6. Can subnational accounting give an early warning of fiscal risks?

Subnational governments provide crucial public services in OECD countries at the state and local levels. They are usually responsible for local infrastructure, including roads and water supply, and they are often responsible for education, health care and other social services. In federations and decentralised unitary countries such as the Nordics, they may account for as much as 50% of government spending (Figure 6.1, Panel A). Their share in government debt is usually lower, but subnational debt is nevertheless more than 10% of GDP in about the half the OECD countries for which data are available (Figure 6.1, Panel B).1 Data on subnational assets are less widely available, but subnational governments’ role in the provision of infrastructure suggests they are likely to own a large share of public non-financial assets in many countries. In Australia, to take one example, they own 89%.2

Generally, solving subnational fiscal problems before they become crises requires subnationals to prepare high-quality accounts (Figure 6.2). Specifically, the accounts should be up to date, which means that they should be produced reasonably frequently and with a reasonably short time lag. They should be reliable, which means, among other things, that they should be audited. They should be easy to understand, which is more likely if, among other things, they follow a common set of standards. And they should be comprehensive. Accounting information on subnational governments’ debt and cash flows is crucial, but early detection of problems is more likely if the accounts include more than this. A subnational balance sheet, for example, may suggest problems in the maintenance and renewal of assets, while data on accrual expenses may show that cash outflows are being controlled only by delaying the payment of bills. Data on liabilities associated with pensions, public-private partnerships, and subnational enterprises may reveal serious fiscal risks even though the conventional direct debt of the subnational government remains modest. Of course, not even the best accounting can reveal all looming subnational fiscal problems: a banking crisis, major natural disaster, or a pandemic can quickly create fiscal problems for subnational governments even with strong finances. Timely, reliable, understandable, and comprehensive accounting should, however, provide early warnings of some subnational financial problems, and also provide information on the extent to which subnationals have the financial strength to weather sudden negative shocks. However, the need for comprehensive, timely and reliable financial reporting for the identification of fiscal risks requires sophisticated accounting standards and practices, which in turn requires capacities and resources for accounting departments and audit institutions that may be difficult to mobilise at subnational level. This leads to a trade-off between cost and quality, which survey results suggest can be handled in a variety of ways.

In certain countries, larger subnational governments’ accounts recognise a broad range of assets and liabilities (including those related to pensions, derivatives and public-private partnerships), but smaller subnational governments are allowed to produce fewer disclosures in their financial statements. Some countries make timely monitoring easier by having subnational governments report on a quarterly or even a monthly basis, but only for headline financial indicators. Concerning external control, while some flexibility may be granted with respect to the modalities for the audit of individual accounts, the supreme audit institution may be tasked with the preparation of a summary report on the financial situation of subnational governments and their annual financial audits, allowing the identification of any widespread problems.

Although systematic data on subnational financial problems are not available, examples are easy to find. Spain’s fiscal problems during the European debt crisis of the early 2010s were compounded by those of the autonomous communities of Andalucía, Catalonia and Valencia (Jenker and Lu, 2014[1]). In the United States, the financial crisis caused Illinois to suffer severe financial difficulties, and the city of Detroit to file for bankruptcy. In Brazil, the State of Rio de Janeiro declared a “state of public financial calamity” in 2016, and some other states have since followed suit (Rodrigues, 2019[2]). Similar problems have arisen, at various times and in varying degrees of severity, in many OECD countries (Allers, 2014[3]; Baskaran, 2014[4]; von Hagen et al., 2000[5]; UK National Audit Office, 2018[6]; Herold, 2018[7]; Herold, 2020[8]). It would not be surprising if the coronavirus pandemic of 2020 also caused serious financial problems for some subnational governments (OECD, 2020[9]).

The importance of subnational governments in many countries means that subnational fiscal problems can have serious repercussions not only for local public services but also for national public finances. For one thing, the national government may feel obliged to step in to remedy the deterioration of public services. It may also have guaranteed the debts of subnational governments. Or, it may simply judge that the national costs of allowing subnational governments to default on their debt are greater than the costs of bailing them out.

