The focus of this brief is on the policy responses that governments can and are taking through the channel of inter-governmental fiscal relations to tackle the Coronavirus (COVID-19) crisis. Subnational governments are playing a crucial role in the current crisis, and past experience suggests they will need sufficient support to ensure they can confront the risks. Coordinated fiscal action across different levels of government is essential to ensuring that responses to the crisis are effective across all regions within countries.

This brief was first published on 21 April 2020, following the special meeting of the OECD Network on Fiscal Relations across Levels of Government on the COVID-19 pandemic. More information is available at http://oe.cd/fiscalnetwork. This revised version incorporates additional comments and suggestions from delegates, information from a rapid survey on responses to the pandemic – that extends the analysis of health-related issues, revenue pressures and equalisation systems – as well as updates based on policy developments as of 31 July.

The world remains confronted with a global pandemic outbreak that is challenging policymakers at all levels of government. Few observers imagined the depth and breadth of the crisis that the world is facing today, with COVID-19 severely affecting health, freedom of movement, food security, employment, education, trade, and leisure – there is hardly any aspect of our social and economic system that has not been shackled or halted by the present crisis. In retrospect, the 2008-09 financial crisis now seems like a small-scale rehearsal of these disruptions to our socioeconomic system, with around one-quarter of total GDP directly impacted by the present crisis (OECD, 2020a).

Subnational governments (SNGs) were at the forefront of the 2008-09 financial crisis, since they were severely hit by it as well as important actors in overcoming it. Today, SNGs are again at the forefront, as they are important actors in the implementation of the public health measures to prevent the spread of the virus (through screening, population monitoring, crisis communications, coordinated care for vulnerable populations, enforcing confinement policies, ensuring public order, among others), and in some countries, in providing the health treatment and services to those infected. Moreover, in many countries, SNGs are also a key player in social protection, providing financial support to citizens who cannot work, and on business support.

Government reactions to the crisis pose coordination risks since all levels of governments need to act coherently and quickly to design and implement policies that, in some cases, were never implemented previously. Moreover, since in some cases different levels of government are tasked with similar responsibilities, coordination is crucial for each level of government to focus on the activities that they are better prepared for, reducing redundancy and improving consistency. Errors may cost both lives and considerable resources. For these reasons, in order to better tackle the COVID-19 crisis, some specific government responsibilities are being re-centralised or decentralised, and institutions are being either created or re-oriented to support vertical and horizontal cooperation.

The necessity of taking reactive measures in conjunction with a reduction in economic activity is affecting SNG fiscal situations, with SNG expenditures surging, while at the same time revenues are collapsing (similar to the situation in the 2008-09 crisis – see Box 2). The crisis is also asymmetric both in terms of its impact and in terms of capacity available within countries; in addition, although most countries have a fiscal equalisation system that aims to level-out SNGs’ capacity to provide public services, those systems are designed to offset structural revenue and cost differences and, thus, can only partially compensate for the extraordinary asymmetries of the current conjuncture. As a result, so as to cope with liquidity needs, most countries are supporting SNGs financially through additional inter-governmental grants, loans, guarantees, and the lifting of fiscal rules.

This note analyses SNGs vulnerabilities and measures in the context of the COVID-19 crisis through the lenses of the classification framework proposed by Forman, Dougherty and Blöchliger (2020) – “Synthesising Good Practices in Fiscal Federalism” (see Figure 1) – and based on the lessons of the 2008‑09 experience.

Figure 1. Linking COVID-19 to typical policy recommendations on fiscal federalism

More specifically, this note: i) shows that SNGs are a key player in addressing the present health and economic crisis, and highlights a few stylised facts on how they are intervening; ii) identifies the main fiscal vulnerabilities of SNGs and actions taken to improve their liquidity and fiscal capacity to react to the crisis; iii) highlights inter-governmental coordination challenges and actions that are employed to improve coordination; and finally, iv) looks at emerging policy challenges based on past experience to ensure that SNGs can effectively react to the health crisis and support economic activity.

 1. SNGs are key players in addressing the COVID-19 crisis

Subnational governments (SNGs) are at the front lines in addressing the COVID-19 pandemic, in terms of its containment, addressing the health challenges, and supporting populations in need.

 Initial responses to control the virus have relied in part on local policy actions

While many of the mitigation and confinement responses have been led at the central government level, many of the early actions in China, Germany, Italy, the UK and the US have been at the local or state/region levels. Initial actions to slow the spread of the COVID-19 virus and “flatten the curve” have relied heavily on targeted confinement actions that were in some cases directed at the national level (e.g., in Lombardy, Italy), with bottom-up responses in other cases. Notably, US states and municipalities have been the lead actors in California, New York and Texas, as well as many other regions. In several US multi-state regions, adjoining states even developed “pacts” to ensure coordination in the relaxing of social distancing controls, in order to try to avoid an un-synchronised adjacent relaxation of controls that would undermine efforts to slow the spread of the virus.

 SNGs are responsible for the provision of essential public goods such as health, public order and safety or social protection

Public services related to health, public order and social protection are potentially going to be more demanded in this stage of virus propagation and lockdowns. On average, SNGs are responsible for 56% of average country expenditure on public order and safety, 39% of expenditure on health, and 16% of expenditure on social protection (Figure 2).

Figure 2. SNGs are key players in the provision of health, public order and social protection
Share of expenditure as a % of general government

Note: Subnational expenditure by function in 2018 shown as a percentage of general government expenditure.

OECD weighted average (weighted by population size of each country) based on available data.

Source: OECD National Accounts COFOG database.

Health is directly affected by the outbreak, since people who get infected might require treatment in case of severe symptoms. Public order and safety are necessary to enforce the lockdown measures that most OECD countries have taken to suppress the spread of the virus. Social transfers are a natural answer to the income reduction and unemployment that commercial and retail business closures are triggering throughout the economy. Thus, many of the policies implemented to tackle the virus outbreak are the responsibility of SNGs.

Health decentralisation varies from zero to almost 100% among OECD countries (Figure 3, Panel A). SNGs are responsible for more than 40% of health expenditure in ten countries, and above 80% in Switzerland, Denmark, Italy, Sweden, Spain and Finland. It is worth noting that Spain and Italy – both of which were hit hard by the virus outbreak – are among the latter group of countries.

The share of public expenditure on public order and safety varies within a range between zero and 90% (Figure 3, Panel B). Federal countries tend to decentralise this function more than unitary ones: Switzerland, Germany, the United States and Belgium are federal countries and are among the countries where SNGs have the highest share on this expenditure function. As a result, SNGs from these countries may face higher financial pressures from this type of expenditure.

Expenditures on social protection are not as decentralised as those on health and public order and safety. Half of the countries analysed decentralise less than 10% of this type of expenditure (Figure 3, Panel C). This follows the common practice of keeping social security and protection more often at the central level.

Figure 3. Decentralisation of health, public order and safety, and social protection expenditure

Note: In Panel c, Denmark appears as an outlier, partly because social security is included with the SNG expenditure.

Source: OECD National Accounts COFOG database.

 SNGs are also a large employer and could therefore play a stabilising role

Public employment may work as a buffer for reducing the negative economic effect generated by the fall in economic activity. In most OECD countries, SNGs are large employers and employ more people than the central government (Figure 4, Panel A). In some countries (usually federal ones), such as Switzerland, Belgium, US and Germany, SNGs are responsible for up to 80% of government employment.

Considering the whole labour market, SNGs also have a prominent role. On average, 10% of OECD countries’ workforce is employed by SNGs (Figure 4, Panel B). This share increases to around 20% in Nordic countries. It thus may be important for SNGs to be able to maintain their personnel, to reduce the negative economic impact from an acute recession.

Figure 4. In many OECD countries, SNGs employ a large share of civil servants and are responsible for a large share of total employment

Source: Calculations based on OECD Government at Glance (2019).

 2. Coordination challenges and responses

Vertical coordination concerns the harmonization of the actions taken by different levels of government with the purpose of providing a consistent and efficient public service. The COVID-19 outbreak is an extraordinary situation that requires this essential inter-governmental framework to function extremely well. It is an important activity given that many key government services are at least partially decentralised – notably education, health, transport, and social services. With different levels of government tasked with similar responsibilities, coordination is crucial for each level of government to focus on the activities that they are better prepared to do, reducing redundancy and improving consistency. In normal circumstances, ensuring a high degree of vertical coordination is not trivial since it involves constitutional, legal and administrative aspects. In extraordinary circumstances, however, it becomes even harder – the necessity of prompt decisions and urgent measures may exacerbate vertical coordination problems and results in the worsening of the public service provided when it is needed the most.

In the specific case of COVID-19, multiple activities can be damaged in case jurisdictions and levels of government do not coordinate themselves. First, variation in SNG procedures may undercut the benefits of scale, particularly in the area of procurement, which is key to meet the demand of medical equipment and medicines necessary to treat patients, and protective equipment to reduce contagion. Second, inconsistent or fragmented regulatory regimes may hinder the quick adoption of new products and procedures discovered/designed elsewhere. Third, lack of monitoring and benchmarking may allow service standards to slip, particularly when data are absent or inaccessible data/monitoring, which is precisely the case for health care, in which data governance is lagging behind other sectors (OECD, 2020d). Fourth, inadequate human and physical capital within the subnational public service can impede policy implementation and, since health care capacity in terms of equipment, staff and space in many countries is insufficient to meet the additional demand from COVID-19 patients, coordination can increase the pool of resources on the hands of SNGs so they can be to better allocated when and where needed.

The typical recommendation on improving coordination from the framework shown in Figure 1 is especially relevant to improve policy alignment and coordination across levels of government, eliminate internal market barriers and reduce regulatory fragmentation, enhance monitoring including through oversight at the national level, benchmark performance nationally and internationally, improve the quality and accessibility of performance measurement data, implement joint procedures and centralise procurement across SNGs and build SNG professional capacity (Forman, Dougherty and Blöchlinger, 2020).

