Chapter 2. Fiscal constitutions

Fiscal constitutions comprise the sets of rules and frameworks that guide a country’s fiscal policy and are enshrined in its fundamental laws. This chapter compares the fiscal constitutions of 15 federal countries by empirically assessing frameworks of intergovernmental relations. It looks at such aspects as the responsibility of sub-national governments for their own policies, their power to shape fiscal policy at the federal level, the strength of intergovernmental budget frameworks, and the stability of fiscal policy arrangements. The chapter also gives a detailed account of how fiscal constitutions evolved between 1917 and 2012, describes historical turning points, and identifies potential drivers of constitutional reform. It then assesses the link between constitutional frameworks and fiscal outcomes, suggesting that the coherence of constitutional arrangements – i.e. the way in which fiscal constitution elements and building blocks fit together – is a crucial factor in the long-term sustainability of public finances. Finally, the chapter suggests a number of policy reform options for making fiscal constitutions more coherent.

Fiscal constitutions: Frameworks for fiscal policy

A fiscal constitution is the body of fundamental rules and regulations that frame decision making in the area of fiscal policy. Fiscal constitutions cover constitutional law as well as selected post-constitutional legislation like basic fiscal and financial laws or constitutional court rulings. A fiscal constitution thus encompasses all legislation that may be amended by stringent voting rules – usually qualified majority – to offer a stable institutional framework for fiscal policy over time. It sets the rules of the public finance game, so providing a framework for policy makers and driving or deterring certain policy patterns (Brennan and Buchanan, 1980). By shaping incentives and limiting arbitrariness, the fiscal constitution determines the course of fiscal policy and outcomes in the long term (Qian and Weingast, 1997). A simplified – albeit somewhat mechanical – causal chain from fiscal constitutions to fiscal outcomes is shown in Figure 2.1.

This chapter focuses on the fiscal constitutions of federal and quasi-federal countries and on how they govern intergovernmental fiscal relations. In federal countries, the fiscal constitution is predominantly concerned with the rules that determine power sharing between the federal government and sub-national jurisdictions at state/regional levels. Federal fiscal constitutions must define the fiscal authority of all tiers of government and how they interact in the conduct of fiscal policy. Moreover, federations often inspire constitutional reform in other places – either in countries that are on a secular path towards decentralisation, or in supra-national entities seeking to strengthen their basic framework. In these entities, almost any potential fiscal policy question has a “who should do what” or “federal” dimension, and fiscal arrangements in federal countries may show the way forward. Altogether, the chapter analyses 15 federations or quasi-federations – Argentina, Australia, Austria, Belgium, Brazil, Canada, Germany, India, Italy, Mexico, Russia, South Africa, Spain, Switzerland and the United States. The fiscal constitution of the European Union, although not a federation, is also assessed.

Assessing the fiscal constitutions of 15 federations

The building blocks of fiscal constitutions

Every fiscal constitution consists of a number of building blocks which together reflect the institutional background of fiscal policy making across all tiers of government. Five building blocks may be distinguished which, in turn, comprise a series of constituent elements (Table 2.1). A constituent element represents a constitutional rule on a specific item of fiscal policy. For instance, “tax autonomy” of the states is a constituent element, while “autonomy” is the building block encompassing tax, spending, borrowing and budgeting autonomy of the states.1 And while numerical fiscal rules are a single constituent element of the budget framework, the framework also includes procedural rules and the functioning of fiscal councils. Some constitutions also incorporate rules for local governments, which are taken into account, if relevant.

Table 2.1. The building blocks of fiscal constitutions
Building block or arrangement	Description	Constituent elements
Autonomy	Extent to which states can conduct their own fiscal policy.	Tax autonomy; spending autonomy in certain policy areas; autonomy to borrow; autonomy to set budget frameworks.
Responsibility	Extent to which states are exposed to budget constraints and must assume responsibility for their own fiscal policy.	Bankruptcy exposure; bailout expectations; responsibility for setting fiscal rules; state revenue mix; dependence on revenue from federal transfers.
Co-determination	Extent to which states can shape fiscal policy at the federal level.	Bicameralism; constitutional courts; intergovernmental executive bodies and meetings; federal transfers.
Budget frameworks	Constraints on discretionary fiscal policymaking at all government levels.	Numerical fiscal rules; procedural fiscal rules; fiscal councils and other independent or arms-length bodies.
Stability	Ease with which constitutional rules affecting fiscal policy can be amended.	Strength of second chamber; power of constitutional courts; the voting rules needed to amend the constitution; scope of direct democracy/popular veto.

Fiscal constitutions, their building blocks and constituent elements are assessed by means of institutional indicators which, together, form an indicator tree (Figure 2.2). Each element – e.g. the extent of state tax autonomy – is measured by a low-level indicator (LLI). The LLIs are then aggregated into intermediate-level indicators (ILIs) which measure the building blocks. For example, one ILI denotes the building block that brings together the extent of states’ tax, spending, borrowing and debt autonomy. The ILIs are, in turn, aggregated to form a summary, or composite, indicator which gauges overall intergovernmental fiscal relations in the fiscal constitution. Indicator values, which range from 0 to 1, measure whether the fiscal constitution features “more” or “less” of a certain element or building block and range from 0 to 1. The coding procedure and values are shown in detail in Blöchliger and Kantorowicz (2015).

The coherence of fiscal constitutions

The coherence of a fiscal constitution is of particular importance. A coherent fiscal constitution combines institutional arrangements in a balanced manner giving states similar degrees of autonomy in various budget items (taxation, spending, borrowing etc.), for example. Similarly, it aligns a certain level of autonomy with a matching level of responsibility, or it strikes a balance between numerical and procedural fiscal rules. Conversely, a less coherent fiscal constitution combines elements and building blocks in an unbalanced way, associating extensive fiscal autonomy with little fiscal responsibility, for example, or strict numerical rules with lenient procedural rules. Coherence is measured as the variance in aggregated indicator values yielded by the random weight technique.

Fiscal constitutions from country to country

This section uses the indicator methodology described above to assess and compare the fiscal constitutions of the 15 federal countries. It considers numerous country examples to complement the empirical exercise and paint a lively, detailed picture of the world of fiscal constitutions.

Autonomy of the states

Fiscal constitutions give states very different degrees of autonomy. In some federations, SCGs are a de facto branch of the federal level, while in others they enjoy wide fiscal autonomy and little interference from federal government. Coherence also varies. While some countries enjoy similar degrees of autonomy in all areas of fiscal policy, others combine considerable spending and borrowing autonomy with little tax autonomy.

Tax autonomy

Tax autonomy varies wildly across federal countries, although constitutionally guaranteed autonomy in levying taxes is seen as an integral part of federalism (Riker, 1975). Some countries grant federal government the power to levy and collect only a few taxes, sometimes combining that power with a residual clause that leaves all other powers of taxation to the states/regions. In other federations, taxation is largely a federal prerogative, often with a general clause allowing central government to change the tax system by ordinary legislation. Some constitutions are very precise about how and which taxing powers are assigned to different levels of government. Others, by contrast, are vague or simply silent.2 In a few countries, post-constitutional legislation and constitutional courts play a crucial role in shaping tax autonomy.

A few country examples illustrate the wide institutional differences in tax autonomy. Austria’s and Italy’s constitutional law assigns most taxing powers to the federal government. There is slightly greater tax autonomy in Russia, where oblasts may reduce the rate of corporate income tax for certain taxpayers. By contrast, the federal government in Switzerland is allowed to levy only those taxes listed in the constitution, and any change in personal income tax rates requires a constitutional amendment. In some countries, constitutional voids are filled by legal interpretation. The Canadian constitution grants power to provinces with respect to natural resource and direct taxes. A creative interpretation of “direct taxes” allows the inclusion of sales taxes, so both federal and provincial tiers of government may concurrently levy all major taxes. As for Australia, judicial review has strongly shaped taxing powers. The constitutional court ruled that all consumption taxes should be considered “trade taxes and excises” – the only levies constitutionally assigned to federal government. The General Sales Tax (Australia’s VAT) is thus a federal tax. Finally, constitutional tax autonomy in Mexico is undermined by a constitutional provision that makes it very costly for states to actually make use of their tax autonomy.3

