Malaysia.” The Report notes, however, that the foregone revenue may be as high as 1.7 percent of GDP, and some of the incentives may have been overly generous or redundant.

IRELAND Over the last 20 years, Ireland has been transformed from a poor backwater of Europe to the continent’s most dynamic economy. Progress was earned through aggressive improvements in economic fundamentals, strengthening the education system, and promoting Ireland as an investment destination, with EU membership and attractive tax incentives as lures. Until the late 1950s, Ireland discouraged foreign investment, and reaped stagnation. In 1959, the government created the Shannon Free Zone to stimulate investment for export. Initially, export profits were entirely untaxed. In 1981 a 10 percent tax rate was established for manufacturing, EPZ operations, and certain service industries, including international

3-4

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

financial service centers. The government also provided financial grants tailored to each project. Still, manufactured exports did not take off until the 1980s, after the government adopted major reforms due to “sheer necessity for economic survival” (Emmons, et al., 1999, 9). The reforms included tight monetary and fiscal policy to achieve macroeconomic stability, a social compact with business and labor, and low overall tax rates. The World Investment Report 1998 (105) states that investment “has been visibly influenced by this policy,” attracting thousands of flourishing new enterprises and creating “new comparative advantage” in sectors such as chemicals, office machinery, electrical engineering and computer software. Since 1987, Ireland has been the fastest growing economy in Europe. By the late 1990s, foreign-owned manufacturing firms accounted for nearly 60 percent of gross output and 45 percent of manufacturing employment, up from a zero base in the late 1950s.

COSTA RICA Thirty years ago the economy in Costa Rica was dominated by coffee and banana exports. Today the country is a dynamic export manufacturer boasting a showcase Intel semiconductor factory. In the 1970s, the government introduced EPZs and investment tax incentives, after previous efforts to stimulate development by protecting infant industries failed to deliver material benefits. Initially, the export promotion effort failed, too, because the export zones were poorly located and overly bureaucratic, and the economy was afflicted by debt crisis and macroeconomic instability. The export boom only gained traction in the mid-1980s after the government stabilized the economy, devalued the currency to improve competitiveness, and liberalized the EPZ laws. By 2000, export earnings reached US$7.5 billion, a six-fold increase (in constant dollars) compared to 1970. Key factors in the transformation included solid democratic institutions, a highly educated and healthy workforce, stable macroeconomic policies, a supportive investment environment, and expanded transportation links to the North American market. To woo the Intel factory, Costa Rica had to compete with Mexico, Brazil, Chile, Indonesia, and Thailand. Intel looked first at factors such as economic and politically stability, human resources, openness to trade, a pro-business environment, cost conditions, the logistics of manufacturing (including air transport links), and a fast track permit process. Taxes entered the picture as an element of the cost evaluation. Though the government did not offer Intel special tax breaks or grants, the standard EPZ package was attractive enough. It included a 12-year tax holiday followed by a 50 percent exemption for 6 more years, a “job bonus” ranging from 7 percent to 15 percent of the payroll for 5 years, and a subsidized on-site training program covering every new direct worker for 3 months. Spar (1998) contends that even with these advantages, Costa Rica made the short-list only because the investment agency had been “assiduously courting” the electronics industry since 1993.

MAURITIUS Mauritius, too, has undergone an economic transformation over the last 30 years, from a low-income monocrop (sugar) economy to an outward-oriented middle-income country with strength in manufacturing, tourism, and financial services, and an emerging ICT sector (in addition to sugar). EPZs were introduced in the early 1970s. Thereafter, through 1999, economic growth averaged 5.9 percent per annum, accompanied by large improvements in health and education and an extraordinary reduction in income inequality and poverty. Important ingredients in the Mauritius success story include a commitment to economic and political stability, a well educated labor force, reasonably efficient infrastructure, effective rule of law and

ECONOMICS OF TAX INCENTIVES

3-5

respect for property rights, and flexible labor market regulations in the EPZs, combined with preferential access to the European and American markets. Mauritius retained a highly protectionist tariff structure, but the export sector was largely insulated from the adverse effects of high tariffs through the EPZ system and various subsidies, including a 10-year tax holiday for most EPZ companies and preferential interest rates on loans. In the late 1990s, most of the tax was replaced by a 15 percent company tax rate as the main fiscal benefit for incentive companies. (See Chapter 7 and the Appendix for details.) A recent IMF study (Subramanian and Roy, 2001) emphasizes that many countries have offered similar fiscal incentives without matching Mauritius’ success. They point out that Mauritius benefited from a well paid and disciplined civil service, which minimized the abuse and leakage that come from weak administration of EPZ privileges and tax incentives. Thus, the authors caution that “attempting to replicate the Mauritius experiment might be hazardous for other countries.” The common theme is that some highly successful developing countries have used investment tax incentives with impressively good effects—but always in conjunction with a broad and consistent framework of other supportive policies and institutions. SOURCES: UNCTAD (1998 and 2002); Shaw (1995); Emmons, et al. (1999); Spar (1998); Fox (2003); Subramanian and Roy (2001).

3.2 Ten Arguments against Investment Tax Incentives This section discusses ten main arguments about costs and problems associated with tax incentives, particularly selective incentives. Most of these arguments are familiar to tax specialists and public finance economists, but they are less well understood by other stakeholders. For that reason, they merit a detailed explanation. 1. Revenue Loss. As explained earlier, the central purpose of tax policy is revenue mobilization. The claim that tax incentives have no adverse impact on revenue assumes that the investments that benefit from tax incentives are additional to what would take place in the absence of the incentives. Full additionality may indeed occur in two cases. The first is where an investment is fundamentally viable in the host country but could earn a higher riskadjusted rate of return in another location, and the profit differential is small enough that a tax break reverses the location advantage. The second is where an investment is not viable under the normal tax code, but becomes so due to the tax break. Investments in this category are inherently those of low productivity. Aside from these special cases, tax incentives do cause a loss of revenue. There are four ways in which this can occur: • Redundancy: A tax incentive is redundant, or superfluous, when it does not materially affect the investment decision. This situation arises if the investment has a sufficient rate of return to be viable under the normal tax code, and cannot relocate easily to another jurisdiction. If the investment makes sense anyway, then the tax incentive simply transfers resources to the investor at the expense of the treasury. If redundancy is widespread, then the incentive program generates little additional investment and the revenue cost is high. An important

3-6

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

example of redundancy is a resource-based investment such as diamond mining projects in Botswana or titanium mining projects in Mozambique. As long as the projects are fundamentally viable, tax incentives are redundant. The investors cannot simply pull up stakes and shift their operations to Costa Rica or Bangladesh. • Partial redundancy: In some cases tax incentives may be essential to attract a particular investment, but the benefit package is more generous than necessary. The incentive is then partially redundant, and a portion of the tax break is a genuine revenue loss to the treasury.27 For example, consider a project with a hurdle rate of 25 percent and a prospective rate of return before tax of 34 percent. If the tax rate is 35 percent, then the after-tax rate of return is just 22 percent. Under the normal tax system, the project will not be implemented. However, if the investor were offered a 10-year tax holiday that boosts the net return to, say, 30 percent, then the project would go ahead. But this tax holiday actually pushes the return well above the hurdle rate of 25 percent. The project would be viable with a tax break less than half this large. So more than half of the benefit is redundant. In real-world situations, governments do not have inside information about the investor’s hurdle rate or expectation of profitability. So it is very difficult to judge whether a given incentive is essential, redundant, or partially redundant. An interesting example arose in Mozambique in 1997−1998 when the government was negotiating fiscal benefits for the $1.3 billion Mozal aluminum smelter project. There is widespread agreement that large tax incentives were essential for landing this huge strategic project. Yet there has also been much debate about whether the deal was too generous. Mozambique did not offer a zero tax rate. Rather, Mozal was subject to a 1 percent tax on gross export revenue.28 If a zero rate was not needed for Mozal, then it is not needed for most other investments. As a rule, a zero tax rate entails an unnecessary loss of revenue. • “Reverse foreign aid”: This case arises when the investor originates from a country such as the United States, Japan, or the United Kingdom, which taxes worldwide income upon repatriation, subject to a tax credit for the imputed tax obligation to the host country. Under these conditions a tax break in the host country—or a tax rate below that in the investor’s home country—is simply offset by additional tax payable in the source country. The tax incentive may have no effect on the investment decision. This situation is called “reverse foreign aid” because the host-country treasury is effectively giving money away to the source-country treasury.

27 The impact of a tax incentive offered by the host country also depends on whether its basic company tax rate

exceeds that of the source country, the technical specification of the tax incentive, whether the company has excess foreign tax credits on its worldwide income, and source country provisions for Controlled Foreign Corporation treatment of earnings retained offshore. See Slemrod (1995) for a careful explanation of the hosthome country tax linkage. 28 This device not recommended for general application because the effective tax rate varies unpredictably from year to year and from company to company depending on the profit margin. If profits are 20 percent of sales, a 1 percent turnover tax is equivalent to a 5 percent tax on profit. If profits are higher the effective tax rate is lower, and vice versa. This is not a sensible system.

ECONOMICS OF TAX INCENTIVES

3-7

As is well known, this problem can be remedied by a “tax sparing” agreement, in which the source-country allows investors to pocket the benefit from host-country tax incentives. The United States is not in the practice of approving tax sparing agreements, and the OECD (1998b) has cautioned members against extending this privilege.29 The problem of reverse foreign aid is also mitigated to the extent that investors postpone the remittance of income to the home country by one technique or another. In general, the interaction between hostand home-country tax codes is a very important consideration for investors. Hence, the effectiveness and the revenue impact of a tax incentive program may vary depending on the origin of an investment. • Indirect revenue costs: Even if investments that benefit from special tax breaks are fully additional and redundancy is zero, significant revenue losses may occur indirectly. This occurs if the tax-favored activities undercut the profitability of other producers who do pay taxes. A classic example is where companies in operating in an export processing zone serve as a conduit for smuggling, which reduces the income of competing domestic producers.30 Indirect revenue losses also arise where producers with tax preferences bid away customers, skilled labor, or raw materials from other producers who pay full tax. An excellent example arose several years ago at a business-government conference, when the president of one SADC country assured a foreign banker that his company would get a 10-year tax holiday for an investment of $10 million. By law, this tax holiday required that certain procedures be followed. But senior tax officers balked when they were instructed to process the President’s verbal commitment. They pointed out that the tax holiday would give the new bank a huge cost advantage over all of the existing banks, which were among the largest taxpayers in the country. It did not make sense, they argued, to subsidize a new non-taxpayer to siphon business away from existing taxpayers. Ultimately the tax-incentive was not rewarded—and the bank came in anyway with a $10 million investment. 31 2. Revenue Leakage through Avoidance and Evasion. Revenue losses can increase many-fold through an entirely different channel. Tax incentives often create opportunities for businesses and individuals to engage in “aggressive tax planning”—a polite term for tax avoidance. It is instructive to cite a few examples of how tax planning can convert well-intended incentives into a revenue drain. • Company “churning.” An existing company (A) can close down all or part of its operations and establish a “new” company (B) that qualifies for a full tax holiday. In the investment

29 The OECD case against tax sparing is that such agreements “provide significant scope for tax planning and

avoidance both in the country of the investor and in the country of the investment.” Also, tax sparing agreements favor the remittance of profits over reinvestment in the host country, and encourage short-term investments. 30 A recent New York Times article about customs reform in Mexico cites a government investigation which found that half of all registered maquiladora companies “were fronts for illegal imports” that have been flooding the informal market and strangling domestic producers who pay tax. Source: “Mexico is Making Headway Against Smuggling,” NYT, June 5, 2003, page W1 and W7. 31 Source: Author’s personal observation while working in the country in question.

3-8

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

promotion statistics, it looks as if the program has stimulated activity B, but in reality the result is a straight revenue loss for the host-country treasury and a windfall for the company. A common variation on this theme occurs when the tax holiday for B comes to an end. The owners may then shut down B and open a “new” company C, which continues the operations under a new name with an additional tax holiday. • Income shifting. A corporate group has a strong incentive to manipulate transactions within the group to reduce revenue and increase expenses in a tax-laden unit (A), to augment the reported income of a related tax-favored unit (B). Company B could be getting a tax holiday, or it could operating in a sector that receives a special low tax rate, such as agriculture in Zambia or manufacturing in Botswana. There are innumerable ways that this can occur, including transfer pricing on purchases and sales, loans at above-market interest rates, management fees and royalty charges, or sale and leaseback of depreciated equipment. • “Interest pump.” Gillis (1989) describes a large revenue drain in Indonesia operating through a tax incentive meant to encourage saving. Prior to tax reforms introduced in 1984, the system exempted income from interest on deposit accounts, but allowed an unlimited deduction for interest on debt. Companies could easily arrange a borrow-and-redeposit scheme with their banks, yielding a deductible interest expense and tax-free interest income. The net effect is a risk-free gain at the expense of the treasury. • False export declarations. In many parts of the world, tax incentives or subsidy schemes that target export performance have triggered false declarations that can be extremely costly. The best known example in recent years is the “Goldenberg affair” in Kenya, in which false export declarations were used to extract hundreds of millions of dollars in export subsidies.32 • Tailor-made loopholes. Exhibit 3-2 provides an intriguing example of how a seemingly small tax incentive turned into a giant loophole because of special features in its technical design. That this happens even in a country with enormous resources for tax administration, such as the United States, means that problems can be far worse in countries with weaker tax administration and less transparent processes.

32 This scheme has been widely reported in the local and international press when it became a central issue in

the loss of IMF and World Bank funding for Kenya in the late 1990s.

3-9

ECONOMICS OF TAX INCENTIVES

Exhibit 3-2 Small Companies, Giant Loopholes In 1954 the United States Congress approved an

potential loss claims.

income tax exemption for small mutual insurance

By creating and heavily capitalizing small

companies collecting less than $350,000 in annual

insurance companies, and making sure that they

premiums. The purpose was to encourage small

don’t sell much insurance, wealthy investors have

start-up

small

sheltered huge amounts of income from tax. Better

businesses, and others who had problems obtaining

yet, the Internal Revenue Service has not been

coverage from major insurers. In 1986 the provision

auditing or even keeping track of filing records for

was amended to cover privately owned companies

these “small” tax-exempt insurance companies.

as well as mutual associations. That opened the

Needless to say, this has not escaped the attention

door for wealthy individuals to avoid huge tax bills.

of accountants and lawyers who provide tax advice

According to a New York Times report in April, 2003,

to wealthy clients.

companies

to

insure

farmers,

one New York billionaire has used this provision to escape more than $100 million in taxes.

The moral of the story is that even in a system with strong tax administration, well intentioned tax

The loophole arose because the legislation did

incentives can open the door to abuses that cause

not set any limit for assets of “small” insurance

enormous revenue losses, from activities that have

companies. As long as they collect just a small

nothing to do with the policy objective. Worse, is it

amount of premium income, they qualify for tax-

possible

free status—even if they earn an enormous income

through the 1986 amendment to take advantage of

by investing reserves that astronomically exceed

the loophole that was then created?

that

special

interest

groups

pushed

any plausible business requirement for covering SOURCE: Based on David Cay Johnston, “Insurance Loophole Helps Rich,” New York Times, April 1, 2003.

Tax advisors everywhere steer clients toward tax-saving techniques like these, wherever the tax code creates opportunities to do so. Large taxpayers, who have the best tax advice and the most to gain from minimizing tax obligations, use these techniques the most. Abuse is especially likely where companies with tax holidays are not audited, by virtue of having no tax obligation. The incentive to misuse tax incentives can be mitigated (at least marginally) by imposing low basic tax rates in the first place, by minimizing tax differentials due to incentive programs, by strengthening tax investigation and audit, by simplifying the tax code to make it easier to administer, and by enforcing economically meaningful tax penalties. Technically, the revenue loss from abusive techniques could be averted with thorough diligent audits and application of provisions in the tax law regarding the use of sham transactions or non-arms-length pricing. In reality, though, tax officers in developing countries rarely detect such practices. Thus, the anticipation of aggressive tax planning should be a central consideration in the design of tax incentives in the SADC region. 3. Impact on Tax Administration. Incentive programs encumber tax administration in several ways. First, selective incentives require applying different rules to different taxpayers, which

3-10

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

inherently complicates the system. Second, preventing and controlling the abuse of loopholes absorbs highly skilled administrative resources. Third, senior tax administrators should be and generally do participate in designing tax incentives, screening applicants, and monitoring performance. Highly trained officers are thus diverted from raising revenue to managing programs designed for other social and economic purposes. As emphasized by Zee, et al. (2002, 1501): “The more scarce resources are devoted to administering tax incentives, the more other important administrative tasks would be impaired—thus jeopardizing tax collection as a whole.” 4. Economic Cost of Fiscal Adjustment. The previous subsections show that investment tax incentives can cause revenue losses through diverse channels, most of which are indirect, obscure, and difficult to measure. To the extent of these revenue losses, other fiscal adjustments are needed to cover the concomitant budget gap. The adjustments take three forms: • Curtailing government expenditure. The government can curtail the growth of spending or reduce the provision of public goods and services. In this case the impact takes the form of less funding for education, health, roads, water systems, rural development, the legal system, social safety nets, public administration, and so on. If the spending cuts impair progress in fixing other serious impediments to investment, then the overall effect of the tax incentives may be the opposite of those intended. • Increasing the tax burden on other activities and persons. The revenue loss can be financed with higher tax collections on other fronts, through increasing tax rates, broadening the tax base, or improving tax administration. If investment incentives are financed this way, then the economic gains in tax-favored sectors come at the expense of a greater tax burden on others, which may impair growth in other areas of the economy. • Resorting to other sources of financing. Instead of taking politically painful steps to close the budget gap, the government can seek instead to finance the gap. Each source of financing, however, imposes a cost on the economy. Compensating for lost revenue by borrowing from the central bank amounts to “printing money,” which contributes to inflationary pressure and economic instability. Issuing additional domestic debt tends to increase real interest rates and divert financial resources from the private sector, thus impairing investment. External borrowing contributes to the build-up of foreign debt, with well known disadvantages. Finally, foreign aid grants come with conditionality constraints and high administrative costs. Put simply, foregone revenue has an economic opportunity cost. The cost may be easy to bear if the revenue loss is small, or if the country has a strong fiscal system and ample financial resources. But the cost can be exceedingly high if the revenue loss is large, or if the country has a weak fiscal position and very scarce resources. Under such circumstances the revenue risk from introducing tax incentives should be a paramount concern in the policy deliberations.

3-11

ECONOMICS OF TAX INCENTIVES

5. Economic Distortions. Economists are very concerned about economic distortions that arise from tax incentive programs. These are not well understood by most stakeholders, yet they can be decisive in determining whether investment tax incentives foster growth and development. Tax incentives reduce efficiency and productivity in two main ways: • Fostering low productivity. Figure 3-1 shows the normal case where tax incentives reduce the effective user cost of capital from UCC0 to UCC1 and therefore increase investment from I0 to I1. The issue at hand is that capital investments to the left of I0 are viable even without the tax break; the investments that are affected by the tax break (in the range I0-I1) are the ones with relatively low productivity. Hence, tax incentive programs tend systematically to foster investments with a low or marginal rate of return. One clear case of adverse impact on efficiency is the use of protective tariffs to stimulate investment in high-cost production activities, so as to shield uncompetitive domestic producers from import competition. Exhibit 3-3 explains how this policy works.

Figure 3-1 Tax Incentives and the User Cost of Capital Value of marginal

User cost of

product of capital

capital

UCC0 UCC1

Tax benefit

I0

I1

Investment

3-12

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Exhibit 3-3 Import Duties, Effective Rate of Protection, and Economic Efficiency The effective rate of protection (ERP) is a measure of the extent to which import duties affect the ability of domestic producers to compete against import competition. Consider a domestic firm that produces cooking oil worth P* per unit in the world market. If the import duty on cooking oil is to, then the firm can charge up Pd = P* (1 + to) and still compete against imports in the domestic market. Suppose the firm uses imported inputs that cost C* in the world market, bearing an average duty rate of ti . The cost of the inputs in the domestic market is then Cd = C* (1 + ti). This firm can meet or beat import competition as long as the margin between the output price and the cost of imported inputs does not exceed Pd – Cd = P* (1 + to) – C* (1 + ti). This is the domestic value added, or VAd. In the world market, competitive producers convert the same inputs into cooking oil at a value added margin of VAw = P* – C*. The ERP is defined as the difference between VAd and VAw, expressed as a fraction of VAw (the international standard): ERP = ( VAd – Vaw) / VAw = [P* (1 + to) – C* (1 + ti)] / [P*– C*] This ratio is called the effective rate of protection because it shows the extent to which domestic producers are protected, in the sense that they can operate less efficiently than international rivals and still compete in the domestic market. The level of the ERP depends on the duty rates on imported inputs and outputs, as well as the domestic content of the activity. Some representative calculations show how powerful the protection effect of tariffs can be: Domestic content (P*-C*)/P*

Case 1 to = 30%; ti = 10%

Case 2 to = 20%, ti = 5%

20%

110%

80%

40%

60%

43%

60%

43%

30%

80%

35%

24%

100%

30%

20%

There are some remarkable lessons here. First, even moderate tariff differentials can shelter highly inefficient investments. Second, a large amount of protection can be delivered without reducing the import duty on inputs to zero. Third, protective tariffs selectively favor investment projects that have low domestic content; this distortion actually discourages the development of backward linkages

The problem is not just that tax incentives may foster investments with low productivity, but that they tilt the playing field and draw resources away from other projects with higher productivity. Exhibit 3-4 provides an instructive example from a classic paper on tax policy in developing countries by Leechor (1986). The example shows how a plausible set of tax incentives can lead to an undesirable result, namely that the least productive projects have the highest after-tax rate of return, while the most productive projects become less attractive to investors.

3-13

ECONOMICS OF TAX INCENTIVES

Exhibit 3-4 Impact of Tax Differentials on the Allocation of Capital In an early paper on tax policy in developing

10 percent tax rate. And sector C has successfully

countries, Leechor (1986) gave a simple numerical

lobbied for a 20 percent protective tariff on

example to illustrate how tax differentials can

competing imports, which boosts the market price

distort the allocation of capital and reduce economic

of the output and increases profit margins. These

efficiency. The example involves projects in three

tax

sectors—A, B, and C. Each has an initial cost of 100

investors.

in year 0, financed entirely by equity, and a ten-year

respectively,

economic life with straight-line depreciation. The before-tax rates of return are •

20 percent for A,



15 percent for B, and



10 percent for C.

differentials The

invert

the

after-tax



11.0 percent for A,



13.8 percent for B



20.1 percent for C.

rates

incentives of

facing

return

are,

Project A is the most productive use of capital resources, but the least attractive sector for

In this example the company tax rate is a flat 50

investment due to the tax differentials. The tax

percent on income net of depreciation. If all three

system pulls capital to sector C, reducing overall

investments faced uniform tax treatment, then the

productivity in the economy. Of course, this

after-tax rates of return would preserve the order of

apparent “distortion” in the allocation of capital

profitability as indicated above. Sector A would be

might be justified if sectors B and C generate large

most attractive to investors.

positive externalities or dynamic benefits to the

But sector B happens to benefit from a special

economy. Is this so in the real world?

Of course, this example begs the question of why a particular sector or activity is granted tax preferences. Recall that tax incentives can be used to enhance efficiency and growth by compensating for market failures such as positive spillover effects or underdeveloped capital markets. The fundamental issue is whether governments can and should use tax incentives as a tool for selective “industrial policy,” in the sense of subsidizing investments that may not be profitable in the short run but are expected to generate large dynamic benefits for the economy. Can governments “pick winners”? Does it make sense to shelter “infant industries”? This is not the place to discuss industrial policy in detail. Suffice it to say that even for countries such as Japan, Korea, and Taiwan, the efficacy of selective industrial policy is endlessly debated. Even experts who believe that government interventions have worked well in the Asian context generally concede that the outcome is due to special circumstances such as high-quality bureaucracy, a well educated labor force, and the linking of benefits to competitive success in export markets.33 Most experts are reluctant

33 See Pack (2000); Pack and Noland (2003); Rodrik (1995); Roemer (1994); World Bank (1993); Porter (1998, pp.

656-675). A recent paper by Hausmann and Rodrik (2003) makes an interesting case for “spurring investment in non-traditional activities” where success can easily be copied. If the market can quickly replicate a successful product or technology, then the benefits of entrepreneurship are heavily diluted and the incentive

3-14

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

to extrapolate from East Asian cases to recommend selective interventions as a tool for industrial policy in countries with weaker public administration and highly politicized procedures for deciding who qualifies for benefits. The main conclusion from decades of experience is that developing countries have largely failed in repeated efforts to nurture infant industries by subsidizing and protecting inefficient operations. The statement that incentives foster low-productivity investment is mitigated by the importance of “footloose” projects that can just as well locate elsewhere. Such projects may be highly efficient, and they can make a large contribution to knowledge transfer, job creation, and development. This argument applies mainly to footloose export-oriented projects.34 Another consideration is that tax preferences may lead to social benefits that counterbalance the economic efficiency costs. For example, Zambia has a 15 percent income tax rate for farm operations and a 35 percent rate for most other companies. This tax structure may draw resources into agricultural investments that have a low rate of return. Yet the policy reflects a political consensus to support farm enterprises, irrespective of productivity considerations. The important thing is that such political decisions should be based on a clear understanding of the economic consequences. • Distorted Technical Decisions. In addition to favoring certain sectors, tax incentives also distort technical decisions. Poorly designed programs have unintended consequences that may operate at cross purposes with other objectives of government policy. For example,  Tax preferences that reduce the cost of capital systematically favor capital-intensive over labor-intensive investments, which reduces the impact of investment on job creation.  Tax holidays will favor short-term over long-term investments.35  Exemptions from import duty on capital goods will favor production activities that depend on imported capital, reducing the incentive to develop local capital goods industries.  Exemptions from duty on imported raw materials and intermediate goods favor activities that use imported inputs, impairing the development of backward linkages.36

to experiment and innovate is suppressed. What appears to be a low-return activity for the entrepreneur can have a high rate of return for the economy overall. 34 To be sure, not all export projects are equal. Some enclave export activities generate few benefits for the host economy, even though they may be efficient in their production technology. 35 UNCTAD (2002), p. 219 cites studies showing that investment tax incentives in Brazil, Indonesia, Malaysia, Mexico, Pakistan, Thailand, and Turkey “led to distorted investment decisions, partly because they discriminated between firms that showed losses in early years and those that did not, and between relatively capital-intensive and relatively labour-intensive activities.” 36 Moran (2001) finds that linkage requirements tend to backfire by making down-stream activities less profitable, which reduces investment and slows growth. He concludes that linkage effects develop more strongly when producers are given the freedom to pursue market incentives.

ECONOMICS OF TAX INCENTIVES

3-15

These serious distortions need to be well understood, particularly in countries that place a high priority on job creation and the promotion of backward linkages to the domestic economy. On balance, investment tax incentives will reduce efficiency and productivity unless they are carefully designed, well targeted, and impartially implemented.37 Recall from Chapter 2 that the quality of investment is at least as important as the quantity in determining the prospects for economic growth and development. 6. Equity. Equity issues usually take a back seat in discussions about tax incentives, or else they are finessed by simply presuming that these programs benefit the general population by generating jobs and promoting growth. In some cases significant job creation does occur, as in Mauritius and Ireland. However, favorable impacts on equity are not a sure thing. In some cases foreign corporations or business elites obtain large benefits at the expense of the general taxpayer. This can occur where tax incentives lead to a revenue loss that is financed by higher taxes on other producers and consumers, 38 or slower progress in expanding and improving social services, or government borrowing that pushes up interest rates and crowds out financing for other users. Also, selective tax incentives that are badly designed or carelessly implemented may create inequities by giving tax-preferred producers an unfair advantage in competing with companies that pay full tax. When this occurs, the unfairness undoubtedly affects tax compliance, as well. The equity effects are often difficult to trace, but in some cases they are clear-cut. For example, Mozambique imposes a high surtax on sugar imports to stimulate investment in rehabilitating the domestic sugar industry. The surtax also led to the creation of jobs for thousands of Mozambicans. But who bears the cost? It is essentially a transfer from domestic consumers to company owners, because the surtax supports a high domestic price of sugar. This raises the question of whether other investment promotion instruments might create jobs more equitably.39 The claim that incentives are needed to foster job creation is less compelling than one would believe from investment promotion brochures. Recall that some investments that benefit from incentives would be implemented anyway. Jobs arising from such investments cannot be attributed to the program as such. Also, if tax-favored producers displace others, jobs are lost

37 A third form of economic inefficiency arises if tax benefits are financed by imposing higher tax rates

elsewhere in the economy. Virtually any tax causes an efficiency loss – called the Deadweight Loss (DWL) – by altering prices and costs faced by households and businesses. The DWL rises with the square of the tax rate. Therefore, offering preferential tax rates to activity A at the expense of imposing higher rates on activity B increases the overall DWL compared to a situation that generates the same revenue with equal tax rates on both activities. 38 Warren Buffet, one of the richest men in the world, and a major beneficiary of the recent reduction in tax on dividends in the United States, objected to this tax cut on equity grounds: “Remember that giving one class of taxpayer a ‘break’ requires – now or down the line—that an equivalent burden be imposed on other parties. In other words, if I get a break, someone else pays.” Washington Post, May 20, 2003, p.A19. 39 A corollary issue is whether the surtax is larger than necessary to stimulate investment in the sugar industry. To the extent that this is so, the redundant tax is an unmitigated burden on consumers.

3-16

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

elsewhere in the economy. Furthermore, most tax incentives draw resources toward capitalintensive investments, which have a high (often very high) cost per job created. Alternative policy tools that avoid this bias can create far more jobs at less cost. Tax incentives rarely make a substantial contribution to solving a country’s employment problem. In Malawi, for example, the labor force of 5 million is expanding by 2 percent per year. Just to break even the economy has to provide opportunities for 100,000 additional workers per year. In Tanzania, the labor force of 18 million is expanding at about the same rate, bringing in 360,000 new workers per year.40 Job creation on this scale requires policies and programs that have a much broader impact. Most tax incentive programs hardly make a dent in the problem. 7. Lack of Transparency. The fiscal cost of a tax break is much less visible than the cost of alternative investment promotion policies that involve actual budget outlays. In addition, the actual economic costs associated with tax incentives are indirect and difficult to monitor. These features may be convenient for short-term political purposes, but they are undesirable features of public policy in an era where good governance, accountability, and fiscal transparency are in demand. When costs are hidden, it is easy for governments to pursue measures that are not cost-effective, and not consistent with the stated development goals. 8. Political Dynamics. Tax incentives are a form of subsidy. They have a direct cash value to recipients, often involving large amounts of money. Consequently, companies and business groups have strong motivation to lobby for tax incentives and to exaggerate the prospective economic benefits. If incentives are provided to selected industries or regions, businesses and politicians representing other industries and regions use these precedents as arguments for pressuring government to broaden the programs and to interpret the eligibility rules liberally. This pressure often leads to a proliferation of tax breaks, which riddle the tax net with holes, multiply the fiscal and economic costs, and undermine the coherence of tax policy.41 And once businesses begin to benefit from special incentives, pressure to perpetuate the programs becomes inevitable.42 The cash value of tax incentives is also an open inducement to bribery and corruption. These abuses are particularly likely where approvals are discretionary, criteria are vague, agreements are confidential, and mechanisms are lacking to track and control the direct and indirect fiscal costs. The lobbying activities themselves represent a loss of economic efficiency, as companies devote resources to seeking profit through political influence rather than through higher productivity and product quality. 9. There Are Other Instruments. Using tax incentives to compensate for deficiencies in other aspects of the investment climate is attractive on the surface. But one of the main conclusions

40 World Bank, World Development Indicators online. 41 Gillis (1989) documents an extreme case of tax-incentive proliferation in Indonesia prior to the 1984 tax

reforms.

42 Porter (1998, 665) refers to the idea of “temporary” subsidies or protection is an oxymoron, because “it

becomes an addictive drug” that creates a strong political dynamic to preserve the benefits.