 Countries face coordination challenges involving different sectors and activities, ranging from public financial management to lockdown enforcement, health input allocation and ownership of policy measures

The coordination challenges faced by countries are summarized in Table 1. The following four challenges were mentioned by more than or half of the countries: financial compensation (56% of countries), lockdown measures, asymmetrical impacts, and overlapping activities between levels of government (the last three were mentioned by 50% of countries). Health inputs allocation, ownership and/or accountability of measures, and public-facing communication were all mentioned by eight countries (44%) whereas cross-border issues were mentioned by four countries (22%). Luxembourg mentioned that through continuous exchange between SNGs and the Ministry of the Interior they are able to coordinate different levels of government successfully, while Poland indicated that providing an adequate and flexible financing of activities has been a challenge.

Table 1. Summary of main coordination challenges (ordered by decentralisation of expenditure)















Expenditure decentralisation ≥ Median





Expenditure decentralisation < Median

























































































Financial compensation

Lockdown measures

Asymmetrical impacts/needs

Overlapping activities between levels of government

Ownership and/or accountability of measures

Health inputs allocation

Public-facing communication

Cross-border issues between SCGs


Source: Based on the COVID-19 Rapid Survey held the Network in June 2020. Only countries that answered this specific question are shown.

 Decentralised countries typically faced additional and different challenges

Countries that are more decentralised than the median1, face 4.3 challenges compared with 2.8 in less decentralised countries2. Notably, two challenges seem to be more common in decentralised countries: overlapping activities between levels of government, and asymmetrical impacts or needs. These two challenges were reported by seven and six countries of the more decentralised group against two and three of the more centralised group, respectively. The only challenge that is more common in the more centralised countries is financial compensation.

These data suggest that depending on the degree of decentralisation, different challenges are dominant. First, more decentralisation seems to be linked to more overlapping responsibilities challenges, which could be potentially explained by the fact SNGs have more autonomy to implement policies that differ from the central government’s, allowing for more flexibility but potentially increasing overlapping. Second, in more decentralised countries, asymmetric impacts are larger. A potential reason for this could be that in the surveyed countries, the decentralised ones are also larger in size, which allows for more variations in demographic, economic sector and culture. Third, with regards to financial compensation, centralised countries seem to be facing more difficulties in coordinating well across different levels of government. One potential explanation for this is that in more centralised countries, subnational governments are more dependent on central government revenue, which might imply that any shortage of resources may need to be compensated by the central authority.

Figure 5. Mechanisms used to address coordination challenges

Source: Based on the COVID-19 Rapid Survey held by the Network in June 2020. Only countries that answered this specific question are shown.

 All countries have used existing institutions to cope with coordination challenges and roughly half of the countries have created new institutions

In order to address these coordination challenges, existing institutions have been put to use in almost all countries surveyed (Australia is the sole exception) and some countries have also resorted to new institutions whose purpose is to address the COVID-19 outbreak (Figure 5).

Country examples of coordination management:

Australia mentioned the creation of a national cabinet, a coordination body comprised of representatives from both state-level and central government. The national cabinet functions as an inter-governmental forum with a direct connection to administrative bodies and considers health and economic matters on a bi-weekly basis. As the crisis has progressed, its scope has been widened to encompass other matters indirectly related to COVID-19 such as transport logistics, law and order, as well as mental health.

Austria put in place a coordination instrument gathering authorities from key ministries, state governments and outside bodies – the National Crisis and Disaster Management. This instrument was originally set up in 1986 and is used to coordinate responses to crisis between federal authorities and provinces.

Belgium extended the existing National Security Council to include the Minsters-President of the Communities and Regions and new advisory bodies with experts. The role of the Council is to address potential risks and design an exit strategy.

In Canada, a committee of federal/provincial/territorial Deputy Ministers of Health is meeting regularly to address challenges in responding to the pandemic and to share best practices. Decisions taken by the federal government are made by Cabinet and its committees. Similar processes occur in provincial and territorial governments. Regular teleconferences and virtual meetings between the Prime Minister of Canada and provincial/territorial premiers, and Ministers at the federal and provincial/territorial levels are held to coordinate actions. Regarding data sharing between governments the federal government uses data and modelling to guide Canada’s response to COVID-19, and to help inform public health and policy decisions to control the pandemic. Some provincial and territorial public health authorities also conduct their own modelling to determine the projected numbers of COVID-19 related cases and deaths to aid in their health system capacity planning. Greater coordination and sharing of information could improve modelling to guide policy responses to the pandemic.

Colombia created a Management Office to respond to the effects of the COVID-19.

Finland created a new institution to share data so that different actors have a comprehensive, consistent and updated picture of the situation.

In Iceland, the existing civil protection authorities’ Coordination Centre coordinates all action and provides instructions and information to the public, institutions and enterprises. Local Authorities by law operate regional civil protection committees that work closely with the Coordination centre, which is ensuring a smooth coordination process throughout the country.

Italy created a scientific-technical committee composed by experts and qualified representatives of local authorities and state administrations that has a supporting role in monitoring the state of the epidemic and providing guidelines on lockdowns and travel restrictions decisions.

The Korean government, in order to bolster government-wide responses to COVID-19 and to provide necessary assistance to SNGs, is involving multiple authorities and bodies such as the Prime Minister, the Minister of Interior and Safety, Ministry of Health and Welfare and bodies specialized in disease control and disasters management. In the SNG level, each local government establishes a Local Disaster and Safety Management Body whose purpose is to secure the provision of health inputs and that can, if necessary, require central government’s support.

Poland: Established the "COVID-19 Counteraction Fund" that, although is not a mechanism per se of fiscal coordination, is facilitating a coordinated response to the COVID-19 outbreak.

South Africa: Is putting to use much more frequently electronic versions of existing coordination forums and has invoked across all levels of government disaster management structures that although were constituted previously are usually inactive.

Spain is improving coordination through the use of two committees that have been meeting weekly during this pandemic: (1) the Conference of Presidents, composed by the Government President and the Presidents of the Autonomous Communities; and (2) the Inter-territorial Council, CISNS, composed by the Minister of Health and the Health Counsellors of the Autonomous Communities.

At the European Union level, the European Committee of the Regions (CoR) has defined an action plan to support and assist local and regional authorities on the forefront of the fight against the Coronavirus pandemic. On 24 March, the CoR released an action plan which includes the launch of an exchange platform to help local and regional leaders sharing their needs and solutions and to enhance mutual support between local communities across Europe.3

 3. Delineation of activities

Delineation challenges and activities are related to but are by no means equivalent as coordination challenges and activities. Good coordination can solve issues generated by poor delineation and vice versa but, in principle they are not the same. Precisely, the latter refers to the harmonisation of activities given each level of government responsibility and capacity, whereas the former refers to “given part of the equation” – the division of responsibility and the provision of the necessary capacity to perform those activities. Thus, measures to improve coordination involves, for instance, standardisation of procedures, monitoring of policies responses, implementation of a multi-level governance system, benchmarking, communication across jurisdiction and levels of government. On the other hand, common recommended measures to improve delineation are: ascribing activities to the level of government that can reduce spill‑overs and maximize economies of scale and scope while providing a service that is flexible enough to capture regional differences and people’s preferences, reducing overlapping responsibilities and providing the necessary financial capacity for each level of government to provide the public services of their responsibility.

In normal circumstances, complex systems like health care, education and social services are particularly susceptible to delineation issues. For instance, in health care, primary care may be allocated to one level of government whereas the responsibility for public hospitals can be retained by other level4 (see Box 1 for a more detailed explanation of activities delineation across levels of government in health care). Such arrangements may work perfectly well, but coordination is required. One good indication of the challenge that different levels of government face to better coordinate an activity is the degree of activity overlap. Figure 7, below, depicts an estimate5 of the proportion of shared responsibilities in health and long-term care among levels of government by country.

Box 1. Decision-making in health care is shared across levels of government

Decision-making in health care tends to rest largely with the central government, which has considerable power across many aspects of the delivery of health services. More specifically, central governments are more likely to be responsible for decisions regarding the policy aspects of health care, but have less control over decisions regarding the inputs, outputs and monitoring of health care services. In most countries, subnational governments have major responsibility for input-related matters, such as determining the outsourcing of services and deciding on the contractual status of staff. While on average, local governments have little decision-making power in the health sector, they have the dominant responsibility with regard to decisions about health care inputs.

Figure 6 shows the allocation of responsibility for decisions in health care across respondent countries. It is calculated as the number of times a country responded that a level of government was responsible for a health decision, and then shows these sub-totals as a proportion of the total ‘yes’ responses, for each country. In certain cases, decisions are shared across more than one level of government.

Figure 6. Health sector spending authority varies widely across levels of government
Share of spending-related decision authority by level of government

Source: OECD Fiscal Network Survey of Spending Power in Healthcare (OECD, 2019).

Due to this decentralised structure in which different levels of government contribute to the decision- making process, collaboration between different levels of government is a good practice to improve the quality and the efficiency of the service provided through a better allocation of resources. Although subnational governments are mostly in charge of health inputs, in critical situations the inputs have an important role since health system capacity can become overwhelmed and inputs scarce. Thus, coordination is ever more crucial to attain a better allocation of resources that makes scarce inputs available to the people in need.

Source: Beazley et al. (2019), “Decentralisation and performance measurement systems in healthcare”, OECD Working Papers on Fiscal Federalism, No. 28.

Figure 7. Shared responsibilities in health and long-term care among levels of government

Source: Spending power survey (based on Dougherty and Phillips, 2019).

 Countries strategically ascribed new activities related to COVID‑19 containment to specific levels of government alone or to both of them in combination with a coordination mechanism

Although for both health and long-term care roughly 25% of the activities are shared among different levels of government, this share is estimated based on activities performed in normal circumstances. In order to deal with the COVID-19 outbreak, new activities that were not done before were implemented, such as enforcement of confinement, regional border controls, detection of infected people and management of scarce health inputs. Keeping delineation issues in mind, countries strategically ascribed those activities to specific levels of government alone or to both of them in combination with a coordination mechanism. Rarely were those activities performed in an uncoordinated manner. This is what Figure 8, below, shows. More specifically, activities that were extraordinarily performed or, for the case of health input management, performed in a different manner than usual, in most cases the central government worked either alone or together in a coordinated manner with the SNGs. Only in five countries did SNGs and central governments manage at least one of these activities in an uncoordinated manner, and four of them are federal countries: Australia, Belgium, Mexico and the United States.

