Global Urban Development Magazine


GUD Magazine Home





Land, Housing, and Transportation: The Global Challenge

Strategizing Slum Improvement in India

Basic Costs of Slum Upgrading in Brazil

Market-Driven Eviction Processes: Kigali and Phnom Penh

Decentralization of Argentina's National Housing Fund

Squatters No More: Singapore Social Housing

Land for Housing in African Cities

Limits to Large-Scale Reconstruction in Honduras

Property Rights in Namibia's 'Extra Legal' Settlements

Impacts of Transportation on Urban Poverty in Colombo, Sir Lanka


Regularization of Informal Settlements in Medellin, Colombia



About the Authors

Editorial Guidelines

Previous Issues

May 2005

March 2006

GUD Home


Published by
Global Urban Development

Executive Editor:
Dr. Marc A. Weiss

Managing Editor:
Nancy Sedmak-Weiss


Volume 3                    Issue 1                    November  2007

Print Version     

Seeking Better Policies or Just Giving Up Responsibility: THE Decentralization of Argentina’s National Housing Fund

Cecilia Zanetta

1.  Introduction

Decentralization has played an important role in the context of the Washington Consensus, figuring prominently in the reform programs implemented in countries around the world during the 1980s and 1990s. While the benefits of decentralization have largely been posed in economic terms, such as its potential for increasing efficiency in the use of public resources, politics has been the force driving decentralization in most countries. As a result, decentralization has often been implemented hastily, paying little attention to the design of the underlying policy framework or the conditions that needed to be in place to ensure its success. Not surprisingly, the potential benefits from decentralization have often not materialized and sub-national governments have consistently failed to improve upon—or even replicate—the levels of performance of central governments.

The decentralization of Argentina’s National Housing Fund (FONAVI) provides a good illustration of the shallow success that decentralization can have when responding primarily to political considerations rather than to the desire to improve service provision and the structure of governance in general. The decentralization of Argentina’s largest housing fund constituted the most significant action in the area of housing and urban policies in the context of the ambitious reform program that was implemented in Argentina during the 1990s. Given the historically poor performance of the program, together with the high diversity exhibited by Argentina’s sub-national governments in terms of their geography, demographic and socio-economic conditions as well as institutional frameworks, the decentralization of FONAVI had the potential to substantially improve public housing programs.  However, improving public housing programs in order to mitigate the negative impact that the overall reform program was having on the more vulnerable sectors of the population was not a consideration in the decentralization of FONAVI. Instead, national authorities used the transfer of the substantial financial resources being channeled under FONAVI to provincial governors as a bargaining chip to obtain their support for the overall reform program. Thus, it is not surprising that national authorities exhibited little interest in designing and implementing a decentralization process aimed at improving the fund’s performance and the transparent and efficient use of FONAVI’s resources. Consequently, the decentralization of FONAVI has failed to yield any significant benefits and, instead, many of the theoretical risks identified in the literature have materialized. 

The paper is organized as follows: first, it examines the definition of decentralization and the theoretical benefits and risks identified in the literature. Second, based on the lessons learned from decentralization experiences worldwide, the paper derives a set of policy prescriptions that could have maximized the chances of success of the decentralization of FONAVI. Third, it contrasts the prescribed policy guidelines with the specific policy framework that underlined the decentralization of FONAVI and its implementation strategy.  Fourth, the paper examines the actual results obtained a decade after the decentralization of FONAVI, showing that it had, at best, only a shallow success. Finally, the paper summarizes the important lessons that can be drawn from this case study, which unfortunately teach us more in terms of its failings rather than its strengths.

2.  A Brief Overview of Decentralization 

In broad terms, decentralization can be defined as the transfer of planning, decision-making, or administrative authority from the central government to other entities, including its local administrative units, semi-autonomous or quasi-public organizations, sub-national governments, non-government organizations, and the private sector (Rondinelli, 1983).  Four major forms of decentralization can be distinguished based on the degree of authority that is being transferred by the central government to the decentralized entities and on the level of autonomy of these decentralized entities: deconcentration, delegation, devolution, and privatization. In this paper, we focus on devolution, which refers to the transfer of authority for decision-making, finance, and management from the national government to autonomous sub-national units of government.  As lower levels of governments can exercise their newly assumed authority in an autonomous manner, devolution has the potential to yield both the benefits and shortcomings commonly ascribed to decentralization, making it both more promising and more risky than other forms of decentralization (Litvack et al. et al., 1998).[1]

The main potential benefit from decentralization is that it can enhance the efficiency and responsiveness of governments (Oates, 1972; Tiebout, 1956; Musgrave, 1983). Devolving resource allocation to sub-national officials can potentially improve efficiency, as they are presumably better positioned to assess the needs and preferences of their constituencies to, in turn, decide on the best resource allocation of public services. It can also improve the management of public services, as sub-national officials can be held more accountable for the performance than more remote national bureaucracies and elected officials (Ostrom et al., 1993). If successfully implemented, decentralization can also reduce red tape and bureaucracy, improve credibility and legitimacy of the government in general, foster innovation while minimizing the risks in case of failure (Rondinelli, 1983) and achieve greater representation of diversity through greater public input in pluralistic political environments (Pauly, 1973).

