Decentralization and Federalism

Hi class,

Decentralization is relevant as a process that tries to address hierarchies. Hierarchies are problematic for economists and public and private managers because they create much more problems while managing resources (especially public resources) than markets. This happens, mostly, because of the way they tend to organize themselves.

Broadly speaking, a hierarchy is any structure of power. Churches, especially the Catholic Church, explicitly acknowledge the existence of such structures and they call them hierarchies. In organizational studies, any structure of power is a hierarchy. They are problematic because when they become too large, they tend to impose political, personal, or clientele criteria in the use and distribution of the resources, reducing the overall efficiency of the processes they perform. This is what economists and managers call externalities.

Historically, the move towards decentralization as a concept goes all the way back to the works of Alexis de Tocqueville, the author of Democracy in America.

Tocqueville wrote:

Decentralization has, not only an administrative value, but also a civic dimension, since it increases the opportunities for citizens to take interest in public affairs; it makes them get accustomed to using freedom. And from the accumulation of these local, active, persnickety freedoms, is born the most efficient counterweight against the claims of the central government, even if it were supported by an impersonal, collective will.

More recently, the observation of bureaucracies and the externalities they face prompted a series of changes in the late 1960s in highly centralized countries as France (a unitary, Presidential polity) after the defeat of Gen. Charles de Gaulle in the plebiscite called in 1968. However, it is possible to find examples of this need to decentralize in large U.S. corporations such as General Motors, Ford, General Electric, and the like during the late 1960s and early 1970s. The U.S. government itself acknowledged this problem when, in the early 1970s, forced the partition of "Ol' Mamma Bell" into the Baby Bells. Not only Big Bell had a monopoly over telephone communications, but it was very inefficient in providing that kind of service.

The tendency spread out to other Western countries during the late 1970s, mixed in many cases with the rise of what here and in Britain is called conservatism and in the rest of the world is called neoliberalism, the case of Britain and Margaret Thatcher is paradigmatic of this convergence. That leads some uncritical observers of Latin American reality to mix-and-match neoliberalism with decentralization.

However, it is necessary to acknowledge that way before the emergence of the International Monetary Fund as the factotum of policy in the region, the governments of Latin America were already aware of the need to promote some forms of decentralization or administrative reform.

Decentralization as a theory rejects the assumption that we need some form of hierarchy or structure of power, seeking to raise specific questions about the specific types of hierarchies that are necessary in any given circumstance. It sees hierarchies (for that purpose any form of authority) as arbitrary. After this first question about the need of structures of power, decentralization asks for the specific forms of structures of power that are better for the processes we are addressing or considering.

Decentralization does not reject authority or hierarchies per se, but acknowledges the need to challenge systematically the specific forms of authority or hierarchy to develop at any given moment the best form of authority possible.

From a very abstract perspective, decentralization de-sacralizes power that is why it rejects fixed structures of power and seeks to develop fluent exchanges of information between service providers and clients.

Our Readings

Falleti is mostly concerned with the relations between executive authorities at the three levels of government in the four largest countries in the region. Willis, Garman and Haggard are concerned with the role that political parties have had in these processes.

Their key findings are:

1. Decentralization is a sequential and path dependent PROCESS. Falleti´s emphasis on the procedural character of decentralization seeks to highlight the fact that we cannot understand these processes as black or white kind of processes. Up until today in Mexico, as an example, there are political actors seeking a broader decentralization, while—of course—there are other actors seeking to expand the role of the Federal Government at the expense of the local and state governments.


2. Decentralization happens or not regardless of: the type of regime (democratic or authoritarian); the Constitutional structure (federal or unitary); and who is in charge (civil or military). The four cases used by Falleti provide different examples of this. Mexico was formally a Federal republic, but it was in actuality a very centralized polity; Colombia was a extreme case of political centralization, ridden with extreme illegal organizations (the guerrilla and the drug lords) challenging the ability of the State to control territory and to provide for the needs of the population. Argentina was a very stiff military regime, unable to open itself to the kind of progressive reform that happened in Brazil, where the military tried to combine a strong authoritarian leadership at the national level, with democratically elected subnational leaders.


