Making Decentralization Work: The Politics of Implementation
One of the longstanding questions among political scientists and constitutional lawyers working on territorial politics concerns the appropriate distribution of power between the center and the periphery. A substantial body of scholarship has sought to explain the origins of federalism and why some states opt for decentralization while others remain more centralized. Some authors emphasize structural factors, such as preexisting ethno-territorial conflicts, regional economic inequalities, colonial legacies, or ethnolinguistic diversity. Others, by contrast, focus on federalism as the outcome of strategic bargaining between central and peripheral elites. According to this perspective, when peripheral elites are strong and the central government is weak, the former can extract greater powers; when the balance of power is reversed, decentralization is more limited. This line of research, known as endogenous federalism, has attracted renewed scholarly attention in recent years.
However, despite the broad literature on why jurisdictions are assigned to one level of government or another, one important question remains largely unanswered and has received far less attention: to what extent and how are decentralization agreements implemented? We can point to several prominent examples of major decentralization agreements, such as the devolution of powers to Scotland and Wales in the late 1990s, the successive rounds of state reform in Belgium, and the various stages of decentralization in Spain, including the 2006 Catalan Statute of Autonomy. Each of these legal reforms required not only formal approval by governments, legislatures, and even citizens’ referenda, but also practical implementation. Once competencies such as healthcare, university education, or forest management are reassigned from the national or federal government to a regional authority, the responsible administration must transfer the corresponding financial, human, and infrastructural resources from one level of government to the other. For instance, when the 2006 Catalan Statute established that the Catalan government would assume responsibility for wastewater management, the Spanish central government transferred control of the Prat de Llobregat wastewater plant to the Catalan authorities.
So, why should we pay attention to how each decentralization process is implemented? In a recent research article in the Journal of European Public Policy, Jorge Fernandes and I argue that for decentralization to be considered complete, it requires not only the formal legal transfer of jurisdiction but also its effective implementation. If the new authorities do not receive the necessary infrastructure, human resources, and financial means to carry out the tasks assigned to them, decentralization remains incomplete. Worse, such partial implementation can generate political backlash, as the original agreement is perceived as unfulfilled. An interesting example is the case of the Basque Nationalist Party (PNV). One of their longstanding claims is that the 1979 Basque Statute of Autonomy has never been fully implemented. They argue that, despite the Basque government’s recognized authority over certain policy areas, many jurisdictions remain under the management of Spanish central authorities. This grievance has repeatedly surfaced and is present in the current parliamentary confidence-and-supply agreement between the PNV and the Spanish Prime Minister, Pedro Sánchez. In such instances, incomplete or unfulfilled decentralization weakens the credibility of central authorities. When peripheral actors perceive that decentralization agreements are not fully implemented, their trust in the central government diminishes, which can lead to renewed calls for greater autonomy – or even independence.
We use the Spanish case to examine how different decentralization reforms are implemented. Spain provides an excellent setting for this study for two main reasons. First, its legal framework for decentralization is highly flexible, and we can observe at least three major waves of reform: an initial wave in the early 1980s, a second in the mid-to-late 1990s, and a third in the late 2000s. Each of these waves involved the decentralization of different policy areas. Second, and critically, Spain offers an unusually high level of transparency. Every time the central government and a region reach an agreement to implement part of a decentralization reform—for instance, transferring a hospital facility or reassigning public employees—the process is formally recorded and published in a public document known as a Royal Decree of Transfer (Real Decreto de Traspasos). We use these legal documents to measure the degree of implementation of the decentralization agreement.
In our article, we follow the literature on endogenous federalism and consider that the implementation of each of these agreements will depend on the strength of the central and peripheral actors at any given moment. In our paper, we test a series of hypotheses about the implementation of decentralization agreements in Spain. We begin by looking at the characteristics of regional governments, and we expect that when the same party governs at both the national and regional levels, the process of implementing decentralization agreements will be smoother. Shared ideological preferences between the two levels of government should reduce potential conflicts. Meanwhile, regional governments may also use the implementation process strategically, positioning themselves as champions of their region to gain electoral advantages. We also look at the characteristics of the national government and propose that decentralization is easier to implement when the central government inherently favors decentralization and when the same cabinet has negotiated both the decentralization agreements and their implementation. Continuity of preferences should minimize uncertainty and prevent renegotiation, thereby enhancing the credibility of the implementation process.
Finally, the bargaining environment is another crucial factor. When the national government is in a minority, the bargaining environment opens up, and other parties, particularly ethnoterritorial ones, can extract concessions, including the implementation of decentralization agreements. For instance, Catalan and Basque nationalist parties often condition their legislative support for the central government on the effective transfer of competencies to their regions. By contrast, when the national government holds an absolute majority, it faces fewer incentives to decentralize and is more likely to resist transfers of power. Other factors, such as bureaucratic bottlenecks, shifts in public opinion, or judicial interventions, may also affect the implementation of decentralization agreements. However, we leave these for future researchers to study.
Using a series of statistical analyses, our findings provide strong support for our expectations. We observe that co-partisan regional governments are more likely to receive transfers, reflecting a smoother implementation process. National governments that are ideologically supportive of decentralization also tend to deliver more transfers. Stability in the cabinet, when the same government is responsible for both negotiating and implementing the agreement, also facilitates the process. Additionally, minority governments at the national level are more likely to transfer power. By contrast, we find little evidence that ethnoterritorial parties systematically extract more transfers when they hold stronger bargaining power in the national arena. Importantly, our results remain robust when disaggregated by specific policy areas, which gives us further confidence in the validity of our findings.
Our paper aims to open a new line of research by moving beyond the traditional focus on measuring the legal distribution of power across regions and the conditions under which regions gain more or less authority. We take the next step by examining implementation: to what extent, and under what circumstances, are decentralization agreements actually carried out? We argue that insufficient implementation can lead to political backlash and worsen pre-existing territorial conflicts. At the same time, we acknowledge that fully unpacking these consequences is beyond the scope of this study. Future research should further explore this area, especially by examining cases where the institutional rules governing the implementation of decentralization agreements differ from those in Spain, and by investigating the mechanisms behind the relationships identified here more closely.
Javier Martínez-Cantó is a political scientist working on comparative political institutions at the Institute of Public Goods and Policies (IPP), part of the Spanish National Research Council (CSIC)