Among the risks from decentralization is the potential decline in the delivery of services as the result of the low technical capacity of sub-national governments (Burki et al., 1999). Similarly, decentralization can also exacerbate the ability of local elites to affect the allocation of public goods (Burki et al., 1999; Wilensky, 1974; Inman, 1997). In particular, if power at the sub-national level is more concentrated and more easily subject to the influence of local elites than at the center, then greater decentralization will not necessarily result in greater democracy or more ‘power to the people’ (Griffin, 1981).  Also, decentralization policies can also result in widening disparities in social spending, such as health and education, with the consequent negative equity implications.  Existing evidence suggests that decentralization often results in an increase in the variance of public service performance, with improvements in some jurisdictions and worsening in others (Burki et al., 1999). Likewise, decentralization can result in the loss of economies of scale and control over scarce financial resources by the central government (World Bank Decentralization Website).[2]

In general, there is little empirical evidence to support any of the theoretical benefits or pitfalls of decentralization (Litvack et al., 1998). As with any public policy, the simple creation of decentralized structures does not guarantee success, greater efficiency, or any other of the theoretical benefits of decentralization (Rondinelli, 1983).  For decentralization to succeed, institutionally and technically adept decentralized structures need to be complemented with adequate systems of incentives purposely designed to induce pre-determined behaviors on the part of sub-national actors as well as safeguards against the potential obstacles that can impede a successful implementation process. Moreover, the analysis of what constitutes sound decentralization policies is country specific, as decentralization often involves drastic reforms cutting across sectors and levels of government. As pointed out by Griffin (1981) ‘it all depends on the circumstances under which decentralization occurs.’  The following sections explore Argentina’s experience with the decentralization of FONAVI, the soundness of the underlying policy framework and the implementation strategy, as well as the main outcomes a decade after it decentralization.

3.  Policy Prescriptions for Successful Decentralization Processes

Valuable lessons can be learned from decentralization experiences across the world.  As shown on Table 1, these lessons can be summarized in four main categories: i) the balance between authority and accountability; ii) the role played by the national government; iii) the validity of the assumptions underlying the decentralization; and iv) the process of implementation.

One of the most important lessons drawn from experiences with decentralization worldwide is the need to strike a balance between responsibility and accountability. In other words, giving sub-national actors both the means (authority) and the incentives (accountability) to fulfill their newly assigned responsibilities is perhaps the most important factor in ensuring the success of decentralization reforms. On the one hand, authority involves granting sub-national officials not only the legal authority to make decisions but also the financial and human resources needed to deliver results. On the other side, accountability aims at making sub-national officials responsible for their performance by tying performance to specific rewards and/or penalties.  Thus, a good balance between authority and responsibility is a key ingredient to ensure the satisfactory performance of politicians and bureaucracies at lower levels of government (Burki et al., 1999). 

Officials at all levels of government have to play their part if decentralization is to succeed. In particular, national authorities—the president, congressmen and political leaders—have a key role in: regulation, redistribution, enforcement, and evaluation and learning. Specifically, national officials are responsible for defining the rules for decentralization—i.e., the policy framework—that, in turn, determines the behavior of politicians at lower tiers of government (Burki et al., 1999). These rules ought to carefully reflect the specific objectives of individual programs as well as national objectives. To ensure that politicians and local officials have an incentive to be responsive, the instruments of decentralization—the legal and institutional framework, the structure of service delivery responsibilities and the level of financing from upper tiers of government—have to be consistent with the political objectives (IBRD Governance website). As indicated by Burki (1999), national officials also have a key role in deciding which of these rules are going to be enforced. While these decisions are not often made explicit, they have an important bearing on the results of decentralization efforts, as they constitute the true incentives—as opposed to the nominal policy framework—determining the behavior of sub-national governments. 

Redistribution is also a valid function of the national government, both geographically—i.e., across sub-national governments—or by income—i.e., within sub-national governments.  Redistribution criteria will be primarily reflected in the way resources are allocated among sub-national jurisdictions and in the rules determining allocation of benefits within each jurisdiction. In addition, given its privileged capacity to overlook at the totality of sub-national governments, the national government has a competitive advantage in promoting learning and horizontal fertilization among sub-national governments. Although often overlooked, this task has great potential in effectively fostering institutional learning by disseminating best practices and innovations while comparing the relative advantages and disadvantages of alternative strategies being tried at the sub-national level. 

Although decentralization has frequently been portrayed as an inherently superior policy strategy, its success depends on the validity of the often-implicit underlying assumptions.  One of the most widely accepted justifications for decentralization is based on the assumption that sub-national officials have a better understanding of the preferences of their constituencies and the determination to honor them. This is often not the case, as politicians’ knowledge of local preferences might not be detailed enough in relation to the delivery of specific services.  Even if sub-national authorities had this knowledge, their actual decisions might respond to the interests of local power elites that do not represent the wider preferences of the population. To ameliorate these risks, it is important to have mechanisms by which citizens can express their preferences in a way that is binding on the politicians—in this way, citizens have a credible incentive to participate.  Participatory budgeting mechanisms, such as the one in Porto Alegre, Brazil, are a good illustration of such mechanisms. In addition, it is important that citizens are well informed about the costs of services and options involved, the resource that are available and their sources, so that the decisions they make are meaningful (IBRD governance site).

Another often-erroneous assumption is that the proximity of politicians at lower levels of government to their constituencies automatically translates into greater political accountability.  In other words, if politicians do not do a good job, their constituencies can vote them out of office.

Table 1.  Golden Rules for Successful Decentralization Policies 

Strike a balance between authority and accountability
  • Grant sub-national governments an adequate degree of authority so that they can fulfill their responsibilities.

ü   Legal authority

ü   Financial resources

ü Human resources

  • Make sub-national governments clearly responsible for their performance—i.e., framework of ‘institutional’ accountability.

ü   Performance tied to mandates/penalties

ü   Performance tied to incentives/rewards

The national government needs to fulfill its various responsibilities

  • Regulation:

ü   The national government is responsible for determining the rules of the game— ‘policy framework.’

