The literature on decentralization, fiscal federalism and local government finance has its foundations in works by political scientists and economists including Paul Samuelson (1954); Charles Tiebout (1956); Ursula Hicks (1961); James Buchanan (1965), Mancur Olson (1965), Wallace Oates (1972), and Dennis Rondinelli (1981). Decentralization is an evolving concept, however, and its role in public sector management and development is increasingly well-understood, including due to more recent efforts by scholars and policy practitioners including Roy Bahl and Richard Bird (2018); Barry Weingast (2009); Paul Smoke (Smoke et al 2011; Smoke 2018); Leonardo Romeo (1999; 2014), Jean-Paul Fauget (2014) and Dorothée Allain-Dupré (2018).

Traditional definition of decentralization. Although there is no single consensus definition of decentralization, most “traditional” definitions are derived from the definition posited by Dennis Rondinelli (1981; 1986; 1999), who defined decentralization as

“the transfer of authority and responsibility for public functions from the central government to subordinate or quasi-independent government organizations or the private sector.”

or in a more detailed manner as the

“transfer of [authority and] responsibility for planning, management, resource-raising and allocation and other functions from the central government and its agencies to (a) field units of central government ministries or agencies; (b) subordinate units or levels of government; (c) semi-autonomous public authorities or corporations; (d) areawide, regional or functional authorities; or (e) nongovernmental private or voluntary organizations.”

Although there are slight variations among Rondinelli’s definitions of decentralization over the years, the core concept of decentralization involves the transfer of authority, responsibility, and resources away from the central government and towards more localized actors.

Forms and dimensions of decentralization. Professor Rondinelli’s definition implies that there are several different forms of decentralization, depending on the nature of the intended recipient of the authority or responsibility that is being decentralized, including:

  • Deconcentration. The transfer of authority, responsibility, and resources from the central government to field administration of central government ministries or agencies; in other words, assigning authority and responsibility to local offices within the central bureaucracy.
  • Delegation. The transfer of authority, responsibility, and resources from the central government to semiautonomous or quasi-public corporations, and the assignment of delegated functions to nongovernment organizations that are ultimately still accountable to the center.
  • Devolution. The transfer of authority, responsibility, and resources from the central government to (elected) local governments.

In addition, decentralization can be segmented is by the nature or type of power that is being decentralized, resulting in three dimensions of decentralization:

  • Political decentralization. The transfer of political authority and oversight responsibility from the central government to citizens and/or their elected representatives at the local level is often associated with pluralistic politics and representative government. This is because it gives citizens, or their representatives, more influence in the formulation and implementation of policies – decisions made with greater participation will be better informed and more relevant to diverse interests in society than those made by national political authorities alone.
  • Administrative decentralization. The transfer of administrative authority and responsibility from the central government to subordinate or quasi-independent subnational governments. This involves the transfer of administrative responsibility for the planning, financing, and management of certain public functions from the central government and its agencies to subnational governments.
  • Fiscal decentralization. The transfer of fiscal authority and responsibility as well as financial resources from the central government to subordinate or quasi-independent subnational governments.

The need for balance between the different dimensions of decentralization. While the three main dimensions of decentralization can be considered in isolation in order to allow for more in-depth analysis, there are clear and strong inter-linkages between these different dimensions. The literature on multilevel governance literature suggests that public sector outcomes or results can be achieved by more centralized as well as by more decentralized public sector arrangements, but that—regardless of the extent of (de)centralization—public sector effectiveness in a multi-level governance context requires that each of these three dimensions are (a) well-structured and internally coherent, and (b) balanced with the other two dimensions (Boex and Simatupang 2015; OECD 2019). For instance, in order for the benefits of devolved public service delivery to materialize, local governments would need to be assigned a reasonable balance of political, administrative as well as fiscal powers. By contrast, assigning local governments with extensive fiscal powers and resources would be unlikely to achieve better service delivery outcomes in the absence of sufficient political accountability or decentralized administrative authority.