Spending autonomy and links between tax and spending autonomy

Spending autonomy captures the extent to which states have sovereign powers of legislation and expenditure in certain policy areas.4 Constitutional spending autonomy varies less than tax autonomy, since there is a core of public functions which, in most federations, is assigned in a similar way. Fiscal constitutions assign legislative and spending authority either exclusively – i.e. to only one tier of government – or jointly, i.e. to both federal government and states. The Austrian and Italian fiscal constitutions grant most spending prerogatives to the federal level, while the provinces and states enjoy extensive autonomy in Canada and the United States, respectively. In the United States, the constitution does not prevent states from spending in any area as long as they do not violate the Commerce Clause.5

Other constitutional provisions further affect the extent and coherence of tax and spending powers:

Rather than dictate every single policy area, some constitutions offer a template for tax and spending assignments. Examples include the principle of subsidiarity, or residual power, enshrined in the constitutions of the United States, Switzerland and Germany.6 Austria also recognises the residual law-making powers of the provinces in its constitution. Conversely, Canada, India, South Africa and Spain assign residual power to the federal government.
Some constitutions link tax and spending autonomy – the so-called “Wicksellian Connection” –, stipulating that a given tier of government should pay for its public expenditure out of its own tax revenues. Italy includes the principle in its constitutional law, although deviating from it in practice. The German constitution applies a weaker form of Wicksellian Connection insofar as it states that funds should be commensurate with tasks.
Although some constitutions have hardly kept pace with the great evolutionary changes in the way that governments spend money, they contain provisions that have been flexible enough to adapt to new public expenditure requirements. Many explicitly enumerated federal powers listed in the United States’ constitution are quite specific to the late 18th century. Yet the power to “build post roads” enabled the federal government to create the highway network of the 20th century.

Borrowing and budgeting autonomy

Borrowing autonomy refers to a state’s ability to borrow on financial markets and/or from public institutions. Federal governments use various instruments to limit SCG borrowing. In the most extreme cases, they are not allowed to borrow. The German constitution, for instance, forbids new state borrowing outright if the state’s budgets are not structurally balanced. There are also less exhaustive borrowing restrictions. The Mexican constitution prohibits the states from borrowing from abroad, while Brazil requires federal approval of state-level borrowing. Several countries use differentiated deficit and debt rules when restricting the power of sub-national states/regions to add further debt, while some – like Brazil – might even restrict or entirely ban borrowing from state-owned banks. In Switzerland and the United States, there are no federal restrictions on state borrowing.

Budgeting autonomy evaluates the extent to which sub-national jurisdictions are able to set their budgets according to their own rules. Federal involvement in the state’s budgeting process varies considerably. Some countries, such as Argentina and Germany, explicitly forbid any federal interference in SCG budgeting. By contrast, some constitutions, like those of South Africa and Spain, allow the federal level to intervene in the substance of the sub-national budgets. Similarly, a number of countries have fiscal rules that reduce sub‐national budgeting autonomy. Failure to comply may bring about further federal intervention in SCGs’ budget processes, as in the Brazilian and Italian constitutions.

Fiscal autonomy indicator values across countries

Fiscal autonomy – sub-national tax, spending, borrowing and budgeting autonomy – varies widely from one country to another. It is greatest in the United States, Canada and Switzerland (Figure 2.3), moderate in Australia, Argentina and Mexico, and low everywhere else. Fiscal autonomy coherence, as denoted by the vertical bars in the figure, also varies. Fiscal autonomy is relatively well balanced in Mexico, where the states enjoy moderate autonomy in all fiscal policy areas, but less so in Argentina, Australia, Austria and Germany. The Argentinian fiscal constitution, in particular, combines low tax autonomy with substantial budgeting and borrowing powers.

Some federal governments restrict sub-national tax, spending and borrowing powers to a considerable degree. Others leave plenty of scope for spending and borrowing, while simultaneously allowing wide-ranging tax autonomy. Both types of institutional settings are coherent. Imbalance between the different forms of autonomy, though, may result in incoherent institutional frameworks and even in undesirable fiscal policy outcomes. In Argentina and Germany, for example, restricted sub-national tax autonomy, allied with extensive borrowing autonomy, has led the states to behave opportunistically and fall short of their budget targets (Tommasi, Saiegh and Sanguinetti, 2011).

Fiscal responsibility of the states

Fiscal responsibility refers to the extent to which sub-national tiers of government have to bear the consequences of their fiscal actions. While autonomy measures the extent of their freedom to conduct their policies, responsibility measures whether they internalise their costs. Responsibility is a central building block in fiscal constitutions since it determines the extent to which states and regions can derail the fiscal position of general government and make fiscal outcomes unsustainable. Responsibility is thus comparable to the concept of budget constraint. An SCG’s responsibility is determined by measuring the likelihood of bankruptcy or bailout, the status (imposed or self-imposed) of fiscal rules, and the strength and size of such transfer mechanisms as grants and equalisation payments. If such constraints are of similar strength, then responsibility arrangements are coherent.

Bailout and bankruptcy exposure

Not being bailed out in the event of bankruptcy is arguably one of the strongest sticks federal governments can wield to enforce sub-national fiscal responsibility. No-bailout clauses and exposure to default are thus core measures for ensuring the strength and credibility of budget constraints. The likelihood of default, the prevalence of insolvency frameworks and the probability of bailout are likely to affect SCG’s long-term behaviour. The extent to which they assume their responsibilities hinges on constitutional provisions as well as on actual experience:

Constitutional provisions. Fiscal constitutions contain rules for dealing with SCGs in fiscal distress. Some countries, like South Africa, forbid state/regional default outright, while others such as Brazil and Switzerland have provisions for an orderly default and an insolvency framework. Bail-out provisions are particularly critical and differ considerably from country to country. The Brazilian and Spanish constitutions forbid them, while those of Argentina and Germany enable them, and the Italian constitution requires them. And, although, some fiscal constitutions do not contain explicit bail-out provisions, they offer alternatives such as federal borrowing guarantees which are akin to an implicit bailout.
Historical experience. The credibility of no-bailout rules is shaped by a country’s experience of past defaults and how the federal government reacted. The experience of defaults may long affect sub-national behaviour. The defaults of several states in the United States in the 1840s and the federal refusal to bail them out may have restrained states’ fiscal policy and shaped collective behaviour to this day. On the other hand, the bailout of two German Länder in the 1990s fuelled further bailout expectations and may have contributed to the fiscal profligacy of some Länder in later years.

Constitutional courts shape bailout expectations extensively. In Germany, a federal court ruled in favour of the Länder of Saarland and Bremen in 1992, pointing to the solidarity principle enshrined in the constitution which it saw as an implicit bailout guarantee (Feld and Baskaran, 2010). In a similar case in 2006, the court adopted an opposite tack and denied the Land of Berlin a bailout on the grounds that it was able to cope itself. A Swiss federal court in 2003 confirmed the non-bailout rule after the bankruptcy of a municipality, thereby cutting the financing costs of the cantons (state level).7

Responsibility for fiscal rules

Fiscal rules can act as a signal to creditors that an SCG is following a prudent fiscal policy (Fall et al., 2015). While fiscal rules are a complement rather than a substitute for well-functioning fiscal frameworks, they help demonstrate that state finances are on a long-term sustainable track. Financial markets tend to reward prudent fiscal behaviour with lower yields (Schuknecht, von Hagen and Wolswijk, 2009). In that respect, a self-imposed rule is supposed to be a stronger sign of responsibility than a fiscal rule imposed by the federal government. An imposed rule assumes that the federal government is ultimately responsible for sub-national finances and that states can shift the fiscal burden onto other governments. Moreover, self-imposed rules create ownership, which is likely to increase the probability that states follow them.

