Mali: Technical Assistance Report - Local Taxation and Decentralization

This Technical Assistance Report surveys the resources of the subnational jurisdictions and proposes a tax reform strategy for Mali. In Mali, the system of local taxes generates insufficient revenue and relies on obsolete taxes, which are particularly difficult to collect. It proposes that the rental value be replaced by a value per hectare indexed on various factors such as access to property or public services; geographic location; and average and georeferenced consumption of electricity, water, cell phone minutes. A real property tax system will also require the elimination of key exemptions under the current real property tax.


This Technical Assistance Report surveys the resources of the subnational jurisdictions and proposes a tax reform strategy for Mali. In Mali, the system of local taxes generates insufficient revenue and relies on obsolete taxes, which are particularly difficult to collect. It proposes that the rental value be replaced by a value per hectare indexed on various factors such as access to property or public services; geographic location; and average and georeferenced consumption of electricity, water, cell phone minutes. A real property tax system will also require the elimination of key exemptions under the current real property tax.

Summary of Recommendations1

This report surveys the resources of the subnational jurisdictions (collectivités territoriales*) and proposes a tax reform strategy. Like most previous studies addressing local taxation in Mali, the mission found that the system of local taxes generates insufficient revenue and relies on obsolete taxes, which are particularly difficult to collect.

The proposed reform takes account of the risks of a poorly managed decentralization process for national public finances. It proposes four complementary pillars designed to increase domestic tax collection (central or local), provide financing for subnational jurisdictions, and improve accountability on the part of local elected officials. The four pillars are:

  • The institution of a property tax based on property value (without awaiting the potential preparation of a modern cadaster) by improving or modifying the current real property tax. This will require the introduction of an alternative to rental value for use in defining the real property tax base. The rental value is not suitable in the Malian context and should be replaced by a value per hectare indexed on various factors such as access to property or public services, geographic location, average and georeferenced consumption of electricity, water, cell phone minutes, etc. A real property tax system will also require the elimination of key exemptions under the current real property tax, such as the tax on the property occupied by the owner or immediate family member, unimproved land held for less than three years, or agricultural land, notably on the perimeter of urban areas. The new real property tax proposed in this report would incorporate particular local taxes such as the business license tax (BLTJ,** and the Regional and Local Development Tax (TDRL), which would then be eliminated.2 Based on international experience, the anticipated revenue from a real property taxes would be on the order of at least 0.5 percent of GDP. The proposed tax would be fairer than the current TDRL and more economically efficient than the BLT. It would also discourage property speculation and optimize the allocation of land.

  • Sharing of VAT revenue between subnational jurisdictions and the central government to ensure stable resource flows to the subnational jurisdictions The share of revenue distributed to subnational jurisdictions and the modalities of defining that share should be given careful consideration, and might be addressed in the context of a specific technical assistance mission.

  • Improving the alternative minimum business tax (impôt synthétique) to rationalize the direct local taxation of small businesses. That tax was significantly simplified in 2014. This tax is an alternative tax, as already established by the 2014 budget, which should lead to the elimination of a number of other local taxes (taxes on livestock, mills, wagons and carts, etc.). It should also justify the transfer of a portion of revenue to the subnational jurisdictions insofar as the Directorate General of Taxation (DGI) or the Treasury is involved in collecting the tax.

  • The replacement of specific local taxes by various fees (redevances).3 Some specific taxes (vehicle exit tax (taxe de sortie des véhicules),, boat tax, etc.) should be eliminated and converted to fees. This would shift the contribution to the user and strengthen accountability on the part of local authorities.

This technical assistance mission was financed by the Tax Policy and Administration Topical Trust Fund. It is part of a three-year program to assist the Malian authorities in their tax policy reform efforts. The following table presents the implementation status of recommendations provided during the June 2014 diagnostic mission and discussed in the report by Rota-Graziosi et al. (2014a).

Table A.

Monitoring of June 2014 tax policy recommendations

article image

CFE: Contribution forfaitaire des employeurs; TFP: Taxe de formation professionnelle; TEJ: Taxet imploi-jeune; TL: 1’axe de logement

I. Introduction

1. The Republic of Mali embarked on a decentralization process in the early 1990s. The process was reinstated with the Algiers Agreement for Peace and Reconciliation in Mali. Article 14 of the February 25, 2015 draft commits the Malian government to transfer 30 percent of government tax revenue to the subnational jurisdictions under an equalization system by 2018.

2. This aide-mémoire surveys the resources of Mali’s subnational jurisdictions (collectivités territoriales), reviews the system of local taxation and intergovernmental transfers, and proposes a tax reform strategy. Numerous studies consulted by the mission, which are identified in the bibliography, attest to the importance of decentralization for Mali. Most, including this report, share the conclusion that the local tax system generates insufficient revenue and is substantially outdated. The proposed reform explicitly incorporates the risks (cf. Chapter II) of decentralization for domestic revenue mobilization. Accordingly, the proposed reform aims to improve the accountability of subnational jurisdictions while increasing domestic revenue collection (central or local).4 The distinction is also mitigated, since the mission proposes a system of VAT revenue sharing as to which the details, proportion of VAT, and modalities of intergovernmental transfers are left to the discretion of the Malian authorities and could be addressed in the context of a specific technical assistance mission. The proposed tax reform also establishes a property tax based on property value without awaiting the eventual modernization of the cadaster. Improvement of the alternative minimum business tax, which was significantly simplified in 2014, should also accompany the effort to modernize local taxation. Finally, a number of local taxes should be eliminated in order to rationalize the Malian (local and central) tax system, either by incorporating them into the AMBT or replacing them by fees administered by local authorities. This report supplements Taiclet et al. (2015), which analyzed the macro-fiscal framework of decentralization in Mali.

II. The Malian Subnational Jurisdictions’ Autonomy and the Risks Decentralization Entails

3. Decentralization is the transfer of revenue and expenditure authorities from the central government to local administrations. The aim of decentralization is to improve the efficiency of public spending, in particular by bringing the communities closer to public decision makers (proximity principle) and by introducing some competition between local elected officials (competition principle).5 The autonomy of local officials is a complex concept involving far more than local tax resources; it includes, for example, the degree of discretion available to local authorities in allocating the transfers they receive. This chapter presents a taxonomy of local authorities’ resources and offers several international comparisons. It then examines the risk of excessive decentralization of powers to define and collect taxes for the Malian economy in general and for the efficiency of the decentralization process in particular.

A. A taxonomy of local resources in Mali and around the world

4. An assessment of the autonomy of Malian subnational jurisdictions must consider all the financial resources available to the authorities and the extent of their discretion in allocating them. In the Malian context, the local administrations’ financial autonomy is not limited to local taxes. Box 1 presents a taxonomy of local government resources and offers several points of international comparison.

5. The overall level of Malian regional and local government’s resources is particularly low, below that of other WAEMU countries (cf. Figure 1). At no time during the period 1995-2008 did total local per capita resources in Mali exceed CFAF 3,100 – compared to CFAF 3,300 in Togo, CFAF 3,900 in Benin, CFAF 11,000 in Senegal, and CFAF 12,400 in Côte d’Ivoire. The ratio of local resources to GDP exceeded the threshold of 1 percent for only two [sic] countries in our study, Côte d’Ivoire and Senegal (1.34 percent and 1.21 percent, respectively). During the period under review, the ratio stood at 1.09 percent in Benin, 1.06 percent in Mali, 0.83 percent in Niger, and 0.52 percent in Togo. Finally, the ratio of local resources to national public resources was 5.88 percent in Benin, 8.91 percent in Côte d’Ivoire, 2.96 percent in Mali, 2.61 percent in Togo, and 2.67 percent in Niger.

Figure 1.
Figure 1.

Comparison of total local resources in WAEMU countries

Citation: IMF Staff Country Reports 2015, 291; 10.5089/9781513580296.002.A001

6. The total local resources of Mali’s subnational administrations significantly increased in recent years but remained modest at CFAF 110 billion, or 13 percent of national tax revenue and 2.05 percent of GDP, in 2013. This increase is primarily due to increased transfers to local administrations, which represented roughly 75 percent of total local resources in 2013. At CFAF 7,200 per capita, however, total local resources are limited.

7. A broader comparison of the share of total local resources6 to total public resources reveals that sub-Saharan African countries have some of the lowest ratios in the world (cf. Figure 2). Ethiopia and South Africa, which are federal countries, have the highest ratios in Africa with 75 percent and 64 percent, respectively. In Latin America, a large number of countries – such as Argentina (44 percent), Brazil (39 percent), Colombia (37 percent), and Bolivia (31 percent)7 – have some of the largest shares of local public resources. They are joined by some Asian countries such as China (67 percent), India (45 percent), and South Korea (39 percent). Among non-OECD European countries, Russia (37 percent) and Ukraine (28 percent) lead the rankings. Among the so-called unitary states of Africa, Algeria stands out in terms of total subnational jurisdiction resources, which represented 4.7 percent of GDP in 2012 (cf. Chambas, 2013).

Figure 2.
Figure 2.

International comparison of the proportion of local resources to total domestic public resources

Citation: IMF Staff Country Reports 2015, 291; 10.5089/9781513580296.002.A001

Source: Indicators14 calculated based on IMF Government Finance Statistics (GFS) data.15

A taxonomy of local financial resources

Local administrations’ resources can be classified in terms of five categories.

Local tax revenue, consisting of taxes (impôts and taxes)8 governed by the General Tax Code and specific laws.9 This revenue may be collected by local authorities or collected by the central government and automatically transferred to the local authorities.

Local nontax revenue, which are all revenue other than taxes, consist essentially of:

  • ➢ Revenue from the operation and provision of services, derived from the remuneration of services provided by local authorities such as the issuance of administrative instruments; removal of household waste; educational or recreational services; roads and highways; water, electricity, and sanitation (pest control, public latrines, etc.); or fuel distribution infrastructure. In certain cases this revenue also includes the shared proceeds of fines.