ü   These rules—together with the corresponding incentives and penalties—should reflect the specific program objectives, as they determine the behavior of lower-tiers of government.

  • Enforcement

ü    National authorities often enforce rules selectively, creating a parallel—although perhaps more important—de facto policy framework.

  • Redistribution:  National governments can pursue redistribution through the allocation of resources

ü   Geographical redistribution: It can be achieved through the allocation of resources among sub-national governments

ü   Redistribution by socio-economic characteristics of the population:  It can be achieved through criteria determining selection of final beneficiaries.

  • Evaluation and learning

ü   Promote institutional learning and horizontal fertilization

ü   Disseminate innovations and best practices

Make sure that the implicit assumptions about the quality of governance of sub-national governments are in place

  • Introduce mechanisms to ensue that sub-national authorities have good knowledge of local conditions.

ü   Promote mechanisms of public consultation as well as partnerships with representatives of civil society

ü   If these mechanisms are binding, there are added incentives for the population to participate and less chances of having local power elites determine policy outcomes

  • Enhance the ‘political’ accountability of sub-national officials with respect to their constituencies.

ü   Development of sound measures of performance that can be clearly understood by the citizens and allows comparisons across jurisdictions

ü   Wide dissemination of performance measures among local constituencies

ü Support the activities of grassroots watchdog organizations

  • Ensuring adequate technical and institutional capacity at the sub-national level

ü   Technical assistance (TA) tends to work best when the emphasis is on promoting the exchange of information among peers rather than relying on top-down models.

ü   Likewise, demand-driven TA tends to be more effective than supply-driven TA.

Devise a sound and realistic implementation process

ü   Incremental or partial decentralization processes can be more effective in minimizing risks.

ü   ‘Certification’ of sub-national governments as an ex-ante condition for decentralization can ensure that key pre-conditions are in place.

This is not always the case, often as a result of the lack of adequate information. Thus, having in place an easily accessible and transparent information system to enable the community to effectively monitor the performance of the local government so they can react appropriately can be an appropriate strategy to enhance accountability (IBRD Governance Website).  Likewise, the wide dissemination of this information among local constituencies and the presence of vital grassroots organizations can also enhance the political accountability of sub-national officials. 

Ensuring that sub-national governments have the institutional and technical capacity needed to take on the newly assigned responsibilities is part of national authorities’ responsibility to ensure that key ‘enabling conditions’ are in place. The success of decentralization often depends heavily on training for both national and local officials on the implications of decentralization and their respective responsibilities. In addition, technical assistance is often required for local governments and local non-governmental groups in the planning, financing, and management of decentralized functions (IBRD governance website). Lessons learned worldwide indicate that technical assistance tends to work best when the emphasis is on promoting the exchange of information among peers rather than relying on top-down models.  Likewise, demand-driven technical assistance, in which sub-national governments determine their own needs in terms of technical and institutional strengthening, tends to be more effective than supply-driven strategies in which the national government defines a one-fits-all strategy. In addition, acquiring minimum levels of technical and institutional capacity can be a requirement for assuming new responsibilities. For example, in the case of Colombia, provincial and local governments had to be ‘certified’ before assuming responsibility for health and education (Burki et al., 1999).

Finally, some countries have been more cautious in their approach toward decentralization, adopting incremental implementation strategies to diminish the risks associated with decentralization—lack of technical capacity at the local level, undue power of local elites, widening disparities among sub-national jurisdictions. For example, Mexico implemented a form of micro-monitored earmarking as it embarked on sector decentralization (Burki at al., 1999).  These and other forms of incremental or partial decentralization can help national governments gauge the success and shortcomings of ongoing decentralization processes and make necessary adjustments while minimizing risks.

In summary, for decentralization to succeed, it is crucial to formulate a national housing strategy in which the national government plays a central role in establishing a sound normative framework that responds to the actual housing needs of the population instead of to the concerns of special interest groups benefiting from ongoing practices. Likewise, the national government needs to set in place—and to enforce—incentives and mandates that foster efficiency and transparent practices among provincial governments. The national government can also play an important role in fostering cross-fertilization among provincial governments, disseminating best practices, and providing technical and institutional assistance. As in the case of the decentralization of health and education, the reassignment of responsibilities of provincial governments was not accompanied by the corresponding redefinition of responsibilities at the national level. This failure is hardly surprising, given that the need to curb public spending at the national level, rather than the vision to drastically reform the delivery of social services, was the driving force behind decentralization policies in Argentina.

4.  The Decentralization of Argentina’s Housing Fund (FONAVI)

Argentina’s FONAVI was created in 1972 to attend the housing needs of lower-income segments of the population. During the two decades following its creation, FONAVI became the primary mechanism for financing low-income housing, commanding considerable financial resources—e.g., 97.3 percent of public resources allocated to housing and infrastructure in 1999 (MECON, 2000). Nevertheless, FONAVI programs satisfied the housing needs of just a small fraction of potential beneficiaries because of its narrow focus on the production of costly, finished units and the chronic mismanagement of its resources.  One of the main factors contributing to FONAVI’s poor performance was the centralized administration of its resources and an inadequate system of incentives that failed to induce provincial governments to administer FONAVI funds efficiently. As a result, the FONAVI system was plagued with structural administrative inefficiencies, including excessive unitary costs, heavy subsidies and insignificant levels of cost recovery (Buckley, 1988, 1991).