3. The reason for it, and this is my own take from the Mexican experience, is that it happens because decentralization as well as federalism and other buzzwords of Latin American political jargon, are usually labels behind the “real” policies. In Mexico’s case, a very centralist and authoritarian government, labeled, however as federalist and democratic, started in the late 1970s a decentralization process. It was their way to deal with hierarchies and their externalities (in simple language with bureaucracies and the red tape they crate). Among the problems were: the inefficiency of local political bosses to aggregate political support for the PRI. With the challenge presented by different political actors, some of them legitimate as the Partido Acción Nacional and some others marginal or even illegal as the guerrilla organizations active during the mid 1970s. In addition, with the externalities (problems) created by the hierarchies. Moreover, the Mexican political elites’ response to these challenge was paradoxical, because on the one hand—as Willis et al. nicely explain—it actually decentralized by allocating more resources to the states and municipalities, although—on the other hand—it also centralized power in the figure of the Secretary of Budget and Planning. Interestingly enough, three of the last four presidents of Mexico were before their election heads of that secretary (Miguel de la Madrid Hurtado, Carlos Salinas de Gortari, and Ernesto Zedillo Ponce de León). That gives you an idea of how powerful that position in the cabinet was. Moreover, it was that secretary the one in charge of the decentralization process in Mexico up until its fusion with the Secretary of the Treasury in 1992.


4. The different outcomes of the decentralization PROCESS are less a result of particular or individual policies, and more the product of the EVOLUTION of such reforms and the type of actors empowered along the way. The path dependent nature of the PROCESS is clear when one considers how their decentralization programme developed within the context of the Argentine military regime and how it has evolved over time. Conversely, the Colombian process reflects the willingness of the national government at the beginning of the PROCESS to introduce deep changes in the structure of the polity. One could argue, considering the evidence provided by Willis et al., that the failure of Venezuelan democracy is correlated, at least partially, with the externalities that a unitary, centralist, presidential government creates in terms of the hierarchies running the show. Is not a surprise that the main problem with Partidocracia (Party-cracy) was precisely in terms of corruption (el cogollo).


5. Willis, Garman, and Haggard’s piece is valuable because it introduces the issue of the role of the national political parties as a key aspect of the decentralization process. It is not a surprise that polities where the parties were under severe constraints (Argentina during the military dictatorship banned the Peronista party, as an example) had poor outcomes. The same can be said of Federalist polities where the President does not enjoy a legislative majority (hence Mr. Fox problems in Mexico nowadays with the fiscal reform). Divided governments (like the situation that Bill Clinton faced here in the States during much of his term) produce (as contemporary Mexico confirms) deadlock.


6. Although decentralization can have positive impacts on poverty and other key topics, we cannot be too optimistic about the outcomes of decentralization processes. The best example of it comes from Colombia. But, of course, as Leah wisely pointed out, the main problem there is that the old centralist, unitary, and presidential system confronted the additional challenges brought by the guerrilla and the drug lords. Here you can get a brief newsletter analyzing some of the negative and positive processes of decentralization worldwide.

We could analyze these processes also by using Foucault’s approach on governmentality. The national governments display their technologies of power, which are confronted by the subnational governments’ technologies of self and their own technologies of power, and out of the interplay of these two forms of technologies, of specialized knowledge expected and unexpected consequences emerge.

I believe, as I already mentioned in class, that when analyzing processes like decentralization or federalization it is necessary to pay attention to the specific policies and path dependent and non-path dependent processes. Moreover, whether as policymakers or as social entrepreneurs or as social scientists or even as simple citizens here in the US, in Latin America or in Europe, we need to pay attention to this interplay of technologies of self and technologies of power and be ready for unexpected consequences.

One way to avoid some of the unexpected consequences brought by large scale planning (the six-year National Development Plan that Mexican presidents must submit to the Congress during the first year of their term could be a good example) is by considering as much voices as possible in the processes: think, as an example of the kinds of audiences that the MTA or many other agencies hold here in NYC everytime they need to raise fares or when they decide to expand a line or to renew or close a station.

If you want to get some more information on the topic, go to the Website of the Organisation for Economic Cooperation and Development.

I also mentioned in class the existence of these new forms of organizations called heterarchies, something in between a market and a hierarchy. A heterarchy can be defined as "horizontal self-organization among mutually interdependent actors." If you are interested in that topic read this brief paper (12 pages).