The purpose of this note is to provide a framework for improving tax policy design in fragile and conflict-affected states, which face political and institutional constraints. This note begins with an overview of experiences in revenue mobilization in fragile states, including relative to other country groups—in particular, nonfragile states and formerly fragile states; that is, countries that exited fragility during the period under study. A discussion follows of how the principles of tax policy design should be applied in fragile states, particularly the relative importance of the revenue objective vis-à-vis other objectives, such as equity and efficiency. The two sections that follow provide guidance on tax policy design in the emergency and consolidation phases, respectively, and discuss how governments can use tax policy to transition from one phase to another, eventually overcoming fragility. The note concludes with key lessons and a set of guiding principles for tax reform in fragile states.


The purpose of this note is to provide a framework for improving tax policy design in fragile and conflict-affected states, which face political and institutional constraints. This note begins with an overview of experiences in revenue mobilization in fragile states, including relative to other country groups—in particular, nonfragile states and formerly fragile states; that is, countries that exited fragility during the period under study. A discussion follows of how the principles of tax policy design should be applied in fragile states, particularly the relative importance of the revenue objective vis-à-vis other objectives, such as equity and efficiency. The two sections that follow provide guidance on tax policy design in the emergency and consolidation phases, respectively, and discuss how governments can use tax policy to transition from one phase to another, eventually overcoming fragility. The note concludes with key lessons and a set of guiding principles for tax reform in fragile states.


The purpose of this note is to provide a framework for improving tax policy design in fragile and conflict-affected states,1 which face political and institutional constraints. It is based on the IMF Fiscal Affairs Department’s research and experience in providing these states with technical support on tax policy. It addresses the following questions:

  • What objectives and principles can tax policy follow in a fragile and conflict-affected state?

  • How should taxes be designed in an emergency situation?

  • How can tax policy evolve as a fragile state recovers from an emergency and, eventually, leaves its fragile situation?

Overcoming fragility is a long and difficult process. It involves all branches of government and a combination of policy actions which raise distinct implementation challenges. Tax policy plays an important part and is never easy to integrate due to difficulties in weighing its various objectives during the fragility phases. These objectives include the following:

  • Revenue mobilization: Sufficient revenue is needed to finance the proper functioning of the government, including tax administration, which constrains tax policy design.

  • Efficiency: In a fragile state’s unstable environment, a transparent, non-distortive tax system provides the visibility private agents need to invest and help the country regain economic momentum.

  • Equity: A tax system that is perceived as just and inclusive contributes to the reestablishment of the state’s legitimacy, which is key to its stability and efficiency in the long term.2

This note begins with an overview of experiences in revenue mobilization in fragile states, including relative to other country groups—in particular, nonfragile states and formerly fragile states; that is, countries that exited fragility during the period under study. A discussion follows of how the principles of tax policy design should be applied in fragile states, particularly the relative importance of the revenue objective vis-à-vis other objectives, such as equity and efficiency. The two sections that follow provide guidance on tax policy design in the emergency and consolidation phases, respectively, and discuss how governments can use tax policy to transition from one phase to another, eventually overcoming fragility. The note concludes with key lessons and a set of guiding principles for tax reform in fragile states.

Level and Composition of Tax Revenue in Fragile States

It can take a country several decades to exit fragility, from the emergency or conflict phase to the consolidation phase. This section examines how fragility and the journey toward normalcy affect the level and revenue structures of fragile and conflict-affected states. The lessons learned from this analysis are used later in the note to help provide a framework for tax policymaking in fragile states.3

What Are Fragile States’ Experiences with Revenue Mobilization?

Fragile states display low tax ratios (on average, 14 percent, as opposed to 18 percent in non-fragile states). These ratios are frequently extremely low, especially after a conflict (the tax-to-GDP ratios of Afghanistan, Chad, the Democratic Republic of the Congo, Guinea-Bissau, Iraq, Libya, Myanmar, and Timor-Leste have been less than 3 percent for several years).

Several fragile states with low tax ratios have been able to compensate for this difference in tax revenue, thanks to nontax revenues from extractive industries.4 Natural resources accounted for 3.4 percent of GDP in the median resource-rich fragile state in recent years. Half of the resource-rich fragile states that later exited fragility collected more than 20 percent of their GDP from natural resources in the early 2000s. However, natural resources have been a mixed blessing for countries lacking strong institutions, given the decline in commodity prices in recent years. Several countries that were not fragile in 2008 but relied heavily on natural resource revenues (in addition to sharing some other characteristics observed in fragile states, such as low GDP per capita, low growth in GDP per capita, and weak tax revenue) became fragile in the subsequent decade.

The general weakness of tax revenue in fragile states reflects lower domestic tax receipts to GDP for all tax revenue sources, except international trade (Figure 1; Annex 2, Table 2.1), which accounts for 19 percent of total taxes, on average, compared with 15 percent in nonfragile states. A few fragile states also made greater use of unorthodox, and therefore unclassified, taxes. Another, more surprising feature of fragile states is how little most of them rely on the taxation of goods and services and sales tax (for example, value-added tax [VAT]). This contrasts with their reliance on direct taxation, especially personal income tax, and differs from the usual negative relationship between the share of indirect taxes in total tax receipts and income levels.5

Figure 1.
Figure 1.

Revenue Structure of Fragile and Nonfragile States

(Simple average for 2013–2016)

Citation: IMF How To Notes 2019, 004; 10.5089/9781513508153.061.A001

Source: IMF, World Revenue Longitudinal Database (WoRLD).Note: Figures are simple averages for each category of country. Figures may not add up because not all series are available every year for all countries. Resource revenue as a percentage of tax revenue is shown for countries that derive 20 percent or more of their net exports from nonrenewable natural resources.

Another feature fragile states have in common is the large variety in their tax-to-GDP ratios and tax revenue composition. This is a characteristic of fragility, which, in addition to economic factors, reflects security and political features. Consequently, tax development and design, and how they interact with capacity building, depend even more on country-specific issues in fragile states than elsewhere.

Are Tax Levels and Composition Different in Formerly Fragile States?

Fragile states that became nonfragile in the decade following 2008 were able to mobilize more revenue—about 5 percentage points of GDP—than those that remained fragile, and revenue from natural resources played an important role in several countries (Figure 2; Annex 2, Table 2.2).6 In the early 2000s, Angola and Niger raised more than 20 percent of GDP from commodities, a share that has fallen significantly since then. Other countries, such as Cambodia and Rwanda, were able to increase the contribution of natural resources to their budgets from initially insignificant levels, which helped them overcome fragility.

Figure 2.
Figure 2.

Revenue Structure of Permanently versus Formerly Fragile States

(Simple average for 2013–2016)

Citation: IMF How To Notes 2019, 004; 10.5089/9781513508153.061.A001

Source: IMF, World Revenue Longitudinal Database (WoRLD).Note: Figures are simple averages for each category of country. Figures may not add up to the total tax revenue because not all series are available every year for all countries. Resource revenue as a percentage of tax revenue is shown for countries that derive 20 percent or more of their net exports from nonrenewable natural resources.

More generally, formerly fragile states’ budgets relied on relatively robust tax revenues (19 percent of GDP around 2006, compared with 11 percent, on average, in states that remained fragile). For the most part, they were able to preserve their revenue performance throughout the decade that followed, despite decreased receipts from natural resources in some countries. All formerly fragile states were able to collect more than 15 percent of their GDP through taxation, and half collected more than 20 percent in some years of the past decade (for example, Angola and Nigeria, where tax revenues collapsed with oil prices in 2014). This is consistent with the empirical finding of an apparent tipping point of about 13 to 15 percent tax revenue to GDP, below which countries have great difficulty rebuilding institutions to overcome fragility.7

Tax composition in formerly fragile states (compared with permanently fragile states) displays the following characteristics (Figure 2):

  • The share of goods and services taxes in total taxes is predominant and increases as the country’s situation improves, eventually reaching a level comparable to that of nonfragile states.

  • The contribution of trade taxes is relatively small and diminishes over time.

  • Some countries rely more on “other taxes”—namely, taxes not classified as “income, payroll, and property taxes,” “goods and services taxes,” or “trade taxes.”

Principles and Objectives for Tax Policy in Fragile States

Tax Policy Priorities

Revenue mobilization is tax policy’s overarching priority in most fragile states; it allows them to finance the reconstruction of their capacity to deliver basic services and infrastructure. To cross the 13 to 15 percent tax-to-GDP tipping point, a prerequisite for overcoming fragility, some fragile states must increase tax revenue by more than 30 percent. This is a challenge for any country, and a particularly daunting one in a fragile situation, where severe institutional capacity constraints limit what tax policy can achieve. That said, the relationship between tax policy and state building goes deeper than just revenue mobilization. Revenue mobilization depends critically on the government’s power and legitimacy; the equity and efficiency impacts of tax reform also determine their robustness to economic and political shocks in the medium term.8

In addition to revenue mobilization, the simplicity, inclusiveness, transparency, and equity of the tax system all play important roles in restoring the state’s capacity and legitimacy. Relative to economic efficiency, these elements must sometimes be given more weight than they would in nonfragile states,9 since their benefits in helping overcome fragility may outweigh the cost of distortive taxes.

