Diversity and Asymmetric Arrangements as Drivers of Fiscal Federalism: A Comparative Overview

Abstract

“Diversity” and “asymmetry” are intimately intertwined with multilevel governance. Asymmetric arrangements often help accommodate various characteristics of diversity in a federation. In turn, asymmetric arrangements may have an impact on scope and level of diversity. This chapter provides an overview on the prevalence of asymmetric arrangements in multilevel governance, particularly fiscal decentralisation, to help address issues of diversity in a federation. The chapter also addresses measurement limits in this area, and presents the framework developed within the OECD to overcome these limits, thereby focusing on expenditure and revenue arrangements, intergovernmental transfers, as well as fiscal constitutions.

Introduction

“Diversity” and “asymmetry” are closely intertwined with the analysis of fiscal federalism, or more broadly, the institutions of multilevel governance. While, diversity, and in particular its accommodation, is seen as one of the reasons why multilevel governance institutions have spread all over the world during recent decades (Gagnon & Tremblay, 2020), asymmetric arrangements, often explicitly enshrined in the institutional legal framework, are used in different countries as tools to pursue diversity accommodation (Sahadžić, 2020).

Although asymmetric arrangements are sometimes considered to be isolated features of a select few federations, it is increasingly recognised in scholarly literature as an intrinsic element of federalism (OECD, 2019; Sacchi & Salotti, 2014). In fact, even most formally unitarian but decentralised countries show some degree of institutional heterogeneity (e.g. United Kingdom, Italy, or Spain). Furthermore, in contrast to the critical view of asymmetric arrangements expressed by scholars such as Tarlton (1965) or Sorens (2015), the position towards them has recently become more favourable (Watts, 2000).

Still, despite the increasingly positive understanding of the core role of asymmetries within federalism, the difficulty of classifying and measuring them makes their analysis a particularly complex task. Identifying and translating cross-jurisdictional heterogeneities into comparable figures might seem challenging, but it is essential to exploring and establishing empirical relationships between federal arrangements on diversity and economic policies.

As pointed out by Romero Caro (2022), research on addressing diversity within countries has traditionally focused on constitutional and legal recognition of national identities and minority rights. Because of this, the literature has not paid much attention to links between diversity and fiscal aspects of federalism—despite their importance (Weingast, 2014). Acknowledging this gap, this book will address questions such as: To what extent has fiscal federalism—or more generally, arrangements of asymmetric fiscal decentralisation—had an impact in accommodating diversity? How does diversity affect intergovernmental financial relations?

This chapter introduces the underexplored fiscal federalism-diversity debate, and is organised as follows: after a brief introduction, the second section provides a definition of diversity and explores its role as a driver for fiscal federalism. It also discusses the limits of measuring diversity and asymmetry, and presents recent proposals to overcome these difficulties. The third section will focus on how asymmetric fiscal federalism arrangements can be used to accommodate diversity. The fourth section addresses how to measure diversity and asymmetric arrangements, presenting the OECD methodology to measure fiscal decentralisation, and proposes that it could be extended to measure asymmetries. Section five explores the relationship between diversity and asymmetries. The final section concludes by inserting the term “menu-federalism” into the discussion as a useful method for approaching the topic.

Drivers of Fiscal Federalism

Why do some countries embrace fiscal federalism? What key factors determine whether a country will become fiscally decentralised? In response to these questions, this paper identifies three main drivers of fiscal decentralisation. The first driver is history. Current institutions are shaped by the past, and sometimes react against an excessively centralised past. Regarding the former, the decentralised character of some European monarchies’ rule, such as the Habsburgs’ (Ingrao, 2019), during the Middle and early Modern Ages can still be perceived in the institutional frameworks of several European countries (e.g. Austria). Regarding the latter, decentralisation was set up in some other jurisdictions to break away from unitary and autocratic periods (e.g. Spain), as decentralisation was identified with transition towards democracy (Hooghe et al., 2016). Finally, traumatic historical events can also produce federated systems—as it was the case in Germany after World War II.