ECONOMICS OF TAX INCENTIVES

3-17

from Chapter 2 is that the tax system is not a major determinant of viability for most projects. Thus, the revenue foregone from offering tax incentives would often be put to better use in improving conditions that are more important to investors, and which contribute more broadly to growth, job creation, and development. Indeed, the provision of tax incentives may divert political attention from fixing more critical problems. For example, tax breaks may be seen as a palliative that reduces pressure to deal with high inflation, overvalued exchange rates, high interest rates, or weak infrastructure and public utilities. Incentives that are not cost-effective overall impact may end up deterring national development. On these grounds, Porter (1998) dismisses the use of tax incentives, tariff protection, and subsidies as tools for national development. He argues that tax incentives shelter companies from competitive pressures rather than promoting innovation and productivity. It is better, in his view, to emphasize policies that facilitate dynamic gains in productivity for prospective centers of excellence. Melo (2001) echoes this point in his survey of the “new approach” to industrial policy in Latin America. He finds that most countries in that region are moving away from subsidies and protection, and instead pursuing programs to improve the economic fundamentals for strategic industries. The recommended policies include • Technical and professional education, • Industrial infrastructure and services, • Establishment of quality standards, • Elimination of restrictive regulations, • Science and technology policies, and • Export marketing support. 10. Experience Shows That Tax Incentives Usually Don’t Work! Success in using tax incentives to spur efficient investment and rapid development is the exception rather than the rule. Most developing countries have tried for many years to promote investment through tax incentives with disappointing results. Even the introduction of export processing zones (EPZs) has failed more often than succeeded.43 Observing that EPZs are often ineffective and commonly become conduits for smuggling into the domestic market, Zee et al. (2001, 1507) conclude that “setting up and/or maintaining export processing zones are rarely advisable.” Furthermore, countries that are successful offer many advantages to investors other than tax breaks. As explained in Exhibit 3-1, Mauritius, Cost Rica, Ireland, and Malaysia used tax incentives in conjunction with stable economic and political conditions, a well educated labor force, good infrastructure, open trade for exporters, dependable rule of law, and effective investment promotion systems. Where these conditions are lacking, tax incentives are generally not sufficient to attract major flows of investment.

43 For recent surveys of experience with EPZs, see UNCTAD (2002), Madani (1999),and Radelet (1999). Radelet

points out that every successful developing country has used some form of export promotion program effectively, but he also emphasizes that EPZs are often unsuccessful in stimulating investments, jobs, or linkage effects.

3-18

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Other cases show that special tax incentives are not necessary for attracting investment. Exhibits 3-5 and 3-6 summarize experience in Uganda and Indonesia, where tax holidays and selective tax incentive programs were terminated in favor of a more attractive general tax regime, with no evident adverse effect on investment flows. Chile also has had outstanding success in attracting both foreign and domestic investment without offering special tax incentives or tax holidays.44

3.3 Summary There are legitimate reasons to favor the use of investment tax incentives, but also strong reasons to believe that the fiscal and economic costs may be very high relative to the benefits. In some prominent cases, incentive programs have met with outstanding success. In many other countries, however, they failed to stimulate much investment. And in a few interesting cases, they have been eliminated without harming investment flows. Results vary according to the circumstances. If anything, the norm is for tax incentives to fail the benefit/cost test. Why, then, are tax incentives so widely used in developing countries? One answer is that incentives are often driven by special interest groups that benefit from tax breaks. But this cannot explain the sincere belief in many quarters that tax incentives are an essential and constructive component of development policy. A more important part of the answer is that the potential benefits are very easy to understand, whereas the fiscal and economic costs are not. Policy decisions on tax incentive are often based on an analysis that exaggerates the likely benefits and seriously underestimates the probable costs (in addition to pressure from parties who stand to benefit from the measures). Thus, efforts to strengthen institutional capacity for economic analysis of tax incentive programs should be a first-order priority for all SADC Member States—individually and jointly. Better analysis will lead to better policies and greater cooperation in the design of tax incentive programs. Meanwhile, skepticism ought to be the starting point for considering any tax incentive proposals. Unless a compelling case can be made, based on a careful economic and financial analysis, policymakers should guard against the temptation of adopting generous tax incentive programs that have little effect on investment or job creation, while adversely affecting revenue, tax administration, productivity, and equity.

44 The website for Chile’s Foreign Investment Committee provides a list of “Reasons to Invest in Chile.” The

list includes, inter alia, strong macroeconomic fundamentals, an outward-looking open economy; high quality human capital; and modern infrastructure. Tax policy is not mentioned. Source: .

3-19

ECONOMICS OF TAX INCENTIVES

Exhibit 3-5 Elimination of Tax Incentives in Uganda, 1997 In 1997, the Government of Uganda implemented a major tax reform program. The previous system was characterized by very high company tax rates alongside widespread provision of tax holidays through fairly arbitrary and cumbersome administrative procedures. The tax reform program included complete elimination of new tax holidays in favor of a uniform 30 percent tax rate on company income, with generous capital allowances for all investors. These include an initial allowance of 50−70 percent (depending on location) for plant and machinery, rapid write-off of the balance, and unlimited loss carry-forward. The government also set to zero the import duty rate on a wide range of capital goods. Thus, the new system provides very attractive general incentives that are heavily front-loaded to assist investors. The elimination of selective incentives also greatly simplified investment licensing. What was the effect of terminating selective tax holidays in favor of moderate overall tax rates and generous investment allowances? The following table provides a remarkable answer. Three-year averages, before and after the 1997 tax reform program

Gross fixed capital formation as percent GDP Foreign direct investment

1994-1996

1998-2000

15.7 %

16.6 %

$110 million

.

$170 million

Tax revenue as % GDP

10.4 %

11.6 %

Income tax other than P.A.Y.E as % total taxes

9.2 %

9.8 %

DATA SOURCE: EIU Country Data Online, and Bank of Uganda, Annual Report 2000/2001 Comparing averages for the three years before and after 1997, the ratio of investment to GDP rose by one percentage point. Foreign investment inflows jumped by 70 percent-- and Uganda was already one of the most successful non-oil-producing countries in sub-Saharan in attracting FDI prior to the reforms. At the same time, tax revenue climbed by one percent of GDP, as income taxes other than P.A.Y.E increased despite the sharp reduction in tax rates. In October 1998, UNCTAD conducted a business survey as part of their Investment Policy Review in Uganda. Respondents ranked investment incentives and the tax system far below factors such as corruption, telecommunications, infrastructure, political stability, and inflation as obstacles to business and investment. This was quite a change from four years earlier when a World Bank survey found that taxation was the top constraint on doing business in Uganda. Even so, when asked in 1998 what the government could do to improve the business environment, businesses appealed once again for tax incentives. Business leaders always keep trying! SOURCE: Chen and Reinikka (1999); UNCTAD (2000).

3-20

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Exhibit 3-6 Elimination of Tax Incentives for Indonesia, 1984 In the early 1980s nearly two-thirds of government revenue in Indonesia came from petroleum. Non-oil revenue amounted to just 7 percent of GDP. A major reason for the low tax yield was that political pressures over many years had produced a proliferation of tax incentives that undermined the tax base and created bountiful opportunities for manipulation of the system. Wary of heavy dependence on mineral revenue, the government spent two years designing an ambitious program of tax reforms that took effect in 1984. A central feature was a cut in the company tax rate from 45 percent to 35 percent combined with the total elimination of selective tax incentives, including tax holidays, preferential tax rates, special investment allowances, and selective accelerated depreciation. The elimination of incentives provoked strong resistance and widespread fear that foreign investors would shun Indonesia in favor of countries like Malaysia and Singapore, which continued to offer generous tax holidays. The government’s decision was driven by technical analysis showing that a uniform 35 percent rate would be simpler to administer, less distortionary, more equitable, and less prone to evasion and corruption. Above all, the analysis showed that many actual investment projects would not be adversely affected by offering lower tax rates in lieu of the former incentives. What happened? The following figure, from Wells and Allen (2001), shows that the number of FDI projects dipped in 1984 but then climbed rapidly for the rest of the decade. In value terms, FDI fell from a plateau achieved the previous two years, but then soared to new heights after 1987. According to Wells and Allen, Indonesia’s share of FDI coming into the ASEAN region was nearly twice as high in 1991-96 as it was before 1984, while the share going to Malaysia and Singapore fell. Equally important, Indonesia continued to enjoy rapid growth and declining poverty until the onset of the Asian crisis in 1997.

Indonesia’s tax reform program demonstrated that special tax incentives are not necessary for attracting investment or stimulating economic development. Despite this success story, pressure to restore tax incentives has been persistent. In 1994 the government negotiated several exceptions, and in 1996 discretionary tax holidays were reintroduced by Presidential Decree. The tax holidays were dropped in 2000, in favor of a new investment allowance and accelerated depreciation. So the controversy continues. SOURCE: Gillis (1989); Wells & Allen (2001).

4. Economic Toolkit for Analyzing Tax Incentives This chapter reviews three kinds of analytical tools that are widely used to assess tax incentive programs: marginal effective tax rates; tax expenditure budgeting; and criteria for screening and evaluating projects to qualify for tax incentives. An economist’s analytical toolkit does not provide precise measurements of the effectiveness and impact of various incentive programs or a definitive determination of whether governments should grant tax breaks to particular investment projects. What the tools can do is reduce uncertainty about costs and benefits, and inform the political judgments.

4.1 Marginal Effective Tax Rate45 The statutory tax rate only partially reveals the extent to which the tax system reduces the rate of return on capital and affects the allocation of capital across alternative uses. This is because the actual tax burden on any given activity is influenced by many factors beside the basic tax rate, including exemptions, deductions, exclusions, allowances, and credits, as well as the nature of the investment itself. A standard technical method for evaluating the impact of the tax system on investment decisions is to estimate the marginal effective tax rate (METR). The METR measures the extent to which the tax system reduces the real rate of return on investment, at the margin. More formally, the METR is defined as METR = (RORbT – RORaT) / RORbT

(4.1)

where RORbT and RORaT are the real rates of return before and after tax. Suppose, for example, that the rate of return on an incremental capital project is 30 percent before tax, and

45 Useful discussions of the METR can be found in Chua (1995), Zee et. al (2001), OECD (2001), Shah (1995), Bird

and Oldman (1990, Part Three), and World Bank (1988, chapter 4).

4-2

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

18 percent after tax. From the equation: METR = (30-18)/30 = .40, or 40 percent. The METR of 40 percent indicates that the tax system diminishes the real rate of return by 40 percent (in this case 12 percentage points are lost out of 30). The METR shows how much the tax system distorts investment incentives by driving a wedge between the underlying profitability of a project and the after-tax return to the investor. The METR can be compared across projects, sectors, and countries. The larger the METR, the bigger the tax wedge. Differences in the METR reveal tax-induced biases in the incentives that drive the allocation of productive resources. In some cases the biases are deliberate aims of policy, such as preferences for agriculture or manufacturing. In many cases, however, the biases are unintended consequences of the tax system. The tax wedge is normally positive, but it can be zero or negative. An interesting example of a METR = 0 is the case of full first-year expensing of capital costs that are financed entirely with equity. To see this, consider the case where the before-tax rate of return on investment is expressed as follows: RORbt = Present discounted value of annual net earnings = PDV(E) Capital expenditure

(4.2)

K

The company tax at rate u reduces the numerator to (1–u) * PDV(E), but full expensing reduces the effective cost of the capital expenditure to (1–u) * K. 46 Hence, the after-tax rate of return is RORat = (1 – u) PDV(E) (1 – u) K

= PDV(E)

= RORbt

(4.3)

K

which implies that METR = 0 despite the positive tax rate. In this case, the government shares equally in the capital outlay (by allowing a full deduction for the expense) and the annual profit (through the tax itself). Consequently, the tax system does not affect the rate of return for by the investor or distort the incentive to invest. Note that government revenue is positive in this case as long as the project is profitable—that is, as long as PDV(E) > K. Thus, METR=0 need not mean zero tax revenue. It simply indicates that the revenue is derived in a manner that does not distort investment incentives. Several SADC countries have adopted tax incentive programs that achieve METR = 0 simply by exempting projects from tax. But in this case, the company revenue yield from the projects

46 Technically, this expression requires that tax losses can be claimed in cash or offset against earnings from

other activities in the year they accrue, rather than being carried forward. The simple formula given here also assumes that: the capital outlay is completed in year 0; there is no tax on dividends; tax depreciation equals economic depreciation; there is no import duty on capital goods; and there is no tax on capital gains (or no capital gain to tax).

ECONOMIC TOOLKIT

4-3

is also zero. With a well designed program the same incentive could be obtained without foregoing all of the company tax revenue. If we modify the example to include debt financing and deductibility of nominal interest expenses along with 100 percent expensing of the capital outlay, then the METR becomes negative. This is because the tax system reduces the real interest rate through interest deductions, effectively subsidizing debt financing, and the project as a whole. The tax wedge appears at two levels—one arising from taxes on the company, and a second stemming from taxes on the remittance of earnings or capital gains to the owners. Thus, the METR can be computed in terms of the returns seen by the company undertaking the investment, or the owners of the company. The second approach gives a better indicator of the impact of the tax system on investment decisions. But the METR computation becomes more complicated, particularly if the investor is a foreign entity because the analysis must then take into account a variety of interactions between host- and home-country tax laws.47 Most empirical studies estimating the METR from microeconomic data rely on the familiar proposition that profit-maximizing firms increase investment to the point where the after-tax rate of return on capital equals the user cost of capital: RORat = UCC. By calculating the UCC with and without tax adjustments (as defined in chapter 2) one obtains an indirect estimate of the before- and after-tax rates of return on capital. A striking result of such studies is that the METR can differ greatly from sector to sector, even if all investors face the same tax regime, with no special tax preferences. The METR differences reflect inherent biases in the basic tax code, as a function of the capital structure (machinery, plant, vehicles, land) and financial structure (debt versus equity), among other things. For example, the World Bank (1988, 92) reports on a study in Malawi which showed that the METR in 1984 varied from 47 percent to 67 percent despite a uniform statutory tax rate of 50 percent. The variations depended on whether an investment was in manufacturing or not, and whether the capital assets were short- or long-lived. Similarly, Chua (1995) cites a Canadian study which found that the METR for investment in machinery in the mid-1980s ranged from 0.4 percent in agriculture and 7.6 percent in manufacturing to 31.6 percent in utilities. A study by Broadway et al. (1995) found that METRs in Malaysia in 1993 varied by more than ten percentage points depending on the sector of activity and whether the projects were financed by bonds or retained earnings. These figures show that tax policy imposes a strong bias on investment patterns even in the absence of selective tax breaks.

47 There are many variations to take into account, including home country provisions for tax paid in overseas

operations, the presence of double tax agreements and tax sparing agreements, company options for deferring repatriation of earnings, and whether the company has an excess or deficit of foreign tax credits on a worldwide basis. These complications go beyond the scope of the present study. See OECD (2001) and Slemrod (1995) for more details.

4-4

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

The reason empirical studies often measure the METR by estimating the UCC is that the data requirements are less formidable for this approach than for the alternative of computing rates of return directly. Even so, estimation of the UCC is no simple exercise. It requires, for a “typical” investment in each sector or category, obtaining data or estimates of the respective cost of borrowing, the required return on equity, the capital composition, and the financing structure, as well as estimates of the difference between economic and tax depreciation rates.48 Policy studies (as distinct from empirical research) typically simplify the data problems by constructing representative examples of investment projects, rather than delving into microeconomic data sets. With this approach one can isolate the impact of various tax provisions by holding constant other factors that affect the METR. Using this approach, it is easy to calculate the METR because the examples are constructed to provide the information needed to compute rates of return before and after taxes. This is a very useful method for analyzing tax policies and comparing tax systems across countries. Exhibit 4-1 gives an example using a representative project to compare how the METR varies under different tax instruments. Starting with a statutory marginal tax rate of 35 percent, the example shows that the METR can vary from 11.1 percent to 45.6 percent depending on what tax incentives are in place. Of the options that were examined, the 20 percent investment credit has by far the largest effect, reducing the METR to less than one-third the statutory rate. Tax holidays have far less effect on the METR. Table 4-1 presents an international comparison of METRs using tax parameters for 1985. This analysis dramatically illustrates the difference between statutory and effective tax rates, as well as the large METR differences between debt and equity financing. For an investment financed entirely with equity (assuming 5 percent inflation), Portugal had a statutory rate of 40 percent and an METR of 45.5 percent. Guatemala had a higher statutory rate of 42 percent, but the METR was just 10.7 percent. With 50 percent debt financing the METR in Guatemala drops to 2.1 percent, compared to 28.7 percent in Portugal.49

48 Data of this sort for all SADC countries were not available for the present study. 49 Inflation also has a big effect on the METRs (not shown in table). High inflation reduces the METR through

the deductibility of nominal interest rates, but increases it through use of historical cost for depreciation, firstin-first-out inventory valuations, and lack of indexing of capital gains.

4-5

ECONOMIC TOOLKIT

Exhibit 4-1 How Different Tax Incentives Affect the Marginal Effective Tax Rate In an influential study for the World Bank, Pellechio, Sicat and Dunn (1987) calculated how different tax incentives affect the marginal effective tax rate (METR) for an illustrative investment project. The project consists of an equity-financed mix of machinery and equipment (40 percent), buildings (40 percent), vehicles and land (10 percent each). The marginal tax rate for company income is 35 percent, and a 10 percent duty is applied to imported capital goods, which account for 20 percent of the building cost and 75 percent of the cost of machinery, equipment, and vehicles. Each asset class is depreciated at a designated rate, and tax losses are carried forward. (Here, we examine only the case of straight-line depreciation.) If standard tax rates apply to the investment, then the METR is 45.6 percent. This means that the tax system reduces the rate of return on capital by nearly half. Here are some of the other results: Tax provisions

METR

1. Standard case, with 35% tax rate

45.6

2. Five-year tax holiday

29.3

3. Eight-year tax holiday

23.2

4. 20% initial investment deduction

33.1

5. 20% investment tax credit (not refundable)

11.1

6. Exemption from 10% import duty on capital goods

37.0

The up-front investment tax credit has by far the most powerful effect in reducing the METR and therefore improving the incentive to invest. Tax holidays are much less effective. The exemption from import duty on capital goods has a surprisingly weak effect on the METR. This is because the example assumes that the duty rate is low, that nearly half the capital assets are not imported, and that the import duty is a deductible expense. With multiple tax incentives, the combined effects are not simply additive. For example, combining the 20 percent investment tax credit with the five-year tax holiday gives a METR of 9.4 percent, which is hardly different from the ITC effect itself. But interacting the five-year tax holiday with the import duty exemption provides much more relief than either instrument by itself; in this case the METR is 19.9 percent. A more complete analysis would show how inflation, debt financing, and other complications, including alternative specifications of the sample project, affect METR differentials. SOURCE: Pellechio, Sicat and Dunn (1987), in Bird and Oldman (1990).

4-6

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Table 4-1 International Comparisons of the Marginal Effective Tax Rate, circa 1985 Selected Countries Ranked by Basic Tax Rate

Statutory Tax Rate circa 1985

METR with all Equity Financing

METR with 50% Debt Financing

Ecuador

20.0

13.5

10.1

Colombia

30.0

28.5

36.9

Korea

30.0

33.3

32.8

Jamaica

33.3

40.6

35.3

Brazil

35.0

54.4

45.9

Thailand

35.0

24.9

20.0

Malaysia

40.0

31.7

24.2

Portugal

40.0

45.5

28.7

Guatemala

42.0

10.7

2.8

Ireland

50.0

5.8

5.5

Note: Calculations are based on the hypothetical project described in Exhibit 4-1. Calculations assume 5 percent inflation. The low METR figures for Ireland reflect full first-year expensing of capital outlays, with no deduction for nominal interest expenses. SOURCE: World Bank (1988), page 93, based on papers cited therein by Pellechio and others.

We have extended the Dunn and Pellechio (1990) model by adding a calculation of the present value of taxes received by the government (PVtax) under each scenario. This allows one to compare the reduction in the METR—which measures the incentive effect—against the reduction in revenue, for each tax measure. One can express this relationship as ratio, with both variables being measured as percentage changes from benchmark values reflecting the standard tax regime. This ratio provides an index of the relative cost-effectiveness (RCE) of various tax incentives. If the RCE index is greater than 1.00, then the incentive effect (reduction in METR) is proportionally greater than the adverse revenue effect (reduction in PVtax). Table 4-2 shows the RCE index for 11 different tax incentive provisions, and four illustrative investment projects. The specifications are explained in notes to the table. The most striking feature of Table 4-2 is that the investment tax credit (ITC) delivers by far the largest incentive effect relative to the revenue loss, with a RCE ratio in the range of 1.43 to 1.81 depending on the characteristics of the investment project. The least cost-efficient instruments are the 10-year tax holiday and a reduction in the general company tax rate from 35 percent to 15 percent. For most instruments the RCE ratio lies between 0.95 and 1.05, indicating that the incentive instrument reduces the METR and the revenue intake in approximately equal proportions. Insights from illustrative calculations must be used with caution, because the results depend heavily on the parameters being used. This is not so much a weakness of the model as a reflection of reality. That is, the impact of any particular tax measure depends on many parameters of the tax system and the investment project. For this reason, it is very important to customize the analysis to fit the circumstances in each country. The best way to do this is to

4-7

ECONOMIC TOOLKIT

have a model in hand that can be adapted to local parameters and applied to policy issues of direct interest. To facilitate this development, a customized version of the Dunn-Pellechio model will be made available to SADC tax officials as an adjunct to the present study (along with detailed technical notes).50 Exhibit 4-2 presents a brief introduction to the model.

Table 4-2 Relative Cost-effectiveness of Various Tax Incentives Scenarios Relative Cost-effectiveness Ratios for Various Tax Incentivesa

1 0% debt; greenfield project c

METR for benchmark tax regimeb

2 50% debt; greenfield projectc

3 0% debt; 100% plant & equipment

4 50% debt; 100% plant & equipment

57.0%

52.6%

59.0%

56.0%

1

Benchmark + Tax rate = 30%d

1.01

1.00

1.02

0.98

2

Benchmark + Tax rate = 15% d

0.99

0.92

0.98

0.90

3

Benchmark + Tax Holiday of 5 years

1.07

1.12

1.05

1.07

4

Benchmark + Tax Holiday of 10 years

0.96

0.88

0.95

0.85

5

Benchmark + Double Declining Balance

1.03

1.30

1.03

1.21

6

Benchmark + 20% ITC (no adjustment to basis) e

1.43

1.72

1.51

1.81

7

Benchmark + 50% IA (adjustment to basis) e

1.06

1.30

1.04

1.21

8

Benchmark + 50% IA (no adjustment to basis) e

1.05

1.07

1.03

0.96

9

Benchmark + Dividend Tax = 0%

1.02

1.07

1.02

1.05

10

Benchmark + Capital Gains Tax = 0%

1.00

0.81

1.01

0.84

11

Benchmark + Import duty on capital goods = 0%

1.03

1.22

1.03

1.21

Notes: aRelative cost effectiveness (RCE) = percentage decline in METR/percentage decline in PV Tax. When RCE > 1, then the incentive effect (% reduction in METR) exceeds the direct revenue effect (% foregone revenue, in present value terms). b Benchmark case: 35% company tax and capital gains tax; declining balance depreciation at rates of 5%, 15%, 25% for buildings, plant & equipment, and vehicles, respectively; 15% dividend withholding tax; unlimited loss carry-forward, but no loss offset; 10% inflation, 25%; nominal interest rate (to accentuate debt effect), without indexing; 10% duty on imported capital goods; sale of company after 10 years. c Greenfield project = 10% land, 40% building; 40% plant & equipment; 10% vehicles d Capital Gains Rate also adjusted to equal to the tax rate e ITC = investment tax credit; IA = initial allowance SOURCE: Author's calculations using Dunn & Pellechio (1990) METR model, modified to include the present value of taxes.

50 This is done with explicit permission from Anthony Pellechio.

4-8

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Exhibit 4-2 METR Model of Dunn and Pellechio The marginal effective tax rate (METR) on an

structure, the inflation rate, and the interest rate. A

investment project can be quite different from the

large side table displays in full detail the various

statutory tax rate on company income, depending

calculations that determine the Before Tax Cash

on many provisions of the tax system, the incentive

Flow and After Tax Cash Flow, for up to 30 years.

regime, and the characteristics of each project. The

Careful appraisal of the cash flow table allows the

METR analysis used in this study is an Excel

user to develop a close understanding of the subtle

spreadsheet

interactions between various tax parameters. For

model

developed

by

Dunn

and

Pellechio (1990). The model allows the user to

example,

one

can

see

exactly

how

a

final

investigate how alternative tax policies affect the

withholding tax on dividends or a heavy tax on

real rate of return on investment for a wide variety

capital gains may offset much of the intended

of tax parameters and types of investment.

advantage of granting a tax holiday.

Specifically, the user defines a scenario by

The utility of the model lies in its flexibility and

setting parameter values for an array of tax policy

simplicity in allowing the user to specify and

instruments including the corporate and personal

evaluate a wide variety of alternative situations.

income tax rates; the tax rate on dividends and

The

capital gains; the capital allowances; the extent and

mechanically. Notwithstanding all of its detail, the

duration of any tax holidays; the availability of

model says nothing about how firms might modify

investment tax credits; the tax on imported capital

their behavior in response to changes in the tax

goods; as well as provisions for loss carry-forward,

system. The METR model simply enables the user

loss offset, and indexing, among other fine points.

to get a fairly complete picture of how alternative

In

tax policies affect the incentive to invest, given any

addition,

the

user

specifies

the

physical

composition of the investment project, the financial

model,

however,

should

not

be

used

particular project structure.

The METR analysis is an excellent tool for assessing the tax wedge, but it does not provide information about how tax breaks will spur investment. If tax reforms or new tax incentives reduce the METR from 40 percent to 20 percent, will that have a large effect on the volume of investment, or hardly any? If non-tax elements of the investment climate are in poor shape, there may be few investments forthcoming even if the METR is very low. And a high METR may have no impact on investments that have very attractive fundamentals, such as major petroleum projects. In other circumstances, many investors may be considering projects where expected returns are close to the hurdle rate; if so, then a high METR will cut them out, and a low METR will bring them in. Another caveat is that the METR analysis does not capture slippage in the enforcement of tax laws. In circumstances where tax administration is arbitrary, ineffective, or corrupt, or where businesses routinely find ways to misuse investment incentives to shelter other income from tax, then investors the formal provisions of the tax law may have little meaning. In these conditions, the METR analysis does not reveal much about the effect of the tax system on investment incentives.

ECONOMIC TOOLKIT

4-9

As an alternative to examining the METR as usually defined, the Division on Investment, Technology and Enterprise Development (DITE) at UNCTAD recently developed a comparative tax model (CTM) to use in comparing the tax burden on investment across sectors and countries. The CTM measures the tax burden in terms of the present value of taxes paid (PV Tax) as a percentage of the present value of the net cash flow for a representative investment project.51 This analysis takes into account corporate tax rates, tax holidays, loss carry-forward provisions, capital allowances, tax on dividends, and customs and excise duties on inputs. In a recent Investment Policy Review for Lesotho,52 DITE presents calculations of PV Tax for six industries: agriculture, hotels, professional services, manufacturing, leisure, and ICT. For manufacturing, which bears a tax rate of just 15 percent, the model shows that the tax burden in Lesotho is highly competitive with tax regimes in South Africa, Botswana, Namibia, Mauritius, and Malaysia, though not Egypt (where the applicable income taxes are zero). Yet the PV Tax in Lesotho is very high in the hotel, professional service, and leisure industries. These results refer to specific types of companies in each industry, and are based on simplifications of the cash flow and the tax code. Nonetheless, the model is an excellent vehicle for testing whether a country’s tax burden is reasonably in line with that in other countries which may be competing for investments. DITE has been working with the World Association of Investment Promotion Agencies to obtain data on tax systems and investment profiles in numerous countries, in order to expand their data base for calculating tax comparisons. DITE plans to offer this service to any government that wishes to furnish relevant fiscal parameters and request a comparative analysis.53 Another frequently used measure of the tax burden is the average effective tax rate (AETR). This is the ratio of actual tax receipts to an appropriate measure of the tax base. The AETR provides is a simple indicator of the how well the tax system is generating revenue. For direct taxes, the AETR can be applied to specific companies at the microeconomic level, or to entire sectors using an aggregate measure of income from the national accounts. The AETR, however, does not provide significant information about the impact of the tax system on investment incentives.

51 OECD (2001a) uses a similar measure: the present value of tax obligations (in constant prices) per unit of

initial investment. This is less adequate because it does not directly compare the tax to the rate of return.

52 DITE has produced thorough Investment Policy Reviews for eleven countries to date, including four SADC

members: Mauritius (2001), Tanzania (2001), Botswana (2002) and Lesotho (2003).

53 Information is available on the internet at http://r0.unctad.org/ipr/ or by email via [email protected] Source:

UNCTAD (2003) and author’s personal communication with DITE.

4-10

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

4.2 Budget Impact—Monitoring “Tax Expenditures” Any waiver or reduction of a standard tax obligation is, in essence, a use of government funds and a form of government assistance to the beneficiary. Because tax preferences are analogous to direct budgetary funding, they are commonly known as tax expenditures. This category includes most investment tax incentives, as well as tax preferences for other economic and social policy purposes such as the promotion of home ownership, charitable donations, medical plans for employees, and child care. In principal, governments should be just as accountable for tax preferences as they are for ordinary government expenditures. To this end, tax expenditures should be reported along with the annual budget presentations. The United States and Germany were the first countries to introduce regular reports on tax expenditures, in the late 1960s. The practice spread among OECD countries, but it is still a rarity in developing countries. Brazil is one exception, and South Africa is developing a tax expenditure budget. That very few countries report systematically on tax expenditures suggests that the exercise is not entirely straightforward. Should it be a norm for SADC countries to report on tax expenditures, particularly for investment tax incentives? Let us see what is involved. The concept of “tax expenditure” can be defined in various ways,54 but the key element is that tax expenditures are foregone revenues due to provisions of the tax code that deviate from the normal tax structure. Most definitions add that the provisions are intended to influence private decisions or provide benefits to designated groups. Operationally, tax expenditures are usually estimated from tax return data by computing the difference between the actual tax liability for each taxpayer, and what would have been due in the absence of each provision. Tax expenditure budgeting has three problems. First, computation is difficult to perform in countries that lack computerized tax information systems. Second, some arbitrary judgments are needed to determine which provisions are deemed tax expenditures. There is an inherent fuzziness in defining the “normal” tax structure as a basis for identifying “deviations.” Deep disputes have arisen over whether the normative system should use the traditional definition of income based on accretions, or modern theories that suggest excluding capital income from the tax base to establish neutrality between incentives for consumption and saving.55 Using the former approach, the elimination of tax on interest, dividends, and capital gains are special tax breaks; using the latter approach, these are all normative principles, rather than tax breaks. Other debates involve technicalities such as whether declining balance depreciation is a tax break or a reflection of geometric obsolescence.

54 A good example is the definition used by the Canadian government: “Tax expenditures are foregone tax

revenues, due to special exemptions, deductions, rate reductions, rebates, credits and deferrals that reduce the amount of tax that would otherwise be payable…to encourage certain kinds of activities or to serve other objectives.” Source: 55 This is emphasized in Hall (1999).

ECONOMIC TOOLKIT

4-11

The most serious problem arises from the usual methodology, which takes the size of the tax base as given. Yet tax breaks alter taxpayer behavior. (That is why they are adopted.) To see the importance of this issue, suppose that the standard tax rate is 35 percent and company A faces a preferential rate of 15 percent. If A’s reported income is 1000, then the tax expenditure is 200 (= 350 payable at the standard tax rate less 150 actually due). This is an accurate measure of the benefit enjoyed by A. But it may or may not be a good measure of the “revenue foregone” by the Treasury. If the incentive is redundant—that is, if the tax preference had no effect on A’s investment decisions—then the tax expenditure calculation is fully accurate as a measure of the direct revenue cost. But what if A exists only because of the tax break? Then the revenue loss is zero: there would be nothing to tax at the 35 percent rate. For the intermediate case of partial redundancy, the revenue loss lies somewhere between 200 and zero. In general, tax expenditure analysis overstates the direct revenue loss because it neglects changes in taxpayer behavior due to the tax differentials. At the same time, it may understate the revenue loss by neglecting indirect effects. Exhibit 4-3 explains more fully the difference between the tax expenditure, as normally calculated, and the overall revenue impact of a tax incentive. In light of these problems,56 is tax expenditure analysis worth using? In countries with adequate data systems, the answer is a qualified yes. The conceptual gray areas are not a major problem because most tax incentives fall clearly into the category of tax expenditure. Also, empirical evidence from many countries (see chapter 2) suggests that most investments are not driven by tax incentives. Hence, the tax expenditure tabulation provides a useful approximation of the direct fiscal cost of incentives. Most important, tax expenditure budgeting subjects incentive programs to public scrutiny and facilitates monitoring of the fiscal costs.57 Far better to measure these costs imperfectly than not to measure them at all, which would leave the programs wide open to misuse. Some simpler techniques can also be used to monitor and control the fiscal cost of investment tax incentives. One method that has been used in Malawi and Mozambique is to establish a ceiling for the value of tax breaks that can be approved by the government each fiscal year. A second method is for the government to issue vouchers that are redeemable in lieu of taxes due, up to a ceiling for each fiscal year. Tanzania is testing this approach, with the intention of generalizing the practice if it works well.58

56 Another issue is the philosophical objection to the presumption that tax breaks are a use of public resources.

The argument is that income inherently belongs to the earner, not to the state.

