Decentralisation is often subject to heated debate. The proponents of decentralisation tend to emphasize the pros and the critics often highlight the cons. Fortunately, after several decades of practical policy implementation, a considerable amount of information has accumulated on the effects of decentralisation. This chapter discusses the benefits and challenges of decentralisation using the best available information on research results and practical policies pursued in various OECD countries. After brief introduction, the chapter begins by describing the benefits and opportunities and then continues to examine the risks and challenges of decentralisation. Both sides of decentralisation are discussed from economic, administrative and political aspects. In the end of the chapter, a conclusion and a short summary table of the main effects are presented.
Making Decentralisation Work
Chapter 4. Decentralisation: Its benefits and challenges
Abstract
The statistical data for Israel are supplied by and under the responsibility of the relevant Israeli authorities. The use of such data by the OECD is without prejudice to the status of the Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of international law.
As highlighted in Chapters 1 and 2, the trend in decentralisation continues worldwide (Hooghe et al., 2016[1]). Moreover, practices in various countries show that the implementation of decentralisation has evolved as new challenges have emerged. Some examples of this are the urbanisation and regionalisation trends. Also, the economic crisis that began in 2008 and the austerity and consolidation measures that followed have affected fiscal decentralisation in many significant ways. This chapter provides a brief discussion of the benefits and challenges of decentralisation, as well as the results of recent empirical research on the topic.
Practical experiences of decentralisation reforms in various countries have accumulated over several decades, with researchers and practitioners debating their effects. One major aspect of this discussion has been the effect of decentralisation on economic growth and well-being. While country statistics do not enable strictly causal conclusions, in recent years it has been found that subnational fiscal power can be positively associated with economic activity (Blöchliger, 2013[2]). Although correlation does not mean causality, it is interesting to note that especially measures such as gross domestic product (GDP), public investments made in physical and human capital, and education outcomes show a positive correlation with decentralisation. Revenue decentralisation appears to be more strongly associated with income gains than spending decentralisation (Blöchliger, 2013[2]). In addition, examples and research results from several countries show that decentralisation can be conducive to public sector efficiency, democratisation and political stability (Faguet, 2014[3]; Ivanyna and Shah, 2014[4]). These effects, however, depend largely on the way decentralisation is designed and implemented. It should be emphasised that decentralisation reform is always ultimately a political choice. Such reforms should always be implemented as part of a larger political reform process, making sure that the judiciary, civil service and regulatory frameworks are capable of dealing with any associated change.
There is an extensive economic theory and empirical research on multilevel governance. The so-called “first generation” economic literature of fiscal federalism, developed in the 1950s and 1960s, emphasised the benefits of decentralisation. The optimistic view of decentralisation was based on the theoretical assumption of the benevolence of public decision-making, or perhaps more realistically, the positive outcomes resulting from electoral pressures in a democratic government system (Ahmad and Brosio, 2006[5]; Oates, 2005[6]). Many decentralisation reforms implemented during the 1970s, 1980s and 1990s were founded on the early fiscal federalism principles. Another important motivation for decentralisation was the desire to advance democratisation processes. The “second generation” fiscal federalism literature, which began to emerge in the late 1990s, mostly relaxes the assumptions of previous research. A more critical and nuanced view of decentralisation is now offered by public choice and political economy research, for instance by allowing for self-interested motives of public decision-making (Weingast, 2014[7]; Lockwood, 2002[8]). The second‑generation literature argues that the first generation literature must be complemented by political economy aspects to give a rigorous account of the preference-matching and accountability benefits of decentralisation (Oates, 2008[9]). Another important strand of second-generation fiscal federalism literature has been motivated by the fiscal crises caused by intergovernmental fiscal behaviour in some Latin American and European countries. This line of research emphasises the dangers of impartial or unbalanced decentralisation reforms and discusses conditions under which decentralised countries could ensure fiscal discipline (Rodden, Eskeland and Litvack, 2003[10]). Moreover, the research focusing on regional development has studied the links between decentralisation and regional development and effects on regional disparities (Ezcurra and Rodríguez-Pose, 2013[11]; Tselios et al., 2012[12]).
Many countries claim that they are decentralised but the reality may be very different. Different measures of decentralisation can lead to very different conclusions regarding degrees of decentralisation (Hatfield, 2015[13]). The way decentralisation is designed and implemented has a major impact on its outcomes. For example, many researchers stress that implementing “partial decentralisation” may lead to unexpected effects and loss of potential benefits. A typical example of partial decentralisation is a policy where spending tasks are widely delegated to subnational governments, but where subnational own-revenues remain highly limited. It should also be noted that spending or revenue decentralisation alone cannot secure or guarantee all the benefits associated with decentralisation. At the same time, adequate capacity of subnational governments, accountability of local public decision-making and good overall governance are required to support positive outcomes of fiscal decentralisation (OECD, 2013[14]; 2017[15]; Allain-Dupré, 2018[16]; Kim and Dougherty, 2018[17]).
The benefits and challenges of decentralisation can be direct or indirect. Direct effects include changes in service levels, plus the quality and efficiency of public services. Indirect effects of decentralisation, such as effects on economic growth or societal stability, result from direct outcomes of decentralisation, such as better education or higher participation in political decision-making. Since the indirect effects of decentralisation are affected by a variety of factors, the role of decentralisation is, of course, harder to separate from other trends and policies.
Opportunities and benefits
There are various economic, political and other reasons behind the upsurge of decentralisation reforms, and the reasons vary considerably across countries (see Chapter 2). For instance, in some countries, decentralisation can be seen as a counter-reaction to previous strong centralisation and even autocratic trends (Hooghe et al., 2016[1]). In these countries, decentralisation has been a way to ensure that democratisation will not be reversed. In other countries, decentralisation has been a method to reform the public sector, for example in order to improve the efficiency of public services, thereby curbing the growth of government spending. As for other motives, decentralisation is often thought to deliver positive effects, such as more accountable and transparent public governance, lower corruption, higher political participation and policy innovation. These benefits are discussed below.
Economic benefits
Decentralisation may facilitate tailoring services to local needs
Perhaps the most important direct benefit of decentralisation is the allocative efficiency resulting from local public service provision. Subnational governments often hold valuable information on local demands and conditions, which enables them to tailor public service provision to meet residents’ needs. Decentralised choice, therefore, gives an opportunity to increase economic welfare by adapting public service provision to the heterogeneous preferences of smaller population groups (Wallis and Oates, 1988[18]). It would be costly for the central government to obtain such information and therefore the central level is often likely to provide a uniform level of public output in all jurisdictions. In the case of heterogeneous preferences and local public goods, subnational governments clearly outperform central government in allocating public sector resources (Oates, 2005[6]).
It is important to note that the highest allocative advance from subnational service provision is usually obtained in the case of “local public services”, which consist of public tasks that have a spatially limited area of benefits.1 Ideally, the benefit areas should match with administrative boundaries of local jurisdictions. Public services with major externalities or services with important economies of scale are less suitable for local provision, although in these cases allocative benefits may also be obtained. In the case of redistributive services, such as education or healthcare, the co‑ordination responsibility is usually maintained with a higher level of government (Box 4.1).
Box 4.1. Assigning allocation, redistribution and stabilisation tasks: Fiscal federalism principles
The traditional fiscal federalism literature divides public tasks into three branches: allocation, redistribution and stabilisation functions (Musgrave and Musgrave, 1980[19]). According to this categorisation the allocation function – i.e. public services provision – can be the responsibility of both the central and subnational levels of government. In allocation, the central level of responsibility is best applied when the services have no specific local interest. Subnational responsibility is justified when the benefits of the goods or services are spatially limited. Moreover, according to the “Decentralisation Theorem” (Oates, 1972[20]), the subnational level is the most suitable level to provide the services and goods, unless the central government has a clear advantage in provision. This could, for example, be where there are considerable economies of scale in the provision.
