Regional Environmental Change

, Volume 18, Issue 6, pp 1655–1666 | Cite as

Federalism, water, and (de)centralization in Brazil: the case of the São Francisco River water diversion

  • Vanessa Lucena Empinotti
  • Wilde Cardoso GontijoJr
  • Vanessa Elias de Oliveira
Original Article


The Brazilian Northeastern semi-arid region’s history has been characterized by drought, water scarcity, and poor living conditions. For the last 100 years, institutions have been established to support the state sector’s different attempts to solve the water scarcity and poverty problems in the region. In the Brazilian federalist model that combines centralized decision power with a decentralized execution of public policies, the federal government promotes large-scale infrastructure interventions, such as the São Francisco River Water Diversion Project (SFRWDP). This article aims to show how intergovernmental relations, under the Brazilian federal system, enable centralized decision-making processes. We will analyze the roles of subnational and national entities and the factors influencing their interaction. Even though decentralized participatory water institutions are in place, the federal government controlled and centralized the process of approval of the SFRWDP as well as its implementation and management. While federal agencies dispute control over decision-making processes and financial resources, the states lack financial and institutional capacity to use the power that is attributed to them by the Water Law. The persistence of a federal system that perpetuates the federal government’s central role and the interstates’ power inequalities, may continue to wreck any attempt to promote decentralization and subnational and interstate collaboration in the Brazilian context.


Water governance São Francisco River Water diversion Brazil Federalism Decentralization 


The management of water in federal countries highlights the challenges of intergovernmental and multi-level water governance (Garrick and De Stefano 2016). Such dynamics respond to the political system in place in which institutional arrangements are designed to coordinate different scales of decision-making processes. In such cases, states and municipalities have autonomy over the use of their resources, being limited (or not) according to distinct federalist models.

To comprehend intergovernmental dynamics, however, it is important to understand the processes that shaped federal systems in different countries. Federalism, as a form of government, entails the autonomy coming from regional units in which power and functions are distributed between central and regional authorities (Smith 1995).

In Brazil, the federal government has taken water as a strategic resource to promote development, and decisions regarding water use and management have been historically a top-down process controlled by the federal government. However, through the Constitution of 1988, the re-democratization of the Brazilian political system promoted a new wave of decentralization of decision-making processes and financial resources, which established a modern federalism in Brazil. In this context, the 9433 Water Law, approved by the Brazilian Congress in 1997, created one of the most concrete opportunities for the Brazilian political system to foster decentralization of decision-making processes as well as recognizing subnational organizations, such as watershed committees and councils. Such changes created the need for the federal government to negotiate with other political spheres to manage water.

This paper aims to explain how intergovernmental relations, under the Brazilian federal system, enable centralized decision-making processes. We will analyze the role of subnational and national entities and the factors influencing the cooperation among these groups. Starting with a discussion on the Brazilian federalist model, which combines a centralized decision power with a decentralized execution of public policies (Arretche 2012), we will look at the São Francisco River Water Diversion Project (SFRWDP), a polemic project that would bring water from the São Francisco River watershed1 to four other states in the Northeast region, in order to foster development, alleviate poverty and drought, and respond to human demand (Lima 2005). The project involves the construction of channels, water pumping stations, small reservoirs, and hydroelectric plants. Its two major channels will supply water to 12 million people. The project was approved in 2005 and partially inaugurated in March of 2017.

This research took place between 2016 and 2017 and used a mixed set of methods, combining participant observation with document analysis and internet research, to gather information on the events, processes, and issues that were involved in the SFRWDP case. We also participated in seven key meetings where the water diversion was being discussed and organized with representatives from the Water National Agency (ANA), São Francisco and Parnaíba Valley Development Company (CODEVASF), the Ministry of National Integration (MNI), and Casa Civil (the Brazilian equivalent of the presidential chief of staff’s office). Further, we conducted textual analysis of relevant documents, such as the SFRWDP reports, resolutions, and government agreements as well as academic articles. From these results, we engaged in an analysis considering the political practices under the federal Brazilian system and how they impact the subnational water management practices in place today.

This article is structured as follows. First, we review water governance literature, discussing how the concept of politics of scale (Swyngedouw and Heynen 2003) helps to understand the reasons and challenges for water management decentralization. We then discuss federal political systems as well as the characteristics of the Brazilian federal model. The article provides a historical overview of the Brazilian Northeast, focusing on the SFRWDP case, and highlights how this infrastructure project has been shaped over time by the institutional agreements created to implement it. We discuss the ways in which the complex interactions among subnational, interstate, and federal spheres in the political practices under the federal Brazilian regime promote the federal government’s persistence as a major protagonist in a federal system.