The second driver is economic efficiency. The Netherlands and some Nordic countries (Bos, 2013), transited towards more decentralised fiscal arrangements for the sake of economic efficiency. Consistently, decentralised arrangements provided local governments with more tax and/or spending power by following the subsidiarity principle, as suggested by the First Generation of Fiscal Federalism (Oates, 1999). Institutional reforms undertaken in recent years in Italy, France, or Portugal have moved in a similar direction and followed the same rationale. In these cases, the intermediate levels of government (regions) are meant to play a more relevant role, presenting a deeper federal tenor than in the former cases. Diversity may also play a role in economic considerations, as pursued efficiency gains will be larger the more heterogeneous jurisdictions are. Therefore, incentives for efficiency-promoting decentralisation will be more powerful the larger and more diverse jurisdictions are (Table 1).

Table 1 Drivers of fiscal decentralisation/federalism

Finally, diversity can be the third driver of fiscal decentralisation. Whether diversity has political, religious, cultural, or ethnic sources, federalism is often an opportunity to accommodate for this by granting minorities the possibility of self-government in policy areas such as education, or by guaranteeing that their voice is heard in federal policy debates. In addition, in recent years, economic disparities in development and per capita income have emerged as a new reason for fiscal decentralisation, as some jurisdictions have started calling for higher tax and spending autonomy (e.g. Lombardy, Veneto, and Emilia-Romagna in Italy, or Catalonia in Spain) or lower levels of fiscal equalisation (e.g. Alberta in Canada, or Catalonia and the Balearic Islands in Spain) based on economic strength.

What Does Diversity Mean?

Since “diversity” is recognised as one of the key determinants of federalism, it is important to define and classify this notion. According to the most basic definition, diversity refers to the existence of differences that can manifest within several of the following areas:

Cultural

Cultural diversity is usually defined by the coexistence of multiple languages within a community. Sometimes, one language will be predominant in the whole country, with other minority languages spoken in specific jurisdictions—such as in Spain, France, or the Netherlands. In other cases, the hegemonic language will differ across jurisdictions because the inhabitants do not all share a common language—as is the case in Belgium, Switzerland, and Canada. Although linguistic diversity is the most evident in the cultural field, cultural diversity can also include spaces such as artistic creation or mass and social media. If the population of a country shares no common language, its common spaces are at risk of fragmentation.

Religious

Faith is another area where diversity can be expressed. In a globalised context where international migration has become more common, no religiously homogeneous country exists anymore. While migration has not led to federalisation on the grounds of religion yet, pre-existing large-scale religious diversity is behind some institutional decentralisation reforms, particularly in African and Asian countries.

Ethnic

Cohabitation of different ethnic groups—such as minorities living heterogeneously across the territory of the country, or minorities that constitute majorities in specific jurisdictions—can also require specific federal accommodation. Ethnic differences are often deeply intertwined with differences in culture and religion. When this sort of diversity is not correctly addressed, tension between ethnic groups can lead to violent clashes and even civil wars or secession. Examples of federations brought about by ethnic diversity can be found in South Africa, Bosnia Herzegovina, Ethiopia, and Pakistan.

Political

These aforementioned differences often encourage political diversity. Probably the most evident sign of political diversity is the creation of so-called non-national political parties and social organisations, such as trade unions. These organisations do not operate in all jurisdictions, and only participate in elections in territories that share common characteristics and goals. These features can oscillate between the existence of a differentiated/separated national feeling, or a deep fragmentation caused by cultural, religious, or ethnic diversity. The Belgian case best exemplifies political diversity, although its fragmentation is caused largely by linguistic differences. Political diversity brought about by contested notions of nationhood also exists in Spain and the United Kingdom.

Economic

Finally, diversity can refer to differences in economic performance. Income per capita and levels of development play a crucial role, but so do the relevance of certain economic sectors, labour market performance, and demographic structure. Low-income jurisdictions will often call for lower levels of decentralisation or higher shares of federal equalisation, whereas high-income jurisdictions will call for the opposite (Blöchliger & Vammalle, 2012). When economic disparities become unbearable, demands for asymmetric arrangements become more likely. Both the asymmetric devolution process in the United Kingdom and the special autonomy status provided to some northern Italian regions are good examples of this phenomenon, both of which grow in significance as territorial inequalities increase.

Measuring Diversity and Asymmetric Fiscal Arrangements

Little attention has been paid to the question of diversity measurement within decentralisation literature. To address this, Mathieu and Guénette (2017) proposed the Societal Culture Index, which consists of twelve indicators grouped into six pillars. This composite index measures the degree of autonomy that multinational federations provide to minorities so they can sustain their “societal culture”. However, this index does not measure diversity itself. The presence of constitutional asymmetries that can be found among the set of proxies is used to measure national recognition. The authors computed the Societal Culture Index using Catalonia, Quebec, and South Tyrol as examples.