57 The IMF Manual on Fiscal Transparency (2001, paragraph 69), puts it this way: “Although there can be

serious difficulties in cost estimation, reporting the approximate cost of tax expenditures and describing the basis of the estimates can significantly enhance transparency.” 58 A similar method is to include the tax benefit in the budget accounts by booking the normal tax liability and then covering it partially or fully as government expenditure on investment promotion.

4-12

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Exhibit 4-3 Foregone Revenue from an Investment Tax Incentive The usual method for estimating the “tax expenditure” cost of a given tax incentive uses a database of taxpayer records to compute the difference between (a) the actual tax due and (b) the tax that would have been due if the incentive were not available, for each filer who has claimed the tax break. Call this difference the Tax Benefit (TBi) for each beneficiary i. Summing over all beneficiaries gives the total tax expenditure:

Σ TB

Tax Expenditure =

i

(1)

{i} Equation (1) accurately measures foregone revenue only if taxpayer behavior (and thus the tax base) is not altered by the incentive, and if there are no indirect revenue effects from the tax incentive program. In reality, the “ifs” are rarely valid. Recalling the discussion in chapter 3 of redundancy and indirect revenue effects, the actual revenue loss from a tax incentive measure can be expressed as follows: Revenue Loss =

Σ (TB

i

x REDi ) + IRE1 + IRE2 + IRE3 + IRE4

(2)

{i} Equation (2) says that the revenue loss depends on: •

TB: The size of the tax benefit. The larger the tax break, the larger the direct revenue loss.



{i}: The size of the set of beneficiaries. A narrowly targeted benefit is less costly than a broad one.



RED: The redundancy ratio. If RED=1.0, then the incentive is superfluous and the investments would be undertaken anyway. The full tax benefit is a loss to the treasury. If RED=0.0, then there is no redundancy. The investments would not exist without the tax benefit, so there is no genuine revenue loss. When RED falls between 0 and 1, then the incentive is partially redundant and the effective revenue loss is a fraction of TB.



Four indirect revenue effects: • IRE1: Tax-favored investments may generate revenue by creating taxable jobs and stimulating other taxable enterprises through linkage effects and positive externalities. This indirect effect on revenue is positive. • IRE2: Expansion of the tax-favored activities may reduce taxable income for other entities by undercutting their market position, bidding away resources, or (in the case of EPZ incentives) providing new channels for smuggling. This indirect effect is negative. • IRE3: The tax incentive may create loopholes that provide new opportunities for tax planning, tax avoidance, and corrupt practices in tax administration. The third indirect effect is also negative, and often very large. • IRE4: The tax incentive may complicate the tax system and require administrative procedures that divert tax officers from their basic task of collecting revenue. This effect is also negative.

Comparing equations (1) and (2), one can see that the tax expenditure is only a partial measure of revenue impact. It will be an overestimate if RED is low or IRE1 is large. Or it will be an underestimate if RED is high and IRE2-4 are large. The redundancy rate and the IREs are hard to measure, so the bias is difficult to determine. If only the tax expenditure can be compiled routinely, the question is whether it is better to have a partial estimate of the revenue loss, or none at all?

4-13

ECONOMIC TOOLKIT

4.3 Screening Criteria Selective tax incentives require screening criteria to determine which projects qualify. In some cases, the screening criteria also determine the nature of the benefits.59 The determination may simply involve confirming objective criteria, such as establishing that a project is a manufacturing operation, or that safari vehicles are being purchased by a tourism company. In many other countries eligibility is limited to investments that are expected to have a favorable economic impact, or even more narrowly to projects that are deemed to be strategically important for the economy. In theory, the screening should be based on a careful cost−benefit analysis that includes an appraisal of both the economic and financial rates of return on a proposed investment project. Again in theory, special incentives should be limited to projects falling in cell 3 of the classification matrix shown in Table 4-3, that is, projects that have a high economic rate of return and—counter intuitively—a relatively low financial rate of return. The reason for the latter condition is to screen out projects where the tax incentives would be superfluous. As Usher (1990, 169) points out, offering incentives to “prize” projects is a waste of public resources. The program “seems to be a glowing success because [tax-]subsidized firms are in the forefront of economic progress,” but in reality the incentives do not affect prize investment decisions and simply sacrifice revenue.

Table 4-3 Conceptual Basis for Screening Selective Tax Incentives

High FRR Æ project viable without incentives

Low FRR Æ project viable only with incentives

High ERR Æ large benefits to economy

Low ERR Æ small benefits to economy

Cell 1

Cell 2

- Excellent project

- Poor project

- Tax incentive redundant

- Tax incentive redundant

Cell 3

Cell 4

- Excellent project

- Poor project

- Tax incentive effective

- Tax incentive effective

Note: ERR = economic rate of return; FRR = financial rate of return

This classification may help to clarify the economic concepts, but it is less helpful in practice because the estimation of economic rates of return (ERR) is a difficult and imprecise exercise. In low-income countries where technical capacity is limited, the analysis may not be feasible

59 Even general incentives may involve discretion. For example, the deductibility of training costs or research

expenses may require prior approval by the tax commissioner.

4-14

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

unless it is based on highly simplified methods.60 If the screening is meant to identify strategic projects that will deliver dynamic spillover benefits through knowledge externalities or technical innovations, then ERR estimates are little more than guesswork in the best of circumstances. Most countries that offer selective incentives sidestep formal economic analysis by using simple criteria, such as limiting incentives to projects in “pioneering industries,” nontraditional export activities, or designated sectors like manufacturing, agriculture or tourism. Other common criteria include the size of the investment, the number of jobs created, the capital cost per job, “significant use” of local raw materials, net earning or saving of foreign exchange, rural enterprises, and positive domestic value added. All of these criteria have been employed in tax incentive programs in various SADC countries. Some of the standards are nonsense. For example, “positive value added” will not rule out anything, because no company will seriously consider an investment where the expected revenues are smaller than the cost of raw materials and intermediate goods. Likewise, nearly every project producing for the domestic market can claim to be “saving foreign exchange.”61 In addition, the threshold levels for eligibility are often unstated, which renders the evaluation highly arbitrary and discretionary. The main problem, however, is that none of these criteria focus on stimulating efficient projects that would not be undertaken in the absence of special incentives. As Krugman (1983) concluded from his review of “popular criteria” for targeting industrial promotion policies, it is very difficult for governments to “devise criteria…which will by and large pick the right industries,” or establish procedures that will not be driven by political interests rather than genuine economic benefits.

60 When Botswana offered investment subsidies through its Financial Assistance Program, one criterion was

that qualifying projects needed an ERR above 6 percent. According to a senior government official, the ERR analysis was rarely carried out. The program was phased out in 1999 after it encountered widespread abuse. 61 The worst criterion that the author has seen appeared in a draft schedule to the Incentives Act in a certain SADC country in 1997. The criterion was: “any activity which in the opinion of the incentives committee promotes economic activity.” In this country, technical discussions about the criteria for awarding incentives held up action on the Incentives Act for several years.

5. Design of Investment Tax Incentives Governments that choose to use the tax system to promote investment can deploy a variety of tax tools. In this chapter we examine the advantages and disadvantages of several common tax incentives, focusing mainly on provisions of the company income tax. Table 5-1 at the end of the chapter summarizes our analysis. To keep the issue of policy design in perspective, it is worth reiterating that most investment decisions are dictated by non-tax considerations. Even a zero tax will not attract much of a response in the face of other serious deficiencies in the investment climate that render projects unprofitable. And if the investment response is concentrated in export enclaves, benefits to the broader economy may be small and possibly non-sustainable.62 Thus, the over-riding concern of investment policy should be to pursue policies and programs that improve investment fundamentals. Within the realm of tax policy, attention should focus first on establishing a basic tax system that is effective, fair, and efficient, since this will stimulate productive investments in all spheres of activity. Adding special incentives to a tax system that is punitive, arbitrary, ineffective or highly distortionary is a poor substitute for improving the basic tax structure— rather like putting attractive decorations on a vehicle with a broken engine.

5.1 Design Considerations The assessment of alternative tax incentive instruments will be viewed here through four lenses:

62 Also, companies that intend to manipulate the tax concession to shelter other income from taxes might

undertake loss-making investments.

5-2

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

• Effectiveness in stimulating investment. The basic indicator of effectiveness is the extent to which each tax tool improves investment incentives by reducing the marginal effective tax rate (METR). This gives a rough approximation but not a pat ranking, since METR differentials vary depending on the type of project and other characteristics of the tax system. Effectiveness also depends heavily on whether the incentives accrue to marginal investments or to projects that would be undertaken anyway. When tax benefits are redundant, the beneficiary investments cannot properly be viewed as arising from the incentive program. • Revenue foregone. To gauge the generosity of each instrument, one can calculate the revenue foregone per beneficiary. But a larger consideration is the number of beneficiaries: does the benefit accrue to every filer, to new investments only, or to a selective subset of new investments? The smaller the set of beneficiaries, the lower the tax loss, other things being equal. A second major factor is whether the incentive is carefully targeted to investments that would otherwise be lost to the economy. If so, then the effective revenue loss is zero. If the incentives are redundant, however, then the tax break is a straight loss to the treasury. (See Exhibit 4-3.) • Economic efficiency. Various tax instruments introduce different kinds of economic distortions, which bias investment decisions and reduce efficiency. • Tax administration. Tax incentives also differ in terms of administrative simplicity, the ease of estimating the foregone revenue (tax expenditures), and the scope for abuse. The latter should be a central consideration for the design of any tax incentive program. Zee, et al. (2001) suggest combining these factors into an overall measure of cost-effectiveness for each tool. This is a useful way to conceptualize the policy analysis, but in practice many relevant factors are difficult to quantify consistently. Beyond these four broad criteria, several other general considerations pertain to the design of a tax incentive program: • Incentives that create large tax differentials between favored and non-favored activities provide the greatest opportunity and strongest incentive for abuse of the incentive as a tax shelter. • The effectiveness and impact of an incentive depends on characteristics of the investor. In the case of investments coming from countries that impose residence-based tax on worldwide income, host-country tax incentives may be neutralized by tax obligations in the home country (see chapter 3). Another important consideration is whether the investor has ongoing operations generating tax obligations that can be offset by tax losses that arise in early years of the new project. • The value of a tax incentive to the investor depends on technical features of the tax code such as loss carry-forward provisions, ring-fencing of tax losses of incentive-eligible activities, the taxation of capital gains, whether the basis for capital allowances and capital

5-3

DESIGN OF INVESTMENT TAX INCENTIVES

gains is indexed for inflation, whether depreciation deductions are mandatory during tax holidays, the definition of start-date for tax holidays, and many others.63 • “Up-front” tax relief is more valuable to investors than a benefit that is deferred or spread out over time. Timing is especially important for companies that may be facing liquidity constraints, including many domestic investors. • Selective

incentives

require

screening

procedures,

which

can

create

significant

administrative costs and delays. • To ensure accountability, full and prompt public disclosure should be a resolute requirement for any system of selective tax incentives. • In theory, discretionary incentives can target projects with the greatest strategic value to the economy, and which would otherwise be lost to the economy. In practice, discretion often invites abuse, since large sums of money are at stake. The risk of abuse is especially acute if incentive agreements are not disclosed to the public. Incentives granted on the basis of well defined technical criteria are less open to abuse.64 • The institutional apparatus is very important. The determination of which investments receive incentives can be controlled by the minister responsible for trade and industry or another designated body. But the fiscal consequences of a tax incentive program should unequivocally be under the control of the minister responsible for finance. After all, revenue mobilization remains the central purpose of taxation, and incentives can impose high revenue costs. Tax breaks are a use of public funds, so their fiscal impact should also be reported regularly to Cabinet, Parliament, and the public. • Temporary tax incentive programs can be used for counter-cyclical fiscal policy, to make investment more attractive and to induce investors to implement projects earlier. • In a system that provides businesses with substantial tax incentives, it is worth considering the introduction of a low-rate minimum tax to ensure that every operating entity contributes at least something to the provision of public services.65

63 Broadway, Chua and Flatters (1995) provide an example from Malaysia where the provision of a 5-year tax

holiday to pioneer industries had the adverse effect of increasing the METR for companies with a high debtequity ratio or large tax losses to carry forward through the holiday period—a description that fits most mining projects. The adverse effect occurred due to an increase in the real cost of bond finance during the zero-tax period, combined with limits on the loss carry-forward to post-holiday years. 64 See the previous chapter for a discussion of technical weaknesses in commonly used screening criteria. Even though the criteria do not generally allocate tax benefits in the most efficient way, they still reduce the scope for abuse. 65 See the essay on this topic by Janet Stotsky in Shome (1995), pp. 263-266.

5-4

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

5.2 Assessment of Alternative Tax Instruments66 DIRECT TAX INCENTIVES Low Statutory Tax Rate Lowering the statutory tax rate is the least distortionary form of investment incentive. A low standard rate applies uniformly to all profitable business activities without biasing the allocation of capital, the choice of production technology, or the form of financing. This measure averts the problem of giving new investors an unfair advantage over existing producers. It also gets top score for administrative simplicity, since no complications are added to the system. A low standard rate is automatic and transparent. And, very importantly, it reduces the incentives for abusive tax planning. Lowering the tax rate can be very effective. For example, the Government of Indonesia found that cutting the tax rate from 45 percent to 35 percent was just as attractive to most investors as the complex system of incentives that were previously in place (Exhibit 3-5). The main drawback is the impact on revenue. This is because the tax relief accrues to all businesses, whether or not they are undertaking new investments. Normally, a substantial reduction in the company tax rate is packaged with other measures that enhance revenues.

Preferential Tax Rates Preferential tax rates are tax-rate reductions that apply to designated sectors, such as manufacturing or agriculture; or to activities with particular characteristics, such as a listing on the stock exchange; or to selected beneficiaries based on discretionary screening criteria. For qualifying enterprises, this benefit affects the METR in exactly the same way as a reduction in the basic tax rate. The revenue cost is lower, though, because the set of beneficiaries is restricted. The administration of multiple tax rates requires determining whether a filer satisfies the designated criteria. Is a particular enterprise engaged in manufacturing or assembly? Are all activities on a farm given the special preference for agriculture? How is the tax computed for an enterprise that combines manufacturing, distribution, and financing? Questions like this are a constant headache for tax administrators. In addition, preferential tax rates open the door for aggressive tax planning by companies engaged in multiple businesses. Thus, the revenue loss can be much larger than it appears on the surface. Furthermore, the greatest tax relief goes to the most profitable activities—exactly the ones that least need a special incentive to be viable. Thus a large portion of the tax benefit is likely to be redundant.

66 This section benefits from the excellent discussion in Zee, et al. (2001).

DESIGN OF INVESTMENT TAX INCENTIVES

5-5

Preferential tax rates are more distortionary than overall rate reductions because they bias investment incentives. As shown in Chapter 3, resources are attracted to projects in the favored sector even when they have a low rate of return. This reduces overall productivity of the economy, unless the designated beneficiaries, on the whole, yield substantial positive externalities or dynamic benefits. That is always the wish, but rarely the outcome. At the same time, a preferential tax rate is less distorting than many other forms of tax incentive because it does not bias the choice of factor intensity or other technical decisions.

Tax Holidays A tax holiday is a preferential tax rate with a limited duration (often five years). Tax holidays are surprisingly popular in developing countries, considering how harshly they are criticized by tax specialists. The revenue loss from a tax holiday depends on multiple factors, including the size and scope of the tax break, the length of the holiday period, and characteristics of each investment project. A zero rate over a long period obviously entails a maximum revenue loss. As with a preferential tax rate, the value of a holiday is greatest for projects that are highly profitable— which are most likely to go ahead even without the tax break. Thus, tax holidays are often redundant to the investment decision (though happily accepted by the investor). The administrative impact of a tax holiday is mixed. Many countries do not require the beneficiary to file returns during the holiday period, which simplifies tax administration. But there are offsetting considerations. First, the grant of a tax holiday almost always requires administrative screening, which can be cumbersome. Second, tax holidays are custom-made for aggressive tax planning to shelter income from other operations of a parent company. Preventing abusive accounting requires vigilant and highly skilled tax administration. To reduce the scope for abuse, firms should be required to file returns during the holiday period; this is also useful for monitoring the respective tax expenditures. Filing during the holiday period is costly because it absorbs administrative skills without directly generating revenue. But if returns are not filed, then difficult transitional problems can arise at the end of the benefit period. Also, many countries have bitter experience with companies that jump in to take advantage of a tax holiday and then shut down at the end of the holiday period. In terms of economic efficiency, tax holidays bias the allocation of resources toward the favored sectors and regions. This can be bad or good depending on the quality of the screening criteria. In addition, tax holidays systematically favor investments that are faststarting, short-lived, and easy to shut down, as well as projects with low capital intensity and low debt costs, which are more likely to generate taxable income during the years covered by the holiday. In contrast, tax holidays are of little value to long-term, capital-intensive projects with long gestation periods, which normally show a tax loss anyway in the early years because of heavy capital allowances.

5-6

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Capital Recovery Incentives--Accelerated Depreciation and Initial Capital Allowances Accelerated depreciation front-loads the capital allowance for tax purposes, relative to the economic rate of capital consumption. The true pattern and rate of depreciation is not generally known, so any designated depreciation schedule is somewhat arbitrary. The most common form is straight-line (SL) depreciation over a plausible number of years for each class of assets. Hussey and Lubick (1996) propose the declining balance method (DBM) as the standard in their model tax code, at annual rates ranging from 5−25 percent.67 A write-off faster than the DBM can be regarded as an incentive, albeit a moderate one because it affects only the timing of cost-recovery, not the amount. Initial capital allowances (ICAs) are special capital write-offs that enhance cost-recovery at the start of a project. The ICA is a percentage of the asset cost that can be written off in the first year (or the first few years). Sometimes the initial write-off involves a corresponding reduction in the basis for depreciation. In other systems the ICA is an extra or additional allowance, over and above full depreciation of the capital asset. This case is equivalent to a subsidy on the purchase price of the capital. The value of an incentive in the form of a capital allowance depends how the exact specification of the tax provision, other aspects of the tax law, and characteristics of each investment. For example, this type of benefit has little value to an investment that would show large tax losses anyway in the early years, unless the losses can be used to offset income from other related sources. Capital allowances therefore favor projects undertaken by existing businesses. Nonetheless, generous investment allowances can strongly reduce the METR on new investments. In particular, an initial allowance of 100 percent—full expensing—can reduce the METR to zero for an equity-financed investment, and less than zero for a debtfinanced project, if interest payments are deductible.68 For most capital-recovery incentives the revenue cost is fairly moderate. Under full expensing, the revenue loss is the tax rate times the difference between the allowable capital cost and the present discounted value of normal depreciation allowances. For large projects this can represent a significant amount of tax money, but it is nonetheless a limited amount of exposure for the Treasury. Capital recovery incentives are relatively simple to administer. However, serious problems can arise if companies abuse the incentives through sham sale and re-purchase of assets, or channeling asset purchases through qualifying companies on behalf of non-qualifying

67 Specifically, they propose annual capital consumption allowances of 5 percent for buildings and structures;

25 percent for automobiles and light trucks, office equipment and computer systems; and 15 percent for all other tangible property. 68 A negative METR means that the after-tax return is higher than the before-tax return. In the case cited here, full expensing reduces the METR to zero and the interest deduction provides a further advantage by lowering the cost of borrowed funds.

5-7

DESIGN OF INVESTMENT TAX INCENTIVES

partners. On this point, Exhibit 5-1 presents an interesting cautionary tale based on experience in India.

Exhibit 5-1 A Cautionary Tale from India The tax code in India has included initial capital

Worse, tax audits uncovered many allowance

allowances since 1945. Since 1955, the investment

claims “which were patently untenable.” Nearly 25

allowance also included a “development rebate” for

percent of the investigated claims were at least

strategic industries, giving a total write-off of 125

partially disallowed.*

percent of capital expenditure. These provisions

In the years after 1990, private corporate

were introduced to offset the disincentive effect on

investment in plant and equipment more than

investment stemming from high income tax rates

doubled as a percentage of GDP. The Government

and the distortion created by lack of an inflation

of India concluded that the “investment allowance

adjustment in computing the basis for depreciation.

was an inefficient method of achieving the objective

The investment allowances were modified

of enabling internal accrual of resources for

many times over the years until 1990, when they

replacement

was discontinued altogether in favor of a reduction

government task force on direct taxes considered

in the basic company tax rate from 52.5 percent to

and rejected appeals by industry for re-introduction

43.2 percent. At the same time depreciation rates

of investment allowance, favoring instead a further

were reduced. Underpinning this decision was a

reduction in the basic company tax rate to 30

detailed empirical study using microeconomic tax

percent.

data, showing that the capital allowance increased

*

and

modernization.”

In

2002

a

The government introduced a special 100 percent

investment by less than 2 percent, while reducing

depreciation deduction for windmills in 1996, as part of a

corporate tax revenue by approximately 13 percent.

new energy policy. One result (perhaps the main one)

Most benefits went to major corporations that did

was to stimulate thousands of false claims for the

not require tax breaks to finance investment.

windmill deduction, which demanded considerable

Furthermore, litigation and political pressure had widened the scope of the incentive program far

administrative resources to challenge. Source: Personal communication with Sebastian James, Department of Revenue, Government of India.

beyond the strategic sectors originally intended. SOURCE: Modi (2003)

Turning to the impact on economic efficiency, the main problem with special investment allowances is that their value is a function of the amount of capital goods being purchased. Consequently, these incentives bias investment decisions in favor of capital-intensive rather than labor-intensive products and processes—which is antithetical to the usual goal of job creation. The incentives also favor short-lived investments and early scrapping of existing capital assets. If the magnitude of the tax break is moderate, then the biases will not be very large. Also, accelerated depreciation can be used to correct (crudely) for a tax bias in the opposite direction caused by inflation, in systems where depreciation is based on historical

5-8

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

cost. In this respect, rapid depreciation (up to a point) serves as a corrective measure rather than an incentive.

Investment Tax Credits Zee, et al. (2001, p.1504) advocate investment tax credits (ITCs) as the preferred form of tax incentive. They note that an ITC is equivalent to an initial allowance (IA) if there is a single company tax rate.69 Hence, the two approaches generally “share the same advantages and shortcomings,” as outlined above. If there are multiple company tax rates, then the ITC is more even-handed than the investment allowance. This is because the IA has greater value to companies that face the higher tax rate, companies that the tax system is designed not to favor. The OECD (2001) cautions that generous up-front incentives, such as investment tax credits or large initial allowances, can endanger revenue by placing the beneficiaries in a large tax-loss position. This makes them attractive candidates for sale to companies with tax obligations, for the purpose of sheltering the latter income from tax.70

Treatment of Dividends To understand how tax policy affects investment decisions, the analysis should examine rates of return from the point of view of the enterprise owners. This requires taking into account the tax treatment of dividend remittances as well as the company tax.71 With a classical tax system, business income is taxed once at the company level and again when remitted to shareholders. The double taxation sharply increases the effective tax rate. It also creates a bias in favor of debt instead of equity financing, and diminishes the incentive to invest. For example, suppose that the company tax rate and the individual income tax rate are both 35 percent. Then the effective tax on company income, after remittance to shareholders, is 58 percent.72 Integrating the company tax and the dividend tax is therefore an important way to improve investment incentives and reduce the tax bias favoring debt.73 The revenue effect of moving to an integrated tax system depends on how much revenue is collected on dividend income. In most SADC countries, this is probably not a major source of revenue. In any case, most economists view the integrated tax system as a normative standard. From this perspective, elimination of the double tax on dividends is an objective of tax reform, not an “incentive” that should be counted as a tax expenditure.

69 The equivalence is: ITC = IA x t, where t is the tax rate. 70 When the United States expanded investment tax credits in the 1981 Economic Recovery Tax Act, one result

was unexpectedly large revenue losses due to this type of tax-loss transaction.

71 The treatment of capital gains also enters the equation here. See the essay on this topic by John King in

Shome (1995, 155-157).

72 Given a pre-tax level of income, I, the double stage tax creates an overall obligation of .35 I + .35 x (.65 I) =

.58 I.

73 Shome (1995, 149-155) contains excellent essays on the arguments for integrating the tax system and methods

for doing so.

DESIGN OF INVESTMENT TAX INCENTIVES

5-9

Extra Deductions Some countries, including several in the SADC region, allow companies deductions greater than 100 percent for certain categories of expenses such as approved training programs, research and development, or export marketing. The extra deduction amounts to having the treasury shoulder a share of the cost via tax relief. In most cases the extra deduction is applicable to all companies, but they can also be part of a selective incentive package, for example in connection with export processing zones. These extra deductions reduce the METR on investment to a minor extent, but their purpose is not to stimulate capital investment as such. Rather, the deduction is intended to strengthen incentives for businesses to promote other policy objectives. The value of the incentive to the beneficiary depends on the marginal tax rate and the size of the extra deduction. For example, Botswana permits companies to deduct 200 percent of qualified training expenses. The benefit has less value to manufacturers, who enjoy preferential tax rates, than to other companies. The revenue impact depends on the same factors, as well as the breadth of applicability and the ability of tax authorities to control abuse of the provision.

INDIRECT TAX INCENTIVES Export Incentives and Export Processing Zones74 The most widespread form of indirect tax benefit for investors is relief from tax on inputs that are used to produce items for export. This vital measure to strengthen incentives for outwardlooking investment is consistent with international norms. Both conventional practice and WTO rules apply the destination principle to indirect taxation of traded goods and services. Put simply, exporters should have access to inputs without paying indirect tax imposed by the country of origin. Eliminating this burden is fully justified and should be viewed as the correction of a tax distortion which would otherwise impede the incentives for efficient export production. The burden of VAT is normally eliminated by zero-rating export sales. The import duty can be eliminated in several different ways. One common approach is to designate certain locations as export processing zones (under various names), or certain facilities as bonded production units. Qualifying producers then obtain inputs and sell outputs as if they were extra-territorial for customs purposes. Import duty relief can also be rendered through duty drawback schemes (also called refunds or rebates), or through provisions to suspend duty on imports by approved exporters, contingent on proof of use for export sales.

74 See survey papers by Radelet (1999) and Madani (1999), the concise discussions in Zee, et al. (2001), and the

excellent summary of benefits and costs of EPZs in UNCTAD (2002).

5-10

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

According to Radelet (1999), every country that has achieved successful export-led growth in the last 30 years has “established and relied heavily on some form of export platform institution to facilitate growth in manufactured exports.” Then again, many other countries have pursued this approach without notable success. In the case of the Dominican Republic and Costa Rica, export platform policies were in place for many years before they began to stimulate a significant response; the policies generated impressive results only after other complementary reforms were well established. In general, success with export platforms (to use Radelet’s term) requires macroeconomic stability, efficient customs operations, reliable utilities and infrastructure, and clear location advantages. Zee, et al. (2001) suggest that the choice of instrument for export promotion should depend on the quality of customs and tax administration. As they put it, “leakage of goods from these zones into the domestic market is usually rampant” in countries with weak or corrupt customs controls. Through this leakage, export platforms can erode revenue while undercutting domestic producers in local markets. The authors caution strongly against introducing duty-free facilities, including EPZs, in such countries, and recommend the use of duty drawbacks instead. Drawbacks have the advantage of collecting duty up front, but they also impose a cash-flow cost on export producers. In addition, they are not simple to administer. For countries with poor customs administration, the main impact of duty drawbacks may be to create red tape and frustration. A totally separate issue is whether exporters should be granted additional tax privileges beyond relief from indirect tax on inputs. Many countries add direct tax incentives for nontraditional exports. In some cases the benefits are extremely generous. One argument in favor of these incentives is that many export activities are “footloose.” When other countries are competing for the same investment, tax incentives may be required to attract the projects. In addition, a plausible case can be made that outward-oriented investments are especially likely to deliver dynamic benefits to the economy through scale economies, technology, training, and competitive pressure for higher productivity. However, the case for additional tax incentives for exporters is not straightforward. Many export enterprises turn out to be enclave activities that generate few benefits for the economy other than a handful of low-skilled jobs. When footloose investments are attracted by temporary benefits such as tax holidays, they may exit quickly once the benefit period ends. In any case, the WTO regards most of these incentives as export subsidies which are prohibited for countries with GNP per capita above $1,000, and will be so in the future for lower income countries.

Reduced Import Duty on Capital Goods and Raw Materials One way to enhance the rate of return and reduce the METR on an investment is to lower the cost of capital by reducing the import duty on imported capital goods. Quite a few countries have reduced these duties to zero. The revenue effect depends partly on whether the low or

DESIGN OF INVESTMENT TAX INCENTIVES

5-11

zero duty rate is offered as a general incentive, or only to qualified investors. In the first case the revenue loss can be very high, especially in developing countries where most capital goods are imported. Cutting the duty on capital goods (relative to other imports) creates significant economic distortions. Most important, it biases resource allocation by favoring capital-intensive investments over those that use labor more intensively. This incentive also hinders the development of a domestic capital goods industry because the benefit is tied to the use of imported capital goods. Domestic machine tool producers often face negative effective protection as a result of this policy because they pay duty on some of their inputs while trying to compete against duty-free imports. Differences in import duty rates create administrative problems as well. Importers have strong incentives to misclassify shipments as capital goods. Besides, many capital goods have dual use as consumer goods; without adequate controls, investors can import the goods duty free and flood the consumer market with products that have escaped tax.75 In cases where the low or zero duty rate applies selectively, the beneficiary may be limited to a pre-approved list of imported capital goods. This approach limits abuse, but at the expense of adding red tape and administrative controls. Some countries also reduce the import duty on raw materials and intermediate goods as an additional investment incentive. If all producers in an industry rely on imported inputs to a similar extent, then the cost reduction from this measure generally accrues to consumers in the form of lower product prices. There is no ultimate improvement in the rate of return for investors. 76 The measure could make a difference if some producers use imported inputs to compete against others who use domestic supplies. But in this case the impact is perverse: the tariff reduction creates a disadvantage for the producers who use domestic supplies—which is probably not what the policymakers had in mind.

Protective Tariffs As shown in Exhibits 3-3 and 3-4, protective tariffs can be highly effective in drawing resources into import-competing production activities. In years past, steep tariff barriers were a common feature of the policy landscape in most developing countries. However, this strategy usually ended up breeding highly inefficient investments, at a high cost to domestic consumers. With few exceptions, highly protective tariffs have proved to be a losing strategy

75 For example, an agricultural investor in Mozambique who was entitled to duty-free capital goods imports

tried in 2001 to bring in hundreds of refrigerators and washing machines, claiming that they were needed for the farm workers. Their claim was denied by customs authorities, but not without public controversy. 76 One often hears that this measure assists liquidity-constrained enterprises by improving their cash flow. There is much less to this argument than meets the eye. Suppose that (a) the tax rate on imported raw materials and intermediate goods is 15 percent; (b) these inputs represent 60 percent of total costs; (c) the nominal interest rate is 30 percent; and (d) the average turnover period is 6 weeks. Then even a total remission of the import duty will reduce costs by just 0.3 percent, relative to the ex-factory product price.

5-12

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

for promoting productive investment and stimulating economic growth. Consequently, most countries have moved in recent years towards lower tariff barriers in order to reduce the economic distortions and foster efficient investment. The revenue effect of reducing duty rates is non-linear. Highly protective duty rates may impair revenue collection because they diminish import volume and create enormous incentives for smuggling and “negotiation” with customs officers. When rates are especially high, say above 40 percent, it is often possible to increase revenue by lowering the customs duties (as Kenya did in the mid-1990s). Further reduction in the maximum duty rate normally entails a substantial revenue loss. To avoid fiscal problems, duty reductions can be phased in and packaged with other measures to enhance revenue (including closing off loopholes by eliminating special exemptions).

5.3 Summary Table 5-1 summarizes the advantages and disadvantages for each tax incentive discussed in this chapter. A simple ranking of the instruments would be inappropriate because the effectiveness and impact of each one is highly dependent on local circumstances, the characteristics of a particular investments, and technical details of the tax code. In addition, every tool has drawbacks as well as virtues, so any ranking involves policy judgments, not just technical analysis. Nevertheless, one may draw a few tentative conclusions about the relative merits of the various instruments: • The soundest tax incentive is a reduction in the overall statutory tax rate to levels that are competitive with most other countries in the region and with rates applying in the main capital-exporting countries. • Relieving exporters of indirect tax on inputs is a must, in order to remove a serious disincentive for export production. The problem lies in devising a practical way to implement and control this export promotion measure in countries with weak tax administration. • Moderate investment tax credits, initial allowances, and accelerated depreciation serve the signaling effect of tax incentives at low cost in terms of the fiscal, economic, and administrative impact. Even these incentives, however, can be abused if not well administered. • The worst form of tax incentive is the imposition of high protective tariffs on competing imports. This may stimulate investment for the domestic market, but it usually turns out to be investment with low productivity and poor development potential. • Eliminating import duties on raw materials and intermediate goods is another poor way to stimulate investment. The measure has a high revenue cost, impairs backward linkages, and has little positive effect on investment for the domestic market since indirect taxes are

DESIGN OF INVESTMENT TAX INCENTIVES

5-13

usually passed along to consumers. (As noted above, mechanisms are needed to eliminate the burden on import duty on inputs used to produce for export.) • Among direct tax tools, tax holidays involving full exemption for a moderate time period are probably the least cost-effective. They favor transitory investment, have only a moderate effect on the METR, and create opportunities for aggressive tax avoidance. • In situations where job creation is a major policy objective, it is very important to bear in mind that many tax incentives work in the opposite direction by favoring capital-intensive investments. • Considerations such as screening criteria and procedures, transparency, and fiscal controls are more important to the ultimate success of a tax incentive program than the exact configuration of the tax provisions. Finally, returning to the starting point for this chapter and the lessons from chapter 2, we reiterate that most investment decisions are driven by non-tax considerations. Even a zero tax regime will not attract much investment if serious deficiencies in the investment climate render most projects fundamentally unprofitable. Concern over tax incentives should not divert attention from other policies and programs that can do much more to improve the returns and reduce the risks for investors.