This is not to say that efficient taxes have no role in fragile states. When such taxes are feasible and compatible with other objectives, they should have priority. Since they are less likely to require further reform than other taxes as countries exit fragility, efficient taxes also serve as useful anchors for investors’ and other economic agents’ long-term expectations.

More generally, the stability of taxes and visibility of their evolution are desirable for attracting investment and motivating the government, tax administrations, and the public to commit to the tax system. In this way, the costs of administering and complying with taxes are kept low. Once the government is less fragile, setting a medium-term revenue strategy—including coordinating international support to build the state’s revenue capacity—could be the key to achieving these desired objectives.

Certain objectives assigned to the tax system in nonfragile states could also be pursued in fragile states. For example, discouraging behavior that harms oneself, others, or the environment can sometimes be achieved through simple efficient taxes, such as excises on imports of alcohol or luxury cars. However, progressive comprehensive income taxation, or comprehensive wealth taxation, can distract governments from rebuilding basic institutional capacity. Such objectives can be addressed more effectively in the medium term as countries become less fragile.

Desirable Characteristics of the Tax System

Given the objective of revenue mobilization and the diversity of fragile states, pragmatism must prevail when it comes to which tax system is desirable in each country (given the nature of fragility and other country-specific factors). Against a general backdrop of weak institutional capacity, policymaking should follow some general principles, including simplicity, transparency, and operability, which translate into a tax system characterized by the following:

  • A small number of taxes: The limited administrative capability should focus on levying a few effective taxes.

  • Easy-to-assess tax bases: Sophisticated tax bases may help reduce economic distortion, but they are costly to assess and audit and may end up opening the way to tax bargaining and avoidance.

  • Few exemptions: Exemptions create loopholes through which parallel markets develop, which is especially costly in a country where an informal economy may already be prevalent, undermining both the government’s revenue and its legitimacy.

  • A focus on large taxpayers and easy collection points: Borders, large companies (especially in extractive industries), and workers in the civil service and formal sector are easier for the tax administration to monitor.

  • A single or very few rates for each tax: This eliminates the bargaining and avoidance that often accompany tax base requalification—to exploit the tax arbitrage that low rates offer.

  • Wide recourse to the presumptive taxation of small taxpayers: For the tax administration to monitor taxpayers effectively under the standard regime, and for the system to be sufficiently inclusive, smaller taxpayers should be taxed on a presumptive or lump-sum basis.

  • Preparation for the future: Even if a fragile state’s tax system cannot comply with the canons of taxation theory, it is preferable that it allows for gradual refinement as the situation improves rather than require a complete overhaul.

The Trade-off between Stability and Reform

As in all countries, reforming tax policy in a fragile state has a cost. Most fragile states’ tax systems were established under assumptions that are overly optimistic about institutional capacity. The government faces a trade-off between keeping the existing tax laws, which the taxpayers and administration are accustomed to, and engaging in reforms that would improve the tax system’s functioning and revenue. The latter might involve significant short-term legislative, administrative, and political costs, even if they can be alleviated by donor resources and technical support.

National authorities should carefully consider the decision to reform the tax legislation. Although there is no one-size-fits-all recommendation, the following points may be considered:10

  • Each country should balance the short-term costs and the short- and long-term benefits of reforming its tax policy. Given fragile states’ vulnerability and possible instability, the short-term consequences should be given more weight than they would in a nonfragile state.

  • Not all taxes need to be reformed simultaneously. Although a broad, up-front reform could yield earlier and more consistent benefits than gradual changes, it also runs the risk of stretching the administration beyond its current capacity and disrupting its effective elements.

  • With a gradual approach, the more dysfunctional or harder-to-administer taxes should be reformed first. For example, if, during a crisis or period of fragility, the customs administration functions better than the internal tax administration, and if tariff policy is adequate, reforming domestic taxes may take priority over changing taxes at the border.

  • Dysfunctional or not, there is no point in maintaining small taxes during a time of scarce administrative capacity. Individual taxes that garner less than 0.05 percent of GDP can be safely suspended, eliminated, or replaced by synthetic taxes to redirect existing administrative capacity toward monitoring taxes that bring more resources to the budget.

  • In fragile states, simple regulations that can be passed directly by the government may be initially preferable to complex legislative changes, which may involve political disputes and economic uncertainty. However, these powers must be kept in check to prevent the government (or ministers) from unbridled discretion over key aspects of tax laws, such as tax rates, and preferential treatment, such as deductions, exceptions, and exemptions.

  • If much of the tax system has already collapsed, after a war for example, the political cost of drafting new tax legislation is relatively low, and is an option that should be seriously considered. This would be better than returning to a prewar system, which is often intrinsically imperfect. Thus, extreme fragility can present an opportunity for much-needed in-depth reform. Rwanda, for example, benefited from the early adoption of a tax system redesigned according to the above principles. The redesigned system subsequently contributed to the stability of the economy, the attraction of foreign investors, and, eventually, the end of fragility.

The rest of this note elaborates on desirable features of the main groups of taxes in fragile states. It follows a stylized chronology according to which, after having reached very low institutional capacity during a crisis, fragile states experience a first stage of rapid economic and political improvement (the emergency phase of fragility) for a few years. During the second stage, they reach a plateau, which may last one or more decades (the consolidation phase), until they have recovered sufficiently to be reclassified as nonfragile. Passing from the emergency to the consolidation phase may be characterized by varying revenue levels and composition, reflecting not only economic structure but policy and institutional choices as well.

Most tax policy recommendations are designed to assist countries during the emergency phase, when progress is needed to boost resource mobilization beyond the tipping point. During the subsequent consolidation phase, efforts can be focused on improving the transparency, equity, and inclusiveness of the tax system, parallel to the tax revenue objectives.

The depth of a country’s crisis when it reaches the nadir of its government’s capacity depends on its specific path toward fragility. An extreme case would be a country that enters fragility as a result of a war, after which budget revenue is so low that resource mobilization is the single objective of tax policy. Such conditions warrant short-term emergency tax measures, which are discussed in the next section. Fragile states may benefit from adopting some of these measures even during less stressful times, given their focus on revenue mobilization.

How to Design Taxes in an Emergency Situation

The Case for an Approach by Collection Point

Tax administration capacity and other factors affecting the collection of taxes, such as education level and business organization, weigh more heavily on tax design in the presence of fragility. During an emergency, when revenue mobilization is the overwhelming concern, the priority should be to assess the available administrative capacity and focus it on effective collection points—namely, the actual locations or chargeable events that are easiest to monitor. Depending on the geography and security of the fragile state, effective collection points may include the following:

  • The border or checkpoints where goods enter the territory under government control (for example, Afghanistan, Central African Republic)

  • The largest companies in the capital city and other cities under the government’s control (for example, Somalia)

  • Mining and extraction sites under government control (for example, Democratic Republic of the Congo)

  • Wages in the civil service and formal sectors (for example, Iraq)

  • Land, if a cadaster is available

During the emergency phase, reorganizing the tax system around these collection points can help optimize receipts when there is administrative scarcity. The system would, however, have to evolve toward a more economically efficient design as soon as the situation improves.

How to Tax at the Border

In an emergency, and especially if a fragile state must begin taxation without a full-fledged administration or tax law to refer to, a single synthetic border tax allows the customs administration to collect revenue at the lowest administrative cost. The gross value of imports is the preferred base for a single tax, but administrative prices may also be applied to goods whose quantities are easier to assess than their value. The rate schedule of a single border tax should reflect, among other factors, the capacity of the customs administration. A moderate flat rate (for example, 20 percent) on all imports would be viable briefly during a state of high emergency and limited administrative capability, but a simple three-rate schedule would enable further development of the tariff system and domestic taxes. To illustrate, a single border tax could be designed with the following elements (similar to the schedule Cambodia adopted in 1994):

  • Low rate—for example about 10 percent, applicable to unprocessed food, medications, agricultural inputs, and equipment goods

  • Standard rate—for example, about 30 percent, applicable to most other consumption goods, raw materials, and intermediate consumption

  • High rate—for example, about 50 percent, applicable to a short list of luxury consumption goods, petroleum products, passenger vehicles, alcohol, and cigarettes

A synthetic border tax would approximate and prepare for a more standard tax system, in which imports are subject to customs duties and fees, as well as VAT and excises. This subsequent tariff and consumption tax system comprises the following elements:

  • A tax rate of 10 to 15 percent on final consumption; low tariffs, with little or no rate differentiation; and excise rates to replace high tariff rates on luxury goods and other selected items (cars, petroleum products, alcohol, cigarettes)

  • Lower tariff (relative to the single border tax) and consumption tax rates (perhaps some exemptions) on staple foods and medications, justified for equity reasons, if social transfers cannot be established

  • Exemption of the agricultural sector, for administrative reasons, given its wide informality in low- and middle-income countries; alternatively, the consumption tax can be designed to exclude, through its registration threshold, most firms in the sector

  • No domestic taxation (or tax deductibility) of capital equipment, with the expectation that the administration will promptly be able to handle the deductibility of the taxes paid on raw and intermediate inputs of formal firms

Under both the synthetic and subsequent systems, exemptions should be kept to a minimum. Some importers, such as the United Nations, Red Cross, and Red Crescent, are exempt from import duties, and sometimes from other taxes, as part of the international obligation to facilitate their operations, principally related to natural disasters and humanitarian relief. These privileges should also be granted under a synthetic border tax. Although other donors often request the same privileges to operate in fragile states, efforts should be made to persuade them that it is counterproductive: it deprives the government of revenue, often as necessary as in-kind assistance itself. More critically, imports under donor privilege may leak into a parallel market, thereby undermining the tax base. Similarly, the imports of state-owned enterprises should be fully taxed, although this does not in itself increase the government’s revenue.