Figure 8. Level of government in charge of selected measures

Source: Based on the COVID-19 Rapid Survey held by the Network in June 2020. Only countries that answered the question are shown. Australia ticked two conflicting options for “Regional border control” and only one option is shown.

This delineation of COVID-19 related activities is in line with OECD recommendations – in response to vertical fragmentation – to consolidate responsibility for an entire public service within a single level of government and, in the case of horizontal fragmentation, enhance cooperation (Forman et al., 2020). Such measures are justified, in part, by the fact that the COVID-19 outbreak respects no geographical and administrative boundaries and, thus, its impact transcend SNGs limits,6 which calls for the reassigning of public functions to higher levels of government in order to reduce spill-overs and to increase economies of scope and scale, or for a coordination mechanism that can align and harmonize the measures taken by different jurisdictions while respecting SNGs autonomy as well as closeness to the people in need.

 Re-centralisation of health care was more frequent than decentralisation

Figure 9 shows that in normal circumstances there is a non-linear effect of administrative decentralisation on health expenditure and health efficiency, in which a moderate degree of decentralisation seems to be preferred in a typical country, since it saves public resources and improve outcomes as compared with high centralisation or “excessive” decentralisation (for more see Dougherty et al., 2019a). Nevertheless, this may not necessarily hold true during an acute health crisis in which extraordinary measures needs to be implemented quickly, by-passing many of steps usually taken in policymaking and implementation.

Figure 9. Marginal effect of decentralisation on public spending and life expectancy

Note: The horizontal axis is the decentralisation index, from the most centralised (0) to most decentralised (6).

Source: Dougherty et al. (2019a).

Figure 10, below, shows that in order to better coordinate healthcare across levels of government, half of the countries surveyed have changed to some extent the health care task division during the COVID-19 crisis, either through re-centralisation or decentralisation. Among the countries that have changed health care task division, the majority of them responded to the crisis through (strictly) more centralisation, followed by the countries that increased both centralisation and decentralisation, depending on the activity. The less common rearrangement was to respond to the crisis through strictly more decentralisation.

Nonetheless, this observation does not mean that the optimal response is to re-centralise health care to tackle an acute health crisis. Both can be justified and work well to successfully address an acute distress. In general, centralised approaches might ensure standardisation of procedures, data collection, and better allocation of resources through the use of health inputs from all regions and, thus, lead potentially to a more equitable service provision across a country. Nevertheless, decentralised approaches might adapt faster to meet regional necessities, and may foster innovation since jurisdictions might approach the problem differently, culminating in a wider range of measures that could be copied in case they are successful. As a result, what is more important than defining the optimal degree of centralisation is to coordinate successfully the responses across all levels of governments, potentially in a manner that there is sufficient flexibility to adapt to local characteristics and consistency to provide comparable services throughout a country (see OECD, 2020b).

Figure 10. Changes in the healthcare responsibilities across levels of government

Source: Based on the COVID-19 Rapid Survey held by the Network in June 2020. Only countries that answered the question are shown.

 4. Fiscal Capacity

Fiscal capacity refers to the capacity of governments to collect revenue with the purpose of executing its function such as providing public goods and maintaining public administrations. In case of SNGs, it is generally recommended that subnational taxation and spending powers should be strengthened to allow governments to respond better to local needs and regional variations (Forman et al., 2020). This general recommendation is even more relevant in the context of COVID-19, since the gap between the revenue collected and the necessary expenditure to tackle the crisis is larger than usual, which, therefore, might require exceptional mechanisms to offset the additional costs and/or the fall in revenue.

 A higher spending and lower tax revenue may hurt SNGs fiscal capacity to take the necessary measures to tackle the COVID-19 outbreak

The COVID-19 outbreak may affect subnational governments’ fiscal position from both the expenditure and revenue side. On the expenditure side, SNGs may suffer severe consequences from the initial phase of virus propagation that pushes up health service demand and public order spending due to lockdowns. Social protection spending, on the other hand, may be affected for much longer, depending on the persistence of the economic crisis. On the revenue side, SNGs may experience a fall in revenues due to the weakening of economic activity and tax policy changes. As a result, SNGs’ fiscal situation may suffer from the combination of higher spending and lower tax revenue (this situation happened during the 2008-09 financial crisis – Box 2 explores this topic in more detail). Not all SNGs are exposed to the same risks – those SNGs reliant on volatile tax bases and charged with large health, public safety and social welfare responsibilities are more vulnerable.

Box 2. The 2008-09 financial crisis generated a combination of a strong reduction in revenue and increase in expenditures, damaging SNGs’ fiscal balances

While the impact of the crisis on subnational government finances varied across countries, most SNGs struggled with a decrease in tax revenues and rise in expenditure, sometimes referred to as a “scissors effect” (Blöchliger et al., 2010). Tax revenues fell sharply due to declining economic activity. In some cases, this was compounded by additional tax cuts included in national recovery packages. Subnational governments that rely the most on taxes faced the greatest deficit in 2009 (i.e., Canada, Iceland, New Zealand, Norway, Poland and Spain). The effect was even sharper in countries in which subnational governments primarily rely on a pro-cyclical tax base, such as corporate or personal income taxes (e.g., Spain and Canada).

At the same time, the crisis led to higher spending on unemployment, social protection and social welfare more generally. In many OECD countries, subnational governments are predominantly responsible for welfare services and social transfers. In EU countries, social protection spending by SNGs increased by 6.4% in 2009 and by 3.5% in 2010, as employment lagged behind economic activity (Dexia, 2011).

As a result, subnational government budget deficits rose in countries where these were allowed, while spending cuts or tax increases were required in countries where subnational governments had to follow balanced-budget rules (e.g., the United States). In Spain, for example, subnational governments were severely hit by the recession, facing both a collapse of their revenues (-7.7%) and a surge in their expenditures (+15.5%) (OECD, 2010). The unemployment rate, which soared from 8.7% in 2007 to 18% in 2009 and 20.7% in March 2011, prevented raising taxes and constrained subnational governments to increase social protection spending (+34.1% between 2006 and 2009).

As subnational governments’ revenues are often based on the previous year’s activity (e.g., for shared taxes, equalisation transfers, etc.), most SNGs saw the situation worsen in 2010 and 2011, despite some recovery at the national level. For US states, the fiscal year 2011 was one of the most difficult in modern times, with little improvement in 2012. According to the United States Centre on Budget and Policy Priorities (CBPP), 44 states projected budget deficits totalling USD 112 billion for Fiscal Year 2012 (Johnson et al., 2011).

Source: OECD/KIPF (2012), Institutional and Financial Relations across Levels of Government.

On the expenditure side, key expenditures to tackle this crisis represent a significant portion of SNGs’ budgets, leaving them vulnerable to spending hikes

Not only are SNGs key players in the provision of health, public order and social protection services (the latter to a lesser extent), but also these services represent a significant portion of SNG’s budgets (Figure 11). On average, 34% of their expenditures are on these three functions. Despite central governments’ having main responsibility for social protection, this expenditure item still consumes a large portion of SNG’s budgets, due to its size. Health expenditures are also very prominent – in Italy, they represent 48% of SNGs’ total expenditure and in countries such as Sweden, Denmark, Finland, Austria, Spain and Colombia its budget share is above 20%. Hence, health and social protection expenditure hikes may disproportionally hurt the fiscal balance of SNGs.

Figure 11. Share of health, public order and social protection expenditures as a % of SNG total expenditure, 2018

Source: OECD National Accounts COFOG database.

SNGs with a higher reliance on taxes linked to economic activity will be hit hard by the crisis and tax policy changes but inter-governmental grants may alleviate the impact

At present, SNGs draw their own-source tax revenues from a range of revenue streams with income, property and consumption taxes being the primary sources (Figure 12). Other taxes play a comparatively minor role. The tax mix itself appears strongly related to the extent of decentralisation. In the countries where payments of SNG taxes are being delayed, SNGs will suffer from revenue losses. The impact of tax deferrals on SNGs will vary in accordance with the relative importance of the deferred tax to the SNG tax base. Similarly, the impact of the crisis on overall SNGs’ revenues will vary in accordance with the income elasticity of the tax base. For example, greater SNG reliance on income relative to property tax is characteristic of countries with higher levels of sub-central fiscal autonomy (Figure 12). Among these countries, there is an inverse correlation between the income elasticity of taxation and property tax contribution to the SNG tax base, consistent with related work by Colin and Brys (2019). This suggests that SNGs that are more dependent on income taxation will face particular risks. In countries where SNGs rely primarily on tax sharing, the effect of the crisis may come with a lag, as tax sharing is usually calculated based on the previous year’s revenues.

Figure 12. Subnational tax base composition, 2018

Note: Countries are grouped into quartiles by SNG share of total government revenue, decreasing left to right. Within each quartile, countries are sorted by income tax in SNG revenue for the top two quartiles, and property tax in SNG revenue for the bottom two quartiles. Where data for 2018 was not available, the most recent data were used.

Source: OECD Revenue Statistics database.

The extent to which SNGs are dependent on inter-governmental grants varies greatly across countries (Figure 13). These grants range from 84% of SCG revenue in Lithuania to less than 10% in Iceland. The ability of such grants to shield SNGs from the fiscal impact of the crisis depends on whether they have pro-cyclical or counter-cyclical design features, with the latter being more protective during an economic downturn. To provide financial support to SNGs during the current crisis, central governments in multiple countries are boosting inter-governmental grants (e.g. Austria, Brazil, Colombia, Estonia, Italy, Japan, Korea, Latvia, Norway, Slovenia, South Africa, Spain, the United States) and in some of these countries the increase in grants are proportional to the reduction of revenue or increase in costs. As a result, inter‑governmental grants have been used as a strong counter-cyclical measure to help SNGs in need.