National governments can try to mitigate the fiscal risks of subnational finances in at least two main ways (Figure 6.2) (OECD/KIPF, 2016[10]; Herold, 2018[7]). They can disclaim responsibility for subnational debts and thus minimise moral hazard and encourage citizens, journalists, researchers, lenders, credit-rating agencies and others to monitor subnational finances (the “no-bailout strategy”). Alternatively, they can establish fiscal rules and other constraints that aim to prevent subnationals from accumulating too much debt or otherwise taking on excessive risks and then monitor the subnationals to check compliance with the rules and to look for signs of trouble even when rules are not broken (“the control-and-monitor strategy”) (Sutherland, Price and Joumard, 2006[11]). Examples of the control-and-monitor strategy in practice are given in Figure 6.2, and illustrate how such a strategy may be implemented by central government following subnational financial distress.

Other strategies are possible, though some run the risk of creating moral hazard without imposing controls strong enough to prevent excessive risk-taking. In addition, the chosen strategy is sometimes complemented by the creation of an insolvency regime for subnationals to reduce the problems created by subnational failures (Herold (2020[8]), and Chapter 8 in this volume).

Having uniform standards is easier in unitary countries. Of the 24 survey respondents classified here as unitary countries,3 10 have a common set of standards for all governments including the national government, and 12 have a common set of standards for subnationals that differs from the standards followed by the national government (Table 6.1). South Africa is distinctive in that the national government and the provinces follow one set of standards, while municipalities follow a different set.

Surprisingly, a high degree of standardisation occurs also in the nine federations surveyed. Four of them have common standards for national and subnational governments, and two have harmonised standards for subnationals that are different from the standards for the national government. In most cases, these common standards are adopted on a voluntary basis, and compliance with them may not be universal. Two federations in which there is less standardisation are Belgium and Germany.

Common standards make information easier to understand and compare, but do not ensure perfect comparability. Sometimes accounting standards allow for more than one option in the treatment of an issue, and even when there is only one option there may be different ways to interpret the standard. Perhaps most important, compliance with the standards may be imperfect, as reportedly in Brazil and Mexico.

The means for defining common standards vary between unitary countries and federations (Figure 6.3). In unitary countries, the national government often sets accounting standards for subnationals, with the responsibility typically attributed to the Ministry of Finance (9 countries) or the Ministry of Interior (7 countries). Ministries of finance are more likely to be involved if there is one set of standards for all governments. In three unitary countries – Iceland, Norway and Sweden – standard-setting responsibility lies with an independent body that includes representatives of national government, local-government associations and sometimes auditors.

In most federations, independent standard-setters establish standards that can be adopted by subnational governments. In Australia, the commonwealth government and the states have agreed to follow standards based on International Financial Reporting Standards (IFRS) set by the Australian Accounting Standards Board (an independent government entity). A similar mechanism exists in Canada. In Switzerland, a set of accounting standards was developed co-operatively and is “widely applied” by cantons according to the questionnaire response. In the United States, many subnational governments follow the standards set by the (private) Governmental Accounting Standards Board. Similarly, the national government in Austria (in particular the Ministry of Finance in agreement with the Court of Auditors) has the authority to set accounting standards for subnational governments based on International Public Sector Accounting Standards (IPSAS).

Accounting standards are generally classified as “cash basis” or “accrual basis” (Moretti and Youngberry, 2018[13]). Cash-based accounts are easier to prepare because they are simpler and require fewer judgements and estimates. That means they are also easier to understand and less vulnerable to manipulation by means of dubious judgments or optimistic estimates. Reliable cash-based accounts and measures of conventional debt also provide much of the information needed for monitoring subnational finances. They show whether a subnational government is generating enough cash to service its debts, and they allow the national government to establish rules that limit subnational deficits and debt. Where cash accounting also requires the recording of contractual commitments, information on committed future cash outflows is also available.