5-14

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Table 5-1 Evaluation of Major Tax Incentive Instruments Instrument

Effectiveness

Impact on Revenue

Economic Distortions

Administrative Impact

Lower statutory tax rate

Moderate reduction in METR; high redundancy rate likely.

High. Must be packaged with revenue enhancement measures.

Very low. Neutral with respect to sector, product, capital-intensity.

Very low. Simple, transparent and automatic. Reduces incentive for abusive tax planning.

Preferential tax rates

Moderate reduction in METR depending on the size of the reduction; high redundancy rate likely.

High, depending on how widely the preferential tax rates apply.

Moderate. Biases resources allocation, but not factor intensity or other business decisions.

Moderate to high. Requires determining which incomes qualify; wide open for abusive tax planning.

Tax holiday

Moderate reduction in METR; depends on type of investment and screening criteria.

Moderate to high; depends on type of investment and screening process.

High. Favors projects that are short-lived, fast starting, easy to shut down. Also projects with low capital intensity and low debt costs.

Moderate to high. Screening procedures a burden; filing may or may not be required during holiday; problems at transition to taxable status; scope for tax planning.

Capital recovery incentives— accelerated depreciation and initial allowances

Low to high; depends on provisions; full expensing can reduce the METR to zero; if initial allowance does not reduce basis for depreciation, then capital purchase is subsidized.

Low to moderate; depends on provisions.

Moderate. Favors capital-intensive investments and activities and projects with shortlived capital assets. Case of full expensing may eliminate tax wedge for equity-financed investments.

Mixed. Simple to administer, but significant scope for abuse through sham capital transactions.

Investment tax credit

Similar to initial allowance.

Similar to initial allowance.

Similar to initial allowance, but avoids differential impact where there are multiple company tax rates.

Similar to initial allowance.

Integration of company tax and dividend tax

High. Ending double tax on dividend income has a large effect on the METR.

Moderate; depends on importance of dividend tax as revenue source.

Low. Enhances efficiency by ending bias in favor of debt financing.

Moderate; depends on method used.

Extra deductions, as for training, R&D, export marketing

Low to moderate; depends on size of deduction and tax rate.

Low to moderate; depends on size of deduction, the tax rate, and breadth of applicability.

Low. Bias created but in direction of explicit policy objective.

Moderate. Complications in validating legitimacy of deductions; scope for abuse through false claims.

5-15

DESIGN OF INVESTMENT TAX INCENTIVES

Instrument

Effectiveness

Impact on Revenue

Economic Distortions

Administrative Impact

Indirect tax relief for exporters— EPZ, bonded production, duty drawback, duty suspension

Mixed. Many successes and many failures, depending on other aspects of the investment climate.

Low to high. The normative standard is not to collect this revenue; in this sense the provision is not a revenue “loss.” But where duty exemptions are channels for smuggling, the revenue cost can be high.

Low. Measures are efficiency enhancing because they correct an anti-export bias.

High. These export promotion mechanisms demand high quality administration to function smoothly without widespread abuse.

Reduced import duty on capital goods

Moderate. Reduces the METR by lowering the upfront outlay for capital assets.

Low to high; depends on the size of the duty reduction, the structure of import duty revenues, the degree of selectivity.

High. Favors investments that are intensive in imported capital goods, over those that are intensive in labor or domestic capital. Hinders development of domestic capital goods industry.

Moderate. The tariff differential creates incentive for misclassification of imports, and dutyfree import goods also used by consumers.

Reduced import duty on raw materials and intermediate goods

Low to nil. Duties on raw materials and intermediate goods are generally passed through to the consumer.

High. These categories may account for a large share of total imports, and the tax relief is almost entirely redundant.

High. Favors import-intensive industries and producers

Moderate – see above.

Protective tariffs

High. Protective tariffs can have a powerful effect in shielding importcompeting activities

Mixed. Excessive tariffs on competing imports can discourage legal imports and stimulate smuggling, to the point that overall revenue suffers.

High. Protected investments tend to be high cost and uncompetitive, reducing productivity and efficiency.

Moderate – see above.

6. Economics of Harmful Tax Competition Tax competition is intrinsic to a world where capital is geographically mobile and many investments can operate viably in alternative tax jurisdictions. This chapter examines the question of where to draw the line between normal tax differentials reflecting political judgments about economic and fiscal policy, and practices that may be deemed harmful tax competition (HTC). The chapter also examines theoretical considerations that are useful in understanding tax competition, and some lessons from international experience in seeking to develop cooperative approaches for dealing with the adverse effects of tax competition. Three postulates underlie the analysis. First, each sovereign jurisdiction has unquestioned authority to define tax policy in accordance with its objectives and interests. Second, when the tax policies of one state impinge seriously on the interests of others, the latter have the right to seek negotiations to change those harmful policies, and to pursue defensive actions either singly or cooperatively. Third, the benefits to a host country of offering extremely low taxation may be more apparent than real; a better understanding of the economic and fiscal consequences of tax incentives may serve as grounds for cooperation in tax competition.

6.1 What is Harmful Tax Competition? The OECD has been working since 1996 to define “harmful tax practices” and develop defensive measures to mitigate the adverse effects.77 Their main focus has been on tax havens that serve as hubs for financial strategies to escape tax on income from other jurisdictions. The OECD’s main report (1998) isolates “four key factors” as defining characteristics for a tax haven: • No taxes or only minimal taxes. This is “the starting point to classify a jurisdiction as a tax haven.”

77 For further information, see OECD (1998, 2000, 2000a, and 2001a).

6-2

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

• Lack of effective exchange of information with tax authorities in other jurisdictions, particularly in the form of secrecy rules. • Lack of transparency in the legal and administrative provisions of the tax system. • No substantial activities, implying that the system encourages business concerns that are merely shells for tax-driven financial manipulations. In 2000, the OECD used these criteria to identify 35 jurisdictions as tax havens,78 and proceeded to request commitments for cooperation in eliminating the alleged harmful tax practices. In the end, this “cooperation” centered on just two criteria: transparency of the legal and administrative regime, and exchange of tax information with other tax authorities. By mid-2002, seven countries were still listed as “uncooperative tax havens” in terms of these minimal conditions.79 Effectively, the OECD initiative has narrowed into an effort to control illegal tax evasion and money laundering, sidestepping the problem of legal tax avoidance (however rampant) and tax competition (however glaring). Although the OECD report is primarily concerned with tax havens, it also identifies attributes of “harmful preferential tax regimes.” Again there are four key factors, which include the first three from the previous list along with “ring-fencing” of tax preferences (so that the tax breaks are not available to resident taxpayers). In this context the condition of “no substantial activity” is omitted because it is pertinent to tax havens but not to tax competition for real projects. The OECD also lists eight other characteristics of a harmful tax regime, claiming that all of them should be taken into account: 1.

Artificial definition of the tax base

2.

Failure to adhere to international transfer pricing principles

3.

Foreign source income exempt from residence country tax

4.

Negotiable tax rate or tax base

5.

Existence of secrecy provisions

6.

Access to a wide network of tax treaties lacking anti-abuse provisions

7.

Regimes which are promoted as tax minimization vehicles

8.

The regime encourages purely tax-driven operations or arrangements.

The SADC Memorandum of Understanding on Co-operation in Taxation (MOU) uses the term “harmful tax competition” (HTC), rather than tax “practices.” This terminology suggests that SADC’s main concern is with competition for substantive activities. Yet Article 1 of the MOU defines HTC as

78 Mauritius was also identified initially as a tax haven, but not included in the list because it committed ahead

of the deadline to eliminating harmful practices. The Seychelles subsequently signed a letter of cooperation as well. 79 Andorra, Liechtenstein, Liberia, Monaco, Marshall Islands, Nauru, and Vanuatu.

ECONOMICS OF HARMFUL TAX COMPETITION

6-3

a situation where the tax systems of a jurisdiction are designed in such a way that they erode the tax bases of other jurisdictions and attract investments or savings originating elsewhere, facilitating the avoidance of taxes in other jurisdictions.

This definition emphasizes the adverse fiscal effect of HTC rather than displacement of real economic activity. Indeed, the clear reference to tax avoidance, taken literally, excludes the impact of tax differentials on the physical location of substantive activity.80 Article 4.3 of the MOU goes on to say that Member States will, in the treatment and application of tax incentives, endeavor to avoid: (a) Harmful tax competition as may be evidenced by: (i)

zero or low effective rates of tax;

(ii)

lack of transparency;

(iii)

lack of effective exchange of information

(iv)

restricting tax incentives to particular tax payers, usually nonresidents;

(v)

promotion of tax incentives as vehicles for tax minimization; or

(vi)

the absence of substantial activity in the jurisdiction to qualify for a tax incentive;

(b) Introducing tax legislation that prejudices another Member State’s economic policies, activities, or the regional mobility of goods, services, capital or labour.

The conditions listed in section 4.3(a) as evidence for HTC include all five key factors cited by the OECD, plus one of the “additional characteristics” of a harmful tax regime: promotion as a vehicle for tax minimization. Of the other seven OECD criteria, three may well be inappropriate for SADC countries because of administrative constraints or special circumstances in low income countries.81 In principle, the remaining four criteria—artificial definition of the tax base; negotiable tax rate or base; existence of secrecy provisions; and encouragement of purely tax-driven operations—are just as applicable in the SADC region as they would be elsewhere. Section 4.3(b) clearly pertains to harmful effects of tax competition. Where tax incentives divert real capital investment from the location where productivity is highest, they distort the allocation of resources and reduce overall efficiency for the region. This inefficiency is a legitimate source of concern for SADC. But without a clear definition of “prejudice,” this provision is too broad to serve as an operational basis for action. One could argue that “prejudice” arises any time an investor chooses country A over country B, where A offers more attractive tax rates (even if those rates are not unusually low). Until such time as the

80 If MegaCorp shifts production from country A to country B because of lower taxes, then its income falls

under the jurisdiction of B. Since the taxable income no longer arises in A, one cannot say that MegaCorp is engaged in avoidance of tax in A. 81 The three points are international transfer pricing principles; exempting foreign source income; and wide network of tax treaties.

6-4

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Community is prepared to pursue full tax harmonization, Member States cannot be expected to “avoid” all such “prejudice.” Is there a more operational way to deal with the real problem of inefficient investment diversions due to tax competition in the region? This question can best be answered after examining some pertinent theoretical concepts.

6.2 Theory of Tax Competition82 A starting point for examining the theory of tax competition is a pioneering model of interjurisdictional tax differentials known as the Tiebout hypothesis. For a decentralized fiscal system, Tiebout postulated that tax competition promotes efficiency by allowing states or localities to cater to different tastes of citizens for taxation and the provision of public goods. Some jurisdictions will choose low taxes and limited public services, while others choose high taxes and more extensive public services. Within a union, individuals can then “vote with their feet” by moving to a jurisdiction that suits their preferences. In the Tiebout model, tax differentials are fundamentally benign. In the international context, individuals cannot so easily move to a compatible fiscal regime. Yet the model aptly describes the fact that some governments choose to restrain the tax burden while others adopt high-tax, high-expenditure outcomes to suit the preferences of their constituents. Equally, jurisdictions differ legitimately in their preference regarding the structure of the tax system. Some countries depend more heavily on consumption-based taxes, while others place more weight on income taxes. Moreover, inter-jurisdictional competition can be a positive force for encouraging tax reform and improving efficiency in the administration of public services. Although it is useful to recognize the healthy side of tax competition, this view obviously omits considerations that lie at the heart of the debate about harmful tax practices. The decision by one state to pursue a low-tax strategy creates fiscal externalities for other jurisdictions. These include83 • Factor movements to the low-tax state (LTS), which erode the tax base in the high-tax state (HTS). Even if HTS offers better public services overall, the additional expenditure may fall in areas (such as social services) that appeal to voters but not to potential investors. If so, the service differentials will not compensate companies for the higher tax rate. Thus, investment, jobs, and taxable income will shift to LTS.

82 This section draws on the survey by Lemgruber (1999). 83 This is only a partial list. Gordon (1983) identifies six inter-jurisdictional tax interactions in a non-cooperative

equilibrium, including the first two above plus equity effects, congestion effects, effects from changes in consumption patterns, changes in the cost of public goods, and changes in the terms of trade (relative prices) between jurisdictions.

6-5

ECONOMICS OF HARMFUL TAX COMPETITION

• The fiscal regime in LTS can constrain HTS to reduce its own tax rates in order to stem the loss of capital. Even if the tax base is fixed, revenues will be lower than in the unconstrained political equilibrium and the supply of public services will be suboptimal. Alternatively, HTS could maintain the level of public services by increasing other taxes, but this approach creates adverse effects in terms of equity and efficiency. • Companies that operate in both states can arbitrage the tax differential by using intracompany transactions and charges to shift income into LTS, at the expense of HTS. Thus, competition from the LTS causes a loss of fiscal resources for the HTS and for the regional fiscal system as a whole. The problem becomes more acute when one takes into account strategic behavior. Here, the classic model is the “prisoner’s dilemma” game, where two players make rational strategic choices which lead to an outcome that is sub-optimal for both in the absence of co-operation. Table 6-1 shows how the logic of the prisoner’s dilemma can affect the decision on tax incentive.

Table 6-1 Prisoner’s Dilemma and Tax Incentives State A, Standard Tax Policy State B, Standard Tax Policy

State B, Tax Concessions

Cell 1

Outcome

State A, Tax Concessions Cell 2

Outcome

A

B .

A

B .

AETR

35%

35%

AETR

25%

35%

Tax base

100

100

Tax base

150

50

Tax revenue

35

35

Cell 3

Outcome A

AETR Tax base Tax revenue

35% 50 17.5

Tax revenue

37.5

Cell 4

Outcome

B .

17.5

A

B .

25%

AETR

25%

25%

150

Tax base

100

100

37.5

Tax revenue

25

25

Notes AETR = Average effective tax rate = tax collected per unit of tax base, taking into account incentives. In this example, each player has a myopic incentive to choose the tax concession policy, regardless of which policy option it expects the other player to select. This myopic incentive lands them in cell 4 (tax concessions), which is a worse outcome for both of them than cell 1 (standard tax policy).

In this simplified example, each state faces a choice of structuring its tax system so that the average effective tax rate is either 35 percent (standard tax policy) or 25 percent (tax concessions). When both states set the same rate, they share equally in investment, and each has a tax base of 100. If just one state sets AETR at 25 percent, then it gets most of the

6-6

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

investment. The tax base in the low-tax state jumps to 150 at the expense of the high-tax state. In the absence of cooperation, the dominant strategy for each state is AETR = 25 percent. Why? Here is the logic from the point of view of state A: Suppose that B chooses the 35% tax; then we are better off setting our tax at 25%. Now suppose that B chooses the 25% tax; we are again better off (in the sense of avoiding a larger loss) by setting the tax at 25%. Either way, 25% is our best choice.

The logic works the same way for state B. Hence, both states are driven to adopt the 25 percent tax rate, leading to the outcome in cell 4. They wind up with the same investments as in cell 1, but with fewer fiscal resources to finance the provision of public services. Once they settle into the low-revenue outcome in cell 4, the identical logic applies to further tax cuts. In the absence of cooperation, each state, acting rationally, has an incentive to reduce the tax to 20 percent. Then to 15 percent. This is the famous “race to the bottom” in tax competition. This adverse result is not inevitable. One way out is cooperation. If A and B agree to hold their tax rates at 35 percent, then they mutually benefit by ending up in cell 1 instead of cell 4. In a dynamic framework there is a chance for implicit cooperation, if forward-looking states recognize that “beggar thy neighbor” competition is a losing game in the long run. The logic of the prisoner’s dilemma is not a theoretical anomaly. It explains why common resources get overused and why cartels break down. More to the point, the model shows that cooperation to fight HTC can be of mutual benefit to all of the participating states. The problem is that even if all of the states in a group like SADC agree to cooperate, countries elsewhere are still pursuing tax practices that divert investment from the region as a whole. Another consideration is that fiscal competition does not necessarily fall into the prisoner’s dilemma pattern. If total investment is highly responsive to the tax rate, then competitive tax cuts can stimulate investment in both countries rather than simply re-shuffling it between the jurisdictions. In Table 6-1, suppose that cutting tax rates from 35−25 percent causes overall investment to rise by half. Then cell 4 would leave both countries with a tax base of 150 and higher revenue due to the tax cuts. In most instances, of course, tax cuts do cause a loss of revenue and create a need for compensating fiscal adjustments.84 Another case where the prisoner’s dilemma would not arise is where investment is highly insensitive to tax differentials. If this is known by both states, then neither has an incentive to cut its effective tax rate to 25 percent, simply because the tax cut will not be beneficial. If one state chooses to do so in response to political pressure or poor policy analysis, the other state would have no incentive to follow suit. For example, Zambia resisted political pressure to

84 Oxfam (2003) attempts to come up with a crude estimate of the revenues foregone by developing countries

due to vying for foreign direct investment through tax relief. Using a 35 percent tax rate as the benchmark, the study estimates that the overall revenue loss to host countries is $50 billion, roughly equal to total of foreign aid flows to developing countries.

6-7

ECONOMICS OF HARMFUL TAX COMPETITION

match a sharp tax cut implemented in Tanzania, because the government’s analysis indicated that the cost would outweigh the benefit (Exhibit 6-1).

Exhibit 6-1 Resisting Tax Competition from Next Door, Zambia, 1963 In December, 1991, a Presidential Commission in

were 15 percent and 20 percent. Tanzania also

Tanzania recommended sweeping tax reforms to

implemented much deeper cuts in excise tax rates.

enhance revenue performance, widen the tax base,

Upon hearing about the Tanzanian budget,

and promote economic efficiency. Some reforms

business leaders in Zambia appealed to the Minister

were implemented in the 1992/93 budget, including

of Finance to match the zero duty and sales tax on

sharp cuts in tax and duty rates. According to the

capital goods and raw materials. They contended

budget program, the revenue loss would be more

that Tanzania did this to help domestic producers,

than offset by improved compliance and tax

so Zambia should do the same. Economists in the

administration,

certain

Budget Office ran the proposal through a revenue

petroleum

estimation model based on import records and

products. But the budget omitted some important

found that the measure would cost nearly one-sixth

revenue

the

of total revenue. They replied to the Minister

Commission had recommended as part of the

saying: Zambia cannot afford these tax cuts without

package.

jeopardizing

exemptions,

and

the

elimination

higher

enhancement

Meanwhile,

Zambia

taxes

of on

measures

was

that

pursuing

a

the

macroeconomic

stabilization

program.

comprehensive tax reform program based on

As it turned out, Tanzania couldn’t afford

recommendations of a public-private Tax Policy

them either. The 1992/93 tax package led to a fiscal

Task Force. Zambia’s 1993 budget included cuts in

crisis, as revenue dropped to 11.5 percent of GDP,

income tax and import duty rates comparable to

from 21 percent the previous year. Much of the

those in Tanzania. One major difference was that

decline was due to falling revenue from sales and

Tanzania set a zero rate of customs duty and sales

excise taxes and import duties. The zero-rated

tax on capital goods and raw materials for use by

import provisions not only caused revenue losses

local

directly, but also created new avenues for tax

industries

to

stimulate

production and

investment. The corresponding rates in Zambia

evasion. Zambia did well to resist the pressure.

SOURCE: Based on Fjeldstad (1995), Ministry of Finance, Zambia (1992) and author’s personal experience in the Ministry of Finance in Zambia, 1993.

Another theoretical insight that strengthens the case for cooperation against tax competition is the concept of the “winner’s curse.” In any competitive bidding process there is a systematic tendency for the highest bidder to pay too much. Put simply, the highest bid tends to come from the contender with the most optimistic appraisal of the object of the bidding—whether it is a painting, a license for oil rights, or a footloose investment project. Applying this concept to the case of tax competition for investment, the “true” value of the project to the host economy is inherently uncertain. The most attractive tax breaks are likely to come from a jurisdiction that is overestimating the benefits. As a result, the net gain to the

6-8

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

economy from “winning” the project may be small or even negative. The tendency to give away the kitchen sink is especially pronounced if policymakers have an exaggerated idea of the benefits and a limited understanding of the costs of tax incentives. Moreover, there are strong political pressures to win projects even when the cost is excessively high, particularly if the adverse effects are hidden or deferred. There is no shortage of evidence that inter-jurisdictional competition for investment often leads to outcomes of dubious value to the “winner.” For example, UNCTAD (2002, 205) lists over a dozen international showcase projects that leveraged fiscal competition to obtain enormous incentive packages. The incentives alone are estimated to have cost more than $340,000 per job created in a Mercedes plant in Brazil, and $420,000 per job in a Ford plant in India. Closer to home, the Ramatex textile factory sought concessions from several countries in the SADC region before accepting a very generous package of incentives from Namibia.85 Even within Namibia questions have been raised about the value of the project to the country that won the competition (see Exhibit 6-2). The main point of this theoretical excursion is that cooperation to stem HTC can serve the interests of the countries that are asked to “concede” to reductions in fiscal incentives as part of a regional tax agreement, as well as the countries that feel “prejudiced” by tax breaks given elsewhere. This observation can be the fulcrum for negotiating a cooperative agreement within SADC.

6.3 Can Jurisdictions Cooperate in Mitigating Harmful Tax Competition? Tax competition is here to stay as a feature of the international economy. Sovereign jurisdictions inevitably make different political decisions about the level and structure of taxes. The resulting tax differentials inherently create competition for internationally mobile capital (and labor). Tax competition can be a healthy inducement for improving efficiency in revenue systems throughout the world and expanding global investment. But this competition also generates harmful fiscal externalities from capital migration. Uncoordinated tax competition (the “race to the bottom”) also creates strategic choice problems that diminish fiscal resources in poor countries and rich countries alike. Experience in many parts of the world suggests that it is very difficult to develop effective cooperative approaches among sovereign jurisdictions to mitigate the harmful effects of tax competition. Within the United States, for example, there has long been concern about the cost

85 Not enough data is available to compute the cost of incentives per job created in this case, but from the

information given in Exhibit 6.2, the subsidy for site preparation alone amounted to $20,000 per job created as of April 2003.

6-9

ECONOMICS OF HARMFUL TAX COMPETITION

of competition among the states in offering costly packages of fiscal incentives to lure investors. The incentives include tax abatements as well as subsidies, soft loans, and targeted

Exhibit 6-2 Shopping for Tax Incentives: The Case of Ramatex Ramatex, a textile manufacturing company based in

free water and electricity service imposed a

Malaysia, was seeking to locate a new production

considerable cost on citizens of Windhoek. Water

facility in southern Africa to benefit from the Africa

usage

Growth and Opportunity Act, which allows for

periodic water shortages, and the municipal

duty free exports to the United States. With a

government has been under financial strain to cover

promised investment of R1Billion (1US$=8.12R), the

the cost of water, electricity, and sewage for the

company sought investment incentives in South

Ramatex site. Also, environmental concerns have

Africa, Madagascar, and Botswana, finally deciding

been raised in the National Assembly and the

to invest in Namibia despite a generous offer by

Namibian Economic Policy Research Unit.

South Africa.

by

Ramatex

reportedly

contributes

to

Based on information to date, it is not clear if

The Government of Namibia devised an

Namibia gained from the offer of generous

incentive package that included a 20-year tax

incentives to attract Ramatex. Indeed, when the

holiday,

charges,

company requested additional land to expand its

subsidized rates for water and electricity, a 99-year

plant, on terms similar to those of the initial

tax exemption on land use, and N$60 million

incentive package, the Windhoek Municipality

(1US$=8.12N$) to prepare the site, including the

turned down the request.

an

installation of

exemption

electricity,

from

wharf

water,

and

sewage

Suppose that Namibia, South Africa, and other

infrastructure. The government anticipated that the

Member States had been bound by an agreement

investment would create 10,000 jobs, boost cotton

restraining tax competition. Would Ramatex have

production, and encourage ancillary industries.

located the factory in the SADC country with the

Production started in middle of 2002. By April 2003,

most efficient fundamentals? Or would it have gone

3,000 jobs had been created.

instead to another AGOA-eligible country outside

According to local press reports, the offer of

SADC, which offered more generous tax breaks?

SOURCE: Background paper by James (2003b) based on news reports from Namibia and Economist Intelligence Unit quarterly reports.

infrastructure and training support. The common conclusion is that the states lack adequate incentives to cooperate in reducing the cost of this competition.86 In Brazil, too, efforts to mitigate competition among subnational jurisdictions has generated more hot air than action (Lemgruber 1999). Even the limited initiative by the OECD to deal with harmful tax practices did not obtain a consensus, as Switzerland abstained and Luxemburg openly disagreed with the findings. The EU has made more progress in this area, but only after decades of

86 For example, see the conference report by the Federal Reserve Bank of Minneapolis (1996).

6-10

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

movement toward economic integration, including fiscal transfers to poorer states and regions. On the other side of the bargaining table, major corporations make a point of playing countries or localities against one another to obtain maximum fiscal concessions, often in the form of tax breaks. In doing so, the companies move toward being international free riders, using the institutional and physical infrastructure of the host countries while letting others bear the cost. When a large investor knocks on the door, though, it is difficult for politicians to refrain from entering the bidding war, especially if they have discretion over the provision of incentives. Vocal and well organized pressure groups from the business community constantly appeal to governments to offer special tax incentives, and to avoid international agreements that would limit such tax breaks.87 In short, there are powerful obstacles to international cooperation to mitigate the harmful effects of tax competition. Nonetheless, the OECD and the EU have shown that constructive steps can be taken.

6.4 Implications for SADC Three dimensions of tax competition affect the SADC region: tax competition between SADC states and the rest of the world; tax competition within SADC; and options for regional cooperation to mitigate the adverse effects of HTC on revenue and real resource allocation within the region, including special consideration to the needs of the least developed Member States.

SADC AND THE REST OF THE WORLD Tax competition between SADC countries and the rest of the world cuts both ways. Some SADC countries offer aggressive tax incentives to attract investment in export activities or global business centers. For EPZ enterprises in Malawi, Namibia, and Zambia, and for offshore businesses in Mauritius (some sectors), the governments offer regimes free of all or most taxes. Several other member states, while not granting full exemptions, also offer fairly generous incentive packages. At the same time, SADC member countries are vulnerable to aggressive tax breaks offered by other countries which compete for footloose investments, as well as the use of outside tax havens by sophisticated domestic entities, including SADC-based operations of transnational corporations. For example, UNCTAD’s Investment Policy Review for Lesotho (2003) reported

87 Most websites that turn up on an internet search for “harmful tax competition” are business lobby groups

opposing any international “tax cartel” to curb access to tax havens.

ECONOMICS OF HARMFUL TAX COMPETITION

6-11

that the country’s important garment sector has paid little tax because of income-shifting strategies rather than fundamentally weak profits. Where such practices are rampant, the domestic tax rate matters little. Few of the SADC countries have strong statutory measures and administrative capacity to deal with abusive transfer pricing, thin capitalization schemes, tax dodges through controlled foreign entities (CFEs), or other forms of aggressive avoidance through international channels. On CFEs, only 5 Member States—Lesotho, Mauritius, Namibia, South Africa and Tanzania— have adopted residence-based tax systems that could bring this income into the domestic tax net. Regional and international cooperation could be very useful in strengthening both the legal provisions and audit systems that are needed to fight potentially huge revenue losses through international tax planning.

COMPETITION WITHIN SADC Within the SADC region, a few cases of head-to-head tax competition have been widely reported, such as the Ramatex example in Exhibit 6-2.88 Undoubtedly many other investments, which do not make the headlines, have crossed borders to take advantage of lower taxes or weaker tax administration. At the policy level, some tax incentive programs have explicitly been introduced as a response to tax competition from other countries in the region and elsewhere. For example, the recent introduction of full tax exemptions for EPZ enterprises in Zambia was motivated in part by a concern to match incentives available in other AGOA-eligible countries.89 In some instances, tax incentives may even have been motivated by a desire to undercut tax rates in a neighboring country. Nonetheless, most tax incentive programs in the SADC region appear to be driven by a legitimate desire to create an attractive investment environment, rather than a deliberate intent to deprive other Member States of investment and revenue. Whatever the motivation for tax competition in SADC, the most important question is whether the effects on regional development are positive or negative. In theory, tax competition could promote development of the entire region—despite the adverse side-effects of tax incentives—if the incentives attract a large amount of new investment. In this scenario there is little reason to pursue restraints on tax competition. But this is not a realistic scenario. The preponderance of evidence (as reviewed in chapters 2 and 3) indicates that tax incentives rarely induce a strong investment response, whereas they do impair revenue mobilization and reduce economic efficiency. Hence, the operational premise is that unrestrained tax

88 Another widely cited example is the case of the Hyundai factory in Botswana, which went into liquidation in

2000. The basic cause was mismanagement and probable fraud, but the company’s demise was precipitated by a decision by South Africa to get tough on what it viewed as tax abuse. Sources: James (2003) and Kebonang (2001). 89 According to a well-informed source in Zambia (in private communication) the initial beneficiaries of the new EPZ regime in Zambia have been companies that were already in existence and paying taxes.

6-12

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

competition is a negative-sum game, which damages development prospects for the region as a whole. If tax competition is a negative-sum game, then a well designed agreement to coordinate incentives can be a win-win arrangement for all Member States. There are two ways this situation can arise: • First, unrestrained tax competition can provoke a “race to the bottom” that impairs development in every country. This may sound unlikely, but it can easily occur as a consequence of the prisoner’s dilemma logic, combined with widespread misperception of the benefits and costs of tax incentives (which feeds the winner’s curse), and political pressure from interest groups that benefit from the incentives. Just as political dynamics often create a tendency for approving budget programs with unsustainable deficits, these common factors create a tendency for jurisdictions to offer tax incentives that harm their own interests, as well as those of the region. In this case, the mitigation of tax competition benefits all Member States. • Second, tax competition may yield winners and losers, even though the net impact is negative for the region. Some states may benefit from offering generous incentives, while the fiscal and economic costs of tax competition impair development prospects in other states. As sovereign states, winners from the status quo will not sacrifice their national interest for the good of the community at large. In this case, an acceptable agreement would require hard bargaining over provisions to compensate the winners for accepting measures to mitigate the tax competition. It is difficult to distinguish between these two cases in practice due to limited information about the costs and benefits of most tax incentive regimes. Indeed, it is quite possible that most countries perceive that they are benefiting from their tax incentive policies, even when they are not. As discussed in chapter 3, decisions on tax incentive policies are often driven by an exaggerated view of the benefits and a serious underestimate of the costs. Improved data systems and solid policy analysis are thus fundamental in achieving more efficient strategies for investment promotion and an effective regional agreement on tax cooperation.

OPTIONS FOR COOPERATION Even though a well designed agreement on tax cooperation can be a win-win strategy for all Member States (compared to the alternative of unrestrained tax competition), reaching an accord is inherently difficult. The usual starting point is a deep distrust of any measures that may compromise state sovereignty, and a strong desire by each country to maximize its share of the benefits. In addition, the same factors that create a tendency for excess tax incentives also work against achieving a cooperative solution.

ECONOMICS OF HARMFUL TAX COMPETITION

6-13

Fundamental differences in political preferences and economic conditions, as well as legitimate concerns about tax competition from outside the region, further complicate bargaining. These issues have particular importance for the poorest Member States, which most need to attract additional investment. Since the entire community has an interest in helping the poorest members to develop, there is widespread acceptance that a viable agreement should include special consideration for these states, allowing them more latitude in structuring their incentive programs. It is important to reiterate, however, that aggressive tax incentives are probably not the best way to achieve these vital development objectives. Despite the difficulties in negotiating an agreement, there are many ways for SADC as a group to pursue regional approaches for mitigating the adverse effects of harmful tax competition. These include 1.

Adopting standards of openness and transparency in the disclosure of investment tax incentives, not only to facilitate regional cooperation, but also to promote domestic accountability. As a starting point, SADC should give serious consideration to adopting the IMF Code of Fiscal Transparency, or a modified version thereof, as a regional standard.