Decentralization of water management, politics of scale, and the need to consider federal practices

Water management studies often consider that the problem of water access would be solved by means of technology and infrastructure planning approaches, consistent with the rational planning tradition. In so doing, they do not account for the political and subjective aspects of such a problem (Furlong et al. 2016; Obertreis et al. 2016). This approach is guided by the so-called hydraulic mission, which celebrates the control of nature through the construction of large-scale infrastructure, such as dams (for water storage) and channels (for water distribution). It creates hydraulic bureaucracies to address the challenges of flood protection, hydropower generation, and large-scale public irrigation (Molle et al. 2009). Management practices work as part of a centralized and expert-driven system.

On the other hand, the water governance approach highlights political and social aspects of water planning and management. The idea is to consider not just the expert view and the state’s apparatus, but also other social actors through a participatory institutional approach and decentralized decision-making processes. This model is heavily influenced by a liberal approach where informed citizens and entrepreneurs can rationally choose and influence political and market options, thus making governments and corporations more accountable and efficient (Hetherington 2011). In this context, multilateral agencies have defined the water crisis largely as a governance crisis (OECD 2015).

Although the governance approach has brought attention to decision-making processes, multilateral agencies understand governance as an instrumental, technical, supposedly neutral management process or policy strategy (Castro 2007). The focus has been on building capacity at local levels, on developing practices of resource allocation, and having clarity of roles and responsibilities (OECD 2015).

However, critical studies on natural resources governance and decentralization have shown that the challenges to implement water governance practices go beyond instrumental fixes (such as tools, rules, capacity building) and instead show the need for political structural changes. Worldwide, studies have found that the decentralization of decision-making processes was not enough to disrupt power relations based on inequality, hierarchical power structures that work through clientelistic centralized political practices which, in turn, permeate the state structures (de Freitas 2015; Ribot and Agrawal 2006). Such political practices translate into the persistence of insufficient transfer of power to local institutions, lack of accountability from representatives, incipient participation of marginalized groups, and the reinforcement of local elites in the control over natural resources (Agarwal 2001; Brosius et al. 1998; Ribot 1996). Critical studies on natural resources governance show how structures of power and their players can readapt and appropriate the decentralized institutions to guarantee their access and control over natural resources (Batterbury and Fernando 2006).

The analysis of relations between state and society in the water context should go beyond local scale and should consider water bio-political characteristics that cross political boundaries, interconnect multiple government scales and ask for multi-governance practices (Bakker 2010). Such characteristics dialogue with the concept of politics of scale that assumes that socio-ecological processes give rise to scalar forms of organization, such as states and interstate arrangements, and to a nested set of related and interacting socio-ecological spatial scales (Swyngedouw and Heynen 2003). These networked spatial scales are never set, as they are periodically transformed, being considered an integral part of social strategies to combat and defend control over limited resources and/or struggles for empowerment (Delaney and Leitner 1997; Marston 2000). In the case of water, each scale evokes different power geometries and may lead to radically different sociological and ecological conditions (Swyngedouw et al. 2002; Kaika 2003; Swyngedouw and Heynen 2003).

Recently, authors have paid more attention to the interaction between politics of scale and water governance (Norman et al. 2012; Norman et al. 2015). The scale where water governance practices occur transcends the physical and hydrological borders of river basins (Budds and Hinojosa 2012; Warner et al. 2008; Lebel et al. 2005). Despite the decentralization of decision-making of the river basin scale, laws and regulations are normally defined in other spheres of negotiation such as the national scale; likewise, productive sectors, such as mining, tend to access water from different river basins, and effluents can also be discharged beyond the original watershed (Budds and Hinojosa 2012; Warner et al. 2008). It is also possible to observe how water resource issues connect players from different scales and how, as a result, new spaces for negotiation are created across the established scales.

Such research contributions were key to understand power asymmetries and to start to unveil political practices influencing decision-making processes in the context of water governance; however they did not consider how current processes and policies play out against the backdrop of specific historical, economic and political trajectories from the federalism perspective. The dependency between historical trajectories and political practices is critical to understanding both regional integration and the related water management processes (Movik et al. 2016). Water management practices should also be understood as process of state-led integration. Thus, it is key to understand how federal system dynamics influence the ways that subnational organizations interact with one another in decentralized systems, and how these impact projects’ outcomes.