  • ➢ Revenue from the use of local public services or property. This includes, for example, market operator fees,10 street lighting fees, parking fees, or license fees. It also includes revenue from the lease or sale of property owned by the cercle or commune.11

Transfers from the central government, which are either conditional or unconditional.

  • ➢ Conditional transfers imply that the funds allocated must be disbursed for a specific purpose, e.g., for the provision of public goods or services defined by the central government.

  • ➢ Unconditional transfers, by contrast, may be disbursed however the local government deems most appropriate.

Borrowing, generally used by subnational jurisdictions to finance local public investment. In general, borrowing capacities are limited strictly by law in Francophone African countries and require the existence of a sufficiently developed financial market.

Grants from the country’s technical and financial partners, which can also represent a significant resource of funds.

8. The concept of autonomy of subnational jurisdictions in Mali is highly ambiguous and subject to dispute. The only evaluation of local government expenditure (or revenue) relative to total public expenditure (or revenue) is inadequate because it does not appropriately capture the extent of discretion or power effectively exercised by local jurisdictions over their expenditure or revenue. For example, in Ethiopia, an apparently highly decentralized country (cf. Figure 2), local administrations have very little autonomy because their resources come mainly from central government conditional transfers, which must be allocated to the expenditures decided by the central government.

9. The Malian subnational jurisdictions’ resources consist of local tax revenue, local nontax revenue, intergovernmental transfers, borrowings, and grants (cf. Box 1). This taxonomy defines the local own revenue (LOR) used in assessing the subnational jurisdictions’ financial autonomy (Chambas, 2010). Local own revenue in the broad sense of the term (LORB) are tax and nontax revenue raised for the subnational jurisdictions’ benefit, regardless of whether it is collected locally. These resources are an imperfect indicator of the jurisdictions’ financial autonomy. Local own resources in the narrow sense of the term (LORN) are tax and nontax revenue collected by subnational jurisdictions for their own benefit. It also measures local decision makers’ revenue collection efforts and promotes a degree of accountability by establishing a link between the local tax burden and the delivery of local public services.

10. LORB averaged 65 percent of total local revenue in Mali over the period 1995-2008 and decreased to 28 percent in 2013 following a significant increase in transfers, particularly transfers from external partners12 (cf. Figure 3). LORB represented only 18 percent of total local resources in 2013.13 In light of Box 1, the sum of LORN, unconditional transfers, and borrowings would serve as a consistent measure of subnational jurisdictions’ financial autonomy.

Figure 3.
Figure 3.

Comparison of local own resources (broadly defined) in WAEMU countries (CFAF thousands)

Citation: IMF Staff Country Reports 2015, 291; 10.5089/9781513580296.002.A001

B. The risks of a poorly managed decentralization process

11. Tax decentralization, particularly improving local taxation, entails a number of risks that should not be overlooked.

Local resources, crowd-out effect on central government tax revenue, and collection costs

12. One of the greatest risks in raising local tax revenue is a resultant reduction in central tax revenue. The human and material resources of central tax administrations are limited, and recent reforms have obliged them to focus on a small number of large and medium-size businesses that are both taxpayers and collectors of central taxes. Collecting local taxes presents a radically different set of problems because they are direct taxes paid by a large number of taxpayers. Recent experience in Côte d’Ivoire shows that giving priority to the collection of local taxes from taxpayers at large may force tax administrations to dissipate resources collecting taxes with little revenue potential, resulting overall in significant loss of tax revenue.

13. In addition to the crowd-out effect on central taxation, local taxation entails the risk of increasing tax collection costs. The proliferation of direct local taxes observed in Mali wastes efforts on collecting obsolete taxes with only marginal revenue potential, such as firearms tax or cart tax.

Local taxes and market fragmentation

14. One of the essential conditions of efficient decentralization is the preservation of the single market (Weingast, 1995). The absence of a common market or restrictions on the mobility of factors of production limit and skew competition between jurisdictions and encourage them to adopt rent-seeking behavior through the application of taxes with high revenue potential that are nonetheless harmful to economic activity. In the case of Mali, a WAEMU member, the objective is also to preserve the single common market.

15. Accordingly, an objective of decentralization is to reconcile the single market and local taxation. The central government, through its role in the coordination and control of taxes applied, plays a critical role in ensuring a single national market. The preservation of a single domestic market assumes that only the central government may tax mobile factors of production or tradable goods; it also assumes oversight of locally assessed taxes and fees.

Competition among subnational jurisdictions

16. Horizontal tax competition between jurisdictions may lead to a race to the bottom that is harmful to all in terms of public finances. For example, since the authority to determine motorcycle registration fees was transferred from the central government to the communes, competition between the communes has exerted downward pressure on these fees.

17. Vertical competition is also a danger, because it could lead to excessive taxation of the same tax base by different levels of government. This phenomenon, observed particularly in federal countries, is the result of ambiguous allocation of taxation authority between two levels of government (central and local) and could ultimately harm economic activity.

Decentralization, soft budget constraint, and raising local resources

18. In the absence of central government oversight, the subnational jurisdictions have already contracted loans for which the cost of a potential default is liable to be covered from central government finances. While the subnational jurisdictions’ borrowing capacity is limited, it entails substantial systemic risks (cf. Box 2). With the exception of Bamako, which has borrowed CFAF 212.6 million, only two regions have contracted loans of relatively small amounts (Kayes, CFAF 2.2 million, and Ségou, CFAF 3.7 million). However, if the subnational jurisdictions’ borrowing capacities increase, the risk of lack of fiscal discipline on their part could result in debt distress for the central government and threaten macroeconomic stability.16 The risk is exacerbated by the fact that the resources available to local jurisdictions are insufficient to cover their expenses and oversight of their borrowing commitments is weak.

Local taxes and corruption

19. One of the risks of local taxation is that it may facilitate corruption, which affects not only local tax revenue but all taxes (Attila et al., 2009). This is because local taxation and tax administration are not subject to the same level of oversight or organization as central taxation and tax administration.

Tax decentralization and unequal distribution of resources between subnational jurisdictions

20. Decentralization can exacerbate inequalities between subnational jurisdictions. Map 1 illustrates the uneven distribution of LOR, which ranged from CFAF 1,131 billion in Mopti, or CFAF 590 per capita, to 7,052 billion in Bamako, or CFAF 3,898 per capita. The revenue potential of the different Malian jurisdictions’ local taxes differs considerably depending on their choice of local tax laws as well as economic and demographic characteristics. Such disparities are not unusual in the region. Dakar’s local resources represented 44 percent of total local resources for a population of 950,000 inhabitants in 2008; in the same year, 90 percent of the Beninese population accounted collectively for only 62 percent of total local resources, with the commune of Cotonou accounting for the remainder.

Map 1.
Map 1.

Local own revenue (broadly defined) in Mali, 2008

Citation: IMF Staff Country Reports 2015, 291; 10.5089/9781513580296.002.A001

Source: mission.

The cost of tax decentralization

21. While the aim of decentralization is to increase the efficiency of public policies, the cost, particularly the administrative cost, should not be underestimated. Decentralization entails a risk of duplication of public goods or services, which is admittedly higher in developed countries. However, the cost of oversight of the subnational jurisdictions is far from negligible in developing countries and generally requires the central government to oversee budget execution. The creation of a subnational civil service is also desirable. However, the experience of certain federal countries like Canada is instructive: municipal government employees earn significantly more than others civil servants holding comparable jobs. This is because the communes have relatively less leverage in negotiating with national unions.

22. Decentralized natural resource management may also exacerbate disparities, weakening the country’s unity. The mining communes of Kayes, a poor region, collect the business license tax (patente) (BLT) from mining companies. Urban jurisdictions such as Bamako with large bases of economic activity, including large businesses, collect BLT and are also able to assess fees that generate substantial revenue. Some of the outlying jurisdictions may also raise BLT revenue and fees generated by dynamic business activities. However, many jurisdictions located in rural or arid areas struggle to raise small or negligible amounts of LOR. No foreseeable innovation in local tax laws is likely to provide the financial resources needed to cover the delivery of sufficient services to reduce disparities in access to public services. Only transfers from the central government can facilitate relative equalization among jurisdictions.

Borrowing and related systemic risks

If the central government is ultimately responsible for fiscal balances, the ability of subnational jurisdictions to borrow entails substantial systemic risks.

These risks arise from the problem of moral hazard, also known as “soft” budget constraints.”17 Under the supervision of the central government, a local government with authority to borrow may circumvent the rules of fiscal discipline to increase its resources and maximize its own objectives. Having the advantage in terms of information, the local government knows that it would be difficult for the central government to detect its behavior and would have no choice but to come to the rescue. The local government anticipates relative fiscal impunity, then, since the impact of its behavior affects all local administrations while it alone profits from its excessive spending.

The mere threat that the central government might not come to the aid of local administrations in financial difficulty is not credible because the local administrations know that the default of one local government could have significant macroeconomic consequences, and the central government would be obliged to intervene – as occurred, for example, in Argentina, South Africa, and Brazil in the 1990s (Prud’homme, 1995).18

The institutional and legal framework for borrowing is critical, then, because it determines the subnational administrations’ capacity to adopt dangerous strategic behavior. However, the surveillance of the subnational administrations’ financial position and borrowing terms under laws enacted to prevent insolvency – the obligations to balance the budget annually, to disclose prudential risk ratios, to repay principal and interest from own resources; and the prohibition on borrowing to finance operating expenditure or refinance outstanding loans – may be insufficient because local administrations retain the incentive to circumvent the rules.

Accordingly, while borrowing is useful to distribute investment expense over time, the risk of undisciplined spending provides yet another argument against the use of borrowing by local authorities. Those arguments include the inefficient operation of financial markets, the unpredictability of future financial resources and expenditure, insufficient managerial capacities and technical and financial knowledge, and an incentive for incumbent policymakers to borrow excessively because they reap the benefits immediately while the costs of borrowing are postponed to the future.