According to the Musgravean distinction, redistribution and stabilisation functions are mostly central government responsibilities. In particular, the central government is considered best suited to deal with monetary or fiscal policy (Musgrave and Musgrave, 1980[19]). It is also widely accepted that the redistribution function should be mostly central level responsibility: the central government is more capable of carrying out income redistribution from the wealthy to the poor and in establishing minimum standards of public services across regions (King, 1984[21]).
“Pure local goods”, such as local infrastructure (street lights, local roads), sewage, land use planning or basic education, are usually considered best suited for subnational government provision. It is nevertheless quite common that subnational governments are also involved in the provision of services with redistributive features, at least in some way. In some countries, such as the Nordic countries, even health, education and welfare services have been delegated from the centre to subnational governments. In the case of a decentralised redistribution, the central government usually retains responsibility for co‑ordination and ensuring equity of citizens in different parts of the country. This can be achieved for example by using transfers from central to subnational governments, or with normative regulation (minimum standards), or both.
Sources: Boadway, R. and J. Tremblay (2012[22]), “Reassessment of the Tiebout model”, http://dx.doi.org/10.1016/j.jpubeco.2011.01.002; Musgrave, R. and P. Musgrave (1980[19]), Public Finance in Theory and Practice, McGraw Hill Kogahusha; Tiebout, C. (1956[23]), A Pure Theory of Local Expenditures, http://www.jstor.org/stable/1826343Accessed:28/07/200805:58; Oates, W. (1972[20]), Fiscal Federalism, Harcourt Brace Jovanovich, New York; King, D. (1984[21]), Fiscal Tiers: The Economics of Multi‑level Government, Allen and Unwin; Allain‑Dupré, D. (2018[16]), Assigning Responsibilities across Levels of Government: Trends, Challenges and Guiding Principles for Policy-makers, OECD, Paris.
Decentralisation may improve the efficiency of public service delivery
Compared with centralisation, decentralisation contributes to better accountability because it alters the incentives of the authorities who serve local populations. In a decentralised model, the elected local authorities are accountable to residents who finance and consume the services. In a centralised public service provision system, the administrators are not accountable to local residents but instead, they report to their superiors in the central government. Decentralisation reverses the accountability chain from a “top-down” to a “bottom-up” relationship and therefore affects the motives of all stakeholders.
Decentralisation allows for many types of political and fiscal competition, which can be efficiency enhancing. From a political perspective, decentralisation increases the number of political arenas and therefore lowers the entry cost for new political candidates. These enlarged political fora give citizen-voters more choice in elections. As a result, voters are better able to express their preferences on service delivery and to inform politicians about problems at the local level.
While local and national elections provide the main channels for citizens to influence policies and express their “voice”, it is important to make sure that there are also other forms of citizen participation (Inman and Rubinfeld, 1997[24]). Such alternative ways include participation in surveys, town meetings, local referenda and direct involvement in service delivery (Azfar et al., 1999[25]).
Increased participation and engagement can contribute to local ownership of public policy programmes. In addition, higher overall citizen participation can lead to tighter political competition not only at the local level but also in national elections. Increased political competition can lead to better overall policies and more efficient implementation of government programmes.
Decentralisation may enhance competition not only within local jurisdictions but also between jurisdictions. The underlying mechanism is the “exit” (or threat of exit) of taxpayers from their current jurisdictions. The more the subnational governments rely on revenues based on mobile resources, the more likely there is to be competition between jurisdictions. This is reflected in the traditional “voting with the feet” model of decentralised government (Tiebout, 1956[23]). Although in reality, the mobility of households and companies is far from perfect, the threat of losing taxpayers creates additional pressure for elected local authorities to ensure that the services match the local demands with a competitive tax burden (Oates, 2005[6]). Such competition is usually assumed to result in a better match between service provision and local preferences, generating a more efficient allocation of resources (Box 4.2).
“Yardstick competition” is a form of competition that does not involve mobility. It occurs when voters can compare the taxes and service quality in their own jurisdiction with those in neighbouring jurisdictions. If voters observe that the service-tax mix is better in the neighbouring jurisdictions (or other similar communities), they punish the elected representatives in their community by voting against them in the next local elections. Yardstick competition can be efficiency improving if voters can truly identify and re-elect politicians whose “type” is “good” rather than those whose type is “bad” (Ahmad and Brosio, 2006[5]). Openly available information on outcomes can foster such benchmarking (Weingast, 2014[7]; Faguet, 2014[3]).
These basic accountability mechanisms work best if local residents have a strong incentive to evaluate the efficiency of their local administration. Such motivation depends primarily on the financing system of locally provided public services and on information available on the service outcomes. If local residents finance a considerable share of local services by paying local taxes, they will have a strong incentive to monitor their local administration.
Box 4.2. Tiebout model and mobility
Charles Tiebout’s famous 1956 paper, A Pure Theory of Local Expenditures is one of the earliest contributions to the modern theory of fiscal federalism and decentralisation. The model was not intended as a complete theoretical description of decentralised government. Instead, the purpose was to demonstrate a mechanism by which voters’ preferences for public services could be revealed by “exit” rather than “voice”. Nevertheless, the concept of “voting with one’s feet” of his model has been later used widely as a theoretical assumption when modelling decentralisation.
In Tiebout’s model, the perfect mobility of households will lead to an optimal allocation of households among local communities. Musgrave and Musgrave described the Tieboutian competing communities as follows: “Those who like sports will want to reside with others who are willing to contribute to playgrounds. Those who like music will join the others who will participate in building a concert hall, and so forth. Each community will do its own thing, and everybody will be happy”.
The assumptions behind Tiebout’s model are strict. In addition to the perfect mobility of households, the assumptions of the model include local public goods, endogenous number of communities and benefit taxation. The model has only one level of government and there is no transfer system.
Since the times of Tiebout’s paper, fiscal federalism theory has been further developed, and many of the strict assumptions made by Tiebout have been relaxed. As Broadway and Tremblay note, fiscal federalism is really about outcomes in a world with more than one level of autonomous government. In reality, the mobility of households is far from perfect, which weakens the effect of competition between communities. The tasks provided by subnational governments may vary less than assumed by Tiebout, because state, regional and municipal levels of governments often provide services that have been delegated from the central level.
Oates argues that the gains from fiscal decentralisation do not depend on the mobility of households. Potential welfare gains from decentralisation would exist even without mobility due to a subnational government’s ability to tailor services to local demands and circumstances. Mobility, if it happens, would, however, strengthen the benefits expected from decentralisation.
Sources: Tiebout, C. (1956[23]), A Pure Theory of Local Expenditures, http://www.jstor.org http://www.jstor.org/stable/1826343\ ;Boadway, R. and J. Tremblay (2012[22]), “Reassessment of the Tiebout model”, http://dx.doi.org/10.1016/j.jpubeco.2011.01.002; Musgrave, R. and P. Musgrave (1980[19]), Public Finance in Theory and Practice, McGraw Hill Kogahusha; Oates, W. (2008[9]), “On the evolution of fiscal federalism: Theory and institutions”, National Tax Journal, Vol. 61/2.