Federalism and water governance: the challenges over borders and domains

Brazil became a federal country through the Constitution of 1891, following Brazil’s independence and the implementation of the republican model (Souza 2002). The Brazilian federal model was influenced by the Portuguese colonial practices. Based on the distribution of capitanias hereditárias,2 Portugal could maintain its control through a centralized model in which local managers responded directly to the Portuguese Crown (Castro 1975). Such distribution was kept during the Imperial period and by 1891 the new states represented one of the 18 former colonial capitanias. Different from other federal countries, the Brazilian states were not autonomous and independent and their population spoke the same language (Schiefler 2010). Therefore, in its origin, federalism emerged as an institutional mechanism to hold together the subnational units that sought separation. This is what Stepan (2004) called “hold together federations”, as opposed to “come together federations”, which have the objective of uniting separate/independent territories. Its genesis explains the subsequent intergovernmental relations, with strong presence of the federal government in the definition of policy and public policies in subnational governments.

As in many federations, Brazilian political system had to deal with significant regional inequalities (Souza 2002) since its beginnings. In order to alleviate such inequalities, the Brazilian federalism is marked by the transference of resources from more developed states and regions to the least socio-economically developed ones. In this sense, Brazilian federalism is more akin to the cooperative model, such as in the German model, than to the competitive one, as the American model (Kincaid 1990; Fischman 2005). Those practices influence how the Brazilian federal government has targeted development in the country, promoting and financing infrastructure programs and incentives that should address the regional inequalities.

Going through times of greater decentralization (democratic periods) or greater centralization (authoritarian periods) throughout the twentieth century, and after 20 years of bureaucratic-authoritarian regime (1964–1984) (Mello 1976; Selcher 1989), in the Constitution of 1988 the Brazilian Federation is transformed into the most decentralized model in Brazil’s history, with autonomy assured not only to the states, but also to all municipalities, in a rather atypical federal model. At the same time, such reforms shared responsibilities between the federal government, the states and the municipalities, regarding the provision of health care, social assistance, education, culture, housing and sanitation provision, environment, cultural and historical protection, and poverty alleviation. Complementary laws should define forms for cooperation among the three levels of government (Almeida 2005). This threefold model entails an important challenge for Brazilian intergovernmental relations, especially in the management of public policies involving different levels of government, as is the case with water resource management.

In order to deal with this challenge, the federation created with the 1988 Constitution combined a broad legislative authority of the federal government with little veto power for subnational units—thus preventing the decision-making paralysis (the “joint decision trap”) (Scharpf 1988). In addition, federal regulation affects the way decentralized policies are implemented. The constitutional rules, the authority of the federal ministries to regulate and supervise implemented policies, as well as the spending power of the federal government are important explanatory factors of the agenda of the subnational governments (Arretche 2012). Thus, the federal government becomes the decision-making sphere (policy decision-making) and the states are the executor sphere (policy-making) (Arretche 2010).

Classified as a “cooperative federalism”, the main criticism of the Brazilian model relates to the unclear division of competences between levels of government, which on one hand favors varied institutional arrangements through the design of public policies but, on the other, makes the management of inter-section and inter-federative policies enormously difficult. This scenario becomes even more complex with the incorporation of new social players in the decision-making process of several policies, something guaranteed by the presence of the spheres of participation incorporated by the Constitution of 1988.

The 9433 Water Law approved by the Brazilian Congress in 1997 reproduces such characteristics for the water sector. As a law rooted in the principles of decentralization, participation, and the acknowledgement of the multiples uses of water, it created new spaces of negotiation, such as federal and state councils and watershed committees (Abers and Keck 2013). It is important to note that this policy has a territorial component, the watersheds, which in their institutional design require a multi-level governance (Pierre and Stoker 2000), which does not follow the political-administrative division of the territory.

While the 9433 Water Law is an opportunity for the Brazilian political system to foster decentralization of decision-making processes (Souza 2005), it also poses challenges. The interdependency among state, municipality and the federal government to manage federal rivers3 can paralyze the system. Finally, the incapability of watershed committees to generate enough revenue to maintain the water system4 reinforces the financial dependency over the federal government’s resources and decreases subnational organizations autonomy and capacity to manage water distribution and allocation (Lemos and Oliveira 2004; Demajorovic et al. 2015).

The questions discussed above reproduce the tensions in implementing a decentralized policy-making system in a centralized policy decision-making federation, in which regional inequalities and historical dependency over federal funds poses challenges on how interstate cooperation and subnational structures should work.