Despite the fact that inter-governmental grants might be a useful tool to help SNGs in need, there are three main issues with their use. First, depending on the rules under which the grants are provided, they may influence the fiscal and health policies implemented by SNGs, causing moral hazard issues. More precisely, in case SNGs don’t bear the costs of their own policy measures, they might choose more popular but costlier measures since these will be financially compensated by the central government.7 Second, in some countries SNGs have substantial tax, expenditure and policy autonomy and, thus, in principle, they should bear the cost of their own policy measures and, if needed, could get loans directly from financial markets.8 Third, depending on the rules under which grants are provided, they may damage the equity across SNGs through the provision of more funds to some regions at the expense of others and, moreover, since financial needs are not policy neutral (the financial needs of SNGs in the context of the COVID-19 outbreak might depend on their own past policy choices). These equity issues are very difficult to be addressed in a manner that satisfies all parties involved. In summary, it is important to carefully analyse SNGs’ needs and circumstances before providing extraordinary grants, and, in case needed, design those grants in a manner that reduces moral hazard and considers SNGs’ autonomy and equity implications.

Figure 13. SNG own-source revenues vs inter-governmental transfers, 2017

Note: This figure depicts SNG own-source revenues as a percentage of the sum of own-source revenues and inter‑governmental transfers.

Source: OECD Fiscal Decentralisation database.

There is a huge uncertainty with regards to SNGs’ revenue in the current fiscal year – forecasts using past data are unlikely to capture the unique dynamic of the current shock

Despite the fact that past elasticities can serve as a proxy to how vulnerable are SNGs’ tax revenues with respect to an economic downturn, forecasts for the current shock that are based on those elasticities might be biased. The COVID-19 shock is asymmetrical in terms of economic activity and, thus, the reduction in revenue depends on the exposure of the regional economy to global value chains and on the dependency on services that are heavily affected to restrictive measures (see OECD, 2020b). For instance, consumption tax elasticity with respect to GDP might change more in comparison to previous years since restrictive measures are hitting retail sectors especially hard. This elasticity change might be even more severe in regions that depend on tourism, since tourism came to a halt during the peak of the pandemic. Thus, it would not be surprising if the past consumption tax elasticity with respect to the GDP widely differs from its elasticity in the COVID-19 crisis, significantly reducing the precision of forecasts based on the former. For these reasons, it is now especially important to use tax bases instead of the GDP as a predictor for each individual tax revenue. In that manner, an asymmetrical impact on, for instance, consumption could be incorporated into the prediction’s value and, thus, be captured by the forecast.9 Moreover, when possible, it is also recommended to base revenue forecasts on sector-specific information to capture sector-specific effects.

Box 3. The problem with SNGs’ revenue elasticities with respect to the GDP in times of crisis

Revenue elasticities and buoyancy (the different between both is that the former control for policy changes) are usually calculated through the use of an ordinary least squares regression that regress the logarithm of the revenue onto the logarithm of the GDP or onto the logarithm of the respective tax base. As a result of this estimation procedure, the parameters are computed in a manner that minimises the sum of the errors throughout the whole dataset. Since most data points refer to periods of mild (positive or negative) GDP growth, it is safe to say that models estimated in that manner are good to forecast revenue under the assumption that GDP will behave in that same manner.

In addition, SNGs’ revenues have two especial characteristics that, in time of crisis, may lead to a non-linear behaviour with respect to the GDP and tax bases. First, a significant portion of their revenue comes from inter-governmental grants or revenue sharing agreements (Figure 13, above), whose values may be calculated through formulas that are based on lagged variables, which, thus, reduce the negative impact in the current year at the cost of an larger negative impact in the next year. Second, property taxes represent a significant portion of some SNGs revenue and the tax base on which property tax are based regards property prices, which do not react promptly to shocks in the short-term and may be valuated only periodically.

Due to these reasons, SNG revenues are very challenging to forecast in times of crisis without having extraordinarily detailed data. Figure 14, below, compares how the SNGs’ and central governments’ revenue buoyancy varied each year. It is clear from those plots that in 2009, the peak of 2008-09 financial crisis, subnational revenue elasticity changed significantly in most countries whereas the central government’s revenue elasticity changed only mildly. SNG revenue buoyancy shrunk in 2009, which means that SNG revenues reacted less abruptly to GDP movement in comparison to previous years. Since in most countries the GDP growth was negative in 2009, SNGs enjoyed a smaller fall in revenue in comparison to central governments.

Figure 14. Revenue buoyancy with respect to the GDP by year and level of government

Note: Each dot represents the revenue buoyancy of that specifically year for a specific OECD country. Min/max ranges are shown with lines, blue boxes are middle two quartiles.

Source: Calculations based on OECD System of National Accounts

One common solution to alleviate this problem is to restrict the dataset to only periods that are representative of the current situation, capturing, thus, the GDP elasticity/buoyancy in times of crisis. For instance, for central governments the decrease in tax revenues during the 2008-09 financial crisis exceed the GDP contraction in most countries, even if normally the tax revenue elasticity with respect to revenue is close to one. More specifically, the average tax-to-GDP ratio in OECD countries fell by 1.4 percentage points (for a more detailed analysis comparing different tax bases and countries see OECD, 2020c) whereas this same ratio for property taxes fell only by 0.2 and for inter-governmental grants increased by 0.97. Under the assumption that the revenue buoyancy/elasticity in the 2008-09 financial crisis captures well the characteristics of the current crisis, this estimation could be used to forecast the impact of the current crisis.

Although indeed past crises represent better the current situation than periods of mild GDP growth, this assumption disregards the fact that the COVID-19 crisis has no parallel in history. Maybe the most similar crisis was the Spanish Flu, which happened roughly 100 years ago, in a far less globalised world. In the current pandemic, contamination occurs rapidly across borders and virtually all countries in the world took drastic actions, with unforeseen consequences. Today’s industrial production occurs worldwide integrated through global supply chains that involve multiple actors across borders – the consequences of disruption are still far from clear. Moreover, in comparison to the 2008-09 financial crisis, this crisis seems to be far more severe – thirty million Americans filed for unemployment in mere six weeks in comparison to one year in the Great Recession (Chen et al., 2020) and even considering that this crisis really started in OECD countries, principally in the second half of the first quarter of 2020, the first quarter of 2020 had an economic performance similar to the bottom trough of the Great Recession. As a result, the uncertainty is immense, and caution is recommended when relying on precise forecasts based on aggregate data and assumptions that consider that the past is alike the current situation.

The impact of the crisis on the reduction in subnational revenues comes from two sources: (1) the reduction in economic activity and (2) changes in tax policies. In a number of countries (e.g., Argentina, Australia, Belgium, Canada, France, Greece, Italy, the Netherlands, Norway and others), central governments are delaying of the payment of taxes or reducing tax bases or rates to support businesses and vulnerable citizens. Some of these taxes are SNG taxes. Deferral of shared personal income taxes and predominantly local property taxes might directly impact local governments.10 Deferring payments on taxes which constitute SNG revenues is sometimes decided by the central government, or the centre is recommending SNGs to do so (Lithuania), or is decided by SNGs themselves (Australia, Canada) (OECD, 2020e). While in some countries the tax authorities have clearly stated they will compensate SNGs for the transitory reduction in revenues – for instance, in Chile (OECD, 2020e) – this is not always so. When possible, these policy impacts should be considered in revenue forecasts.

SNGs budget rigidity, limited access to borrowing and compliance with fiscal rules tends to reduce their room to manoeuvre and, they thus react less counter-cyclically than central governments

In the great majority of countries, the fiscal situation of subnational governments is expected to deteriorate, which will put pressure on subnational budget balances and debt (Figure 15). Notable exceptions are Latvia and Luxembourg, which are not expecting an increase in neither subnational debt nor deficits; Mexico, which is not expecting increase in subnational debt; and Spain, which is not expecting an increase in subnational debt.

Figure 15. Expected impact of the COVID-19 crisis on SNGs finances

Source: Based on the COVID-19 Rapid Survey held by the Network in June 2020. Only countries that answered the respective question are shown.

The fiscal balances of SNGs are negative in eight of the countries (United Kingdom, Austria, United States, Canada, Norway, Sweden, Finland and Spain) and, in those with fiscal surpluses, in eight countries the surplus exceeds 5% of their revenues (Figure 16). A substantial part of subnational expenditures is either mandatory or regards essential services and SNGs have limited autonomy to increase revenues since they depend on inter-governmental transfers and have limited autonomy to change tax policies (see Dougherty et al., 2019b). As a result, in the short-term the amplification of fiscal imbalances is probably not going to be compensated by neither expenditure reductions nor increase in revenues which may aggravate fiscal imbalances.

Figure 16. Subnational fiscal balance as a percentage of their revenues, 2016

Source: Calculations based on consolidated data in the OECD Fiscal Decentralisation database.

This question of SNGs facing higher budget rigidity in terms of expenditure and revenue can be examined through a comparison between central government’s and SNGs’ reaction to economic cycles. Figure 17 shows that subnational fiscal balances tend to fluctuate less than central governments’, which might occur because SNGs tend to have more budget rigidity. This fact also makes SNGs’ reaction less counter-cyclical than the central government’s. Aside from budget rigidity, two other characteristics of SNGs that give them less room to manoeuvre regards their limited access to financial markets in comparison to central governments and the fact that – particularly in unitary countries – SNGs may follow strict fiscal rules.

Figure 17. SNG fiscal balances usually fluctuate less than central governments’
Average central and subnational fiscal balance as a percentage of their revenues for OECD countries

Note: Fiscal Balance/Net Borrowing (GB9) as a percent of Total Revenue (GTR) of the respective level of government. Non-consolidated data.

Source: Calculations based on the OECD National Accounts database.

SNGs’ debts are expected to increase

The negative impact of the crisis on subnational fiscal balances may be covered by an increase either in inter-governmental transfers or in SNGs’ debts. While, on average, SNGs’ debts only represent about 15% of total public debt and 12% of GDP (Figure 18), these may represent a high burden on SNGs’ budgets.

Figure 18. Composition of public debt as a share of GDP, 2016

Note: This figure shows the respective contributions of SNG and central government plus social security (CGSS) to overall public debt as a percentage of GDP.

