This Selected Issues paper assesses macroeconomic fiscal risks and the benefits of improved fiscal risk management in Angola. Angola faces fiscal risks coming from multiple sources, such as volatility in oil prices and production, macroeconomic shocks, weak macroeconomic forecasting; weaknesses in public fiscal management, energy subsidies, potential delays of oil revenue transfers from the state-owned oil company Sonangol to the Treasury, and contingent liabilities from state-owned banks and enterprises. Addressing these risks requires action in various fronts, including more transparent fiscal reporting, improved forecasting of fiscal aggregates and other macroeconomic variables, developing a fiscal stabilization fund with more flexible deposit and withdrawal rules, strengthened public expenditure controls, and more timely oil revenue transfers from Sonangol to the Treasury.


This Selected Issues paper assesses macroeconomic fiscal risks and the benefits of improved fiscal risk management in Angola. Angola faces fiscal risks coming from multiple sources, such as volatility in oil prices and production, macroeconomic shocks, weak macroeconomic forecasting; weaknesses in public fiscal management, energy subsidies, potential delays of oil revenue transfers from the state-owned oil company Sonangol to the Treasury, and contingent liabilities from state-owned banks and enterprises. Addressing these risks requires action in various fronts, including more transparent fiscal reporting, improved forecasting of fiscal aggregates and other macroeconomic variables, developing a fiscal stabilization fund with more flexible deposit and withdrawal rules, strengthened public expenditure controls, and more timely oil revenue transfers from Sonangol to the Treasury.

Non-Oil Tax Expenditures1

This chapter provides a roadmap for properly measuring in Angola non-oil tax expenditures, understood as the potential revenues that the government forgoes as a result of tax reductions and exemptions. It also highlights the importance of discussing in the annual budget the costs and benefits for the country of tax expenditures in the same manner as public expenditures are discussed. This information is critical for policymakers to take decisions on whether to maintain or eliminate tax benefits and can prevent the tax system from deteriorating over time under the accumulation of tax benefits that may have ceased to address the country’s objectives.

The rest of this chapter is organized as follows: Section A highlights the importance of tax expenditure budgets. Section B discusses the methodological foundations for identifying and measuring tax expenditures. Section C proposes “benchmark” taxes for each of the main Angolan levies and identifies the tax expenditures relating to them. The final section presents conclusions and recommendations.

A. Introduction2

1. Tax expenditures are potential revenues that the government forgoes as a result of tax reductions and exemptions in deviation from the “normal” tax structure.3 Tax expenditures may take different forms, such as reductions in the tax base (exclusion, exemption, deduction, etc.), deductions from tax liability, tax rate reductions, and tax deferrals (for example, accelerated depreciation). As tax expenditures result from a deviation from the “normal” structure of the tax, estimating them thus requires that we define the “normal” structure, i.e., benchmark taxes that will serve as the basis for comparison.

2. Tax expenditures are not subject to the same scrutiny as government spending, unless the annual budget is accompanied by a tax expenditure budget (TEB). Traditionally, the document submitting the annual budget for consideration by the legislature contains an estimate of proposed revenues and expenditures. Tax expenditures—and the additional revenues that would be collected if they did not exist—remain hidden. Notwithstanding the fact that the same policy objectives targeted by tax expenditures could be achieved through subsidies or other outlays, the relevance of budgetary expenditures is debated every year, while the appropriateness of tax expenditures is discussed only at the time the legal rules giving rise to them are approved. It is highly likely that tax expenditures created in the past (very often in the distant past) will be irrelevant for current policy purposes.

3. In the absence of yearly scrutiny, the tendency is to perpetuate tax benefits that are no longer necessary, while at the same time introducing new ones as instruments for current policies. The resulting increase in the value of tax expenditures over time undermines the efficiency of the tax system and creates long-term budgetary problems. As circumstances change over time, the effectiveness of tax expenditures is reduced. Likewise, the growing cost in terms of lost revenues obliges the government to neglect new priorities and/or to raise tax rates and/or to incur greater deficits and debt. This reduces the effectiveness of government policy and can also undermine the efficiency, the fairness and the simplicity of the tax system.