Decentralisation may enhance economic growth
Several theoretical and empirical studies have been carried out to test the link between growth and decentralisation. Theoretical research on this topic can be grouped roughly into three aspects: increased competition, better preference matching and enhanced accountability (Hatfield, 2015[13]). While the results of theoretical work are somewhat mixed, the general conclusion is that decentralisation can be conducive to growth when it is properly designed and implemented (OECD, 2016[26]), and especially if competition between jurisdictions is allowed to work in a way that promotes efficiency, accountability and preference matching at the subnational level.
There is extensive empirical research on the effects of decentralisation on growth (Hatfield, 2015[13]). The largest group of empirical work consists of studies that use cross-country data to regress a measure of economic growth on measures of decentralisation, such as local revenue share or local expenditure share. For instance, recent studies analysing data from OECD countries find that subnational fiscal power is positively associated with economic activity (Blöchliger, 2013[2]; OECD, 2016[26]). According to the study, the positive impulse seems to stem both from productivity and human capital improvements. Decentralisation can, for instance, improve the efficiency and productivity of the public sector, which in turn may contribute to higher productivity in the private sector. Decentralisation could also result in more educational investment and enhanced human capital, both of which are important factors behind economic growth (Blöchliger, Égert and Fredriksen, 2013[27]).
Investment in physical and human capital as a share of general government spending is significantly higher in more decentralised countries. It has been estimated that on average, a 10 percentage point increase in decentralisation is associated with 3% to 4% higher share of investment in total government spending (Blöchliger, Égert and Fredriksen, 2013[27]). The relationship is stronger for investment in human than physical capital and stronger for revenue than for spending decentralisation.
Recent research results suggest that revenue decentralisation, in particular, is related to growth (Blöchliger, 2013[2]; OECD, 2016[26]). The link with spending decentralisation and economic growth seems to be less clear, however. This empirical finding may reflect that “true” fiscal autonomy is better captured by the subnational revenue share (instead of the spending share), as a large part of subnational spending may be mandated or regulated by central government (Blöchliger, 2013[2]).
The second group of studies focuses on how variation in the local share of government revenues and expenditures across provinces or states affects outcomes within a single country. The third group of empirical research examines the effect of inter-jurisdictional competition by using the number of jurisdictions within a geographic unit as a measure of competition. The results of the second and third groups of empirical research are mixed. Moreover, because of the endogenous relationship between growth and decentralisation, it not possible to draw causal conclusions. That said, one general result of the empirical studies seems to rise above others: it is important that a certain level of taxes be levied at the local level for fiscal decentralisation to contribute to economic growth (Weingast, 2014[7]; Hatfield, 2015[13]; Ivanyna and Shah, 2014[4]; Blöchliger, 2013[2]). While a general rule for the optimal degree of tax autonomy is obviously difficult to define, it is usually argued that local authorities should rely on their own revenues for financing their services at the margin. Such a principle would help ensure that decisions to expand public programmes are made keeping in mind the additional costs (Oates, 2008[9]).
Decentralisation may contribute to lower regional disparities
Recent empirical evidence indicates that revenue decentralisation could be associated with smaller regional economic disparities (Bartolini, Stossberg and Blöchliger, 2016[28]). Although correlation does not imply causality, a possible explanation is that own-source revenue may spur growth especially in poorer regions and enhance the convergence process towards the best performing regions (Blöchliger, Bartolini and Stossberg, 2016[29]) (Figure 4.1). In some countries such as Mexico, decentralisation is motivated by the fight against poverty and territorial disparities (Faguet, 2014[3]).
Figure 4.1. Subnational government tax autonomy tends to be associated with lower regional GDP disparities
Note: The sample covers 20 OECD countries (19 for tax autonomy) and the years 1995 to 2011. Each point reflects the coefficient of variation of regional GDP per capita in one country in one year. The lines indicate the results of a bivariate ordinary least squares (OLS) regression. Variables are normalised using the respective country means to net out differences between countries that are persistent over time.
Source: Blöchliger, H., D. Bartolini and S. Stossberg (2016[29]), Does Fiscal Decentralisation Foster Regional Convergence?.
High share in intergovernmental transfers of total subnational government revenues has been found to correlate with disparities in regional GDP per capita. Fiscal autonomy and reliance on own-source revenues, therefore, appear to help the catching-up regions more than those above the national average (Blöchliger, Bartolini and Stossberg, 2016[29]). This result is supported by a recent study by Tselios et al., who observed that decentralisation is positively associated with smaller interpersonal inequities at the regional level. The relationship seems to weaken as overall income rises (2012[12]). These results do not necessarily mean that a causal relationship exists between decentralisation and regional disparities (Box 4.3). More research is certainly needed on this topic.
While incentives for developing own source revenues at the subnational government level contribute to regional growth policies, many subnational governments nevertheless need substantial central government financing to provide the services they are assigned (Shah, 2017[30]). Transfer systems are, therefore, often designed to equalise both cost differences and differences in revenue base. Well-designed equalisation systems can help ensure reasonably comparable levels of public services at comparable burdens of taxation. At best, equalisation can contribute to economic growth by creating a level playing field, which facilitates inter-regional movement of labour and business in response to economic stimuli, but slows down their movements in response to fiscal considerations alone. Equalisation transfers support national/state objectives in creating a common economic and social union.
Box 4.3. Linking decentralisation and development: Some evidence at world level
At the world level, the 2016 OECD-UCLG study shows that the wealthiest countries tend to be more decentralised. The correlation between the GDP per capita (measured in USD PPP) and the level of spending decentralisation (measured by the share in GDP) is again positive (Figure 4.2).
Figure 4.2. Decentralisation of spending responsibilities is a feature of development at world level, 2014
Note: Luxembourg is not represented on the graph as it is an extreme case due to its high GDP per capita.
Note by Turkey: The information in this document with reference to “Cyprus” relates to the southern part of the Island. There is no single authority representing both Turkish and Greek Cypriot people on the Island. Turkey recognises the Turkish Republic of Northern Cyprus (TRNC). Until a lasting and equitable solution is found within the context of the United Nations, Turkey shall preserve its position concerning the “Cyprus issue”.
Note by all the European Union Member States of the OECD and the European Union: The Republic of Cyprus is recognised by all members of the United Nations with the exception of Turkey. The information in this document relates to the area under the effective control of the Government of the Republic of Cyprus.
Source: OECD-UCLG (2016[31]), Subnational Governments Around the World: Structure and Finance, http://www.oecd.org/regional/regional-policy/sngs-around-the-world.htm.
Decentralisation can be a lever for regional development
Another potential benefit of decentralisation is the ability to carry out better regional development policies (Morgan, 2006[32]). To mobilise the regional productivity catch-up potential as well as to ensure that growth and productivity will be more balanced and inclusive across the territory, an efficient multi-level governance system is therefore required – one that is based on an enhanced role for subnational governments, especially regions, capable of designing and implementing context-sensitive interventions (OECD, 2017[33]).
The links between decentralisation and regional development may at least partly explain why the regionalisation process was so strong in the Central and East European countries that were planning to join the European Union. Even if the EU did not promote a particular model of decentralised governance, the prospect of entering the EU led several countries to create self-governing regions to access and manage EU funds for regional development.
Regional development was also a strong motivation of the first decentralisation reforms in Japan, in particular, to correct the excessive population concentration in Tokyo. In the creation of new regions, many saw the need to correct interregional disparities, to give regional and local actors the means to implement regional development policies adapted to the new economic realities (functional areas of prefectures were outdated), and to and realise economies of scale in terms of infrastructure facilities and services (OECD, 2016[34]; 2017[15]). In Chile, the regionalisation process that created self-governing regions with directly elected governors was linked to the need to correct high regional disparities (OECD, 2017[33]). In Chile, “hyper-centralisation” has gone hand in hand with a “hyper-concentration” of population, resources and powers in the metropolitan region of Santiago. In Korea, addressing the regional imbalance between Seoul and surrounding regions is also at the core of the decentralisation programme launched in 2017-18 (Annex C).