The São Francisco River Diversion Project: challenging decentralization

In 2001, the Ministry of National Integration proposed the São Francisco River Water Diversion Project (SFRWDP), the latest but certainly not the last attempt to deal with the problem of drought, poverty, and lack of access to water in the Northeast region of Brazil. It is key to understand that the Brazilian Northeast is one of the poorest regions of the country, in which one third of the 45 million population lives in absolute poverty (IBGE 2010). The reasons for such conditions are a combination of social and environmental factors. On the one hand power structures were historically molded through access and control over natural resources, such as water and land (Furtado 2000). Strong elites comprised of leading families and politicians at local, state, and federal levels, who control access to land, water, and development, are responsible for promoting a clientelistic way of governing resulting in poor quality of public administration and inequality (Tendler 1997; Kenny 2002; Nelson and Finan 2009). On the other hand, the region experiences a semi-arid climate, with the earliest recorded occurrence of drought dating back to the sixteenth century (Greenfield 2001). More specifically, drought occurs in parts of eight Northeast states (Bahia, Alagoas, Sergipe, Pernambuco, Paraíba, Rio Grande do Norte, Ceará and Piauí) and parts of the northern Minas Gerais. This region covers 57% of the Northeast, where annual average precipitation is less than 800 mm/year (Rebouças 1997).

Understanding that the problem of water scarcity, erratic rainfall and recurrent drought as essentially a supply problem, the Brazilian State had proposed the diversion of the São Francisco River waters since the late 1800s (Rebouças 1997). The persistence of such techno-economic perspective reflects how the State’s attempts to deal with the consequences of drought have been heavily influenced by the hydraulic mission approach, combined with centralized decision-making occurring to the late 1990s. Under a centralized and paternalistic development approach, water has been a major factor in promoting development in the Northeast region and in Brazil as whole. The São Francisco river plays a key role as it is the most important river in the Northeast (de Freitas 2015). With a length of 2800 km and an average flow of 2880 m3/s, the river supplies water to approximately 19 million people. Its watershed is comprised of 521 municipalities across the Federal District, Goiás, Minas Gerais, Bahia, Pernambuco, Sergipe, and Alagoas (IBGE 2010), within an area of 640,000 km2, reservoirs and six hydropower plants provide water mainly for irrigation and energy production responding to 12% of the national electricity demand (CBHSF 2016).

It was in 2003, under President Lula that the water diversion option returned to the political agenda. Lula was aided by the Minister of National Integration and former governor of the state of Ceará, Ciro Gomes, who had a significant role in implementing the new water policy in that state in 1993 and who represented a new generation of business-oriented politicians.5 The project’s goal was to bring water to four states in the Northeast region (Ceará, Paraíba, Pernambuco, and Rio Grande do Norte) to guarantee water access to 12 million people in 391 municipalities, to foster regional development mainly at urban areas, to alleviate poverty and drought (Soares 2013). It aimed to diverge 2% of available water from the São Francisco River via two axes: North (Eixo Norte) and East (Eixo Leste), which will take water from this river and transport it through a channel system6 (see Fig. 1). There were two main factors supporting the project. One was that 28% of the Brazilian population living in a region where only 3% of the Brazilian water is available. The second factor tries to address the uneven water distribution within the Northeast region. According to the SFRWDP proposal, the São Francisco River watershed presents water availability which ranges between 2000 to 10,000 m3/hab/year, while the recipient area has 400 m3/hab/year, much below the 1500 m3/hab/year recommended by the UN (de Castro 2011). The main goal was to guarantee water supply for population consumption.
Fig. 1

The São Francisco Basin and the São Francisco River Water Diversion Project

The process of approving the SFRWDP was controversial because there was no consensus that such a project would solve the water scarcity problem in the region (Rebouças 1997). The distribution between supporters and non-supporters followed geographical features in which recipient states’ governments and civil society organizations supported the SFRWDP and the ones placed in the donors’ area were against it,7 including the members of the São Francisco River Watershed Committee (SFRWC)8 created in 2001 (Roman 2017).

When the federal government decided to put into effect SFRWDP, ANA—who regulates federal waters—became responsible to prepare the São Francisco River Watershed Plan (CBHSF 2016). According to the law, the watershed plans should be approved by their respective watershed committees. When ANA presented the plan to the SFRWC in 2004, which included the SFRWDP, the document was not accepted at first (Abers and Keck 2013). In response, a working group composed by watershed committee members analyzed and rewrote a new version of the plan. The SFRWC also elaborated the Deliberation 18 that prioritized the water use in the basin. Yet, the use of water outside the basin is allowed so long as it is for human consumption and animal needs (PHBSF 2004). Such rules made the SFRWDP first proposal impractical.

The federal government did not accept the SFRWC deliberation and decided to bring such issue to the Water National Council (WNC), which has the final decision and where the federal government maintains 51% of the seats (Abers and Keck 2013). At the same time, civil society organizations initiated administrative and judicial proceedings against the execution of the SFRWDP (Roman 2017). In January 2005, before the WNC section that approved the SFRWDP, the Casa Civil suggested that all the federal government representatives vote for SFRWDP approval, overriding the watershed committee’s decision (Empinotti 2011; Abers and Keck 2013). Finally, the Brazilian Judiciary system also supported the federal government approval of the diversion plan and in 2007 the SFRWDP construction started (Roman 2017).