How to Tax Companies

In many developing economies, fragile or not, the largest 1 percent of businesses account for more than half of total tax payments, including those imposed at the border.11 Focusing the scarce administrative capacity of a fragile state on these taxpayers is warranted, especially in the emergency phase, when revenue concerns take precedence over other policy objectives. A single synthetic tax can be designed to proxy or prepare for the adoption of the main taxes large firms would generally have to pay under a more complex tax system; namely, VAT and corporate income tax. VAT would be based on value added and income tax on profit.

Which Base Should Be Used?

A convenient base for such a tax—and by far the most widely used—is a firm’s turnover. It is a broad base, defined as the value of a firm’s annual gross sales (possibly including sales of ancillary services, perhaps with adjustments for returns and irrecoverable receivables). It is easy for the firm to assess and for the tax administration to audit, using invoices, and therefore limits compliance and litigation costs. In addition, it anticipates a more granular future tax system, since turnover broadly coincides with the gross base of a consumption tax on goods in the retail sector. Turnover taxes are also used in most countries as the base for the presumptive taxation of small businesses, for both VAT and income tax.

Determining which firms must pay the synthetic tax depends on the country’s administrative capacity and economic fabric. The largest formal firms in the economic capital and other cities under the government’s control should certainly be included. Smaller firms, whose tax liabilities would not cover the costs of credibly monitoring their tax bases, or the cost of diverting auditing resources from the largest firms, could be excluded. Moreover, to reduce both tax avoidance and economic distortion, the line between firms that are liable and those that are not should be drawn in a manner that does not allow taxpayers to shift the tax base away from the synthetic tax.

In practice, a firm can be considered taxable if its turnover exceeds a particular threshold, it belongs to a particular sector, or it has some distinct legal form:

  • The turnover threshold should be set high enough to enable the tax administration to monitor the tax base effectively. While the optimal threshold depends on each country’s particular features, experience suggests that it is often set too low in fragile states. Governments tend to assign excessive value to catching as many firms as possible in the tax net and to overestimate their tax administrations’ capacity. This may result in the dilution of administrative efforts, inadequate monitoring of the largest enterprises, and, eventually, lower tax receipts and tax law credibility. Therefore, it is usually preferable to err on the high side when setting the initial threshold, with a view to reducing it gradually as the tax administration’s capacity improves. Without being overly prescriptive, setting an initial turnover threshold that is 50 times above the national GDP per capita should not be considered outlandish in the emergency phase, and significantly higher figures can be envisaged in some cases.12

  • A few sectors, such as banking and telecommunications, may account for a large share of value added and profits of nonresource private firms. Firms in these sectors are formal by nature, as they must register and are subject to specific regulatory requirements for running their businesses. Therefore, they can be taxed regardless of their turnover levels. However, specific aspects of their activities (for example, the banking sector’s relationship between turnover and value added and profits, which is different from that of other sectors) justify their being taxed differently from other firms. Another justification for such taxation stems from their ability to extract some rents (for example, through bank licenses and spectrums and infrastructures for telecommunications, which give operators the benefits of limited competition).

  • Finally, based on the country’s legal business framework, corporations and other forms of limited liability companies (and some types of partnerships) should be taxed without regard to turnover and sector classification. Such legal forms of business organization presuppose adequate levels of accounting and financial reporting capacity, and these firms are likely to be captured under the first two dimensions above. However, it may be necessary to include them explicitly if the turnover threshold is set too high.

Which Rate Should Apply?

A single-rate tax on turnover must balance several concerns. To yield revenue, the rate should not be set too low. To deter evasion, reduce distortion from differences in profit margins, and limit distortion between liable and nonliable firms, it cannot be too high. Moreover, the rate should reflect the revenue expected from the more granular future tax system targeted after the emergency. Finally, the rate should account for the deductibility (or nondeductibility) of the tax paid on imported or domestic purchases:

  • The synthetic tax rate should be kept relatively low if deductibility is not allowed—say, because of compliance costs. A rate of about 7 to 14 percent, for example, would mimic (1) a future consumption tax, initially set at a low level (5–10 percent); (2) a future presumptive enterprise income tax, calculated to be equivalent to 2–4 percent of turnover; and (3) a rough provision for taxes already embedded in the price of inputs or paid at the border.

  • The rate can be set at a higher level if the tax paid on certain inputs—for instance, capital—can be deducted. In most cases, keeping it under 15 percent would be justified by the initially low consumption tax and its role in discouraging avoidance.

Although the turnover tax rate differentiates between the retail and other sectors (or between manufacturing and services) to account for different shares of intermediary inputs in many fragile and nonfragile states, a synthetic tax need not be differentiated during an emergency. Retailers (or service providers) are unlikely to be among the largest firms and would have already been taxed on imported goods they resell in the domestic market. In rare cases, large retailers (or service providers) may be taxed at higher rates on their turnover if they have some monopoly powers.13 Alternatively, they could break themselves up into parts, each below the threshold, without significantly damaging the economy.

Taxing the Banking Sector

Special provisions should be made for the banking sector. Banks’ turnover is driven by loans and deposits and does not reflect the value of the services they sell (nor their value added or profits). In addition, given banks’ accounting and reporting obligations, their activities and possible tax bases can be controlled at lower additional costs than in other sectors. Rather than subjecting banks to synthetic tax, taxation could consist of the following:

  • An income tax, which, when taxable profits are not readily defined or available, can be calculated as a cash flow tax. The calculation would be based on the difference between the interest received on loans plus fees for services, on the one hand, and the interest paid on deposits plus capital investments, on the other, at a rate approximating a future corporate income tax rate (for example, 20–25 percent).

  • A tax based on banks’ value added, which may be defined either as the sum of the wage bill and cash profits or as the difference between the interest the bank receives on loans and the interest it pays to depositors, at a rate approximating the normal rate of a possible future VAT (for example, 10–15 percent).14

Taxing Natural Resources

A resource-rich fragile state in the emergency phase might be inclined to tap extractive industries for more revenue. Besides their revenue potential, there are good reasons to use natural resources to help fund the budget over and above their contributions through general taxes.15 One reason is that the resource does not reflect any private agent’s economic efforts and can therefore be considered the property of the public at large. Furthermore, resource-rich countries usually need private investors to extract and exploit natural wealth efficiently. However, property rights for the resources, once allocated, often generate rent that is above the normal return on invested capital. A tax on this rent is one of the least detrimental to the economy.16 However, overreliance on resources and the volatility of commodity prices make resource-rich countries as prone to fragility as other countries, if not more.

Maintaining the existing tax and fee arrangements between the government and extractive firms could be the best route in the emergency phase, provided they generate reasonable revenue shares for the government and can be (easily) managed by the administration. This does not mean there is no scope for improving natural resource taxation in fragile states. In fact, given its importance to recovery, prioritizing reform of the taxation of the sector has been the recommended strategy in many cases. However, unless there is obvious leeway to enforce improvements realistically, these reforms are often better left to the future, after the emergency pressure has subsided somewhat.

Taxing the Telecommunications Sector

Firms in the telecommunications sector benefit from limited competition arising naturally from economies of scale attached to a network industry. These benefits also come from a licensing process that allocates spectrum or access to the existing physical infrastructure. There is a rationale for taxing the rent generated through the government’s license sales or through specific taxes.

A recurrent tax on telecommunications activity may be preferable to a one-time licensing fee, as some telecommunications operators are usually already licensed, and new entrants may be in a strong bargaining position vis-à-vis the government when negotiating access. This is because they have more information regarding future rents and because the government is fragile.

Different methods are used to tax telecommunications firms. In a very fragile country, the most convenient, though not the least distortive, method would involve additional taxes on turnover, to be paid in addition to the common synthetic taxes applied to enterprises in general, and equivalent to excises on calls.17 Given the importance of telecommunications in the production process, especially for foreign investors, and the relatively high price elasticity of calls, the additional telecommunications turnover levy should be kept at a moderate level (about 10 percent) until more growth-friendly taxation of operators’ rent can be established.18

Should Exceptions Be Made for Specific Sectors?

Fragile states may allow sectoral taxes to proliferate, in addition to general consumption and income taxes. This sometimes happens under the authority of line ministries or government agencies, which find it convenient to finance themselves directly, especially if the Treasury’s single account is not functioning to their satisfaction. Such taxes can be presented as fees for services, compensation for negative externalities, or taxation of sectoral rents. Similarly, and often simultaneously, fragile states are tempted to offer tax advantages to firms investing in sectors considered important to structuring the economy or kick-starting a recovery. This fragmentation in the approach to tax policy is undesirable on many levels:

  • It results in an inefficient fragmentation or duplication of tax responsibilities between the Ministry of Finance and other sectoral ministries, or the ministry responsible for investment—hence hindering coherent policymaking.

  • It leads to distortion (sometimes intended, but in most cases, not fully acknowledged) and opens up room for tax shifting (for example, through transfer prices) among firms facing a weak administration. The budget loses revenue as a result.