A suitable design for a Federal Diversity Index could calculate a composite indicator based on items for the four diversity categories enumerated in previous paragraphs (cultural, religious, ethnic, and political), together with proxies for economic disparities. We propose structuring this index around three main pillars: existence of diversity, depth of diversity, and fragmentation (Table 2).

Table 2 Proposed criteria for a federal diversity index

Asymmetries in Fiscal Federalism Arrangements

Adequate diversity management is often key for countries to prevent secessionist pressures and remain united (Alesina & Spolaore, 2005; Tarlton, 1965). Because federalism provides multiple tools to respond to the need to account for diversity within the institutional framework, it is often presented as an intermediate alternative to both unitary and secessionist calls. Federalism’s main benefit is its capacity to provide differentiated responses to diverse situations. Indeed, according to the first generation of authors of the theory of fiscal federalism (Oates, 1999), the main difference between the unitary and decentralised models is the unitary state’s lack of capacity to provide geographically tailored policies. Policy-tailoring is just one kind of beneficial asymmetric arrangement available for federations.

What Is Meant by Asymmetry?

Congleton (2015) defines asymmetric decentralisation as an arrangement whereby different subnational governments of the same level have different political, administrative, or fiscal powers. Other authors have a broader understanding of asymmetry. According to Watts (2000), asymmetries are the varying degrees of autonomy and power that represent the political expression of diversities. Accordingly, Watts (2000) classified asymmetries according to four criteria:

• Politically or capacity-driven asymmetries: while the former would be driven by diversity, the latter would respond to economic efficiency reasons. For example, better fiscal or human capital endowments could encourage the recognising of higher decentralisation levels within large municipalities.

• De jure or de facto asymmetries: constitutionally declared institutional heterogeneities versus asymmetries which emerged from the distinct application of a homogenous legal framework.

• Central or peripheral asymmetries: Canada or Australia are some examples of federations that provide a lower level of decentralisation to peripheral “centrally administered territories” where fewer people generally live, in contrast to central or core jurisdictions. This criterion could also be applied to analyse federal models where it is the federal/central government that governs the capital region, as is the case of the United Kingdom.

• Temporary or permanent asymmetries: asymmetries that are temporarily enforced during the first years of federal reforms and which become consolidated over time (e.g. in Spain).

In accordance with this classification of asymmetries, Watts (2000) proposed nine main areas from which asymmetries are likely to emerge:

• Population, territory, and wealth.

• Degree of autonomy and powers.

• Fiscal power.

• Representation in federal institutions.

• Particular bilateral negotiation for intergovernmental relations.

• Role of regional political parties.

• Degree of uniformity on the application of the constitutional block.

• Importance placed on the constitutional amendment process.

• Degree of uniformity imposed on subnational constitutions.

This list goes far beyond fiscal institutions, as it provides a comprehensive overview of potentially asymmetric institutional arrangements in federal or quasi-federal/regional countries. Table 3 provides a list of asymmetries in the fiscal domain in several federal, quasi-federal, or regionalised countries. Since this chapter deals mainly with asymmetries in the fiscal domain, we will follow the categories as established by the OECD (2020).

Table 3 Examples of asymmetric fiscal institutions in decentralised countries

Measuring Fiscal Decentralisation

It is said that as many federal models exist as there are federal countries, meaning that although similarities across federal arrangements occur, identical fiscal decentralisation frameworks do not exist. These country-by-country specificities and nuances add complexity, which makes measuring and comparing fiscal decentralisation arrangements difficult. In addition, the multi-faceted character of decentralisation makes it challenging to capture existing distinctiveness through a single indicator (Blöchliger, 2015). In this section, some of the most commonly used quantitative variables will be enumerated and their limits discussed, then recently developed qualitative indicators will be proposed as an alternative approach. This will be done for four main fiscal decentralisation arrangements examined by the OECD (2020): public expenditures, revenue, fiscal constitutions, and asymmetries.