The decentralization of FONAVI took place in 1992, as part of the reform program that was implemented in Argentina during the Menen administration. As in most countries in the region, the reform program was aimed at tackling fiscal imbalances and reducing the role of the state. Ideologically, Menem’s reform program sought to dramatically reverse the economic model that had been in place for over five decades—one of heavy state interventionism, inward-looking trade orientation and disregard for macroeconomic equilibrium—replacing it by an economic strategy based on competition and economic openness. Although not an end per se, decentralization was an important element of the reform program.  The role of the central government was drastically reduced both as a regulator of economic activities and as a provider of services. Equally important, decentralization efforts were motivated by the need to reduce spending at the national level to sustain the fragile macroeconomic stability that had been achieved as part of the so-called Convertibility Plan. In this context, sub-national governments and the private sector became more active in the provision of services, as the national government completed the privatization of most public enterprises and transferred the responsibility for some public services—i.e., public hospitals and secondary education—and programs—such as FONAVI—to the provinces (Zanetta, 2004, 2004b).

The decentralization of FONAVI was part of an agreement between the national and provincial governments that modified the terms of the automatic revenue system—the so-called first Fiscal Pact of 1992.  Under the Fiscal Pact, provincial governments agreed to forgo 15 percent of their shared revenues to finance national social security reform in exchange for a minimum level of transfers—set at US$725 million per month or US$8.7 billion per year.  In exchange, the national government also guaranteed a minimum level of FONAVI transfers of US$75 million per month—or US$900 million a year—in case the receipts from the gasoline tax were lower than the minimum established (Viola, 2000; Ministry of Economy, 2000).  To make the agreement more palatable for provincial governments, the national government also agreed to transfer to the provinces the financial resources corresponding to four national funds, including FONAVI (Vetter and Zanetta, 2000; Cuenya, 1997). In this way, provincial governments assumed control of all FONAVI funds—amounting to approximately US$1 billion per year—and full responsibility for defining and administering their own housing programs, including their technical and financial characteristics (Cuenya, 1997; Martínez de Jiménez, 1997).

Given the historically poor performance of the program as well as the wide variations exhibited by Argentina’s provinces in terms of most demographic, geographic, and socio-economic variables, the decentralization of FONAVI provided an opportunity to enhance the efficiency and responsiveness of public housing programs. As pointed out in the decentralization literature, the type and mix of housing programs could have been improved, as provincial officials are generally better positioned to assess local needs and preferences. Likewise, the decentralized administration of FONAVI funds could have resulted in better management—such as increased cost recovery—as provincial officials could potentially be held more accountable for their performance. If successfully implemented, FONAVI’s decentralization could also have resulted in less red tape and bureaucracy, increased innovation, better representation through greater public input, and enhanced credibility of provincial governments in general. If adequately instrumented, the decentralization of FONAVI could have potentially benefited those segments of the population that were being negatively affected by the structural adjustment program (Zanetta, 2004b).

5.  Policy Framework and Implementation Strategy for the Decentralization of FONAVI

As with the decentralization of other public services in Argentina, the decentralization of FONAVI was not the result of a well thought-out reform strategy aimed at improving efficiency in resource allocation, enhancing transparency or fostering public participation. Instead, the national government used FONAVI as a bargaining chip in the negotiation of the Fiscal Pacts to gain the support of provincial governors for the reform program and, thus, ensure the sustainability of the newly achieved macroeconomic stability. Through the Fiscal Pact of 1992, the national government succeeded in reducing its automatic transfers to the provinces, a major step toward controlling fiscal deficits at the national level. In exchange, the transfer of FONAVI funds gave governors de facto control over this important mass of resources, with very little oversight on the part of the national government (Zanetta, 2004).  With other actors unwilling or unable to exert sufficient influence, the decentralization of FONAVI had a narrow political focus, at odds with larger social and economic objectives (Zanetta, 2004b).[3]

Although not necessarily incompatible with technical considerations, the political motivations driving the decentralization of FONAVI did not engender a process of decentralization aimed at enhancing the efficiency, effectiveness and transparency of FONAVI.  Although the technical staff at the Secretariat of Housing made a substantial effort to introduce sound mandates in the new legal framework, national authorities exhibited little commitment to their enforcement. Moreover, many of the lessons learned from decentralization processes elsewhere were ignored, consequently repeating many mistakes that could have been easily avoided.  Equally importantly, Argentina’s authorities overlooked many opportunities that arose from the decentralization of FONAVI that could have resulted in substantially better housing programs without compromising the support of provincial governors.

Authority:  In 1995, a new legal framework (Law 24.464) was enacted to adapt the original FONAVI legal framework to the terms agreed under the Fiscal Pact of 1992.  The new law established a Federal Housing System (Sistema Federal de Vivienda) that put FONAVI under the jurisdiction of provincial governments. Provincial housing entities (Institutos Provinciales de Vivienda, IPVs) were given full responsibility for the administration of the resources, and a newly created National Housing Council (Consejo Nacional de la Vivienda, CNV) was charged with the coordination of normative, planning and evaluation functions (Cuenya, 1997).  In practice, the provincial housing entities received full authority to manage their share of the funds, without much interference from either the national government or the National Housing Council. Arguably, provincial governments received too much authority, considering that they were hardly accountable for their performance.