The extent to which states self-impose fiscal rules varies across countries. While in Canada, Switzerland and the United States federal governments have no power over state fiscal policy, they do in Brazil, Germany and Russia. In countries like Argentina, Austria and Spain, the different tiers of government negotiate rules, which affords SCGs more leeway, though not full responsibility. In the United States, self-imposed rules are a strong anchor for state fiscal policy. Following a series of defaults during the 1840s and 1870s, virtually all the states enshrined relatively tight fiscal rules in their constitutions, which has provided them with a rules-based framework for sustainable fiscal policy to this day.8

Revenue responsibility

Revenue responsibility is the extent to which states draw on their own revenue sources. It is inversely related to external financing from the federal government – the greater the transfers, the lower the responsibility. High reliance on transfers and other common pool resources may ease the budget constraints on sub-central governments, create moral hazard, and distort tax enforcement (Rodden, Eskeland and Litvack, 2003). Fiscal constitutions usually comprise four types of federal support to the states:

Fiscal equalisation. The purpose of equalisation is to narrow differences in fiscal capacity between states. It may be enshrined in the constitution – in Canada and Switzerland, for example – or in secondary legislation only, as in Australia and Mexico. While countries like Australia and Germany require almost full equalisation of regional disparities, others (e.g. Mexico and Spain) require them to be equalised to a certain degree only. Some constitutions, like that of South Africa, require an independent council for equalisation policy.9
Tax sharing. In a number of federal countries, certain taxes are shared between tiers of government. Some constitutions, such as those of Argentina and Germany, have tax-sharing rules that apply to the major taxes, while ordinary law has provisions for tax sharing in Australia. Although there was a tax-sharing system in place in the United States in the 1970s and 1980s, but was easy to abandon because it was not enshrined in the constitution. The South African and Spanish constitutions require an independent body to set and adjust tax shares.
Stabilisation. Transfers for stabilisation purposes constitute another way of co-financing sub-national jurisdictions. Like tax-sharing arrangements, federal stabilisation policy may be enshrined in constitutional law, as in Germany and Italy. Alternatively, ordinary law may carry provisions to that effect, as in Australia and Canada. However, constitutional provisions attach a different thrust to stabilisation. While some, like those of Italy and Russia, stipulate that stabilisation “must” be pursued, others advise that it “should” be pursued (e.g. Germany and Spain).
Other intergovernmental transfers. Federal governments often support state activities through different forms of intergovernmental grants. The German and Swiss constitutions prescribe grants, while ordinary legislation shapes them in countries like Mexico and the United States. Some constitutions, e.g. those of the United States (again) and Belgium, have extensive provisions for grant systems, while Australia and Canada hardly mention grants or intergovernmental collaboration.

Generally, transfer systems are complements to each other rather than alternatives. In other words, more equalisation tends to go together with more stabilisation and more tax sharing.

Fiscal responsibility indicators across countries

Sub-national fiscal responsibility, as measured by institutional indicators, again varies widely from country to country. While levels of fiscal responsibility are high in Australia, Canada, Switzerland and the United States, they are low in Germany, Italy, South Africa and Belgium (Figure 2.4). Germany, India and Italy have relatively coherent responsibility arrangements, while in South Africa, Australia, Switzerland and the United States they are less coherent. Coherence in Switzerland is low because, while there is little likelihood of a bailout, the system of federal support to sub-national state governments in distress is extensive.

States’ power to co-determine federal policy

Co-determination is the extent to which SCGs can shape fiscal policy making at the federal level. While state autonomy refers to an SCG’s power to legislate for its own jurisdiction, co-determination refers to the degree of influence that an SCG, or group of SCGs, can exert on the fiscal policy of the whole country (Hooghe, Marks and Schakel, 2008). States can influence overall fiscal policy through different channels, the most important of which is often the second chamber of the federal parliament, or Senate. The coherence of co-determination reflects the extent to which certain channels of influence are complements rather than alternatives to each other. In a balanced co-determination framework, sub-national governments use all channels. In a less balanced setting, they use some channels heavily, while others are barely available.

Bicameralism

The two-chamber, or bicameral, system is a core attribute of federalism. The second chamber acts as a forum where the interests of state governments are explicitly represented and where they can co-determine national policies. The extent to which they can co-exert influence at the federal level, however, depends on the institutional strength of the second chamber and how amply represented their interests are.

Institutional strength. The Senate formally represents the states. It has full legislative and veto power in most federal countries, although some limit its power. In Canada and India, for instance, the second chamber can be excluded from initiating legislation. In Germany, the Senate can veto only legislation concerning the Länder. In Austria, the first chamber can overrule almost any Senate decision.
Representing states’ interests. States’ interests are represented to a different extent in the second chamber. Senators elected or appointed with a mandate by a state legislature or executive – as in Germany – generally represent its interests faithfully. They are less likely to do so, however, if they are elected at the ballot box. In Canada, senators are appointed by the governor general on the advice of the prime minister, which thwarts effective state representation. If the distribution of seats is proportional to population size – as in Belgium and South Africa – the federal government is more likely to align policies with the preferences of the largest jurisdictions.

While strong sub-national participation in federal policy making may balance the interests of more stakeholders, it may also lead to a joint decision trap where no policy decisions are taken at all (Scharpf, 2006). Reducing the degree of joint decision-making was the main aim of the 2006 federalism reform in Germany, where overlapping competencies, intertwined responsibilities and political bargaining led to frequent policy deadlocks. The German reform strengthened both the federal government’s decision-making capacity and the autonomy of the Länder (Moore, Jacoby and Gunlicks, 2010).

Judicial review

The second channel through which sub-national jurisdictions can co-determine federal policy is judicial review by the supreme or constitutional court. Apart from Switzerland, all the countries considered here have some form of constitutional review of federal laws, which can even void unconstitutional laws. Sub-national governments can influence constitutional review in two ways.

By challenging federal laws. States can trigger judicial reviews of federal legislation either directly, as in Italy and Spain, or through the second chamber, as in Germany and South Africa. The law under scrutiny may be repealed or modified as a consequence. The sole right of states and the second chamber to challenge federal legislation in court can be perceived as a deterrent against legislation that is unfavourable to states (Stone Sweet, 2000).
Through judicial appointment. In countries like Argentina and Germany, the sub-federal level or second chambers are often involved in appointing and approving judges to courts. There is broad evidence of close alignment between appointers and judges.10 Thus, when SCGs or second chambers play an important role in nominating or approving judges, they are likely to rule in their favour more often.

In Argentina, Australia, Germany, Mexico and the United States, judicial reviews play an important part in resolving conflicts between federal and sub-federal tiers of government. The United States Supreme Court is a particularly important crossroads of federal and state interests. A number of court rulings have reigned in the power of the federal government by insisting on a narrow interpretation of key clauses in the constitution.11

Other channels of co-determination

There are three other channels through which states co-determine policymaking at the federal level.

Constitutional amendment. All federal constitutions require that either sub-national jurisdictions – the provinces in Canada and the states in the United States – or the second chamber – as in Belgium and India – approve constitutional amendments. In some countries, e.g. Mexico and in Russia, both the second chamber and the states have to approve changes. In others, like Brazil and Mexico, the sub-national governments or, as in Italy and Switzerland, the second chamber may propose constitutional reform. That right is denied them, however, in countries such as Canada and Argentina.
Intergovernmental executive meetings. Intergovernmental meetings serve as a forum for negotiations between tiers of government and the co-ordination of national and state policies. Executive meetings are often institutionalised and take place as a matter of routine. Examples include the Consultation Committee in Belgium and the State Council in Russia. Some intergovernmental bodies wield considerable authority and their decisions formally bind the participants – e.g. the Council of Australian Governments (COAG) and Argentina’s Consejo Federal de Inversiones (Federal Investment Council) and Consejo Federal de Educación (Federal Education Council).
Intergovernmental transfers. Transfers are often determined through intergovernmental bargaining in which recipient governments have a say. Transfers are frequently the result of lobbying by sub-national politicians and interest groups. Indeed, SCGs can work to secure transfer through various channels, depending on how such channels are constitutionally anchored (OECD, 2014).

Co-determination indicators across countries

Most countries’ constitutions have provisions for extensive co-determination, as in Argentina, Germany and Brazil, though not in Canada, the United States or South Africa (Figure 2.5). The most coherent institutional setting prevails in India, where the states influence national policy moderately through all channels. Co-determination in Belgium, Canada, Russia and Switzerland is less coherent, with Belgium, for example, boasting a strong executive branch but a relatively weak second chamber. The combination of a weak federal court – one which is not allowed to review federal laws – with a strong second chamber makes co-determination less coherent in Switzerland.

Strength of budget frameworks

Budget frameworks govern the budget process and seek to restrain discretionary fiscal policy. Frameworks are defined by three elements: numerical fiscal rules, procedural fiscal rules and fiscal councils. Tight fiscal frameworks impose a set of well-defined numerical fiscal rules, involve top-down, transparent procedural and budgeting rules, and include fiscal councils or other arms-length agencies. Coherent budget frameworks are those where the three elements have similar weight. Less coherent frameworks are those where instruments are not uniform – where tight numerical fiscal rules co-exist with weak procedural rules, for example.