2.

Agreeing to negotiate bilateral or multilateral protocols or treaties for the effective exchange of tax information, including coordination on tax audits and investigation of tax crimes.

3.

Agreeing to introduce tax expenditure budgeting or similar tools to improve the monitoring and control of the fiscal costs of tax incentives, when and where information technology systems will permit.

4.

Agreeing to a periodic joint review of existing tax incentive programs to evaluate compliance with any SADC agreement on tax cooperation, and also to help every Member State avoid the winner’s curse by providing (non-binding) recommendations to improve the cost-effectiveness of their programs.

5.

Establishing a joint SADC training program to improve the quality of the tax policy analysis, and educational programs to help government officials and other stakeholders better understand the economics of investment tax incentives. Better information is essential for improving the cost-effectiveness of many investment incentive programs.

6.

Participating proactively in international forums dealing with harmful tax practices.

7.

Supporting the concept of an International Tax Organization with a mandate to evaluate and facilitate international responses to harmful tax practices.

8.

Adopting clear and effective legal provisions on transfer pricing, thin capitalization, controlled foreign entities.

6-14

9.

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Jointly pursuing defensive actions to minimize abusive tax avoidance schemes using cross-border tax differentials.

10. Developing programs for regional fiscal transfers, similar to the structural funding arrangement in the European Community, to help poorer Member States pursue more constructive approaches to improving their investment climate, in lieu of aggressive tax concessions. 11. Finally, any regional agreement to mitigate the harmful effects of tax competition requires an effective mechanism for resolving disputes and enforcing remedial actions in response to violations. Otherwise, it will be difficult to sustain cooperation because the logic of the prisoner’s dilemma creates incentives for each state to deviate from the accords, to the detriment of overall economic development in the region.90

90 An analysis of dispute and enforcement mechanisms is beyond the scope of the present study.

7. Tax Systems and Tax Incentives in the SADC Region This chapter provides an overview of investment tax incentives throughout the SADC region, based on country-specific information from the SADC tax database and other supplementary data sources.91 At the outset, it is worth reiterating an important lesson from preceding chapters: that tax considerations are not the main determinant of viability for most investment projects. Generous tax incentives rarely stimulate a substantial investment response where the basic climate for doing business is seriously deficient due to political and macroeconomic instability, infrastructure problems, unreliable legal and judicial systems, or inefficient public services. Incentives can even have a detrimental impact on development by impairing revenue mobilization, eroding fiscal capacity to address fundamental problems, and distorting the allocation of investment resources. The present chapter starts with a brief discussion in section 7.1 of the fiscal situation in the SADC countries. Section 7.2 examines the basic company tax structure and presents illustrative calculations of the Marginal Effective Tax Rate (METR) for all 14 member countries. Section 7.3 describes the widespread use of export promotion schemes, tax holidays, differential tax rates, special capital allowances, investment tax credits, and other special deductions. Section 7.4 discusses some successes and failures with investment tax incentives in the region. Section 7.5 examines the framework for tax incentives, including the objectives, procedures, and the need for fiscal transparency. A table at the end of the chapter presents a comparison of key provisions of the various tax systems; more details on the main tax incentive policies of each country are provided in the appendix.

7.1 Revenue Performance Sustainable revenue mobilization is a paramount consideration in any analysis of tax policy. Concerns about revenue were explicitly addressed in Mozambique’s recent decision to scale

91 The SADC database used here is from August, 2003. The supplementary sources are listed in the notes at the

end of Table 7-5 below and in the Appendix.

7-2

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

back tax incentives as part of its tax reform program in 2002. Similarly, Lesotho turned away from tax holidays after experiencing substantial abuse. The government of Tanzania has been studying whether current incentive programs have been impairing revenue performance. In Namibia a study by a major policy institute in 2003 suggested a link between tax incentives and weak company tax revenues, while a second study in 2002 concluded that revenue foregone from tax incentives made it difficult to lower the general company tax rate.92 However, the tide flows both ways. Zambia, for example, introduced new tax benefits in 2003 for companies in export processing zones. SADC countries differ greatly in terms of their revenue performance and their sensitivity to revenue risks that would arise if tax incentives are poorly designed, inadequately targeted, or widely abused. Table 7-1 presents selected fiscal indicators for 1999-2001 (this is the latest three-year period for which complete data are available). The figures show that DCR, Tanzania, and Mozambique have been facing very low revenue flows, with taxes yielding no more than 12 percent of GDP. Other states are mobilizing large revenue flows. Angola tops the list at 46 percent of GDP, while in Botswana, Lesotho, Namibia, Seychelles, Swaziland, and Zimbabwe revenue collections exceed of 25 percent of GDP (inclusive of SACU fiscal transfers, for SACU members). Yet even some of the high-revenue countries have nonetheless been incurring large budget deficits. In such cases, expenditure control is the first place to look to restore fiscal balance, but protecting revenue is a vital policy issue. Four SADC countries—Botswana, Namibia, South Africa, and Swaziland—enjoy a solid fiscal position of strong revenue yield, sustainable budget balance, and low dependence on foreign aid. In other SADC countries, the risk of revenue loss from tax incentives must be taken seriously. Revenue loss from any source has a high opportunity cost in terms of foregone public services and development programs, or a greater tax burden on other taxpayers.

7.2 Standard Tax Rates Having in place a basic tax structure that is reasonably competitive is at least as important as offering special incentives. Figure 7-1 shows the standard statutory tax rates in each country, excluding tax incentives. The figure includes the company tax, the dividend tax on payouts to resident shareholders, and the combined effect of both levels of taxation (assuming that aftertax profits are fully remitted). The standard statutory rate of company tax clusters at 30–35 percent, with a simple average of 33 percent. Botswana and Mauritius have the lowest standard tax rates in the region, at 25 percent, whereas the Seychelles and DRC have high rates of 40 percent. The tax rate on dividend distributions to residents clusters around 15 percent. On the low end, Mauritius,

92 See IPPR Opinion No. 8, April 2003, and the report of the Namibian Tax Consortium, December 2002.

7-3

TAX SYSTEMS AND INCENTIVES IN THE SADC REGION

Namibia, and Seychelles do not tax these dividends, while Malawi and Swaziland charge 10 percent. On the high end, Zimbabwe and DRC impose a 20 percent tax, while in Zambia the rate is 25 percent. Every SADC country uses withholding to collect tax on dividends, except Swaziland, which taxes dividends in the hands of the recipients. With the exception of Botswana, the combined statutory tax rate is derived from the formula: TT = TC + TD * (1 – TC), where TT, TC and TD are the total, company, and dividend tax rates, respectively. Botswana has an integrated tax system designed to avoid double taxation of dividend income.93 The company tax rate of 25 percent consists of two elements: a 15 percent standard rate plus 10 percent additional company tax (ACT). The payment of ACT is offset

Table 7-1 SADC Fiscal Indicators, Three- year Averages, 1999-2001(as % of GDP)

Country

Tax Revenue

Total Revenue (excluding grants)

Budget Balance (excluding grants)

Budget Balance (including grants)

Net ODA from all sources

Angola

46.3

46.6

-17.4

-7.7

4.2

Botswana

35.4

42.6

3.3

3.7

1.1

DR Congo

2.4

5.1

-2.3

-2.3

0.8

Lesotho

32.4

40.6

-8.8

-6.5

4.8

Malawi

16.0

17.8

-13.1

-5.6

24.6

Mauritius

17.3

19.8

-4.7

-4.6

0.6

Mozambique

11.4

12.5

-15.7

-3.6

23.0

Namibia

30.3

33.1

-3.6

-3.2

4.4

Seychelles

34.8

43.6

-13.2

-12.1

2.5

South Africa

23.7

24.2

-1.9

-1.9

0.4

Swaziland

26.3

28.0

-2.9

-2.0

1.8

Tanzania

9.7

10.9

-4.3

-0.5

12.0

Zambia

17.7

18.7

-11.3

-4.9

18.2

Zimbabwe

24.7

26.5

-14.9

-14.1

2.9

SOURCE: World Bank Africa Database 2003 CD-ROM

93 Malawi had an integrated tax system until the 2001/02 budget. The system worked through a “dividend tax

account,” which was widely considered to be too complicated. Also, companies enjoying tax holidays objected to facing a dividend tax on income that was untaxed at the company level. The new system is simpler, but it increases the overall tax on dividend payouts for companies that are not receiving holidays as priority industries or exemptions as EPZ enterprises.

7-4

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Figure 7-1 Standard Company and Dividend Tax Rates, SADC Member Countries, 2003

(a) Standard Company Tax Rates SADC Average MAURITIUS BOTSWANA ZIMBABWE TANZANIA SWAZILAND SOUTH AFRICA MALAWI ZAMBIA NAMIBIA MOZAMBIQUE LESOTHO ANGOLA SEYCHELLES DR CONGO

0%

10%

20%

30%

40%

50%

(b) Standard Dividend Tax Rates (case: resident shareholders) SADC Average SEYCHELLES NAMIBIA MAURITIUS SWAZILAND MALAWI SOUTH AFRICA TANZANIA LESOTHO BOTSWANA ANGOLA MOZAMBIQUE ZIMBABWE DR CONGO ZAMBIA

0.0%

zero zero zero

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

(c) Combined: Company & Dividend Tax (case: resident shareholders) SADC Average MAURITIUS BOTSWANA NAMIBIA SWAZILAND MALAWI SOUTH AFRICA SEYCHELLES TANZANIA ZIMBABWE LESOTHO ANGOLA MOZAMBIQUE ZAMBIA DR CONGO

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

TAX SYSTEMS AND INCENTIVES IN THE SADC REGION

7-5

against the 15 percent withholding tax on dividends. For example, a company with taxable income of P100 is subject to a basic tax of P15 and ACT of P10, for a total of P25. Any dividend tax due up to P10 is offset by the ACT. Thus, the ACT eliminates tax on dividends for payouts up to P67. Suppose, though, that after-tax profits of P75 are fully distributed. In this case the 15 percent dividend tax of P11.25 exceeds the ACT; thus the company owes dividend tax of P1.25, which increases the total tax rate to 26.25 percent. The point of the offset arrangement is to ensure that income would not go untaxed if a company were to report zero income for tax purposes while paying out profits to shareholders. In this case, the company tax would be zero but the dividend withholding tax would be 15 percent (since the ACT offset is zero). Statutory tax rates tell only part of the story, because the tax burden facing an investor depends on many other elements of the tax system, as well as the type of investment itself (not to mention the quality of tax administration). Recall from chapter 4 that the marginal effective tax rate (METR) can be used to gauge the extent to which the tax system affects the incentive to invest. In essence, the METR shows the extent to which the tax system reduces the rate of return on investment, taking into account not only the basic tax rates but also many technical features of the tax system such as capital allowances, loss carry-forward provisions, and the capital gains tax on disposal of capital assets. Table 7-2 presents a comparison of the METR for each SADC country under the standard tax system (including withholding tax on dividend payments to non-resident shareholders), using four illustrative project scenarios. The scenarios differ according to the composition of the capital stock (machinery only, or a combination of building and machinery) and the form of financing (equity only, or a combination of debt and equity). Under all four project scenarios, the standard tax system in Mauritius offers the lowest METR to investors. The standard systems in Namibia, Botswana and Zambia are also attractive to investors, with the METR under 35 percent. In contrast, the standard tax system in Angola, DRC, Mozambique,94 Swaziland, and Zimbabwe all impose METRs above 50 percent, which would deter many investors. Highly leveraged investments generally bear a lower METR since nominal interest payments are deductible, whereas dividends are not. Machinery investment tends to face a higher METR than investments combining building and machinery, despite faster depreciation on the former. This is because the examples here assume that machinery has a high import content and that imported capital goods face a 10 percent duty.

94 Most of the tax information for Mozambique was not available in the version of the SADC Tax Database

used for this study. Tax parameters for Mozambique were therefore taken from other sources, which may not be fully up to date.

7-6

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Table 7-2 Illustrative Marginal Effective Tax Rate Calculations, SADC Member Countries, Standard Tax Systems (non-exporters, without special incentives)

Country

Project Scenario 1

Project Scenario 2

Project Scenario 3

Project Scenario 4

0% Debt Financing; 50% Building + 50% Machinery & Equipment

50% Debt Financing; 50% Building + 50% Machinery & Equipment

0% Debt Financing; 100% Machinery & Equipment

50% Debt Financing; 100% Machinery & Equipment

METR, Basic Tax System Angola

55.5

50.0

58.4

54.9

Botswana

34.5

30.5

36.3

33.7

DR Congo

63.2

56.5

65.1

60.1

Lesotho

62.8

59.6

66.3

65.5

Malawi

47.5

43.1

50.8

48.8

Mauritius

26.0

19.6

25.1

20.9

Mozambique

56.2

51.8

58.9

56.2

Namibia

30.2

22.3

21.4

28.9

Seychelles

46.7

35.7

42.4

32.9

South Africa

41.3

34.2

40.3

35.8

Swaziland

50.9

47.3

53.9

52.3

Tanzania

44.5

38.6

49.3

46.3

Zambia

31.1

35.0

34.1

40.7

Zimbabwe

44.2

39.2

47.1

44.6

Notes: To isolate the effects of company tax, calculations are based on four illustrative project scenarios, as defined in column headers. For each case, it is assumed that investment is foreign owned with a 20% target real rate of return, inflation rate of 10%, and nominal interest rate of 25% (to accentuate the effect of interest deductibility), and a 10% import duty on capital goods. These calculations use the tax system in each country without special incentives. The model takes into account the standard company tax, dividend tax (non-residents), loss carry-forward, capital gains tax (assuming business sold at year 10), and standard depreciation for industrial buildings and machinery, and other generally available capital allowances. SOURCE: Author's calculations using Dunn-Pellechio METR model described in Chapter 4. The main data source is the SADC Tax Database, version August 2003. For several countries this version of the SADC database was incomplete or not fully up-to-date. Other sources used to supplement the database are cited in the notes to Table 7-5 and in the Appendix. In view of data limitations, these figures are illustrative only, subject to a full assessment by country specialists.

TAX SYSTEMS AND INCENTIVES IN THE SADC REGION

7-7

One interesting feature that emerges from the illustrative calculations is that the tax treatment of capital gains has a relatively large effect on the METR. Each scenario assumes that the investor exits after 10 years by selling off the project. In this situation, the capital gains tax can be a major part of the overall tax burden, especially where profits escape tax through tax holidays or generous capital allowances, and where gains are not adjusted for inflation.95 Most SADC countries include capital gains in the tax base at normal tax rates, but Mauritius, Namibia, the Seychelles, and Zambia do not tax gains from the disposal of business assets.96 Tanzania, South Africa, and Zimbabwe tax capital gains from the disposal of corporate assets at a reduced rate (10, 15, and 20 percent, respectively).

7.3 Investment Tax Incentives Every SADC country, without exception, offers special investment tax incentives. Table 7-3 shows which countries offer each of 7 tax instruments, in one form or another.97 Lesotho has the most streamlined program (2 instruments). At the other end of the scale, Mauritius, Mozambique, Namibia, Swaziland, and Zimbabwe offer the broadest buffet of tax incentives. Many countries have a complicated patchwork of incentives, and could surely achieve equal or better results with a simpler regime. To be sure, Mauritius has demonstrated that a complicated program can succeed in attracting investment—at least in a country with efficient administrative systems. But some of the other examples demonstrate that a complicated system can also fail. Most important, there is no need for a complicated system to achieve results. No single package of investment tax incentives is suitable for all SADC countries, because they are heterogeneous in terms of economic conditions, fiscal requirements, administrative capabilities, and political preferences. In any case, as discussed in chapter 5 above, each instrument has advantages and disadvantages in terms of effectiveness and impact. Building on that analysis, the remainder of this section describes how each tool is being used in the region, taking them in order of popularity.

INITIAL CAPITAL ALLOWANCES (12 COUNTRIES) Initial capital allowances (ICAs) are widely popular in the SADC region. Only Angola and Lesotho do not report provisions in this category. South Africa offers an ICA between 50 and 100 percent in addition to normal depreciation, as the sole tax incentive offered under the Strategic Industrial Projects (SIP) program (Exhibit 7-1).

95 Angola and Botswana allow an inflation adjustment to the basis for capital allowances, but this feature was

not included in the present calculations. South Africa includes only 50 percent of the capital gain in the tax base. 96 Zambia imposes a 3 percent property transfer tax instead. 97 Table 7-5 and the Appendix provide more information.

Table 7-3 Investment Tax Incentives in the SADC Region (2003)-- Who Does What?

Count: Types of Incentive (max = 7)

Tax Holiday, Full or Partial, incl. export incentives

Favorable Tax Rates or Exemptions

Angola

3

1

1

1

Botswana

4

1

1

DR Congo

4

1

Lesotho

2

Malawi

5

1

1

Mauritius

6

1

1

1

1

1

1

Mozambique

6

1

1

1

1

1

1

Namibia

6

1

1

1

1

Seychelles

4

1

1

1

South Africa

5

1

1

1

1

1

Swaziland

6

1

1

1

1

1

1

Tanzania

3

1

Zambia

5

1

1

Zimbabwe

6

1

4.6

11

Country

SADC Count

Initial Capital Allowance, Some Categories of Companies

Export Incentives (beyond standard indirect tax relief)

Accelerated Depreciation <6 yrs for industrial machinery

Investment Tax Credit

1 1

Special Deductions For Employment Or Training

1

1

1

1

1 1

1

1

1

1 1

1

1

1

1

1

1

1

1

12

10

12

Note: 1 indicates that the incentive is available in some form. SOURCE: SADC Tax Database and other sources noted in Table 7-5 at the end of this chapter. For details, see the Appendix.

2

1

1

7

11

7-9

TAX SYSTEMS AND INCENTIVES IN THE SADC REGION

Exhibit 7-1 Strategic Industrial Projects Program in South Africa In November 2001, the government of South Africa



Cost Limits. The program sets a ceiling (up to

issued regulations defining a new regime of tax

R600 million) on the cost of the industrial

incentives for Strategic Industrial Projects (SIP) to

assets that may qualify for the ICA for any one

encourage selected industrial investments and

project. Separately, the law sets a ceiling of R10

stimulate growth, development, and competitive-

billion on the cumulative amount of ICA

ness. Compared to many incentive programs, the

benefits that can be granted under the

SIP is well designed. The favorable features include:

program.



Coherent Targeting. The sectors targeted by

Transparency. Under the SIP program, the

the SIP program are not atypical: new projects

qualifying criteria are explicit and substantive,

in manufacturing, computer technology, and

applications are to be gazetted promptly,

research and development. But the screening

awards are to be reported annually, and

system applies a scoring system that is highly

revenue costs are to be monitored.

coherent with the policy goals. Point

are

allotted for new products or processes, for filling a critical gap in an industrial cluster, for value added of at least 35 percent, for procurement from small, medium and micro enterprises, for infrastructure provision, and for full-time jobs created per million Rand of



Clawback Provisions. In addition to standard provisions for canceling benefits due to noncompliance with performance or reporting requirements,

the

program

provides

for

possible tax penalties in the event that benefits are disallowed.

determines

To be sure, the program has weakness. First, the

qualifying status and the level of benefits. Job

Minister of Trade Industry must take into account,

creation can account for up to 4 points on a

but not necessarily heed, the recommendations of

scale of 10.

the adjudication committee. Thus, decisions may

investment.





Attractive

The

point

Benefits

score

that

Still

Generate

Revenue. The sole tax benefit is an initial capital allowance (ICA) of 50 or 100 percent, depending on the qualifying point score. The ICA

is

additional

depreciation.

This

to

normal

is

very

accelerated

attractive

to

investors and yet fiscally reasonable. That is to say, the initial allowance substantially lowers the METR for most projects, while yielding

still be discretionary. Second, the critical job criterion includes “indirect jobs,” for which figures are easy to manipulate and difficult to substantiate. Third, the benefits strongly favor projects with a rapid

payback

period,

and

projects

run

by

companies with other industrial income against which to offset tax losses in early years. For standalone projects with a long payback period, the present value of the allowance may be small.

significant revenue in the medium run.

SOURCE: Section 12G of the Income Tax Act, and DTI promotional documents.

7-10

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Several SADC countries grant full first-year expensing—an ICA of 100 percent—for certain capital outlays. The most common categories are farm works and mining investments. Tanzania allows full expensing for approved investments in lead sectors and priority sectors, while Mozambique recently introduced full expensing for advanced technology equipment. Other ICAs in the region range from 10 to 60 percent, covering categories such as industrial buildings, export manufacturers, computer equipment, tourism and hotels, small industry, agriculture, and companies operating in designated regions. In Zimbabwe, a 50 percent special initial allowance covers a wide range of activities. In some countries the ICA reduces the basis for subsequent depreciation. In several cases, like South Africa’s SIP program, the ICA is an additional allowance, which amounts to direct subsidization of capital outlays. Zambia offers both: a 10 percent initial allowance and a 10 percent additional investment allowance. Several countries provide an ICA as the first step in a rapid depreciation schedule. A few, notably South Africa and Mozambique, set a cap on the amount of ICA that can be claimed for a given investment. Finally, a few ICA benefits are granted as a built-in general incentive of the tax system, while others apply only to projects that qualify under particular investment promotion programs.

PREFERENTIAL TAX RATES (12 COUNTRIES) Most SADC countries grant preferential tax rates to certain classes of companies. In some instances, the beneficiaries can obtain complete exemptions. Specific cases are EPZ companies in Namibia, Malawi, and (recently) Zambia, certain global financial services companies in Mauritius, and offshore companies established in the Seychelles. In other cases, special tax rates range from 10 percent (companies designated under Development Assistance Orders in Swaziland) to 25 percent (mining companies in Zambia and Zimbabwe). The most frequent preferential tax rate is 15 percent. This rate applies, for example, to manufacturing companies in Botswana and Lesotho, and agricultural enterprises in Lesotho and Mozambique. Viewing the tax differentials by sector, special tax rates are variously targeted to tax incentive companies in general, agriculture, manufacturing, and mining. In addition, Botswana offers a 15 percent tax rate to international financial service center (IFSC) companies, while Zambia grants 5 percentage points of tax relief to companies listed on the stock exchange. Mauritius has by far the most complicated list of targets for preferred status. The island nation applies a 15 percent tax rate to export enterprises, pioneer enterprises, strategic local enterprises, industrial buildings, companies in agriculture, tourism, export services, information and communications technology (ICT), and numerous other categories.

TAX HOLIDAYS (11 COUNTRIES) Many experts regard tax holidays as poor instruments for stimulating sound and sustainable investments. Nonetheless, this instrument is very popular in the region. Eleven countries

TAX SYSTEMS AND INCENTIVES IN THE SADC REGION

7-11

currently offer tax holidays to certain types of investors. The exceptions are Lesotho, the Seychelles, and South Africa. In two of these cases—Lesotho and South Africa—tax holiday programs were withdrawn fairly recently. Nine SADC countries—Angola, Botswana,98 DRC, Malawi, Mauritius, Swaziland, Tanzania, Zambia, and Zimbabwe—offer qualifying companies a full exemption from company tax during the holiday period. Holidays with partial tax reductions are available in six countries— Angola, Mozambique, Namibia, Swaziland, Zambia, and Zimbabwe. The programs in Zimbabwe and Namibia phase out tax benefit in stages, rather than ending them abruptly at the end of the initial holiday period. The most common tax holiday period is 5 years, but the benefits range from a low of 3 years for approved activities in Angola and DRC to as much as 20 years for some EPZ companies in Tanzania. For certain types of beneficiaries in Swaziland, Mozambique, and Malawi, the tax holiday includes an exemption from withholding tax on dividends. In most SADC countries, tax holidays are granted to licensed investors under specific investment promotion schemes. The beneficiaries variously include, among others, exporters, enterprises in priority industries (Malawi, Mauritius), registered manufacturers (Namibia), and companies in particular locations (Zimbabwe, Zambia). The case of Mozambique is especially interesting. In 2002 the government eliminated a variety of full tax holidays in favor of alternative incentives. For example, manufacturers operating in an Industrial Free Zone (IFZ) were previously exempt from income tax; under the new Fiscal Benefits Code they get a 60 percent reduction in tax for up to 10 years. Enterprises operating in four provinces of the Zambezi region, in a broad range of eligible sectors, previously obtained a full 5-year tax holiday followed by a reduced tax rate; they now qualify instead for investment tax credits. This program of incentives is less costly and much easier to monitor than the former system. A provocative feature of the program is an explicit provision for “exceptional incentives” that can be negotiated as a contract between the State and an entity investing at least US$500 million in agriculture, aquaculture, livestock, forestry, agroindustry, manufacturing, infrastructure, or tourism activities. This discretionary approach provides latitude for the government to pursue truly transformational mega-projects similar to the prototype, Mozal. At the same time, discretion can lead to kickbacks and revenue losses through the granting of excessive or redundant benefits. For example, there is no reason to offer large tax breaks to fundamentally viable resource-based projects that are location specific—as Botswana and Namibia have recognized to great advantage. The actual effect of discretionary authority depends on the quality of the government’s technical analysis and bargaining acumen, the rectitude of key decision makers, and the transparency of any such agreements.

98 Tax holidays in Botswana are obtained through a Development Approval Order. Very few orders have been

granted since the early 1970s.

7-12

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

SPECIAL EXPORT INCENTIVES (11 COUNTRIES) Every SADC country except DRC has mechanisms in place to provide exporters with relief from duty and indirect tax on inputs, consistent with the internationally accepted destination principal.99 Ten countries do this through free trade zones (FTZs), under various names. Other exporters can obtain relief from indirect taxes through bonded manufacturing arrangements, duty drawbacks, rebates, or exemptions. Exports are zero-rated for VAT in every SADC country that uses this tax. These provisions are a standard part of a sensible tax code, and should not tallied as tax incentives. In addition, 11 countries provide other tax benefits for exporters—all but Angola, Botswana, and Lesotho. Qualifying exporters can benefit from tax holidays or tax rate reductions in 8 SADC countries. Companies in FTZs can obtain full exemptions in Malawi, Namibia, and Zambia, and tax holidays in Mozambique (60 percent reduction for 10 years), Tanzania (exemption for 10-20 years), and Zimbabwe (exemption for 5 years). Reduced tax rates of 15 percent are available to exporters in Mauritius and the Seychelles. Three countries grant special benefits to exporters in calculating taxable income or the amount of tax due. Malawi provides an allowance equal to 12.5 percent of gross export sales; in addition, bonded exporters can deduct an additional 25 percent of their transportation costs. Namibia and Mauritius allow exporters excess deductions for export marketing and promotion costs. Mauritius also provides a tax credit of 15 percent to 40 percent of export volume, subject to the condition that the tax rate does not fall below 15 percent. South Africa offers special export tax incentives through the Motor Industry Development Program (MIDP). The MIDP incentive operates not through reductions in the income tax, but through import duty credits that are granted as a function of the domestic content of export sales. Swaziland has a similar program, called the duty credit certificate scheme, for the textile and clothing industry. The value of these duty credits arises from the fact that imports, say of automobiles, can be procured duty free and sold in the domestic market at duty-inclusive prices. The difference is a cash benefit, as a function of export sales. The scope for targeting tax incentives to exporters is constrained by WTO Agreement on Subsidies and Countervailing Measures (SCM). For countries with per capita GNP above US$1000 (in constant 1990 dollars), the SCM agreement requires the phasing out of export subsidies by 2002, including fiscal incentives for export enterprises as such. The main obligation of lower income countries under the SCM agreement is to avoid introducing new export subsidies. Existing export subsidies in low-income countries were not prohibited under the SCM agreement,100 but they can be “actionable.” This means that they can be

99 Goods purchased and sold between countries of the SACU region are not treated as imports and exports for

tax purposes.

100 The prohibition was to take effect for all WTO members at the end of 2002, including all developing

countries. The WTO is considering appeals to extend the deadline, and in the meantime allowing exceptions.

TAX SYSTEMS AND INCENTIVES IN THE SADC REGION

7-13

subject to challenge if they adversely affect the interests of another WTO member. SADC Member States need to ensure that their respective incentive schemes comply with the SCM conditions, to avoid the threat of countervailing action by their trading partners.

ACCELERATED DEPRECIATION (10 COUNTRIES) Strictly speaking, depreciation rates are “accelerated” if the write-off rate exceeds the real economic rate of capital consumption. Since the true economic life of a capital good cannot easily be determined, tax systems apply standardized write-off rates to broad categories of assets. To simplify the comparison across the SADC countries, the present analysis focuses on standard depreciation rates for plant and equipment in manufacturing. Depreciation is viewed as accelerated here if the allowable rate exceeds the (rather arbitrary) figure of 15 percent per year—a figure that Hussey and Lubick (1996) have suggested as an international norm. On these criteria, 10 SADC countries offer accelerated depreciation on industrial plant and machinery. Excluding full expensing (discussed above), the fastest write-off is 50 percent over 2 years in Zambia. Namibia offers a 33.3 percent per year write-off for mining and petroleum, while South Africa recently adopted a 40-20-20-20 depreciation schedule over 4 years for manufacturing machinery and equipment. In the Seychelles investments are written off in 3 years under a 5-year schedule that sums to 150 percent of the actual investment cost. Swaziland and Zimbabwe apply a depreciation rate of up to 25 percent to companies operating multiple shifts, and in Angola applies this rate to companies with investment licenses.

DEDUCTIONS FOR TRAINING AND EMPLOYMENT (7 COUNTRIES) Every SADC country supports training programs through budget outlays. On the tax side, seven countries go beyond allowing training and employment costs as a qualifying business expense. Botswana and Swaziland allow a 200 percent deduction for the cost of such programs. Smaller excess deductions are granted in Lesotho (25 percent), Malawi (50 percent) and Namibia (25 percent). Two countries offer tax incentives tied to employment or to wage costs. In Namibia, an extra 25 percent deduction applies to the cost of production line wages, while manufacturers in Zimbabwe can claim a 200 percent deduction on wages and salaries for additional employees. South Africa pursues an alternative approach to fostering training by charging companies a Skills Development Levy equal to 1 percent of total remuneration paid to employees. The levy generates earmarked funding to support training programs, but it also adds to labor costs, which creates a (small) disincentive for job creation.

7-14

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

INVESTMENT TAX CREDIT (2 COUNTRIES) The least popular form of tax incentive in the SADC region is the investment tax credit (ITC). This is odd because the ITC is widely regarded as the most cost-effective and transparent form of investment incentive (see chapter 5). In Mauritius, several categories of investment qualify for a 10 percent investment tax credit, subject to the condition that the tax payable is no lower than 15 percent. This minimum tax provision is very prudent. It deserves serious consideration in other SADC countries that are juggling the need for improved revenue performance and pressure to offer tax incentives. Mozambique’s major tax reform program in 2002 adopted ITCs as a central instrument for stimulating investment. The amount of the ITC ranges from 5 percent to 30 percent depending on the sector, location, and size of investment. Unused credits can be carried forward for a maximum of 5 years. This restriction limits (or eliminates) the value of the ITC for projects with large capital outlays and long payback periods, since they normally incur tax losses in the early years of operation.

OTHER INCENTIVES Other tax benefits prevailing in the region include exemptions from import duty on capital goods, capital gains tax, withholding tax on royalties and management fees, or personal taxes on key employees. The marginal effective tax rate faced by investors is also affected by other elements of the tax system such as provisions for loss carry-forward, loss offset against tax due on income from other sources, and indexing of the basis for depreciation and capital gains. Information on all of these issues can be obtained from the responsible authorities in each country.

ILLUSTRATIVE MARGINAL EFFECTIVE TAX RATES FOR MANUFACTURING COMPANIES WITH EXPORT TAX INCENTIVES Table 7-4 recalculates the illustrative METR for each country, but this time using tax parameters reflecting incentive programs for priority manufacturers who produce for the export market, with EPZ status where applicable. For example, using the SIP program (Exhibit 7-1) for South Africa, the METR calculation in Table 7-4 incorporates a100 percent initial allowance that is additional to the normal depreciation schedule. (The main parameters used for each member country, as well as the underlying assumptions, are summarized in Table 7-6.) Note that the METR calculations do not capture the full complexity of various tax systems. Also, the analysis does not take into account interactions between host- and home-

TAX SYSTEMS AND INCENTIVES IN THE SADC REGION

7-15

country tax systems, which can have a major influence on the tax burden faced by foreign investors.101 The METR figures show that Malawi, Namibia, and Zambia offer incentive regimes that reduce the METR to zero by fully exempting manufacturing enterprises in an EPZ from company income tax, dividend withholding, and capital gains tax, as well as import duty on capital goods. In Tanzania, a zero METR is achieved through a 10-year tax holiday, since we assume here a 10-year life for the investment. While a zero METR is very attractive to investors, one must not forget that these countries obtain zero revenue from their EPZ operations.102 Other incentive programs—in Botswana, Lesotho, Mauritius, Mozambique, the Seychelles, South Africa, Swaziland and Zimbabwe--strike a better balance by offering attractively low effective tax rates to investors, without foregoing all of the revenue. At the other end of the spectrum, the METR remains quite high under the incentive programs prevailing in Angola and DRC. From Table 7-4 and Figure 7-2 one can see that tax incentives greatly reduce the METR facing investors in all SADC countries. Incentives that sharply reduce the METR may have a strong effect on designated categories of investment if the incentives are well targeted, effectively administered, and complemented by other favorable investment policies. Otherwise, they may simply induce a misallocation of resources and large revenue losses through political maneuvering, corruption, and abusive tax avoidance. As emphasized throughout this report, the “best” package of tax incentives is a matter for each country to judge in light of national conditions and priorities. Nonetheless, one is tempted to single out Botswana and South Africa as having a particularly well balanced tax regimes characterized by competitive standard tax rates, along with incentive packages that are attractive to investors while raising significant amounts of revenue, without creating extreme economic distortions.