Accountability:  The 1995 Law included some incentives and mandates aimed at ensuring that the potential benefits of decentralization would occur. For example, the Law provided that the allocation of FONAVI funds among provinces, which is determined by formula explicitly determined by law, could be modified every other year according to changes in the housing deficit and provincial performance, including adequate use of funds, levels of cost recovery and the levels of provincial investment.  Likewise, as part of the 1992 agreement, provincial governments were required to maintain a Registry of Applicants (Registro Permanente de Postulantes), including data on the applicants’ household structure, income and date of application, as a way of enhancing the transparency in the allocation process (Zanetta, 2004). In addition, and despite strong resistance on the part of provincial governments, the executive branch—the Subsecretaría de Vivienda, SVN—later audited provincial FONAVI accounts to ensure that the funds were not channeled to other uses, partly due to pressures from the construction industry and the World Bank (Cuenya, 1997; Holubeck, 2002)

Redistribution and other program objectives: The 1995 legal framework somewhat expanded the narrow political objectives that drove the decentralization of FONAVI, as it offered provincial governments greater flexibility to adapt their low-income housing strategies to respond to their specific needs and established new guidelines for the use of resources, encouraging greater diversification. For example, instead of restricting financing to housing units built under FONAVI, the new law set spending targets requiring provinces to provide mortgage loans to individuals purchasing housing in the market, as well as upgrading and expanding their current units. Specifically, it required that loans to final beneficiaries account for at least 15 percent of all FONAVI spending, with a goal of 45 percent by 1998 (see Table 2).  Likewise, it encouraged the construction of community facilities and the provision of urban infrastructure, up to 20 percent of all FONAVI funds (Cuenya, 1997).  The 1995 FONAVI Law, however, fell short of providing any incentives or enforcement mechanisms aimed at ensuring larger social coverage, a wider spectrum of housing solutions and the wider participation of economic and social actors, including medium and small construction firms, cooperatives, and non-government and community organizations (Cuenya, 1997).

Enforcement:  The most serious shortcomings associated with the decentralization of FONAVI are related to the lack of enforcement and weak implementation. In particular, the national government showed an astonishing lack of political will to enforce the mandates of the 1995 Law. As illustrated by the consolidated FONAVI spending in 1999 (see Table 2), the almost exclusive focus on the construction of new housing has persisted, absorbing most of the resources (89 percent of total funds when excluding operation costs). Likewise, the possibility of offering individual loans to final beneficiaries, one of the main innovations of the 1995 FONAVI Law, has remained untapped. Only 4.5 percent of the resources were allocated to this category in 1999 compared with the minimum 15 percent stipulated for 1995 and the 45 percent anticipated for 1998. Similarly, there has been very little investment in community facilities or basic infrastructure. Only 2.4 percent of all FONAVI funds were being used for infrastructure in 1999, which is negligible compared to the upper limit of 20 percent defined by the FONAVI Law (SVN, 1997). The mismatch between the Law’s spending targets and the actual allocation of FONAVI resources has been particularly large in some provinces.  Likewise, as of 2000, only nine out of 23 provinces have in place up-to-date registries and, in most provinces, the selection process is not based on clearly defined criteria or a transparent process (MECON, 2000).  Despite the wide differences in performance among provinces, no changes have been made in the original allocation, with FONAVI funds continuing to function as fixed transfers (Zanetta, 2004).

Table 2.  Spending Targets Established by 1995 FONAVI Law and

Consolidated Provincial FONAVI Spending by Category, 1999 


1999 FONAVI Spending


Expenditure Category

$ Million


Law 24.464/95

Housing construction




Loans to final beneficiaries



Minimum 15 percent;

45 percent by 1998

Community facilities







Up to 20 percent

Administrative costs








 Source: Zanetta (2004)

In summary, although the policy framework underlying the decentralization of FONAVI provided some incentives and enforcement mechanisms to diversify investments and enhance transparency in the selection of final beneficiaries, the national government was unwilling to enforce them. Motivated by narrowly defined political interests, Argentina’s authorities failed to capitalize on the window of opportunity that opened with the decentralization of FONAVI, which could have resulted in substantially better housing programs targeting the poor, ameliorating at least in part the high social costs of the overall reform program.

6.  Results after a Decade of Decentralization 

Not surprisingly, the decentralization has failed to render any substantial benefits, except for reducing the bureaucracy at the national level—at the expense of forgoing any policy making technical and institutional capacity at the national level—and fostering new partnerships with alternative actors, such as labor unions, municipal governments and non-government organizations.  Other improvements include an increase in the level of provincial funding and a reduction in construction start-up times, as most of the administrative processes were now within provincial jurisdiction.  Overall, annual production has increased from approximately 30,000 units between 1983 and 1992 to 48,000 units between 1993 and 1999, including 6,600 alternative housing solutions (MECON, 2000) (see Table 3).

There have been, however, important shortcomings as a result of the decentralization of FONAVI. For example, there has been a considerable expansion in provincial bureaucracies, which, in turn, has translated into substantial increases in administrative costs. In 1999, administrative and operational costs accounted for 22.4 percent of all consolidated FONAVI spending, compared to the typical 7 to 10 percent of similar programs (MECON, 2000). Likewise, decentralization has also widened technical and institutional differences between provinces. More important, the decentralization of FONAVI has failed to overcome the traditional shortcomings, including high unitary costs and the consequent lack of affordability for low-income households, lack of transparency in the selection of final beneficiaries, narrow emphasis on finished housing, chronic low levels of cost recovery and systematic administrative delays (MECON, 2000) (see Table 3).

The poor performance of FONAVI in the decade following its decentralization should not come as a surprise.  After all, improving sector policies, increasing the efficiency of the system, or promoting wider participation among various societal actors were not the driving forces behind its decentralization. Likewise, the fact that old practices have continued at the provincial level should not be surprising either, as decentralization did not respond to the demands of wider sectors of society calling for greater participation (Aguilar and Sbrocco, 1997). Instead, the decentralization of FONAVI was indeed successful when examined through political lenses, as it succeeded in rallying the support of provincial governments for the Convertibility Plan, which in 1992 depended heavily on the signing of the Fiscal Pact (Zanetta, 2004b). 