Numerical fiscal rules

Numerical fiscal rules constrain policy makers’ fiscal policy discretionary powers. The main rationale for fiscal rules is a perceived spending and deficit bias and the reluctance of states to commit to fiscal discipline. As such, numerical fiscal rules have gained popularity since the 1990s when they were put in place to curb irresponsible fiscal behaviour at all levels of government. There are four main types of numerical fiscal rules: those that focus on the budget balance (deficit), on debt, on expenditure and revenues, and those that focus on a combination of them all. Policy makers are most constrained when a country uses all four types.

There are four criteria for assessing the strength of numerical rules:

Legal basis. While some rules are enshrined in federal constitutions – e.g. debt brakes in Germany and Switzerland – others, like rules in Argentina and Australia, are set forth in secondary legislation. Constitutional fiscal rules are more difficult to amend and may entail high reputation costs for the government if breached. The use of a constitutional fiscal rule signals that fiscal discipline is perceived as a fundamental policy objective (Drazen, 2002).
Status. Sub-national rules are either federally imposed or self-imposed. Imposed rules are more likely to reflect a consistent, harmonised budget framework, whereas self-imposed rules may differ from one sub-federal jurisdiction to another. In Germany, for example, the federal government imposes a budget balance rule on the Länder. In Spain, too, central government imposes rules like the budget balance and debt and expenditure rules on the regions. In some countries, the different tiers of government negotiate rules. In Austria, for example, the various fiscal rules for the Länder are negotiated, while the regions and communities in Belgium discuss budget balance rules. In Switzerland and the United States, state fiscal rules are self-imposed and mostly enshrined in state constitutions.
Enforcement and sanctions. The constitutional setting may underpin the enforcement of fiscal rules. Rules are self-enforced in Australia, Canada, Switzerland and the United States. They tend to be stronger when enforced by a higher tier of government or by external bodies. Sub-national fiscal rules are enforced by the federal government in Spain and Russia, for instance. Courts and audit institutions have a prominent role in Brazil and the United States.12 In some countries, though, enforcement is not explicitly specified. The German constitution, for instance, does not mention sanctions and their enforcement.
Coverage. Wide coverage limits policy makers’ ability to bypass the rules and re-allocate fiscal resources between governments (Milesi-Ferretti, 2003). Some numerical fiscal rules cover the general government budget, as in Spain, whereas others apply only to federal or sub-federal finances. For example, the deficit rule in Mexico applies only to central government and a number of rules in South Africa govern only the provinces. In some countries, like Switzerland, separate federal and state fiscal rules complement each other.

Procedural fiscal rules

Procedural fiscal rules ensure that budget planning, approval and execution is subject to proper control and accountability, and that the annual budget law is consistent with medium- and long-term fiscal plans and objectives. Two elements help assess the bite of procedural fiscal rules: 1) the extent of top-down budgeting and 2) the transparency of the budgeting process.

Top-down approach. Top-down procedural rules aim to empower a single actor in the budgeting process as a way of addressing the problem of common-pool resources in public decision making (Ljungman, 2009). The top-down approach gives strong prerogatives to the federal executive over the legislature in the approval stage of the budget and/or to the federal prime or finance minister over other spending ministers, as in Argentina or Brazil (Alesina et al., 1999). In some countries, such as Belgium or Germany, constitutional law restricts the federal legislature’s power to amend a budget proposed by the federal government.
Transparency. Constitutions strive for budget transparency in three ways: 1) Some, like the German and Spanish constitutions, require federal and sub-national budgets to be assessed by an independent audit institution. Brazil and India, too, require their states to be audited. 2) The Brazilian and Russian constitutions require that states draw up medium-term budget frameworks to ensure planning transparency over several years. Medium-term objectives are further co-ordinated between the federal government and Belgium’s regions and South Africa’s provinces, for instance. 3) Constitutions may require uniform accounting standards across all levels of government, as in Italy and India.

Fiscal councils and other arms-length agencies

Fiscal councils and other arms-length bodies can help strengthen the budget framework. Fiscal councils independently analyse and review governments’ fiscal projections. To be more precise, they may assess compliance with fiscal rules and sustainability requirements or issue recommendations on specific items of budgetary policy. Councils thus raise awareness of the short- and long-term costs and benefits of budgetary measures both among policy makers and the general public (Calmfors and Wren-Lewis, 2011).

Three yardsticks help measure the strength of fiscal councils.

Institutional anchoring. Fiscal councils are likely to enjoy more stability, legitimacy and recognition if enshrined in the constitution, as in Germany or Spain, rather than in ordinary law, as in Australia or Canada. In some countries, such as Argentina and Brazil, they have been enshrined in the constitution for many years, yet are still not in place.
Prerogatives. In some countries, fiscal councils have such broad-reaching prerogatives as the right to assess the fiscal stance of both federal and state governments. Examples include the Parliamentary Budget Office in Italy and the High Council of Finance in Belgium. The Public Council in Russia and the Congressional Budget Office in the United States are allowed to evaluate the federal budget only.
Independence. The degree of independence that a fiscal council enjoys depends on its make-up – members of parliament, representatives of states, ministers, independent experts, etc. – and the body (e.g. parliament) to which it reports. Fiscal councils are largely independent, for instance, in Germany and Spain, while the parliamentary budget offices in Italy and the United States are only partially so.

Budget framework indicators across countries

The strength of budget frameworks varies considerably from one country to another. It is substantial in Spain, South Africa and Germany, where frameworks are well integrated (Figure 2.6), but weak in Argentina, Switzerland and Canada. Budget frameworks are coherent in Austria and Italy, but less so in Brazil, South Africa and India. Brazil’s budget framework is less coherent because, although numerical and procedural fiscal rules are very robust, there is no fiscal council. Fiscal rules and councils work best if they are combined (Coletta et al., 2015).

Stability of constitutional arrangements

The stability of fiscal constitutions depends largely on the number and strength of actors and their powers of veto. Veto powers increase the transaction costs of reforms and bias the institutional framework towards the status quo (Tsebelis, 2002). And, while stable institutions may provide a basis for long-term fiscal planning at all government levels, they may also prevent reform and the ability to adapt to changing circumstances. Very stable constitutions may slow down the pace of structural reform and fiscal adjustment. Stability is thus a two-edged sword.

Bicameral veto

In most federal countries, the second chamber is involved in the legislative process and can veto reform of the fiscal constitution. The strength of a bicameral veto can be gauged by the extent of the second chamber’s powers of veto and distribution of seats.

Extent of veto powers. The Senate enjoys full power of veto when it can veto any law initiated by the first chamber, which is the case in most federal countries. In others, though, like Germany and Mexico, it has only partial veto power.
Distribution of seats. A Senate with an even distribution of seats between states is usually more veto prone than a fragmented chamber, where it can be difficult to garner support for reform. Under a proportional, or partially proportional system (e.g. Germany and India), the federal government must secure support only from the larger jurisdictions.

Judicial veto

The strength of a judicial veto depends on the degree of constitutional review and the power to strike down unconstitutional legislation (Gutmann, Hayo and Voigt, 2014). It can be evaluated against four criteria.

Coverage. Certain laws are off-limits for judicial review. The Swiss Federal Court may rule only on laws passed by the cantons and not by the federal government.
Right to petition. The power of judicial review depends on the number of actors who may initiate a petition in court. In some countries, a broad range of political actors may do so, as in Austria and Brazil. By contrast, in Argentina and Australia, few have access to the constitutional court.
Timing. Judicial influence is broadest where the court is competent to check the constitutionality of laws both before and after a law has been adopted and implemented, which is the case in India and South Africa. Judicial influence is weaker, however, if the court can review legislation only after adoption, e.g. Canada and Mexico.
Unconstitutionality. Countries have different procedures for laws deemed unconstitutional. The offending piece of legislation is repealed automatically in Russia and Italy, while elsewhere it is returned to the legislature for revision. In countries like Canada and India, the constitution is silent on what happens with unconstitutional laws.

Powers of judicial veto are strongest when no laws are excluded from court adjudication, a wide range of actors can challenge federal legislation in the court, legislation can be challenged before and after a law is enacted, and when an unconstitutional law is automatically void.

Direct democracy

Direct democracy makes use of referendums and initiatives which might have opposite effects on the stability of the fiscal constitution. Referendums are an additional power of veto and make the status quo more difficult to change. Unlike referendums, citizens’ initiatives introduce additional instability into fiscal frameworks since they translate changes in public opinion directly into policy. While the Swiss constitution enshrines the fundamental right to launch a popular initiative, the Argentinian constitution explicitly bans citizens’ initiatives that address constitutional reform, international treaties, taxes, the budget and penal matters, for example.

