Decentralization of Education and Institutional Change


A Look at Mexico

by | Apr 7, 1999

“Citizens are becoming accustomed to holding policy makers accountable,” writes Gustavo Merino.

In the national independence day parade in Oaxaca last September, as is tradition, hundreds of the city’s schoolchildren marched alongside rescue workers, police and soldiers. Leading each school contingent were two children carrying the school banner bearing its name and in most cases the words Escuela Póblica Federal and its number. The legend struck me as odd. Eight years before the federal government had transferred to the states the responsibility for the operation of all but a handful of schools at the “basic” level (preschool through grade nine plus teachers training. Schools, since that time, are no longer “federal.” Had the schools been slow to change their name or merely failed to update their banners? While a minor issue, it is related larger questions I am working on as part of my doctoral dissertation and which had brought me on a research tour to Oaxaca and other states: How had state governments responded to the decentralization of education? Did decentralization result in significant innovations, changes in education policy, finance or schooling techniques?

The decentralization of education has the stated goal of improving the quality and access to educational services. It’s expected to promote better resource allocation because state authorities have more information on local conditions than federal bureaucrats and can foster innovation. Ideally, it will also elicit greater financial contributions from the states. The reform could improve administrative efficiency within the Ministry of Public Education (SEP), which had become extremely large and inflexible, and curb the influence of powerful groups within, especially the National Union of Education Workers (SNTE).

Decentralization efforts in education were not new to Mexico. Previous attempts in the seventies and eighties had been significantly scaled down in the face of opposition, primarily from SNTE, so that the educational system remained highly centralized by the early nineties. 65% of all schools and 75% of those in basic education (preschool through ninth grade) were federally funded and controlled. Federal involvement in education was unequally distributed, however. In some states like Aguascalientes, Oaxaca and Hidalgo, all but a handful of schools at the basic level were federally operated and funded. At the other extreme were states with a strong educational tradition and large education budgets that enrolled a significant number of students in state schools. The state of Mexico, Baja California and Nuevo León fall in this category. The rest of the states fall in between these extremes. Everywhere, however, the federal government designed education policies, set the calendar, assigned textbooks and other activities.

By 1992, President Salinas was able to overcome much of the opposition and the decentralization, or “federalization” agreement was signed by both levels of government and the SNTE. By design and political compromise, its scope was limited although wider reaching than previous attempts. States would now control and operate all schools in basic education, but the federal government kept the main regulatory and policy responsibilities and remained as the principal source of funding. Teachers became state employees but remained affiliated with SNTE under a largely unchanged labor contract and centralized wage and benefit negotiations.

Preliminary analysis suggests that the response of state governments to decentralization has been mixed and generally weak. State governments have not, in general, carried out significant reforms in the operation of educational services now under their responsibility, nor in the allocation and magnitude of monetary and human resources devoted to education. This result is common across states in spite of the large differences between them regarding educational levels and economic, social and demographic characteristics.

This is not to say that states have been idle or uninterested in promoting policy changes. 23 states out of 31 have drafted or reformed their education legislation and most have created educational development plans. Several states have promoted reforms seeking to improve administrative procedures and reduce costs or instituted training programs for teachers and supervisors. Some have funded scholarships and special programs such as the use of computers in schools. While necessary and praiseworthy, these efforts nevertheless do not appear very far reaching as they do not significantly change the way public education is provided.

Three highly important policy areas were reform has been lacking refer to education finance, integration of state and federal systems and control over human resources. First, while public education expenditures have risen since decentralization, much of the increase is driven by federal aid rather than state expenditures. Additional funds are used primarily to cover the costs imposed by decentralization, mainly the equalization of teacher wages and benefits between ex-federal and state teachers, and not in other educational inputs currently under-funded. Further, the structure of spending, at least with regard to educational levels, has not changed significantly and most changes can be explained by longer term trends.