III. Local Taxation

23. Local taxes in Mali are governed by the CGI, the law governing subnational jurisdictions (Code des Collectivités Territoriales), Law 96-058 of October 16, 1996, and Law 211-034 of July 15, 2011.19 Table 1 identifies the local taxes and the respective rates and tax bases. The revenue figures for each tax, compiled by the National Directorate of Treasury and Public Accounting (DNTCP), are provided in Table 2. The local taxes are broken down by the type of tax base: individuals, economic activity, extractive activity (mining), vehicles, and other.

Table 1.

Taxes assessed by subnational jurisdictions

article image
article image
Sources: CGI et Lois.
Table 2.

Subnational jurisdictions’ tax revenue (CFAF millions)

article image
Source: Trésor.

24. The DGI collects 40 percent of the local jurisdictions’ revenue (cf. Table 3). In 2014, the BLT accounted for 93 percent of local revenue collected by the DGI. The ordinary BLT accounts for the majority of this revenue (82 percent), with receipts from the BLT applicable to administrative contracts (5.53 percent) and the BLT components of the AMBT (3 percent) the road transport tax (TTR) (2.5 percent) representing relatively small amounts.20 The road maintenance tax (taxe de voirie) represented 4.43 percent of local taxes collected by the DGI in 2014, most of which is collected with the professional BLT. There are inconsistencies between two laws governing the collection of taxes that the authorities indicated are currently being resolved. Specifically, the Code des Collectivités Territoriales provides that taxes are collected by the Treasury personnel, while Law 036 attributes this responsibility to the DGI. The differences in the legal texts may be due in part to the negligible amounts at issue.21

Table 3.

Subnational jurisdictions’ tax revenue collected by the DGI

(CFAF millions)

article image
Source: DGI

A. Local taxes assessed on business activities

25. Local taxes on business activities represented 45 percent of local revenue in 2013, with the BLT accounting for nearly 86 percent of that amount (cf. Table 2). The predominant role of the BLT is common in most of the region’s countries (cf. Figure 4). This tax is considered a license to practice, applicable to all professionals regardless of the scale of their activity. The BLT applicable to all business entities consists of a fixed fee and a proportional fee:

  • ➢ The fixed fee is determined according to the nature of the activity, the location where it is practiced,22 and the annual turnover; the fee is capped at CFAF 1 million and becomes applicable to a broad range of activities if they generate turnover of CFAF 1 billion or more.

  • ➢ The proportional fee is assessed at 10 percent of the rental value of the business premises, machinery, and various installations.

Figure 4.
Figure 4.

Composition of local tax revenue, 1995-2008 (CFAF thousands)

Citation: IMF Staff Country Reports 2015, 291; 10.5089/9781513580296.002.A001

Figure 5.
Figure 5.

Composition of local tax revenue in WAEMU countries, 1995-2008

Citation: IMF Staff Country Reports 2015, 291; 10.5089/9781513580296.002.A001

Alternative methods of calculating the BLT are established for certain economic activities. For example, the BLT applicable to public procurement is 2.5 percent of the contract amount. For the supply of water and electricity, it is 1 percent of turnover. Exemptions from the BLT are also provided for public entities, entities exempt from corporate income tax (1991 Mining Code, Investment Code), and livestock exporters.

26. Beginning in 2014, businesses with turnover of less than CFAF 50 million no longer pay the BLT directly, but continue to pay it indirectly in the form of the AMBT. The tax was simplified considerably in 2014 and is now a 3 percent turnover tax, in accordance with best international practices (cf. Rota-Graziosi et al., 2014a). The DGI distributes 10.45 percent of receipts to the subnational jurisdictions as the BLT. Some activities generating turnover below the CFAF 50 million threshold are taxed on actual income and are required to pay a BLT. Examples include tax consultants, accountants and auditors, businesses eligible for the provisions of the Investment Code, and certain professionals.23

27. A similar tax, the contribution des licences, is assessed on the sale of alcoholic beverages. Businesses required to pay the licence are also required to pay the BLT. Like the BLT, the licence consists of a fixed fee ranging from CFAF 9,000 to CFAF 150,000 and a proportional fee of 10 percent.

28. The tax base and rates of the BLT and licence are governed exclusively by the CGI. The fixed fee varies from one commune or region to another and is defined in the CGI, with no leeway provided for the subnational jurisdictions. In practice, the BLT is a capital tax for businesses generating turnover of more than CFAF 50 million. The DGI lacks comprehensive information on rental values, which are the basis for this tax (and incidentally the real property tax). Instead, it uses the original cost of fixed assets, which are taxed at the rate of 0.5 percent (CGI, article 140). No provision is made for depreciation of assets. This practice is applied in particular to mining companies, whose initial investments are substantial (several hundred million dollars).

29. Total BLT revenue represents a significant source of local jurisdictions’ revenue (40 percent of LOR), but remains marginal in terms of GDP (0.13 percent of GDP in 2014). Of the CFAF 7.731 billion total BLT receipts, the professional BLT accounted for 82 percent; the BLT for administrative contracts, 5.5 percent; the BLT included in the AMBT, 3 percent; and the TTR, 2.5 percent.24

30. BLT revenue is not concentrated solely in Bamako but distributed more or less throughout the Malian territory. As an exception to the rules organizing the tax administration, BLT revenue is not managed centrally by the Directorate of Large Business Taxpayers (DGE) or the Directorate of Medium-Size Business Taxpayers (DME), but is collected by local tax offices. The District of Bamako produces 47.6 percent of BLT revenue (with 24.3 percent and 7.7 percent corresponding to the DGE and the DME, respectively). The BLT for the regions of Kayes (a mining region) and Sikasso (an agricultural region with dynamic economic activity) represent 31 percent and 12.5 percent of BLT revenue, respectively. In the regions of Timbuktu, Gao, and Kidal, BLT collections, historically low in view of limited economic activity, have become negligible.

31. Receipts from the BLT component of the AMBT totaled CFAF 250.1 million in 2014, which is extremely low in comparison to the CFAF 7.731 billion generated by the BLT. The distribution of receipts is similar to the BLT applied to businesses taxed on actual income. More than half, 51.6 percent, come from the District of Bamako. Among the regions, Kayes and Sikasso generate the largest proportions while those of Timbuktu, Gao, and Kidal are insignificant. Finally, receipts from licences are negligible for all jurisdictions.

B. Local taxes on individuals

32. Local taxes paid by individuals consist primarily of income tax, the Regional and Local Development Tax (TDRL), and the real property tax. Revenue from the tax on wages and salaries (ITS) (16 percent of local tax revenue in 2016) comes from income taxes paid by civil servants of the subnational administrations, particularly employees of the national education system working at schools located within the commune, cercle or region. The ITS rates established by the CGI are applied uniformly throughout Mali. In other words, local ITS receipts are not a local surtax, as is the case in some federal countries, but the distribution of centrally managed revenue.

33. The TDRL is a flat, per capita tax payable by individuals over age 14. The rate is fixed by the CGI and ranges from CFAF 875 in the region of Timbuktu to CFAF 3,000 for the District of Bamako. A subnational authority may modify its own rate within a limit of plus or minus 25 percent. The TDRL is supplemented by the road maintenance tax, which is paid by families and businesses that pay the BLT. The rate of this tax is the larger of 5 percent of the BLT or CFAF 3,000 per family in the District of Bamako (CFAF 2,000 elsewhere).

34. TDRL receipts represented 21 percent of local tax receipts in 2013. The TDRL generates considerable revenue in rural areas but, like the road maintenance tax, produces very little in urban areas (Bamako) as shown in Figure 6. A combination of factors account for much of the disparity: the TDRL is collected by village leaders, it is one of the only taxes paid directly in some rural areas, and it has proven more difficult to collect in denser urban areas where taxpayers are protected by a degree of anonymity.

Figure 6.
Figure 6.

Geographic distribution of per capita taxes on individuals (TDLR and public road tax)

Citation: IMF Staff Country Reports 2015, 291; 10.5089/9781513580296.002.A001

Note. 2013 except for the three regions Kidal, Gao, Timbuktu (2010).Source: DNTCP and mission calculations.

C. Real property tax in Mali

35. The real property tax established by Law 034 in 2011 is assessed at the rate of 3 percent of the rental value of buildings or unimproved land held for more than three years. It applies to individuals as well as legal entities. Property tax receipts collected by the DGI are particularly low (CFAF 195 million in 2014, cf. Table 3) in regard to the experience of other countries in the region (cf. Figure 4).

36. The local jurisdictions’ performance in terms of property tax varies considerably (cf. Figure 7). The District of Bamako generated over 80 percent of property tax receipts in 2014 and posted the highest real property tax per capita (CFAF 65 per capita).

Figure 7.
Figure 7.

Property tax revenue by region in 2014 (per capita)

Citation: IMF Staff Country Reports 2015, 291; 10.5089/9781513580296.002.A001

Source: DGI and mission calculations.

37. The real property tax distributed to the subnational jurisdictions is actually derived from sharing of the Property Income Tax (IRF). According to the mission’s discussion, the DGI collects property tax only on rental income. The IRF tax rate was 15 percent prior to the introduction of the law establishing the real property tax. The rate was reduced to 12 percent thereafter, with the difference of 3 percent representing the real property tax rate.

38. The failings of the real property tax can be attributed to the combination of a tax base (rental value) that is inappropriate to the Malian context and broad exemptions. The issue of the tax base is addressed in the following chapter. The real property tax exemptions apply in particular to the buildings occupied by the owner or dependent family members and agricultural land.

D. Local taxation of natural resources and other local taxes

39. The local taxation of natural resources (mines, forests, fisheries) involves multiple ministries and sector codes (Mining Code, Forest Code, etc.). The various taxes assessed on the exploitation of natural resources should be consolidated in order to rationalize them. This is particularly important given that institutional fragmentation in Mali – where the Ministry of Government Lands is separate from the Ministry of Finance – undermines the management of the different tax bases.

