3.1 Relevance of revenue assignments and own revenue sources

Although revenue sources are often less decentralized than expenditure responsibilities, tax revenues are an important source of income for subnational governments, accounting for one-third of total subnational government revenue or roughly 3.3 percent of GDP on average (OECD/UCLG 2019: 71, 77).[12] Other (non-tax) own revenue sources such as user charges, fees, and property income, account for another 11 percent of subnational revenue or approximately an additional one percent of GDP. Naturally, the importance of subnational own source revenues, and the breakdown between the different types of own source revenues, vary considerably from one country to another.

3.2 An overview of devolved (local government) revenue assignments and administration

The economics of local taxation under fiscal federalism. Unlike central government taxes (which are generally defined as compulsory payments to the central government for which there is no quid pro quo), local government taxes in a well-designed intergovernmental fiscal system are more appropriately seen as quasi-user fees for locally-provided services. Indeed, in order to maximize social welfare and improve the allocative efficiency of resources in a decentralized public sector, the goal of local taxation is not to maximize the volume of local revenue collections, but rather, to ensure that local taxpayers in different local jurisdictions only pay local taxes commensurate to the level of locally-provided services that they demand from and get supplied by their local government.[13]

In line with the concept that “finance should follow function,” local taxes and user fees should be considered appropriate funding sources to pay for exclusive local government functions—where the benefits of local government services largely or wholly are received by residents of the local government jurisdiction itself. As noted in Section 3.3 below, to the extent that concurrent functions partially or largely benefit residents outside the local government jurisdiction, it would be conceptually more appropriate to fund such concurrent government functions in part or in whole through intergovernmental fiscal transfers.

Assignment of own revenue sources. Public finance theory prescribes a number of rather stringent conditions to determine which taxes and revenue sources should be considered good candidates for assignment to the local or regional level (Bird 2000). In fact, in line with the subsidiarity principle, the only taxes and revenue sources that could be suitably collected by subnational governments are revenue sources that (a) can be administered efficiently at the local or regional level; (b) are imposed solely or mainly on local residents;[14] and (c) do not raise problems of harmonization or competition between subnational governments or between subnational and national governments.[15]

The only major revenue source usually seen as passing these stringent tests for assignment to the local level is the property tax; the second-largest category of local revenues in many countries tends to be user fees and charges. In fact, for all other high-yielding tax sources—including personal income taxes, corporate income taxes, value-added taxes or sales taxes, and trade taxes—it could reasonably be argued that the central government is the lowest level of government able to collect those revenue sources without causing inefficiency. As a result, it is no surprise that the vast majority of revenues in most countries is collected by the central government.

Tax autonomy and the assignment of shared revenue sources. Because the practical scope for autonomous subnational taxation—in a way that ensures efficiency—is limited, some countries assign local governments the right to collect different revenue sources, while limiting the control of subnational governments over one or more aspects of these taxes. This results in a spectrum ranging from own source revenues fully under the control of local decision-makers to tax sharing arrangements over which local governments have no control (Table 3.1).

Table 3.1 A taxonomy of tax autonomy (OECD)

ClassNature and extent of tax autonomy – relative to higher-level government (HLG)
a.1 The recipient subcentral government (SCG) sets the tax rate and any tax reliefs without needing to consult a HLG.
a.2The recipient SCG sets the rate and any reliefs after consulting a HLG.
b.1  The recipient SCG sets the tax rate, and a HLG does not set upper or lower limits on the rate chosen.
b.2The recipient SCG sets the tax rate, and a HLG does sets upper and/or lower limits on the rate chosen.
c.1The recipient SCG sets tax reliefs – but it sets tax allowances only.
c.2The recipient SCG sets tax reliefs – but it sets tax credits only.
c.3The recipient SCG sets tax reliefs – and it sets both tax allowances and tax credits.
d.1Tax sharing arrangement in which the SCGs determine the revenue split.
d.2Tax sharing arrangement in which the revenue split can be changed only with the consent of SCGs.
d.3Tax sharing arrangement in which the revenue split is determined in HLG legislation (less frequently than once a year).
d.4Tax -sharing arrangement in which the revenue split is determined annually by a HLG.
e.Other cases in which the central government sets the rate and base of the SCG tax.
f.None of the above categories of a, b, c, d, or e applies.