7.  Conclusions

Although the decentralization of FONAVI resources was an appropriate step in a country as large and diverse as Argentina, the national government did little to ensure that the potential benefits of decentralization did indeed materialize. Specifically, the national government failed to create an environment of accountability by tying the allocation of FONAVI funds to provincial performance, thus fostering inefficiency among provincial housing programs. Likewise, it did not enforce the targets established in the 1995 FONAVI law, such as minimum levels of spending on urban services and individual loans, effectively relinquishing its policy making responsibility.  Finally, it failed to enforce minimum standards, including ensuring transparency in the selection of final beneficiaries.

It is not surprising that the theoretical benefits of decentralization largely failed to materialize in the case of FONAVI, as the national government effectively relinquished its responsibility to implement it adequately. As a result, the traditional shortcomings of FONAVI have persisted, including high unitary costs, lack of transparency, narrow emphasis on finished housing and chronic low levels of cost recovery. Furthermore, FONAVI’s experience illustrates some of the potential risks associated with decentralization, such as widening disparities among provinces and the enhanced ability of local elites to affect the allocation of public goods.

Overall, Argentina’s decentralization of FONAVI indicates that, as with any public policy, the simple creation of decentralized structures does not guarantee greater efficiency or any of the theoretical benefits of decentralization.  For decentralization to succeed, it needs to be supported by clear policy guidelines, effective technical and institutional assistance, sound monitoring and control mechanisms and, most importantly, a willingness to enforce them on the part of the national government.  All these elements have been missing in the decentralization of FONAVI as it was implemented during the 1990s.

As a manner of epilogue, it is interesting to point out that the collapse of Argentina’s economy in December 2001 brought about a drastic policy shift, not only at the macroeconomic level but also within individual sectors, including housing. In contrast with the Menen administration that advocated for reducing the role of government in the economy, the current administration believes in the government playing a strong role guiding economic growth and actively participating in the provision of infrastructure and social services. Within this model, the construction of pro-poor housing is one of President Kirchner’s flagship initiatives to reduce unemployment and poverty. Since taking office in May 2003, the Kirchner administration has launched a number of ambitious housing programs to support the construction of over 440,000 new housing units and the expansion and upgrading of over 140,000 additional units with approximately US$7.4 billion in financing from the national government. These programs, which operate outside the sphere of FONAVI, reflect a trend toward ‘re-centralization,’ with the national government playing a stronger role in their implementation than under the ‘decentralized’ FONAVI model. While provincial IPVs and, in the case of some programs, also municipalities, play an important role in implementation, being responsible for providing vacant land, project design, bidding and contracting, and supervision of works, the Under-Secretariat for Urban Development and Housing (Sub-Secretaría de Desarrollo Urbano y Vivienda – SDUV) at the national level is responsible for approving individual project proposal submitted by the provinces. Likewise, funds are transfers to the provinces only after appropriate certification of work (certificados de obra) is submitted to the SDUV.

While the recentralization of housing programs may have been partly motivated by the poor results achieved under the decentralized FONAVI model, there are clear indications that the current implementation model cannot keep abreast the ambitious scale of these housing programs. As of June 2006, official data show that only 42,400 new housing units have been completed, equivalent to roughly 10 percent of the overall target. The proportion is even smaller for expansion and upgrading programs, which show that only 6,600 housing units have been completed, equivalent to 5 percent of the original target (SDUV, 2006). While provincial IPVs and municipal governments are undoubtedly far from being without blame, delays on the part of the national government to transfer funds to the provinces are not only causing a slow-down in implementation but also the financial asphyxia of the private construction firms that have been awarded construction contracts under these programs, or that of those provincial governments that have accepted to advance the funds from their own treasuries.[4] In this way, Argentina is once again experiencing the limitations associated with centralization. This, in turn, suggests that re-centralization might not be the best way to respond to the unsuccessful decentralization of FONAVI that took place during the 1990s. Instead,  a better approach might be to take the decentralization process further, incorporating the lessons learned during the 1990s.

Table 3.  Outcomes resulting from the Decentralization of FONAVI 

Positive Outcomes

Bureaucracy has decreased at the national level

  • There has been an important reduction in the number of employees working on housing and urban issues at the national level, with only 15 employees remaining in 2000. However, the price tag for such reduction has been high, as the policymaking capacity was virtually destroyed.

There is more diversity in the institutional arrangements

  • Alternative approaches for the production of new units have been developed, including: i) co-financing with other institutions, such as labor unions, ii) decentralization to municipalities, which are responsible for providing the land and selecting final beneficiaries; iii) financing of individual credits; and iv) titling through the BHN, S.A.

  • In turn, this institutional diversity has translated into an increase in annual production, from approximately 30,000 units between 1983 and1992 to 48,000 units between 1993 and 1999, including 6,600 alternative housing solutions.  

There has been an increase in provincial contributions

  • There has been a significant increase in provincial funds allocated to complement FONAVI transfers in each province, equivalent to a 57 percent increase between 1996 and 1999.

  • Provincial contributions constituted 22.5 percent of all FONAVI funds in 1999, compared to only 13 percent in 1996.

Construction start-up times have decreased

  • Provincial governments were given responsibility for evaluating firms’ technical and financial profile as part of the bidding process, previously in the hands of the national government. This has resulted in shorter administrative times.

Negative Outcomes

There has been an increase in administrative costs

  • Administrative costs—such as personnel, fees, and legal expenditures—have climbed steadily, from almost 18 percent in 1996 to 22 percent in 1999.