Source: Calculations based on OECD Fiscal Decentralisation database (for the debt) and OECD National Accounts database (for the GDP). When the data was not available for 2016, the ratio for the closest year is shown.

In some countries, average levels of SNGs’ debts (measured as their debt to revenue ratio) are high (Figure 19). In 20 out of 31 countries SNGs’ debt is higher than one entire year of fiscal revenues and only in one country (Hungary) does SNGs’ debt represent less than half of one fiscal year revenue. It is worth noting though that the task of defining a debt level in which a SNG may be considered insolvent is a non-trivial and subjective task.11 In addition, SNGs are generally considered to have a softer budget constraint than central governments since the later can, under extraordinary circumstances, rescue SNGs in need.12

Figure 19. Subnational debt to revenue ratio by country, 2016

Source: Calculations based on OECD Fiscal Decentralisation database using consolidated data. When the data was not available for 2016, the ratio for the closest year is shown.

SNGs face borrowing limitations since subnational bond market are usually less developed and illiquid than national bond markets, while there are borrowing rules that forbid subnational borrowing under some conditions, and SNGs have limited autonomy to change fiscal policy. More specifically, SNGs usually have limited power to change their tax policy, are dependent on inter-governmental transfers, have little discretion over spending, and are in charge of key public services that cannot be adjusted abruptly (e.g., education, health, public safety). As a result, SNGs cannot easily roll-up debt nor adjust their budgets to deal with acute shocks, which makes them more vulnerable to liquidity crises than central governments.

In that light, it is worth analysing the impact of SNG debt service on their budgets. All other things held equal, the higher the debt of a SNG, the higher the interest rate that it must pay to service the debt, which can further reduce their room to manoeuvre in critical situations. In 16 out of 31 countries SNGs’ interest payments as a share of their revenues are lower than 2% (Figure 20). Only in six countries (the United States, Canada, New Zealand, the United Kingdom, Iceland and Spain) do SNGs have interest payment to revenue ratios above 4%, with the highest being the United States, where interest payments represent about 10% of SNGs’ revenues.

The interest payments share of SNG budgets is, on average, approximately half of the same share for central government. The average spent on interest payments as a percentage of their respective budget is 2.6% for SNGs in comparison to 5% for central governments (Figure 20). Despite this fact, the interest rate may constrain more SNGs’ fiscal policy than central governments’ since SNGs have less room to manoeuvre and more difficulty to roll out debt.

Figure 20. Central government’s and SNGs’ interest paid to revenue ratio by country, 2016

Note: Calculations based on OECD National Accounts for interest rates (transaction GD41P) and consolidated revenue (from the OECD Fiscal Decentralisation database). For the USA, SNG data refer to the state level only.

Source: OECD Fiscal Decentralisation and National Accounts databases.

Budgetary pressures on SNGs are likely to be proportional to (1) reliance on volatile tax bases and consumption taxes, (2) health, social protection and public order expenditure and (3) fiscal deficits

Table 2, below, depicts how SNGs in OECD countries are in terms of expenditure share of key government functions to tackle the current crisis, fiscal balance to revenue ratio and property tax as a share of tax revenue. SNGs from three Nordic countries (Finland, Sweden and Norway) might suffer more budgetary pressure in the current crisis considering that they do not rely on the more stable property tax, are in charge of health, social protection and public order functions, and have a negative fiscal balance. On the other hand, SNGs from Czech Republic, Estonia, Greece, Israel and Slovak Republic might suffer less than SNGs from other countries considering that they rely significantly on property tax, are not in charge of those key government functions and are in a positive fiscal situation.

Table 2. Budgetary pressures on SNGs in OECD countries considering three indicators


Health, Social Protection and Public Order Budget Share

Fiscal Balance to Revenue Ratio

Share of tax revenue that does not come from property taxes









Czech Republic




































































Slovak Republic




















United Kingdom




United States




Note: It is worth noting that this analysis is simplified and does not control for policies, severity of the health crisis and dependency on specific sectors of the economy. For instance, SNGs in countries that depends on tourism or that were more severely hit by the health crisis might be affected even more even if their SNGs’ revenue sources are not as volatile as others and their SNGs’ expenditures are not directly affected. Moreover, exogenous policies, which were not considered in that analysis (e.g. tax exemptions), can have a greater impact than the reduction in economic activity.

Thresholds for colours are, by column, respectively: (1) >50% red, <50% and >25% yellow, <25% green; (2) >2% green, <2% and >0% yellow, <0% red; (3) >50% red, <50% and >25% yellow, <25% green.

Source: For the column (1) OECD System of National Accounts (as of 2018); for the column (2) OECD Fiscal Decentralisation Database (as of 2016 or closest year); and for the column (3) OECD Revenue Statistics (as of 2018).

The result of the necessary increase in spending in 2020 combined with the reduction in revenues might lead SNGs to cut investments as they did during the 2008-09 financial crisis, which may further aggravate the economic crisis

SNGs tend to invest more than central governments as a proportion of their revenues. The investment share as percentage of total revenue is around 15% at the subnational level and 5% at the central level (Figure 21). Economic downturns that create fiscal imbalances are corrected to some extent at the subnational level through a significant reduction in investment13 since most of their expenditures are rigid and since they have to comply with fiscal rules. This situation is not ideal since SNGs are responsible for nearly half of public investment which aggravates the crisis due to a pro-cyclical effect of reducing investments in times of economic distress14 (OECD, 2011).

Figure 21. SNG investment usually fluctuate more than central governments
Average central and subnational investment as a percentage of their revenues

Note: Gross capital formation is used as a proxy for investment (GP5P) and is presented as a percent of Total Revenue (GTR) of the respective level of government.

Source: Calculations based on the OECD National Accounts database.

Flexible compliance with fiscal rules can be considered as an option to prevent pro-cyclical fiscal policy at the subnational level and to increase SNGs’ capacity to take the necessary measures

Fiscal rules are institutional arrangements whose purpose is to mitigate subnational fiscal risks through the imposition of constraints on fiscal policy, usually operationalised by limits on certain aggregates. Their main goal is to ensure fiscal sustainability at the subnational level in order to prevent subnational debt crises that might pose systemic macroeconomic risks and require centralised bailouts. Fiscal rules that constrain SNG budgeting are very common across the OECD (Table 3), and, in many countries, these were further strengthened after the 2008-09 crisis. In particular, most OECD countries require SNGs to balance their budgets, and many countries only allow SNG borrowing for financing public investment.

Table 3. SNG fiscal rule practices

Subnational government

Budget balance rule

Expenditure limit

Taxation limit

Borrowing restraint

Subnational government

Budget balance rule

Expenditure limit

Taxation limit

Borrowing restraint

Australia - State level





Latvia - Districts and Municipalities

Austria - State level

Lithuania -Municipalities

Belgium - Regions and Communities



The Netherlands - Municipalities

Brazil - States and Municipalities

New Zealand - Local authority


Canada - Provinces





Poland - Municipalities

Estonia - Municipalities


Portugal -Municipalities

Germany - Lander

Slovenia - Municipalities

Iceland - Municipalities

South Africa –Provinces

Ireland - Local governments

Spain - Regions and Municipalities

Italy - Provinces, Regions and Municipalities

Switzerland - Cantons





Korea - Municipalities

United States of America - States





Note: This table is based on a preliminary assessment of the 2019 Fiscal Network Survey of Subnational Fiscal Rules. * indicates that the fiscal rule is self-imposed by the SNG, without the intervention of a higher level of government. ** Such restraints have not been imposed in practice by the federal government on regions and communities.

Source: OECD Fiscal Network Survey of Subnational Fiscal Rules (2019).

The restrictive nature of fiscal rules in general, and balanced budget rules in particular, necessarily triggers concerns about pro-cyclicality (OECD, 2013). More rigid rules and rules that apply to shorter timeframes are even more susceptible to pro-cyclical tendencies. During a time of crisis, it may be possible to relax such rules with a reduced risk of moral hazard problems emerging later. Relaxation may take the form of both formal escape clauses that can be triggered by prescribed circumstances and/or effective suspension of the rules in practice when it is unreasonable to expect SNGs to comply. The latter case may be particularly pertinent where SNGs self-impose their own fiscal rules, implying that they are equally in a position to lift or adjust them.

Despite fiscal rules’ fundamental role in ensuring fiscal sustainability, during times of crisis compliance with such rules might not be attainable or even desirable. The use of escape clauses and flexibility can help to mitigate the pro-cyclicality of fiscal rules and ensure an adequate fiscal policy during times of crisis. A large portion of countries that have adopted subnational fiscal rules have similarly implemented escape clauses (Table 4). In some cases where formal exemptions to fiscal rules are not specified in legislation, mechanisms exist whereby central governments take a holistic view of circumstances in assessing fiscal rule compliance.

Similar to escape clauses in fiscal rules that allow for exceptional borrowing, withdraws from rainy day funds may serve as a counter-cyclical fiscal instrument. By enabling jurisdictions to save funds in a reserve when revenues allow, they create some fiscal space to react during emergencies. While rainy day funds are uncommon at the subnational level, there are some notable examples, such as the US and Canada where 48 of 50 states and 6 out of 10 provinces possess them, respectively.

Table 4. Existence of SNG escape clauses in fiscal rules


Escape clauses

Rainy day funds

Australia-State level



Austria-State level

Exemptions possible when finances are adversely affected by local economic conditions, disasters, supreme court rulings, or EC decisions.

Belgium-Regions and Communities




Exemptions possible for budget balance objectives and borrowing constraints in case of low economic growth or public calamity.



Exemptions possible when finances are adversely affected by falls in revenue, local economic conditions, or disasters.

Yes, 6 of 10 provinces


Sanctions only take effect when SNG violates limits two-years in a row


Germany -Lander

Exemptions possible for borrowing constraints in case of exceptional emergency conditions.





Ireland-Local governments



Italy- Regions and Municipalities




Exemptions possible for borrowing constraints when finances are adversely affected by falls in revenue, local economic conditions, or disasters.