Constitutional amendment

All constitutions contain sections that spell out the rules that govern amendment. They determine the frequency of actual amendments and, therefore, the stability of the constitution (Rasch and Congleton, 2006). There are five ways to change constitutions and which make it more or less easy to do so.

Qualified majorities. In most countries a qualified majority is required to approve a constitutional reform. Issues pertaining to the federal level can be amended with the consent of a three-quarters majority in both chambers in Russia. Belgium and Germany require two-thirds majorities in both chambers, and India and Italy absolute majorities.
Referendums. National referendums on constitutional reforms are required or allowed in Australia, Austria, Italy and Switzerland.
Consent from the states. In some countries, such as Australia and Switzerland, constitutional change must be approved by a majority of voters nationally and by a majority of states.
The number of actors that can propose a reform. In some countries a wide range of institutions and officials may propose constitutional reform. In Brazil the president, the first and second chambers (separately), and the states can do so. In Germany, only the federal parliament has that power.
Non-amendable parts. The most radical constraint on constitutional amendment is non-amendability. The first 12 articles of the Italian constitution, for example, cannot be modified. One concerns the principle of local autonomy and how administrative decentralisation has to be implemented.

Indicators that measure the stability of countries’ fiscal constitutions

Constitutional stability varies relatively little from one country to another compared to other building blocks. Generally speaking, though, the fiscal constitutions of Australia, Russia and the United States are more stable than those of Austria and Switzerland (Figure 2.7). Germany has the most coherent institutional environment. Brazil’s is less so, as it has strong judicial and bicameral veto powers, while allowing multiple actors to propose constitutional change.

Overall constitutional patterns

The final step in the empirical assessment is to aggregate the five building blocks, compare fiscal constitutions with each other, then give some statistical underpinning to terms like “competitive”, “co-operative”, “executive”, “dual” or “integrated” federalism. Two different methods help discern similarities as well as differences between the fiscal constitutions of the 15 countries under scrutiny. The first is clustering, which identifies countries whose fiscal constitutions are similar, but genuinely different from those of other groups. The second method is factor analysis, which makes it possible to calculate a summary, or composite, indicator that measures the degree of decentralisation granted in the fiscal constitution. Statistical analysis provides some evidence that the 15 fiscal constitutions can all be said to deliver some “constitutionally guaranteed decentralisation”, albeit to widely different degrees.

Clustering distinguishes between “decentralised” and “integrated” fiscal constitutions

This section describes the results of cluster analysis, which assesses whether some fiscal constitutions are similar to each other but different from others. It reveals two distinct groups – “decentralised” and “integrated” fiscal constitutions (Figure 2.8).

The United States, Canada, Switzerland, Australia, Argentina and Mexico have decentralised fiscal constitutions. They combine institutions that give rise to states with considerable autonomy, relatively high levels of responsibility, low powers of co-determination, and weak budget rules and frameworks. Decentralised constitutions tend to be quite stable as well, although Switzerland is an exception. Despite forming a cluster, decentralised fiscal constitutions still differ significantly from each other when it comes to degrees of responsibility. In the United States, Canada and Switzerland states, provinces and cantons are highly accountable for their actions. However, levels of responsibility are lower in Argentina, Australia and Mexico, which form a separate cluster of quasi-decentralised federations.
Austria, Belgium, Brazil, Germany, India, Italy, Russia, South Africa and Spain are characterised by co-operative or integrated fiscal constitutions. They are a mirror image of the decentralised cluster, generally combining low levels of autonomy and responsibility with high powers of co-determination and strong fiscal rules and frameworks. Typically, integrated fiscal constitutions are less stable. Some outliers should be pointed out, nevertheless. South Africa’s provinces have relatively weak powers of co-determination, while Belgium’s regions and communities and Russia’s oblasts boast quite stable fiscal constitutions.

A summary indicator of fiscal constitutions

The second method of gauging similarities and differences between fiscal constitutions is to develop a summary, or composite, indicator that reflects the degree of constitutionally enshrined decentralisation. The first step is to conduct a factor analysis to determine whether fiscal constitutions arrange their building blocks in the same combinations. Technically speaking, factor analysis tests to what extent variances in the building blocks’ indicator values co-move. Results suggest that the different building blocks are indeed highly correlated, with two single factors accounting for around 85% of the total variation (Table 2.2).

Factor 1 makes up around 66% of the variation in the original building blocks. It is strongly associated with autonomy, responsibility, co-determination and budget rules. These four building blocks are reduced to a single dimension, or factor, which may be termed “the extent of decentralisation”.
Factor 2 explains roughly 19% of the variation in the original variables. It is associated chiefly with the stability of the fiscal constitution. “Different degrees of stability” are thus the second characteristic, or factor, that helps differentiate fiscal constitutions.

Table 2.2. Commonalities between the building blocks of fiscal constitutions
Results of factor analysis
Variable	Factor 1	Factor 2
Autonomy	0.95	0.17
Responsibility	0.92	0.20
Co-determination	-0.81	0.05
Budget frameworks	-0.86	-0.03
Stability	0.10	0.99
Proportion of total variance	66%	19%

The summary indicator method, which aggregates four building blocks (not including the stability building block) delivers results similar to those of the clustering method, except that it is more fine-grained and also incorporates coherence (Figure 2.9). The United States, Canada and Switzerland are federations with highly decentralised fiscal constitutions and sometimes referred to as “competitive federalism”. Spain, Germany and Russia boast relatively well integrated, or co-operative, fiscal constitutions. Mexico, Argentina and Australia lie somewhere in between. Confidence intervals – the vertical bars around indicator values – indicate the level of coherence between building blocks. Spain and Canada have the most coherent fiscal constitutions, while Argentina has the least. Again, constitutional coherence is independent of whether a federation is decentralised or integrated. The fiscal constitution of the European Union, sometimes considered a proto-federation, leans towards decentralisation and is less coherent (Box 2.1). Decentralisation and stability are not correlated – i.e. they are independent of each other – as stable or unstable fiscal constitutions are to be found in both decentralised and integrated federations.

Box 2.1. The EU’s fiscal constitution: How decentralised, how coherent?

Although the European Union is not usually considered a true federation, some aspects of its institutional design and governance are akin to those observed in nation-state federations. As in all federations, many policy issues are related to institutional questions such as “Who does what?” and “How is co-ordination organised?” The single market, the (albeit small) EU budget and majority voting in selected policy areas also suggest that the European Union has some of the attributes of a federation. Against that background, it is possible to assess and rank the European Union’s constitutional design and coherence by applying the same methodology used for the 15 nation-state federal countries.

The EU fiscal constitution is moderately decentralised and less coherent than those of most federal countries (Figure 2.10). It is less decentralised, though, than those of the United States, Canada and Switzerland, and also less than those of quasi-decentralised federations like Mexico, Argentina and Australia. It gives the member states relatively high levels of autonomy and responsibility, as well as extensive powers of co-determination and strong hierarchical budget rules and frameworks – building blocks of co-operative federalism. In other words, the EU fiscal constitution combines elements from both competitive and co-operative federal systems.

This assessment seems to reflect the EU’s constitutional set-up well. Although the member states enjoy wide fiscal autonomy, a good number of EU policies help co-ordinate fiscal policy across countries and limit national discretionary powers. Since the EU budget is small, fiscal co-ordination is achieved through a set of stringent fiscal rules (OECD, 2014). They are applied under the terms of the Maastricht Treaty’s Excessive Deficit Procedure, the Stability and Growth Pact, the Fiscal Compact, and Six Pack and Two Pack regulations. Policies are also co-ordinated and kept under surveillance by the European Commission and the European Council as part of the annual cycle of the European Semester. Recent changes to the EU’s fiscal constitution further reinforced fiscal and economic governance by amending surveillance procedures, sharpening sanction mechanisms and setting intermediate fiscal and economic targets and adjustment procedures. This binding framework has been put in place over the last 20 years, particularly in the wake of the economic and fiscal crisis. It stands in contrast to the extensive autonomy of the member states in tax and spending matters.

How fiscal constitutions have evolved

Fiscal constitutions evolve over time, which raises three questions: “What changes?”, “How does it change?”, and “Why does it change?” (Benz and Broschek, 2013). The first and second questions relate to the five building blocks and the third to the mechanisms that produce constitutional reform. Fiscal constitutions may change because the fiscal or economic environment changes (at times of boom and bust), because the political setting changes (with sharp swings in the political climate or the advent of military or authoritarian regimes), or because there is a rise in separatist movements or a country’s break-up looms large. Fiscal constitutions thus reflect not only fiscal policy considerations, but the wider environment within which countries thrive.