40. The taxation of mining also includes local components. The BLT applicable to mining companies, which is a tax on the mining companies’ initial investments, represents more than half of total BLT receipts, or CFAF 3.4 billion (cf. Table 2). The three Mining Codes (1991, 1999, and 2012) establish the surface fees, which distinguish between the types of permits25 (CFAF 1,000/sq. km. to CFAF 2,000/sq. km. for exploration permits under the 2012 Mining Code, and CFAF 100,000/sq. km. for production permits) and the category of mineral exploited. The Directorate of Government Lands collects the surface fees for production permits, which are in rem rights similar to property deeds, while the Directorate of Mines collects the fees on exploration permits. The Ministry of Mines is in the process of consolidating the collection of the two fees.

41. Artisanal quarry and mining activities are taxed under special local provisions such as the additional mining tax or the quarry or gold panning tax. The receipts are insignificant and illustrate the difficulty of properly taxing artisanal mines26 under the current tax system.

42. The other principal local taxes are vehicle taxes. These taxes vary according to vehicle capacity, and generated 6.5 percent of local tax revenue in 2013. The lack of tax revenue from construction permits or private use of the public domain compromises the institution of real property taxation. Efficient management of construction permits is essential for the control of real property. The lack of revenue, in particular from a construction permit fee, is of particular concern in the context of Bamako given its sustained growth rate (over 7 percent per annum). This indicates failings on the part of the administrations involved in the management and utilization of government lands and/or a possible source of corruption.

IV. Proposed Reform Strategy

43. The local taxes described in the preceding chapter fail to generate sufficient revenue or promote accountability on the part of the subnational jurisdictions. The taxes concerned are assessed on bases that are especially difficult to establish.27 The yields of these taxes are particularly low, a finding widely shared by multiple studies on decentralization in Mali and in West Africa in general.28 The few taxes such as the BLT or TDRL that generate appreciable resources are either highly distortionary because they increase the cost of capital invested, as in the case of the BLT, or their revenue potential is quite limited if not declining with Mali’s urban development.

44. This chapter proposes a reform strategy designed to rationalize local taxes by enhancing the accountability of subnational jurisdictions and strengthening the capacities of central and local governments to raise additional resources. The strategy is based on four key components:

  • The institution of a real property tax that takes account of the failings of the existing cadaster and taxes;

  • The sharing of VAT among subnational jurisdictions and the central government to ensure stability in the allocation of resources among local jurisdictions (the modalities of transfer are discussed in the following chapter);

  • The improvement of the AMBT which, being an alternative tax, has become the principal means of direct taxation of business activities (for businesses with turnover of less than CFAF 50 million); and

  • The elimination of specific local taxes, to be replaced by fees for the provision of public services or access to certain local public goods.

A. The institution of a real property tax based on property value

45. The proposed property tax reform aims to increase public revenue collection on the order of 0.5 percent of GDP, or CFAF 30 billion. While property tax receipts in Mali are particularly weak (CFAF 194 million in 2014),29 they represent an average of 0.6 percent of GDP in developing countries (cf. Bahl et al. 2008) compared to over 2 percent of GDP in OECD countries and as much as 4 percent in Canada. Property tax revenue represents close to 0.4 percent of GDP in Côte d’Ivoire despite the difficulties of implementing this tax, the lessons of which are taken into consideration in the proposed reform.

46. The property tax review should incorporate existing taxes such as the property tax, the BLT, and the TDRL, which should be eliminated or combined with the real property tax to be created. The tax base should vary depending on whether the owner or occupant is an individual or entity, and could incorporate elements of the previous taxes to ensure that the transition does not entail any loss of revenue for the central government or local jurisdictions.

47. In Mali, as in most developing countries, it is not possible to establish a real property tax regime similar to those of developed countries, which are based on the rental or market value of the tax base. The Ministry of Government Lands plans to institute a modern cadaster over a four-year horizon with support from Mali’s development partners. The failings of the cadaster are widely shared in Africa and beyond; cadastral reform, programmed several times in the past, will entail a long-term effort. The existence of a real property market outside the urban centers is far from certain. Finally, the issue of property rights, particularly in rural areas, raises a number of difficulties.

48. The point of departure for real property tax reform is to base the calculation on an indexed value per hectare rather than the rental value. The indexed value could be adjusted according to geographic location and other factors. Such factors could include the type of building, the economic use of the property, proximity to public or central goods and services, connection to the electric or water distribution network, average consumption of electricity, cell phone minutes, etc., in the area. The central government (in particular the DGI) and local elected officials should be involved in determining those criteria.

49. A real property tax based on an indexed value of owned or occupied property would be fairer than the current TDRL because it would reflect differences in access to public infrastructure. Real property values incorporate local public goods and services. Even the villages of the same rural commune have different levels of access to public goods, such as distance from wells, schools, and the like. Moreover, the proposed approach would incorporate and thereby strengthen the planned DGI reform that would institute a dwelling or residence tax for cities, a so-called urban TDRL.30

50. The second part of the property tax reform is to eliminate the current exemptions from property tax, in particular the exemptions for buildings occupied by the owner or family members and agricultural land. In addition to significant revenue gains, the taxation of owner-occupants under the real property tax, like the taxation of agricultural lands, would replace the TDRL. The current exemptions are not justified in terms of economic efficiency because they encourage speculation and the holding of land as a nonproductive asset. China and Singapore explicitly increased the taxation of real property to fight against property speculation, which was a serious risk for the stability of their respective economies (cf. Norregaard, 2013). The taxation of agricultural property would in turn promote the more efficient use of land, thereby stimulating economic development (particularly agricultural development). Namibia recently instituted a 0.75 percent tax on the value of agricultural land; the authorities’ stated initial objective was to promote efficient use of this factor of production.

51. Mali’s sustained rate of urbanization and the rapid growth of the city of Bamako justify the taxation of agricultural property, particularly in the periurban areas. Those lands have high economic value, and converting them into buildable lots creates significant gains that are not adequately taxed by the Directorate of Government Lands responsible for taxing gains on the sale of real estate (less than CFAF 900 million in 2014). An indexed real property tax would allow the substantial real property income generated each year by the development of urban centers to be taxed.

52. The proposed real property tax would incorporate the BLT currently paid by businesses generating turnover of over CFAF 50 million. The elimination of the BLT, a tax created in France in 1791 that has undergone multiple reforms,31 is needed to promote investment in Mali because the tax is assessed directly on businesses’ productive investments and fixed assets. To limit the loss of revenue associated with the reform, the authorities could estimate the real property tax payable by business and calibrate the applicable index values based on the historical BLT revenue.

53. In fact, since the introduction of the AMBT, the current BLT system introduces distortion to the detriment of modern, formal activities. The BLT penalizes businesses that invest because the amount of the tax automatically increases with the original value of businesses’ fixed assets. The tax falls most heavily on the industrial sector because those businesses are more capital intensive than service providers and commercial activities. The BLT thus compromises the neutrality of the tax system by increasing the cost of fixed assets and introducing distortions between activities according to their capital requirements.

54. Exemptions from the BLT created for certain activities, notably those benefiting under the Investment Code or conventions, serve to reduce distortions for certain activities. However those exemptions increase the complexity of the BLT and create distortions between exempt and nonexempt activities. Moreover, they increase uncertainty in regard to local jurisdictions’ resources because local authorities are not involved or even aware of the central authorities’ decisions to implement BLT exemptions.


  • Create a real property tax based on property value.

    • Introduce an alternative to the rental value in the form of an indexed value based on the area of land or type of real property, which could take account of certain factors of production or consumption (access to local public goods and services, geographic location, etc.).

    • Eliminate property tax exemptions for agricultural land (particularly on the periphery of urban areas), unimproved land held for less than three years, and property occupied by the owner or family members.

    • Include a revised TDRL in the real property tax and eliminate the BLT.

B. Sharing of VAT between subnational jurisdictions and the central government

55. The VAT is a modern tax applied to the final consumption of households and the intermediate consumption of businesses with turnover below the VAT threshold of CFAF 50 million. In Mali, as in most developed or developing countries, the VAT is more than a simple consumption tax. The mechanism for collecting VAT throughout the production process serves to secure revenue and create a distinction between taxed businesses that can deduct VAT paid on inputs and businesses that are not taxed. If the untaxed businesses are similar to informal businesses, the VAT is a tax on inputs imported or provided by formal businesses consumed by the informal businesses. Analyses by the IMF Fiscal Affairs Department indicate that the proportion of VAT receipts based on the intermediate consumption of informal businesses can represent close to 70 percent of total VAT receipts, as in Benin (60 percent in Mauritania).

56. The sharing of VAT receipts between the subnational jurisdictions and central government not only ensures stable revenue for the former but also reinforces the role of VAT in the Malian tax system. The instability of intergovernmental transfers described in the following chapter compromises the success of decentralization. Converting the VAT (which is regularly threatened by exemptions for certain interest groups) into a shared tax would strengthen it because any erosion of the VAT base would affect both the central and subnational governments. The expected advantages and costs of those exemptions would then be more widely debated in the political sphere.

57. Although VAT revenue is shared, it is nonetheless collected exclusively by the central administrations of the DGI and the DGD. The tax rate and base are governed by the CGI and the so-called central tax system, which would become more a national tax system. Local management of the VAT is liable to entail a number of difficulties in administering the tax, which would undermine its objectives of revenue and neutrality. Experiences with VAT in federal countries like Brazil or India illustrate the failings of a VAT managed simultaneously by local (provincial) and central governments. Canada would appear to be an exception, with the majority of VAT managed by the federal administration except for the province of Québec.