Inconsistent application of definitions and understanding of the concept. Unfortunately, the traditional definitions of decentralization are not universally understood or or consistently applied, even within the academic literature. The inconsistent use of terminology tends to cause confusion and takes away from the consistency of Rondinelli’s definitions.[5]

Perhaps the most common source of confusion is the interchangeable use by some of the terms like “decentralization” and “devolution.” Traditionally, discussions of decentralization have tended to focus on devolution of political, administrative, and fiscal authority and establishing political accountability linkages between citizens and local politicians. In that sense, the term decentralization is often but incorrectly equated to devolution.

It is not unusual, for instance, for a Cabinet to approve a Decentralization Policy or law which spells out a series of decentralization reforms empowering local governments over a number of sectoral functions (i.e., devolution reforms), only for sectoral ministries to implement a series of “sectoral decentralization” reforms by which the ministries empower their own lower-level administrations or frontline service delivery facilities through deconcentration or delegation.[6] As these different types of decentralization reforms are very different in nature, it is critical to clearly specific the type of decentralization to be pursued. Consistent use of the appropriate terms helps to ensure that everyone is on the same page when discussing a reform, particularly one of such a political nature.

Box 1.1 What is a local government?

Devolution is defined as the transfer of authority, responsibility, and resources from the central government to local governments. Thus, in order to have a clear understanding of the meaning of devolution, it is important to clearly define local governments as opposed to, for instance, deconcentrated local administration entities.

Although there is no single consensus definition of local governments, they are often understood to be defined by these four characteristics: (a) separate legal entity or body corporate; (b) authoritative decision-making power over one or more public functions in a local jurisdiction – that is, with its own political leadership; (c) with control over its own officers and staff; and (d) responsible for preparing and executing its own budget (WDR 2004; PEFA 2013).

Local governments may be formed at one or more levels of territorial administration; may have different legal status in urban versus rural areas; and may go by different names in different countries – including local governments, local authorities, local councils, district governments, municipalities, and communes. Some countries recognize two types of local governments, general-purpose local governments or special-purpose local governments. Whereas general-purpose local governments have a broad range of functions or responsibilities, the special-purpose ones typically only have functional responsibility over a single function, such as water boards in the Netherlands or school districts in the United States.

There is no absolute dividing line between what a local government is and what it is not. For instance, based on the characteristics defined above, not all entities traditionally accepted as local government adhere closely to all four characteristics.[7] In other cases, depending on their adherence to the characteristics noted above, some types of local entities, such as local school committees, may be considered quasi-local government authorities.  

From decentralization to multilevel governance. During the second half of the twentieth century, decentralization (and devolution) in particular, was a dominant public sector reform around the world, particularly in post-colonial developing countries in Africa and Asia. Similarly, decentralization was an important part of public sector reforms in former centrally planned economies and other authoritarian regimes that sought to transition away from central planning and central government dominance to more decentralized, market-based economies. Use of the term, “democratic decentralization,” became especially common in the decade after the collapse of the Berlin Wall (Manor 1999; Crook and Manor 2000). Indeed, the twentieth century saw a string of decentralization reforms around the world focused on democratization, including major decentralization reforms in the Philippines (1991), the Russian Federation (1993), South Africa (1993), and Indonesia (1999).

Despite its continued importance as a major public sector reform, the momentum of decentralization reforms around the world has waned somewhat during the first two decades of the new millennium. Within the global development community, attention shifted from the (largely politically driven) decentralization reforms of the 1990s towards a greater sector focus based on the Millennium Development Goals (MDGs). While major decentralization reforms continued to be introduced in countries such as Kenya (2010) and Nepal (2015), others such as the Russian Federation, and other countries in Central and Eastern Europe experienced a degree of re-centralization.

To some extent, the increased focused on development results – first under the MDGs, and now, under the Sustainable Development Goals (SDGs) – has been accompanied by a shift in tone when speaking about decentralization. Although (democratic or political) decentralization was seen by some as a goal in its own right, decentralization is increasingly understood as a means to an end, with the goal of greater public sector efficiency, a more inclusive and responsive public sector, greater political empowerment, or better service delivery results.