Expenditure Autonomy

Subnational public expenditure share is the most common variable used to quantitatively describe degree of decentralisation within government budgeted spending. A higher rate of subnational spending is interpreted as higher spending autonomy for local and intermediate government levels. In addition to aggregate spending figures, recent OECD data offers separate numbers for ten policy areas under the Classification of Functions of Government (COFOG) criteria.

$$\frac{{{\text{Subnational}}\,{\text{public}}\,{\text{spending}}}}{{{\text{Total}}\,{\text{public}}\,{\text{spending}}}}.$$

However, although these ratios can offer a glimpse of the extent to which subnational governments are decentralised, they fail to capture the actual degree of autonomy (Blöchliger & King, 2007; OECD, 2021), and can lead to misleading results. Ratios explain which level of government executes expenditure or provides funding for public services, such as healthcare or education, but neglect to explain which level of government sets access and provision requirements, determines staff compensation, purchases supplies, or how much oversight higher levels of government exert. All these elements, not just funding, define what constitutes the actual level of subnational spending autonomy.

Therefore, considering the limitations of purely quantitative methodologies based on National Account Statistics, scholars have proposed alternative approaches based on qualitative measures in order to shed some light into this issue. The spending autonomy indicators developed by Dougherty and Phillips (2019), based on the Bach et al. (2009) classification of spending autonomy, represent a good example. This approach identifies a series of measurable indicators by codifying survey responses (Fig. 1).

Fig. 1

(Source Dougherty & Phillips, 2019)

Set of indicators used by OECD’s spending autonomy composite index

This exercise is carried out for five policy areas: education, long-term care, transport services, housing, and healthcare, which are among the most frequently decentralised policy areas. Both the sub-indicators within policy areas and the aggregate indices for the five policy areas are computed by applying two alternative aggregation methods. First, there is the random weights method—which in this case assumes that all indicators should have the same relevance in determining indicator’s final values. Next is the country-product dummy method—which is meant to address problems of missing values by taking into account country, autonomy aspect, and policy area fixed effects. Because both approaches are very similar, only the outcomes of the latter approach (country-product dummy method) will be reported here.

As noted in Fig. 2, there is a strong correlation between the degree of spending decentralisation reported by both approaches. Overall, the quantitative approach usually overestimates the actual subnational spending capacity, as it disregards SNG’s (subnational governments) actual decision-making capacity over the money spent. This upward bias is particularly noticeable in the case of Iceland. Conversely, the quantitative approach seems to underestimate the actual sub-central spending power in the United Kingdom and New Zealand, where low budgetary figures do not adequately represent actual spending autonomy.

Fig. 2

(Source Authors’ calculation, based on OECD [2021] dataFootnote

Note Quantitative figures report subnational spending out of total public spending for 2019, while qualitative figures report estimates of the composite indicator for spending autonomy using the country-product dummy approach. Line represents the linear trend.

)

Qualitative (y-axis) vs. quantitative (x-axis) spending decentralisation (2019)

Following the same rationale, Kersting et al. (2020) adapt OECD’s methodology to the context of Catalonia in Spain and provide figures for spending autonomy on healthcare, education, and social protection. The overall results portray actual autonomy as being lower than suggested by purely quantitative variables, as is the case in most other OECD countries. Finally, other qualitative spending autonomy measurement methodologies can be found for education (OECD, 2004), healthcare (Paris et al., 2010), and the aggregate of public policies (Abdelhak et al., 2012).

Despite the unquestionable usefulness of composite indicators to describe complex information in a synthetic way and to allow cross-country comparisons, some authors have expressed concern about its implementation. Usually, this criticism points out the high discretion of their design and that aggregation can sometimes lead to important information losses (Mazziotta & Pareto, 2013).

Regarding asymmetric arrangements in expenditure autonomy, the variable to observe would be the variance of expenditure autonomy across regions. To calculate this, it would be necessary to access data regarding levels of expenditure autonomy not only at the country-aggregated level, but also on a region-to-region basis. Although this approach is currently possible for individual countries, the lack of cross-regional comprehensive databases on subnational public finances makes it impossible to compute quantitative results.

Revenue Autonomy

The share of subnational tax revenues is the variable that has traditionally been used to measure the degree of tax autonomy granted to local and regional/state governments. Naturally, the higher the share of total tax revenue determined, collected, or owned by local and regional governments, the higher the degree of tax autonomy.

$$\frac{{{\text{Subnational}}\,{\text{tax}}\,{\text{revenue}}}}{{{\text{Total}}\,{\text{tax}}\,{\text{revenue}}}}$$

However, this quantitative indicator has some shortcomings when determining the actual level of tax autonomy. For instance, it does not clarify whether the design and/or collection of decentralised tax revenues are shared across levels of government or not, or which tax elements (rates, brackets, bases, breaks, reliefs) and steps of the taxing process (legislation, administration and collection, inspection) the SNGs control.