4. The purpose of TEBs is to mitigate these problems by making tax expenditures just as visible as budgetary outlays. They contain information on the amount, the purpose and the use of tax benefits, thus allowing for scrutiny and facilitating control. They increase the transparency of government policies, for they allow policymakers and legislators to obtain a full overview of the costs of each activity or project. Estimating tax expenditures is also essential for conducting cost benefit analyses and for assessing the potential revenue that eliminating the tax benefit would produce. Identifying tax expenditures also allows the tax administration to pinpoint gaps or loopholes in the legislation that could be used for aggressive tax planning (with its attendant harm to the Treasury) or for evasion.

5. Although there is no standard format for TEBs, the steps necessary to put one together are the same in all cases. These involve: (a) choosing the methodology for measurement; (b) defining the benchmark tax structure; (c) identifying tax expenditures; (d) estimating the value of tax expenditures, i.e. the amount of potential tax revenue that was not collected; and (e) preparing the document listing tax expenditures, which in some countries is an integral part of the annual budget, while in others it is annexed to the budget.

B. Methodology

6. The first step in estimating tax expenditures is to specify the benchmark taxes. By definition, tax expenditures result from the difference between the treatment actually accorded the taxpayer and the treatment that would result from adopting the benchmark tax. There is no consensus in the tax expenditure literature as to how the benchmark taxes should be specified. However, most countries use either the conceptual approach or the legal approach for defining them.4

7. In the conceptual approach, benchmark taxes are ideal taxes that follow the theoretical principles of taxation. For example, the benchmark for the consumption tax would be a tax that is applied at a uniform rate to all goods and services consumed in the country; in the case of individual income tax, the benchmark would be a progressive tax on global income. This approach is obviously not applicable to Angola, where there is no tax on global income, but rather separate taxes on different types of income, and where the consumption tax is not a general tax on consumption but rather a tax with a less comprehensive base and with selective rates.

8. In the legal approach, benchmark taxes are those with the general characteristics of taxes defined in the country’s laws. The definition of tax expenditures is more restrictive under the legal than under the conceptual approach because, in defining the tax bases, the laws usually determine what is subject to the tax and what is not. With the conceptual approach, exclusions as well as exemptions are considered tax expenditures, while under the legal approach only exemptions constitute tax expenditures. On the other hand, the legal approach is simpler than the conceptual one and may better reflect the country’s reality, in contrast to a theoretical ideal. For this reason, the legal approach is the one adopted by the majority of developing countries.

9. Once the benchmark tax structure has been specified and tax expenditures identified, the next methodological issue is how to estimate their values. The literature offers three approaches that differ with respect to the objective, the degree of difficulty in the estimation, and the estimated value of the tax expenditures.

10. The simplest method to use is that of foregone revenue. This method measures the treasury’s revenue losses on the assumption that there is no change in the behavior either of taxpayers or the government as a result of the tax benefit granted. This assumption represents the main shortcoming of the method: obviously, a tax incentive is granted with a view to changing taxpayer behavior. For example, a reduction in the tax on a product is very likely to boost demand for that product and, given the household budget constraint, will tend to reduce consumption of other goods. Another shortcoming of this method is that it does not consider the effect of one tax expenditure on all the others. For example, if firms that have set up in a given region enjoy a reduction in the industrial tax and a similar reduction is then created to benefit firms that set up in other regions, it is very likely that future investment in the first region will be less than what it would have been in the absence of the second incentive.

11. The revenue gain method seeks to shed light on a question: how much would revenue rise if the legal provision that created a given tax expenditure was repealed? To answer this question one must consider the possible behavioral reactions of the taxpayer, and this requires a solid database and knowledge of the elasticities in the economy. In addition, one must make assumptions about the behavior of taxpayers and the government, which could have a significant impact on the estimate.

12. The third alternative is the equivalent outlay method. It seeks to estimate how much the government would have to spend through a subsidy or transfer in order to provide the taxpayer with an equivalent benefit, i.e., to leave the taxpayer with the same after-tax income that he or she enjoys through the tax expenditure that is being replaced.

13. Despite its shortcomings, the revenue foregone method is the one used by nearly all countries in the world due to its simplicity. A few countries use the revenue gain method and, according to OECD (2010), only Sweden estimates tax expenditures using the equivalent outlay method.