Political benefits
Decentralisation may strengthen citizen participation in government
Decentralisation has the potential to support and expand citizen participation by bringing government closer to citizens and by making government more accessible. Citizen participation is an essential part of successful decentralisation. High voter turnout in local elections helps to ensure that spending reflects the preferences of residents. Active citizen engagement also supports the accountability of local public decision-making.
Political participation has traditionally been an important justification for decentralisation reform. This was the case in many former communist countries such as Poland (OECD, 2008[35]; Regulski and Drozda, 2015[36]), and many developing countries (e.g. Bolivia, Cambodia, Peru and Uganda) (Faguet, 2014[3]). Recent empirical evidence from both developed and developing countries has shown that there is a positive association with decentralisation and political participation (Stoyan and Niedzwiecki, 2018[37]; Michelsen, Boenisch and Geys, 2014[38]). Moreover, an earlier study using data from 80 counties also found a positive relationship between fiscal decentralisation and citizen participation (Huther and Shah, 1998[39]).
Decentralisation may improve political stability
Decentralisation can be a “glue” that holds countries together (Bird, 2003[40]). By decentralising powers to regions and subnational governments, the tensions arising due to various cultural, historical or political reasons may be mitigated. This can also happen because decentralisation might be asymmetric, thus making it easier to take into account certain territorial specificities.
Since decentralised governments are well positioned to offer services suited to the local needs and preferences, the destabilising demands promoted by separatist movements may be alleviated. It has also been argued that a decentralised system may promote the rise of types of political leaders who are willing to work co-operatively within the state (Faguet, 2014[3]).
Since decentralisation generates a larger number of political arenas and government layers (compared with a centralised model), pressure on candidates to win elections at the national level at any cost may be diminished. This can reduce the overall political tensions. In addition, decentralised systems usually result in a higher number of independent actors in government. This can create stability because rules, laws and policies cannot be easily and frequently changed (Faguet, Fox and Pöschl, 2014[41]).
Administrative benefits
Decentralisation may constrain rent-seeking and corruption
Decentralisation may diminish the opportunities for rent-seeking and corruption in public administration (Boadway and Tremblay, 2012[22]). There are alternative and complementary explanations for why this might happen. One explanation is simply that in a decentralised setting the scale of government is smaller. This reduces the size of rents available and makes rent-seeking less interesting. An alternative explanation is that the enhanced competition fostered by decentralisation reduces opportunities for rent-seeking and corruption. In other words, decentralisation brings a local aspect to lobbying, and this reduces the “monopoly power” of national level rent-seeking (Bordignon, Colombo and Galmarini, 2008[42]).
Empirical research provides some support to these theoretical arguments. For instance, a larger subnational share of public expenditures has been found to correlate with lower corruption (De Mello and Barenstein, 2001[43]). Using cross-country data on governance and fiscal indicators from 78 countries, the study found that in particular, a higher local government share of public revenue2 was associated with lower corruption. These results may depend on the decentralisation model that was chosen and on its implementation. For example, very complicated multi-level governance models with several government tiers and unclear assignments of responsibilities have been found to be more prone to corruption (Fan, Lin and Treisman, 2009[44]). Using data on a survey of business managers conducted in 80 countries3 and data on several fiscal indicators including the number of tiers of government, the study found that in countries with a larger number of government tiers, reported bribery was both more frequent and costlier to firms. According to the study, the degree of a country’s development affects the relationship between governance and corruption. In developed countries, the association was weaker than in developing countries (Fan, Lin and Treisman, 2009[44]).
Decentralisation enables experimenting and policy innovation
Decentralisation provides a useful platform for experimenting with public policies. At best, such a “learning by doing” process of decentralised policy innovation can result in important information spillovers from good practices. The “information externalities” created by decentralisation can benefit not just subnational governments themselves but also central government.
Subnational innovation activities are best motivated if local jurisdictions are responsible not only for spending but also for raising financing. Innovation activities at the subnational government level can be further enhanced if systematic frameworks are created to support and encourage subnational governments to introduce their own programmes (Oates, 2008[9]). Yardstick competition between subnational governments can foster the adoption of the most effective methods.
In many countries, central governments have adopted practices that were first implemented in the best performing subnational governments. Oates describes several examples of such cases from the United States (Oates, 2008[9]). For instance, unemployment insurance, gasoline taxation and environmental regulation were state-level policies before the federal government introduced similar measures or standards for the whole country. There are numerous such examples from all over the world, including Participatory Budgeting which was first implemented in the city of Porto Alegre, Brazil, and is now practised in several countries (Campbell and Fuhr, 2004[45]). Such “laboratory federalism” may enhance the efficiency and quality of public policy in general.
Some countries have adopted bottom-up models of experimental governance, i.e. state-sponsored experimentalism. For example, governments are increasingly receptive to and supportive of public sector innovation hubs to promote territorial development and public service reform. The UK innovation foundation, NESTA, is one of the most prominent pioneers of public and social labs dedicated to addressing societal challenges through evidence-based local experiments (Morgan, forthcoming[46]).
Risks and challenges
Despite the numerous potential benefits of decentralisation, there are also potential risks in such reforms that need to be properly addressed. Some of these challenges relate to decentralisation in general but problems may also arise because of partial or unbalanced implementation of decentralisation processes. The question is not whether decentralisation is good or bad in itself – it is rather under which conditions decentralisation can be conducive to regional development and citizen engagement. The outcomes of decentralisation reforms depend extensively on the way decentralisation is conceived and put in place.
Decentralisation presents a challenge for subnational governments because it requires certain economic, political and administrative capacities. Unless the capacity challenge is addressed, there is a risk that decentralisation intensifies differences between jurisdictions in a way that could jeopardise equal access and service quality. Partial decentralisation, such as high subnational government spending responsibilities combined with strong normative regulation and low revenue-raising autonomy, could put at risk the subnational government ability to adjust public-good levels to suit local demands. A high reliance on central government transfers may reduce a subnational government’s incentives for responsible fiscal behaviour. Unbalanced decentralisation, i.e. situations where tasks that are closely linked or complementary are decentralised to varying degrees, can weaken multi-level governance. These aspects and other challenges of decentralisation are briefly discussed below.
Economic challenges
Lack of resources and underfunded mandates
One of the most frequent challenges is the misalignment between responsibilities allocated to subnational governments and the resources available to them. The most extreme cases, unfunded mandates, where subnational governments have the responsibility to provide services or manage policies but without the requisite resources, are common.
The fiscal dimension is very often the weak, or even missing, link of decentralisation. The transfer of spending responsibilities should normally involve transferring equivalent resources to subnational governments in order to allow them to perform their new tasks correctly. This can be done through central government transfers (general or earmarked), or through the allocation of a share of national taxes such as the personal income task (tax sharing arrangements). It can also be accomplished through the right to levy own-source revenues, such as local taxes and user charges and fees, and to raise revenues from subnational financial and physical assets (e.g. natural resources, dividends from local companies, sales of property assets, etc.). This is the “finance follow functions” principle, also called the “connection” or “matching principle”.
In practice, there are often imbalances between the assignment of spending responsibilities and the assignment of revenues, resulting in unfunded or under-funded mandates.