The OECD tried to address this issue by designing a taxonomy of taxing power that could help better understand what is the actual power of SNGs to determine revenues. According to this definition, the highest level of tax autonomy is where the SNG has full authority over tax rates and tax bases, which includes the power to set the rates of taxation and to define what constitutes the tax base. By contrast, tax-revenue sharing schemes are considered to be the decentralisation approach which provides the lowest level of autonomy. It should also be noted that, as pointed out by Blöchliger and Kantorowitz (2015), even when qualitative indicators suggest high levels of tax, SNGs could be discouraged to make use of it through centrally set tools such as intergovernmental transfer schemes (Table 4).

Table 4 OECD taxonomy of taxing powerFootnote

Alternatives within each category are ordered from highest to lowest tax autonomy level.

Figure 3 compares the subnational revenue shares with subnational tax autonomy of a selection of OECD member countries. While the former variable includes all tax revenue entering SNGs’ budgets (including those coming from tax-sharing schemes, thus without autonomy over tax design), the latter variable includes only revenue streams over which SNGs can determine tax rates or tax reliefs. Again, there is a strong correlation between results for both approaches. Larger differences are recorded in the cases of Germany, Latvia, United States, Poland, and Canada, with little to no variance in the rest of the countries considered. The German case is particularly remarkable, as it ranks fifth using the revenue share (quantitative) approach but is considered highly centralised if the tax autonomy indicator is used. This can be explained by the fact that the largest revenue flows for German SNGs come from a share in the PIT and VAT, whose design SNGs can only change collectively in the Bundesrat.

Fig. 3

(Source Own elaboration, based on OECD [2021] dataFootnote

Note subnational tax autonomy is measured as the share of subnational revenue over which SNGs do have power regarding tax rates and reliefs, or just tax rates.

)

Subnational revenue share (% out of total) vs. subnational tax autonomy (2018)

The same difficulties when measuring asymmetric arrangements in spending autonomy also influence measuring revenue autonomy. Information about the decision-making power each subnational jurisdiction is vested with regarding each tax figure is necessary for this model. Once this information is gathered, tax autonomy variance can be computed as a proxy for revenue asymmetries. The highly data-demanding nature of this exercise explains why it has not been attempted yet.

Intergovernmental Fiscal Transfers

Intergovernmental fiscal transfer schemes are intrinsic to every federation or fiscally decentralised jurisdiction. Indeed, when spending autonomy is larger than tax autonomy—a phenomenon called “vertical fiscal imbalance”—transfers are required to address this imbalance. Transfers can also reduce disparities in tax capacity and spending needs across jurisdictions. Such transfers are usually known as “fiscal equalisation”. It is relatively common for transfers designed to close vertical and horizontal imbalances to be intertwined within the same intergovernmental transfer framework.

Traditionally understood quantitative ratios measure the share of subnational budget covered by intergovernmental transfers. Dependence ratios can be computed based on both income and expenditure statistics. However, the actual autonomy which transfers grant to SNGs can vary according to their design. For example:

• Are grants earmarked for a specific policy area or programme?

• Are transfers paid as block grants, or as matching grants?

• How are transfers funded? Which level of government pays for transfers (vertical/horizontal)?

The transfer programmes that provide the greatest level of autonomy are non-earmarked grants, as they come without strings attached. Generally, fiscal equalisation transfers fit this description. Block grants also provide significant levels of autonomy. Conversely, earmarked grants, which must be devoted to pursuing centrally defined policy objectives, and matching grants, whose amount varies together with the funding devoted by SNGs, allow the central government to exert a stronger control over subnational entities. Finally, when transfers are horizontal, meaning they are funded by revenue raised by the same level of government that will receive them, less central control is exerted—although even horizontal transfer schemes are usually arranged by the federal government.