58. A shared tax system as proposed here exists in a number of federal and unitary countries and has often served as a means of reducing central government’s grants to the local administrations. In Germany, for example, VAT revenue is shared between the local districts and cities (2 percent), the states (46 percent), and the central government. In Spain, Law 22/2009 of December 18, 2009 reformed the system of autonomous communities by increasing the amount of national taxes redistributed to them from 33 percent to 50 percent for income tax and VAT, and from 40 percent to 58 percent for special taxes ([alcoholic] beverages, petroleum products, tobacco). In France, the departments have received a portion of the domestic tax on petroleum products and the special tax on insurance contracts since 2005. In Great Britain, a portion of corporate income tax is transferred to the subnational administrations. The 2011 reform in Belgium eliminated some grants from the center and the ability to assess so-called “extended” additional centimes on revenue from taxes on individuals. In Senegal, 2 percent of VAT revenue is allocated to local jurisdictions (cf. Box 3), while in Cameroon the principal central taxes (VAT and corporate income tax) are subject to additional centimes of 10 percent.

59. The Malian government already shares some central taxes with subnational jurisdictions. The current real property tax is in fact assessed on the real property income tax (IRF) or in the same manner an additional tax of 25 percent (3 percent in regard to the 12 percent IFR rate) for the benefit of local jurisdictions. Similarly, the remittance to communes of ITS paid by government officials working there could be viewed as a sharing of that central tax based on the civil service wage bill.


  • Share VAT receipts in order to guarantee a stable flow of resources to local authorities. The share to be remitted to the subnational jurisdictions and the modalities of distribution among those jurisdictions remain to be determined (cf. following chapter).

C. Improvement of the AMBT

60. The AMBT was significantly simplified in 2014. The taxes assessed at the rate of 3 percent on the turnover of businesses generating less than CFAF 50 million. It contributes to the rationalization of direct taxation in Mali because firms generating turnover above the threshold must pay the larger of a 30 percent profit tax and a minimal 1 percent tax on turnover.

61. The subnational jurisdictions should be involved in managing the AMBT. Article 73 of the CGI provides that payment of the AMBT releases the taxpayer from the obligation to pay “all other taxes and fees established in Titles 1 and 2 of this Code.” This provision exempts small and medium-size businesses from most local and central taxes (other than VAT).

62. The distribution of AMBT revenue should be discontinued because it leads to confusion. In addition to the BLT component representing 10.45 percent, the AMBT also represents the profit tax (50.90 percent), the road maintenance tax (0.55 percent), the Chamber of Commerce contribution (1.1 percent), VAT (33 percent), the fixed employer contribution (2.47 percent), the tax on housing (0.35 percent), the vocational training tax (0.18 percent), and finally, income tax withholding on wages and salaries (1 percent). Including the VAT among the taxes that make up the AMBT invites confusion: an individual who pays the AMBT may believe he is not required to pay the VAT on inputs because it is included in the AMBT. Moreover, the previous distribution of AMBT revenue makes [the AMBT a complex hybrid of central and local taxes]. Such complexity makes it difficult for many small economic operators to fully understand the tax and is especially pointless in view of the small amounts collected (AMBT revenue totals CFAF 2.7 billion).

63. Allocating all AMBT revenue to the subnational jurisdictions would promote clarity and simplification. This could be implemented by deducting management fees on the order of 5 percent to 10 percent if the DGI continued to manage the tax. The Treasury and its regional units could also collect the tax. Individual jurisdictions could also decided to collect the tax directly or delegate collection to the DGI. This simplification measure would promote a relationship of accountability between small operators and local authorities.

64. If AMBT receipts were allocated to the subnational jurisdictions, a number of local taxes could be eliminated. For example the taxes on livestock, mills, wagons and carts would be integrated into the AMBT. In fact, the elimination of those taxes and others would make it clear that the AMBT is a minimum tax.


  • Improve the AMBT by allocating some or all receipts to the subnational jurisdictions, by involving them in the modalities of collection (DGI, Treasury, etc.), and by eliminating certain specific taxes (taxes on livestock, mills, wagons and carts, etc.).

D. Replacement of some local taxes with fees

65. The subnational jurisdictions play an important role in local development by providing collective services or infrastructure. Numerous examples exist, such as marketplaces, bus stations, facilities for transhumant livestock herders, docks and quays, wells or water supply systems, local electricity production, and collection of household garbage.

66. Some local taxes should be replaced by user fees. The advantage of a fee over a tax is a higher willingness to pay because the taxpayer then becomes a paying user. Numerous surveys among populations have found that users are prepared to pay for services provided the service meets their needs.32 However, low and unstable incomes, particularly in poor rural areas, represent obstacles to the acceptance of fees for services.

67. The vehicle exit tax (taxe de sortie des véhicules) applicable to the public transportation of persons of merchandise is an example of a local tax contrary to the free circulation of persons and merchandise that undermines the unity of the national market. The tax is paid by bus drivers and public passenger vehicle operators upon exiting the communes. It is a nominal amount per vehicle that varies between the communes. This tax should be eliminated and replaced by fees for the use of bus stops and/or merchandise loading and unloading areas. The boat tax could also be converted to a fee for access to anchorage rings or pontoons.

68. A commune’s allocation of buildable urban land within the public domain to private individuals should serve as an important resource for urban local authorities. Because of continuous urban expansion, this resource has a recurrent character in major cities, particularly in Bamako. Land allocation procedures are often accompanied by speculative conditions and a lack of transparency.

69. Local jurisdictions increasingly demand the right to collect royalties for the exploitation of natural resources – both renewable (forests, fisheries resources) and nonrenewable (quarries, mines) – for their own benefit. Law 00-044 of July 7, 2000 includes fees for artisanal gold production permits, fees for the opening of artisanal quarries, and taxes on wood harvested from the central government’s forest domain in Malian local jurisdictions’ tax resources. Local jurisdictions also receive the proceeds of trade and professional licenses. This leads to substantial disparities – in favor of communes in the Kayes region in particular – which can fuel a sense of injustice and result in waste. Some communes may also be encouraged to overexploit resources (quarries) without ensuring that royalties are effectively collected.

70. The Ministry of Mines and the other ministries concerned are compiling an inventory of revenue collected on behalf of communes located in the vicinity of mining operations. This inventory is indispensable in order to contemplate reform in the allocation of tax and nontax revenue from mining activities, which this report does not directly address. The allocated revenue may be direct, tied to mining taxes (surface fees); general (BLT, real property tax); or indirect (e.g., the deductibility of mining companies’ outlays in connection with grants or public interest works for subnational jurisdictions [for purposes of ] corporate income tax). Revenue transparency is also crucial to establish accountability on the part of local elected officials vis-à-vis their constituencies.

71. Market fees and parking fees represent the majority of income from the local public domain. The fees on market stalls represented 9 percent of nontax revenue in Mali, and taxi and transport vehicle parking fees represented 12 percent in 2013. Market fees are the principal nontax resource in West African countries (35 percent in Côte d’Ivoire, 24 percent in Togo, 22 percent in Niger, 13 percent in Senegal, and 10 percent in Benin). Revenue from household garbage collection represents a smaller share of receipts, totaling 17 percent in Senegal and 3 percent in Côte d’Ivoire. In Mali, they represented less than 2 percent of nontax revenue in 2013.

72. Income derived from the public domain represents the largest proportion of nontax revenue. In 2013, the contribution of this income to local nontax resources was 47 percent compared to 10 percent for income derived from operation and the delivery of services. In contrast with Mali, Benin, Togo, and Niger derived the majority of their nontax revenue from operation and delivery of services during the period 1995-2008 (CEF Figure 8).

Figure 8.
Figure 8.

Comparison of the composition of nontax revenue in 2008

Citation: IMF Staff Country Reports 2015, 291; 10.5089/9781513580296.002.A001

Source: Treasury and mission calculations.


  • Eliminate certain specific taxes such as the vehicle exit tax and the boat tax, and convert them to fees.

V. Fiscal Transfers

A. A taxonomy of transfers and their respective effects on the behavior of subnational jurisdictions

73. Any analysis of intergovernmental transfers requires that a distinction be drawn at the outset between conditional and unconditional transfers (cf. Box 1). Conditional transfers must be allocated to specific expenditures, i.e., the delivery of goods or services previously defined by the central government. In that sense they are a means for the central government to direct the actions of local governments. In contrast, unconditional transfers are not subject to any conditions. The local authorities allocate these resources according to their own criteria. They may use them for any good or service or even to reduce the local tax burden.

74. If transfers are not allocated in a transparent and predictable way to subnational jurisdictions, they could undermine the efficiency of local public policies. Transfers that are opaque to taxpayers do not promote accountability on the part of local authorities.33 The relationship between available resources and the delivery of local public services is obscured and citizens are unable to assess the efficiency of local public policies, which effectively deprives them of their power of oversight and sanction. Accordingly, an opaque system of transfers undermines the positive effects expected of decentralization, which are based on the proximity of (and therefore the reduction of information asymmetries between) citizens and local authorities. An effective system of transfers should be transparent, then, in terms of the basic amounts transferred and the system of allocation (cf. infra).

75. Central government transfers change the fiscal behavior of decentralized authorities. The taxation of residents entails costs (political, economic, and social) such that local authorities may prefer to scale back efforts to raise LOR if transfers increase. This crowd-out effect of transfers on LOR varies according to the type of transfer. In Benin, for example, unconditional, predictable transfers that are allocated transparently stimulate LOR (Caldeira and Rota-Graziosi, 2014a).

76. In addition to the crowd-out effect, the very existence of a system of transfers entails a risk for local authorities’ fiscal discipline. Local administrations may spends beyond their resources knowing that central government transfers will finance their deficits after the fact. The local government alone benefits from the spending, while the national population ultimately bears the cost. There is therefore an element of moral hazard in local authorities’ strategic behavior and the central government’s inability to assume a credible ex ante commitment not to provide ex post budget support to a financially troubled local administration. This risk may explain the preponderance of conditional transfers aimed at controlling subnational jurisdictions’ spending.