Risks of partial decentralisation: Lack of fiscal autonomy
Lack of fiscal autonomy for subnational governments to exert their responsibilities forms a major challenge to decentralisation. However, a minimum degree of fiscal autonomy is required to allow subnational governments to make better use of public resources, targeted to local needs. Fiscal decentralisation is not only about assigning tasks to subnational governments, but also about granting some autonomy to the subnational government to raise and manage resources.
While in a strict sense decentralisation means devolving true spending and taxing powers from central to subnational governments, in reality, decentralisation is often implemented in a way that satisfies only part of that definition. For instance, seen from the fiscal decentralisation perspective, spending is usually much more decentralised than revenues (OECD, 2018[47]). This is the case in all OECD countries (Figure 4.3). If subnational governments have little freedom to choose the levels of local public goods, especially when transfers are accompanied by mandates that specify how the money is to be allocated across spending categories, or if service provision is tightly regulated with norms and laws, the decentralisation is only partial (Brueckner, 2009[48]; Borge, Brueckner and Rattsø, 2014[49]). This is especially the case in developing countries but it is also the case also in many developed countries. In particular, own revenue-raising capability by subnational governments is often limited, for instance, because central government has reserved the most valuable tax bases for central taxing only. Central governments also commonly regulate subnational government tax bases and tax rates. Furthermore, if subnational service provision is strongly steered by normative regulation, also the spending autonomy is limited. Implementing such “partial decentralisation” could lead to unwanted effects and loss of potential benefits of decentralisation.
Figure 4.3. Vertical fiscal imbalances in the OECD, 2016
Note: Australia and Chile: estimates from IMF Government Finance Statistics. 2015 data for Mexico, New Zealand and Turkey.
Source: OECD elaboration based on OECD (2018[47]), Subnational Governments in OECD Countries: Key Data (brochure and database), http://dx.doi.org/10.1787/region-data-en.
Partial decentralisation can have a variety of undesirable effects on subnational government incentives and decision-making. The main disadvantage resulting from highly restricted subnational spending and revenue autonomy is the weakened ability of subnational governments to allocate public resources according to local demands and local conditions. According to recent theoretical research, policies that limit subnational government taxing power weaken allocative efficiency (Brueckner, 2009[48]).4 In addition, a carefully implemented empirical study using data from Norway found that dropping central regulations on subnational spending resulted almost immediately in important changes in local service provision. This suggests that there was a clear demand for local discretion (Borge, Brueckner and Rattsø, 2014[49]).
A high reliance on central transfers can incentivise subnational governments to overspend and to incur debt especially if subnational governments operate under soft budget constraints. Soft budget constraints may develop if subnational governments expect that central government will bail them out in the case of economic troubles or insolvency (Bordignon, Giglio and Turati, 2015[50]). The soft budget constraint problem is exacerbated if subnational governments are responsible for services that are of national importance, like healthcare. This is because in these cases central governments may find it politically difficult to commit to a no-bailout policy. In addition, some subnational governments, such as the largest cities, can be too important for the central authorities to allow bankruptcy (the “too big to fail” argument). At worst, subnational government fiscal problems coupled with soft budget constraints can destabilise the entire economy (Rodden, Eskeland and Litvack, 2003[10]). In this respect, the experience of some Latin American countries (especially Argentina and Brazil) is usually mentioned. In these countries, powerful provincial authorities were able to run major fiscal deficits and accumulate large debts up to a point that led to the bailout of provinces by the central government, and eventually to a national financial crisis (Oates, 2008[9]).
The importance of grants and shared taxes in subnational funding results in a significant vertical fiscal imbalance, i.e. the difference between subnational governments own revenues (own-source tax, user charges and property income) and spending obligations. To measure vertical fiscal imbalances, a proxy comparing the level of tax revenues and that of spending can be used.
Benefit spillovers and inadequate scale
While the traditional theoretical assumption is that a jurisdiction’s administrative boundaries coincide with the service benefit areas (the so-called “fiscal equivalence principle”), in practice benefit spillovers across jurisdiction borders form a common problem of decentralised service delivery. Benefit spillovers can lead to under-provision of public services if subnational governments do not take into account the benefits received by residents in other jurisdictions. Central government may intervene in these situations by paying matching grants to subnational governments to encourage extended service delivery. Other solutions to under-provision and spillovers include mergers of the smallest subnational governments and enhanced co-operation between local jurisdictions. These reforms can be voluntary, “bottom-up” reforms or centrally-led, “top-down” restructurings. The issue of benefit spillovers is not an easy challenge to solve, however, not least because each public service has a different optimal benefit area and internalising externalities in an existing structure is often difficult.
Decentralisation may result in loss of certain economies of scale (Oates, 1985[51]). This could happen if subnational governments are unable to co-operate with each other or to outsource production to larger, neighbouring subnational governments or private companies. The risk of inefficient outcomes is smaller if subnational governments are responsible for both the spending on and financing of services because subnational self-financing provides a strong incentive to select the most efficient service production technology. In countries with small subnational government units, such as the Nordic countries, France, the Slovak Republic and Spain, subnational partnerships, joint authorities and outsourcing are frequently used to utilise scale economies. Municipal merger reforms are also often on the political agenda, especially in decentralised countries, to respond to economies of scale and benefit spillover problems.
Determining optimal subnational unit size is a context-specific task; it varies not only by region or country but by policy area, as well. The efficient size differs between waste disposal, schools or hospitals. In Finland, research on scale benefits of expanding the size of local governments found that large municipalities were less efficient at service delivery and the optimal size was between 20 000 and 40 000 inhabitants (OECD, 2017[15]; Moisio, Loikkanen and Oulasvirta, 2010[52]). Yet in Japan, unit costs of public services bottomed out at about 120 000 inhabitants and increased at both higher and lower municipal sizes (OECD, 2017[15]).
Municipal mergers may be considered effective in larger conurbations with a high degree of municipal fragmentation, but ineffective when the size of municipalities is already large or in remote areas where service delivery is largely determined by geography. Often, efficiency gains can be made without abolishing tiers or merging municipalities (OECD, 2017[15]).
Decentralisation may increase disparities between subnational governments
Without specific measures that strengthen the capacity of poor regions, the benefits of decentralisation may accrue only to the most developed and prosperous regions, hence contributing to increased regional disparities. There can be important differences between subnational governments in a financial capacity and administrative skills, which can endanger the ability of lagging regions to catch up. It has also been argued that decentralisation may foster agglomeration effects, which could increase disparities (Bartolini, Stossberg and Blöchliger, 2016[28]). These results may depend on the type of country, and its level of economic development. According to some studies, decentralisation may reduce regional disparities in high-income countries, while increasing them in low-income countries (Ezcurra and Rodríguez-Pose, 2013[11]). On the other hand, these results also show that when the inefficiencies within a system are properly addressed by decentralisation processes, interpersonal inequality could decrease faster in less developed than in more developed areas. This would be despite the large capacity constraints (Ezcurra and Rodríguez-Pose, 2013[11]).
In this context, it is also interesting to note the results from Albouy (Albouy, 2012[53]), who argues that fiscal equalisation systems which are currently designed to support poorer areas, may, in fact, slow down much wanted economic growth both in lagging and in more prosperous regions. This could be because transfers could constrain inter-regional migration, which is an important component of economic growth, and because these transfers usually do not take into account the costs that accrue to major urban areas. Blöchliger and Pinero-Campos argue that equalisation transfers reduce fiscal disparities in the short term, but may reduce development incentives in the long run (Blöchliger and Pinero-Campos, 2011[54]).