C. Defining Tax Benchmarks and Identifying Tax Expenditures Revenue from the main non-oil taxes

14. The share of non-oil tax revenue in total tax revenue is relatively small. The eight taxes shown in Tables 1 and 2 accounted for nearly all non-oil tax receipts in recent years. Nevertheless, despite the high rates of growth in non-oil tax collections in the last two years and the sharp drop in oil tax revenue in 2014, non-oil taxes were 9.3 percent of GDP in 2014, or the equivalent to about 30 percent of total tax receipts.

Table 1.

Angola: Revenue from the Main Non-Oil Taxes

(millions of Kz)

article image
Sources: Angolan authorities; and IMF staff estimates.

15. The tax expenditures dealt with in this chapter are only one of the factors behind the relatively small share of non-oil receipts in the total. There are other factors that are not considered in this chapter. First, there are indications of a high level of tax evasion, and the recently-created General Tax Administration (AGT) is already taking steps to control it. Second, the reduced rates for the consumption tax—2 percent applied to a long list of products and 5 percent on all taxable services with the exception of hotel services—are low. Third, the tax bases, especially for the consumption tax and the industrial tax, are narrow, underscoring the importance and indeed the urgency of a tax reform that would broaden the tax base.

Table 2.

Angola: Revenue from the Main Non-Oil Taxes

(percent of GDP)

article image
Sources: Angolan authorities; and IMF staff estimates.

Industrial tax

16. The structure of the industrial tax is similar to that of taxes on corporate profits in most countries. Residents are taxed on their worldwide income, and nonresidents on their territorial income. The normal rate, applicable to the great majority of sectors, is 30 percent. There is a “deemed profit” regime in which the tax base is the volume of sales of goods and services, with the rate set at 6.5 percent.5

17. Tax expenditures related to the industrial tax take various forms. Table 3 presents a wide-ranging but unlikely to be exhaustive list of such tax expenditures.

Table 3.

Angola: Tax Expenditures under the Industrial Tax

(Law No. 19/14)

article image
Sources: Angolan authorities; and IMF staff compilation.

Tax on labor income

18. The tax on work income applies to three groups of taxpayers: workers in a relationship of employment, self-employed workers, and individuals engaged in industrial and commercial activities. Each of these groups is taxed in accordance with rules exclusive to that group. In this chapter, we consider as benchmark the tax defined in Law 18/14, excluding its Chapter II which deals with exemptions, and article 7.2a (deduction of the mandatory social security contribution), as social allowances paid by the National Social Security Institute are not subject to taxation. As the counterpart, the amount of tax not collected on the income sources mentioned in Chapter II as well as the additional amount of tax that would result from not deducting social security contributions are considered to be tax expenditures.

Tax on capital income

19. The benchmark tax for the tax on capital income has all its characteristics defined in Presidential Legislative Decree no. 2/14, except its Chapter II, which deals with exemptions. This is a tax imposed on interest, dividends and other distributed profits, royalties and net capital gains and losses as well as on lottery prizes and earnings from gambling. The rates are 5 percent, 10 percent, and 15 percent, depending on the type of income. Tax expenditures are measured as the revenues not collected by virtue of the exemptions mentioned in Chapter II and in the private investment laws of 2011 and 2015, namely:6

  • Interest on merchants’ credit sales.

  • Interest on life insurance policy loans made by the insurance company.

  • Interest on instruments intended to foster savings, provided the capital does not exceed 500,000 kwanzas per person.

  • Interest on home-purchase savings accounts.

  • Reduction or exemption from the tax on profits distributed to partners by firms covered by the Private Investment Law for a term that, pursuant to the 2011 law, depended on the location of the project and, under the 2015 law, on the score obtained by the project in question.

Consumption tax

20. The consumption tax, as the name indicates, represents a tax on the final consumer. However, those liable to pay this tax are producers, importers, providers of taxable services, and those who sell or dispose of goods at public auctions. The tax base consists of all products, except non-processed products of agriculture, livestock, mining, forestry and fisheries activities, and the services contained in a specific list. In addition, exports, raw materials and subsidiary materials incorporated in the production process and capital goods and spare parts are exempt. The taxable value is the production cost for goods produced in Angola, the customs value for imported goods, the price paid for services, and the selling price at auctions. The production cost of goods produced in the country is defined as the cost of raw materials and intermediate inputs, labor, technology and other goods or services necessary for production, excluding the costs of distribution, transportation, insurance or others that are incurred after warehousing. The tax rate is 10 percent for most products, and 5 percent for all taxable services, except for hotel services, where the rate is 10 percent, but there are also rates of 2 percent, 20 percent, and 30 percent applicable to the projects listed in annexes to the consumption tax regulation.7