  • These costs are excessive high, as compared to the average 7 to 10 percent benchmark typical for this type of programs.

  • Large differences in the incidence of administrative costs among provinces—ranging from 0.07 percent to 47 percent of overall expenditures—point to severe inefficiencies in some provinces.

Technical and institutional inequalities among provincial governments are exacerbated

  • Large variations in the number of provincial employees—even when taking into account differences in number of beneficiaries and number of finished housing units—indicate strong differences in technical and institutional capacity across provinces. For example, the number of finished units per employee ranged from 1 to 56 in 1999.

Persistent Shortcomings

Cost recovery remains low

  • When measured as the ratio between actual payments received versus the payments due on an annual basis, provincial performance in terms of cost recovery shows almost no improvement between 1996 and 1999—48.14 percent and 48.7 percent respectively.

  • There are extreme variations in the level of cost recovery among provinces, ranging from just 15 percent to almost 79 percent.

  • Low levels of cost recovery seriously jeopardize the sustainability of the system.

Unitary costs continue to be excessively high

  • Unitary housing costs remain excessively high, at an average of $26,000 per unit.

  • High housing costs continue to pose a barrier to improved affordability and targeting of the program to low-income groups.

There is a persistent lack of transparency in the selection of final beneficiaries

  • As part of the 1992 agreement, provincial governments were given the responsibility for defining the criteria to select final beneficiaries.   

  • In addition, provincial governments are required to maintain a Registry of Applicants (Registro Permanente de Postulantes), which includes data on the applicants’ household structure, income and date of the application as a way of enhancing the transparency in the allocation process.

  • As of 2000, only nine provinces have up-to-date registries. In most provinces, the selection process is not based on clearly defined criteria or a transparent process. Thus, the historical lack of transparency of FONAVI has not been overcome with decentralization.

Ineffective targeting of low-income sectors of the population

  • An analysis of the income eligibility criteria for the various FONAVI programs indicates that those programs implemented by local governments and other intermediaries are more likely to serve low-income groups.

  • However, these decentralized programs account for only one-third of all FONAVI housing units being built in a year, with the remaining two-thirds of the units goes to households whose income is higher than FONAVI’s intended target population.

There is a persistent emphasis on the construction of new units

  • FONAVI’s historical bias toward new housing units still persists.  Of all units completed or financed in 1999, 83 percent corresponded to new units compared to only 17 percent for alternative housing solutions, such as expansion and upgrading of existing units, sites-and-services, and basic sanitary modules.

It has been difficult to break away from the old FONAVI models (Martínez de Jiménez, 1997)

  • Most newly built housing complexes are located outside the existing urban structure.

  • Almost all provincial governments continue to use the same housing prototypes, with no innovative use of materials or non-traditional building techniques.

  • There has been very little community participation or involvement of NGOs in the definition of housing strategies at the provincial level.

Subsidies are still highly regressive (IDB, 2001)

  • While the interest rates vary greatly among provinces, from 0 percent to 12 percent, they are significantly below comparable market rates—i.e., 16.9 percent in February 2001.

  • Highly subsidized interest rates translate into heavy indirect subsidies, which represent approximately 54 percent of all FONAVI resources.

  • There are additional subsidies, such as hidden land and construction costs, as well as low repayment rates from final beneficiaries.

 Source:  Zanetta (2004)


Cecilia Zanetta is an Adjunct Professor in the Department of Geography at the University of Tennessee, Knoxville. She is the author of The Influence of the World Bank on National Housing and Urban Policies.  During the past 15 years, she has worked as a consultant for the World Bank and the Inter-American Development Bank in Mexico, Argentina, and other Latin American countries.


[1]       In contrast, deconcentration involves the transfer of administrative responsibilities—but not of authority—to lower levels of governments.  Delegation, as its name indicates it, involves the delegation of decision-making and management authority of public functions from the central government to local governments or semi-autonomous organizations that are not totally under the control of the central government but are ultimately accountable to it (Rondinelli, 1983; Litvack et al., 1998).  Thus, as a result of the lack of authority and full autonomy of lower levels of government, these two forms of decentralization are less likely to lead to the potential benefits and shortcomings of decentralization (Litvack et al., 1998).

[2]       Fiscal decentralization—as opposed to the decentralization of public services—also poses potential risks to macroeconomic stability. Decentralization might make it harder for central governments to use fiscal policy to adjust to economic shocks, as fiscal decentralization reduces central control over the aggregate public sector revenues and expenditures (Tanzi, 1996). Also, it might result in an over-expanded public sector when there is a mismatch between revenues and expenditures assigned to each level of government.  Finally, decentralization might pose incentives for excessive sub-national borrowing if there is an expectation of bailouts by the national government (Burki et al., 1999). The risks associated with fiscal decentralization were given substantial attention by the decentralization literature of the 1990s, as a result of the high priority assigned to macroeconomic performance in the context of the Washington Consensus.

[3]       As noted by Zanetta (2004b), the decentralization of FONAVI funds to the provinces received the support of the construction industry, as it left intact—or even enhanced—its ability to influence the adoption of favorable housing strategies. Alternatively, other social actors, such as labor unions, popular urban movements and the poor in general, lacked enough political leverage to introduce any special concessions or considerations as part of the housing policies implemented in the context of the reform program.

[4]       See, for example, ‘Más suspensiones de obreros que construyen viviendas socials,’ La Capital, 1 July 2006, Rosario, Santa Fe; ‘Admiten demoras en el Plan Federal,’ La Opinión, 17 July 2006, San Rafaela, Santa Fe; ‘Bell Ville aún espera 24 casas del Plan Federal,’ La Voz del Interior, 3 October 2006, Córdoba.