Different visualisation strategies help to better compare intergovernmental fiscal transfer arrangements across countries. Figure 4 displays the amount of fiscal equalisation measured as a percentage of total government expenditure. Among the selected OECD countries, Australia, Japan, Germany, and Korea are the ones where the size of equalisation is largest with respect to total expenditure carried out by the public sector. Still, there are also differences across these countries—as Japan and Korea heavily rely on cost equalisation, while the German equalisation system is mainly based on revenue equalisation.

Fig. 4

(Source OECD, 2021)

Cost and revenue equalisation (% of total government expenditure) (2017)

Some countries deliberately apply several equalisation models or formula when disbursing transfers across jurisdictions, resulting in different levels of per capita funding. Examples of such asymmetries can be found at the regional level in Spain, the United Kingdom, Canada, and Italy. They are even more common at the local level, when rural, depopulated, or small municipalities are supported by a different transfer scheme than large and dynamic cities and capitals—as is the case in Canada or Spain. Some asymmetries might also be implicit and the result of inconsistent transfer formulas, such as when revenue disparities increase rather than decreasing after transfers.

When looking for a feasible measure for asymmetric intergovernmental fiscal transfer arrangements, variance among SNGs in funding per capita—adjusted to differences in spending needs–may represent the best choice. This could be a less data-intensive approach and would capture disparities coming from both explicit and implicit asymmetries.

Fiscal Constitutions

Fiscal constitutions are defined as the set of rules and frameworks which guide fiscal policy and are enshrined in fundamental laws (Blöchliger & Kantorowitz, 2015), such as federal constitutions or high-ranking institutional laws—often known as basic or organic laws—which enjoy a quasi-constitutional status. As this concept is so broad, scholarly literature has devoted efforts to systematise its analysis by elaborating a list of measurable elements of fiscal constitutions. For instance, following the rationale for qualitative composite indices previously examined, Blöchliger and Kantorowitz (2015) identified five building blocks of fiscal constitutions, and a set of indicators that could help to quantify “decentralisation” in each of those blocks:

• Autonomy: the extent to which SNGs can design and implement their own fiscal policy.

• Responsibility: the degree to which SNGs are subject to budget constraints and must face the consequences of their own fiscal policy.

• Co-determination: the extent to which SNGs participate in federal fiscal policy determination.

• Budget frameworks: the degree to which fiscal rules and budgetary stability and financial sustainability frameworks constrain discretionary fiscal policy across government levels.

• Stability: ease with which fiscal constitutions can be reformed (Fig. 5).

Fig. 5

(Source Blöchliger & Kantorowitz, 2015)

Fiscal constitutions, Indicator tree

By pairing each fiscal constitution block with a series of indicators, the authors can build a composite index. The proposed composite index allows the classification of fiscal constitutions into three different types: integrated, quasi-decentralised, and decentralised fiscal constitutions. In addition, Blöchliger and Kantorowitz (2015) compute two more related variables. First, the “completeness” index measures whether a certain feature (e.g. transfers, fiscal rules, councils) is listed in the fiscal constitution or not. Second, the “coherence” index measures whether fiscal constitutions provide similar degrees of autonomy across the five building blocks, with more balanced and coherent decentralisation schemes being preferable.

Figure 6 presents the results of the degree of constitutional decentralisation and its coherence between the five building blocks. While the United States, Canada, and Switzerland reach the highest constitutionally guaranteed values of decentralisation (represented by the numerical value of the diamond), India, Canada, and Spain feature the highest level of constitutional coherence (represented by the length of the vertical bar). Argentina’s and Switzerland’s constitutions are relatively incoherent, in that they combine elements of high and low decentralisation across constitutional building blocks.

Fig. 6

(Source Blöchliger & Kantorowitz, 2015)

Degree of constitutional decentralisation and coherence

Fiscal constitutions emerge as the key explanatory element in all the asymmetric arrangements examined before—as tax and spending autonomy, intergovernmental transfers, and fiscal rules are determined by fiscal constitutions. Indeed, replicating the constitutional decentralisation degree composite indicator across subnational jurisdictions would provide an answer to the question of the level of asymmetries within a country in qualitative terms.

Exploring the Relationship Between Diversity and Asymmetry

Do asymmetric arrangements necessarily mean higher inequality? And do asymmetries feed diversity? These questions have become a key issue in the debate on the political economy of federalism (Bird & Ebel, 2006; Martínez‐Vázquez et al., 2017). To empirically explore the links between diversity and asymmetries, it would be necessary to compute and to measure the degree of asymmetry in fiscal federalism frameworks. Since empirical studies on this topic have not been carried out so far, we will consider the topic from a theoretical point of view.