21. The consumption tax is likely borne by the final consumer, as intended, despite the fact that it is not paid directly by consumers. This is due to the fact that, by exempting producers’ purchases of inputs and capital goods, it avoids double taxation and, by taxing imports and exempting exports, it allows the producer to receive tax-free the full international market price for sales abroad and requires the consumer to pay the international market price plus the consumption tax (in addition to customs duties) when purchasing an imported product. In this way, it creates conditions that encourage the tax, for which the producer is responsible for paying, to be passed on to the final consumer.

22. The discussion in the previous paragraphs provides a roadmap for the identification of the tax expenditures related to the consumption tax. As this chapter adopts the legal approach, the revenue forgone because of the exclusion from the consumption tax of agricultural, livestock, forestry, fisheries and mining products does not constitute tax expenditure. Similarly, the difference between what would be collected through application of a 10 percent rate instead of 2 percent or 5 percent rate does not constitute tax expenditure. This is because, in contrast to general sales taxes, such as the VAT or the retail sales tax, the consumption tax in Angola is deliberately designed to play a dual role as a general sales tax and specific excise tax. Moreover, the exemptions for exports and goods required for production, although they are legally mandated exemptions, are in fact elements of the tax structure, necessary for avoiding multiple or cascading taxation and to give this levy the nature of a true consumption tax.8

23. Nevertheless, there are some tax expenditures relating to the consumption tax, although in terms of foregone revenue they are less important than exclusions and reduced rates. These tax expenditures flow from subjective exemptions granted under international agreements (i.e., diplomatic and consular missions and international organizations) as well as to persons (i.e., former combatants) who are given the right to import goods similar to the ones produced domestically, and to persons who pursue activities using artisanal processes.

24. In addition, the discussion above has implications for tax revenue and social policy.9 The consumption tax regulation ignores a big slice of the value of consumption in the country. In addition to excluding nearly all domestic production in the primary sector and a good portion of the tertiary sector, it excludes from the tax base a portion of the value of consumption of domestic industrialized products corresponding to producers’ profit margins as well as the entire value added by merchants. To this must be added taxation at reduced rates for all taxable services, except those in the hotel industry, and a good share of industrialized food products and of medications and pharmaceutical articles. Using input-output tables for the year 2007 (the latest available), a rough estimate of the additional revenues that could be collected if these items were all taxed at a rate of 10 percent is of the order of 2.7 percent of GDP.10

25. A portion of the potential revenue forgone can be traced to an inefficient social policy. As consumption taxes tend to be regressive, Angola and many other countries exclude, exempt, or apply reduced tax rates to certain goods and services that are deemed essential and that represent an important fraction of spending on consumption by the poorest families. In fact, this type of social policy does benefit those families, and it benefits them proportionately more than wealthier families, thereby mitigating the regressive nature of the tax. Nonetheless, this type of social policy is extremely inefficient, for the wealthiest families consume preferentially taxed products and services in much greater quantities and values. Consequently, it is the wealthiest families who appropriate the bulk of the amount that the government forgoes.11 Government action would be more beneficial for the poorest families if the foregone taxes were to be collected and their revenues used in well-targeted programs to combat poverty.

The stamp tax

26. The benchmark for the stamp tax is as defined in Presidential Legislative Decree 3/14 with the exception of its Article 6, which deals with exemptions. Tax expenditures are equal to the tax revenues forgone by virtue of the exemptions under Article 6 and the tax exemptions or reductions contained in other laws, which have their primary impact on revenues from import taxes. The customs tariff schedule does not grant any benefits under the stamp duty and provides for it to be collected on all imports, even those exempt from import duties, at a rate of 1 percent of the customs value. Article 6, however, provides exemptions for: the State, its services, establishments and agencies, except for public enterprises; public welfare and social security institutions, associations of public utility and religious institutions, except for their economic activities of a business nature; and various financial and insurance transactions, labor contracts, and exports.