The World Development Report 2004, “Making Services Work for Poor People,” was an important inflection point in the global policy debate surrounding decentralization. Prior to this point, decentralization, or specifically, devolution, was generally pursued as a governance-motivated reform, and the capacity of local government institutions was regarded as the main binding constraint to successful devolution. WDR 2004 articulated the more nuanced notion that decentralization is a multilevel governance reform that has the potential to improve public sector performance by shortening the “long route of accountability” between people, central government policy makers, and providers, but explicitly recognized that decentralization is not a one-size-fits all solution. Instead, in order to achieve the effective delivery of (pro-poor) public services in a multilevel public sector, simultaneous interventions would be required at three levels: empowering intergovernmental systems; effective, inclusive and responsive local institutions; and an engaged civil society, citizenry, and private sector. The WDR (2004: 75) further explicitly acknowledged that under different stages of economic and democratic development, resulting in different degrees of client empowerment, different approaches to decentralization and localization would be appropriate.

The conceptual evolution in thinking about decentralization was accompanied by an evolution in the terminology used to discuss decentralization and intergovernmental relations. Whereas the terminology and definitions of decentralization are seen by some to imply a value-judgement that more decentralization is better, some global development actors – including some UN agencies and United Cities and Local Governments (UCLG) – increasingly speak of “localization,” especially when it comes to the localized achievement of the SDGs. Similarly, Boex (2012) adopted the more neutral terminology of the “local public sector,” while others consider the “territorial approach to local development” and “community-led development” (Romeo 2014; THP).

Researchers and practitioners also increasingly use the term “multilevel governance” to refer to different aspects of intergovernmental (fiscal) relations, recognizing that public services are often co-produced by stakeholders at different government levels, often simultaneously relying on different approaches to decentralization or localization, even within the same sector. In turn, the effectiveness of subnational governments is largely defined by the nature of multilevel governance arrangements and by the institutional background of the stakeholders at different government levels (Charbit 2011; Enderlein, Wälti and Zürn 2011).

In this context, the key underlying question is not necessarily whether countries should decentralize or not, or even what model of decentralization should be followed, but rather that public sector effectiveness requires practitioners to focus on identifying ways to improve capacity and coordination among public stakeholders at different levels of government to increase the efficiency, equity, and sustainability of public spending in the context of a multilevel public sector (e.g., Charbit 2011; OECD 2019).

Emerging alternate definitions. In line with this evolving view of decentralization, Roy Bahl (2005) offered, by way of alternative working definition for decentralization, that the concept entails “the empowerment of people by the empowerment of their local governments.” This formulation was slightly generalized by Boex and Yilmaz (2010) to suggest that “decentralization is the empowerment of people through the empowerment of the local public sector.”

These emerging alternate definitions should be seen as complements rather than as substitutes to Rondinelli’s original definitions of decentralization. Whereas Rondinelli offered considerable detail on the “what” and “how,” the more recent definitions focus more on the “why” (empowerment), which is a driving force not only behind today’s inclusive global Sustainable Development Agenda, but also fundamental to desire of government leaders and global development actors to promote public sector performance.

[5] Two common inconsistencies in terminology are, first, to equate devolution and political decentralization, and second, to equate deconcentration and administrative decentralization (Boex 2012). While political decentralization is an important dimension of devolution (without which in would be impossible to have an effective devolved system), these two terms do not share the same meaning. Likewise, while administrative decentralization is an important dimension of deconcentration (without which in would be impossible to have an effective deconcentrated system), these two terms do not mean the same thing.

[6] This was the case during the implementation of Nepal’s Local Self-Government Act (1999). Despite the fact that the legislation clearly empowered local government bodies over primary education, the Ministry of Education transferred responsibility for the operation of primary schools to facility-level school committees – while retaining authoritative decision-making and resource-distribution powers at the ministerial level – based on the argument that the school committees were “closer to the people” than the elected local governments.   

[7] For instance, municipalities in Afghanistan lack an elected mayor or council, but are nonetheless viewed as being local governments. In the Netherlands, while municipal councils are elected, mayors are appointed by the central government. In other countries, such as Botswana and Uganda, local governments lack an independent budget or control over their own officers and staff. In fact, it is quite common in many countries for local governments to be (largely) staffed by officers seconded by the central government.