  • It creates opportunities for firms to bargain with the government over their tax obligations, which may be more profitable than developing their production, and which may easily absorb too much of the government’s limited capacity.

  • Because sectoral provisions will (inevitably) prove inefficient at certain points, they will be reversed.19 Therefore, special provisions tend to attract selective investors intending to make quick profits in the short term, rather than stable investors who contribute to long-term recovery.

Fragile states should limit sectoral provisions to those that are efficient. Such provisions include excises on petroleum products, tobacco, alcohol, and selected luxury goods, which can, in most cases, be collected efficiently at the border, as well as special taxes for the banking and telecommunications sectors. By default, other sectors should be taxed under the general tax law. Justified exceptions, if any, would depend on each country’s specific situation and would likely be very limited. The following are examples of justified exceptions:

  • Sectors characterized by economic rent, such as high-end hotels and restaurants in a country where the fragile situation dissuades new openings, can be subject to a sectoral levy, in addition to general taxes on income and consumption or the synthetic taxes replacing them. This may be of particular relevance in a crisis country where peacekeeping forces, NGOs, and technical assistance spur high demand for such services. At the same time, it is in the country’s interest to keep the additional levies low enough to avoid discouraging aid providers.

  • Investment incentives in some sectors can be granted under special conditions—for example, construction of essential infrastructure, such as ports, roads, and pipelines, which would become economic assets for agents other than the investing firm. After a conflict or long period of fragility, lack of infrastructure is often a bottleneck, and economic recovery is delayed when the government cannot finance its construction. In such cases, absent external public funding, the private sector is necessary, but an open-ended uncontested monopoly over essential infrastructure should be avoided. Maintaining this balance may require tax concessions, which entail only limited risks in cases of long-term immovable investments.

A reasonable policymaking approach in these cases might look like this:

  • The proposal of sectoral tax provisions should remain under the responsibility of the Ministry of Finance in consultation with the relevant line ministry.

  • The receipts from sectoral taxes should be collected centrally and accrue to the Treasury.

  • Sectoral tax provision should be codified as soon as possible in the tax law. Fragile states (and all others) should try to avoid enacting tax provisions in nontax laws.

Should Personal Income Be Taxed?

Taxing income at the individual level promotes the perception that the tax system is inclusive and that all citizens contribute to financing the government. In an emergency phase, however, a tax on personal income is costly to administer and may be outside the tax administration’s reach. Nevertheless, even if initially embryonic, some sort of income tax must be established. It can be a simplified dual income tax, taking the following form:

  • A withholding tax at a progressive rate on wages, which can be collected at low administrative cost. The top rate should not exceed 25 percent, and the tax-free personal allowance should be sufficiently large (for example, approximately the GDP per capita). This tax would serve as a proxy for a more comprehensive future personal income tax and; as such, it is a signal of tax inclusiveness. To be administratively feasible as a final withholding tax, it must be limited to civil servants and large enterprises (including state-owned enterprises). The net revenue potential of such a tax remains low, given the narrowness of its base and the fact that civil servants may be able to preserve their net incomes by an offsetting increase in public wages.

  • Refraining from taxation of nonwage income sources under the personal income tax: This may be best during the emergency phase so as not to divert administrative resources from higher-return taxes.

Should Real Property Be Taxed?

Although property taxation offers only limited scope for revenue in a very fragile country, it has other merits that may make it worth considering, if and where a cadastre is available, which is more likely in middle-income fragile states. Among 42 fragile states, only Lebanon collected more than 0.5 percent of GDP through property taxes in 2017, and 34 collected no revenue at all.

Property taxation may be simple to administer and is one of the least distortive taxes. Introducing it at a low initial rate early after a crisis will pave the way for an efficient future tax system. The tax can be focused on larger property owners by providing tax-free surface allowances. This introduces an element of equity in the emergency tax system, which may facilitate public acceptance, especially if a progressive tax on wages in the private sector is not possible.

Property taxes can be earmarked for local governments, providing them with some leeway to implement policies tailored to their jurisdictions. They can thus be used to calm political tension in fragile situations and pave the way, in the longer term, for tax sharing between the central and local governments.

How to Deal with Precrisis Legacies

One question that arises during or immediately after an emergency situation is how to treat the precrisis legacy. It may be extremely costly or even impossible to audit or document taxpayer claims concerning liabilities, allowances, and time-dependent tax provisions predating the crisis, even if the claims are made in good faith. Failing to resolve these issues hinders the return to normality—and possibly the implementation of an effective emergency tax system—since large taxpayers can be tempted to net their alleged positions vis-à-vis the tax administration before making postcrisis tax payments.20

In such circumstances, it is best to address the problem immediately. Depending on the tax administration’s conditions—particularly when determining the feasibility, cost, and duration of an audit—the following options may be viable:

  • Total tax amnesty: A radical solution, which should not be dismissed up-front, is the immediate zeroing out of all tax liabilities and allowances for all taxpayers. It may be the best course of action after a conflict, especially one that lasted over a year.

  • Partial tax amnesty: The cancellation of tax liabilities and allowances may be the preferred option for small taxpayers, individuals or firms, regardless of the source of the emergency situation. This allows for auditing only taxpayers whose liabilities (and economic activities) matter most. The amnesty threshold should reflect the administration’s capacity and priorities, but for consistency’s sake, should be equal to the tax threshold for large enterprises. Similarly, sectors subject to specific taxation (banking or telecommunications) need not benefit from partial amnesty.

  • Depreciation allowances: Amnesty may not benefit taxpayers if they had recently invested in, and were still entitled to, depreciation allowances running for several years. Despite its immediate cost to the budget, it may be in the government’s long-term interest to demonstrate its commitment to investors’ rights by permitting companies to keep deducting allowances related to ongoing investments made before the crisis (provided there is a profit tax).

  • Changes in legislation: Articulating the treatment of the precrisis legacy by installing an emergency tax system may prove especially challenging if the emergency follows a slow descent into institutional fragility (during which time the accumulation of unpaid taxes and unresolved credit claims blur the budget situation and clog the administration). Because the administration is unable to keep pace with the accruals of tax obligations, there is good reason to consider changing the tax legislation. Doing so would extend the use of final charges, thereby releasing taxpayers from other obligations.

How to Reform Tax Policy in the Consolidation Phase

In the second stage, after a country has exited extreme fragility and sufficiently restored its budget and institutional capacity, more weight can go to non-revenue objectives, such as tax efficiency and equity. Although this is a time when there is more room for diversity in national approaches, the persistence of institutional fragility remains a constraint and calls for prudence. This section focuses on the key elements of a good transitional tax system while recognizing that both the shape and speed of the transition are country-specific.

What Policy and Governance Process Should be Used?

Strengthening fiscal institutions is a priority for carrying out tax system reform during the consolidation phase. The process is facilitated by the following two steps:21

  • Establishing a sound legal framework for fiscal management: During a crisis, a fragile state may resort to emergency fiscal measures. For example, spending ministries or local entities may determine or directly collect taxes, or a multiplicity of agencies may grant tax advantages. This results in a fragmented legal framework across several laws, sectoral codes (for example, mining, forestry, petroleum, environment, health, tourism, commerce), and investment codes. Diluted, overlapping responsibilities for operating the tax system hamper changes and foster political dissension. Consolidating all tax receipts into the Treasury’s single account (even if some remain earmarked), and entrusting their management to the Ministry of Finance, are prerequisites for postemergency reform.

  • Establishing a central authority responsible for tax reform: Even within the Ministry of Finance, which is typically responsible for tax policy as well as tax and customs administration, certain sources of inefficiency may complicate the reform process. Establishing a central authority privy to all relevant information and with the final say in reform proposals would help overcome this hurdle. This authority could take the form of a high-level committee with a direct reporting line to the highest authority, often Parliament.

Once these two steps have been taken, policymakers will be expected to make a clear outline of their proposed tax system’s characteristics. However, a balance ought to be found between their commitment to a medium-term target tax system and the necessary flexibility in timing. The adoption and publication of a medium-term revenue strategy may be instrumental in this regard. Desirable characteristics of such a strategy include the following:22

  • Clarity of objectives and means: Reducing uncertainty is the main benefit of adopting a medium-term revenue strategy; it stabilizes the expectations of taxpayers, investors, and the tax administration itself. Government should not announce changes to the tax system too vaguely, and should assure the public of the administration’s capacity to manage them adequately.

  • Joint treatment of tax policy and tax administration issues:23 By timing reforms in legislation and administration consistently, the medium-term revenue strategy would signal how each step of the plan depends on the successful completion of previous steps. This would clarify the feasibility of each change and provide flexibility if improvements are faster or slower than expected.

  • Political commitment: A medium-term revenue strategy—adopted with as large a consensus as possible across the political spectrum—reduces the probability of tax objective upheavals after a political change during the transition period. A fair discussion of the fiscal, economic, and distributional implications of the Parliament’s tax reforms would help the public and politicians understand the rationale behind them. Early feedback on reform projects would also reduce the risk of sudden, unexpected consequences.

  • Good data and information: To gather political support, the medium-term revenue strategy should include projections of its effects on the budget and, ideally, on growth and income distribution. This may be unlikely in many fragile states, even with external technical assistance, but data collection and forecasting capacity should be developed as early as possible to improve decision making and build trust in the process.