101 Strictly speaking, the analysis applies to investments originating from a country that has a source-based

income tax or a tax sparing agreement with the SADC host country.

102 Of course EPZ companies may still be withholding tax on wages and generating revenue indirectly by

supporting taxable ancillary operations and linkage effects.

7-16

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Table 7-4 Illustrative Marginal Effective Tax Rate Calculations, SADC Member Countries: Case—Tax Incentives for Projects Manufacturing Goods for Export

Country

Project Scenario 1

Project Scenario 2

Project Scenario 3

Project Scenario 4

0% debt financing; 50% building + 50% machinery & equipment

50% debt financing; 50% building + 50% machinery & equipment

0% debt financing; 100% machinery & equipment

50% debt financing; 100% machinery & equipment

(METR with standard tax system in parentheses) Angola

40.9 (55.5)

34.3 (50.0)

40.9 (58.4)

34.5 (54.9)

Botswana

15.5 (34.5)

9.2 (30.5)

15.9 (36.3)

10.1 (33.7)

DR Congo

40.5 (63.2)

37.7 (56.5)

40.5 (65.1)

37.7 (60.1)

Lesotho

18.2 (62.8)

13.9 (59.6)

19.3 (66.3)

15.8 (65.5)

Malawi

0.0 (47.5)

0.0 (43.1)

0.0 (50.8)

0.0 (48.8)

Mauritius

10.6 (26.0)

3.5 (19.6)

8.9 (25.1)

1.4 (20.9)

Mozambique

19.4 (55.4)

9.4 (50.5)

16.0 (58.9)

5.0 (56.2)

Namibia

0.0 (30.2)

0.0 (22.3)

0.0 (21.4)

0.0 (28.9)

Seychelles

19.5 (46.7)

14.8 (35.7)

15.9 (42.4)

14.5 (32.9)

South Africa

20.0 (41.3)

14.6 (34.2)

17.7 (40.3)

12.6 (35.8)

Swaziland

8.3 (50.9)

4.2 (47.3)

8.4 (53.9)

4.3 (52.3)

Tanzania

0.0 (44.5)

0.0 (38.6)

0.0 (49.3)

0.0 (46.3)

Zambia

0.0 (31.1)

0.0 (35.0)

0.0 (34.1)

0.0 (40.7)

Zimbabwe

4.5 (44.2)

2.9 (39.2)

4.3 (47.1)

3.1 (44.6)

Notes: To isolate the effects of company tax, calculations are based on four illustrative project scenarios, as defined in the column headers. For each case, it is assumed that investment is foreign owned with a 20% target real rate of return, inflation rate of 10%, and nominal interest rate of 25% (to accentuate the effect of interest deductibility). These calculations are based on the tax system in each country with special incentives geared to manufactured exports. The model takes into account the standard company tax, dividend tax (non-residents), loss carry-forward, capital gains tax (assuming business sold at year 10), and depreciation for industrial buildings and machinery, and other capital allowances. For exporters, the import duty on capital goods is assumed to be zero. The calculations exclude excess deductions other than capital allowances and other technicalities that would require customizing the METR model. SOURCE: Author's calculations using Dunn-Pellechio METR model described in Chapter 4. The main data source is the SADC Tax Database, version August 2003. For several countries this version of the SADC database was incomplete or not fully up-to-date. Other sources used to supplement the database are cited in the notes to Table 7-5 and in the Appendix. In view of the data limitations, these figures are illustrative only, subject to a full assessment by country specialists.

7-17

TAX SYSTEMS AND INCENTIVES IN THE SADC REGION

Figure 7-2 Illustrative METR with Standard Tax System and with Tax Incentives

34.5

Angola 10.1

Botsw ana

33.7 37.7

DRC

65.5

0.0

48.8

1.4

Mozam bique

20.9 5.0

Mauritius Nam ibia

60.1

15.8

Lesotho Malaw i

54.9

Standard

28.9 14.5

South Africa

32.9

12.6

Seychelles

35.8

4.3

Sw aziland Tanzania

0.0

Zam bia

0.0

52.3 46.3 40.7

3.1

Zim babw e 0.0

Incentive

56.2

0.0

44.6 10.0

20.0

30.0

40.0

50.0

60.0

70.0

Marginal Effective Tax Rate (METR) (Project Scenario 4 -- see Table 7-4)

7.4 Successes and Failures Experience with investment tax incentives in the SADC region is distinctly mixed. In some circumstances, tax incentives have helped to stimulate important investments: • The clearest and best known example is the Mauritius Miracle, where generous tax benefits are widely regarded as having played an essential part in attracting investments that generated tens of thousands of jobs and transformed the economy. • Lesotho has had considerable success in creating tens of thousands of taxpaying jobs in export manufacturing, by applying a simple 15 percent company tax rate for this sector, combined with the AGOA provision that allows low-income African countries to have duty-free access to the American market for garments made from third-country fabrics. • Analysts generally agree that tax incentives were essential in bringing the giant Mozal aluminum smelter into Mozambique as the first major investment since the end of the civil war and the introduction of democracy. • In South Africa, the MIDP program is widely regarded as a success in stimulating investment and rapid growth of automotive exports.

7-18

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

• The Ramatex textile factory in Namibia unquestionably based its location decision on the incentive package that included both tax and non-tax concessions. • In Botswana, the government considers as reasonably successful its 15 percent tax rate for manufacturers, and more recently for International Financial Services Centers. Since 1996, 120 manufacturing companies have been established under this program, and the new IFSC program is generating a good early response. One can find many other instances throughout the region where investment incentives have delivered results. But these positive examples are the exception rather than the rule. In most SADC countries, tax incentives have not generated substantial volumes of investment nor large gains in employment. As discussed in Chapter 2, taxes do affect investment flows, but tax incentives are a distinctly secondary consideration for most (but not all) investors. This pattern shows up in large-scale surveys studies, such as the World Bank’s World Business Environment Survey, and microeconomic studies such as that by Macamo (2002), who found that three-fourths of the recent investors he contacted in Mozambique would have pursued their projects without obtaining the tax incentives that they did receive. Even in the showcase examples noted above, important qualifications apply. In Mauritius, tax incentives were accompanied by very favorable economic and political conditions. So the island nation’s experience does not generalize to most other countries.103 In the case of Lesotho, it remains to be seen how well the garment industry will hold up to the scheduled lapse of the AGOA entitlement in 2004 and the termination of other textile and garment quotas in 2005. Temporary advantages may not produce sustainable success. With regard to Mozal, some have suggested that the government could have obtained a better bargain and more revenue. In the case of the MIDP, even proponents acknowledge that the investment has been highly capital intensive and that the boom in automotive exports has not created net new jobs. Finally, the Ramatex decision was clearly a case of intraregional tax competition, and questions have been raised about the net benefit for the winner (see Exhibit 6-2). In short, facts in the SADC region, and internationally, refute any simple claim that tax incentives are highly effective, or never effective. The real issue is, first, to understand the conditions in which tax incentives are likely to succeed or fail, and second, to design tax incentive programs, where appropriate, that will maximize the positive effects and minimize the negative effects. This requires careful policy analysis in the context of each country’s economic and fiscal conditions, institutions, and political preferences.

103 Furthermore, Mauritius has been facing what the Minister of Finance called “severe budget constraints” in

his Budget Speech 2003-2004 (paragraph 211). Budget deficits have exceeded 5 percent of GDP in recent years. In the latest budget, income taxes account for only 14 percent of total recurrent revenue.

TAX SYSTEMS AND INCENTIVES IN THE SADC REGION

7-19

7.5 Institutional Framework for Tax Incentives Tax incentive programs are characterized not only by tax rates and technical provisions of the tax law, but also by the institutional framework. Without attempting a detailed review, this section briefly discusses experience in the region on three issues: the statement of objectives; procedures; and fiscal transparency.

OBJECTIVES In official government statements on tax incentives or investment promotion policies in the SADC region, the most frequently named objectives are, first, the promotion of exports (9 countries); second, the development of manufacturing and industry, or more broadly, economic diversification (8 countries); and third, employment creation (7 countries). Other expressed objectives include “upliftment” of disadvantaged groups, spatial development, and promotion of small business development. In four countries, investment itself is stated as an objective, though it is better viewed as a means to more fundamental ends, not as an end in itself. Some of these official statements may be largely rhetorical, but many of them have substance. In particular, most of the incentive programs target export promotion and manufacturing development. In addition, countries that identify small business development and spatial development as objectives generally follow through with specific programs aimed at these targets (though in some cases the programs involve financial assistance rather than tax breaks). The employment objective is a genuine goal in every SADC country. Ironically, many of the tax incentives work at cross purposes to the goal of job creation by favoring capital-intensive activities and technology choices, by virtue of structuring benefits as a function of capital costs.104 A more coherent policy design would not depend on tax breaks for capital, as such, but would focus in the first instance on lowering tax rates to stimulate efficient investment generally, without tilting the field toward capital intensity. As for uplifting disadvantaged groups, tax incentives for manufacturing, exports, and so forth are a very roundabout way to achieve this end. Direct expenditure programs can be far more efficacious. If tax breaks cause a significant loss of revenue, then they can produce the opposite effect by reducing fiscal resources for expenditure programs aimed at poverty reduction.

104 Differences in factor intensity can be huge. Consider, for example, the fact that the $1.3 billion Mozal project

in Mozambique created fewer than 700 direct jobs; whereas $2.1 billion of labor-intensive investments by Taiwanese companies in South Africa, Swaziland, and Lesotho have created 110,000 direct jobs. Source on the Taiwanese investment: Speech by Taiwan liaison office in South Africa, June 12, 2003. http://www.roctaiwan.org.za/press/20030612/2003061201.html.

7-20

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Three premises underlie the use of selective tax breaks to enhance the profitability of certain private investments. The first is that the activities that gain support are likely to generate substantial economic and social externalities, more so than equally productive investments in other sectors. The second is that most of the favored investments would not be forthcoming in the absence of the tax benefits. And the third is that the external benefits from such projects outweigh the adverse side effects that arise from granting special tax breaks. The first premise is arguably valid for manufacturing and exports, if the activities are not highly protected or subsidized. But the second and third are often invalid in the real world, where the effectiveness and impact of tax incentives depend greatly on political and economic circumstances and the procedures for targeting the benefits. Thus, most tax incentive programs may ultimately not be serving their intended purposes.

PROCEDURES In every SADC country except Lesotho, a degree of discretion is involved in granting at least some of the main tax incentives. For the investor, the system generally involves filing an application with detailed information about the business plan and then passing a screening process. The screening may be tightly rules-based, as in Botswana and South Africa, or largely discretionary, in countries where the criteria are broad, vague, or loosely interpreted, as in Zambia and Tanzania. Approval is required even in a seemingly straightforward program such as the 15 percent tax rate for manufacturers in Botswana. To qualify, companies must apply for the benefit and demonstrate to a technical committee (headed by tax officials) that they are genuinely involved in a transformational manufacturing activity. Once through this screen, though, eligibility for the preferential tax rate is automatic. In many countries, certified investors are also subject to monitoring requirements intended to ensure that they deliver benefits to the economy in line with the business plan submitted at the time of registration. In most cases, the monitoring appears to be completely ineffectual. In any case, the prospect of having government monitor an investor’s business plan is itself an administrative impediment to entry. Some of the tax incentives, of course, are automatic. Examples include special capital allowances that are not tied to obtaining an investment certificate, such as the 40-20-20-20 depreciation schedule for manufacturers in South Africa, and certain preferential tax rates, such as the 15 percent rate for farming activities in Zambia and for manufacturing enterprises in Lesotho. Where screening and certification are required, the process is often administered by an investment promotion agency, but the formal approval is normally the responsibility of the ministry responsible for trade and industry or the ministry responsible for finance. In most countries, the investment agency is also responsible for promoting and facilitating investment

TAX SYSTEMS AND INCENTIVES IN THE SADC REGION

7-21

projects, often in the face of daunting bureaucratic hurdles.105 These days, most countries in the region are doing a good job of providing prospective investors with information about the application procedures, the screening criteria, and the approval process. In countries where decision to grant tax incentives are not controlled by the ministry of finance, officials representing that ministry are almost always involved in the process. Representation on a committee, however, may not be an effective guard against fraying of the tax net. In all circumstances, the ministry of finance should be empowered to set cost ceilings to limit the amount of revenue foregone.

FISCAL TRANSPARENCY Transparency is an issue at two levels. First, the granting of any discretionary tax benefits should be subject to public disclosure, as a guard against abuse. Some SADC countries have formal requirements for the gazetting of projects that are approved under special investment promotion programs. But this dissemination system is often honored in the breach, or with long delays, and the public information rarely includes a quantitative disclosure of the value of the tax breaks accorded. Second, the aggregate revenue cost of special tax breaks should be monitored and reported on a regular basis to ensure that the use of public funds is subject to public scrutiny and budgetary control. Transparency in accounting for the fiscal cost of tax incentives is a recent idea in the region. South Africa is in the forefront. The main tax incentive program, SIP (Exhibit 7-1), operates with a cap on the cumulative allocation of benefits, which demands monitoring of the taxes foregone. South Africa is also developing an annual statement of tax expenditures to be presented along with the budget. In Mozambique, the new Code of Fiscal Benefits explicitly regards tax benefits as “fiscal expenses,” and requires beneficiaries to file a declaration of benefits used in each tax year. It is not yet clear whether the mandate will work in practice, given the administrative weaknesses in the system. Tanzania has also moved in the direction of cost monitoring by introducing a Treasury Voucher System in 2003 to document and control the cost of exemptions granted to organizations and institutions. Early signs are that the procedure is working effectively. This prototype may evolve into a system for budgeting other tax expenditures. In countries where taxpayer files are not yet computerized and where investment promotion agencies do not have a record of which projects are being implemented, it would be very difficult to track annual tax expenditures or even the cost of newly approved fiscal benefits.

105 Djankov, et al. (2000) analyze the results of red tape studies, investor road maps, and other documents on

administrative barriers to investment in 74 countries. Based on a red tape study from 1996, Mozambique was ranked at the bottom for the number of days required to obtain all approvals needed to open a simple new business.

7-22

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Where information technology systems are reasonably efficient, the monitoring of tax expenditures or tax incentive commitment costs is an idea whose time has come.

7.6 Summary SADC Member States face critical budget constraints and economic development requirements, which underscore the need to ensure that investment tax systems and investment incentives are effective, efficient, and fiscally responsible. Tax incentive programs differ widely across the region, creating a variety of experience with different approaches. On balance, the evidence suggests that tax incentives are not an elixir for achieving rapid and sustainable growth or large gains in job creation, especially if investors perceive the investment climate to be fundamentally unfavorable. Most countries in the region would benefit from a critical review of their tax incentive programs, applying the lessons of experience and insights from careful policy analysis, with full regard to local circumstances. In many cases, this review is likely to indicate that the Member States would benefit by restructuring, scaling back, and simplifying their tax incentive programs, while introducing stronger fiscal controls and greater transparency.

Table 7-5 Elements of Tax Structure Facing Companies in SADC Member Countries (Sheet 1 of 4)

Country Angolaf

Basis

Basic Rate (%)

Source

35

Treatment of Dividends and Capital Gains on Business Assets Dividends—15% w/holding; CG—Taxed

Combined Company plus Dividend Tax (residents) (%)e 44.8

Memorandum : Personal Income Tax Marginal Rates 2-4-6-8-10-12.5-15; 20% for self employment

Botswana

Source, but worldwide for IFSC

25 (=15% + 10% add’l company tax)

DR Congo

Source

Lesotho

Malawi

Dividends—15% w/holding with set-off against the 10% additional tax; CG—at company tax rate, indexed.

26.3

5-10-15-20-25

40

Dividends—20% w/holding; CG— included in tax base.

52.0

3-5-10-15-20-25-30-3540-45-50

Residence

35

Dividends—15% w/holding on payments to residents; 25% to non-residents; payments by manufacturing companies exempt.; CG—gains on disposal of business assets included in tax base.

44.8

25-35

Source

30

Dividends—10% w/holding; companies in EPZ exempt.; CG—included in tax base.D16

37.0

10-20-30-40

Residence

25

Dividends—No tax on dividends; CG— No tax on capital gains.

25.0

15-25

Mozambiquea

Source

35

Dividends—18% w/holding; exemption under IFZ and regional incentive schemes; CG—information not available.

46.7

10-15-20 [Changed in 2002, details not yet available]

Namibia

Source

35

Dividends—No tax on payments to residents; 10% w/holding on payments to non-residents; CG: zero tax.

35.0

18-30-35 [top rate reduced in 2003 budget]

Seychelles

Source

40

Dividends—0% w/holding on payment to residents; 15% on payment to nonresidents; CG—zero tax.

40.0

No personal income tax

Residence

30

Dividends—12.5% secondary company tax (SCT) charged on declared dividends; CG—included in tax base.

38.8

18-25-30-35-38-40

Swaziland

Source

30

Dividends—15% w/holding on payments to non-residents (12.5% for SACU based companies); residents pay 10% tax; CG— "receipts and accruals of a capital nature are excluded unless specifically included by law."

37.0

12-19-26-33

Tanzania

Residence

30

Dividends—15% standard w/holding rate; 10% for agr. or certificate of investment; 0% for mineral sector & EPZ; CG—10%.

40.5

17.5-20-25-30

Zambia c

Source

35

Dividends—25% w/holding; exemption for mining sector; 5 year exemption for farming; EPZ companies exempt; CG— zero, but 3% property transfer tax.

51.3

30

Zimbabwed

Source

30

Dividends—20% w/holding; 15% for quoted companies; CG—20% (after inflation adjustment)

44.0

20-25-30-35-40-45

Mauritius

South Africa

Table 7-5 (continued) Sheet 2 of 4 Country

Tax Holiday (temporary reduction/exemption/rebate)

Exempt Companies (0% rate)

Differential Tax Rates

Angolaf

(1) 50% tax reduction for 3-5 years for new industries and commercial activities in key areas

None indicated

20% for agriculture, forestry, livestock; up to 70% tax on income from oil production

Botswana

5 years typically with Development Approval Order (DAO)

None

(1) 15% (5%+10% additional company tax) for mfg; (2) 15% for IFSCs; (3) for mining companies with profit rate > 33.3% tax rate > 25% based on formula

DR Congo

(1) 3 -5 year exemption for new companies; (2) Discretionary exemptions under contractual regime

None indicated

None indicated

Lesotho

None indicated

None indicated

15% for manufacturing and farming

Malawi

Priority industries have option of 5-10 year tax holiday (depending on size of investment) or fixed 15% tax rate

EPZ companies

21% life insurance; 35% branch of foreign company

Mauritius

10 years for foreign income of certified regional headquarters; holiday to June 2008 (or 15% tax rate) for investment under ICT scheme

Companies in freeport zone; financial services company under Global Business License; sugar cane cultivation on new land.

15% for tax incentive companies in a variety of categories.

Mozambiquea

(1) For agriculture 80% reduction until 2012; (2) for IFZ operations 60% reduction up to 10 yrs; (3) for mining and petroleum 25% reduction for 5 and 8 years; (4) exceptional incentives for investments > $500 million for up to 10 years

None indicated

10% for agriculture

Namibia

50% abatement for 5 years, phasing out over following 10 years for registered manufacturers

EPZ companies

55% diamond mining; 37.5% other mining; petroleum 35% + variable Additional Profits Tax

Seychelles

None indicated

(Offshore) companies in

15% rate for export companies under IPA and companies in special growth areas; 25% and 35% brackets for small businesses

International Trade Zoneb South Africa

6 year for export companies in Industrial Development Zones ended in 1999

None indicated

15% rate for small manufacturers (taxable income < R150,000; turnover < R5 million)

Swaziland

(1) 5-year holiday new export mfg industry*; (2) Development Approval Order gives 10 year holidays at 10% tax rate + exemption from dividend w/holding. * Note: Tax applies to excess income, as per formula.

None indicated

10% Development Approval Order

Tanzania

(1) 10 -20 year holiday for EPZ companies, followed by 25% tax rate (details depend on location); (2) Full remission until 2008 in Dodoma Capital Development Area

None indicated

None indicated

Zambiac

(1) Special agreements for tourism in Livingstone; (2) 5 year exemption for some small scale industry; (3) one-seventh reduction for rural enterprises for 5 years.

(1) EPZ companies; (2) Activities deemed by the Minister "to assist in the development of the Republic" can be wholly or partially exempt.

15% for farming, fertilizer, non-traditional exports; 25% for mining; 30% for listed companies; 45% on excess income in banking.

Zimbabwed

(1) 5-yr holiday + 5 yrs at 15% for qualifying investors; (2) 5 yrs for EPZ companies, followed by 15% rate; (3) 5 yrs each at 0%, 15% and 20% for BOOT arrangement and tourism facility in tourist zone:

None indicated

15% for licensed investor (after 5 yr holiday) or new infra. project in growth point area; 25% for mining; 20% for export manufacturing or processing and some tourist facilities; 10% for new manufacturing in growth point area.

Table 7-5 (continued) Sheet 3 of 4

Country

Standard Depreciation: case of industrial plant and machinery g

Loss Carry Forward

Initial Capital Allowance (ICA) or Accelerated Depreciation

Angolaf

12.5% SL; doubled with investment license; asset valuation adjusted for inflation via indexation coefficients

3 years

None indicated

Botswana

15% SL

5 years; except unlimited for mining and agriculture

25% ICA on industrial buildings; full expensing of capital assets for mining and agriculture

Dr Congo

DB (sliding scale) 5 years

5 years; unlimited for depreciation charges

60% ICA for manufacturers exporting > 20% of output

Lesotho

10% DB

Unrestricted

None indicated

Malawi

15% heavy machinery; 10% light machinery -- DB \f

Unlimited

Full expensing of farm works, industrial buildings, railway lines; 40% ICA for manufacturers (additional 15% in designated areas)

Mauritius

20% SL

Unlimited

25% additional ICA on industrial premises, plant and machinery, computer software, and state-ofthe-art technology in manufacturing; for ICT eqpt 50% ICA plus 3-yr write-off at 33.3% per year.

Mozambiquea

10% or 16.66%; rate doubled under investment license, tripled for tourism

5 years

Full expensing of special equipment for advanced technology, up to max of 15% of taxable income; 120-150% deduction (depending on location) for investment in public utility infrastructure

Namibia

33.3% SL

Unlimited

20% ICA on buildings, with 8% per year write-off of balance in manufacturing; full expensing of farm works; 3-year write-off for dev. Expenditures in mining and petroleum

Seychelles

20% SL

5 years on trading losses, unlimited on losses from capital allowances

20% additional ICA on manufacturing plant; Under IPA: 45-40-30-25-10% (total 150%) for capital assets in mfg, tourism and small industry; 45-40-2015-5 (total 120%) in agric, marine resources and professional services

South Africa

40-20-20-20% (since 2002 budget)

Unlimited

(1) Full expensing of capital for farm development and mining; (2) 50-30-20% for farm machinery and eqpt; (3) 50% to 100% additional allowance for industrial assets in qualifying strategic projects (subject to cap)

Swaziland

10% - 25% depending on number of shifts, DB

Unlimited

(1) 50% additional ICA for plant and machinery in manufacturing, for infrastructure assets, for hotels; (3) 20% ICA for farm buildings and employee housing; (4) Full write-off of capital for mining and farm development.

Tanzania

12.5%SL

5 years; unrestricted for mineral sector and economic infra- structure companies

50% ICA for approved investments in lead and priority sectors (reduced from 100% in 2002); 20% investment deduction for industrial building and machinery and farm works.

Zambiac

50% SL

5 years; 10 for mining; unlimited for certified Investor

10% investment allowance plus 10% ICA for industrial buildings; full expensing of farm works; 10% ICA for investment in certain tree & bush crops

Zimbabwed

10-25% depending on number of shifts, SL or DB

maximum 6 years, except for mining companies

(1) Expensing of certain farm works and mining investment; (2) 15% ICA for companies in growth point areas; (3) 25% ICA per year for 4 years on industrial and commercial buildings and machinery; (4) 50% Special Initial Allow on most capital assets

Table 7-5 (continued) Sheet 4 of 4 Country

Investment Tax Credit (ITC)

Excess Deductions for Employment and Training

Export Tax Incentives

Angolaf

None indicated

None (employment subsidy instead)

No free trade zones; exporters exempt from excise tax

Botswana

None

200% deduction for cost of approved training for citizen employees

No export zones or export incentives other than rebate on raw materials sourced outside SACU for exports outside SACU

Dr Congo

None indicated

None indicated

No free trade zones; approved companies exempt from export duties and taxes; see initial allowance (previous page)

Lesotho

None indicated

125% for training or tertiary education costs for manufacturing companies

Duty free imports for manufacture of goods to be exported outside SACU

Malawi

None indicated

Additional 50% of training cost for employee to earn degree, diploma or certificate

Exporters may have EPZ or bonded status; zero corporate tax and dividend tax; zero import duty and indirect tax inputs; allowance of 12.5% of gross export sales; additional 25% allowance on transport costs

Mauritius

10% on investment by companies in certain categories other than tax incentive companies, such that tax payable is not less than 15%

None indicated

Free trade zones; 15% company tax rate. Other exporters zero duty and VAT on scheduled inputs; double deduction of export marketing costs; tax credit of 15-40% on export volume such that tax not less than 15%

Mozambiquea

5%, 10%, 15%, 20% or 30% ITC depending on location, sector, and size of project) on eligible fixed assets; add 3% for tourism; credit c/fwd expires after year 5

None indicated

Industrial Free Zones function as EPZ except labor laws apply; also 60% reduction on company tax for 10 years

Namibia

None indicated

Manufacturers qualify for additional deduction on training expenses [amount?], and 25% additional deduction for production line wages

EPZ companies exempt from all taxes and duties; additional deduction of 25% to 75% on costs for export promotion and marketing; 80% allowance on taxable income from export of manufactured goods (excl. fish and meat)

Seychelles

None indicated

None indicated

15% tax rate for export companies under IPA; zero tax on offshore companies operating in international trade zones \b

South Africa

None indicated

Additional deduction up to R50,000 per employee under approved learnership programs.

(1) Companies in Ind. Dev. Zones zero duty and VAT on inputs. (2) Automotive exports (MIDP) obtain import duty credits as function of domestic content of exports. (3) Other companies that manufacture or process exports get rebate or drawback of duty on some imported inputs.

Swaziland

None indicated

200% deduction for cost of approved training expenses

(1) Free trade zones; (2) duty credit certificate scheme for textile and clothing exporters

Tanzania

None indicated

None indicated

(1) 10 -20 year holiday for EPZ companies, followed by 25% tax rate (details depend on location); holiday includes withholding tax; outside EPZ, duty drawback scheme applies

Zambiac

None indicated

None indicated

EPZ Act implemented 2003 provides; provides for standalone EPZ sites; in addition to standard EPZ tax benefits, also provides full exemption from corporate tax, withholding tax, capital gains tax and excise duty.

Zimbabwed

None indicated

Double tax deduction on wages and salaries for additional employees in manufacturing [SADC review]

EPZ at industrial parks or stand-alone: 5 year tax holiday then 15% rate; no w/holding tax on dividends etc.; tax reduced by 10 percentage pts for manufacturer exporting > 50% of output; 200% deduction for export development market costs.

Notes for Table 7-5 are on the next page.

Notes a Mozambique information Council of Ministers Decree 16/2002, CPI, IMF, and Globeafrica. For inconsistencies, the Decree is assumed to be correct. b Information on offshore business in the Seychelles is from SIBA website. c Information about EPZ incentives in Zambia obtained from MCTI website; 10 year c/fwd for mining is from ZIC via ipanet website. d Information on Zimbabwe supplemented from Zimbabwe Revenue Authority website, IMF tax summary 2001 and Ernst & Young Budget Proposals 2003. e Assumes that after-tax income is fully distributed at standard dividend tax rate to residents. For Botswana the combined tax is computed as follows: On income of 100, ACT = 10 and 15% of dividend payout (75) = 11.25. After offset, the dividend tax is 1.25, which is added to the company tax. f SADC database information supplemented from recent IMF Tax Summary g SL = straight line; DB = declining balance; table assumes SL if not otherwise indicated SOURCE: SADC Tax Database, August 2003, unless otherwise noted

Table 7-6 Parameters used for METR Calculation in SADC Countries, with Tax Incentives Case: Export manufacturing, foreign shareholders

Angola

Botswanaa

DR Congo

Lesotho

Malawi

Mauritius

Mozambique

Namibia

Seychelles

South Africa

Swaziland

Tanzania

Zambia

Zimbabwe

35.0%

15.0%

40.0%

15.0%

0.0%

15.0%

35.0%

0.0%

15.0%

30.0%

30.0%

30.0%

0.0%

20.0%

Building

5.0%

2.5%

5.0%

5.0%

5.0%

5.0%

4.0%

4.0%

4.0%

5.0%

4.0%

5.0%

10.0%

5.0%

Machinery & Equipment

12.5%

15.0%

20.0%

10.0%

15.0%

20.0%

20.0%

33.3%

25.0%

20.0%

15.0%

12.5%

50.0%

17.5%

SL

SL

DB

DB

DB

SL

SL

SL

SL

SL

DB

SL

SL

SL

15.0%

0.0%

20.0%

0.0%

0.0%

0.0%

18.0%

0.0%

15.0%

12.5%

0.0%

0.0%

0.0%

0.0%

3

5

5

30

30

30

5

30

5

30

30

5

5

6

Capital Gains Tax Rate

35.0%

15.0%

40.0%

15.0%

0.0%

0.0%

35% ??

0.0%

0.0%

15.0%

30.0%

10.0%

0.0%

20.0%

Tax Holiday (period 1)

5

0

5

0

0

0

10

0

0

0

10

10

0

5

50.0%

0.0%

100.0%

0.0%

0.0%

0.0%

60.0%

0.0%

0.0%

0.0%

66.7%

100.0%

0.0%

100.0%

0

0

0

0

0

0

0

0

0

0

0

20

0

5

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

16.7%

0.0%

50.0%

0.0%

25.0%

60.0%

0.0%

100.0%

25.0%

0.0%

20.0%

45.0%

100.0%

50.0%

50.0%

10.0%

50.0%

-

Yes

Yes

-

Yes

Yes

-

Yes

No

No

No

Yes

No

No

0.0%

0.0%

60.0%

0.0%

40.0%

25.0%

0.0%

0.0%

45.0%

100.0%

50.0%

50.0%

0.0%

50.0%

-

Yes

Yes

-

Yes

Yes

-

Yes

No

No

no

No

No

No

Building

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

10.0%

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

Machinery & Equipment

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

10.0%

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

0.0%

Company Tax Rate Tax Depreciation

Depreciation Method Dividends Withholding Tax for Non-Resident Carry-forward

Rate reduction Tax Holiday (period 2) Rate reduction Initial Allowance Building Adjust Base ? Machinery & Equipment Adjust Base? Investment Tax Credits

Additional Assumptions (held constant) Inflation—10%, Nominal interest rate—25%; Required real rate of return—20%; Import content of capital—25% for building; 75% for machinery & equipment Loss Offset (against other income)—No; Carry-over-Credits—Yes; Indexation of Tax Parameters—No; Import Duty on Capital Goods—0% aFor Botswana, the calculation assumes (for simplicity) that dividend payouts are less than the amount due under the 10% secondary tax; thus, the effective dividend tax rate is zero. SOURCE: SADC Tax Database, version August 2003 supplemented by other sources noted in Table 7-5 and the Appendix.