B. The problems of defining the system of allocating transfers

77. Equalization is a priority objective in establishing a system of allocating transfers. Because decentralization is generally associated with an increased risk of disparities (cf. Section II.B), there [must] be consensus as to the need to incorporate a system of intergovernmental transfers in the decentralization process. The allocation of transfers should then offset the unequal access to public goods and services produced by the unequal distribution of wealth – and the resultant inequitable distribution of the tax base – over the national territory. The pursuit of this objective implies that transfers should be directed toward the poorest local jurisdictions. The equalization transfers are then perceived as ensuring national economic and social cohesion.

78. The equalization objective may run counter to the pursuit of efficient public policies. From the standpoint of economic theory, the economic efficiency objective behind the distribution of transfers is associated with the correction of externalities34 between jurisdictions in the supply of certain local public services. Spillover effects would otherwise lead to a suboptimal supply of local public goods and services. Local authorities are not encouraged to fully assume the cost of a policy that partly benefits neighboring jurisdictions.35 There is also the issue of encouraging the mobilization of LOR such that in addition to fiscal capacities and local requirements, the local authorities’ budget efforts become one of the transfer allocation criteria. From a practical standpoint, it is difficult to simultaneously satisfy the objectives of equity and efficiency. In fact, the poorest jurisdictions are also generally those with the most limited capacities and the weakest performance in terms of raising LOR.

79. Allocation formulas that are overly complex or lack transparency leave room for political manipulation in the allocation of transfers. A wealth of empirical literature reveals that the distribution of transfers between jurisdictions is not based on the sole objectives of offsetting inequalities and reducing inefficiencies, even where allocation formulas are used. The central government may use fiscal transfers in a “tactical” manner to maximize chances of reelection, or in a “partisan” manner to reward political supporters. This phenomenon of manipulating allocation policies has been identified in Ghana by Banful (2010) and Miguel and Zaidi (2003); and in Senegal by Caldeira (2012). Fixed, simple allocation formulas should be preferred, then. It is also important to ensure that local jurisdictions have access to transparent information on the allocation of resources and have the right to challenge any errors in the distribution.36

C. Comparative analysis and assessment of the Malian system of transfers

80. A comparative analysis of WAEMU countries reveals that the amounts of intergovernmental transfers are relatively small. Over the period 1995-2008, central government transfers averaged 20 percent of total local resources in the region. At 35 percent, Mali’s percentage was the highest in our sample. The ratio of transfers to total public resources remained barely marginal until 2008: 2.41 percent in Côte d’Ivoire, 1.05 percent in Mali, 1 percent in Benin, 0.34 percent in Togo, and 0.13 percent in Niger.

81. Transfers to Malian local jurisdictions increased sharply since the 2009 creation of the National Support Fund for Regional and Local Authorities (FNACT), managed by the National Local Government Investment Agency (ANICT). The funds disbursed via ANICT alone totaled CFAF 26 billion in 2009, CFAF 23 billion in 2010, and CFAF 17 billion, or 2.25 percent of national tax revenue, in 2011 (cf. Table 3). Direct subsidies from the central government have decreased, since it now contributes directly to the FNACT. Since 2011, all government transfers are channeled through ANICT, with the exception of conditional operating transfers (civil service salaries).

82. Unlike other countries in the region (cf. Box 3), Mali devotes the largest share of transfers via ANICT to investment transfers, which may penalize the most vulnerable local jurisdictions. Investment transfers represent 80 percent of total transfers37 (cf. Table 4). In addition to investment grants (DIN) are technical support grants (10 percent) intended to strengthen the subnational jurisdictions’ technical capacities. Finally, the remaining 10 percent of grants are divided between three subsidies, a grant to support the operations of subnational jurisdictions, an inter-jurisdictional transfer, and a grant to guarantee subnational jurisdictions’ loans. While investment subsidies help finance essential investments to improve access to basic social services, they could exacerbate the already marked disparities in Mali. Not all local communities have the resources (financial, technical, human) to invest, or even submit investment projects for financing by the FNACT. Support for the most vulnerable communities is essential, then, and should be one of the ANICT’s key missions.

Table 4.

ANICT investment and operating subsidies in Mali, 2009–2011 (CFAF thousands)38

article image
Source: Données de l’ANICT.

83. The Malian system of transfers is less than predictable owing, notably, to its dependence vis-à-vis external partners. There are two causes for the unpredictability of transfers dispersed by ANICT: first, the transfers from the central government are the result of discretionary decisions rather than a primary grant being pegged to the distribution of one or more national taxes. Secondly, 90 percent of FNACT disbursements come from external partners and are therefore subject to those partners’ respective criteria for allocating aid (cf. Table 5). Assistance from Luxembourg, for example, delivered through its Program for improved access to drinking water, basic sanitation, and urban development (PEA), represented 17 percent of financing sources in 2011. Also, the 2012 crisis in Mali prompted a large number of partners to temporarily suspend contributions to the FNACT.

Table 5.

Investment grants (DIN) in 201140

article image
Source: Données de l’ANICT.

84. The risk of instability of transfers relative to LOR exists in most countries of the region. Transfers are particularly unstable in Togo and Niger (cf. Figure 9). The amount of resources transferred by the central government is also unstable, varying in particular with the approach of local and central elections. According to Worthington and Dollery (1998), transfers increase during local elections because they are more politically useful at that time, but decrease during national elections because the yield in terms of political capital is smaller relative to direct central spending. In Senegal, the magnitude of the variations is smaller, and affects transfers as much as LOR. The system of transfers in Senegal, based in part on the transfer of 2 percent of VAT receipts, is undoubtedly a factor in this relative stability. In effect, the use of a peg for the primary transfer or a system of shared taxation has allowed certain countries to improve the predictability, stability, and transparency of the transfers by the central government. In Ghana, for example, transfers cannot be less than 5 percent of tax revenue, while local jurisdictions in Algeria and Morocco receive 20 percent and 30 percent of VAT proceeds, respectively.

Figure 9.
Figure 9.

Instability of transfers relative to local own revenue

Citation: IMF Staff Country Reports 2015, 291; 10.5089/9781513580296.002.A001

Different types of transfers in the region

In Benin, the Communal Development Support Fund established in 2008 is administered by an interagency commission (National Commission on Local Finances). It manages unallocated grants (operating and investment) and investment grants. The unconditional investment grants are divided into three components: (i) a structural grant, representing 35 percent of unallocated investment grants, is a fixed amount per commune regardless of population size; (ii) the equalization transfer, representing 50 percent of unallocated investment grants, consists of share (28 percent) that increases with the population and another share (40 percent) dependent on the commune’s index of non-monetary poverty; (iii) a third share (32 percent) that increases with the size of the commune; and (iv) a performance appropriation (15 percent) determined according to the jurisdictions’ fiscal effort, the “assessment of the evolution of own revenue and the quality of expenditure.” However, the evaluation criteria are opaque.

In Côte d’Ivoire, an overall operating grant is distributed partly in equal proportions to each jurisdiction (considering population size) and partly based on disparities in the jurisdictions’ resources. In addition, a general decentralization grant is distributed to offset expenses associated with the powers and responsibilities devolved from the central government. There are also modest investment subsidies.

A similar system exists in Niger with the general decentralization grant and an equalization fund.

In Togo, grants are composed of a general transfer and an investment grant.

In Senegal, decentralization transfers and facilities grants (investment subsidy) are allocated to the local jurisdictions. There is also a central government transfer representing 2 percent of VAT receipts. The modalities of distribution are determined each year by the National Council on Local Development (CNDCL) and decided by the Ministry of Decentralization and the Ministry of Economy and Finance, with the stated objection of reducing inequality and encouraging sound local management. However, the relatively fluid criteria leave room for discretion.

85. The allocation formula for ANICT investment transfers is based on relatively transparent equity and efficiency criteria. The criteria for allocation of DIN include the jurisdiction’s wealth, the number of inhabitants, the distance from major supply centers, and a communal poverty index determined by the Observatory on Sustainable Human Development. There are also criteria relating to fiscal effort calculated on the basis of TDRL collection rates and sound local governance (number of regular meetings held, percentage of minutes forwarded to the supervisory authority, date of most recent administrative accounts forwarded). Maps 1, 2, and 3 present the allocation of DIN drawing rights in 2011 by region, cercle, and commune.41

Map 2.
Map 2.

Allocation of DIN drawing rights by region in 201142

Citation: IMF Staff Country Reports 2015, 291; 10.5089/9781513580296.002.A001

Map 3.
Map 3.

Allocation of DIN drawing rights by cercle in 2011

Citation: IMF Staff Country Reports 2015, 291; 10.5089/9781513580296.002.A001

86. The maps of total amounts (Maps 2-4) do not point to a clear equalization objective. The map by region (Map 2) does not evidence an allocation to the poorest regions. The regions in the North, the poorest in terms of own resources, are those receiving the lowest DIN. The allocation by cercle (Map 3) seems to play more of an equalization role. The Goundam cercle in the Timbuktu region and the Gao and Kidal cercles received the largest investment grants. Yet the cercles of the Kayes’ region also receive some of the most generous grants, although their LOR are among the highest.

Map 4.
Map 4.

Allocation of DIN drawing rights by commune in 2011

Citation: IMF Staff Country Reports 2015, 291; 10.5089/9781513580296.002.A001

87. The per capita amounts of the allocations illustrate the equalization objective more clearly, but could be improved (cf. Table 6). The Kidal region raises 0.69 percent of total own resources (excluding Bamako) and receives 74 percent of per capita investment transfers from the ANICT. In contrast, Kayes, which raises over 30 percent of LOR, receives less than 3 percent of per capita transfers. Timbuktu does not appear to have been adequately considered in terms of equalization; with 4 percent of national LOR, it receives less than 6 percent of total per capita subsidies. It is likely that the poverty criteria are not given adequate consideration or are offset by the population criterion. Jurisdictions with low population density, such as the Timbuktu region, seem to be at a disadvantage. The introduction of a fixed proportion per jurisdiction in the allocation formula (as applied in Côte d’Ivoire or Benin), or consideration of the area or population density would address the intention of considering problems specific to local jurisdictions with low population densities. Including a fixed proportion in the allocation of transfers would also provide the advantage of a simple, specific criterion to limit discretion in the allocation of transfers and make the resource flows more understandable and predictable. Care should be taken, however, to avoid overly complex allocation formulas.