Adapting Taxation Principles to the Transition

A fragile country’s transition from a state of emergency to normality is often protracted (more than a decade) and bumpy (with temporary exits from, and relapses into, fragility). Tax policy during this period can contribute to the acceleration and stabilization of the transition’s path in several ways. This may be accomplished while accounting more thoroughly for equity and other objectives hampered by priority on continuity or simplicity in revenue collection, and by recourse to single-stage taxes during the emergency.

Addressing Efficiency Concerns

As fragile states recover, most adopt slightly more sophisticated systems, reorganizing them more systematically into the usual categories: international trade, sales and consumption, personal and corporate income, property, natural resources, and so on. This unbundling allows authorities to determine how each revenue source affects agents, behavior, and, ultimately, economic growth (see below).

Many taxpayers’ liabilities, as well as the actual economic incidence of the tax system, would be modified during the unbundling process. Careful planning and gradualism can help prevent political deadlock and unnecessary economic shocks. Experience in fragile states suggests that, although there is no lack of gradualism in tax legislation reform, such reform may still be too ambitious regarding administrative capacity. Although the other objectives can be assigned increasing levels of importance, authorities must not lose sight of parsimony as the reform process progresses. This points to the advantages of simple taxes with moderate rates applied to broad bases with few exemptions.

Addressing Equity Concerns

Enhancing equity through progressive taxation, particularly personal income tax, poses challenges to fragile states, which cannot easily assess individuals’ total income if it is derived from different sources. However, implementing a progressive, comprehensive personal income tax is not the only way to strengthen a tax system’s equity:24

  • An easy-to-tax legitimate tax base is perceived as more equitable than one that is difficult to tax. Tax evasion or even (legal) avoidance is a more blatant source of inequity than the persistence of numerous little-taxed informal activities. Keeping taxes for formal businesses simple reduces tax evasion and, therefore, makes the system more equitable. It is important to pay particular attention to how taxes apply to the self-employed and high-income individuals in order to balance the need for equity with taxation that discourages entrepreneurship.

  • A general consumption tax with a high threshold lowers the tax burden on the poor, who purchase more from smaller retailers. Conversely, exemptions from consumption tax or lower rates for some products benefit the rich, who, in most cases, consume more of those products.

  • Excise taxes on luxury goods are an effective way to compel the rich to contribute more to the budget, as long as these taxes are easy to collect (for example, at the border). • A property tax is progressive (even at a flat rate), especially if the rate applies above an adequate threshold. Moreover, it is easier to manage a multiple-rate property tax than a comprehensive personal income tax, since property taxes are based on predetermined values, unlike comprehensive personal income taxes, whose bases are very difficult to verify.

Taking Country Characteristics into Account

The following are country characteristics that should be considered when defining the target tax system:

  • Development level: In lower-income countries, agriculture usually involves a large portion of the total population, and informal and low-productivity jobs engage significant shares of the urban population. Aiming for broad personal income tax coverage or a value-added tax that encompasses the small retail sector would be premature in such conditions. More generally, a lower-income country may wish to keep tax design and tax collection focused on larger taxpayers for longer, given the high cost of covering a large share of all potential taxpayers.25

  • Administrative capacity: Fragile states are often overconfident regarding tax administration and compliance. This can lead to tax policy mistakes, such as setting thresholds too low too early, defining taxes using too many rates, or letting the tax base shrink through hard-to-monitor exemptions. Policy options that help minimize the cost of compliance and collection should be favored in fragile states.

  • International commitment: A fragile state that is party to a regional agreement (for example, West African Economic and Monetary Union [WAEMU], Central African Economic and Monetary Community [CEMAC]) that has implications for tax policy can take advantage of such an arrangement.26 While honoring these agreements is not always possible, nor advisable, during a crisis, they do provide anchors for future tax policies. They also help with implementation of shared objectives through peer pressure from, and experiences of, neighboring countries.27

  • Natural resources: Natural resources can finance large shares of public spending during the consolidation phase in some countries (for example, Iraq, Libya, Timor-Leste, Yemen), but diversifying budget revenue sources brings stability and security in the longer term. This may be challenging during commodity boom periods when there is little pressure to increase taxes from other sources (for example, Angola). A medium-term revenue strategy geared toward increasing taxes on nonresource bases would strengthen the commitment to medium-term diversification, regardless of commodity price fluctuations.

  • Peacekeeping forces and international aid: The presence of aid or NGO workers in fragile states may affect the level and composition of tax revenue positively through the consumption of domestic services and goods, and sometimes negatively through imports of untaxed goods that leak into parallel markets. Tax policy reforms can limit such leakage in the short term (by persuading donors to pay taxes) and prepare for withdrawal in the medium term (by anticipating decreased consumption of excised items, such as alcoholic beverages, cigarettes, and hotel stays).28

How to Reform Taxes

The variety of fragile states means diverse approaches to tax policy, but the paths to their objectives are often similar. They also frequently learn the same lessons during their different experiences.

Unbundling the Tax System

Fragile states often have bundled tax systems in which taxes pursue several objectives at the same time, and the same objective is pursued by several taxes. Common examples abound: consumption taxes (instead of tariffs) are used to protect domestic production, corporate income taxes are differentiated to capture sectoral rents (instead of sectoral surcharges), presumptive levies are imposed on extractive industries (instead of disaggregated taxes), and several taxes are imposed on the same base. While most presumptive taxes are unwound as soon as the country overcomes the emergency situation, other forms of bundling often survive for extended periods.

Unbundling the tax system does not, by itself, address new efficiency or equity concerns, but it lays the groundwork for doing so. Once properly unbundled, the tax system can be gradually improved, tax by tax, with clear objectives and fewer unintended consequences. This reform process is efficient if the initial unbundling incorporates a few simple principles:

  • Taxable enterprises should be subjected to the same income tax, regardless of sector. If sectoral differentiation is made, it should be done through specific sectoral taxes that supplement the regular income tax. This principle is also recommended for extractive industries, even though the bilateral nature of many rent-sharing agreements may make it difficult to apply immediately.

  • If there is a VAT or a sales tax, it should be applied at the same rate both for goods that are imported and for those produced domestically. Protecting domestic producers is the main rationale behind customs duties and should be left to the external tariff.

  • Similarly, excises should not depend on whether the goods are imported.

  • Once customs duties, the personal income tax, the corporate income tax, VAT (or the sales tax), and possibly property tax have been clearly identified, the authorities should consider consolidating the remaining taxes (such as excises) if they are levied on identical bases.

General Enterprise Taxation

After initially focusing on larger firms, fragile states should concentrate on including smaller firms in the tax system. Inclusiveness reduces some economic distortions, increases the perceived fairness of the system, and improves revenue. This can be accomplished through simple levies on the turnover of small and medium enterprises and possibly through lump-sum taxation of microenterprises. Simultaneously, an increasing number of larger enterprises can be taxed on their actual incomes by lowering the thresholds for presumptive taxes.

Simultaneous, somewhat contradictory efforts often attempt to attract investment through special tax regimes for particular firm categories (for example, exporting firms, firms in lagging areas, and firms in sectors seen as key to economic development). In fragile states, such special provisions typically exceed the tax administration’s control capacity, which allows for tax avoidance and abuse. While not always unappealing to investors, tax incentives oppose the corporate income tax’s inclusiveness objective and distract the administration from collection of actual taxes. Furthermore, in the long term, countries known to be more neutral in their taxes (for example, Rwanda) receive more revenue and are more attractive than countries known for particular tax advantages.

Sectoral Taxes

Taxes specific to firms in sectors that benefit from rents (for example, telecommunications and natural resources) should have higher revenue, efficiency, and distributional objectives. Ideally, only rent above the normal remuneration of invested capital should be taxed. In practice, defining the base is tricky, and simpler, more manageable sectoral taxes are used, such as surcharges on the general corporate income tax, ad valorem excises (taxing the value of sales), specific excises (based on quantities sold), and tariff surcharges (taxing imported inputs).

An ad valorem excise is probably the most cost-effective way to capture sectoral rents in fragile states, but it generates distortions. A corporate income tax surcharge is preferable, because its base is closer to the rent’s value. However, a corporate income tax surcharge makes manipulating the taxable profit even more lucrative than under the general corporate income tax, and such manipulation may be difficult to prevent. Avoiding specific excises would be advisable, since they do react to the rent per unit sold. Tariffs and other input surcharges are also unnecessarily distortive because they tax investment.29

For high-rent sectors, an overall legal framework is advisable, not only for taxation but also for regulation (for example, new mining, forestry, telecommunications, and hydrocarbon codes). From an investor’s perspective, the legal framework’s stability is as desirable as that of the tax system. To capture revenue in the early stages of extraction, the resource sector should be subject to a general tax on profits (or production-sharing agreements in oil and gas), a royalty, and a rent tax. In many cases, this regime may not differ greatly from previous ones (in terms of the policy tools used to extract revenue for the government). However, it could be significantly improved in terms of tax parameters (for example, bonuses or the balance between profit-sensitive and profit-insensitive taxes), including the parameters’ broader relationship with the sector’s legal framework.

As governments undergo these very complex and difficult reform exercises, the implications for revenue and the economy should be carefully assessed. Rushing to reform simply to encourage investors is not always desirable. Fragile states, even in the consolidation phase, are frequently at a disadvantage when negotiating long-term investment projects with large multinationals.