Aguilar, M.A. and Sbrocco, M.E. 1997. ‘Salta frente al POSFONAVI ¿Una historia circular?,’ in B. Cuenya and A. Falú (eds.), Reestructuración del Estado y Política de Vivienda en Argentina. Buenos Aires: Centro de Estudios Avanzados, Universidad de Buenos Aires.

Buckley, R. 1991. ‘The Measurement and Targeting of Housing Finance Subsidies: The Case of Argentina,’ Public Finance, Vol. 31, pp. 355-72.

Buckley, R. 1988. The Measurement Control and Targeting of Housing Subsidies: The Case of Argentina. Washington, D.C.: Policy Planning and Research Staff, Discussion Paper, World Bank.

Burki, S. J, Perry, G., and Dillinger, W. 1999. Beyond the Center: Decentralizing the State. Washington, D.C.: World Bank.

Consejo Nacional de la Vivienda (CNV). 2002. Declaración de la XXVIII Asamblea Ordinaria del Consejo Nacional de la Vivienda, 5-6 December 2002. Puerto Madryn, Argentina.

Cuenya, B. 1997. ‘Descentralización y política de vivienda en Argentina,’ in B. Cuenya and A. Falú (eds.), Reestructuración del Estado y Política de Vivienda en Argentina. Buenos Aires: Centro de Estudios Avanzados, Universidad de Buenos Aires.

Holubeck, A. 2002. A Shallow Success Story: The Decentralization of Argentina’s National Housing Fund, Manuscript, Center for Latin American Social Policy, University of Texas at Austin.

Inman, R. P. 1997. ‘Rethinking Federalism,’ Journal of Economic Perspectives 11(4): 43-64.

Litvack, J., Ahmad, J., and Bird, R. 1998. Rethinking Decentralization in Developing Countries. Washington, D.C.: Sector Studies Series, Poverty Reduction and Economic Management, World Bank.

Martínez de Jiménez, L. 1997, ‘Nuevas Líneas de Acción Desarrolladas por los Organismos Provinciales,’ in B. Cuenya and A. Falú (eds.), Reestructuración del Estado y Política de Vivienda en Argentina. Buenos Aires: Centro de Estudios Avanzados, Universidad de Buenos Aires.

Ministerio de Economía (MECON). 2000. Evaluación del Fondo Nacional de Vivienda (FONAVI), 29 December 2002. Buenos Aires: Dirección de Gastos Sociales Consolidados Ministerio de Economía.

Musgrave, R. A. 1983. ‘Who Should Tax, Where, and What?,’ Tax Assignments in federal Countries.  Canberra: Centre for Research on Federal Financial Relations, Australian National University.

Oates, W. 1972.  Fiscal Federalism. New York: Harcourt Brace Jovanovich.

Ostrom, E., Schroeder, L. and Wynne, S. 1993. Institutional Incentives and Sustainable Development. Boulder: Westview Press.

Pauly, M. 1973. ‘Income Distribution as a Local Public Good,’ Journal of Public Economics (2): 35-58.

Rondinelli, D. A. 1983. ‘Decentralization,’ in Cheema, S, and Rondinelli, D. A. (eds.), Decentralization and Development: Policy Implementation in Developing Countries. Beverly Hills: Sage.

Shah, Anwar. 1998. ‘Fostering fiscally responsive and accountable governance: lessons from decentralization,’ in Picciotto, Robert and Eduardo Wiesner (eds.), Evaluation and Development: The Institutional Dimension. New Brunswick, USA and London, U.K.: Transaction Publishers.

Sub-Secretaría de Desarrollo Urbano y Vivienda (SDUV). 2006. Programas.  de Vivienda.

Sub-Secretaría de Desarrollo Urbano y Vivienda (SDUV). 2006. Programas Federales de Vivienda, según Jurisdiccion al 30/06/2006, Documentos y Estadísticas de la Dirección Nacional de Políticas Habitacionales.

Subsecretaría de Vivienda (SVN) 1997), ‘Bases de una Política Integral de Vivienda,’ in B. Cuenya and A. Falú (eds.), Reestructuración del Estado y Política de Vivienda en Argentina. Buenos Aires: Centro de Estudios Avanzados, Universidad de Buenos Aires.

Tanzi, V. 1996. ‘Fiscal Federalism and Decentralization: A Review of some Efficiency and Macroeconomic Aspects,’ In Bruno, M., and Pleskovic, B. (eds.), Proceedings from the 1995 Annual World Bank Conference on Development Economics. Washington, D.C.: World Bank.

Tiebout, C. 1956. ‘A Pure Theory of Local Expenditures,’ Journal of Political Economy 64 (5): 416-24.

Viola, A. 2000. ‘La situación del sector vivienda en la República Argentina,’ Colección, Escuela de Ciencias Sociales, Universidad Católica Argentina, Año VII, No. 11, pp. 227-234,

World Bank. 2005. Decentralization and Subnational Regional Economics - What, Why, and Where?

Wilensky,  H. 1974. The Welfare State and Equality: Structural and Ideological Roots of Public Expenditures. Berkeley: University of California Press.

Zanetta, C. 2004. The Influence of the World Bank on National Housing and Urban Policies: A Comparison of Mexico and Argentina during the 1990s. London: Ashgate.

Zanetta, C. 2004b. ‘Housing Policy and Economic Reform: Comparative Case Examples from Mexico and Argentina’, in N. Hamdi (ed.), Urban Futures: Economic Growth and Poverty Reduction. London: ITDG Publishing.


Return to top