However, having more information than is provided by traditional cash accounts should allow subnational fiscal problems to be detected earlier. Sooner or later, most fiscal problems show up in cash deficits and in difficulties repaying outstanding debt, but problems may build up well before these things happen. To take one example, a struggling subnational government may respond to declining revenues by delaying payments to suppliers, as happened during the European debt crisis in countries such as Italy and Spain. For a while, the struggling subnational’s cash accounts may remain in balance, thereby concealing the looming problem. Accrual-based accounts will reveal the problem earlier since they record spending when the government is invoiced (for instance), and not when it pays.

The merits of accrual accounting in providing a more comprehensive picture of the finances of local government are recognised in most OECD countries (Figure 6.4). In a large majority of the countries surveyed, subnational accounts prepared on an accrual basis or subnationals are transitioning to the accrual basis (e.g. Brazil). In addition, in those countries where subnational governments prepare their main accounts on a cash basis, complementary accrual financial statements may be prepared on mandatory (e.g. Portugal) or voluntary (e.g. Japan) bases.

There are some exceptions. Israel, Luxembourg and Norway use a mix of accrual and cash concepts. This means in most cases that expenses are recognised on accrual basis, but the depreciation of assets is not included in the annual profit and loss. In other countries that do not have harmonised accounting standards for subnationals, the type of accounting basis used depends on the subnational government. In Belgium, for example, Brussels, Flanders and the German-speaking community have accrual accounts, but Wallonia does not. In Germany, Hesse and Hamburg and most local governments have accrual accounts, but most Länder and some local governments do not.

In most of the countries that use the accrual basis for subnational accounts (Table 6.2), the transition took place more than a decade ago – that is before the 2008 financial crisis. The adoption of accrual standards did not prevent a series of recent problems with subnational government finances. In some cases, the problems may have been hard to prevent with any kind of accounting because the main cause was a deep recession and a credit crunch. In other cases, the quality of the accounting standards may have been part of the problems. For example, in Portugal or Greece, subnationals were required to move from cash to accrual accounting in the wake of the financial crisis. Case studies show that some countries have recently strengthened the accrual-accounting requirements for subnationals (Brazil, France and Iceland).

The debate on characteristics for “quality” of accrual accounting standards in the case of the public sector is not settled. However, ensuring that subnationals follow standards that are established by an independent standard-setter with appropriate technical capacities or are aligned with recognised international accounting standards can go a long way towards ensuring that their accounts provide comprehensive and relevant information.

Countries in which accrual-based standards for subnationals were adopted before the 1990s often based those standards on national standards for the private sector (Figure 6.5). Later, a few countries adopted standards based on international standards for the private sector, namely International Financial Reporting Standards (IFRS) (e.g. Australia and the United Kingdom). Recently, it has become more common to adopt standards based on International Public Sector Accounting Standards (IPSAS) (Brazil, Slovak Republic and Lithuania). Some early adopters of accrual accounting have also started aligning their framework with IPSAS (e.g. Latvia).

Countries that use IFRS or IPSAS as a reference for their national framework sometimes mention adaptations or exceptions, due for example to the mismatch between the requirements of the standards and available resources and capacities at subnational level. In Portugal, for example, so-called local “micro-entities” are not required to prepare accrual-basis financial statements in addition to their cash accounts. Similarly, in the United Kingdom, “small entities” have simpler reporting requirements.

Early detection of problems requires timely and reliable accounts. To be timely, accounts must be published or transmitted to the national government on a sufficiently frequent basis and within a reasonably short lag after the end of the period. To ensure reliability, the annual accounts need to be audited. Cash accounts can be manipulated by changing the timing of cash flows without greatly changing the substance of subnational finances. Accrual accounts introduce opportunities to manipulate accounts by making dubious judgments and optimistic estimates. In both kinds of accounts, some form of quality assurance is needed to help detect fraud and ensure that the accounts reflect the rules. However, the importance of estimates and judgments in accrual accounting increases the onus on auditing to maintain the reliability of the accounts.