8. Conclusions and Recommendations We opened this study with a question about investment tax incentives from 25 years ago: “Why should schemes whose impact is either slight or unknown be so widespread and popular in developing countries?” This question is just as relevant today because international experience with tax incentives remains decidedly mixed. Some countries, such as Mauritius and Malaysia, have used incentives effectively in conjunction with other policies to stimulate efficient investment, rapid growth, and structural transformation. Many others, including several SADC countries, have offered generous tax incentives with little to show for it in employment, growth, and productivity. Still other countries, such as Uganda and Indonesia, have ended special tax incentives in favor of improving the basic tax system without diminishing investment or growth. Furthermore, even in some of the success cases, tax incentives have been associated with revenue loss, fiscal imbalance, economic inefficiency, abusive tax avoidance schemes, and pressure from interest groups seeking to benefit at the expense of the treasury. The objective of this study has been to (1) review the use of tax incentive policies in the SADC region and (2) provide officials and stakeholders in the region with a balanced explanation of the issues so they may better understand the conditions under which tax incentive programs are likely to succeed or fail. We have examined the basic economics of investment tax incentives; some of the tools used in economic policy analysis; the advantages and disadvantages of alternative tax instruments for stimulating investment; and the problem of harmful tax competition. In doing so, we have drawn widely on theoretical and empirical literature about determinants of investment, the economics of taxation, and lessons from international and regional experience. Our general conclusions are as follows: 1.

Non-tax factors are far more important than tax incentives in determining the level and quality of investment flows. Non-tax factors include the stability of the political system, the macroeconomic policy environment, the skills of the labor force, the condition of the infrastructure, the legal and judicial framework, and the efficiency of the banking system.

8-2

2.

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

The effect of incentives on productivity and efficiency is at least as important as the effect on the amount of investment. Productivity growth accounts for much of the difference in economic growth performance across countries. Modern theories of competitiveness emphasize the importance of policies in fostering productivity growth. Hence, productivity effects should be a central concern in any assessment of tax incentive policies.

3.

Investment tax incentives work well in some countries and poorly in others. The effectiveness and impact of any package of incentives depends on local economic and fiscal conditions, characteristics of the incoming investment projects, details of the tax code, and political judgments about tradeoffs among competing policy objectives. Thus, decisions about tax incentives must be country-specific.

4.

The benefits of investment tax incentives are widely exaggerated, while the costs are widely underestimated. This bias arises partly from weaknesses in tax policy analysis, and partly from political pressures that inherently favor special interests. By implication, it is advisable to err on the side of caution in establishing tax incentive policies, so as to avoid the risk of revenue losses and economic distortions. This is especially so in countries with stringent revenue constraints and problems with tax administration.

5.

Capacity building to strengthen tax policy analysis is a central priority. Every SADC country can benefit from developing stronger capacity to analyze the effectiveness and impact of investment tax incentive programs.

Given that the effectiveness and impact of any program depends on country-specific judgments and conditions, no single system of tax incentives can suit all SADC members. Nonetheless, our analysis suggests numerous principles and actions that SADC members may pursue jointly or individually in order to coordinate and improve their tax incentive policies. The rest of this chapter outlines the recommends.

8.1 General Principles for Agreement DEFINITION OF TAX INCENTIVES The SADC Memorandum of Understanding on Co-operation in Taxation and Related Matters should adopt a definition of “tax incentives” that recognizes both general and selective incentives. The definition should continue to encompass both direct and indirect tax measures.

CONCLUSIONS AND RECOMMENDATIONS

8-3

POLICY COHERENCE SADC members should agree to review the coherence of their tax incentive programs. One element of coherence is consistency between goals, criteria, and instruments. For example, the goal of promoting growth and development may be served by tax measures that carefully target activities that create substantial positive externalities through modern technology, research and development, and technical training. But this goal is poorly served by policies that support inefficient and uncompetitive activities or seriously distort the allocation of resources. The goal of employment creation may be served by incentives that are designed to stimulate labor-intensive activities, but not by measures that sharply reduce the cost of capital, thereby favoring capital-intensive projects and technologies. Likewise, the goal of bettering the lives of the poor might be served by well designed incentives to stimulate laborintensive investments, especially in rural areas, but not by programs that award excessive or indiscriminate fiscal benefits to wealthy business interests at the expense of funding for poverty reduction programs. A second element of coherence is consistency between tax policies and other policies to improve the investment climate, including macroeconomic policies, structural reforms, and institutional reforms. Tax incentives should be viewed as one element of an integrated program for promoting productive investment. The analysis of policy coherence should encompass incentives operating through both direct and indirect taxes, including protective tariffs and, in some countries, the remission of import duty on capital goods, raw materials, and intermediate goods.

TAX POLICY ANALYSIS SADC member states should mutually agree to develop the analytical capacity, organizational arrangements, and institutional procedures necessary to conduct a professional review of existing and proposed tax incentive measures and other tax policies. The review should address the effectiveness of each measure and its impact on revenue, tax administration, economic efficiency, and equity. This does not require highly sophisticated technical analysis. The purpose of strengthening systems for policy analysis is to ensure that policy decisions are based on full information about their likely impact. Otherwise, tax incentive policies tend to be driven by an overly optimistic vision of benefits and disregard of important costs.

REVENUE MANAGEMENT AND THE FINANCE MINISTRY Because revenue mobilization is the central purpose of the tax system, SADC member states should agree in principle that tax incentive programs must be designed to ensure prudent revenue management.

8-4

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Countries facing serious revenue constraints should be especially cautious about offering incentives that create substantial revenue risks, either directly or indirectly. This principle is important not only to maintain macroeconomic stability, but also to prevent tax incentives from undercutting the state’s fiscal capacity to improve more basic elements of the investment climate. To ensure proper consideration of revenue objectives, the ministry responsible for finance must be centrally involved in formulating tax incentive policies. Furthermore, to preclude legal inconsistencies, tax incentives should be incorporated in or carefully coordinated with the relevant tax legislation. Another ministry or agency may have the authority to manage and administer the tax incentive program, but the finance ministry must control the budgetary consequences. One way to shift authority while retaining fiscal control is to set an explicit budgetary ceiling on the amount of tax breaks that can be granted in any fiscal year.

8.2 Structure of Investment Tax Incentives CHOICE OF TAX INCENTIVE INSTRUMENTS The advantages and disadvantages of each tax incentive instrument must be judged in light of local conditions and priorities, subject to systematic policy analysis to support well informed decisions. Nevertheless, our analysis suggests ten broad conclusions about the choice among tax incentive instruments: 1.

The soundest tax incentive, overall, is to establish standard tax rates that are fair and moderate in comparison with rates prevailing elsewhere in the region and in the main capital-source countries.

2.

Relieving exporters of indirect tax on their inputs should be a top priority, with due regard to the need for effective procedures to prevent abuse of tax and duty remissions.

3.

The most cost-effective tax incentives are the investment tax credit (ITC) and the initial capital allowance. These tools yield a large reduction in the marginal effective tax rate (METR) relative to the revenue cost, with a minimum of administrative complexity. However, these incentives are not costless. They create moderate distortions by favoring capital-intensive projects and short-lived assets, and they can serve as avenues for tax evasion if not well administered.

4.

Tax holidays with a full exemption are far less cost-effective. The revenue loss is large relative to the improvement in investment incentives via reduction of the METR. Tax holidays also favor transitory rather than sustainable investments, and they create glaring opportunities for aggressive tax avoidance.

CONCLUSIONS AND RECOMMENDATIONS

5.

8-5

The worst form of tax incentive is the imposition of high protective tariffs on competing imports. This may stimulate investment oriented to the domestic market, but it usually turns out to have low productivity and poor development potential.

6.

In states with serious revenue constraints, special tax benefits should be narrowly targeted to activities that are likely to deliver a high payoff in terms of the policy objectives, and which would not be undertaken in the absence of incentives. The broader the applicability of tax incentives, the more revenue is foregone.

7.

Avoid zero tax rates. The vast majority of viable investment projects do not hinge on getting a total exemption from tax. Low or moderate tax rates, even during a tax holiday period, serve the purpose with a smaller revenue costs, less distortion of economic decisions, and lower incentives for tax avoidance.

8.

Eliminating import duties on raw materials and intermediate goods is a poor approach to stimulating investment. This measure usually has a very high revenue cost and little effect on investment (because indirect taxes are usually passed along to consumers rather than being borne by the producers).

9.

Location-dependent investments that are fundamentally viable, especially resource-based projects, should not receive special tax preferences. On the contrary, governments should negotiate carefully to capture a fair share of resource-extraction rents.

10. A low-rate alternative minimum tax can ensure that every company contributes at least minimally to the cost of basic public services.

TAX ADMINISTRATION In designing or reforming tax incentive programs, SADC members should place a high priority on the administrative implications. Countries with weak tax administration should shun high-value tax breaks that invite aggressive tax planning and abusive tax avoidance. Programs that involve relatively few, easily controlled tax incentives are preferable to offering a panoply of benefits.

AUTOMATIC VERSUS DISCRETIONARY INCENTIVES As with the choice among tax incentive instruments, member states must judge for themselves the scope for discretion in light of local conditions and priorities. In a country with highly disciplined administrative systems (e.g., Korea), discretionary programs can have advantages in targeting the use of incentives to enhance their effectiveness and impact, while minimizing the loss of revenue. In other circumstances, however, discretion usually creates more problems than it solves by increasing the administrative costs for both the authorities and the beneficiaries, politicizing decisions, and creating incentives for corrupt practices.

8-6

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Because these problems are often rampant and severe, automatic mechanisms are favored. If discretion is employed to screen beneficiaries or negotiate fiscal benefits, it is essential to have in place effective rules for timely public disclosure to ensure accountability.

8.3 Transparency FISCAL TRANSPARENCY SADC member states should embrace the need for fiscal transparency and move aggressively to implement suitable provisions and systems. The IMF Code of Good Practices on Fiscal Transparency should be the starting point for developing SADC standards. Specifically, member states should commit to: 1.

Transparent tax incentive systems, procedures, and criteria. In accordance with Article 2 of the SADC MOU on Co-operation in Taxation and Related Matters, the SADC tax subcommittee has been developing a comprehensive database on tax systems in the region. This work needs to be completed so that the database can be posted for public access in the near future. Member states should then establish sustainable systems for maintaining the database to ensure that the information is fully accurate and up to date. In addition, Member States should agree to publish and post on government and IPA websites both summaries and comprehensive descriptions of the current tax system and tax incentive programs. The publications and postings should include all information contained in the SADC Tax Database plus comprehensive information on procedures and criteria for obtaining tax incentives under each existing program.

2.

Public disclosure of discretionary tax incentives granted. Wherever tax incentive regimes involve discretionary screening of beneficiaries or negotiation of fiscal benefits, member states should ensure full and timely public disclosure of decisions on any tax relief so granted. This practice is essential to ensure accountability and guard against misuse of authority.

3.

Transparent information on revenue costs. SADC member states should regard revenue foregone through special tax incentives as a policy use of public resources. The amounts involved should therefore be monitored, reported, and subject to public scrutiny, including parliamentary review. The granting of discretionary tax incentives should be subject to budget ceilings consistent with the fiscal program, and reported to Parliament as such. As information technology systems permit, member states should monitor the annual cost of on-going tax incentive programs by developing and implementing tax expenditure budget reports, suitable to local conditions. Governments should convey these reports to Parliament on an annual basis and ensure public access to the information. As far as possible, the information systems should require from beneficiaries

CONCLUSIONS AND RECOMMENDATIONS

8-7

the submission of normal tax returns and other appropriate forms to report the use of special deductions, allowances, credits, rebates, remissions, or other tax benefits. 4.

Monitoring mechanisms. Member states should take seriously the need to establish effective systems for monitoring the implementation of projects that receive discretionary fiscal benefits contingent on performance criteria, at least on a random audit basis. Failure to monitor encourages false applications, places public revenue unduly at risk, renders useless any penalty or sanction clauses, and makes it virtually impossible for the government or the public to ascertain the effectiveness of the tax incentive programs.

8.4 Capacity Building TECHNICAL ASSISTANCE To implement the Capacity Building section of the MOU (Article 3), Member States should establish intraregional technical assistance and training programs for improving the analysis of tax incentive policies. Member States should also develop educational programs to help government officials and stakeholders better understand the economic and financial impact of investment tax incentives.

ANALYTICAL TOOLS The SADC tax subcommittee should develop a standard kit of analytical tools that Member States can use to analyze the effectiveness and impact of their tax incentive programs, and for the Community to use in comparing tax incentives programs within and outside of the region. The tool kit should include, 1.

A model for evaluating the marginal effective tax rates under different regimes and situations.

2.

Data systems for monitoring average effective tax rates by sector and size of firm.

3.

Guidelines for monitoring and disclosing tax expenditures.

4.

Guidelines for screening applicants for selective discretionary incentives, to maximize the cost-effectiveness of incentive programs throughout the region.

8-8

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

8.5 Harmful Tax Competition DEFINITION OF HARMFUL TAX COMPETITION With reference to the OECD (1998) list of “characteristics of a harmful tax regime,” Member States should consider adding several criteria to the six conditions listed in Article 4.3(a) of the MOU as evidence of “harmful tax competition.” While some of the OECD conditions may be inappropriate for low-income countries with weak systems for tax administration, four of the criteria—artificial definition of the tax base; negotiable tax rate or base; existence of secrecy provisions; and encouragement of purely tax-driven operations—are fully applicable to the SADC region.

AREAS FOR COOPERATION Article 4.3(b) of the MOU calls on Member States to avoid “introducing tax legislation that prejudices” another Member State. Member States should supplement this broad provision with an agreement to pursue more concrete steps to mitigate harmful tax competition. Many of the recommendations given above serve this purpose by improving policy formulation, transparency, and accountability. In addition, Member States should consider agreeing on the following measures to reduce the harmful effects of tax competition: 1.

Negotiate bilateral or multilateral protocols or treaties for the effective exchange of information, including coordination on tax audits and investigation of tax crimes.

2.

Establish a periodic joint review of existing tax incentive programs to evaluate compliance with any SADC agreement on tax cooperation, and provide every Member State with information and (advisory but non-binding) recommendations to improve the cost-effectiveness of their programs.

3.

Participate pro-actively in international forums dealing with harmful tax practices

4.

Jointly support the concept of an International Tax Organization with a mandate to evaluate and facilitate international responses to harmful tax practices.

5.

Develop standards to establish clear and effective legal provisions in each country on transfer pricing, thin capitalization, controlled foreign entities.

6.

Jointly pursue defensive actions to minimize abusive tax avoidance schemes using crossborder tax differentials.

7.

Initiate programs for regional fiscal transfers, similar to the structural funding arrangement in the European Community, to help poorer Member States pursue more

CONCLUSIONS AND RECOMMENDATIONS

8-9

constructive approaches to improving their investment climate, in lieu of aggressive tax concessions. 8.

Finally, any regional agreement to mitigate the harmful effects of tax competition requires an effective mechanism for resolving disputes and enforcing remedial actions in response to violations. Otherwise, it will be difficult to sustain cooperation, because the logic of the prisoner’s dilemma creates incentives for each state to deviate from the accords, to the detriment of overall economic development in the region.

References Tax policy and tax incentives Asher, Mukul (1997), “Reforming the Tax System in Indonesia,” pp. 127-166 in Thirsk (1997): Bird, Richard M. (1992), Tax Policy & Economic Development. Baltimore, Johns Hopkins. Bird, Richard M. and Oliver Oldman (1990), Taxation in Developing Countries, Fourth Edition. Baltimore: Johns Hopkins. Bosworth, Barry P. (1984), Tax Incentives and Economic Growth. Washington: Brookings. Broadway, Robin, Dale Chua and Frank Flatters (1995), “Investment Incentives and the Corporate Tax System in Malaysia,” pp. 341-374, and “Indirect Taxes and Investment Incentives in Malaysia,” pp. 375-398, in Shah (1995). Burgess, Robin and Nicholas Stern (1993), “Taxation and Development,” Journal of Economic Literature, Vol. XXXI (June), pp. 796-818. Casanegra de Jantscher, Milka (1990), “Administering the VAT,” in Malcolm Gillis, Carl Shoup and Gerardo Sicat, eds. (1990), Value Added Taxation in Developing Countries. Washington: World Bank. Chen, Duanjie and Ritva Reinikka (1999), Business Taxation in a Low-Revenue Economy: A Study on Uganda in Comparison with Neighboring Countries, Africa Region Working Paper Series No. 3. Washington: World Bank. Chua, Dale (1995), “The Concept of Cost of Capital: Marginal Effective Tax Rate on Investment,” in Shome (1995), pp. 161-165). Dunn, David and Anthony Pellechio (1990), Analyzing Taxes on Business Income with the Marginal Effective Tax Rate Model, World Bank Discussion Paper 79. Washington: World Bank. Flatters, Frank (2002), From Import Substitution to Export Promotion: Driving the South African Motor Industry. Queens University, Canada.

R-2

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Gillis, Malcolm (1989), “Comprehensive Tax Reform: The Indonesian Experience,” Chapter 4 in Malcolm Gillis (ed.), Tax Reform in Developing Countries. Durham: Duke University Press. Gordon, R. (1983), “An Optimal Taxation Approach to Fiscal Federalism,” Quarterly Journal of Economics volume 98, pp. 567-586. Hall, Joshua (1999), Tax Expenditures: A Review and Analysis, Joint Economic Committee Study. Washington: United States Congress. Hines, James R. Jr. (1998), “Tax Sparing” and Direct Investment in Developing Countries, NBER Working Paper No. 6728. Cambridge: NBER. Hussey, Ward and Donald Lubick (1996), Basic World Tax Code. Cambridge: International Tax Program, Harvard University. James, Sebastian (2003a), The Hyundai Case. Background Note on the SADC Tax Incentives study. Cambridge: Kennedy School of Government. James, Sebastian (2003b), The Ramatex Case. Background Note on the SADC Tax Incentives study. Cambridge: Kennedy School of Government. Kebonang, Zein (2001), Economic Cross Roads: Tax Incentives or Tax Competition: The Case of Botswana, Master’s thesis, Harvard University. Leechor, Chad (1986), Tax Policy and Tax Reform in Semi-Industrial Countries, Industry and Finance Series Volume 13. Washington: World Bank. Legwaila, Thabo (2003a), [complete this]. Background Note on the SADC Tax Incentives study. Legwaila, Thabo (2003b), [complete this]. Background Note on the SADC Tax Incentives study. Lemgruber, Andréa (1999), A Competição Tributária em Economias Federativas: Aspectos Teóricos, Constatações Empíricas e Uma Análise do Caso Brasileiro [Tax Competition in Federal Economies: Theoretical Aspects, Empirical Evidence, and an Analysis of the Case of Brazil], Master’s thesis, Department of Economics, University of Brasília. Brasília Modi, Arbind (2003), Capital Allowances: The Indian Experience, Background Note for the SADC Tax Incentives Study. Delhi: Department of Revenue, Government of India. Madani, Dorsati (1999), A Review of the role and impact of export processing zones,

Policy

Working Paper No. 2238. Washington: World Bank. Morisset, Jacques and Neda Pirnia (2001), “How Tax Policy and Tax Incentives Affect Foreign Direct Investment: A Review,” in Wells, et. al, pp. 69-108. Newbury, David and Nicholas Stern (1987), The Theory of Taxation for Developing Countries. New York: Oxford University Press, for the World Bank. OECD (1998a), Harmful Tax Competition: An Emerging Global Issue. Paris.

REFERENCES

R-3

OECD (1998b), Tax Sparing: A Reconsideration. Paris. OECD (2000), Towards Global Tax Cooperation: Progress in Identifying and Eliminating Harmful Tax Practices, Committee on Fiscal Affairs. Paris. OECD (2000a), Framework for a Collective Memorandum of Understanding on Eliminating Harmful Tax Practices. Paris. OECD (2001a), Corporate Tax Incentives for Foreign Direct Investment, OECD Tax Policy Studies No. 4. Paris. OECD (2001b), The OECD’s Project on Harmful Tax Practices: The 2001 Progress Report. Centre for Tax Policy and Administration. Paris. Oxfam Great Britain (2003), Tax Havens: Releasing the Hidden Billions for Poverty Reduction, Policy Paper 8. London. Parthasarathi Shome, ed. (1995), Tax Policy Handbook, Fiscal Affairs Department, IMF, Chapter 1, pp. 3-21. Pellachio, A.J., G.P. Sicat and D.G. Dunn (1987), “Effective Tax Rates under Varying Tax Incentives,” excerpted as Chapter 14 in Bird and Oldman (1990), pp. 181-188. Radelet, Steven (1999), Manufactured Exports, Export Platforms and Economic Growth, CAER-II Discussion Paper no. 43, Cambridge: HIID. Richard Bird and Oliver Oldman, eds. (1990), Taxation in Developing Countries, Fourth Edition, Johns Hopkins University Press, Chapters 12-16, covering “A Survey of Tax Incentives,” “The Economics of Tax Incentives,” “Comparing Tax Incentives,” “A Cost-Benefit Analysis of Tax Incentives,” and “Tax Incentives and Tax Expenditures,” pp. 151-210. Shah, Anwar, ed. (1995), Fiscal Incentives for Investment and Innovation. Oxford: Oxford University Press. Shah, S.M.H. and J.F.J. Toye (1978), “Fiscal Incentives for Firms in Some Developing Countries: Survey and Critique,” in J.F.J. Toye (ed.), Taxation and Economic Development. London: Frank Cass. Sholes, Wolfson, Erickson, Maychar & Shevlin (2002), Taxation and Business Strategy. Englewood, NJ: Prentice Hall. [incomplete reference] Slemrod, Joel (1005), “Tax Policy toward Foreign Direct Investment in Developing Countris in Light of Recent International Tax Changes,” chapter 6 in Shah (1995), pp. 289-307. Stotsky, Janet (1995), “Summary of IMF Tax Policy Advice,” pp. 279-283 in Shome (1995) Toft, Graham S. (1996), “Doing Battle over the Incentives War,” pp. 37-40 in Federal Reserve Bank of Minneapolis, The Region, special issue on The Economic War Among the States. Minneapolis. Vito Tanzi and Howell Zee (2000), Tax Policy for Emerging Markets: Developing Countries, IMF Working Paper WP/00/35.

R-4

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Wayne Thirsk, ed. (1997), Tax Reform in Developing Countries. Washington: World Bank. Wells, Louis, Jr and Nancy Allen (2001), “Tax Holidays to Attract Foreign Direct Investment: Lessons from Two Experiments,” in Wells, Allen, Morisset and Pirnia (2001), pp. 1-67. Wells, Louis, Jr and Nancy Allen, Jacques Morisset and Neda Pirnia (2001) , Using Tax Incentives to Compete for Foreign Investment: Are They Worth the Costs?, Foreign Investment Advisory Service Occasional Paper No. 15. Washington: International Finance Corporation. World Bank (1988), World Development Report 1988, Section II on “Public Finance for Development.” New York: Oxford University Press. World Bank (1991), Lessons from Tax Reform. Washington. Zee, Howell, Janet Stotsky and Eduardo Ley (2002), “Tax Incentives for Business Investment: A Primer for Policy Makers in Developing Countries,” World Development 30:9, pp. 14971516.

Investment Pigato, Miria (2001), The Foreign Direct Investment Environment in Africa, Africa Region Working Paper Series No. 15. Washington: World Bank. Basu, Anupam and Krishna Srinivasan (2002), Foreign Direct Investment in Africa – Some Case Studies, Working Paper WP/02/61. Washington: IMF. Batra, Geeta, Daniel Kaufmann and Andrew Stone (2002), Voices of the Firms 2000: Investment Climate and Governance Findings of the World Business Environment Survey (WBES), World Bank Discussion Draft. Devarajan, Shantayanan, William Easterly and Howard Pack (2002), Low Investment is not the Constraint on African Development, Center for Global Development Working Paper No. 13. Washington. Easterly, William and Ross Levine (2001), “It’s Not Factor Accumulation: Stylized Facts and Growth Models,” World Bank Economic Review 15:2, pp. 177-227. Emmons, Willis, Adele Cooper and J. Richard Lenane (1999), 1 800 Buy Ireland, HBS Case Study No. 9-799-132. Cambridge: Harvard Business School. Fox, James (2003), Successful Integration into the Global Economy: Costa Rica and Mauritius. Washington: Nathan Associates, Inc. Hadjimichael, Michael, Dhaneshwar Ghura, Martin Muhleisen, Roger Nord and E. Murat Ucer (1995), Sub-Saharan Africa: Growth, Savings, and Investment, 1986-93, Occasional Paper No. 118. Washington: IMF. Hausmann, Ricardo and Dani Rodrik (2002), Economic Development as Self-Discovery. Kennedy School of Government Working Paper. Cambridge: Harvard University.

REFERENCES

R-5

Hubbard, R.G. (1994), “Investment under Uncertainty: Keeping One’s Options Open,” Journal of Economic Literature 32:4, pp. 1816-1831. Macamo, José Luís, “Barrieras Administrativas ao Investimento em Moçambique: Liçoes Aprendidas da Experiencia de Investidores Recentes,” in Cassio Rolim at al., (2002), A Economia Moçambicana Contemporânea: Ensaios. Maputo: Gabinete de Estudos, Ministry of Planning and Finance. Mankiw, N. Gregory (2002), Macroeconomics, fifth edition, New York: Worth, 2002. Chapter 17 (Investment), pp. 453-474. McMillan, Margaret, Selina Pandolfi and B. Lynn Salinger (1999), Promoting Foreign Direct Investment in Labor-Intensive Manufacturing Exports in Developing Countries, CAER-II Discussion Paper No. 42. Cambridge: Harvard Institute for International Development. Melo, Alberto (2001), Industrial Policy in Latin America and the Caribbean at the Turn of the Century, Working Paper #459. Inter-American Development Bank. Michael Porter (1998), The Competitive Advantage of Nations, New York: Free Press. Moran, Theodore (1998) Foreign Direct Investment and Development: The New Policy Agenda for Developing Countries and Economies in Transition. Washington: Institute for International Economics. Moran, Theodore (2001), Parental Supervision: The New Paradigm for Foreign Direct Investment and Development. Washington: Institute for International Economics. Noland, Marcus and Howard Pack (2003), Industrial Policy in an Era of Globalization: Lessons from Asia. Washington: Institute for International Economics. Odenthal, Ludger (2001), FDI in Sub-Saharan Africa, OECD Development Centre Technical Paper No. 173,. Paris: OECD Pack, Howard (2000), “Industrial Policy: Growth Elixir or Poison?” World Bank Research Observer 15:1, pp. 47-67. Pack, Howard and Marcus Noland (2003), Industrial Policy in an Era of Globalization: Lessons from Asia. Washington: Institute for International Economics. Porter, Michael (1998), The Competitive Advantage of Nations. New York: Free Press. Pigato, Miria (2000), Foreign Direct Investment in Africa: Old Tales and New Evidence, Africa Region Working Paper Series No. 8. Washington: World Bank. Rodrik, Dani (1995), “Getting Interventions Right: How South Korea and Taiwan Grew Rich,” Economic Policy volume 20, pp. 53-107. Roemer, Michael (1994), “Industrial Strategies: Outward Bound,” in David Lindauer and Michael Roemer, eds., Asia and Africa: Legacies and Opportunities in Development. San Francisco: Institute for Contemporary Studies.

R-6

EFFECTIVENESS AND IMPACT OF TAX INCENTIVES IN THE SADC REGION

Spar, Deborah (1998), Attracting High Technology Investment: Intel’s Costa Rican Plant. FIAS Occasional paper 11. Washington: IFC and World Bank. Stern, Nicholas, “A Strategy for Development,” Keynote Address to the Annual Bank Conference on Development Economics, World Bank, May, 2001. Subramanian, Arvind and Devesh Roy (2001), Who Can Explain the Mauritian Miracle: Meade, Romer, Sachs, or Rodrik?. IMF Working Paper No. 01/116. Washington. UNCTAD (1998), World Investment Report 1998. New York: United Nations. UNCTAD (2000), Investment Policy Review: Uganda. Geneva: United Nations. UNCTAD (2002), World Investment Report 2002. New York: United Nations. UNCTAD (2003), Investment Policy Review: Lesotho. Geneva: United Nations. Wei, Shang-jin (2000), “How Taxing is Corruption on International Investors?” Review of Economics and Statistics 82:1, pp. 1-11. Wilhelms, Saskia (1998), Foreign Direct Investment and Its Determinants in Emerging Economies, EAGER Discussion Paper No. 9. Washington: USAID. World Bank (1993), The East Asian Miracle: Economic Growth and Public Policy, Oxford: Oxford University Press. World Bank (1997), World Development Report: The State in a Changing World, Oxford: Oxford University Press. World Bank (2002), Global Development Finance: Financing the Poorest Countries. Oxford: Oxford University Press. World Economic Forum and Center for International Development (2002) at Harvard University, World Competitiveness Report 2001/2002. New York: Oxford University Press.

Appendix. Investment Tax Incentives in the SADC Region This appendix extends the discussion of SADC tax systems in Chapter 7 by summarizing the main features of the investment tax incentive programs in each Member State. The summaries are based on official descriptions from the SADC Tax Database (version: August 2003), supplemented with information from members of the tax subcommittee, and internet sites providing country budget documents, IMF country reports, local investment promotion agencies, U.S. embassy country commercial guides, and tax reports from major accounting companies. Some passages are extracted directly from official documents. Further details can be obtained from the SADC Tax Database (once it is available to the public) and from official agencies in the respective countries. Each summary covers official objectives of the tax incentive programs, the program coverage, the main types of direct tax incentives, and (in most cases) the administrative process and fiscal context. The discussion does not cover incentives in the form of protective tariffs.

Angola 106 Emerging from a long and destructive war, the Government of Angola recognizes that there are serious deficiencies in the investment climate for projects outside the enclave petroleum sector. In addition to prudent macroeconomic policies and rehabilitation of public infrastructure, the Government sees a need to provide tax and financial incentives to spur economic modernization and productive investment, especially on the micro, small and medium scale. The current investment incentive law was introduced in 1997 (Decree 73/97), but it has not been widely implemented. A Presidential Working Group is completing a proposal for a new investment act.

106 Supplementary sources include information from SADC Tax Subcommittee, data from the Angola Ministry

of Finance and National Bank of Angola; IMF country Report No. 292, September 2003.

A-2

APPENDIX

The present investment code includes four tax benefits for approved investment projects: • 50 percent reduction for 5 years in the industrial tax rate (the company tax) • Exemption from excise duties on raw materials and equipment • Double the rate of amortization • Exemption from the transfer tax on acquisition of fixed assets. Eligible investments must produce goods for export or import substitution, with local value added of at least 30 percent. Approval is at the discretion of the National Tax Directorate. The basic tax code also provides an incentive for investment in agriculture, forestry, and animal husbandry by setting a preferential tax rate of 20 percent, compared to the standard “industrial tax” rate of 35 percent. A tax of up to 70 percent applies to income from oil production. There are no free trade zones, but exporters are exempt from the excise tax. The SADC database provides no indication of whether exporters have access to inputs free of import duty. No special investment allowance or investment tax credits are available. In addition to tax benefits, the incentive program includes preferential interest rates on financing, an employment subsidy, and subsidies for infrastructure works. Angola has been suffering triple-digit inflation in recent years, while running large government deficits. But the fiscal problem is not on the revenue side. Government revenue has ranged from 44−49 percent of GDP, mostly from the petroleum sector. Non-petroleum income taxes are a minor source of funds, and there is no immediate need to increase this component of revenue.

Botswana107 Botswana provides incentives mainly in the form of attractively low tax rates, and a simple system for integrating the company tax and dividend tax. The major goals of the incentive programs are employment, development of the manufacturing sector, citizen empowerment, and development of small businesses. Tax incentives apply to foreign and domestic investors alike. Since 1972, special incentives have been available through Development Approval Orders (DAOs), normally in the form of a 5-year tax holiday. Only a few companies have taken advantage of DAOs. In 1996, the Government passed a general DAO for manufacturing, reducing the basic tax to 5 percent plus a 10 percent additional tax, giving an overall standard rate of 15 percent for manufacturing. In 1999 a 15 percent tax rate108 was established for International Financial Service Centers (IFSC). IFSC companies are also exempt from

107 Supplementary sources include information from the tax subcommittee; IMF Article IV Consultation report

for 2002 and Country Report No. 02/243; PWC Botswana tax information summary 2002 and Botswana Budget 2003-2004; Kebonang (2001); and interviews conducted in Botswana. 108 This is a simple 15 percent rate; the additional company tax does not apply to IFSC enterprises since they are not subject to withholding tax on dividends.