Import duties and general customs fees

27. The benchmarks for customs duties and for general customs fees are defined in the customs tariff schedule with the adjustment noted below. The new customs tariff schedule12 adopted the 2012 version of the World Customs Organization’s Harmonized System Nomenclature (HS 2012), replacing the 2007 version. The customs tariff schedule of Angola added a chapter to those listed in the HS 2012, namely, Chapter 98 on goods imported for special purposes. It contains exceptions to the provisions of the other chapters and, consequently, the tax expenditures relating to import duties, as well as those relating to the consumption tax with respect to imports. Concerning the general customs fees, the customs tariff schedule provides no exemptions, even for the State. Exemptions may, however, exist by force of law or of agreements, conventions or contracts that are binding on the government.

28. Imports under Chapter 98 were responsible for 11.7 percent of the total value of imports in 2014. A small portion of that value corresponds to temporary imports. Exemptions for temporary imports are not tax expenditures. All other exemptions under that chapter are tax expenditures. In that chapter, the column showing import duty rates indicates that the imports are free. Thus, to calculate the tax expenditure corresponding to a good classified in that chapter, there is a need to apply to the import value the rate that is shown for the same good in the chapter where it would be classified, if it were not imported for a special purpose.

29. Tax expenditures relating to import duties and general customs fees are not limited to the exemptions contained in Chapter 98. There are a great number of exemptions specified in legal rules and administrative acts that are not related to the tax system. Circular 131/DPP/SNA/2014, which defines codes of regimes, procedures, treatment and exemption, mentions nearly 50 codes of exemption.

Urban property tax

30. The law on the urban property tax provides few exemptions that are granted to: 13

  • The State, public institutes and associations that enjoy public utility status.

  • Foreign governments with respect to buildings intended to house their diplomatic or consular establishments, provided there is reciprocity.

  • Religious institutions, for properties devoted exclusively to worship.

D. Concluding Remarks

31. This chapter has highlighted the importance of tax expenditures budgets as a tool to provide policymakers and legislators with information on the cost of tax benefits in terms of forgone revenues. An understanding of that cost, which remains hidden in Angola throughout the traditional budgetary process, is important for taking decisions on whether to maintain or eliminate those benefits. This can prevent the tax system from deteriorating over time under the accumulation of tax benefits that, while they may have been useful when they were created, have ceased to be important in terms of the country’s current objectives. Consequently, tax expenditure budgets allow public funds to be put to more efficient use.

32. The Angolan authorities are thus encouraged to adopt:

  • The legal approach to defining its benchmark taxes. As the taxes currently in force bear little similarity to the theoretically ideal taxes discussed in the public finance literature, selecting benchmark taxes via the conceptual approach would not produce realistic results.

  • The forgone revenue method for estimating tax expenditures. Despite its shortcomings, this method is much simpler and requires much less economic information than the other methods, and has therefore been adopted by the great majority of countries.

33. The methodological base will only be useful in practice if it is associated with a database that allows the methodology to be applied so as to obtain the desired estimates. It was not possible, unfortunately, to obtain, at the time of the 2015 Article IV mission, the data that would have allowed an estimation of tax expenditures. Customs data are available, but the AGT officials responsible for extracting and preparing the data were not able to provide reliable data. With respect to domestic taxation, there is no information available that could be used for estimating tax expenditures. There are at least three reasons why not:

  • The beneficiaries of tax incentives, who have no tax to pay, tend not to comply with their accessory obligations, such as submitting annual tax returns, and they have not as yet been forced to comply with those obligations.

  • The tax return forms are not sufficiently detailed to yield the relevant information.

  • The tax returns are submitted across the entire country, many of them in paper format, and the scanty information that could be available is not compiled.

34. In the absence of a proper database to estimate tax expenditures in Angola, this chapter has defined benchmarks for each of the main non-oil taxes and identified the existing tax expenditures in the country. The lists are fairly comprehensive but they are certainly not exhaustive, as many tax expenditures are created in non-tax legislation and administrative acts, agreements, and contracts. Although it has not been possible to assess their importance in terms of value, there are many exemptions that generate tax expenditures through import duties. There are also significant tax expenditures in the industrial tax. Therefore, it is recommended that:

  • Tax expenditures associated with import duties should be not only pinpointed but also analyzed carefully.