Latvia-Districts and Municipalities

Exemptions possible when finances are adversely affected by natural or other disasters



Exemptions possible to balanced budget when finances are adversely affected by a deterioration of local economic conditions.


The Netherlands-Municipalities



New Zealand-Local authorities



Poland- Municipalities




Exemptions possible when finances are adversely affected by natural or other disasters



Exemptions possible for expenditure limitations when finances are adversely affected by falls in revenue, local economic conditions, or disasters.


South Africa-Provinces

In the case of an emergency, provinces are allowed to spend without following all the normal budgeting process, but are only allowed to spend up to 2% of the total provincial budget.


Spain–Regions and Municipalities

Exemptions possible for balanced budget objectives when finances are adversely affected by falls in revenue, local economic conditions, or disasters.


Switzerland- Cantons

About half of the cantonal fiscal rules foresee exemptions for special circumstances, but the rules vary by the type/design of the fiscal rule. The rules of the Canton of Bern foresee no exemptions.

None, however some Cantons maintain some reserve on their balance sheets which are not referred to as rainy day funds.

Unites States of America



Source: OECD Fiscal Network Survey of Subnational Fiscal Rules (2019).

There is already some evidence of asymmetric regional impacts of COVID-19 on SNG finances

SNGs within countries differ substantially in terms of budget composition, revenue sources, dependency in certain economic activities, capacity to tackle the crisis and how severely their people were hit by the pandemic (see OECD, 2020b). Therefore, it is also expected that SNGs may face different challenges within countries both in terms of magnitude and nature. Approximately half of countries (10 out of 21) indicate that asymmetric impacts have been observed across regions (Figure 22). While a significant portion (7 out of 20) are not yet able to determine whether impacts are asymmetric, only a small minority of countries (Australia, Latvia, Switzerland and Austria) indicate that they have not observed asymmetries.

Among countries that have observed asymmetry, a near equal proportion has observed asymmetric cost increases (10 out of 10) as has observed asymmetric revenue losses (9 out of 10). However, when it comes to the magnitude of these impacts, the majority of respondents (11 out of 19) indicate that SNGs’ revenues, rather than costs, will be harder hit. Only one country, Spain, indicates that they anticipate costs will have a greater impact on SNG finances.

Figure 22. Asymmetric impacts on subnational revenue and expenditure and fiscal equalisation

Source: Based on the COVID-19 Rapid Survey held by the Network in June 2020. Only countries that answered the respective question are shown.

Most countries have equalisation systems that might help mitigate regional disparities, but this mitigation might be limited and depends heavily on the equalisation system distribution formula

Equalisation systems aim to mitigate regional differences in fiscal capacity and expenditure needs by transferring resources from a central government or wealthier SNGs towards SNGs that face higher costs or lower revenues. These grants aim at reducing inequality on the revenue side, due to the uneven distribution of taxable activities across jurisdictions of a country, and on the expenditure side, due to inter-regional demographic or geographical differences that may require a larger expenditure on a specific function within one region (for instance, a region with an ageing population or high unemployment may need more health and social protection resources than others).

Fiscal equalisation systems vary significantly among countries and are shaped by the extent of fiscal decentralisation and subnational fiscal autonomy, as well as the nature of the vertical fiscal gap and historical responsibility assignment to SNGs, among other factors. Normally, equalisation systems are redistributive rules-based mechanisms which aim to compensate for regional imbalances that may be exacerbated by decentralisation. Figure 23, below, depicts how countries differ in the total amount of equalisation funds as a share of SNGs’ expenditure.

Figure 23. Total amount of equalisation funds as a share of SNGs’ expenditure

Source: OECD Fiscal Network fiscal equalisation survey.

In theory, fiscal equalisation may work as an insurance against asymmetric shocks by providing revenues collected in one jurisdiction to those more severely affected. However, in practice equalising transfers are typically sensitive to changes in tax collection either because they are funded by earmarked revenue streams or are capped at a certain growth rate. Thus, a fall in central government revenue (or subnational revenue) is spread through the equalisation system to other jurisdictions, which may exacerbate the fluctuation, acting pro‑cyclically. At the same time, it is common for countries to link equalisation transfers to lagged fiscal capacity or to a moving average, which may mitigate this pro-cyclicality, but at the cost of reducing equalisation effects in a time of crisis – for instance, if a jurisdiction is experiencing a dramatic fall in revenue, through the equalisation formula it should receive more transfers, but since the formula is lagged, this jurisdiction will only level its revenue in the next fiscal year, leaving it vulnerable in the current year (OECD, 2013). As a result, given their pro-cyclical tendencies and typical reliance on lagged measures, equalisation systems may not be well suited to meeting the extraordinary needs of SNGs during a time of crisis, necessitating additional measures.

Often, equalising transfers are funded by buoyant revenue sources which may reduce the total amount paid during times of crisis

The asymmetric impacts of COVID-19 may interact with equalisation systems by changing the geographic distribution of the demand for equalising payments and, in principle, an efficient equalisation system could alleviate asymmetric differences of the current crisis. However, where funds for equalising transfers are tied to dedicated revenues streams, the available revenues may shrink as a result of the pandemic, with a resulting negative impact on the size of the transfers themselves. The vast majority of respondents (17 out of 21) indicate that equalisation systems exist in their countries (Figure 24, covering these). Three federal countries (Austria, Mexico, and the United States) indicated that they do not have equalisation systems.

Figure 24. Fiscal equalisation funding streams

Note: Only countries that answered the respective question are shown. Poland and Latvia were considered to have a buoyant revenue stream since both mentioned that the equalisation formula involves tax revenue collected.

Source: Based on the COVID-19 Rapid Survey held by the Network in June 2020.

Within many equalisation systems (10 out of 17), funding for transfers comes from a source that fluctuates with revenue receipts (Figure 25). In 8 cases, equalisation is at least partially funded by appropriations from central government revenues and in 6 cases it is at least partially funded by horizontal transfers among SNGs. Depending on the income elasticity of revenue at the central and subnational levels, transfers funded by central government and horizontal transfers may both be susceptible to contractions in economic activity. This helps to explain Figure 25, which shows that roughly half of respondents (8 out of 17) anticipate a fall in total equalising transfers, whereas only one country, Canada, anticipates an increase to one of its two equalising transfers (the Territorial Financing Formula). Overall, this suggests that equalisation systems may have a pro-cyclical impact on subnational finances.

In cases where no effect on equalising transfers is anticipated, the impact of the pandemic may be expected to be symmetrical (e.g. Australia, Switzerland), difficult to assess at this point in time (e.g., Australia, Belgium, Estonia, Norway, and Spain) or bear no direct impact on the determination of equalisation payments (e.g. Italy and Canada’s provincial equalisation formula). Importantly, several respondents indicated that the use of lagged variables when computing equalisation payments means that the effect of the pandemic may not be reflected in the actual sums paid to SNGs for several years (e.g., Australia, Colombia, Finland and Poland).

Figure 25. Impact on total amount transferred by fiscal equalisation

Note: Only countries that answered the respective question are shown. Canada expects no direct effect on provincial equalisation transfers and an increase in payments made via the Territorial Financing Formula.

Source: Based on the COVID-19 Rapid Survey held by the Network in June 2020.

The following list provides country-specific examples of the anticipated impact of the crisis on fiscal equalisation transfers:

Australia: The base on which payments are made (GST pool) will be affected by economic activity, likely to fall initially and then recover, although it is not clear if it will return to pre-COVID-19 levels in the next two fiscal years. In terms of the equalisation, the revenue and expense impact on SNGs will be different based on their infection levels and policy responses. However, it is not yet clear how the Commonwealth Grants Commission will interpret and assess differences across states. The lag in assessment also means the impact of any redistribution will occur over at least 5 years.

Brazil: The most important fiscal transfers, including those with an equalising effect,15 are funded by a percentage of the income tax and the VAT on industrialised products. Due to the expected recession, the resources collected through these taxes would decrease, which would consequently reduce the transfers to SNGs.

Canada: The Territorial Formula Financing, which addresses horizontal disparities in the northern territories, is expected to increase more quickly over the next two fiscal years since growth in the program's funding envelope is tied to provincial and local government spending. Spending is expected to increase in 2020 in response to the COVID-19 pandemic, before returning to more normal levels thereafter. Canada's provincial equalisation program is legislated to grow in line with the economy. The annual increase in equalisation is determined by a three-year moving average of nominal GDP growth. At this point, the three-year moving average is still expected to generate positive growth in the transfer in the following years, although that growth is lower than it would have been in the absence of the economic shock from the pandemic.

Colombia: Initially, the most prosperous regions will be directly affected by reductions in local tax revenue. Lagged effects are apparent with respect to fiscal transfers as in 2021, all SNGs will face a reduction in the national grants that fund most health and education investments. The most impoverished regions of the country are highly reliant on these grants, even for administrative expenses. The grant system that constitutes the equalizing transfers to SNGs depends on the amount of revenue collected by the CG. A reduction in economic activity and tax revenue will reduce the growth of transfers to SNGs.

Finland: The cost increases have been asymmetric across regions, with the number of COVID-19 patients outside the capital region being very limited. The impact on tax revenues has been greater this year, but it is expected to return towards its normal level by next year. In the Finnish state grants system, there is a correction made to the cost base based on the final fiscal data, so if there is a permanent increase in costs due to COVID-19, it will be corrected with a two-year delay.

Iceland: The income of the equalisation fund in 2020 will be, according to projections, 10‑11% lower than envisaged in the 2020 Budget. Income comes from current tax revenue of CG and SNGs. Taxes, such as on capital income and corporate income tax, are levied on previous years. Tax revenue from these sources will most certainly fall significantly, impacting the fund's income in 2021.

Italy: The total amount of resources transferred by fiscal equalisation remains unchanged because equalisation (both horizontal and vertical) is based on the relative positions of municipalities (distribution coefficients) and not on the absolute amount of fiscal capacity and standard need. The total amount of the equalisation fund is established by law and at the moment there are no changes in the allocation.