Personal Income Tax

Governments willing to make their tax systems more equitable and inclusive are keen on enhancing personal income tax. Although the distributional effects of a comprehensive personal income tax may be somewhat overstated compared with those of well-targeted transfers or robust property taxes,30 bringing more citizens into the tax net does help reconstruct the state through a shared sense of belonging. But a comprehensive, progressive personal income tax is one of the costliest to administer, and fragile states risk devoting too much administrative capacity to it.

This constraint can be eased through a dual income tax system, which is used in many developing and developed economies, and which be considered in fragile states. Dual income tax can be defined in a variety of ways. It is a tax withheld from wages at a progressive rate and applied to other sources of income at a flat rate, often also through withholding. Therefore, it does not require the actual monitoring of each individual’s comprehensive income. As such, it can build directly on a withholding tax on wages set up during an emergency by supplementing it with other levies during the consolidation phase.31

The push to relax bank secrecy and exchange data across tax jurisdictions gives fragile states opportunities to use dual income tax more effectively. This is because high incomes, mostly from capital sources, are often reinvested abroad to avoid the hazards of fragility and can therefore be identified through international information sharing.

General Consumption Tax

Domestic taxes on consumption (that is, after proper unbundling of the general consumption tax and excises) are the main sources of tax revenue in nonresource fragile economies, but they are still untapped, especially VAT. In the past few decades, VAT has been adopted in most countries, including those that are fragile, sometimes significantly boosting tax revenue (for example, Bosnia and Herzegovina, Rwanda). More generally, the revenue effects of adopting VAT have been positive in low- and middle-income countries and are associated not only with increased taxes from consumption but also with more efficient administration.

However, VAT receipts fall when a country is suddenly hit with fragility (from 4.8 percent to 4.3 percent of GDP in the median newly fragile state) and remain low in permanently fragile states. This raises the question of whether VAT may be too difficult to administer in a fragile state. Experiences in recently fragile states, and the accumulation of nonrefunded credits, which accompanies the descent into fragility, suggest that countries threatened by or suffering from fragility should design VAT to be more robust to institutional shocks. The following guidance can be applied: (1) fragile states that already have VAT may wish to amend it in order to increase its robustness to shocks during the consolidation phase (see below); (2) fragile states without VAT may consider adopting it, but not before their administrative capacity has recovered sufficiently.

VAT design that limits the risk of faltering in the face of shocks includes the following:

  • Limiting the number of taxpayers liable for the full VAT regime without changing the rest of the VAT legislation: The turnover threshold, below which a presumptive regime (or mere exemption) applies, can be adapted to the country’s weaker capacity, then lowered again as it recovers. This reduces the number of taxpayers whose VAT credit claims must be monitored.32

  • Simplifying VAT legislation: The VAT design may have to deviate from what is recommended for a higher-income country. The general recommendation is to adopt a single VAT rate on a large base with few, if any, exemptions (beyond the one based on size). This reduces compliance and administrative costs, limits the scope for refunds, and preserves or improves VAT revenue. In addition, reducing the need for coordination between customs and the inland tax administration can alleviate compliance and administrative burdens. This can be accomplished by exempting imported capital goods from VAT at the border, which would significantly limit the scope for VAT refunds other than those related to exports.

Excise Taxes

It is surprising how infrequently excises are used in fragile states, given the simplicity of their administration and their revenue potential. Their underutilization may reflect the perceived distributional effects and potential political difficulties of introducing or increasing excises on products consumed directly or indirectly by the poor (for example, petroleum products) in countries where direct targeted transfers may not be feasible.

Still, excises are low-hanging fruits among revenue-enhancing tax reforms in fragile states. While the list of excisable products does not change during an emergency period, petroleum products, cigarettes, alcohol, and cars are likely to have higher revenue potential. Therefore, policymakers could consider higher rates during the consolidation phase as the capacity to battle the development of parallel markets increases.

Property Taxes

Establishing a cadastre is an important step in assessing property rights, securing them for real estate transactions, and facilitating their use as collateral for credit and investment. A cadastre is therefore an objective to be pursued for its own sake in a fragile state, where these rights may be contested more than they would be elsewhere.

A property tax can be considered wherever, and as soon as, a cadastre is available, since it outperforms most other taxes in terms of efficiency and equity. Its desirable characteristics need not differ from those recommended for emergency situations.

Key Lessons and Guiding Principles for Tax Reform

The experiences of fragile states reveal a few key lessons for tax policy, one of many components that contribute to overcoming fragility. Some of these lessons may appear trivial, yet they have proved elusive and challenging in many fragile states.

First, stability-oriented policies have played key roles in luring private investors to fragile states. During the past two decades, political and tax policy stability have endowed some fragile states with comparative advantages in the competition for the relocation of global value chains. For example, Cambodia, Lao P.D.R., Rwanda, and, more recently, Niger have benefited from relatively stable tax systems established after the onset of fragility. This enabled tax administration reforms and revenue mobilization to progress without further legislative changes.

Second, drafting tax policy from scratch has not been a disadvantage for countries hit by particularly deep crises. In Cambodia and Rwanda, promptly adopting tax systems that were viewed as manageable, investment-friendly, and reasonably inclusive has provided protracted periods of tax stability, which has been favorable to economic development.

Third, natural resource revenue can provide the time and means needed for institutional reconstruction but cannot delay tax reform indefinitely. Having taken advantage of their natural resource revenue for a long time, Angola and Lao P.D.R. are now engaging in tax system reforms to respond to the pressures of lower commodity prices and competition for foreign investment from their neighbors.

Fourth, political stability, which is often needed for improving tax policy, may require compromises with tax purism for some time. For example, for many years, Bosnia and Herzegovina had maintained an agreement for sharing revenue between their separate entities using a far-from-perfect consumption tax system. However, once the political ground matured, the country was able to introduce a modern and highly efficient unified VAT.

Fifth, externally anchoring tax policy has helped in several cases. The WAEMU constraints on customs and domestic tax legislation, though not always binding in emergency situations, have served as a rebalancing and stabilizing force in Niger. Similarly, the prospect of accession to the European Union has been a useful anchor for tax reform in Albania, Bosnia and Herzegovina, and Serbia.

Sixth, several countries that exited fragility have benefited from intensive technical support in tax policy and administration from various providers. Aside from the level of technical support, the coordination among providers has also played an important role in successful tax reforms, such as in the design of Bosnia and Herzegovina’s VAT.

From these lessons, and the analysis provided in this how-to note, it is possible to derive a set of broad guidelines for tax policymaking in fragile states—recognizing that success in overcoming fragility is to a large extent determined by how countries adapt guidelines and experiences to their own situations, and how the links between taxation and other factors affecting fragility are integrated in policymaking.

The following guidelines are appropriate during the emergency phase:

  • Identify and define taxes to raise revenue according to collection points, favoring points with a high concentration of collection—for example, imports and very large taxpayers.

  • Use simple tax bases, such as the gross value of imports or turnover.

  • Exempt enterprises from turnover taxation if their turnover is below a high threshold, but do not provide other forms of exemptions.

  • Apply additional levies on natural resources, banks, and telecommunications enterprises.

  • Keep the level and number of tax rates low—say, below 20 percent for taxes on imports and well below 10 percent for taxes on turnover.

  • Tax formal wages (for example, public sector wages and wages paid by enterprises subject to turnover taxation) but not other sources of personal income.

During the consolidation stage, adapt the guidelines of the emergency stage to raise more revenue (if needed) and to account for nonrevenue objectives, such as improving efficiency and equity in taxation:

  • Consolidate all taxation responsibilities (both policy analysis and collection of revenue) under the Ministry of Finance. (This is in addition to the central authority overseeing tax reform proposed earlier, which plays a distinct but complementary role.)

  • Design a medium-term reform strategy with implementation conditional on progress in revenue administration.

  • Unbundle the tax system to differentiate (1) taxes on imports into tariffs, general taxes on goods and services, and excise taxes; and (2) taxes on turnover into general taxes on goods and services and corporate and personal income taxes.

  • Gradually move the income tax system toward a form of dual income taxation, with personal income taxes largely based on final withholding.

  • Gradually lower the exemption thresholds (for general taxes on goods and services and income taxes) to bring more taxpayers into the tax net.

Annex 1. Definition and Characteristics of Fragile States

A fragile state is a state whose government is unable to provide its population with basic services and security. Whatever the initial cause(s), institutional failures tend to snowball and impact different areas of activity, undermining the state’s credibility, legitimacy, and stability. This, in turn, affects economic activity. Fragility in one country can also spread to other countries, through emigration, civil unrest, or war.

Despite the variety of its manifestations, a clear-cut definition of fragility is helpful in studying these countries. Such criteria cannot be grounded in hard science. In particular, there are different ways of assessing security issues, and the choice of criteria and of a threshold for assessing institutional capacity is somewhat arbitrary. The World Bank’s Country Policy and Institutional Assessment (CPIA) is used, as it measures the extent to which a country’s policy and institutional framework supports sustainable growth and poverty reduction, and consequently the effective use of development assistance. The CPIA was developed and first employed in the mid-1970s. Over the years, the World Bank has periodically updated it to reflect the lessons of experience and the evolution of thinking about development. Whatever the definition of the fragility boundary, some countries cross back and forth before reaching safer territory.