Second, states have been very slow to integrate their own educational systems, where in existence, with the ex-federal system in their jurisdiction. Hence, they cannot take advantage of economies of scale or administrative efficiencies. By 1998, two systems still remained in 9 out of 20 states, each with its separate head, schools, teachers and students. Even where there was formal integration of the educational systems under one Ministry of Education or its equivalent, the two sub-systems were often administratively separate or treated as such in practice. Third, closely related to the lack of effective integration is the restructuring of the system’s human resources to achieve policy goals. Most states have not attempted any changes in the distribution of teachers among schools, levels or districts, even though there might be a surplus in some levels or areas and a shortfall in others.

Why the weak response of state governments to decentralization? I think it reflects inappropriate economic and political incentives for significant policy reform at the state level, combined with the federal government’s failure to enact complementary institutional reforms to reinforce the mechanisms by which decentralization supposedly leads to better service delivery. Moreover, some of the major policy tools directly or indirectly affecting the provision of education remain centralized.

As was mentioned earlier, the 1992 decentralization did not grant state governments much autonomy with regards to educational policy since the federal government kept most regulatory and policy-design functions. This affects educational planning on both tiers: the federal government cannot exercise the same authority over state departments of education as it could previously over its own delegates, and state authorities face uncertainty regarding policy directives. It also reduces accountability as the blame for suboptimal performance can easily be shifted between levels of government given the shared responsibility for education.

Furthermore, the fiscal system remains highly centralized. State governments are highly dependent on the federal government not just for education finance but for most of their income as their taxing powers are very constrained. With limited ability to raise additional funds, increasing educational spending means less expenditures for other public services. Because much of federal assistance is discretionary, states might also fear lower aid receipts if they increase their own expenditure in any public service. Such fiscal arrangement can further weaken accountability structures.

The persistent centralization of the SNTE and the political influence at its disposal through its control of over a million members and its traditional alliance with the PRI, makes it a formidable opponent few governors willingly challenge. In addition, decentralization went as far as the state level, with no provisions for greater autonomy at the school level. Last, low technical capacity at the state level reflecting decades of centralization and bureaucratic inertia, slows-down change. The new educational authorities in many states were formerly employed by SEP and therefore their policies might not differ much from those to which they are accustomed.

The limitations mentioned above should not be taken to imply that decentralization brought no benefits. It is perhaps too early to make definitive judgments. Some reforms have been mentioned already and there is evidence from a few states that decentralization led to the social and political reevaluation of the importance of education for regional development. The federal authorities, now free from some administrative burdens, have also developed new programs to raise educational quality, access and equity. Time and the consolidation of democracy might help solve some of the problems that limit effective policy response. Demands for fiscal decentralization are louder. The national SNTE leadership will loose influence in line with the PRI’s electoral fortunes and as the regional section leaders gain prominence. Citizens are becoming accustomed to holding policy-makers accountable.

What the evidence so far suggests, however, is that if additional reforms do not take place, the process might be too slow. Education is too important to wait for these developments to happen on their own. Institutional change and the generation of appropriate incentives can be achieved faster if there is political will.

Spring 1999

Gustavo Merino Juarez is a Ph.D. candidate in Public Policy at Harvard University. This article refers to preliminary results of work in progress for his doctoral dissertation on decentralization and federalism in Mexico. Some of the material was gathered through field research financed in part through a Summer Research Travel Grant from DRCLAS. Comments welcome at

Related Articles

Latino Families and the Educational Dream

Latino Families and the Educational Dream

Insisting that we do our interview in English, Ricardo Robles reluctantly recalls the dreams he had for his children when first arriving nine years ago to the U.S. from Zacatecas, Mexico.” They…

LASPAU Expands Fulbright Partnerships

LASPAU Expands Fulbright Partnerships

LASPAU Expands Fulbright Partnerships An Array of Partners Since 1975, LASPAU has administered Fulbright Program grants for Latin American and Caribbean faculty members, professionals, and researchers. Fulbright programs have provided academic opportunities for...

Harvard Immigration Project

Harvard Immigration Project

A refrigerator is the most important thing in life, the 10-year-old immigrant child reported in a matter-of-fact sort of way…

Print Friendly, PDF & Email