Table 6.

DIN allocated to regions, cercles, and communes in 201143

article image
Source: Calculs faits par les auteurs à partir des données de l’ANICT et du Trésor.

88. The allocation of transfers in WAEMU countries often reveals marked disparities and does not systematically further and equalization objective. Prior to 2008, it was not uncommon for local jurisdictions in a number of WAEMU countries to receive no transfers at all.44 Transfers are generally concentrated in several communes, and the size of jurisdiction alone cannot explain the disparities in transfers received. Per capita transfers to Pèrrèrè, Ranérou, and Djado are respectively four, seven, and 38 times the national average in Benin, Senegal, and Niger. In Benin, half the population receives close to 80 percent of total transfers. Similar findings were observed in all the countries studied. A simple comparative analysis of data fails to evidence a clear equalization objective in the allocation of transfers (cf. Table 7). While overall operating transfers seem better focused on this objective, individual situations vary. In Benin, transfers – in particular operating transfers – achieve a certain amount of equalization because total and per capita LOR are negatively correlated with transfers. It is likely, then, that the allocation formulas are applied and help reduce inequalities. In Mali, a different result was observed prior to the creation of the ANICT, since positive correlations indicated that transfers and inequalities evolved in the same direction.

Table 7.

The equalization objective in the allocation of transfers

article image
Source: Calculs des auteurs à partir des données du PDM


Create a transfer mechanism that emphasizes simplicity and therefore the predictability and stability of resources allocated. A dedicated technical assistance mission could help simulate the redistributive effects of different allocation mechanisms.

Bibliographic References

Books and articles

  • Akin, J., P. Hutchinson and K. Strumpf (2005). “Decentralisation and government provision of public goods. The public health sector in Uganda,” Journal of Development Studies, 41(8), 14171443.

    • Search Google Scholar
    • Export Citation
  • Attila J., G. Chambas and Combes J-L (2009). “Corruption and mobilisation des recettes publiques. une analyse économétrique,” Recherches économiques de Louvain, Vol. 75, issue 2, pages 229268.

    • Search Google Scholar
    • Export Citation
  • Bahl, R., Martinez-Vazquez J. and J. Yougman (eds), 2008, Making the Property Tax Work, Lincoln Institute of Land Policy, Cambridge, Massachusetts.

    • Search Google Scholar
    • Export Citation
  • Banful, A. B. (2010). “Do formula-based intergovernmental transfer mechanisms eliminate politically motivated targeting? Evidence from Ghana,” Journal of Development Economics, 94(1).

    • Search Google Scholar
    • Export Citation
  • Caldeira E. (2012). “Does the System of Allocation of Intergovernmental Transfers in Senegal eliminate Politically Motivated Targeting?,” Journal of African Economies, 21(2). 167191.

    • Search Google Scholar
    • Export Citation
  • Caldeira E. and G. Rota-Graziosi (2014a). “The Crowding-in Effect of Simple Unconditional Central Grants on Local Own-Source Revenue. The Case of Benin,” Journal of African Economies, 23 (3). 361387.

    • Search Google Scholar
    • Export Citation
  • Caldeira E. and G. Rota-Graziosi (2014b). “Decentralization dans les pays en développement. une revue de la littérature,” Revue d’économie du développement, 2014/4 Vol. 22, p. 537. DOI. 10.3917/edd.284.0005.

    • Search Google Scholar
    • Export Citation
  • Caldeira E., G. Rota-Graziosi and M. Foucault (2012). “Does decentralization facilitate access to poverty-related services? Evidence from Benin,” NBER Working Paper No. 18118.

    • Search Google Scholar
    • Export Citation
  • Caldeira E., G. Rota-Graziosi and M. Foucault (2014). “Decentralization in Africa and the nature of local governments’ competition. Evidence from Benin,” International Tax and Public Finance, December 2014.

    • Search Google Scholar
    • Export Citation
  • Chambas, G. (2010).). Mobiliser des local resources en Afrique subsaharienne., Economica, Paris.

  • Kornai, J. (1979). “Resource-constrained versus demand-constrained systems,” Econometrica, 47 (4), 80119.

  • Kornai, J., E. Maskin, and G. Roland (2003). “Understanding the soft budget constraint,” Journal of Economic Literature, 41(4), 10951136.

    • Search Google Scholar
    • Export Citation
  • Norregaard, J., 2013, “Taxing immovable property. Revenue potential and implementation challenges,” IMF Working Paper WP/13/129.

  • Prud’homme, R. (1995). “The dangers of decentralization,” World Bank Research Observer, 10(2), 20120.

  • Qian, Y., and G. Roland (1998). “Federalism and the soft budget constraint,” American Economic Review, 88 (5), 114362.

  • Rodden, J. (2002). “The dilemma of fiscal federalism. Grants and fiscal performance around the world,” American Journal of Political Science, 46 (3), 670687.

    • Search Google Scholar
    • Export Citation
  • Weingast, B. R. (1995). “The economic role of political institutions. Market-preserving federalism and economic development,” Journal of Law, Economics and Organization, 11(1), 131.,

    • Search Google Scholar
    • Export Citation
  • Wibbels, E. (2000). “Federalism and the politics of macroeconomic policy and performance,” American Journal of Political Science, 44 (4), 687702.

    • Search Google Scholar
    • Export Citation
  • Wildasin, D. E. (1988). “Nash equilibria in models of fiscal competition,” Journal of Public Economics, 35 (2), 229240.

Technical assistance reports

  • Benoit Taiclet, Marie-Laure Berbach and Moussé Sow, 2015, Mali. Réussir decentralization financière, IMF, January.

  • Grégoire Rota-Graziosi, Anne-Marie Geourjon and Gilbert Menard, 2014a, Mali. Politique fiscale (diagnostic), IMF, September.

  • Grégoire Rota-Graziosi, Bertrand Laporte, Maude Lavoie and Oana Luca, 2014b, Mali. Fiscalité minière (diagnostic), IMF, September.

Technical notes and other documents provided by the authorities

  • Groupe de Suivi Budgétaire, Mali, Rapport d’Etudes. État des lieux de la fiscalité locale et des transferts aux 6 communes du District, de Kati, de Ouéléssébougou, de Sanankoroba, de Sadiola et de Kayes, May 2015.

    • Search Google Scholar
    • Export Citation
  • Ministère de l’Administration Territoriale et des Collectivités Locales. Guide pratique de recouvrement des impôts, taxes et des redevances des collectivités territoriales.

    • Search Google Scholar
    • Export Citation
  • Ministère de l’Administration Territoriale et des Collectivités Locales. Mobilisation des resources propres et redevabilité des collectivités territoriales. Approche et stratégie d’intervention du PADK II, December 2013.

    • Search Google Scholar
    • Export Citation
  • Ministère de l’Administration Territoriale, de decentralization et de l’Aménagement du Territoire. Note technique sur la fiscalité locale au Mali, May 2012.

    • Search Google Scholar
    • Export Citation
  • Sy, Awa Diallo, 2015. Note technique, Ministère des Domaines de l’État et des Affaires Foncières, Direction Nationale des Domaines et du Cadastre, Juin.

    • Search Google Scholar
    • Export Citation
  • Djenontin, Georgette, 2002. Analyse comparative des taxes municipales dans les pays de l’UEMOA, Partenariat pour le Développement Municipal.

    • Search Google Scholar
    • Export Citation
  • Programme de Développement Institutionnel. Etude sur la révision du système de fiscalité locale, August 2010.

  • Djibril Semega, 2014, Étude sur la mobilisation des resources fiscales des collectivités territoriales, Direction Générale des Collectivités territoriales

    • Search Google Scholar
    • Export Citation


Annex 1. The promotion of the real property tax in Côte d’Ivoire and the decline of tax revenue.

Ivoirien legislation provides for a tax on real property assets (“real property tax”) applicable to all buildings regardless of whether they generate income. Furthermore, rental income is subject to tax on real property income. The ordinary rate of the real property tax, which is 11 percent of the rental value, is reduced to 4 percent for a principal residence or vacant properties. The rate is increased to 15 percent for property owned by legal entities and allocated to businesses. Unimproved urban land and agricultural land used for the production of rubber trees, oil palms, coffee, cocoa, cashews, cotton, bananas, mangoes, sugar cane, lemon, or papaya and exploited by legal entities or agro-industrial businesses [sic]. The ordinary rate is 1.5 percent of the market value of unimproved land; for agricultural land, a tariff per hectare is applied (e.g. CFAF 7,500 for rubber trees, CFAF 5,000 for cocoa).

Beginning in 1992, the Ivoirien authorities focused on increasing real property tax receipts in order to increase tax revenue during the difficult period prior to the devaluation of the CFA franc. Until then, the tax had been paid primarily by businesses that were taxed on actual income. The authorities used sophisticated land registry programs in urban areas and modern information processing resources. Despite the resources deployed, the results were disappointing because of significant resistance and the difficulty in collecting a broad-based real property tax.

Beginning in 2005, the difficulties in raising tax revenue during the crisis years prompted the Ivoirien authorities to focus their priorities on the real property tax. It became such a priority that in 2012, the DGI strategic action plan distinguished two categories of taxes, “real property taxation and other taxes.” To raise real property tax revenue, the Ivoirien DGI embarked on a strategy of systematic coverage of the entire Ivoirien territory. To this end, the DGI establish numerous tax offices and dedicated tax base units and revenue units to assess and collect the real property tax. This approach led to management by type of tax – the very opposite of management by taxpayer, which is the standard adopted by modern tax administrations.