This problem can be attenuated, but not fully eliminated, by adopting the IMF approach,1 according to which a country is fragile in a given year if:

  • it is a low-income country with little adminis-trative capacity, in the sense that the three-year average CPIA (for the most recent years available) is below 3.2, or

  • international peacekeeping or peacebuilding forces have been operating in the country, whatever its income level, during the previous three years.

In 2017, there were 42 fragile states, among which sub-Saharan Africa and the Pacific islands were overrepresented (respectively 24 and 8 countries). Among low- and middle-income countries, one in three was fragile. No high-income country was fragile. More than 530 million people (7.3 percent of the world population) lived in fragile states in 2017, but they accounted for only 1.7 percent of global GDP, although a significant proportion of them (18 out of 42) were resource-rich countries (Annex Table 1.1). The number of fragile states declined by only one between 2008 and 2017, but fragility decreased in Europe and Asia and increased in the Middle East, while Latin America remained untouched by fragility. More than 40 percent of fragile states had security problems, and new conflicts were the immediate cause of one-third of fragility cases.

Annex Table 1.1.

Fragile States, 2008 and 2017

article image
Source: Authors, based on World Bank’s Country Policy and Institutional Assessment index.Notes: *denotes countries where peacekeeping or peacebuilding forces were present in 2013, 2014, or 2015 or countries where a conflict prevented computation of the World Bank Country Policy and Institutional Assessment (CPIA) index.°denotes countries where peacekeeping or peacebuilding forces were present in 2006, 2007, or 2008 or countries where a conflict prevented computation of the CPIA index. Countries in italics derive 20 percent or more of their net exports from natural resources.

Statistical data on Somalia are not available.

South Sudan became independent from Sudan in 2011.

Kosovo became independent in 2008.

These are countries that were fragile in 2008 but were no longer fragile in 2017.

What Characterizes Fragile States?

As noted above, a state may be fragile for different reasons. Some young countries are still building their institutions (for example, Eritrea, Kosovo, South Sudan, Timor-Leste). Others have been affected by war (for example, Afghanistan, Iraq, Mali, Syria), natural disasters (for example, Haiti), or epidemics (for example, Liberia, Sierra Leone). Others have slipped into fragility as a result of economic shocks, such as declines in natural resource prices (for example, Cameroon, Democratic Republic of the Congo, Papua New Guinea), or because of protracted weak policies (for example, Madagascar, Maldives, Myanmar).

How Different Are Fragile States?

Fragility is associated with weak institutions, unstable political regimes, low state legitimacy, and war, any of which may be amplified by natural disasters or epidemics. These features often constrain the shaping of tax policy more than economic factors do.2

Fragile states share some characteristics with low-income countries, but they differ in certain key factors. Unlike the 82 nonfragile lower- or middle-income states in 2017, fragile states exhibited several characteristics (Annex Table 1.2):

  • GDP per capita is about one-third lower than that of nonfragile states, with a high prevalence of informality.3

  • The growth of GDP per capita is also lower, on average, and even negative in some countries.

  • The current account balance is often more negative, but not always.

  • Fragile states depend much more on official development aid.

Annex Table 1.2.

Fragile and Nonfragile States: Selected Indicators

article image
Source: IMF.Note: The list of FSs and NFSs is for 2017. Figures represent simple 2013–15 averages in each category, with median in parentheses. FS 5 fragile state; NFS 5 nonfragile state; PPP 5 purchasing power parity.

Fragile states do not look much different from non-fragile states in terms of inflation or budget balance.4 However, some economic weaknesses may not be apparent from simple indicators. For example, GDP is boosted where peacekeeping forces spark economic activity, but fades out once security is restored. The current account deficit is often more than covered by official development assistance, while private capital flows out of the country. Net private investment is generally low or negative during fragility periods, and drawing foreign direct investment back is a challenge. All these imbalances must be addressed to overcome fragility and restore economic growth and independence.5

How Long Do Countries Remain Fragile?

It takes a long time to overcome fragility. According to several studies, 15 to 30 years may be needed to rebuild the institutional capacity of a country after a sharp deterioration, and not all fragile states succeed within that time frame.6 Once it reaches the lowest level of fragility (hereafter the “crisis”), a fragile state typically makes rapid progress for about five years, during which time the World Bank’s CPIA indicators, tax revenue, and the security situation improve steadily (a period called hereafter the “emergency phase”). Progress then slows, and it takes another 10 years or more to graduate from fragility status (a period hereafter called the “consolidation phase”).

About 30 percent of the states that were fragile in 2008 have succeeded in their efforts to exit fragility, in the sense that they were no longer (technically) so by 2017. None of these countries had security problems immediately before 2008, but half had some in the previous decade. This indicates that security is a prerequisite for consolidation, and that once security is reestablished (as it was in 2017 in Chad, Eritrea, and Timor-Leste) several additional consolidation years are needed to conquer fragility.

Before 2008, besides security, formerly fragile states exhibited higher GDP per capita, more GDP per capita growth, less dependence on official development assistance, and lower government debt than states that remained fragile (Annex Table 1.3). Formerly fragile states’ governments were also able to increase their gross debt during the decade, which may indicate restoration of investor confidence.

Annex Table 1.3.

Economic Trends in Formerly and Permanently Fragile States

article image
Source: IMF staff calculations.Note: Figures are simple averages for each category, with medians in parentheses. FFS 5 formerly fragile state; PFS 5 permanently fragile state; PPP 5 purchasing power parity.

Annex 2. Comparative Statistics on Fragile and Nonfragile States

Annex Table 2.1.

Revenue Structure

(Simple averages for 2013–15 based on 2017 list of fragile states1; median in parenthesis)

article image
Source: IMF, World Revenue Longitudinal Database.Note: FS 5 fragile state; NFS 5 nonfragile state.

The figures for the various taxes may not add up because not all series are available every year for all countries.

The average in percent of tax revenue is for countries that derive 20 percent or more of their net exports from nonrenewable natural resources.

Annex Table 2.2.

Revenue Trends in Formerly and Permanently Fragile States

(Simple average of each category of country,1 median in parentheses)

article image
Source: IMF, World Revenue Longitudinal Database.Note: FFS 5 formerly fragile state; PFS 5 permanently fragile state.

The figures for the various taxes may not add up to the total tax revenue because not all series are available every year for all countries.

The average in percent of total tax revenue is for countries that derive 20 percent or more of their net exports from nonrenewable natural resources.


  • Acemoglu, D., and J. Robinson. 2012. Why Nations Fail: The Origins of Power, Prosperity, and Poverty. New York: Crown Publishers.

  • Acosta-Ormaechea, S., and J. Yoo. 2012. “Tax Composition and Growth: A Broad Cross-Country Perspective.” IMF Working Paper 257, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation
  • Baer, K., O. Benon, and J. Toro. 2002. Improving Large Taxpayers’ Compliance: A Review of Country Experience. IMF Occasional Paper 215, Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation
  • Besley, T., and T. Persson. 2014. “Why Do Developing Countries Tax So Little?Journal of Economic Perspectives, 28 (4).

  • Bird, R., and E. Zolt. 2005. “Redistribution Via Taxation: The Limited role of the Personal Income Tax in Developing Countries.” UCLA Law Review Vol. 52.

    • Search Google Scholar
    • Export Citation
  • Bird, R., and E. Zolt. 2010. “Dual Income Taxation and Developing Countries.” Columbia Journal of Tax Law 1.

  • Boadway, R., and F. Flatters. 1993. “The Taxation of Natural Resources: Principles and Policy Issues.” World Bank Working Paper 1210, Washington, DC.

    • Search Google Scholar
    • Export Citation
  • Boadway, R., and M. Keen. 2010. “Theoretical Perspectives on Resource Tax Design.” In The Taxation of Petroleum and Minerals: Principles, Problems and Practice, edited by P. Daniel, M. Keen, and C. McPherson. New York: Routledge.

    • Search Google Scholar
    • Export Citation
  • Brautigam, D., F. Odd-Helge, and M. Moore, eds. 2008. Taxation and State Building in Developing Countries. Cambridge, UK: Cambridge University Press.

    • Search Google Scholar
    • Export Citation
  • Brun, J.-F., G. Chambas, and M. Mansour. 2015. “Tax Efforts of Developing Countries: An Alternative Measure.” In Financing Sustainable Development—Addressing Vulnerabilities, edited by M. Boussichas and P. Guillaumont. Paris: Economica.

    • Search Google Scholar
    • Export Citation
  • Cilliers, J., and T. Sisk. 2013., “Assessing Long-Term State Fragility in Africa: Prospects for 26 ‘More Fragile’ Countries.” Institute for Security Studies Monograph 188, Pretoria.

    • Search Google Scholar
    • Export Citation
  • Crivelli, E., and S. Gupta. 2014. “Resource Blessing, Revenue Curse? Domestic Revenue Effort in Resource-Rich Countries.” IMF Working Paper 14/5, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation
  • Everest-Phillips, M. 2010. “State-Building Taxation for Developing Countries: Principles for Reform.” Development Policy Review, 28 (1).