Some of the reasons to promote a decentralized water governance system is to increase proximity between decision-makers and those affected by governance decisions, an increase in subnational level democratic participation, greater access to local knowledge and expertise, as well as to intensify responsiveness to citizen’s needs and concerns (Cohen and Davidson 2011). However, even though the SFRWDP approval process took place under decentralized and participatory institutions, federal scale was activated through the alignment among federal state agencies and ministries, unveiling asymmetries of power and control using the State structure to override such decision.

The SFRWDP case also unveils how watershed boundaries are overruled and the institutional disparity among states. On the one hand, we observed that the negotiation over the SFRWDP crossed the watershed boundaries once it dealt with water transfers among different watersheds, as well as water distribution and access between regions with different levels of water scarcity. Such networking and power flows, basically created to guarantee access to water, defined a hydrosocial territory (Boelens et al. 2016). The complexity and spaces for negotiation should reflect this new territory that goes beyond the watershed boundaries.

This brings us to the second point: the institutional imbalance among states. Up to 2005, during the SFRWDP approval process, there were few state watershed committees9 in place at the recipients’ watersheds that could represent their own agenda. Thus, the federal government had the opportunity to assume the goal for defending water distribution beyond the SFRW borders. The nonexistence of watershed committees in the majority of recipient states reflects the lack of technical support as well as the low interest from state governments to create the water system institutions (Moreira 2010). The institutional inequality indicates different stages of institutional maturity among states and calls the federal government to counterbalance such inequalities. Such an arrangement reproduced the federal government practices of privileging massive infrastructure implementation in order to decrease inequalities among states. Yet, such practices reinforce the federal state control over decision-making (Souza 2005), which in turn reifies centralized practices in a federal nation.

Federal and subnational scales of decision to manage the São Francisco River Water Diversion Project implementation

The federal government’s control over the SFRWDP continued towards the project’s implementation as we will discuss in this section. In September 2005, based on the proposal drawn up by the inter-ministerial working group (Brasil 2005a), the Term of Agreement was defined between the federal government—represented by the MNI, the Ministry of Mines and Energy (MME), the Ministry of the Environment (MMA), the Casa Civil—and the four beneficiary states (Brasil 2005b). In this model, the work would be carried out by the MNI, which would also be in charge of the articulations within the federal government and with the external players involved. When the work was completed, the federal operator would take care of the operation and the maintenance of the infrastructure, delivering the distribution of water in each state to the states’ operators. The states’ operators should be created or qualified for such a function by means of nominations or legal adjustment of competencies. They would be responsible for all operation costs and by the charging of tariffs for such services and the water usage. The regulation of the water usage of the São Francisco River and the adduction service provided by the project operator, including the approval of the corresponding tariff, would be carried out by ANA (Fig. 2).
Fig. 2

The São Francisco River Water Diversion Project Management System

In addition, the MNI had been working on the Management Board (MB) initially composed by representatives of six federal ministries, and of the four recipient states. Established by Decree No. 5995 in December 2006 (Brasil 2006), the board main objective was the general coordination of the project implementation, ensuring the fulfillment of the initial goals, as well as the proposal elaboration for the management model after the completion of the work. However, in 2014 a new agreement10 was signed, which also included in the MB, one representative from the São Francisco River Watershed Committee, and another from the Salgado River Watershed Committee from the recipient region (Brasil 2014).

In so doing, we can infer the prevalence of the public sector representatives at state and national scale over the watershed scale. However, the Term of Agreement also indicated the centralization of decisions within the national scale, by which the MNI became the main project manager and also responsible for the implementation and distribution of the federal government’s financial resources (Brasil 2005b). The dispute over the projects’ control started to take place first among different ministries and federal agencies, and later within the states representatives that activated new networks and scales of action. Such tensions became apparent through the definition of the system’s federal operator.

According to the Term of Agreement, the MME was responsible for creating the legal conditions for the São Francisco Hydropower Company (Companhia Hidrelétrica do São Francisco, CHESF) to become the federal operator of the SFRWDP. However, this option did not please the MNI due to the fact that CHESF would also be the water provider for other uses, and the user of the same water to generate energy in the São Francisco River basin. Another reason for the opposition to CHESF was its subordination to MME, who was competing with the MNI in the management of the project construction (Banco Mundial 2011). It is important to notice that, historically, the MNI is controlled by politicians from the Northeast region since it is the Ministry that promotes regional development and distributes financial resources for water-related infrastructure. In order to counteract this decision, the MNI proposed one of its own agencies, CODESVASF, to become the federal operator. As a result of this strategy, the MNI became the sole Ministry to manage and control the diversion project’s implementation and further operation.