For instance, central legislation might provide local governments with the power to collect a certain tax – a corporate income tax, for example – while defining the base of this tax uniformly across the entire national territory in order to limit the administrative burden of local taxation on taxpayers. Similarly, central legislation may limit the tax rates that local governments may impose on local taxpayers for different taxes – for example, by setting lower and upper bounds – in order to prevent territorial or vertical tax competition. Alternatively, central authorities may simply decide to share the revenue collected from certain revenue sources with subnational governments. For example, this may be done on a derivation basis (based on where the revenue is collected) without giving subnational governments any control over the tax base, the tax rate, or the sharing rate.[16]

In addition to property taxation, another area of focus for subnational revenue mobilization efforts could be on user charges and fees. The ability of local governments to collect these types of revenues depends considerably on the assignment of functional powers; local institutions’ ability to deliver local services in way that provides value-for-money; and on the capacity and willingness of users to pay for these services.

3.3 An overview of non-devolved revenue assignments

Traditionally, the discussion of revenue decentralization and the assignment of revenue powers has focused almost exclusively on the local property tax and any other local tax and non-tax revenue funds that are part of the local government budget. Virtually no systematic attention has been paid to the assignment of revenue powers to non-devolved actors in the intergovernmental system. This includes any discussion or analysis of revenues collected by national parastatal entities, authorities and funds—revenues collected by entities that are funders or providers of delegated services. Also overlooked are revenues collected by local government-owned public companies, delegated service providers, and other “last mile” providers such as local water utilities, transit companies, or fee-collecting local health facilities. All these revenues are typically excluded from measures of revenue decentralization, as traditional measures of revenue decentralization focus exclusively on national government revenue collections and local government revenue collections. Any revenues collected by off-budget entities at both the central government and local government levels are often simply overlooked.[17]

While the reliance on non-devolved revenue sources is likely to vary significantly from country to country and from sector to sector, these revenues are likely to play a much more significant role than commonly recognized. For instance, in the provision of public health services, how much do local health facilities collect in terms of user fees or private or social health insurance payments in a way that is not captured by local government accounts? In turn, how much revenue do national or local health insurance schemes collect from the public? Similarly, to the extent that schools collect school fees from parents and/or to the extent that school committees or parent-teacher committees, as quasi-public entities, contribute to the provision of primary education, how significant is this funding?[18]

In the provision of water and sanitation services, what is the total revenue collected each year and subsequently spent for recurrent operation and capital purposes by off-budget urban water utilities? Similarly, in rural areas, what revenues are collected by water user committees which, in many countries, serve as the de facto provider of rural water services? Both of these questions should be answered fully to get a comprehensive picture of water and sanitation revenues. It is not unusual, however, for the accounting of water and sanitation revenue and spending to focus exclusively on capital investment spending, and to ignore the revenues and expenditures needed to operate and maintain water and sanitation infrastructure.

Likewise, to the extent that roads and other transportation infrastructure may be operated in an off-budget manner by a national road fund (often funded by a fuel levy) or by dedicated transportation authorities or public-private partnerships (PPPs), what are the fuel levies or road tolls that are collected by these authorities or entities that operate and/or maintain public sector roads or bridges?

3.4 Common obstacles in domestic revenue mobilization and subnational revenue administration: technical challenges

Local own source revenues are seen by many as a preferred source of funding for local government services. This is not only because there is a stronger conceptual link between the benefits and costs of locally-provided services,[19] but also because local taxpayers are expected to exert stronger oversight over the efficient spending of their own local tax contributions. Furthermore, revenue decentralization gives subnational governments a fiscal stake in the economic success of their jurisdictions. As a result of these factors, the failure to decentralize revenue powers while decentralizing expenditure responsibilities is generally assumed to result in greater local fiscal indiscipline and risk taking.

But, the evidence on this point is mixed. Given the fact that the collection of most major revenue sources—with the exception of property taxes—is generally assigned to the central government in line with the subsidiarity principle in revenue administration, virtually every country in the world faces a significant primary vertical fiscal imbalance. In many countries, the assignment of shared revenue sources on a derivation basis, or the introduction of local surtaxes or piggy-back taxes is often able to reduce the vertical fiscal gap in a way that provides resources to subnational governments without the potential inefficiencies associated with full revenue decentralization (Hunter 1977).