    • Search Google Scholar
    • Export Citation
  • Gaspar, V., L. Jaramillo, and P. Wingender. 2016. “Political Institutions, State Building and Tax Capacity: Crossing the Tipping Point.” IMF Working Paper 16/233, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation
  • Gelbard, E., C. Deléchat, E. Fuli, M. Hussain, U. Jacoby, D. Glaser, M. Pani, G. Ramirez, and R. Xu. 2015. Building Resilience in Sub-Saharan Africa’s Fragile States. Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation
  • Gupta S., S. Tareq, B. Clements, A. Segura-Ubiergo, R. Bhjattarcharya, and T. Mattina. 2005. Rebuilding Fiscal Institutions in Post-Conflict Countries. IMF Occasional Paper 247. Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund (IMF). 2010. “Financial Sector Taxation: The IMF’s Report to G-20 and Background Material.” Washington, DC.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund (IMF). 2011. “Revenue Mobilization in Developing Countries.” IMF Policy Paper, Washington, DC.

  • International Monetary Fund (IMF). 2017a. “Building Fiscal Capacity in Fragile States.” IMF Executive Board Paper, Washington, DC.

  • International Monetary Fund (IMF). 2017b. “Multi-Country Report: Building Fiscal Capacity in Fragile States—Case Studies.” IMF Executive Board Paper, Washington, DC.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund (IMF). 2018. “Sub-Saharan Africa: Domestic Revenue Mobilization and Private Investment.” Regional Economic Outlook, Washington, DC, April.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund (IMF), Organisation for Economic Co-operation and Development (OECD), United Nations (UN), and World Bank (WB). 2016. “Enhancing the Effectiveness of External Support in Building Tax Capacity in Developing Countries.” Paper prepared for submission to G20 finance ministers, Washington, DC.

    • Search Google Scholar
    • Export Citation
  • International Tax Dialogue. 2006. Tax Treatment of Donor-Financed Projects.

  • Keen, M., and J. Slemrod. 2017. “Optimal Tax Administration.” IMF Working Paper 8, International Monetary Fund, Washington, DC.

  • Mansour, M., and G. Rota-Graziosi. 2013. “Tax Coordination, Tax Competition, and Revenue Mobilization in the West African Economic and Monetary Union.” IMF Working Paper 163, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation
  • Matheson, T., and P. Petit. 2017. “Taxing Telecommunications in Developing Countries.” IMF Working Paper 247, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation
  • Organisation for Economic Co-operation and Development (OECD). 2014. Fragile States 2014: Revenue Mobilisation in Fragile States. Paris.

    • Search Google Scholar
    • Export Citation
  • Porter-Peschka, M. 2011. “The Role of the Private Sector in Fragile and Conflict-Affected States.” World Development Report 2011. Washington, DC: World Bank.

    • Search Google Scholar
    • Export Citation
  • Schneider, F., and D. Enste. 2013. The Shadow Economy. Cambridge, UK: Cambridge University Press.

  • World Bank. 2011. World Development Report—Conflict, Security, and Development, Washington, DC.

  • Zee, H. 2004. Taxing the Financial Sector: Concepts, Issues, and Practice. International Monetary Fund: Washington, DC.


  • Besley, T., and T. Persson. 2013. “Taxation and Development.” Handbook of Public Economics Vol. 5, edited by A. Auer-bach, R. Chetty, M. Fieldstein, and E. Saez. Amsterdam: North Holland.

    • Search Google Scholar
    • Export Citation
  • Carnahan, M. 2007. “Options for Revenue-Generation in Post-Conflict Environments.” University of Massachusetts Political Economy Research Institute Working Paper 137, Amherst.

    • Search Google Scholar
    • Export Citation
  • Collier, P. 2007. The Bottom Billion: Why the Poorest Countries Are Failing and What Can Be Done About It. Oxford: Oxford University Press.

    • Search Google Scholar
    • Export Citation
  • Deléchat, C., W. Clark, P. Gupta, M. Kabedi-Mbuyi, M. Koulet-Vickot, C. Macario, T. Orav, M. Rosales, R. Tapsoba, D. Zhdankin, and S. Yang. 2015. “Harnessing Resource Wealth for Inclusive Growth in Fragile States.” IMF Working Paper 15/25, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation
  • Deléchat, C., E. Fuli, D. Mulaj, G. Ramirez, and R. Xu. 2015. “Exiting from Fragility in Sub-Saharan Africa: The Role of Fiscal Policies and Fiscal Institutions.” IMF Working Paper 15/268, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation
  • Di John, J. 2010. “Taxation, Resource Mobilisation, and State Performance.” London School of Economics Crisis States Research Centre Working Paper 84, London.

    • Search Google Scholar
    • Export Citation
  • Drummond, P., W. Daal, N. Srivastava, and L. Oliveira. 2012. “Mobilizing Revenue in Sub-Saharan Africa: Empirical Norms and Key Determinants.” IMF Working Paper 108, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation
  • Ebrill, L., M. Keen, J. P. Bodin, and V. Perry. 2001. The Modern VAT. Washington, DC: International Monetary Fund.

  • Fjeldstad, O.-H., and M. Moore. 2007. “Taxation and State-Building: Poor Countries in a Globalised World.” Christian Michelsen Institute Working Paper 11, Bergen, Norway.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund. 2015. “The Commodity Roller-Coaster: A Fiscal Framework for Uncertain Times.” Fiscal Monitor, Washington, DC, September.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund. 2016. “How to Adjust to a Large Fall in Commodity Prices.” IMF How-To Note 1, Washington, DC.

  • Keen, M. 2012. “Taxation and Development—Again.” IMF Working Paper 220, International Monetary Fund, Washington, DC.

  • Keen, M., and J. Mintz. 2004. “The Optimal Threshold for a Value-Added Tax.” Journal of Public Economics, 87 (4).

  • Maier, R. 2010. Growth and Equity in Fragile States. The Hague: Conflict Research Unit, Netherlands Institute for International Relations.

    • Search Google Scholar
    • Export Citation
  • Skinner, J. 1987. “Taxation and Output Growth: Evidence from African Countries.” NBER Working Paper 2335, National Bureau of Economic Research, Cambridge, MA.

    • Search Google Scholar
    • Export Citation
  • Tanzi, V., and H. Zee. 2000. “Tax Policy for Emerging Markets: Developing Countries.” IMF Working Paper 35, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation

Mario Mansour is the Coordinator of the IMF Middle East Regional Technical Assistance Center. Jean-Luc Schneider is with France’s Ministry of Economy and Finance, Directorate General of the Treasury.


Annex 1 provides a brief analysis of the economic differences between fragile and non-fragile states, which complements the taxation discussion in this note.


See Crivelli and Gupta (2014); and Brun, Chambas, and Man-sour (2015). Governments may collect nontax revenue from natural resources through sharing in the equity of the extractive firms or receiving licensing fees.


States that exited fragility during the study period are referred to as formerly fragile states, and those that remained fragile are referred to as permanently fragile states. It is important to note that this characterization says nothing about the probability of a formerly fragile state sliding back into fragility or a permanently fragile state exiting fragility in the years following the study period.


In Haiti, the IMF recommended setting the VAT liability threshold at 450 times the GDP per capita.


In developing economies, taxes generally applicable to the resource sector are often found in mining sector licenses or production-sharing agreements in oil and gas, with long tax stability clauses. Changing these agreements involves a very complex process of renegotiation, with highly uncertain outcomes.


In 2015, the volume of calls in Guinea fell by 15 percent following the introduction of tax representing about 15 percent of call prices.


Experience suggests that this may take a long time, as vested interests tend to be well organized to lobby for keeping such provisions.


Countries that have slipped into fragility without a crisis may experience low-intensity versions of this problem if VAT credits accumulate while the tax administration’s capacity to audit them deteriorates. This creates a liability for the government, often left of budget, and a loophole in the tax system.


For a more detailed discussion of core medium-term revenue strategy elements, see IMF and others (2016).


More broadly, governments are advised to link tax policy and administration to the budget’s expenditure side through clear statements. These statements should be backed by spending allocation and execution, illustrating the general benefits taxpayers may enjoy.


Regional agreements usually cover the taxation of trade between participating countries, thereby providing guidelines for tariff policy more generally. In addition, the WAEMU has directives constraining the rates, exemptions, and excisable products for consumption taxes (VAT and excise), as well as for corporate income tax. See Mansour and Rota-Graziosi (2013).


Niger, for example, has benefited from its WAEMU membership, which accelerated the adoption of a VAT in the 1990s. Focusing on administrative improvements, without having to change the law, allowed the country to more than double VAT revenue in the past decade.


For further discussion, refer to International Tax Dialogue (2006).


As the threshold is lowered, some of the distributional benefits mentioned earlier are likely to become negligible or disappear, and may need to be replaced by other mechanisms to cushion the impact on the poor.


See IMF (2017a, b).


Government gross debt has become lower in fragile than in nonfragile states recently, in contrast to the situation 10 years ago, mainly as a consequence of the Heavily Indebted Poor Countries debt relief initiative.

How to Design Tax Policy in Fragile States
Author: Mario Mansour and Mr. Jean-Luc Schneider
  • View in gallery

    Revenue Structure of Fragile and Nonfragile States

    (Simple average for 2013–2016)

  • View in gallery

    Revenue Structure of Permanently versus Formerly Fragile States

    (Simple average for 2013–2016)