Centralization of decision-making took place within the federal agencies when they competed among each other for the project’s control, unveiling political agendas and reinforcing historical power distribution within the federal government (Abers and Keck 2013). Such a dynamic was overpowering the institutional arrangement created to implement and coordinate the project. This was the case in relation to ANA. In response to the delays in the work of the SFRWDP between 2006 and 2009,11 ANA drew up a report which showed that the commitments made were not being fulfilled. Resolution ANA No. 714, of September 28, 2009, was edited, determining the completion of actions by MNI and the states, with details and deadlines, but without any penalty clause for non-compliance with the requirements. The work and actions had not been fulfilled by the MNI, as shown by the technical note No10/2011/SRE-ANA. Since June 2011, the MNI requested the extension to the deadlines for implementation of the project three times. The deadline in place today is March 26, 2018 defined by the technical note ANA No. 1133.

Clearly, the institutional arrangements created to coordinate and to implement the project were not working. The main players made isolated decisions and deadlines were not being met. The lack of dialogue among federal government agencies, states and watershed committees become more obvious when we analyze the MB. The dynamics within the MB reproduced federal regulation practices in which the federal government is the decision-maker and the states are the executors (Arretche 2010). Until 2014, the MB discussed and followed the progress of the work without presenting any solutions in regard to the delays which persisted. Subsequently, the MNI used the MB to present the costs of the work to be paid by the respective states as indicated in the 2005 Term of Agreement (Brasil 2005a, b). The relationship between the federal government agencies and recipient states resembled one of clients rather than partners. The Term of Agreement, compiled and presented by an inter-ministerial working group, defined the states roles without negotiating with them first. For example, the states should be responsible for investing in capacity building, they should pay the costs of operation and maintenance, and they should guarantee such payments and be subject to the delivery conditions defined by ANA in the grant of water rights usage (Brasil 2005a, b). However, there was no negotiation with the states to verify if they had the economic condition to execute the agreement. So, even though the institutional structures offered opportunities for collaboration, the centralized decision practices prevailed. The federal government, in this case the MNI, controlled the whole process and neutralized any opportunity for collaboration and dialogue.

Such disparity also reflects Brazil’s choice of adopting a symmetric federalism in an asymmetric federation (Souza 2005). The lack of financial state autonomy and the recurrent dependency of poor states over the federal government is a characteristic of the Brazilian Federalist System. Hence, states in Brazil are continually struggling for their autonomy as they are not able to exercise it in practice. For example, Pernambuco, Paraíba, and Rio Grande do Norte state agencies are not ready to become state operators. The charges for the use of water were implemented only in part of the state of Paraíba and the amount collected does not exceed 3 million reais. In the states of Pernambuco and Rio Grande do Norte, this provision has yet to be employed. The state compliance improves when such conditions are measured in the state of Ceará. Only the state of Ceará has defined its state operator, elaborated a financial support model for the operation, and partially created infrastructure to receive the water coming from the São Francisco River basin. It should be noted, however, that such a situation was already present in this state even before the signing of the Terms of Agreement.

As an attempt to renegotiate the terms and commitments established in 2005, in February of 2017 (Estados 2017) the four state governors sent a letter to the federal government suggesting that payments should only start after the completion of the work, and after 5 years of operation the states would assume the total cost of the project; and a total exemption of federal taxes on the income from water tariffs. In addition to these suggestions, they have argued for the expansion of the federal government’s investments in the implementation of the distribution of the water in the states, far beyond what was agreed in the 2005 Term of Agreement. On March 21, 2017, the MNI and Casa Civil sent an answer to the Governors reaffirming their commitment to the Term of Agreement and the beginning of charges by March 28, 2018, as determined by ANA.

When looking at the interaction among the state recipients we observe that they were not totally aligned in the beginning. Having the Ministry of National Integration as Ciro Gomes, a former governor of Ceará, the balance among recipient states was also dubious. On the one hand, Ceará state already had in place all institutional, financial and infrastructure structures to receive the diverted water, been ready to benefit from it. In certain way it created a powerful alliance between MNI and the state of Ceará. However, when Ciro Gomes left the ministry, Ceará aligned with the water states in order to renegotiate former commitments as presented above. Such interactions exemplify how networked spatial scales are never set, but are periodically transformed to guarantee access over limited resources such as water. The shifts in power change the importance and the role of certain geographical scales and also possibly creating entirely new scales (Cox 1998; Marston 2000).