Annual reporting to national government is mandatory in almost all unitary states. A large majority of countries do not require that this reporting be done with audited accounts. Where audited accounts are required, the time lag for the transmission of the accounts tends indeed to be longer – for example, more than six months (Greece, Israel and Norway).

Although this information was not required in the survey, a number of countries indicated that in addition to the transmission of the annual accounts, subnational governments are required to report some data on a more frequent basis. That is the case for example in Greece, Estonia, Lithuania, Brazil (monthly) and in Iceland and Israel (quarterly).

In most federations, states and local governments are not required to report financial data to the federal government (except for statistical purposes) (Table 6.3). They often have the obligation however, to publish their annual accounts, according to detailed rules specified in national or subnational legislation. The publication of accounts by subnational governments is mandatory in most unitary states. The exceptions are Korea, Luxembourg, France and Norway. In the two latter countries, all citizens have however access to this information upon request.

There are many variations in the auditing of subnational accounts (Figure 6.6). While the audit of the national government’s accounts by the supreme audit institution (SAI) is the norm in the OECD, only a small group of countries require that their SAI also audit subnationals’ accounts each year (7 of 33 countries). In other countries, there are limits on what type of entities the SAI can audit, for constitutional or other reasons. For example, the SAI may have discretion as to how many individual entities to audit each year, in order for them to accommodate other types of audits within their resource constraints.4

The survey answers reveal two dominant practices. In 10 countries, independent external audit is required, and it often involves both the SAI and private audit firms. Under this model, the SAI or a state auditor may provide an audit opinion on the consolidated financial statements of the state, while audit firms may audit the accounts of local governments. In 12 countries, other audit models are used. Audits may be carried out by audit departments, inspections or the like, operating under the auspices of either national or subnational government. In Sweden, politically elected auditors conduct controls and report their findings and recommendations to the audit committee of the municipal council. In Germany, local governments are authorised to self-audit.

Very few countries indicated widespread audit issues with subnational accounts. This is understandable, as discovering such issues could require reviewing hundreds if not thousands of audit reports. In addition, audit qualifications may be difficult to analyse or compare where accounting standards have not been fully harmonised. However, a recent report of the Austrian Court of Audit on the fiscal data of municipalities identified many issues with incomplete data. Two countries indicated that their SAI produces a synthesis of key audit issues identified in the local sector (France and New Zealand). Such reports are useful in that they may help identify common problems and risks that may have gone unnoticed through the usual monitoring mechanisms.

The accounts that subnationals produce are often monitored by national governments. National governments that pursue the control-and-monitor strategy must of course monitor subnational finances, but governments that pursue the no-bailout strategy may also choose to do some monitoring, while being careful to avoid any suggestion that the monitoring implies responsibility.

The approach that national governments take depends in part on their authority over subnational governments, which tends to depend in turn on whether the constitution is federal or unitary. As a rule, national governments in unitary countries have authority over subnational governments and tend to follow the control-and-monitor strategy, while national governments in federations are less likely to have such authority and more likely to follow the no-bailout strategy. Yet there are several exceptions to this rule, and some governments follow strategies that do not fit neatly into either of the two categories.

In the countries surveyed for this study, 18 national governments reported that they had the authority to monitor subnational governments and take steps to remedy their problems (Table 6.4). As might be expected, these countries are almost all unitary. The exception is Spain, which has been described as quasi-federal. In 12 countries, the national government reported that it did not have the authority to monitor subnational governments and take steps to remedy their problems. As might be expected, all the federations apart of Spain fall in this category. However, four unitary states are also in this category: the Czech Republic, Portugal and two Nordic countries that, despite being unitary, have a tradition of autonomous local government - Iceland and Sweden.