Does Diversity Accommodation Necessarily Lead to Larger Asymmetries?

Previous sections have presented asymmetries as a tool to address diversity. But does diversity have to be addressed by setting up asymmetric arrangements? In our view, two conditions need to be fulfilled: diversity needs to be “geographically concentrated” and the baseline degree of decentralisation of the country needs to be lower than that demanded by the jurisdictions that host the minority group. Our analysis assumes the existence of democratic inclusive institutions, as it is difficult to envision that calls for higher decentralisation driven by diversity would be taken seriously by an authoritarian regime.

When only one jurisdiction is different, and thus diversity is heavily concentrated (e.g. the French-speaking minority in Canada in the province of Quebec, and, to a lesser extent, political diversity across entities in the United Kingdom), asymmetries are more likely to arise—as long as the overall degree of decentralisation within the federation as a whole does not satisfy the desire for autonomy within that jurisdiction. When diversity is scattered across virtually all jurisdictions (e.g. ethnic diversity in India, linguistic diversity in Belgium), then federalism is likely to be symmetric. Even if diversity is “concentrated”, no asymmetric arrangements will be needed if the level of decentralisation provided to all jurisdictions is at least as high as the level of autonomy claimed by the dissenting jurisdiction (Congleton, 2015) (Fig. 7).

Fig. 7

(Source Own elaboration)

When are asymmetric arrangements used to address diversity?

However, the fact that asymmetries are less likely to be established when diversity is scattered across jurisdictions does not necessarily imply that managing diversity in that federation will be easier. In fact, each minority group could have different needs and the focus of their demands could be very different. Vice versa, when diversity is concentrated, it can be easier to identify a region’s specific needs and provide targeted solutions to accommodate them, even if they translate into higher asymmetries.

Does Asymmetry Contribute to Increased Disparities?

When focusing on fiscal federalism, political diversity and economic disparities are the two types of diversity directly involved. In particular, the discussion surrounding diversity and asymmetries usually deals with attitudes towards interjurisdictional inequality. In addition, during recent decades, a decrease in cross-country inequalities has been accompanied by increasing levels of inequality within countries themselves (Bartolini et al., 2016). Because of this, the relationship between decentralisation, asymmetries, and the rise of economic disparities has gained increasing attention.

Some scholars have tried to disentangle the fiscal decentralisation-inequality link, with results being contradictory and often dependent on the context, data, or methodology used (Martínez‐Vázquez et al., 2017). Additionally, a lack of data has hindered studies which focus on the relationship between asymmetry and inequality. At first glance, explicit asymmetries that provide different treatment to some subnational jurisdictions could result in an increase in economic and/or political disparities. However, even if asymmetries represent a policy response to accommodate diversity, a vicious circle could still emerge: to accommodate diversity, asymmetric arrangements would be set up, which could then lead to disparities rising further, until any scope for further asymmetry would be exhausted. For instance, this is the conclusion reached by Ishiyama (2022) regarding Ethiopian asymmetric ethnic federalism (Fig. 8).

Fig. 8

(Source Own elaboration)

The diversity-asymmetry vicious circle

However, there is no empirical evidence that asymmetric fiscal federal arrangements result in greater political or economic disparities than symmetric fiscal decentralisation (see Martínez‐Vázquez et al., 2017 about the likelihood of emergence of subnational political parties). In addition, the implicit asymmetric results of formally symmetric institutional arrangements also potentially lead to similar consequences. As an example, regional policy provided by unitary governments, although being formally “symmetric” as the same investment allocation formula is applied to all territories, translates into uneven allocation. Indeed, low-income jurisdictions will receive proportionately more funding, as the aim of the policy is to close the development disparities among different regions. This is just an example of how formally symmetric policies could provide asymmetric results while, if successful, producing a disparity-decreasing effect at the same time.

As it is not clear which cause determines the outcomes of the asymmetry-diversity relationship, literature on the subject is divided. Watts (2000) or Brand (2008) defends the positive aspects of asymmetries for diversity accommodation—offering a long list of examples of de facto and de jure asymmetries. On the other hand, Tarlton (1965) or Sorens (2015) warns about the risks of asymmetries for the political stability of federations. In their view, asymmetries might make preserving harmony and unity difficult, and thus strengthen calls for symmetrical arrangements. Similarly, Bird and Ebel (2006) postulate that asymmetries could heighten feelings of comparative affront or discrimination within the regions where special regimes are not applied, thus aggravating the political conflict that asymmetries aimed to ameliorate.