The SFRWDP implementation process helped us to understand how federal system dynamics influence the ways that subnational organizations interact with one another in decentralized systems, and how these dynamics impact project outcomes. The control of the water diversion project by the MNI neutralized other players’ influences and at the same time created a top-down relationship mainly between the MNI and recipient states reproducing the control of the federal scale over others. The MNI centralizing behavior was so severe that it became the only federal player, and subjugated all others to it. At the same time, donor states input was diminished during the negotiations, only being marginally represented by a SFRWC member and easily outnumbered by other members at the Management Board.

As presented above, these dynamics of Brazilian federalism definitely influenced the prevalence of the federal sphere over others; the fragile participatory institutions and institutional and financial asymmetries within the Brazilian states. The consequences of these political practices activated alliances and networks in order to guarantee power through water control and access (Swyngedouw and Heynen 2003). The concentration of power and control by the MNI also shaped the scale and created specific networks that were mainly contained at the national and state level in order to promote a development model strongly aligned with the politicians of Ceará’s agenda. These politicians were able to assume positions at federal and state level and influence how the water diversion project was approved and implemented, indicating that networks and the capacity of political mobility are behind the activation of scales (Budds and Hinojosa 2012).


Socio-technical change such as the decentralization of decision-making processes is seen less as a transition from one path to another than as a largely messy, contested, and discursive process, strongly framed by contexts of actions and contingent events (Obertreis et al. 2016). In the water sector in Brazil, such changes are influenced by political practices which characterize the Brazilian federal system. On one hand, water management still follows the hydraulic mission when it relies on major infrastructure projects to guarantee water supply. On the other hand, the struggle to promote a decentralized system reflects the interdependencies among states and the federal government. In this article, we have shown how current processes and policies play out against the backdrop of specific historical, economic, and political trajectories (Movik 2016). These dynamics highlight how decentralization of decision-making processes is affected by controversies.

The decisions towards the SFRWDP led by the federal agencies overruled the watershed physical boundaries, empowered the federal scale and gave the opportunity for the recipient watersheds and their states to influence the implementation and management processes. However, two key points should be highlighted. First, the federal government as the decision-maker sphere struggled for power not only with states but also among federal agencies. Such control perpetuated political practices of paternalism and clientelism between federal and state governments and maintained a close presence of Northeast politicians at federal agencies. This in turn shapes intergovernmental relations and reinforces politics of scale in which the federal government is key in controlling flows of financial and institutional resources towards specific programs and regions.

Second, the lack of institutional and financial capacity from recipient states did not allow them to exercise the power that is attributed to them by the Water Law as the executor’s sphere. Such convenient dependency on the federal government influenced the final outcome in which decentralization is challenged. Finally, states with more capacity are able to align their own agenda and influence the federal government actions through the presence of politicians assuming key positions at the federal government and creating a strong network to guarantee access to water.

The role of subnational and national entities directly impacts how intergovernmental relations create new political scales that reinforce a centralized federal system. The option of placing the federal government as the central player imposing and implementing the water diversion project exposed the weakness of the water management system in Brazil. When the federal government imposes its preferences over interstate and subnational institutions, it discourages any attempt to exercise decentralization and public participation, and becomes one of the reasons for the disengagement of civil society participation within watershed committees and councils (Empinotti 2011). On the other hand, since the federal government defined the states commitments without any prior negotiation, three of the recipient states did not assume their responsibilities in operationalizing and monitoring the water system, as well as in investing in infrastructure and expert training.

The political practices constructed around a federal system with a strong federal government and regional inequalities contribute neither to subnational cooperation nor to the elaboration of common information system and the definition of financial and operational responsibilities, as proposed by multilateral agencies. These practices challenge the efficiency of the new Brazilian water policy to a point where such a policy is incapable of responding and managing water crises (Gontijo 2013; Jacobi et al. 2015). The persistence of a federal system that accommodates the federal government’s central role and the states’ power inequalities will continue to wreck any attempt to promote decentralization and subnational and interstate collaborations in the Brazilian context.


  1. 1.

    The São Francisco River watershed covers six Brazilian states and the Federal District: Goiás, Minas Gerais, Bahia, Pernambuco, Sergipe and Alagoas (ANA 2013).

  2. 2.

    Capitanias hereditárias were a feudal system in which the King’s vassals were responsible for managing large amounts of land representing the Portuguese Crown’s interests. This system was put in place to defend the Colony and to better exploit it by capturing natives, taking their land, imposing Europeans beliefs and transforming them into submissive agents of the colonizers’ domain (Castro 1975).

  3. 3.

    According to the Constitution of 1988, rivers that cross two states are in federal domain while the others are in domain of the states. So, in watersheds where the main river is under the federal domain, at least three different domains will be in place (Lemos and Oliveira 2004). This is the case in 75% of the Brazilian territory, which pressures for interstate water negotiation (Gontijo and Trigo 2013).

  4. 4.