The proportion of national governments that monitor subnational finances in some way is higher because some national governments monitor even though they do not have legal authority to intervene to solve subnational fiscal problems (Table 6.5). In particular, 27 of the 33 countries responding to the survey said that the national government monitored the finances of subnationals. As might be expected, there is again a relationship between whether national governments monitor and the constitution of the country. Monitoring is almost universal in unitary countries, and most of the countries in which the national government does not monitor are federal.

One exception to the expected pattern is that national governments reported that they monitored subnational governments in four unambiguously federal countries – Australia, Austria, Brazil and Canada – as well as in Spain. In Brazil, the monitoring includes verifying whether subnationals have complied with fiscal rules set at the national level.

Another exception to the expected pattern is that two national governments of unitary countries, the Netherlands and New Zealand, reported that they did not monitor the finances of subnational governments. In New Zealand, the system was said to be “sufficiently robust that active monitoring is not required”. The response for the Netherlands did not mention the reason for the approach, but Allers (2014[3]) notes that subnational governments in the Netherlands sometimes receive bailouts from the national government and that the fiscal rules governing subnationals are weak or do not bind. Nevertheless, Allers argues, the approach works reasonably well because local politicians want to avoid the loss of autonomy that comes with a bailout.

In the member countries of the European Union, the finances of subnational governments matter to national governments partly because of the countries’ treaty commitments to restrict the debt and deficit of general government (i.e. national and subnational government combined, as defined in macroeconomic statistics). As a result, some monitoring of subnational finances is necessary even if the national government does not have authority over subnational governments. In the federations of Belgium and Germany, the problem has been addressed by a form of co-operative monitoring. In Germany, the accounts of both the national government and the Länder are monitored by the Stability Council (Stabilitätsrat), which was jointly created by the federal government and the Länder. In Belgium, the High Council of Finance performs a similar function.

When monitoring is done by the national government, it is almost always done according to laws or guidelines (Table 6.6). The two exceptions are Canada and Sweden. In unitary countries, it is typically done by the finance ministry or the interior ministry. Often, responsibility for monitoring accompanies responsibility for setting accounting standards (Figure 6.3). Sometimes, both the finance and interior ministries are involved. In France, the Ministry of the Interior monitors local governments’ compliance with a golden rule (current spending cannot exceed revenue), while the Ministry of Economy and Finance looks for warning signs of financial trouble by monitoring subnational governments’ debt, expenditure rigidity, self-financing capacity, and ability to increase local taxes. As these examples illustrate, monitoring may need to go beyond accounting indicators to include the economic and legal context.

Independent fiscal institutions also play a growing role in monitoring subnational finances, as mentioned in the questionnaire responses from Austria, Lithuania, Spain and Sweden. Independent fiscal institutions tend to examine not only subnational governments’ recent fiscal performance, but also their forecast financial performance, providing an important complement to monitoring that looks only at fiscal outcomes. One of their main tasks is typically to report on compliance with fiscal rules, so when the fiscal rules apply to subnationals, as in the European Union, their monitoring of subnationals is especially significant. In Spain, for example, the Independent Authority for Fiscal Responsibility, established in 2013, now plays a major role in monitoring subnational finances (von Trapp et al., 2017[19]).

The term “monitoring” encompasses a broad variety of activities, and the nature of the monitoring varies across countries. In some countries, the main purpose of the monitoring is to determine whether subnational governments have complied with fiscal rules and other legal requirements. If the monitoring suggests non-compliance, the national government is likely to have the ability to intervene. In Israel, for example, local governments’ budgets must be approved by representatives of the national government and quarterly data on budget execution are reviewed by the Ministry of the Interior, which can intervene in the budgets if it deems this to be necessary.

In other countries, such as Australia and Canada, the nature of the monitoring is different. The national government gathers information, but it is not verifying compliance with fiscal rules with a view to possible interventions in the budgets of the subnational governments. In some cases, the objective of the monitoring may be to identify early warning signs of subnational financial problems, and this may be separate from the monitoring of compliance with rules. Since financial problems can show up first in indicators not governed by rules, this monitoring generally looks at a wide range of indicators. In the cases of Australia, Canada or France, it can be said that the ministry of finance performs an advisory function vis-à-vis subnationals, rather than a surveillance one.