Korea: The direct cost impact of COVID-19 is relatively low due successful management with a low level of casualties. However, revenue reductions related to COVID-19 will be very high due to its impact on overall economic activity. A sizable amount of SNG revenue comes from general grants which are fixed at 19.24% of CG tax revenues (except earmarked taxes). So, with the great economic shock caused by the COVID-19 crisis, SNG revenue will be greatly affected in the coming years.

Mexico: Federal transfers to SNGs, which account on average for almost 80% of their incomes, are associated with the Federal Tax Collection which it is presumed will behave differently next year, due to the national economy recovery period.

Poland: The reduction in expenditure planned by local governments is steeper than the decrease in income. Variables used in determining equalisation payments are calculated on the basis of lagged (t-2) data. Based on current forecasts, it should be expected that in 2021 the equalisation fund will increase compared to 2020 due to the increase in tax revenues of local government units in 2019 compared to 2018. Forecasting the situation in 2022 is subject to high risk due to the high uncertainty associated with the COVID-19 pandemic. At present, a decrease in the equalization fund should be expected due to lower tax revenues of local government units in 2020.

South Africa: The revenue impact will be larger for municipalities, but it is not yet clear how long this will impact last. Conversely, the spending impact will be higher for provinces which are responsible for the health system. Transfers to municipalities will rise in the first year, but subsequently the economic impact of the crisis will constrain CG revenue and lead to larger reductions in transfers to municipalities and provinces in the second year.

Spain: In 2020, cost increases related to COVID-19 will have a greater impact than revenue losses. While 2021 projections are highly uncertain, the fiscal revenues of the local entities are relatively stable, as they are not associated with the economic cycle.

Switzerland: A symmetric impact on revenue side is expected, which should not have an effect on fiscal equalization.

 To level out SNGs’ fiscal capacity, central governments and central banks are stepping in, often through additional grants, loans and guarantees, and/or temporarily lifting of fiscal rules

In most countries, central governments and/or central banks are stepping in to support SNGs in that critical moment. Even in countries that SNGs are believed to have sufficient reserves to avoid a short-term liquidity problem (Spain, Latvia and Australia), SNGs are receiving/may receive some support from central governments or from the Central Bank. Switzerland is the only country that reported that SNGs should respond to the crisis themselves, including potentially through insolvency.

The most common response from central governments was providing extraordinary grants (13 countries). Nine countries have temporarily lifted fiscal rules (in particular in the EU countries, this was allowed by the “general escape clause” by the European Commission16), and in six countries, the central government is providing additional loans and guarantees to SNGs (Figure 26).

Figure 26. Measures employed by central governments to improve subnational liquidity

Source: Based on the COVID-19 Rapid Survey held by the Network in June 2020. Only countries that answered the respective question are shown.

Examples of specific country responses:

The Australian Reserve Bank provided support through purchase of semi-government bonds, during the period of instability in markets.

In Brazil, the central government is providing fiscal relief by: (1) suspending the service of SNG debt owed to the central government; and (2) allowing SNGs to renegotiate with banks to suspend their debt payments in 2020.

The Canadian Central Bank is providing liquidity support to SNGs.

The central government of Colombia adopted four additional measures: (1) increased the maximum value of short-term and long-term loans for SNGs and SOE; (2) offered subsidized-rate credits from public financial corporations to SNGs; (3) enabled SNGs to draw from local pension funds; and (4) temporary removed restrictions to reallocate earmarked revenues.

In Iceland, central government measures regarding finances of local authorities are being contemplated, and will be decided following the findings of a special task force, due in June/July 2020.

Italy exempted CIT (corporate income taxes) for small companies and property tax for the tourism sector. As these are a source of SNG revenues, three measures were taken to compensate SNGs: (1) compensating revenue reductions caused by these tax measures; (2) providing cash advances to local authorities for the payment of commercial debts; and (3) providing extra funds to SNG to ensure the funding of their fundamental functions.

Latvia’s central government will compensate SNGs for decreases in the subnational personal income tax share. Nevertheless, so far, income tax collection has been in line with projections.

Luxembourg fiscal rule ensures that annual repayments of all local government loans are less than 20% of their operating income. This thus leaves local governments room for manoeuvre to absorb shocks and avoid over-indebtedness. The central government has strongly encouraged the public sector not to cut investment and has assessed that municipalities have sufficient resources.

In Mexico, central government has ensured liquidity by paying the federal transfers set out in the Law of Fiscal Coordination promptly.

Norway established a working group composed of central government ministries and the Norwegian Association of Local and Regional Authorities (KS) to address the fiscal impact of the COVID-19 outbreak on local governments. Several municipalities faced difficulties to refinance their loans on the market. The equity capital of Kommunalbanken17 (KBN) was increased by NOK 750 million “to help the markets to function as well as possible and to prevent municipalities’ refinancing of short-maturity securities from contributing to further stress in the markets”.

Poland created a COVID-19 Counteraction Fund, which can provide grants to SNGs to finance or co-finance activities to counteract the COVID-19 pandemic. Moreover, disbursements of general grants may take place earlier than required by law, while local government’s payments into equalisation mechanisms scheduled for June and July 2020 may be postponed to the period between August and December 2020. Actions taken to respond to the COVID-19 outbreak will not be taken into account for assessing compliance with fiscal rules, (1) both revenue losses and additional current expenditure resulting from the COVID-19 outbreak will be excluded from calculations of the deficits in the balanced budget rule; and (2) loans taken to compensate revenue losses will be excluded from the debt service ratios in the borrowing constraints rule.

In Spain, among the specific measures adopted in order to strengthen the financing of the sub‑central governments, it is important to emphasize the creation by the Central Government of the Covid-19 Fund. The purpose of these 16 billion funds is to increase the funding for the Autonomous Communities, so that they can deal with the increase in expenditure and the income drop caused by the current crisis. The distribution among the Autonomous Communities is based on criteria related mainly to expenditure in health, education and the reduction of revenues, although the funds are non-earmarked.

United States: the Federal Reserve is lending and providing liquidity to SNGs while the Treasury is guaranteeing the loans given through that mechanism.

Additional funds are mainly provided with no conditions attached but, in most cases, SNGs must comply with increased transparency requirements

Table 5 shows that these extra resources are provided by the central government mostly with no conditions attached (15 out of 21 have no conditions) but extra accountability is required in most cases (17 out of 21). Conditions were imposed by Australia, Brazil, Finland, Iceland, Norway, and Spain and are further described for Spain and Brazil below:

Brazil: SNGs benefiting from exceptional central government transfers cannot increase the salaries of public workers until the end of 2021.

In Spain, conditions of the Autonomous Communities Financing Fund apply for autonomous communities.

Regarding transparency requirements, a little over half of responding countries have set up or are planning to set up ad-hoc reporting mechanisms on how the additional COVID relief funds are being used by SNGs. Austria and Colombia are designing the reporting scheme, and information was not yet available when answering the survey.

The Italian central government has set up technical tables to monitor the effects of the COVID-19 emergency on the adequacy of revenues to expenditure needs for SNGs (municipalities, provinces, metropolitan cities and regions).

Table 5. Conditions and transparency requirements for benefitting from CG financial support

Source: Based on the COVID-19 Rapid Survey held by the Network in June 2020. Only countries that answered this question are shown.

There are three main reasons for countries not setting up special reporting systems:

Luxembourg: Due to the small size of the country, crisis management is centrally organised.

Korea: Existing reporting systems for SNG expenditure are sufficient to ensure the sound spending of the exceptional COVID relief funds.

Norway and Finland: Exceptional funds are transferred as general purpose grants (not earmarked for the COVID response).

In more detail these countries did the following/have the following transparency scheme in place:

In Korea, the transparency of funds allocated to SNGs is ensured by the existing accounting standards and financial management information system. Relief funds are spent by designated credit cards, and information on spending patterns (where and on what) is gathered in real time.

Finland did not set up specific reporting obligations or indicators on how the additional funds are used in the local level. Most of the additional funds for 2020 are non-earmarked and are being paid to municipalities on a per capita basis. Some of the additional funds will be paid on application, and the aim is to direct these funds to hospital districts or municipalities to cover for the extra costs due to the COVID-19.

In Norway, additional funds are provided based on preliminary estimates of additional spending needs and/or loss of revenue, which is mainly based on reported figures from local governments. The additional funds are non-earmarked, municipalities and counties can prioritise their revenues and allocate fund according to their needs.

Only 6 out of 18 countries mentioned they intend to evaluate the performance of the exceptional COVID relief funds ex-post. However, 5 out of these 6 countries have special reporting in place (Australia, Canada, Japan, US and South Africa). Australia and the United States were the only countries that further detailed the transparency oversight framework.

Australia: Reporting requirements on additional health funding for the purposes of responding to COVID-19, to confirm funding is being utilised on the pandemic response.

In the United States, transparency/oversight will be done through the legislative branch (oversight committees) and internal ombudsmen within the Federal government (agency Inspectors General). In the latter case, there is a specific Pandemic Response Accountability Committee (https://pandemic.oversight.gov/).

 5. Conclusion: A three-phase reaction to the Covid-19 crisis

Today, all players (central and subnational governments, as well as the private sector) are taking emergency measures to save lives and protect employment and citizens. In this first moment (the “emergency phase”), coordination mechanisms are essential, as is delineating properly the activities to tackle the COVID-19 outbreak across levels of government and ensuring sufficient liquidity to avoid bankruptcies and pro-cyclical policies (see Box 4, below). Although some countries are already leaving this first phase, the possibility of a second wave should be carefully monitored and extraordinary measures might have to be taken again in order to avoid it. Hopefully, lessons learnt from the first wave would help countries and SNGs cope better with a possible second wave.

As the pandemic gets under control, the key priority will shift to addressing the economic impact of the crisis (Figure 27) in order to bring the economy as fast as possible to normality (the “recovery phase”). Current OECD projections show that the initial impact of containment measures may result in losses of up to 30% of GDP. To address this, the OECD is calling for a “Global Marshall Plan”, supporting workers and individuals, and keeping businesses afloat. The experience from the 2008-09 financial crisis shows that it is essential to ensure that SNGs are not forced to carry out pro-cyclical policies, which could worsen the economic crisis. SNGs are key actors for public investment, and a large employer. They should therefore be key actors in the medium-term, to help rebuild our economies through the channels of public investment, social protection and business support, among others.