Despite its efforts, the DGI achieved only half of its objective property of collecting CFAF 100 billion in property tax revenue in 2012, raising only CFAF 49 billion (0.36 percent of GDP). This figure may be compared to the CFAF 556.8 billion collected through the VAT. Due to dispersion of resources, gross VAT revenue remained stable for several years at roughly 4 percent of GDP despite a remarkable economic recovery since 2011.45 By giving priority to the real property tax, the DGI hampered its own ability to raise major tax revenue representing 95 percent of total tax revenue.

Rather than establishing new tax units throughout the country, the DGI should have focused its efforts on managing major tax payers. In regard to real property taxes, it would have been far more effective to focus reasonable resources on collected real property taxes owed by legal entities (estimated at CFAF 29 billion for the real property tax by a 2013 study by the Office of the Inspector General of Tax Services) and by the owners of high-value properties.

A notable improvement in VAT revenue is expected to result from the reorganization of the tax administration, effective since 2014, with the creation of two Directorates of Medium-Size Business Taxpayers to complement the Directorate of Large Business Taxpayers established previously.


The original version of this report was in French. The present English translation is for consultation purpose only.


[Tr.: The administrative divisions of Mali are the District of Bamako and associated communes; and eight regions, divided into cercles, which are themselves divided into communes. Communes may also include one or more associated fractions.]


[Tr.: “Patente” or “contribution de la patente.” A tax assessed on the exercise of a profession or trade other than by an employee.]


The bases for calculating the TDRL – particularly in rural areas – and the BLT could be maintained temporarily to eliminate any risk of loss of revenue during the transition period. Reform of the TDRL in urban areas would include the definition of a new real property tax.


Fees, such as royalties or license or user fees, are payments by the users of public services or works. A fee is paid in consideration of the benefits of such public services or the use of public works. Unlike a tax, a fee is paid only by a user, and the amount of the fee is proportional to the service rendered.


WAEMU Directive 04/99 does not specify the place of local own revenue in the second-order convergence criterion relating to minimum tax receipts equivalent to 17 percent of GDP. In the absence of indications to the contrary, the subnational administrations are a component of the central government and therefore local tax revenue should be included in total tax revenue to determine whether the criterion of 17 percent of GDP is met. This interpretation is corroborated by WAEMU Directive 10/2009 concerning the government financial operations table (TOFE), which includes local administrations as a component of the public administration.


See Caldeira and Rota-Graziosi (2014b) for a survey of academic literature on the effects of decentralization in developing countries.


This report does not distinguish between resources allocated to different levels of subnational administrations (communes and cercles here, and the regions below). The distribution of revenue among the three levels and between administrations of the same level is addressed in the final chapter of this report. It is determined largely by the definition of the transfer equalization mechanism, which could be addressed in the context of a dedicated technical assistance mission.


Latin America is composed of mostly unitary governments, with the exception of four countries: Argentina, Brazil, Mexico, and Venezuela have federal systems of government.


While an impôt is an unrequited payment, a taxe is used to fund the government’s provision of a public service or public work. In contrast with a fee, however, the consideration received by the taxpayer is secondary, since the taxpayer must pay the taxe regardless of whether the individual has made use of the service, and there is no proportionality between the amount of the taxe and the service rendered.


In Mali: Law No. 96-058 of October 16, 1996 determining the tax resources of the District of Bamako and its constituent communes; Law No. 2011-36th of July 15, 2011 determining the tax resources of the communes, cercles, and regions of Mali; and Law No. 2011-034 of July 15, 2011 instituting the real property tax.


Certain markets, notably in the major capitals, may be government owned and therefore entail the sharing of revenue from the market.


West African communes rarely own real or movable property. While the proceeds of sales of communal property are small, they contribute to investment revenue, together with borrowing.


Authors’ calculations based on the Programme de Développement Municipal (Municipal Development Programme) data prior to 2008 and Treasury data thereafter.


In 2013, since tax revenue accounted for 80 percent of LORB and 42 percent of tax revenue was provided by the DGI, only 66 percent of LORB can be considered LORN. World Bank, Fiscal Decentralization Indicators.


World Bank, Fiscal Decentralization Indicators.


We consider average data for the period 1995-2011 (subject to availability).


This risk was previously identified in Taiclet et al. (2015).


The theoretical discussions of soft budget constraints (Kornai, 1979, Qian and Roland, 1998, and Kornai et al., 2003) were adopted by the literature on fiscal federalism (Wildasin, 1988) from the analysis of the behavior of public enterprises in planned economies.


Few empirical studies have evaluated the macroeconomic effects of decentralization. Among them, Wibbels (2000) demonstrates a negative effect of fiscal federalism on macroeconomic performance, resulting in higher levels of inflation and fiscal imbalances. Rodden (2002) shows that persistent deficits occur when local governments are simultaneously dependent on transfers and permitted to borrow.


Law 96-058 of October 16, 1996 determines the tax resources of the District of Bamako and associated communes. Law 2011-03/6 of July 15, 2011 establishes the tax resources of Mali’s communes, cercles, and regions. Law 2011-034 of July 15, 2011 institutes the real property tax.


Receipts generated by licences (an excise tax applicable to alcoholic beverage sales), which were already quite low in 2011 (0.41 percent) have been negligible since 2012 (0.03 percent in 2012, 0.02 percent in 2013, 0.07 percent in 2014).


It should be noted that real property tax revenue collected by the DGI (Table 3) is not explicitly presented in the Treasury accounts (Table 2).


The fixed fee is determined by area: I, the District of Bamako; Zone II, regional seats of government; and Zone III, other locations.


For example, security companies, private educational institutions, and surveyors.


The TTR based on the number and characteristics of vehicles was replaced by the 3 percent turnover tax assessed under [CGI provisions applicable to all businesses].


According to the GIZ, which is conducting technical assistance in regard to the mining cadaster, there were 562 active permits at the time of the mission’s visit (cf.


The quantity of gold produced by artisanal miners in 2014 is estimated at 5 tons or more.


The central government’s transfer of responsibility for those taxes to the subnational jurisdictions could be explained in part by the low yields of those taxes.


A comparison of local tax revenue in WAEMU countries comes to the same conclusion: local taxes other than the BLT and licence and the real property tax and TDRL in Mali [sic], are numerous and generate only a negligible share of tax revenue.


The recent trend is positive, however, as property tax receipts have doubled on average each year since the 2011 introduction of the real property tax: CFAF 53 million in 2012, CFAF 134 million in 2013, and CFAF 195 million in 2014.


The DGI is in the process of analyzing a dwelling or residence tax for cities or so-called urban TDRL, which would be fully incorporated in the proposed reform.


The patente in France was instituted by the Decree-Law of March 2 and 17, 1791. At that time, it already included fixed and proportional components, similar to the tax applied in Mali. The fixed component was a rate based on the nature of the activity exercised and the population of the commune. The second component was determined according to the rental value of the taxed establishments and their equipment. The tax was widely criticized and was replaced in 1975 by the professional tax (taxe professionnelle), which was also substantially modified over time and recently replaced by the local economic contribution (contribution économique territoriale).


Users’ unwillingness to pay fees is basically due to deficiencies in the services delivered (e.g., disorganization or even lack of household garbage collection, unhealthy or unsafe conditions at markets, quality of conditions of slaughterhouses, cluttering of sidewalks by haphazard street vendor installations, inadequate maintenance of roads and drainage works).


Greater accountability is [considered] more likely [when] the effectiveness of public policy determines the reelection of the incumbent administration.


An externality is an economic situation in which one agent’s act of consumption or production positively or negatively affects the situation of another agent not involved in the action who does not bear the full cost or receive full compensation for the benefit or damage caused.


Aking, Hutchinson, and Strumpf (2005) citing evidence of the problem of spillover effect between jurisdictions following the decentralization of health services in Uganda, resulting in the reduced supply of a public good.


See Réseau Rhône-Alpes d’appui à la coopération international, “Livre blanc de la décentralisation financière dans l’espace UEMOA” [White paper on financial decentralization in the WAEMU area] (2014).


Investment financing is provided in the form of drawing rights. Drawing rights are used exclusively for cofinancing, within the limits of appropriated funds, of 80 percent of subnational jurisdictions’ public facilities projects. The jurisdictions contribute 20 percent to each investment, of which 10 percent may be provided in kind by the beneficiary community.


Source: ANICT data.


ABS-PARADDER: Sector Budgetary Support, Program to Support Administrative Reform, Decentralization, and Regional Economic Development. PIDRN: Investment and Rural Development Program for the Northern Regions of Mali. LUX-DEV-MLI 16 PASAB: Luxembourg Program to Support Basic Healthcare. LUX-ODEV-MLI 17 PEA: Access to Drinking Water and Sanitation Program.


ABS-PARADDER: Sector Budgetary Support, Program to Support Administrative Reform, Decentralization, and Regional Economic Development. PIDRN: Investment and Rural Development Program for the Northern Regions of Mali. LUX-DEV-MLI 16 PASAB: Luxembourg Program to Support Basic Healthcare. LUX-ODEV-MLI 17 PEA: Access to Drinking Water and Sanitation Program.


For transfers to the communes, we present the sum of transfers received by communes, aggregated by cercle.


Maps prepared by authors based on ANICT data.


Source: authors’ calculations based on ANICT and Treasury data.


One commune in Senegal, four in Benin, and 12 in Togo received no subsidies in 2006. The absence of investment subsidies is also common, as occurred in 40 percent of Beninese communes in 2006.


In 2014, VAT revenue in Mali represented 5.6 percent of GDP.

Mali: Technical Assistance Report - Local Taxation and Decentralization
Author: International Monetary Fund. Fiscal Affairs Dept.