The indicators that are most commonly monitored are debt and the cash deficit (Figure 6.7). Subnational governments’ debt is monitored in all 27 countries in which subnational finances are monitored, and the cash deficit is monitored in 21 of those countries. The fact that the cash deficit is not systematically monitored in all the countries is a possible concern, because although the accrual deficit can pick up early warning signs of problems not apparent in the cash deficit, the cash deficit is also informative. Subnational governments need to be generating enough cash to meet debt-service and other requirements, and a large unexplained difference in the two surpluses might be a sign that the accrual surplus is being mis-measured (for example, by over-estimating receivables).

In many countries, indicators available only in accrual accounts are also monitored. The accrual deficit is monitored in 15 countries, and liabilities (defined as including not just ordinary debt but also accounts payable, leases and other items) are monitored in 17 countries. Monitoring a broad measure of liabilities is especially useful since subnational governments may borrow in unconventional ways if conventional debt is closely scrutinised or restricted by fiscal rules. Australia, for example, experienced this when conventional borrowing by the states was controlled by the Loan Council (see Box 6.1, above, and Table 6.8). Under the heading of other monitored indicators, Brazil mentioned pensions, Greece arrears, and Latvia guarantees.

Overall, the survey results are encouraging in the extent to which the monitoring of traditional fiscal indicators is now supplemented by the monitoring of non-traditional indicators. Moreover, almost all survey respondents said that they believed that the data they received was satisfactory. Nevertheless, the results suggest a few possible remaining problems (in addition to the issues discussed above, including the unavailability of accrual indicators in some countries) and some areas where further progress might be sought.

Only 13 countries monitor subnational governments’ net financial worth or net worth. This may prevent identifying early signs of future problems, in relation for example to the losses of entities owned or otherwise controlled by subnational governments. Although serious financial problems in these entities are likely ultimately to show up in the accounts of the subnational government itself, problems may be detected earlier if monitoring encompasses the fair value of the shareholdings as reported in the balance sheet or, of course, if the subnational governments publish accounts that consolidate all controlled entities. Australia and Iceland are examples of countries in which subnationals publish accounts that consolidate controlled entities, as well as showing accounts for a narrower definition of subnational government.5

Another possible problem is that there is little evidence of the monitoring of the subnational governments’ fixed assets. This may make early detection of problems of deteriorating infrastructure more difficult – a problem analysed, for example, in New South Wales.6 Conversely, there may be cases (as perhaps in some subnational governments in Spain before the European debt crisis of the early 2010s) where monitoring would reveal riskily high levels of investment in infrastructure.

Although the survey did not include any question on the preparation of forecasts by subnationals, some answers noted that historical information is necessary but insufficient for good monitoring. It is possible for the accounts of a subnational to be healthy even though its prospects are poor, and conversely, for its prospects to be good even though its accounts suggest that it is in bad shape. What also matters is what will happen to its revenue, whether in the form of transfers from a higher tier of government or in the form of its own taxes, and what will happen to its spending obligations, either because of changes in local conditions or because of changes in its responsibilities. Moreover, the subnational’s ability to respond to changes is also crucial. Can it raise taxes or cut spending, or are its hands tied?

It is useful, therefore, if subnational governments publish fiscal forecasts, discuss the factors that could cause outcomes to differ from the forecasts, and explain their ability to cut spending or raise revenue if outcomes turn out to be worse than forecasts. Some subnational governments provide medium-term forecasts. The State of Victoria in Australia, for example, publishes “forward estimates” for the three years following the budget year for various categories of revenue and spending, along with a discussion of the economic outlook, and the risks surrounding the economic and fiscal forecasts.7 New Zealand’s local governments publish audited 10-year plans that include financial forecasts.