Nonetheless, lackluster collection of local taxes and other own source revenues in many local jurisdictions is common, particularly in developing and transition countries. Analyses of local revenue performance frequently attribute the lack of local revenue effort to an amorphous “weak local revenue administration” which, in turn, is often attributed to a “lack of local political will.” Instead, weak local revenue performance is often caused by a combination of factors, including the fact that local governments are assigned unpopular taxes that are relatively costly to collect, and have weak enforcement powers and weak political incentives and/or the absence of hard budget constraints.[20]

As a result, most real-world interventions related to revenue assignment and local revenues are intended to ensure that subnational governments administer the limited revenue instruments assigned to them as efficiently as possible. Efforts to improve local property tax administration (particularly in urban areas), often play on outsized role in development partner interventions related to local government revenues (Kelly, White, and Anand 2020).

3.5 Political economy considerations: common obstacles in revenue assignments

Empowering intergovernmental (fiscal) systems: revenue assignments. Public sector revenues tend to be much less decentralized when compared to public sector expenditures. As noted in Section 2, when we apply the subsidiarity principle to the function of public taxation and revenue administration, most revenues are efficiently collected at the national level. An additional reason for this pattern is that political economy forces cause revenues to be highly centralized. Most Finance Ministers will be hesitant to give away high-yielding revenue instruments to subnational governments, and thereby reduce the ability of the national fiscus to ensure macro-fiscal stability.

Furthermore, it is common for central government politicians—ahead of their next election—to abolish local taxes that are unpopular with the electorate, allowing central politicians to cut taxes for voters without a negative impact on their own (central) budget. More often than not, these local revenue sources are reinstated after the election, when locally elected leaders appeal to the national party that local revenues are an important foundation for the financial survival of local governments.

Efficient, inclusive and responsive revenue assignment. In response to news that local governments are collecting only x percent of the revenue that they could be collecting (where x is a small number, sometimes even as small as 10 percent), it is not unusual for national-level politicians or policy researchers studying local revenue administration to condemn local government officials for lacking the political will to collect own source revenues.

Such criticism may or may not be warranted, and if nothing else, it does not necessarily point to a problem with local tax administration. It is useful to start by acknowledging the political economy argument that local revenue collections are not intended to be maximized, but rather, that local revenue collections are optimal where the marginal cost to local taxpayers of additional taxation equals the marginal benefits from additional public services. In an effectively decentralized system, if the chain of accountability is working, locally elected officials are the arbiters of the level of local taxation at which this optimum is achieved. The “lack of political will” may simply reflect a rational political response to a situation where it might be politically easier for a mayor to get additional resources as a special grant from central government compared to collecting from local constituents. Local leaders may also exhibit a lack of political will to collect own source revenues results if the efficiency or responsiveness of local government spending is relatively low. A low level of lack of political will is only a real concern if local politicians are setting effective tax rates – through a combination of formal tax rates and weak revenue administration and enforcement – that result in a level of local taxation that falls below what is considered optimal by local constituents.

A bigger concern may actually be when predatory local taxation, the opposite of inadequate revenue mobilization, occurs.[21] Another serious problem occurs when the local government administers local taxes and revenues in a patently inequitable manner for example, enforcing taxes on political opponents, but not on political supporters, or when pervasive inefficiency or corruption exists in local tax administration. It is not just local politicians who are to blame at the local level for weak local revenue administration. As long as local politicians and taxpayers are satisfied to remain at an equilibrium of low taxation and low service delivery performance, the tax administration apparatus does not face strong incentives to improve its collection performance. Perhaps unsurprisingly, then, most local revenue mobilization efforts focus on other local administration improvement efforts such as improving land administration and property valuation, while basic revenue collection activities, such as billing systems and enforcement and collection of arrears, are frequently overlooked.

Engaged civil society, citizens, and business community: revenue assignments. While the long term success of any public sector depends on its ability to generate revenues from which to fund public sector expenditure, it is equally important to consider the perspective of the (local) taxpayer in determining the assignment of revenue sources and the optimal level of taxation at different levels. In most countries, even under the best of circumstances, taxpayers are unlikely to pay their (local) taxes if payment can be avoided without negative consequences. Tax collection and enforcement issues aside, local taxpayers’ willingness to pay taxes in return for local public services is likely to be limited if the local government’s decision making is unresponsive, or if the local government’s capacity to efficiently deliver services is weak.