A short history of fiscal constitutions

Over time, most fiscal constitutions have become less decentralised and more integrated since their inception, although there were a few countervailing episodes (Figure 2.11). To assess how fiscal constitutions have evolved, Blöchliger and Kantorowicz (2015) calculate the average of the summary indicator values of all 15 countries for the years 1917 to 2013.13 While the autonomy and responsibility of sub-federal entities trend downward – despite some increases in autonomy in the 1980s and 1990s – co‐determination and budget frameworks have strengthened over time. The degree of stability has remained – well – stable.

What drives constitutional reform?

This section affords further insight into the changes in the individual building blocks of fiscal constitutions. As the federal country club has grown over the last 100 years, the changes reflect both trends within old federations and the birth and subsequent evolution of new ones.

Trends in autonomy

Changes in the autonomy of sub-federal jurisdictions can be divided into three periods:

1917-80

During this period, losses of autonomy occurred chiefly in times of crisis. Autonomy declined during the Great Depression in the early 1930s, World War II and, less clearly, during the oil crises at the beginning of 1970s. Economic shocks and crises often prompted federal interference in sub-national autonomy. In Switzerland, the federal government expanded its powers of taxation at the expense of those of the cantons. After World War II, however, trends varied from country to country. Australia’s states never regained the power to tax income, which is today largely a federal prerogative. In Canada, by contrast, provinces like Quebec and Ontario ended the tax rental agreement with the federal level and established their own tax base again in the post-war period.

1980 to mid-1990s

During this period, sub-national autonomy grew. States started regaining power in the 1980s. In Australia, the credit limitations imposed by the Loan Council were phased out and the monitoring of states’ debt was left to financial markets. Mexico experienced a considerable surge in state autonomy, with several policy functions delegated to the states as part of the education reform in the early 1990s. In the 1990s, the states in the United States gained more power in the wake of a series of Supreme Court rulings and welfare reforms. The rulings devolved substantial responsibility to the states in the implementation of welfare policies.

Mid-1990s onwards

Since 1990, there has been a renewed decline in state autonomy, stemming chiefly from the financial crisis in emerging economies in 1998 and the global crisis of 2008. In the early 2000s, following a debt crisis and the bailouts of sub-national governments, the federal government in Brazil passed the Fiscal Responsibility Law in 2000, curbed sub-federal fiscal autonomy and re-centralised fiscal policy. In the European Union, the financial crisis of 2008 and the debt crisis of 2010 led many countries to introduce or refine numerical fiscal rules. In 2009, Germany adopted a constitutional debt brake that encompassed all levels of government. Spain and Italy soon followed suit.

Authoritarian regimes tend to restrict state autonomy, with governors (heads of sub-national executives) no longer being directly elected but appointed by central government. The military regimes of Argentina and Brazil began appointing governors in the 1930s, interspersed with democratic episodes when they were elected. The restoration of democracy in the 1980s in both countries resulted in a surge of sub-national autonomy. In Mexico, state autonomy was curbed by de facto single-party rule in all tiers of government between 1929 and 1989. Single-party rule was also the norm at both levels of government in India until the early 1990s.

Trends in responsibility

The responsibility indicator has trended downward since 1917, much in line with autonomy. The small upward movement in the late 1980s can be related to institutional changes in Australia, Canada and the United States. During the 1980s, the federal government in the United States abolished its revenue-sharing and equalisation mechanism and reined in the intergovernmental grant system. In Australia and Canada, states and provinces self-imposed a set of fiscal rules, although such rules actually play only a minor role in Canadian fiscal policy.

There are two main reasons for sub-national fiscal responsibility shrinking throughout the 20th century. The first is bailouts. The second is the rise of intergovernmental transfers in all forms in response to crises, regional disparities between sub-national jurisdictions, and the growth of inequality as a policy issue:

Tax sharing. Tax sharing is a popular means of sharing risk across all tiers of government levels in federal countries. It often dates back many decades. Argentina introduced its coparticipación system in the mid-1930s through ordinary legislative channels. Finally, in 1994, after pooling more and more taxes under federal supervision, it anchored coparticipación in the constitution. Germany’s tax-sharing mechanism has been constitutional since 1955 and was further broadened by an amendment to the constitution in 1969. The United States introduced a tax-sharing-cum-equalisation mechanism in the mid-1970s, before abolishing it in 1986.
Equalisation. Redistribution across sub-national jurisdictions has become wider-reaching and more institutionalised over time, as the policy objective of cutting inequality has gained importance. Switzerland made equalisation constitutional in 1958 and Canada in 1982, although it had fully established it in 1962. Australia has operated full horizontal fiscal equalisation since the 1980s, though not on a constitutional footing, while Russia introduced an equalisation formula in its Budget Code in 2004. Indeed, apart from the United States, all federations have explicit equalisation systems nowadays.
Stabilisation. Stabilisation policy and counter-cyclical transfers were introduced in the second half of the 20th century in the wake of the Great Depression. In 1947, the Swiss were the first to adopt a constitutional provision for coping with a slowdown, before further strengthening measures in 1978. In Germany, the power to pass counter-cyclical policy measures became constitutional in 1967, while Spain’s new constitution of 1978 empowered federal government to co-ordinate general economic planning. In Canada, the Federal-Provincial Fiscal Arrangements Act of 1985 enables the federal level to stabilise provincial revenues, with earlier legislation dating back to 1967.
Other transfers. Federal countries gradually established grant systems, either co-financing policy areas under state jurisdiction or, more recently, compensating for decentralised spending responsibilities. The German and Swiss constitutions contain multiple provisions which state that the federal level “should” or “must” support state activities. In 1999, Russia increased transfers to oblasts in support of spending mandates, while Belgium has funded spending decentralisation over the last two decades chiefly through more grants. In 1998, Mexico’s fiscal co-ordination law added transfers for education, health and infrastructure. In general, institutional anchoring of grants has increased with time.
Bailouts. Bailouts are a specific form of ad hoc intergovernmental transfer. Apart from Switzerland, all federations have bailed out a sub-national jurisdiction at some time. In Australia, New South Wales and in Canada Alberta, respectively, were bailed out during or after the Great Depression of the 1930s. In the United States, New York was bailed out in 1975, shortly after the first oil crisis, while in Latin America a first round of bailouts occurred after the fiscal crisis and sudden-stop episodes of the late 1980s. Although patterns were similar in Argentina and Brazil throughout the 1990s, they seemed to diverge in the 2000s. While Argentina bailed out two more states in 2003 and 2011, Brazil managed to stick to its constitutional no-bailout strategy.

Trends in co-determination

There are two episodes in the dynamics of co-determination that are worth discussing. The first was a surge in powers of co-determination in the 1950s and 1960s, the second a slower upward trend that commenced in the early 1980s.

Stronger powers of co-determination in the late 1940s and early 1950s followed the end of authoritarian rule in some countries. The evolution of co-determination during the 1960s was largely driven by Germany and Brazil. In the late 1960s, the power of the second chamber in Germany (Bundesrat) increased further, making the country one of the federations with the strongest joint decision-making powers. However, in 2006, the second chamber saw some of its powers reduced. As for Brazil, strong co-determination was established after the authoritarian rule during the 1960s and 1970s.
From the 1980s onwards, a number of countries strengthened sub-federal jurisdictions’ power to co-determine federal policy. In the 1980s, Belgium introduced the Consultation Committee, made up of the prime ministers of both tiers of government. Among its other tasks, the committee approves each government’s contribution to the effort to reduce the overall deficit in compliance with the Stability and Growth Pact. The establishment of a Constitutional Court in 1980 also strengthened Belgium’s regions and communities. In 1994, the Supreme Court in Mexico was given the power to review federal legislation to ensure compliance with the constitution. Similarly, in 1982, the Canadian provinces secured the right to approve constitutional amendments. As for Australia, it established the Council of Australian Governments (COAG) in 1992. Finally, since 1997, the Standing Conference for the Relationship between the State and the Regions has been the forum for political negotiations between the two tiers of government in Italy.

Changing patterns in budget rules and frameworks

Budget rules and frameworks have been beefed up at an unprecedented scale over the last decade, after long having changed very little. Numerical and procedural rules and other fiscal institutions have undergone sweeping reform, driven chiefly by the recent financial and debt crises.