Decentralisation may risk uncoordinated public investment
Recent research results suggest that there is a positive association between decentralisation and total regional public investment. In particular, decentralisation in terms of revenue autonomy increases public investment in infrastructure (Kappeler et al., 2013[55]). Public investment is nevertheless a shared responsibility among levels of government (OECD, 2013[14]). Yet these are not all and not always equipped with the same level of capacity to ensure successful investment planning and implementation. Due to the long-term effects of investment decisions, errors made in the decision-making process are difficult and costly to correct. Without effective vertical and horizontal co‑ordination, decentralisation may lead to inadequate scale and loss in public investment opportunities (OECD, 2013[14]).
Finding the right balance between own-source revenues and transfers
How can we determine the level of own-source revenues, in particular of own-source taxes? There is no ideal model or rule for thresholds to determine a good level of revenue autonomy. It depends on the overall system, including the type of responsibilities managed by subnational governments and the overall institutional, economic, fiscal and social conditions framework. For example, one major difficulty of developing own‑revenues is a country’s level of development. It is difficult to solicit the taxpayer or the users of local services in poor countries where there is a high level of informality; where potential contributors are partially registered or not registered at all and have a very low level of income; where the land registry or the company register is non-existent or under-developed; where there are few local services or infrastructure that can be priced and generate user fees or charges; where tax administration is weak at the national and local levels; and where subnational governments lack the capacity to raise and manage revenues, etc. These challenges call for careful planning and implementation of decentralisation policies.
In addition, increasing own-source revenues, in particular through greater tax autonomy, may bring many benefits but also difficulties. There are pros and cons in some counter-arguments that can justify limiting tax decentralisation to a certain extent (Ter-Minassian, 2015[56]) (Table 4.1).
Table 4.1. Increasing own-source revenue of subnational governments: Benefits and drawbacks
Sources: OECD (2017[15]), Multi-level Governance Reforms: Overview of OECD Country Experiences, http://dx.doi.org/10.1787/9789264272866-en; Ter-Minassian, T. (2015[56]), “Promoting responsible and sustainable fiscal decentralisation”, in Ahmad, E. and G. Brosio (eds.), Handbook of Multilevel Finance, Edward Elgar Publishing.
Political risks
The political dimension behind decentralisation reform – to favour local democracy, improve the quality of governance, promote citizen participation and control, show accountability and transparency – has been somewhat left on the backburner in favour of a more economic approach focused on policy-relevant outcomes, such as education, health or fiscal sustainability (Faguet, 2011[57]). This is partly explained by the fact that, in some countries, decentralisation has in reality been hijacked by local and national elites, as a strategy for mobilising and maintaining regional power bases (Devas and Delay, 2006[58]). This “elite capture” of local power structures has produced counter-productive outcomes, for example in terms of poverty reduction and corruption (Crook, 2003[59]), but has also damaged the credibility of political arguments.
Local or central elite takeovers and corruption may compromise decentralisation benefits
It has been argued that in some cases the benefits of decentralised service delivery may primarily go to local elites. Particularly in developing countries, local jurisdictions may be vulnerable to capture by predatory pressure groups, who may be able to receive a disproportionate share of spending on public goods (Bardhan, 2002[60]). Such unhealthy development would also increase opportunities for corruption. This underlines the importance of the institutions of local democracy and mechanisms of overall political accountability.
There is also the possibility of predatory central government, not just local elites. Predatory central government can compromise all or most of the benefits of decentralisation (Weingast, 2014[7]). For instance, the ruling parties at the national level may try to affect local decision-making to avoid competition from most successful subnational governments. A central government transfer system may be used to affect local decision-making. At worst, predatory governments may even use their powers to reduce the authority of the local government, for example by taking over the local government and reversing its policies.
Administrative challenges
Unclear assignment of responsibilities
An important challenge of decentralisation is formed by overlapping assignments among levels of government. This challenge is repeatedly mentioned as a critical one in OECD Territorial Reviews and Economic Surveys, as well as in the International Monetary Fund (IMF) and World Bank country studies for developing countries. The challenge is true for unitary countries, but also some federal countries, such as Australia or Germany, for example.
A lack of clarity in the assignment of responsibilities makes service provision and policymaking costlier. It also contributes to a democratic deficit by creating confusion among citizens regarding which agency or level of government is responsible for a specific service, activity or policy. Without a clear assignment of responsibilities, it becomes almost impossible to hold leaders accountable for shortcomings or policy failures. It also hinders efforts at transparency and citizen engagement (Allain-Dupré, 2018[16]).
This problem may arise especially in a multi-level governance system with several government tiers and with a large number of subnational governments. While the challenge can be mitigated with well-planned service assignments and taking into account the varying capacities of subnational governments, it is a very difficult situation to completely avoid in a decentralised environment. Overall, the assignment of functions to the most appropriate government tier plays an important role in minimising both unnecessary spillovers and overlapping responsibilities.
Several countries grapple with unclear assignments. In Brazil, the division of responsibilities is unclear in a number of areas, including healthcare, education, social security, welfare, agriculture and food distribution, environmental protection, sanitation and housing. In Chile, municipalities have few exclusive competencies, and there are 13 national/municipal shared competencies with unclear or poorly established lines of interventions, which are also subject to change. In France, a 2017 report from the Cour des Comptes suggested going further in the clarification of the allocation of competencies between departments and intermediate governments and strengthening efforts to mutualise functions across municipalities (OECD, 2017[15]).
The unclear allocation of responsibilities and functions is particularly notable for the policy areas which are most frequently “shared” among levels of government, in particular, infrastructure (transport), education, spatial planning, health and labour market policy:
Transport: in Mexico, the competency for road construction and maintenance are divided between the three levels of government, with construction mainly executed by federal and state governments, and maintenance mainly carried out by the states or municipalities. In Ukraine, municipalities are responsible for infrastructure and its maintenance. At the same time, the road agency of the national government is responsible for roads, including their paving and repaving. In the case of a bridge (infrastructure), if its road-surface requires repaving, there is a high risk of inaction as neither level of government is compelled to act (OECD, 2018[61]).
Education: in most OECD countries, lower levels of government are responsible for managing and funding lower levels of schooling (mainly pre-elementary, primary and sometimes lower secondary education), whereas responsibility for secondary, and in particular upper secondary, schooling is more often retained at provincial/regional or central levels. Such arrangements, where sub-sectors of schooling operate under different political and administrative jurisdictions, may raise significant challenges concerning efficient use of resources (risks of competition, duplication and overlaps) and co-ordination of policies and actors.
Health: healthcare systems seem to suffer greatly from an unclear division of responsibility, duplication, cost shifting and scale inefficiencies .
Labour market policy: is often shared, with frequent overlaps among levels of government. Greater difficulty arises when several levels of government share the same functional responsibilities, within the same responsibility sector.
Unclear allocation of roles between subnational governments and deconcentrated central state administrations
Another type of challenge is linked to the often unclear allocation of roles between subnational governments and deconcentrated central state administrations in territories. Although the prevalence of deconcentrated central state administrations alongside autonomous subnational governments in the same territory has diminished in the past decades, the challenges remain important in some countries, such as Estonia, Finland, France, Hungary, Greece, Italy, Poland, Sweden, Turkey and Ukraine. Reconsidering these complex and often opaque arrangements could help facilitate the clear assignment of responsibilities.