In the longer-term, as economies recover, fiscal deficits accumulated throughout the first two phases will be reflected in higher indebtedness that, at some moment, will have to be addressed in order for governments to remain fiscally sustainable. Therefore, SNGs will need to participate in national consolidation policies (the “fiscal consolidation phase”). Escape clauses which are justified in present circumstances will need to be replaced by post-pandemic rules, when economic conditions allow, with a necessary correction within an adequately lengthy period.

Finally, keeping the lessons of the COVID-19 crisis in mind, inter-governmental fiscal relations can be permanently improved through the adoption of successful measures that were employed to tackle the outbreak, potentially leading to more resilient policies.

Figure 27. The role of SNGs now, in the short and medium-term to address the Covid-19 crisis
Box 4. Guidance when considering next steps
Fiscal capacity measures (Type I)
  • Don’t hold back: SNGs may not have the necessary liquidity to cope with the crisis and it is critical to support them in fulfilling their mandates during the pandemic.

  • Fiscal support has been provided in the form of inter-governmental grants, lifting of fiscal rules, guarantees or loans from CGs, or facilitated access to external financing.

  • Fiscal equalisation funds are expected to fall or remain constant in the context of increasing asymmetries from the COVID-19 outbreak – additional support to the more severely affected regions might be necessary to fill the gap.

  • Additional transparency requirement may facilitate tracking how the funds are being spent and monitoring SNGs situation, which can help provide support to regions according to their needs.

  • Keep lessons of the past in mind: The 2008-09 financial crisis provides a point of reference when analysing the extent and impact of policy responses to the pandemic.

Delineation measures (Type II), and Monitoring and Coordination measures (Type III)
  • Coordination is essential: The virus respects neither national nor subnational borders, underscoring the importance of coordination across levels of government. Governments must both coordinate vertically and adapt responses to regional needs.

  • The re-orientation of existing or the creation of new institutions involving experts and different levels of government can be put into play in order to design and implement coherent and consistent policy solutions with a well-established role for each body and level of government.

  • Frequent communication between, and data sharing across, regions and levels of government can play an important role to obtain broad enough view of the situation to design better policies.

  • Overlapping and uncoordinated activities can be avoided through the use of enhanced communication mechanisms and guidelines.

  • A phased response is key. Ultimately, the policy response will require multiple phases meaning that it is important to start considering next steps now.

  • Variation allows for benchmarking. Asymmetrical impacts between and within countries highlights the importance of differentiated policy responses. It also creates rich possibilities for cross-sectional benchmarking as different regions within countries move towards a gradual exit from the virus-related restrictions.

Positive developments should be encouraged to persist. The positive changes brought about in response to COVID-19, such as enhanced inter-governmental coordination, should be retained even as the pandemic subsides.


Ahrend, R., Curto-Grau, M., & Vammalle, C. (2013), “Passing the Buck? Central and Sub-national Governments in Times of Fiscal Stress”, OECD Regional Development Working Papers, No. 2013/05.

Allain-Dupré, D., C. Hulbert and C. Vammalle (2012), “Public investment at central and subnational levels: an adjustment variable for OECD countries in the present context of austerity?”, OECD, www.archive.org.

Beazley, I., S. Dougherty, C. James, C. Penn and L. Phillips (2019), “Decentralisation and performance measurement systems in health care”, OECD Working Papers on Fiscal Federalism, No. 28.

Blöchliger, H., C. Charbit, J-M Piñero Campos and C. Vammalle (2010), “Sub-central Governments and the Economic Crisis: Impact and Policy Responses”, OECD Economics Dept. Working Papers, No. 752.

Chen, S., D. Igan, N. Pierri and A. Presbitero (2020), “The economic impact of Covid-19 in Europe and the US”, VoxEU, https://voxeu.org/article/economic-impact-covid-19-europe-and-us.

Colin, C. and B. Brys (2019), “Population ageing and sub-central governments: long-term fiscal challenges and tax policy reform options”, OECD Working Papers on Fiscal Federalism, No. 30.

Dougherty, S., L. Lorenzoni, A. Marino and F. Murtin (2019a), “The impact of decentralisation on the performance of health systems: A non-linear relationship”, Working Papers on Fiscal Federalism, No. 27.

Dougherty, S., Harding, M. and Reschovsky, A. (2019b). “Twenty years of tax autonomy across levels of government: measurement and applications”, OECD Working Papers on Fiscal Federalism, No. 29.

Dougherty, S. and L. Phillips (2019). “The spending power of subnational decision makers across five policy sectors”, OECD Working Papers on Fiscal Federalism, No. 25.

Forman, K, S. Dougherty and H. Blöchlinger (2020), “Synthesising Best Practices in Fiscal Federalism: Key Principles from 15 Years of Country Surveys”, OECD Economic Policy Papers, No. 28.

Herold, K. (2018), “Insolvency Frameworks for Subnational Governments”, OECD Working Papers on Fiscal Federalism, No. 23.

OECD/Korea Institute of Public Finance (2012), Institutional and Financial Relations across Levels of Government, OECD Fiscal Federalism Studies, OECD Publishing, Paris.

OECD (2011), Making the Most of Public Investment in a Tight Fiscal Environment: Multi-level Governance Lessons from the Crisis, OECD Multi-level Governance Studies, OECD Publishing, Paris.

OECD (2013), Fiscal Federalism 2014, OECD Publishing, Paris.

OECD (2014), OECD Economic Surveys: Australia 2014, OECD Publishing.

OECD (2019), OECD Journal on Budgeting: Special Issue on Health, OECD Publications, Paris.

OECD (2020a), “New OECD outlook on the global economy”, Tackling Coronavirus (COVID-19), www.oecd.org/coronavirus/.

OECD (2020b), “The Territorial Impact of COVID-19: Managing the Crisis across Levels of Government”, Tackling Coronavirus (CONVID-19).

OECD (2020c), “Tax and Fiscal Policy in Response to the Coronavirus Crisis: Strengthening Confidence and Resilience”, Tackling Coronavirus (CONVID-19).

OECD (2020d), “Beyond Containment: Health systems responses to COVID-19 in the OECD”, OECD Tackling Coronavirus (COVID-19).

OECD (2020e), Initial Budget and Public Management Responses to the Coronavirus (COVID-19) Pandemic in OECD Countries”, OECD Tackling Coronavirus (COVID-19).

Price, R., T. Dang, J. Botev (2015), “Adjusting Fiscal Balances for the Business Cycle: New Tax and Expenditure Elasticity Estimates for OECD Countries”, OECD Economics Working Papers, No. 1275.

Vammalle, C. (2012), “Recent trends in subnational finances: limited room for manoeuvre”, OECD Workshop on effective public investment at subnational level in times of fiscal constraints, OECD, Paris.


Sean DOUGHERTY (✉ [email protected])

Camila VAMMALLE (✉ [email protected])

Pietrangelo DE BIASE (✉ [email protected])

Kass FORMAN (✉ [email protected])



The edian based on the sample of countries that answered the question which was 39%.


It is worth noting that in no way does those conclusions imply that a centralised response is better than a decentralised one since neither the severity of the challenges nor the benefits of the responses are being considered. For more on that please section 3 on delineation of activities.


This is the case for Australia, in which primary care was allocated to the Commonwealth while States retained responsibility for public hospitals (OECD, 2014).


Estimate based on how often more than one box was ticked in a questionnaire regarding activities done by level of government (for more details see Dougherty and Philips, 2019).


Actually, it even transcends countries boundaries, and, for that reason, international cooperation could be justified by the same reasoning.


For instance, in case SNGs are in charge of enforcing lockdowns, they might choose the level of enforcement to maximize their own benefits. Heavily enforced lockdowns might damage more the local economy but might reduce the health costs and, thus, could be ideal for SNGs that are fully financially compensated by the central government and that bear health costs. The opposite is also true: SNGs that do not bear health costs and that are not compensated by central governments could simply enforce weak lockdowns in a manner to transfer the health costs to the central government.


In countries in which SNGs enjoy more tax autonomy, they commonly are more indebted (Ahrend, Curto-Grau and Vammalle, 2013), which might indicate that they also have a better access to financial markets, dispensing/reducing with the need for central government support.


See Price et al. (2015) for more on elasticities estimates using tax bases in OECD countries.


It is worth noting that not all deferral of taxes that are shared with SNGs are going to impact SNGs finances. For instance, in case the deferred tax is transferred to SNGs only after the deferral end date, SNGs might not face any reduction in revenue.


That is because the solvency is generally assessed by the budget constrain framework, which in theory requires a perpetual forecast of the fiscal balance of a government. Since perpetual forecasts are impossible to be made in a credible manner, it also is impossible to assess fiscal sustainability in practice. Moreover, forecasts also depend on many assumptions regarding the future that cannot be defined deterministically. As a result, subjective assumptions are necessary to overcome this difficulty, leading to an assessment that is, to some extent, subjective.


It is worth noting that when central governments rescue SNGs they may generate an incentive for SNGs to overspend or undertax local citizens or companies. Thus, although in theory the central government can rescue SNGs, there might be nasty consequences of unjustified rescues. In case SNGs do become insolvent, one interesting solution is to have an insolvency framework that can deal with that insolvency in a manner that neither damage the central government credibility nor the regional economy, while securing the payment of the creditors (see Herold, 2018).


Allain-Dupré, Hulbert and Vammalle (2012).


This was for example the case in some countries during the 2008-09 crisis (Vammalle, 2010).


Such as the States Participation Fund, which is computed partially based on state-level per capita income.


At the European level, the European Commission proposed on the 27 March to activate the general escape clause of the Stability and Growth Pact as part of its strategy to respond quickly, forcefully and in a coordinated manner to the COVID-19 pandemic (http://ec.europa.eu).


Kommunalbanken is a state-owned limited company, and the largest provider of debt financing to Norwegian municipalities.