Most respondents said that monitoring compared subnational financial indicators with benchmarks (Table 6.7). Specifically, of the 27 national governments that monitor the finances of the subnational governments, 19 said that they monitored against benchmarks, while only seven said that they did not. Most governments publish the results of their monitoring. The Norwegian government said it did not set benchmarks against which to monitor subnational finances, but that it published financial indicators for local governments and compared them to regional or national averages. In many countries, the benchmarks are just the fiscal rules. Only in a few cases, it seems, is the benchmarking used for analysis of the strength of the finances of subnationals that goes beyond checking compliance with fiscal rules.

The use of benchmarking and comparative analysis is likely to be most useful if there are many subnational governments, as for example in France, where there are more than 36 000 communes. If one tier of subnational government contains only a few entities, as with the states and territories in Australia and the provinces in Canada, paying close attention to the special characteristics of each government is feasible and benchmarking is less valuable (on the benchmarking of non-financial performance, see Chapter 4 and Phillips (2018[20])).

The more subnational governments there are, the more attractive is statistical analysis. If there are many subnationals and sufficient time series data on their finances, circumstances and financial problems, it may be possible to develop a statistical model that identifies which indicators have best predicted subnational fiscal problems in the past and to use the model to help predict future problems.

If there are many subnational governments, the form in which their accounts are published or submitted to the national government starts to matter. If there are only a few subnationals, officials in the national government can read the accounts in a form such as a pdf file and enter data as needed into their own files. If there are many subnationals, however, monitoring is much easier if the accounts are submitted or published in a way that facilitates the downloading of the financial data from each subnational.

If there are only a few subnational governments or only a few years of historical data, developing a reliable model may not be possible. More use must then be made of analysts’ judgments. Even in this case, however, it may be helpful to ask expert analysts to assign weights to various indicators as predictors of financial trouble and then to use a spreadsheet model to calculate for each subnational an index of risk that employs those weights. It may also be possible to make use of estimates made in other contexts, such as estimates of the links between credit ratings and debt defaults. If the subnationals have credit ratings, it may be possible to apply these estimates directly. In other cases, it may be possible to estimate the credit ratings they would receive.

Categorising subnationals by apparent risk can help identify which ones merit closer scrutiny. Even if there are too many subnationals for close analysis of the finances of each one to be feasible, a model can be used to identify those that seem most at risk. Then analysts can look more closely at the finances of the smaller group to determine whether they are actually performing reasonably or whether remedial action should be taken.

Finally, national governments pursuing the control-and-monitor strategy can enlist the help of analysts outside government by requiring subnationals to publish their accounts. This allows citizens, journalists, and others to investigate subnational finances if they wish to do so. The justification of fiscal decentralisation is that local citizens are often more knowledgeable and more concerned about what is happening in their area than are officials in the national government. For the same reason, local citizens, journalists, and others may sometimes be better at detecting local problems than is the national government.


[3] Allers, M. (2014), “The Dutch local government bailout puzzle”, Public Administration, Vol. 93, pp. 451–470.

[16] Angelo, P. (2013), Understanding the valuation of public pension liabilities: Expected cost versus market price, American Enterprise Institute.

[4] Baskaran, T. (2014), “Bailouts and austerity”, CEGE Discussion Papers, No. 212, University of Göttingen, Center for European, Governance and Economic Development Research.

[21] Dougherty, S. and L. Phillips (2019), “The spending power of sub-national decision makers across five policy sectors”, OECD Working Papers on Fiscal Federalism, No. 25, https://dx.doi.org/10.1787/8955021f-en.

[17] GAO (2014), Pension Plan Valuation: Views on Using Multiple Measures to Offer a More Complete Financial Picture, The Government Accountability Office, http://www.gao.gov/assets/670/666287.pdf.

[8] Herold, K. (2020), “Insolvency Frameworks for State and Local Governments”, OECD Journal on Budgeting, https://dx.doi.org/10.1787/4fe1859f-en.

[7] Herold, K. (2018), “Insolvency Frameworks for Sub-national Governments”, OECD Working Papers on Fiscal Federalism, No. 23, OECD Publishing, Paris, https://dx.doi.org/10.1787/f9874122-en.