These risks could be particularly serious in cases where calls for asymmetric arrangements do not emerge due to cultural, religious, ethnic, or political diversity, but instead as a consequence of economic disparities. In fact, recently observed movements calling for differentiated fiscal arrangements for subnational jurisdictions have sprung up in high-income regions, such as in northern Italy or in the Spanish region of Catalonia. Following a similar rationale which aimed at lowering federal redistribution, the high-income Canadian province of Alberta held a referendum in 2021 debating whether equalisation payments should be eliminated from the Constitution.

Accommodating Diversity: Is Menu-Federalism the Solution?

This chapter has examined the existence of several kinds of diversity and the potential ability of asymmetric federal arrangements to accommodate them. However, asymmetric arrangements also carry risks. In this context, the key challenge lies in how to make use of asymmetry for diversity accommodation in a way that is fair. Fiscal federalism institutions have two tools that could contribute to the solution: fiscal equalisation and menu-federalism.

Fiscal equalisation represents the main instrument for horizontal federal redistribution. This transfer programme aims to ensure that disparities in tax capacity and needs across territories do not translate into large differences in access to and quality of local public services. In addition, to correct efficiency-related issues—such as intergovernmental spillovers or excessive tax competition—fiscal equalisation transfers could ensure that the consequences of asymmetries in fiscal decentralisation do not reach levels that are harmful for political stability. Indeed, even if different arrangements would be applied in different jurisdictions, disparities in results would be nuanced, making asymmetries compatible with the perception of common regime territories that the system is fair.

However, there are some concerns about the usefulness of fiscal equalisation. First, asymmetries could also alter the redistributive power fiscal equalisation arrangements have. Second, fiscal equalisation is commonly contentious. Finally, even if fiscal equalisation achieves equitable results in terms of access to and quality of local public services, that does not necessarily imply that perceptions will match reality. In fact, as it is quite common to mix up fiscal equalisation’s actual objectives with those of regional policy—namely to decrease territorial economic disparities—(Dougherty et al., 2022), a successful fiscal equalisation programme that operates in a country with large economic disparities may be wrongly perceived as ineffective. Regional policy might emerge as a viable alternative to fiscal decentralisation institutions as a whole, because this arrangement would be perceived as being “fairer”, despite the existence of asymmetries.

Menu-federalism involves designing fiscal federalism arrangements à la carte for each subnational jurisdiction. For example, it allows each SNG to unilaterally decide on the desired level of tax autonomy, and thus on the degree of transfer dependence desired. Interestingly, Bell et al. (2021) propose a creative approach to menu-federalism that combines autonomy-on-demand with fiscal co-responsibility. This means that incentives are employed to discipline the use of autonomy by requiring that SNGs face the consequences of their fiscal policy. This way, menu-federalism would allow for the existence of asymmetrical fiscal arrangements, without incurring the risk that the public would perceive these arrangements as discriminatory. In fact, all jurisdictions would have the possibility to devise their own arrangement, thereby reinforcing the legitimacy of the model.

These alternatives demonstrate that, overall, the existence of asymmetries does not necessarily have to be regarded as undesirable, as there are several ways to prevent redistributive side-effects. If asymmetries are coupled with fiscal equalisation, regional policy, or menu-federalism, they could preserve systemic fairness. Despite this, it must be noted that asymmetries are likely to make federalism a more complex institutional framework.

As this chapter notes, more data on diversity and asymmetries is needed. Otherwise, it is impossible to produce empirical evidence on this relationship. A novel approach could depart from the current index on fiscal constitutions by extending the current methodology used to determine the degree of decentralisation to the subnational level. Comparing variance across jurisdictions could provide a glimpse of the extent to which asymmetric arrangements are in place. This exercise could be complemented by adding new indicators to calculate the existence, relevance, and depth of differential treatment given by fiscal federalism frameworks to certain subnational jurisdictions. Asymmetries would thus be detected where tax and spending autonomy differs across territories, or where special/charter regimes were reported. The more pronounced these heterogeneities and the greater the number and size of jurisdictions affected, the more significant the asymmetries are.