Coordination of service delivery is a challenge especially in the case of redistributive services
In many countries, the provision of redistributive services, such as healthcare and secondary education, has been assigned to subnational governments. In such cases, central government is usually concerned with equity issues and co-ordination of service delivery. Without adequate central steering of subnational governments and with no mechanism that transfers resources to poor jurisdictions, decentralisation may lead to considerable differences in regional disparities in health, education and social service outcomes (Martinez-Vasquez, 2011[62]). With the intention of securing access to services in all parts of the country and to guarantee a minimum quality of services for all residents, laws and other normative steering mechanisms are frequently used to regulate their provision and financing. The problem is then to strike a balance between co‑ordination and local autonomy. With strict regulation, the benefits from decentralised provision may be radically diminished. Moreover, the bureaucracy costs can be high from such steering. On the other hand, with loose regulation, it is possible that differences between subnational governments become politically unacceptable.
It should also be noted that in a multi-level governance setting, the policy-makers are often confronted by a series of policy “gaps” (Charbit and Michalun, 2009[63]) which can present a considerable challenge to governance.
Lack of subnational government capacity may limit benefits received from decentralisation
The lack of sufficient administrative, technical or strategic capacities is probably one of the bigger challenges in the field of decentralisation. Inadequate capacity may, therefore, be an argument for limiting or delaying decentralisation.
Beyond the fiscal capacity issues, reviews repeatedly report the lack of adequate capacities – in terms of staff, expertise, scale – to address complex issues such as strategic planning, procurement, infrastructure investment, oversight in local public service delivery, performance monitoring, etc. The institutional capacities of subnational governments vary enormously within countries, even the most developed ones (Tselios et al., 2012[12]; OECD/CoR, 2015[64])
Sometimes subnational governments may lack adequate human resources to plan, implement and manage public services. The professional quality (e.g. level of education and job experience in particular) of civil servants working in subnational governments is therefore key. If subnational governments were for some reason not able to hire the quality staff needed to provide services that meet the minimum standards, then decentralisation would pose a risk. In these cases, central government could support local capacity building with training and financial incentives. However, local leadership, community participation and local ownership of the service programmes are equally important factors behind successful local capacity building (Fiszbein, 1997[65]). Without enough local commitment, the resources invested in strengthening subnational capacities may be wasted.
It should be acknowledged that building capacities takes time and therefore needs a long-term commitment from both central and subnational government levels. Even among the developed countries, the capacity differences remain considerable despite efforts to close the gap. For instance, the joint Consultation of Sub-national Governments among the OECD countries by the OECD and EU Council of Regions showed that two-thirds of subnational governments (65%) reported that the capacity to design adequate infrastructure strategies is lacking in their city/region. More than half of the subnational governments (56%) reported a lack of adequate own expertise on infrastructure. These results were particularly strong in the case of small municipalities and inter-municipal structures (OECD/CoR, 2015[64]).
Conclusion
While each country must find its own way to utilise the benefits and tackle the challenges of decentralisation, the policy experiences and research results that have accumulated over the past decades can help policy-makers to implement decentralisation reforms in a way that avoids the major pitfalls. For this purpose, the OECD has developed ten Guidelines for implementing decentralisation, which apply both to federal and unitary countries. These are described in the following section of this report. The guidelines are more than just recommendations. Each section covers the rationale of each guideline, key trends, good practices and not-so-good practices. The themes are elaborated with examples from federal and unitary countries.
Table 4.2. Benefits and challenges of decentralisation
References
[5] Ahmad, E. and G. Brosio (2006), Handbook of Fiscal Federalism.
[53] Albouy, D. (2012), “Evaluating the efficiency and equity of federal fiscal equalization”, Journal of Public Economics, http://dx.doi.org/10.1016/j.jpubeco.2012.05.015.
[16] Allain-Dupré, D. (2018), Assigning Responsibilities across Levels of Government: Trends, Challenges and Guiding Principles for Policy-makers, OECD, Paris.
[25] Azfar, O. et al. (1999), Decentralization, Governance and Public Services: The Impact of Institutional Arrangements.
[60] Bardhan, P. (2002), “Decentralization of governance and development”, Journal of Economic Perspectives, http://dx.doi.org/10.1257/089533002320951037.
[28] Bartolini, D., S. Stossberg and H. Blöchliger (2016), “Fiscal Decentralisation and Regional Disparities”, OECD Economics Department Working Papers, No. 1330, OECD Publishing, Paris, http://dx.doi.org/10.1787/5jlpq7v3j237-en.
[40] Bird, R. (2003), Asymmetric Fiscal Decentralization: Glue or Solvent?.
[2] Blöchliger, H. (2013), “Decentralisation and Economic Growth - Part 1: How Fiscal Federalism Affects Long-Term Development”, OECD Working Papers on Fiscal Federalism, No. 14, OECD Publishing, Paris, http://dx.doi.org/10.1787/5k4559gx1q8r-en.
[29] Blöchliger, H., D. Bartolini and S. Stossberg (2016), Does Fiscal Decentralisation Foster Regional Convergence?.
[27] Blöchliger, H., B. Égert and K. Fredriksen (2013), “Unclassified ECO/WKP(2013)43 Organisation de Coopération et de Développement Économiques Organisation for Economic Co-operation and Development”, http://www.oecd.org/eco/Workingpapers.
[54] Blöchliger, H. and J. Pinero-Campos (2011), “Tax Competition Between Sub-Central Governments”, OECD Working Papers on Fiscal Federalism, No. 13, OECD Publishing, Paris, http://dx.doi.org/10.1787/5k97b1120t6b-en.
[22] Boadway, R. and J. Tremblay (2012), “Reassessment of the Tiebout model”, Journal of Public Economics, http://dx.doi.org/10.1016/j.jpubeco.2011.01.002.
[42] Bordignon, M., L. Colombo and U. Galmarini (2008), “Fiscal federalism and lobbying”, Journal of Public Economics, http://dx.doi.org/10.1016/j.jpubeco.2008.05.003.
[50] Bordignon, M., A. Giglio and G. Turati (2015), Soft Budget Constraints: The Case of Municipal Bonds in Italy, OECD and KIPF.
[49] Borge, L., J. Brueckner and J. Rattsø (2014), “Partial fiscal decentralization and demand responsiveness of the local public sector: Theory and evidence from Norway”, Journal of Urban Economics, http://dx.doi.org/10.1016/j.jue.2014.01.003.
[48] Brueckner, J. (2009), “Partial fiscal decentralization”, Regional Science and Urban Economics, http://dx.doi.org/10.1016/j.regsciurbeco.2008.06.001.
[45] Campbell, T. and H. Fuhr (2004), Leadership and Innovation in Subnational Government Case Studies from Latin America.
[63] Charbit, C. and M. Michalun (2009), “Mind the Gaps: Managing Mutual Dependence in Relations among Levels of Government”, OECD Working Papers on Public Governance, No. 14, OECD Publishing, Paris, http://dx.doi.org/10.1787/221253707200.
[59] Crook, R. (2003), “Decentralisation and poverty reduction in Africa: The politics of central–local relations”, Public Administration and Development, Vol. 23/1, pp. 77–88.
[43] De Mello, L. and M. Barenstein (2001), Fiscal Decentralisation and Governance: A Cross-Country Analysis, IMF.
[58] Devas, N. and S. Delay (2006), “Local democracy and the challenges of decentralising the state: An international perspective”, Local Government Studies, Vol. 32/5, pp. 677-695, http://dx.doi.org/10.1080/03003930600896293.
[11] Ezcurra, R. and A. Rodríguez-Pose (2013), “Political decentralization, economic growth and regional disparities in the OECD”, Regional Studies, http://dx.doi.org/10.1080/00343404.2012.731046.
[3] Faguet, J. (2014), “Decentralization and governance”, World Development, Vol. 53, pp. 2-13, http://dx.doi.org/10.1016/j.worlddev.2013.01.002.