[18] IMF (2018), Austria: Fiscal Transparency Evaluation, International Monetary Fund, Washington, D.C.

[12] Irwin, T. and D. Moretti (2020), “Can subnational accounting give an early warning of fiscal risks?”, OECD Journal on Budgeting, https://dx.doi.org/10.1787/be73a937-en.

[1] Jenker, E. and Z. Lu (2014), Sub-national credit risk and sovereign bailouts – Who pays the premium?, WP/14/20.

[13] Moretti, D. and T. Youngberry (2018), “Getting added value out of accruals reforms”, OECD Journal on Budgeting, https://dx.doi.org/10.1787/budget-18-5j8l804hpvmt.

[15] Novy-Marx, R. and J. Rauh (2009), “The liabilities and risks of state-sponsored pension plans”, Journal of Economic Perspectives, Vol. 23/4, pp. 191–210.

[9] OECD (2020), “COVID-19 and fiscal relations across levels of government”, Coronavirus Policy Briefs, http://www.oecd.org/coronavirus/.

[10] OECD/KIPF (2016), Fiscal Federalism 2016: Making Decentralisation Work, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264254053-en.

[20] Phillips, L. (2018), “Improving the Performance of Sub-national Governments through Benchmarking and Performance Reporting”, OECD Working Papers on Fiscal Federalism, No. 22, OECD Publishing, Paris, https://dx.doi.org/10.1787/ffff92c6-en.

[2] Rodrigues, E. (2019), “Seis Estados já declararam situação de calamidade financeira”, O Estado de S. Paulo, https://economia.estadao.com.br/noticias/geral,seis-estados-ja-declararam-situacao-de-calamidade-financeira,70002683568.

[14] Sanderson, R., G. Dinmore and G. Tett (2010), “Finance: An exposed position”, Financial Times, http://www.ft.com/content/0d29fbdc-2aef-11df-886b-00144feabdc0.

[11] Sutherland, D., R. Price and I. Joumard (2006), “Fiscal rules for sub-central governments: Design and impact”, OECD Working Papers on Fiscal Federalism, No. 1, https://dx.doi.org/10.1787/3e6551ae-en.

[6] UK National Audit Office (2018), Financial Sustainability of Local Authorities 2018, http://www.nao.org.uk/wp-content/uploads/2018/03/Financial-sustainabilty-of-local-authorites-2018-Summary.pdf.

[5] von Hagen, J. et al. (2000), “Subnational Government Bailouts in OECD Countries: Four Case Studies”, Inter-American Development Bank Research Network Working Papers, R-399.

[19] von Trapp, L. et al. (2017), Review of the Independent Authority for Fiscal Responsibility (AIRef), OECD, Paris.


← 1. The OECD Fiscal Decentralisation database is available at http://oe.cd/FDdb.

← 2. Australian Bureau of Statistics, Government Finance Statistics 2017-18 (estimate excludes non-financial assets not allocated to a particular jurisdiction).

← 3. Whether a country should be classified as federal or unitary is sometimes unclear. Borderline cases include Spain and South Africa. Following Dougherty and Phillips (2019[21]), this study classifies Spain as federal and South Africa as unitary.

← 4. Audits conducted by SAIs may include regularity audits (encompassing compliance and financial audits), performance or value for money audits, environmental audits, and forensic audits.

← 5. See annex of Irwin and Moretti (2020[12]).

← 6. See annex of Irwin and Moretti (2020[12]).

← 7. See Victorian Budget 19/20, Strategy and Outlook: Budget Paper 2, available at https://s3-ap-southeast-2.amazonaws.com/budgetfiles201920.budget.vic.gov.au/2019-20+State+Budget+-+Strategy+and+Outlook.pdf (accessed 11 July 2019).

Metadata, Legal and Rights

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

© OECD 2021

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at https://www.oecd.org/termsandconditions.