The introduction of second-generation numerical fiscal rules was probably the most salient element of budgetary reform. Switzerland introduced a constitutional debt brake (in effect, a balanced budget rule) in 2001. Germany, Italy and Spain followed suit in 2009, 2011 and 2012, respectively, although the rules they introduced were more encompassing because they covered general government and not only, as in Switzerland, the federal level. A Spanish organic law also sets debt and expenditure rules for general government, while Italy’s Internal Stability Pact introduced expenditure and deficit rules in 1999. As a post-crisis and preventive measure, Russia amended its Budget Code in 2012 and introduced a fiscal rule that set a cap on federal government expenditure. As early as 1982, Germany ushered in expenditure rules that operated on the basis of political commitment to control spending.

Budget institutions and frameworks were also strengthened. The 2012 fiscal reform in Spain introduced medium-term budgetary frameworks for the top-two tiers of government and enabled the Supreme Audit Institution to scrutinise the budgets of the autonomous regions. In the same year, Italy harmonised budgetary frameworks across all tiers of government and unified accounting methods. Both countries also established fiscal councils. Canada’s Parliamentary Budget Office started operating in 2008 and Russia’s Public Council in 2011, although both bodies still boast only limited powers. In Germany, the 2009 constitutional amendment established the Stability Council, which has had the duty of monitoring compliance with the debt brake since 2013. In the United States, the Congressional Budget Office was established in 1974, while in Belgium the Public Sector Borrowing Requirement Section of the High Council of Finance came into being in 1989.

Coherence of fiscal constitutions increased over time

Overall, the coherence or alignment of fiscal institutions remained flat over long periods, but has considerably increased since the 1980s (Figure 2.12). The increase in the last 30 years can be traced back to the strengthening of the budget framework in many federations, often in reaction to low state responsibility and, to a lesser extent, to closer alignments of autonomy and responsibility. Decentralised federations have evolved less than integrated ones. Incoherence is most acute in times of war and under authoritarian regimes. Some constitutions, such as those of Argentina or the United States, have seen practically no change in their levels of coherence.

The clear-cut distinction between decentralised and integrated fiscal constitutions that can be observed today is actually quite recent. A cluster analysis of fiscal constitutions in 1980 and 1996 found that federations could not be neatly divided into the two groups (Blöchliger and Kantorowicz, 2015). In 1980 and 1996, at least four clusters of federations were distinguished, with no characteristic dividing line between them. Over the last 20 years or so fiscal constitutions have moved towards either the decentralised or the integrated model. In other words, distinction has become more pronounced.

Fiscal constitutions and fiscal outcomes

This section considers a few simple bivariate correlations between selected features of a fiscal constitution and fiscal outcomes between 1980 and 2013. The correlations link fiscal outcomes to both the level and coherence of constitutional decentralisation or, in other words, to both indicator values and variances. Correlation does not mean causation, but rather interaction. Fiscal institutions may affect fiscal outcomes at certain times, while outcomes might trigger changes to basic fiscal frameworks at others – as Section 2.5 shows. Average indicator values for 1980-2013 help capture long-term effects.

The correlations suggest that fiscal outcomes are hardly related to the level of constitutional decentralisation (Figure 2.13), while they are more closely related to coherence of constitutional decentralisation (Figure 2.14). In other words, the extent to which fiscal constitutions are decentralised has less impact on outcomes than the extent to which the various arrangements within a fiscal constitution fit together. As a result, the results of the simple bivariate correlations linking coherence to outcomes can be summarised as follows:

Coherence and spending. Primary spending growth seems to be positively correlated with less coherence. An unbalanced setting might allow jurisdictions to shift the consequences of excessive spending to other government levels or general government.
Coherence and debt. The growth of debt seems to be positively correlated with less balanced fiscal constitutions. In less coherent settings – when autonomy and responsibility are not aligned, for example – sub-national governments may be able to shift the consequences of fiscal profligacy to the federal level or other jurisdictions.
Coherence and crises. There is a correlation between the incoherence of fiscal constitutions and the frequency of crises a country underwent. Less balanced settings may be more prone to the build-up of deficit and debt, leading to a greater likelihood of crisis.
Coherence and economic growth. Coherence is positively correlated with growth rates. Conversely, unbalanced fiscal constitutions can harm the economic fabric.

As for levels of constitutional decentralisation, linking them to the same fiscal outcome variables as coherence (growth of debt, growth of primary spending, economic crises and GDP growth) delivers almost no relationship. The one possible exception is the relationship between the level of decentralisation and debt growth, which is positive, albeit only slightly.

Summary and conclusions

This chapter has sought to analyse the fiscal constitutions of federal countries and how they define intergovernmental relations and sub-central governments’ power and authority over fiscal resources. The results suggest that constitutional set-ups vary widely from one country to another. Yet their differences boil down to a single underlying dimension – the degree of constitutionally guaranteed decentralisation – within which three groups of federations can be distinguished: decentralised, integrated and (somewhere in-between) quasi-decentralised. All three comprise the same components (or building blocks) – sub-national fiscal autonomy, fiscal responsibility, influence on federal fiscal policy, and the strength of intergovernmental fiscal frameworks. Yet they are combined in distinctive arrangements in each group. Very tentative empirical evidence suggests that the constitutional set-up of a federation – the extent to which it delegates fiscal power to its sub-federal jurisdictions – has hardly any impact on such core policy outcomes as its budget balance, debt levels, economic growth or the number of economic crises it has undergone. At first glance, no federal model is better than another.

However, further empirical evidence suggests that coherence of constitutional arrangements does indeed have an impact on fiscal and economic outcomes. Coherence is the extent to which constitutional arrangements fit together – to what extent spending autonomy matches tax autonomy, for example, or whether fiscal rules are aligned with the fiscal responsibility of state and local governments. Less coherent constitutional arrangements tend to be associated with greater budget deficits, higher debt growth, lower economic growth, and more economic crises. Like a jigsaw whose pieces can fit only a certain way, the elements and building blocks of a fiscal constitution must form a coherent whole if they are to yield sound fiscal and economic outcomes. It is important to note that coherence is independent of whether a federation is decentralised or integrated: in both cases building blocks can be well or poorly aligned.

Some mismatches in fiscal constitutions are recurrent. Policy makers might wish to re-align certain building blocks and elements through constitutional or other policy reforms. The most frequent reforms would probably include:

Aligning sub-national taxing and spending autonomy – especially by increasing sub‐federal and local governments’ share of total tax revenue and their power to tax.
Aligning the autonomy of sub-national governments with their responsibility – particularly by strengthening budget constraints; adjusting deficit, debt and spending rules; and enforcing no-bailout clauses through an insolvency framework.
Aligning sub-national co-determination prerogatives with autonomy – either by strengthening co-determination if fiscal autonomy is to be reduced, or restricting it if fiscal autonomy is being extended and the over-fishing of common-pool resources has to be prevented.
Establishing a well anchored budget framework and a set of good budget practices – especially by introducing medium-term budgeting and aligning substantive (numerical) and procedural fiscal rules.

Other coherence and alignment issues may be more specific to individual federations. For each country has its own way towards a coherent fiscal constitution.

References

Alesina et al. (1999), “Budget Institutions and Fiscal Performance in Latin America”, Inter-American Development Bank Office of the Chief Economist, Working Paper Series, No. 394.

Benz, A. and J. Broschek (2013), Federal Dynamics.Continuity, Change, and the Varieties of Federalism, Oxford University Press.

Brennan, G. and J. Buchanan (1980), The Power to Tax: Analytic Foundations of a Fiscal Constitution, Cambridge.

Blöchliger, H. and J. Kantorowicz (2015), “Fiscal Constitutions: The Fundamental Drivers of Intergovernmental Fiscal Policy”, in: Kim, J. and H. Blöchliger, The Institutions of Intergovernmental Fiscal Relations, OECD, Paris.

Calmfors, L. and S. Wren-Lewis (2011), “What Should Fiscal Councils Do?”, Economic Policy, Vol. 26, Issue 68, pp. 649-695.

Dove, J. (2014), “Financial Markets, Fiscal Constraints, and Municipal Debt: Lessons and Evidence from the Panic of 1873”, Journal of Institutional Economics, Vol. 10, Issue 1, pp. 71-106.