  • After taking steps to improve the availability of data, the AGT should expand the study of tax expenditures under the industrial tax to include other tax benefits beyond those granted by the National Private Investment Agency (ANIP).

  • Tax expenditures under the tax on labor income and the tax on capital income should be scrutinized over the medium term.

35. In the case of the consumption tax, much more important than the tax expenditures associated with it is the revenue forgone due to the narrowness of its tax base and the existence of reduced rates for services, processed foods, and drugs. The revenue forgone estimate by virtue of these two reasons is in the order of 2.7 percent of GDP. Thus, revenues from the consumption tax, which currently amount to 1.8 percent of GDP, could reach 4.5 percent of GDP.

36. The expansion of the tax base that could generate the revenue increase estimated above appears equivalent to replacing the consumption tax with a VAT and product-specific excise taxes, which is already called for in Angola’s tax reform plans. Replacement with a modern VAT seems to be a better alternative than attempting to reform again the consumption tax, which is a tax handle that has been largely abandoned around the world. For this reason, it is recommended that:

  • Work should start immediately to make careful preparations for implementing a modern VAT and product-specific excise taxes.

  • The VAT to be adopted should have a single positive rate and zero rate applicable exclusively to exports.

  • The VAT should contain few exemptions (preferably none).

  • The specific excise tax should be confined to a limited number of products, preferably tobacco and its derivatives, alcoholic and nonalcoholic beverages, fuels, and vehicles.

  • The products included in the excise tax base should also be subject to the VAT.

  • The tax base for the specific excise tax should include, in the case of imports, the amount of import duties and general customs fees.

  • The tax base for the VAT should include, in the case of imports, the amounts of import duties and general customs fees and, in all cases, the amount of the specific excise tax.


  • CIAT (2011), Handbook of Best Practices on Tax Expenditure Management: An Iberoamerican Experience (Panama: CIAT).

  • IMF (2007), Manual on Fiscal Transparency, revised edition (Washington: International Monetary Fund).

  • OECD (2010), Tax Expenditures in OECD Countries (Paris: Organization for Economic Co-operation and Development).

  • Villela, Luiz, Andrea Lemgruber, and Michel Jorrat (2010), “Tax Expenditure Budgets: Concepts and Challenges for Implementation” (Washington: Inter-American Development Bank).

    • Search Google Scholar
    • Export Citation

This chapter was prepared by Ricardo Varsano (FAD expert).


This section and the next use material contained in CIAT (2011), IMF (2007), OECD (2010), and Villela, Lemgruber and Jorrat (2010).


This definition, similar to that contained in IMF (2007), is one of the various definitions of tax expenditures found in the literature.


A few countries, among them Germany, use a third approach, that of the analogous subsidy, which is not considered in this paper.


The “deemed profit” regime can be used by commercial and single-person companies if their corporate capital is less than 2 million kwanzas, if their total annual profits are less than 500 million kwanzas, and if they do not keep organized accounts.


It should not be considered as tax expenditure the foregone revenue due to the exemption granted for profits or dividends distributed to corporations, subject to the industrial tax, which hold no less than 25 percent of the capital of the entity that is distributing the profits. These profits comprise the basis of the industrial tax. Therefore, without the tax on capital income exemption, these profits would be taxed twice.


Presidential Legislative Decree 3-A/14.


In the case of the VAT, as with the Angolan consumption tax, it is necessary to tax imports and exempt exports for it to be a true consumption tax. The VAT credit applicable to inputs replaces the exemption so as to avoid cascading taxation. In the case of the retail sales tax, it is not necessary to tax imports, because they are taxed once they enter the retail trade sector, nor is it necessary to exempt exports, as they are rarely conducted by retailers. On the other hand, it is necessary to exempt purchases that producers make on the retail market.


There are also implications for competition between imported and domestic goods, which are not taken into consideration here.


This estimate is necessarily rough, because the products and services in the input-output tables do not correspond precisely to the products and services exempt from the tax, and also because the tables are for the year 2007 and the structure of Angolan economy has undergone important changes since then.


Data on the consumption of family units obtained from the National Statistics Institute (INE) show this fact clearly, in line with what emerges from similar data for other countries.


Rectification 1/14 of Presidential Legislative Decree 10/13.


Article 5 of Legislative Act 4044 of October 13, 1970, in the wording given by Law 18/11.