INVESTMENT TAX INCENTIVES IN THE SADC REGION

A-3

withholding tax on dividends, interest, royalties, or management fees. The law has a broad provision for the Minister to enter into special tax agreements, but this authority has rarely been invoked. The government is also developing a new investment code that will liberalize restrictions on foreign investment. Mining and agriculture companies receive a tax preference via full deductibility of capital expenditures with unlimited loss carry-forward (compared to the standard 5-year limit). (The tax rate on mining companies ranges from 25 percent to 55 percent, depending on the profit ratio.) Another depreciation benefit is a 25 percent initial allowance on new industrial buildings, as well as repairs and improvements to such buildings. Companies can also claim 200 percent deductions for the cost of approved training programs for citizen employees. For companies to obtain DAO benefits, their applications must be screened by a technical committee headed by the Ministry of Finance and Development. The screening is based on criteria that emphasize employment and training of Botswana citizens. Even the general DAO for manufacturing requires screening to ensure that beneficiaries are genuine manufacturing operations. Similar careful screening applies to IFSC enterprises, through the Botswana Development Corporation. Thus, the tax incentives are discretionary but rules-based. The approval process generally takes six months to one year from the date of application. Since 1996, 120 companies have been approved under the manufacturing DAO. Since 1999, 13 applications have been approved for IFCS companies. The Department of Taxes is planning to form an audit division to minimize the abuse of these incentives, which could result if beneficiaries use avoidance schemes to shift income from related activities that bear the full 25 percent tax rate. Botswana has no free trade zones, but companies may obtain rebates or drawbacks for inputs imported from outside the SACU region for exports to non-SACU countries. Other investment incentives take the form of financial support rather than tax relief. These include programs of the Citizen Entrepreneur Development Agency and the Citizen Entrepreneur Assistance Equity Fund, as well as preferences for indigenous suppliers in government procurement. The Government formerly provided up-front grants through the Financial Assistance Program, but this program was widely abused by businesses that accepted cash payments without delivering tangible, sustainable results. The Government phased out the program in 1999. The Government of Botswana has no problem with revenue mobilization due to high receipts from the mining sector. Revenue generally exceeds 40 percent of GDP, and the budget is well under control. There is concern, however, to diversify the revenue stream away from dependency on mining, which currently accounts for about two-thirds of total revenue. To this end, the Government introduced a 10 percent VAT in 2002.

A-4

APPENDIX

Democratic Republic of the Congo109 A new Investment Code adopted in 2002 aims to support four key areas: • Civil engineering companies involved in the construction and maintenance of roads and motorways as well as public transport for passengers and goods; • Investments to develop mechanized agriculture and agri-processing; • Heavy investment to create a solid industrial base; and • Investments to develop the country’s natural resources, raising the added value and increasing export capacity. Investments in four sectors are explicitly excluded: mining and hydrocarbons; banking and insurance; commerce; and arms and military equipment. The Investment Code also has special provisions for qualifying small and medium enterprises, in the range of US$10,000200,000. The main tax benefit is a full exemption from business tax for a period of 3 to 5 years, depending on location. Approved investors are also exempt from payment of turnover tax on locally sourced equipment, industrial inputs, and construction; tax on land and building directly connected to the approved investment; and ad valorem duty on registration of their capital. Approved investors are also exempt from duties and taxes on imports of new machines, plant and equipment and spare parts (and used imports, in the case of small and medium enterprises). The import duty exemption is capped at 10 percent of CIF value, except for heavy-duty vehicles, ships, and aircraft. Companies that export finished or semi-finished products are exempt from export duties and taxes. Manufacturers who export at least 20 percent of production qualify for a special 60 percent initial allowance on investment, and declining balance depreciation of the remainder. According to the Foreign Investment Advisory Service of the World Bank, the 2002 Investment Code provides “international best practice” in areas such as legal guarantees, equal treatment of national and foreign investors, dispute mechanisms, and profit repatriation. It also greatly simplifies the tax incentive regime and streamlines the procedures for investment approval, compared to the 1986 Code. The new regime also has a provision for financial assistance funded by an Investment Promotion Tax on imports and domestic production. Prospective beneficiaries must file an application for approval by interdepartmental decree from the ministers responsible for planning and finance, following an assessment by the National Investment Promotion Agency. One generally criterion is a minimum 35 percent value added.

109 Supplementary sources include U.S. Embassy Country Commercial Guide 2002; FIAS project report 2002;

IMF country report 03/175

INVESTMENT TAX INCENTIVES IN THE SADC REGION

A-5

The investment climate in DRC is impaired by continuing problems with security, infrastructure, corruption, a weak legal system, and steadily negative economic growth reflecting the loss of capital and declining productivity. However, recent macroeconomic management has been outstandingly good, with inflation declining from more than 500 percent in 2000 to less than 10 percent in 2002. Yet revenue remains extremely low, at about 6 percent of GDP (in 2001). Despite the highest income tax rates in the SADC region, income and profit taxes account for only 1 percent of GDP.110

Lesotho111 The government’s stated objective for investment policy is to create a highly competitive environment for export-oriented manufacturing industries, and to increase employment opportunities for Basotho. The primary agency for promoting and facilitating investment is the Lesotho Investment Promotion Center. The incentive system in Lesotho is simple and clearly targeted. Manufacturers benefit from a special 15 percent tax rate and are exempt from withholding tax on dividends paid to residents or foreigners. Manufacturers also benefit from an exemption from sales tax on capital machinery and equipment, a full rebate on imported raw materials and components used in production for export outside the SACU region, and a 125 percent deduction for training or tertiary education costs. The 10 percent sales tax was replaced by a 14 percent VAT on July 1, 2003, with zero rating for exports. The 15 percent income tax rate also applies to farm enterprises. All of these incentives are automatic, and they apply equally to domestic and foreign investors. According to a March, 2003 Investment Policy Review by UNCTAD, the tax incentive program makes Lesotho highly tax-competitive in the manufacturing industry and in agriculture. For other sectors, UNCTAD finds that the tax burden is “quite unattractive,” especially for foreign investors who face a 25 percent withholding tax on dividends. The UNCTAD report mentions that the Ministry of Trade, Industry and Marketing is proposing additional incentives for manufacturing, and suggests that this will not be efficacious because the sector is already lightly taxed. Lesotho used to offer 5−10-year tax holidays for pioneering industries and export incentives under the GEIS scheme. The programs were cumbersome to administer, and reportedly encountered serious abuses. As a result, the government replaced them with the current streamlined system.

110 IMF Country Report No. 03/175, June 2003. 111 Supplementary sources include the Lesotho government website; IMF Country Report 02/97, 2002 and

Fourth PRSP Review, 2003; UNCTAD Investment Policy Review 2003; and interview with former senior government official.

A-6

APPENDIX

Government revenue in Lesotho has averaged around 40 percent of GDP. More than half of this comes from customs revenue sharing through SACU. Income taxes account for just under one-fourth of the total. In recent years the budget balance, including grants, has been well under control. Thus, the country does not face serious revenue constraints.

Malawi112 Malawi offers a variety of tax incentives with the aim of encouraging development, enhancing output, earning or saving foreign exchange, and expanding employment opportunities. In 1991 the Government passed an Investment Promotion Act providing large tax benefits to exporters and pioneering industries in agriculture, agro-processing, manufacturing, tourism, and several other sectors. At the same time the Malawi Investment Promotion Agency was established to promote, and facilitate both domestic and foreign investment. The current tax system retains generous incentives for exporters and priority industries. Exporters can qualify for EPZ privileges at an approved location, or be licensed to manufacture under bond. The geographic dispersal of duty-free operations makes it more difficult to control the leakage of imports into the domestic economy. Both export mechanisms provide standard free trade benefits, including exemption from excise tax on purchases of raw materials and packaging made in Malawi. EPZ companies also benefit from zero corporate tax and an exemption from withholding tax on dividends. Exporters also obtain a 25 percent tax allowance on international transport costs, and an allowance of 12.5 percent of gross export sales. Separately, investments of more than $10 million in priority industries are eligible for a 10year tax holiday or an indefinite 15 percent tax rate. (The holiday option is 5 years for investments between $5 and $10 million.) Other incentives include duty free importation of qualifying capital goods used in manufacturing, tourism, mining, horticulture, IT and telecommunication, and agriculture. For manufacturing, the duty-free benefit applies also to raw materials. On top of this, manufacturers qualify for a 40 percent initial allowance (plus 15 percent in designated areas). Agricultural enterprises obtain favorable capital allowances in the form of full expensing of farm works. Finally, companies receive a 50 percent additional deduction for training costs. Designation as an EPZ, a bonded manufacturer, or a priority industry requires approval by an appraisal committee chaired by the Ministry of Commerce and Industry. The tax incentive regime is exceedingly favorable for the investment in export activities, yet the incentives have not been notably effective due to other pressing problems with the

112 Supplementary sources include investment promotion documents from MIPA; IMF Country Report 02/182,

2002; and www.sadcreview.com.

INVESTMENT TAX INCENTIVES IN THE SADC REGION

A-7

investment climate. Tax collections have averaged about 15 percent of GDP over the past decade, which is a bit above the average for HIPC countries, but the has government faced serious budgetary constraints.

Mauritius113 Mauritius is known worldwide for transforming its economy and sustaining rapid growth through a strategy of creating an attractive and supportive investment climate. A wide range of tax incentives, including liberal EPZ provisions, have been a vital part of the investment promotion strategy since shortly after Independence in 1968. The incentive program was reformed in 1993, partly to reduce the abuse of tax holidays. A new Investment Promotion Act enacted in 2000 established a Board of Investment (BOI) with responsibility to promote Mauritius as an international investment, business, and service center. The BOI handles and expedites all proposals for investment. The objective of the investment incentive regime is to attract FDI, create employment, promote exports, and diversify the industrial base. The current incentive regime provides special fiscal benefits for both foreign and domestic investors in 22 categories including export and export service enterprises; global (offshore) businesses; pioneer enterprises; strategic local enterprises; modernization and expansion enterprises; industrial building enterprises; small and medium enterprises; regional headquarters; as well as investments in agriculture, tourism, leisure, financial services, venture capital, fishing, health, and ICT. The basic tax benefit for incentive enterprises is a 15 percent company tax rate in place of the standard 25 percent rate. Dividends and capital gains are untaxed altogether. Tax holidays are only available for foreign income of regional headquarters (for 10 years) and ICT companies (until June 2008). These businesses may elect a 15 percent tax indefinitely instead of the tax holiday. Export incentives include the standard exemption from duty and indirect tax on inputs for companies operating in a free trade zone. Export companies also get a double deduction for export marketing and promotion costs. Certain global business companies are considered as non-resident for tax purposes; others face the 15 percent income tax but receive a deemed foreign tax credit of 80 percent on chargeable income. A non-incentive company engaged in exports of goods or services qualifies for a tax credit of 15-40 percent depending on volume, such that the effective tax rate does not fall below 15 percent. Fiscal benefits for other incentive companies vary by the type of investment. They include exemption from customs duty on machinery, equipment and spare parts; a reduction in

113 Supplementary sources include government website documents, including Budget Speech 2003/2004; IMF

Article IV consultation report, 2002 and country report 02/144, 2002; and Subramanian et al (2001).

A-8

APPENDIX

income tax payable by key employees; and reduced registration duties. Special capital allowances are also available for certain sectors, including: a 20 percent annual allowance for hotel buildings; a 50 percent investment allowance and 3 year write-off for ICT equipment; and a 10 percent tax credit (spread over three years) for modernization and expansion enterprises. In addition to tax incentives, Mauritius also provides financial assistance to small and medium enterprises. Tax revenue in Mauritius has averaged about 20 percent of GDP. The company income tax, however, accounts for just 1-2 percent of GDP. Considering the highly developed condition of the economy, this is a very low yield. In recent years, expenditure demands have driven the government budget deficit to unsustainable levels above 5 percent of GDP. The government has recognized the need for revenue enhancements, but does not appear to be considering any rollback of tax incentives. Indeed, the 2003 budget adds education and training companies to the list of categories qualifying for the 15 percent tax rate; provides a tax credit of up to 60 percent for equity investments in spinning plants; and allows a 10 percent accelerated rate of depreciation on capital expenditure for golf courses.

Mozambique114 The Government of Mozambique enacted a major income tax reform in 2002. The reform included an overhaul of the 1993 Fiscal Benefits Code (amended in 1995 and 1998) and other special tax regimes which had been widely criticized as costly and complicated. The new Code simplifies and consolidates previous programs, and reduces the extent of tax relief for most investors. The 2002 Code states that the purpose of the tax incentives (“fiscal benefits”) are to benefit investments that promote economic development and activities “having a recognized public social or cultural interest.” Provisions of the Code apply to all investments that are duly registered and authorized for tax purposes, excluding most wholesale and retail commercial activities. Operationally, the Investment Promotion Center (CPI) serves as the central institution for promoting and facilitating investments. The main fiscal benefits, which apply equally to domestic and foreign investors, are • Exemption from import duties on capital goods, except where suitable capital goods can be obtained from local producers;

114 The SADC tax database had incomplete information for Mozambique, which was supplemented from the

following sources: Council of Ministers Decree 16/2002; IMF country report 02/139, 2002, 02/140, 2002, and 03/98, 2003; and the Mozambique country report for 2003 at www.sadcreview.com.

INVESTMENT TAX INCENTIVES IN THE SADC REGION

A-9

• Investment tax credits ranging from 5 percent to 30 percent of the total investment (excluding certain assets such as passenger vehicles), depending on the sector, location, and size of project; • Accelerated depreciation at double the normal rate for new immovable assets. Also, full expensing is allowed on special equipment for advanced technology (up to a maximum of 15 percent of taxable income); • Extra deductions of 20-50 percent for investments in public infrastructure or utilities, depending on the location; and • Tax holidays for certain types of investment, involving partial remission of income tax ranging from 25 percent to 80 percent. There are specific provisions for Industrial Free Zones (ZFI), large-scale projects (exceeding $500 million), designated Rapid Development Zones (ZRD), and investments in agriculture, hotel and tourism, mining, and petroleum. The ZFI are free trade zones that offer standard exemptions from duty and indirect tax on inputs, and a 60 percent remission of company income tax for 10 years. (The previous regime charged ZFI enterprises a tax of 1 percent on gross receipts in lieu of company tax.) For large-scale projects a contractual regime of exceptional fiscal benefits is subject to negotiation and approval by the Council of Ministers. This special treatment is available to projects that materially promote national economic development, materially reduce regional imbalances, and create at least 1000 jobs within 3 years, in five sectors: agriculture, aquaculture, livestock and forestry; agro-industry; manufacturing; infrastructure construction; and tourism. One notable feature of the new Code is that the fiscal benefits are considered fiscal expenses. Beneficiaries are required to submit an annual declaration of the benefits used each tax year. This is a major innovation, but the implementation remains to be demonstrated. The restructuring of fiscal benefits in 2002 was an integral part of an overall reform program aimed at modernizing the tax system. Revenue enhancement is one of the major concerns behind this process. In recent years the tax system has been generating 12−13 percent of GDP, leaving the government dependent on foreign aid for over half of its financing.

Namibia 115 The goals of the tax incentive program in Namibia are employment creation, industrial and economic development, export expansion, and uplifting of previously disadvantaged citizens. Tax incentives target exporters and manufacturers and apply equally to domestic and foreign investors. Companies enjoying EPZ status are exempt from all taxes and duties in Namibia

115 Supplementary sources include government website documents; IPPR Opinion No. 8, April 2003; report of

the Namibian Tax Consortium, December 2002.

A-10

APPENDIX

(though they are obligated to file returns to facilitate control). The government has also been offering additional benefits such as grants, loans, land, and infrastructure for major investment projects. Non-EPZ companies are given an 80 percent allowance on taxable income derived from the export of manufactured goods (excluding fish and meat), which reduces the tax on this income to 7 percent; and excess deductions ranging from 25 percent to 75 percent, depending on the export turnover) for export marketing and promotion expenses. Registered manufacturers who are not exporting are granted a 50 percent abatement of taxable income for 5 years, which reduces the tax rate to 17.5 percent; the abatement is phased out over the following 10 years. Manufacturers also benefit from accelerated depreciation for industrial buildings at 20 percent the first year and 8 percent for the next 10 years, and a 25 percent additional deduction for training expenses and production line wages. The processing of investment applications is handled by the Investment Centre, which is an agency of the Ministry of Trade and Industry, in close consultation with the Ministry of Finance. A recent review of the tax system by the Namibian Tax Consortium concluded that the generous EPZ provisions have been reasonably successful and should be retained. However, an analysis of the 2003 budget by the Institute for Public Policy Research116 points out that corporate tax revenue from non-mining companies has been declining in real terms in recent years, which could be a reflection of the incentives regime. Overall tax revenue has averaged about 30 percent of GDP, and budget deficits have been running at 3-4 percent of GDP.

Seychelles117 The Investment Promotion Act (IPA) of 1995 was adopted to develop the economy through diversification. The IPA prescribes programs for export-oriented units, special growth areas, and general businesses in five sectors:

agriculture and marine resources; industry and

manufacturing; professional services; small scale industry; and tourism. IPA concessions are provided equally to foreign and domestic businesses under a Certificate of Approval issued by the Ministry of Finance. Applications are processed by the Division of Trade and Commerce in that ministry. The basic incentive is a 15 percent income tax rate for export-oriented units (excluding tourism), and for tourism development in special growth areas. Certified investors in all categories obtain special depreciation rates on capital investments other than land and building. The rates total 120 percent to 150 percent over five years, depending on the sector, including 45 percent the first year and 40 percent the second.

116 IPPR Opinion No. 8, April 2003. 117 Supplementary sources include Government and SIBA website documents; and www.sadcreview.com.

INVESTMENT TAX INCENTIVES IN THE SADC REGION

A-11

Businesses operating in the Seychelles Industrial Trade Zone are exempted from business tax, trade taxes on business inputs, withholding tax, employer’s contribution to social security and the pension scheme. In addition, the Seychelles has heavily promoted the offshore business sector, which is administered through the Seychelles International Business Authority.

South Africa 118 Stated goals of the investment incentive programs in South Africa include employment creation, small business development, industrial development, export growth, spatial development and the “upliftment” of previously disadvantaged people. The use of income tax incentives in South Africa was sharply criticized by the Katz Commission report in 1994.119 Nonetheless, the government introduced tax holidays in 1996. Since 1999, however, the government has been minimizing such incentives and relying on more targeted schemes to provide financing, grants, training programs, marketing support, and infrastructure development. The Department of Trade and Industry and the Industrial Development Corporation administer most of these schemes. Several tax incentive programs remain in effect, applying equally to domestic and foreign investors: • Industrial Development Zones are zones in which companies obtain automatic duty-free access to imported inputs, exemption from VAT on inputs from the domestic economy, exemption from property and local taxes, and any incentives applicable under other provisions or programs. The government emphasizes that IDZ companies are not exempt from labor laws. Exporters outside the IDZs are eligible for rebate or drawback of duties on imported inputs used in manufacturing or processing for export. • Strategic industrial projects in manufacturing, computer related activities, and research and development activities can obtain additional capital allowances of 50 percent to 100 percent on the cost of qualifying industrial assets, subject to caps of R300 million and R600 million, respectively. The applicable allowance depends on a point scoring system based on an admirably clear checklist of criteria, as evaluated by a technical committee headed by the DTI. The additional deduction can only be claimed against income from the respective project. The SIP allowance provides a substantial effective subsidy for strategic investors. • Motor Industry Development Program. This program provides an incentive for exports of automobiles and automotive equipment, in the form of import rebate credit certificates as a function of the domestic content of export sales. The program also provides a Productive

118 Supplementary sources include SARS, 2003 Budget Tax Proposals; National Treasury, Budget Review 2003;

DTI website documents; IMF country report 03/17 and 03/18, 2003; Legwaila (2003a and 2003b); Flatters (2002); www.sadcreview.com; and interviews conducted in South Africa. 119 Thabo Legwaila, Tax incentives in South Africa: Findings of the Katz Commission, Background paper for the present study. University of Stellenbosch, May 2003.

A-12

APPENDIX

Asset Allowance in the form of an import duty credit equal to 20 percent of the value of qualifying new capital assets.120 The MIDP is widely regarded as a major success story in terms of promoting export growth, but not job growth.121 A similar temporary program is in place for exporters of textiles and clothing. • Since 2002, an accelerated 40-20-20-20 depreciation schedule applies to capital expenditures for manufacturing firms. In 2003 this was extended to scientific research and development enterprises. Special capital allowances are also available for agricultural enterprises, small business corporations, buildings in certain urban development zones, and several other categories. As part of its tax reform program the National Treasury is introducing an annual tax expenditure statement to increase transparency and control of the fiscal costs of tax incentives.122 Looking at the overall fiscal picture, total revenue of the national government in South Africa averages about 24 percent of GDP, and the overall budget balance is well under control.

Swaziland 123 Investment policy in Swaziland is driven by the need to promote and facilitate both foreign and domestic investment. The policy relies mainly on establishing attractive conditions for investment, financing programs, and export credit guarantees. The main investment programs are administered through the Swaziland Investment Promotion Agency (SIPA) Although the government prefers not to emphasize special tax concessions, several incentive schemes are available, applying equally to domestic and foreign investors. Their stated purpose is employment creation, development of small to medium businesses, industrial development, exports, and the “upliftment” of living standards. The main tax incentives are as follows: Free trade zones are authorized, providing the standard benefits to exporters. • Exporters of textiles and clothing also qualify for a Duty Credit Certificate Scheme which is scheduled to operate until March 2005. The credits are a function of export sales, and can be used to offset duties on the importation of designated textile and clothing products. • New businesses that are approved by the Minister of Finance as Development Enterprises are eligible for a DAO which grants a concessionary tax rate up to 10 percent for 10 years, plus exemption from dividend withholding tax. This provision applies to investments in manufacturing, mining, international services, and tourism.

120 Flatters (2002). 121 DTI, Monitoring the Implementation of the MIDP: Current Developments in the Automotive Industry, Draft 2000. 122 For a listing of current tax expenditures, see:

http://www.finance.gov.za/documents/budget/2003/review/annexure_c.pdf

123 Supplementary sources include IMF country report 03/22, 2003.

INVESTMENT TAX INCENTIVES IN THE SADC REGION

A-13

• Tax holidays for 5 years are also available to new businesses in manufacturing industries that are not already present in Swaziland, and which predominantly produce for export. These holidays are also subject to the approval of the Minister of Finance. Income in excess of a formula-based limit is still subject to tax during the holiday period. • Accelerated depreciation and additional initial capital allowances of 50 percent are provided for industrial buildings for manufacturing, machinery and equipment for manufacturing, infrastructure investments, and hotels. A 20 percent initial allowance applies to the construction of employee housing and farm buildings. Other tax incentives include a double deduction for approved employee training expenses in designated industries, and a 133-150 percent deduction for expenses incurred by handicraft and cottage industries to increase exports. The fiscal position in Swaziland is fairly good. Tax revenues have averaged more than 25 percent of GDP, with SACU receipts accounting for about half of the total. Company tax amounts to about 8 percent of government revenue, or 2 percent of GDP, which is a very low effective tax rate. The budget deficit has been running at 4−5 percent of GDP in the past two years.

Tanzania 124 The aim of Tanzania’s tax incentive programs is to attract productive investment, create employment and enhance exports. The Tanzania Investment Act of 1997 provides the basic framework for investment promotion, though associated tax measures are incorporated into the respective tax legislation. The main change in 1997 was to end income tax holidays outside of export processing zones, in favor of expensing of capital assets and remission from customs duty on capital goods for holders of a Certificate of Investment from the Tanzania Investment Center. The investment regime distinguishes two main categories of beneficiaries: lead sectors and priority sectors. Lead sectors include agriculture, minerals, economic infrastructure, tourism, and petroleum and gas, while priority sectors are manufacturing, natural resources such as fisheries and forestry, aviation, commercial building, financial services, transport, broadcasting, human resource development, and exports. Incentives apply equally to qualifying domestic and foreign investors. Lead sectors qualify for a 50 percent capital allowance (reduced from 100 percent in 2002), with accelerated write-off of the remainder, and a reduced 10 percent withholding tax on dividends (0 percent for mineral companies and economic infrastructure companies). They

124 Supplementary sources include information provided by tax subcommittee; Budget Speech 2003/2004;

PWC Budget 2003 report; IMF country reports 02/1 and 03/2, 2003.

A-14

APPENDIX

also get zero import duty and a VAT deferment or exemption on imported capital goods. Similar benefits apply to companies in the priority sectors, upon obtaining a Certificate of Investment. More generous incentives apply to exporters. Companies operating in EPZs obtain standard free trade benefits and exemption from domestic indirect taxes for production inputs. In addition, • Companies licensed under the Zanzibar Free Economic Zones Authority obtain a 10-year tax holiday from corporate tax and withholding tax, followed by a reduced tax rate of 25 percent for 5 years. • Companies operating under the Zanzibar Free Port Authority are exempt from corporate tax for 20 years. • Companies operating in EPZs on the mainland receive a 10 year holiday on company tax and withholding tax, and a maximum company tax rate of 25 percent thereafter. Other exporters are eligible for duty drawbacks on inputs used to produce exported goods. Special incentives were introduced in 1988 to attract investment into the Dodoma Capital Development Area. The package included remission of import duty and sales tax/VAT on inputs, and remission of income tax, together with a 50 percent remittance of power and water charges. These incentives have had no significant effect. Revenue collection has been a serious concern in Tanzania, as revenue yields are only 12−13 percent of GDP and over half of the budget is financed by foreign aid. To raise revenue, the government in 2002 eliminated all tax exemptions issued prior to 1997, and in 2003, introduced a Treasury Voucher System to document and cost all exemptions granted to organizations and institutions. If this process is successful, the voucher system may be extended to other tax incentives in the future. This is a first step toward explicit budgeting of tax expenditures. A recent evaluation of the current investment promotion strategy found that heavy incentives granted to agriculture have negatively affected revenue without generating much new investment. Incentives for the mining sector have also eroded revenue performance. To modernize the system, the Minister of Finance has announced that the government is planning to repeal the current Income Tax Act (1973) and replace it with a new act as of January 2004.

INVESTMENT TAX INCENTIVES IN THE SADC REGION

A-15

Zambia 125 The Investment Act of 1993 is under review in Zambia, so changes in the incentive regime may be expected. The current program is designed to establish an attractive environment for domestic industrial growth, export promotion, and the development of a market-oriented economy. Investment certificates are negotiated on a rather discretionary basis through the Zambia Investment Centre, which assists investors in obtaining the necessary licenses, authorizations, and permits. The two main forms of tax incentive are preferential tax rates and special capital allowances. The standard tax rate of 35 percent, combined with the dividend withholding tax of 25 percent, produces a 51 percent overall tax on remitted profits, nearly the highest in the region. A tax rate of 15 percent applies to enterprises producing non-traditional exports, as well as farm enterprises and companies producing chemical fertilizer. A 25 percent tax rate applies to mining, and a 30 percent rate to companies listed on the stock exchange. Mining companies are also granted an exemption from withholding tax on dividends, as are farm enterprises for five years. In addition, the mining sector benefits from duty free importation of capital equipment. Special capital allowances include full expensing of farm works, 20 percent per year accelerated depreciation for farm improvements, and an additional 10 percent initial allowance on buildings used for manufacturing, mining or hotels, as well as capital expenditures for growing certain tree and bush crops. Plant, machinery, and implements used in farming, manufacturing, leasing, or tourism can be depreciated at an annual rate of 50 percent. Tax holidays are available in limited situations. Small scale enterprises and village enterprises registered under the Small Industries Development Act are exempt from income tax for three to five years. Other rural enterprises receive a one-seventh reduction in tax for five years. Some special tax holiday agreements have been negotiated with tourism enterprises in the Livingstone area. In addition, the Minister has authority to approve whole or partial exemptions for activities that the Minister may deem to assist the development of the economy. A major change in the tax incentive regime was introduced under the EPZ Act in 2002, which also established an EPZ Authority Board under the Ministry of Commerce, Trade and Industry. The government now grants a full exemption from corporate tax and withholding tax to licensed EPZ enterprises, in addition to the standard EPZ remission of import duty and other indirect tax on inputs. Zambia’s Poverty Reduction Strategy Paper (PRSP) states that the

125 Supplementary sources include Zambia PRSP; ZIC website documents; ZRA website documents; IMF Letter

of Intent 2002; and news articles on the recent EPZ provisions.

A-16

APPENDIX

EPZ measure was motivated as a means of stimulating industrial activity.126 The PRSP indicates that other measures under consideration include duty relief on imported raw materials, and the designation of certain locations as tax-free areas. Overall tax revenue is averaging about 18.5 percent of GDP, which is good performance for a low income country. Nonetheless, the government has faced extremely high fiscal deficits of over 7 percent of GDP due to excessive expenditure.

Zimbabwe127 Zimbabwe provides investment incentives with six objectives in mind: employment creation; small business development; industrial development; export promotion; spatial development; and “upliftment” of the disadvantaged. Many of the incentives take the form of financing arrangements, which operate through the Ministry of Industry and International Trade, the Industrial Development Corporation and the Zimbabwe Investment Centre. The most extensive tax incentives accrue to exporters. Under the Export Processing Zone Act of 1995, enterprises in manufacturing, processing or services that are licensed by the EPZ Authority to operate in an EPZ obtain a 5 year tax holiday, followed by a rate of 15 percent. EPZ companies also receive the standard duty-free access to imports and refunds on sales tax for domestically procured goods and services. In addition they are exempt from capital gains tax, shareholder’s taxes and non-resident taxes on interest, fees, royalties and remittances. Other exporters outside EPZs qualify for a rebate or drawback of certain duties on imported inputs. Since January 2003, manufacturers that export 50 percent or more of their volume are taxed at 20 percent. Finally, exporters can take a double deduction for export marketing costs. Tax holidays apply to other activities as well. Tourism operators in approved tourist development zones benefit from a 5 year holiday, followed by a 15 percent tax rate. The same provisions apply to industrial park developers. Build-own-operate-transfer projects obtain a 5 year holiday, followed by 15 percent for 5 years, 20 percent for 5 years, and then the normal tax rate. In growth point areas, approved manufacturers get a 10 percent tax rate, while certain infrastructure projects get a 15 percent rate. Special investment allowances also apply to a limited set of beneficiaries. Operators in a growth point area get a 15 percent initial allowance for investment in commercial or industrial buildings, staff housing, and machinery

126 According to a local press report (Times of Zambia, December 5, 2002) the EPZ measure was driven by

disenchantment with the effectiveness of former provisions for manufacturing under bond and duty drawbacks, and by the example of Kenya and Mauritius, as “living testimonies of the success stories of the export processing zones.” These examples are surprising, because the success in Mauritius depended heavily on a very supportive environment aside from tax breaks, and EPZs in Kenya have not been successful. This is the clear conclusion of Radelet (1999) and Madani (1999). 127 Supplementary sources include Zimbabwe Revenue Authority website; IMF country report 02/126, 2002; Ernst & Young Budget Proposals 2003 and Tax Facts 2002.

INVESTMENT TAX INCENTIVES IN THE SADC REGION

A-17

and implements. (This allowance reduces the basis for further depreciation.) Full expensing is available for farm works and mining investments. In the second half of the 1990s, overall tax revenues in Zimbabwe normally ranged between 24 percent and 28 percent of GDP, but in recent years revenues have fallen due to the economic contraction and instability. Expenditures have been far higher, creating budget deficits above 10 percent of GDP. Neither revenue collection nor investment are likely to recover until the government achieves its stated aim of creating a stable macroeconomic environment as a foundation for investment promotion.

Loading...

Effectiveness and Economic Impact of Tax Incentives in the SADC

TECHNICAL REPORT Effectiveness and Economic Impact of Tax Incentives in the SADC Region SUBMITTED TO USAID/RCSA SADC Tax Subcommittee, SADC Trade, ...

2MB Sizes 75 Downloads 7 Views

Recommend Documents

Effectiveness of tax incentives for venture capital and business angels
Such tax incentives have become an increasingly important part of the investment and innovation policy mix in the EU and

Tax incentives in India - EY
various incentives such as tax holidays, investment allowances, tax credits, rebates and so on. ▻ Prior to expansion/

The Economic Impact of President Bush's Tax Relief Plan | The
The President has proposed to lower marginal income tax rates, reduce the marriage penalty, increase the child tax credi

The impact and effectiveness of ministerial reshuffles
Jun 14, 2013 - and with Parliament, and highlighting the importance of the Ministerial and Civil Service Codes. In addit

The Impact of Tax Revenue on Economic Growth - IOSR journals
individual from the services rendered by the government has rendered this theory inapplicable (Ahuja, 2012). ..... Ahuja

the economic impact of trade in counterfeit goods and - EFZG
Nov 23, 2016 - uncertainty signals a difference between the beliefs and values of the population and the ..... internati

The Impact of Communication on Marketing Effectiveness
ACKNOWLEDGEMENT. The quest for knowledge is a universal one and it is what sets man apart from lower animals. The pursui

Utility Incentives and Services - Iowa Economic Development
Utility companies provide energy to Main Street districts, but they can also be a valuable resource when considering ene

Tax Planning Dalam Pemanfaatan Tax Incentives - Documents
Nov 8, 2015 - MAKALAH MAKALAH TAX PLANNING DALAM PEMANFAATAN TAX INCENTIVES Dosen Pengampu: Ali Irfan, M.Si., Ak. TUGAS

Oklahoma Business Incentives and Tax Guide
The State of Oklahoma was awarded $13,168,350 by the U.S Treasury for the State Small Business Credit. Initiative. The a