Drazen, A. (2002), “Fiscal Rules from a Political Economy Perspective”, in G. Kopits (ed.), Rules-Based Fiscal Policy in Emerging Markets, Background, Analysis, and Prospects, pp. 15-29, Palgrave MacMillan.

Fall, F. et al. (2015), “Prudent Debt Targets and Fiscal Frameworks”, OECD Economic Policy Papers, No. 15, OECD Publishing, Paris, https://doi.org/10.1787/5jrxtjmmt9f7-en.

Feld, L. and T. Baskaran (2010), “Federalism Commission II – Recent Reforms of Federal-Lander Financial Relationship in Germany”, Review of Law & Economics, Issue 3.

Feld, L. et al. (2013), “Sovereign Bond Market Reactions to Fiscal Rules and No-Bailout Clauses – The Swiss Experience”, CESifo Working Paper, No. 4195.

Gutmann, J., B. Hayo and S. Voigt (2014), “Determinants of Constitutionally Safeguarded Judicial Review: Insights Based on a New Indicator”, in Y. Chang, (ed.), Empirical Legal Analysis: Assessing the Performance of Legal Institutions, pp. 216-254, Routledge.

Hooghe, L., G. Marks and A. Schakel (2008), “Operationalizing Regional Authority: A Coding Scheme for 42 Countries”, Regional and Federal Studies, Vol. 18, Nos. 2-3, pp. 123-142.

Ljungman, G. (2009), “Top-Down Budgeting – An Instrument to Strengthen Budget Management”, IMF Working Paper, WP/09/243.

Milesi-Ferretti, G.M. (2003), “Good, Bad or Ugly? On the Effects of Fiscal Rules with Creative Accounting”, Journal of Public Economics, Vol. 88, pp. 377-394.

Moore, C., W. Jacoby and A. Gunlicks (2010), “Introduction: German Federalism in Transition?”, in Moore, C. and W. Jacoby (eds.), German Federalism in Transition. Reforms in a Consensual State, Routledge.

OECD (2014), Fiscal Federalism 2014: Making Decentralisation Work, OECD Publishing, Paris, https://doi.org/10.1787/9789264204577-en.

Qian, Y. and B. Weingast (1997), “Federalism as a Commitment to Preserving Market Incentives”, Journal of Economic Perspectives, Vol. 11, No. 4.

Persson, T. and G. Tabellini (2003), Economic Effects of Constitutions, MIT Press, Cambridge.

Raudla, R. (2010), “Constitution, Public Finance, and Transition. Theoretical Developments in Constitutional Public Finance and the Case of Estonia”, Finanzsoziologie 4, Peter Lang.

Reinhart, C. and K. Rogoff (2010), This Time is Different. Eight Centuries of Financial Folly, Princeton University Press.

Riker, W. (1975), “Federalism”, in F. Greenstein and N. Polsby (eds.), The Handbook of Political Science, Vol. V: Government Institutions and Processes, Addison Wesley, Reading, MA.

Rodden, J., G. Eskeland and J. Litvack (2003), Fiscal Decentralization and the Challenge of Hard Budget Constraints, MIT Press, Cambridge.

Scharpf, F. (2006), “The Joint-Decision Trap Revisited”, Journal of Common Market Studies, Vol. 44, No. 4, pp. 845-64.

Schuknecht, L., J. von Hagen and G. Wolswijk (2009), “Government Risk Premiums in the Bond Market: EMU and Canada”, European Journal of Political Economy 25, pp. 371-384.

Smyth, R. and P. Kumar Narayan (2004), “Hail to the Chief! Leadership and Structural Change in the Level of Consensus on the High Court of Australia”, Journal of Empirical Legal Studies, Vol. 1, pp. 399-427.

Stone Sweet, A. (2000), Governing with Judges, Constitutional Politics in Europe, Oxford University Press.

Tommasi, M., S. Saiegh and P. Sanguinetti (2001), “Fiscal Federalism in Argentina: Policies, Politics, and Institutional Reform”, Economia, pp. 147-201.

Tsebelis, G. (2002), “Federalism and Veto Players” in U. Wagschal and H. Rentsch (eds.), Der Preis des Föderalismus, NZZ Verlag, pp. 295-318.

Vanberg, G. (2005), The Politics of Constitutional Review in Germany, Cambridge University Press, Cambridge.

Notes

← 1. The term “state” is used interchangeably with the country-specific terms for the intermediate level. In Austria and Germany, a state is a Land. In Canada and South Africa states are called provinces. In Belgium, Italy and Spain, the intermediate level is a region or community, in Switzerland a canton, and in Russia an oblast. Finally, in Australia, Brazil, India, Mexico and the United States, the intermediate level of government is a state or estado.

← 2. The questionnaire asked about assignment rules for personal and corporate income tax, indirect taxes, natural resource taxes and social security contributions.

← 3. Although states could in principle exit the revenue sharing system and start levying their own taxes, there is a provision in the law that makes exit prohibitively expensive (Convenio de Adhesion). If a state exits the system, federal taxes continue to be levied at the sub-national level, and revenue shares are calculated as though the state were still in the system. So the federal government keeps residual revenue fully.

← 4. For the purpose of this study, 16 spending categories (policy areas) were selected from the OECD’s Classification of the Functions of Government, second-level (COFOG-2). These are 1) national defence, 2) police services, 3) law courts, 4) prisons, 5) public transportation, 6) environmental protection,7) housing development, 8) out-patient services, 9) hospital services, 10) primary education, 11) secondary education, 12) tertiary education, 13) sickness and disability, 14) old age, 15) family and children and 16) unemployment.

← 5. The Commerce Clause says that states may not adopt regulations or taxes that place an “undue burden” on interstate commerce. Similar provisions are enshrined in Austrian and Swiss constitutional law.

← 6. The principle of subsidiarity holds that decentralising economic functions to lower levels of government should be favoured unless convincing arguments can be advanced for centralising them. Residual legislative power means that, unless some policy area is assigned to one level, it is automatically within the authority of the other level.

← 7. The court ruled that the canton of Valais was not liable for the debt of Leukerbad municipality, which went bankrupt in 1998. With the court’s ruling the no-bailout clause was confirmed. The court’s decision cut the relation between cantonal risk premiums and the financial situation of the municipalities, and reduced cantonal risk premiums by around 25 base points. The ruling showed that weak no-bailout commitments impose high costs on potential guarantors (Feld et al., 2013).

← 8. Since the beginning of 19th century the states had accumulated a large amount of debt chiefly to finance infrastructure projects. After fiscal panic due to shrinking revenues, several states were unable to service their debt. A bailout was discussed in the federal legislature but ultimately rejected. This decision sent a clear message thatstate debt was a state responsibility. In order to tap into the credit markets again, states made substantial reforms including the introduction of various balanced budget requirements. In the 1870s, after a banking panic and ensuing depression, states tightened their fiscal rules further (Dove, 2014).

← 9. In Australia, the Commonwealth Grants Commission, a non-partisan body, is responsible for equalisation. Unlike South Africa, the Commission’s role is not defined in the constitution but only by ordinary law.

← 10. For Australia, see Smyth and Narayan (2004); for Canada, see Songer et al. (1989); for Germany, see Vanberg (2005); for Italy, see Della Pellegrina and Garoupa, 2012; for Spain, see Garoupa, Gómez-Pomar and Grembi (2013).

← 11. See for instance, the 1995 court case United States v. Lopez. Also in 2012, in the case National Federation of Independent Business v. Sebelius the justices decided in favour of the states, ruling that the significant expansion of Medicaid was not a valid exercise of Congress’ spending power, since it forced states to accept the expansion at the risk of losing existing Medicaid funding.

← 12. In Brazil, state and local public officials are subject to criminal prosecution for non-compliance with the Fiscal Responsibility Law of 2000. The law limits new funding for sub-national governments and denies credit guarantees in case of systematic violation (Goldfajn and Guardia, 2004). A special Fiscal Crime Law (Lei dos Crimes Fiscais) sets out a range of penalties for budget mismanagement such as fines, removal from office, ineligibility for public office up to five years and even imprisonment.

← 13. The indicatorsfor the following countries and periods are coded: Argentina (1853-2013), Australia (1901-2013), Austria (1945-2013), Brazil (1891-2013), Belgium (1969-2013), Canada (1867-2013), Germany (1949-2013), India (1949-2013), Italy (1948-2013), Mexico (1917-2013), Russia (1993-2013), South Africa (1996-2013), Spain (1978-2013), Switzerland (1848-2013) and the United States (1791-2013).