[57] Faguet, J. (2011), Decentralization and Governance, Economic Organisation and Public Policy Programme, http://dx.doi.org/10.2139/ssrn.1892149.
[41] Faguet, J., A. Fox and C. Pöschl (2014), Does Decentralization Strengthen or Weaken the State? Authority and Social Learning in a Supple State, http://eprints.lse.ac.uk/60631/.
[44] Fan, C., C. Lin and D. Treisman (2009), “Political decentralization and corruption: Evidence from around the world”, Journal of Public Economics, http://dx.doi.org/10.1016/j.jpubeco.2008.09.001.
[65] Fiszbein, A. (1997), “The emergence of local capacity: Lessons from Colombia”, World Development, http://dx.doi.org/10.1016/S0305-750X(97)00020-X.
[13] Hatfield, J. (2015), “Federalism, taxation, and economic growth”, Journal of Urban Economics, http://dx.doi.org/10.1016/j.jue.2015.01.003.
[1] Hooghe, L. et al. (2016), Measuring Regional Authority: A Postfunctionalist Theory of Governance, Oxford University Press, Oxford.
[39] Huther, J. and A. Shah (1998), Applying a Simple Measure Decentralization governance.
[24] Inman, R. and D. Rubinfeld (1997), Rethinking Federalism.
[4] Ivanyna, M. and A. Shah (2014), “How close is your government to its people? Worldwide indicators on localization and decentralization”, Economics, http://dx.doi.org/10.5018/economics-ejournal.ja.2014-3.
[55] Kappeler, A. et al. (2013), “Does fiscal decentralization foster regional investment in productive infrastructure?”, European Journal of Political Economy, http://dx.doi.org/10.1016/j.ejpoleco.2013.03.003.
[17] Kim, J. and S. Dougherty (eds.) (2018), Fiscal Decentralisation and Inclusive Growth, OECD Fiscal Federalism Studies, OECD Publishing, Paris/KIPF, Seoul, https://doi.org/10.1787/9789264302488-en.
[21] King, D. (1984), Fiscal Tiers: The Economics of Multi-level Government, Allen and Unwin.
[8] Lockwood, B. (2002), Distributive Politics and the Costs of Centralization.
[62] Martinez-Vasquez, J. (2011), The Impact of Decentralization: Issues in Theory and Challenges in Practice, http://www.adb.org.
[38] Michelsen, C., P. Boenisch and B. Geys (2014), “(De)centralization and voter turnout: Theory and evidence from German municipalities”, Public Choice, http://dx.doi.org/10.1007/s11127-013-0061-2.
[52] Moisio, A., H. Loikkanen and L. Oulasvirta (2010), Public Services at the Local Level - The Finnish Way.
[32] Morgan, K. (2006), “Devolution and development: Territorial justice and the north-south divide”, Publius, Vol. 36/1, pp. 189–206, http://www.jstor.org/stable/20184949.
[46] Morgan, K. (forthcoming), Experimental Governance and Territorial Development.
[19] Musgrave, R. and P. Musgrave (1980), Public Finance in Theory and Practice, McGraw Hill Kogahusha.
[9] Oates, W. (2008), “On the evolution of fiscal federalism: Theory and institutions”, National Tax Journal, Vol. 61/2.
[6] Oates, W. (2005), “Toward a second-generation theory of fiscal federalism”, International Tax and Public Finance, http://dx.doi.org/10.1007/s10797-005-1619-9.
[51] Oates, W. (1985), Searching for Leviathan: An Empirical Study.
[20] Oates, W. (1972), Fiscal Federalism, Harcourt Brace Jovanovich, New York.
[61] OECD (2018), Maintaining the Momentum of Decentralisation in Ukraine, OECD Multi-level Governance Studies, OECD Publishing, Paris, https://doi.org/10.1787/9789264301436-en.
[47] OECD (2018), Subnational Governments in OECD Countries: Key Data (brochure and database), OECD, Paris, http://dx.doi.org/10.1787/region-data-en.
[33] OECD (2017), Making Decentralisation Work in Chile: Towards Stronger Municipalities, OECD Multi-level Governance Studies, OECD Publishing, Paris, https://doi.org/10.1787/9789264279049-en.
[15] OECD (2017), Multi-level Governance Reforms: Overview of OECD Country Experiences, OECD Publishing, Paris, http://dx.doi.org/10.1787/9789264272866-en.
[26] OECD (2016), Decentralisation and Quality of Public Finance: Intermediate Report, OECD, Paris.
[34] OECD (2016), OECD Territorial Reviews: Japan 2016, OECD Territorial Reviews, OECD Publishing, Paris, http://dx.doi.org/10.1787/9789264250543-en.
[14] OECD (2013), Investing Together: Working Effectively across Levels of Government, OECD Multi-level Governance Studies, OECD Publishing, Paris, https://doi.org/10.1787/9789264197022-en.
[35] OECD (2008), OECD Territorial Reviews: Poland 2008, OECD Territorial Reviews, OECD Publishing, Paris, http://dx.doi.org/10.1787/9789264049529-en.
[64] OECD/CoR (2015), Infrastructure Planning and Investment across Levels of Government: Current Challenges and Possible Solutions, OECD, Paris.
[31] OECD-UCLG (2016), Subnational Governments Around the World: Structure and Finance, http://www.oecd.org/regional/regional-policy/sngs-around-the-world.htm.
[36] Regulski, J. and J. Drozda (2015), Building Local Government: Lessons of Experience from the Polish Transition, International Bank for Reconstruction and Development/The World Bank.
[10] Rodden, J., G. Eskeland and J. Litvack (2003), Fiscal Decentralization and the Challenge of Hard Budget Constraints, MIT Press.
[30] Shah, A. (2017), Horizontal Fiscal Equalization in Australia: Peering Inside the Black Box, http://dx.doi.org/10.13140/RG.2.2.29867.54564.
[37] Stoyan, A. and S. Niedzwiecki (2018), “Decentralization and democratic participation: The effect of subnational self-rule on voting in Latin America and the Caribbean”, Electoral Studies, http://dx.doi.org/10.1016/j.electstud.2017.12.001.
[56] Ter-Minassian, T. (2015), “Promoting responsible and sustainable fiscal decentralisation”, in Ahmad, E. and G. Brosio (eds.), Handbook of Multilevel Finance, Edward Elgar Publishing.
[23] Tiebout, C. (1956), A Pure Theory of Local Expenditures, JSTOR, http://www.jstor.orgURL:http://www.jstor.org/stable/1826343Accessed:28/07/200805:58.
[12] Tselios, V. et al. (2012), “Income inequality, decentralisation, and regional development in Western Europe”, Environment and Planning A, http://dx.doi.org/10.1068/a44334.
[18] Wallis, J. and W. Oates (1988), Decentralization in the Public Sector: An Empirical Study of State and Local Government.
[7] Weingast, B. (2014), “Second generation fiscal federalism: Political aspects of decentralization and economic development”, World Development, http://dx.doi.org/10.1016/j.worlddev.2013.01.003.
Notes
← 1. Examples of such services include a wide collection of “neighbourhood services”, such as local infrastructure, parks, kindergartens, elementary schools and water and sewage.
← 2. As measured in the IMF’s Government Finance Statistics.
← 3. The survey was conducted in 1999 and 2000 by a team from the World Bank. Managers from over 9 000 firms in more than 80 countries were surveyed with a standard questionnaire.
← 4. In the same vein, compared with fully centralised taxation, partial revenue decentralisation is a superior alternative, assuming benevolent decision-makers and provided that subnational governments are able to adjust public-good levels. This is mainly because of the allocation benefits from local discretion (Brueckner, 2009[48]).