A final political economy consideration regarding local revenue collection is how the money gets spent. Wealthier taxpayers might be willing to pay local taxes if they perceive benefits from higher local taxes. However, the willingness of wealthy taxpayers to support pro-poor local services is often limited by the strength of local social contract. Thus, local revenue compliance may decline over time when local governments pursue redistributive policies beyond the level supported by those contributing most to the local treasury.

Box 3.1 Background and resources on revenue assignments and local revenue administration

Subnational Taxation in Developing Countries: A Review of the Literature. Richard Bird: World Bank, 2010.
Sub-central Tax Autonomy. Hansjörg Blöchliger and Maurice Nettley: OECD Fiscal Federalism Studies, 2015.
Municipal Finances: A Handbook for Local Governments. Catherine Farvacque-Vitkovic and Mihaly Kopanyi: World Bank, 2014.
Property Tax Diagnostic. Roy Kelly, Roland White, and Aanchal Anand: World Bank, 2020.

[12] According to the OECD definition used, tax revenue is not made up only of own-source taxes, but includes shared taxes as well. Even with this more expansive definition of subnational government tax revenues, subnational taxes account for only 14.9% of public tax revenues. As discussed further below, the main funding source for subnational governments (on average) is formed by intergovernmental fiscal transfers.

[13] In this sense, decentralized provision of locally-provided goods mimics market-provision of private goods, where consumers opt to consume a private good up to the point where the marginal benefit from the good equals the marginal cost. Basic economic analysis (for instance, in the context of a representative agent or median voter model) suggests that in addition to the local governments’ responsiveness to constituent preferences, other key determinants of the optimal level of local taxation include the relative price (i.e., efficiency or inefficiency) of local service provision and the presence or absence of general-purpose grants.

[14] An efficient assignment of revenue sources should prevent the possibility of “tax exporting”, by which a local or regional government is able to impose a tax burden on residents outside its jurisdiction. For this purpose, it is important to recognize that the burden of a tax may be borne by someone other than the person who pays the tax. For instance, while import duties are paid by the importer, the actual burden of the tax is typically borne by the final consumer (because the cost of the import duty raises the final sales price). As such, assigning the power to levy import duties to local governments (or the practice of charging an octroi on the trans-shipment of goods through a local jurisdiction) would effectively allow local government to tax the residents of other local governments without providing commensurate services to them.

[15] Tax competition between different subnational jurisdictions as well as duplicative taxation by different levels (resulting in cumulative high marginal tax rates) would have the potential for economic distortion and inefficiency.

[16] As noted in Section 3.3 below, economists consider that such shared revenues are in fact intergovernmental fiscal transfers. Nonetheless, it is not unusual for the domestic Chart of Accounts to register such shared revenues as own source revenues rather than as intergovernmental revenues in order to give the appearance of tax autonomy.

[17] Compared to other sectors, the health sector offers a positive example, as the World Health Organization’s accounting of Total Health Expenditures seeks to incorporate different funding flows, including public sources (government spending); private (out of pocket) spending; social health insurance; and donor organization spending. Despite the extensive guidance in the sector, however, it is often still difficult to entangle how much is being collected and spent of health services, and, by whom, at the subnational level.

[18] Boex and Vaillancourt (2014) point to the case of education spending in Madagascar. Primary education is formally a central government responsibility provided in a deconcentrated fashion following a classic French model. In 2010‐2011, centrally hired primary school teachers (either as permanent civil servants or contractual employees) accounted for only 32% of all public school teachers; of the remaining 68% (called FRAM teachers), 48% were hired and paid in part by parental committees and in part by a subsidy paid directly to teachers by the central government and 20% were hired/paid by parent’s committees, often with in kind payment (rice).

[19] The link between local taxes and local expenditures and accountability at the local level is called Wicksellian connection. See Bird and Slack (2013).

[20] National revenue authorities don’t necessarily do any better job when asked to collect local revenues (Fjeldstad, Ali, and Katera 2019).

[21] The definition of predatory taxation is often in the eye of the beholder. However, most people would be concerned about the efficiency and equity of local revenue assignments if a major share of local revenues would benefit tax collectors, or if these local revenues are mainly used to pay for the sitting allowance of local officials.