    The 9433 Water Law defined that the water system financial resources should be provided through charging water from users. Such revenue should be able to financially support the respective river basin water agency, research projects and infrastructures to improve water quality and availability. The amount to be charged is defined by the watershed committee’s members (Lemos and Oliveira 2004).

  5. 5.

    Since 1985, a group of young industrial entrepreneurs started to push for a new agenda towards industrialization, the use of market tools to promote efficiency to promote a self-sustainable production system. Participation was key to promote and legitimize such structural shifts. Considering water as a key supply to promote economic activities, the 1993 water state law became an essential fuel for industrial growth and services concentrated in larger urban areas. For these politicians, the SFRWDP was key to promote such development model in Ceará (Gontijo 2013).

  6. 6.

    The water will be redirected and distributed to four states through 477 km of canals, 4 tunnels, 13 aqueducts, 9 pumping stations, and 30 reservoirs. The total investment as for 2013 was over 2 billion dollars and is being financed by federal resources (Soares 2013).

  7. 7.

    To understand the reasons that lead donors to go against the project, check Roman (2017).

  8. 8.

    Comprised of 62 voting members representing the state, water users, and civil society organizations, the SFRWC is responsible for designing a new system for the allocation of water rights, creating different forums for public participation and implementing water charging at the watershed level (De Freitas 2015).

  9. 9.

    The majority of existing state watershed committees were at Ceará state (ANA 2013).

  10. 10.

    In 2006, the first agreement did not include as members of the MB representatives from watershed committees in place at the recipient watersheds because they did not exist. On the other hand, the SFRWC was against the SFRWDP and refused to participate at the MB. By 2014, there were watershed committees in place at the recipient watersheds as well as the SFRWC now was on board after implementing water charging from the water diversion project. Because of that, two seats were created.

  11. 11.

    Between 2006 and 2009, there were many implementation difficulties due to gaps in the technical project, legal tribulations related to environmental suits, and even climatic considerations: it rained more than normal (ANA 2009).


  1. Abers RN, Keck M (2013) Practical authority: agency and institutional change in Brazilian water politics. Oxford University Press, OxfordCrossRefGoogle Scholar
  2. Agarwal B (2001) Participatory exclusion, community forestry, and gender: an analysis for South Asia and a conceptual framework. World Dev 29:1623–1648CrossRefGoogle Scholar
  3. Almeida MHT (2005) Decentralization and centralization in a federal system: the case of democratic Brazil. Rev Social Polit 24:29–40CrossRefGoogle Scholar
  4. ANA (2009) Relatório de Fiscalização da Comissão Especial de Acompanhamento do PISF criada pela Portaria ANA n° 97, de 30 de abril de 2009Google Scholar
  5. ANA (2013) Conjuntura dos Recursos Hídricos no Brasil 2013. ANA, BrasíliaGoogle Scholar
  6. Arretche M (2010) Federalismo e igualdade territorial: uma contradição de termos? Dados 53:587–620. CrossRefGoogle Scholar
  7. Arretche M (2012) Democracia, federalismo e centralização no Brasil. Fiocruz, Rio de JaneiroCrossRefGoogle Scholar
  8. Bakker K (2010) Politics and biopolitics: debating ecological governance. In: Bakker K (ed) Privatizing water: governance failure and the world’s urban water crisis. Cornell University Press, London, pp 190–212Google Scholar
  9. Banco Mundial (2011) Relatório de Análise das Alternativas de Gestão da Operação da Infraestrutura de Transposição do rio São Francisco. Relatório n° 3 – Contrato n° 0.129.01/2010. Brasília. DFGoogle Scholar
  10. Batterbury SPJ, Fernando JL (2006) Rescaling governance and the impact of political and environmental decentralization: an introduction. World Dev 34:1851–1863. CrossRefGoogle Scholar
  11. Boelens R, Hoogesteger J, Swyngedouw E, Vos J, Philippus W (2016) Hydrosocial territories: a political ecology perpective. Water Int 41:1–14. CrossRefGoogle Scholar
  12. Brasil (2005a) Relatório do Grupo de Trabalho criado pela Portaria Interministerial n° 24, de 11 de novembro de 2004, publicado em agosto de 2005Google Scholar
  13. Brasil (2005b) Termo de Compromisso Firmado entre a União, por intermédio dos Ministérios da Integração Nacional, de Minas e Energia, do Meio Ambiente e da Casa Civil da Presidência da República, e os Estados do Ceará, Paraíba, Pernambuco e Rio Grande do Norte, para garantia da operação sustentável do Projeto de Integração do Rio São Francisco com as Bacias Hidrográficas do Nordeste Setentrional - PISF. BrasíliaGoogle Scholar
  14. Brasil (2006) Decreto n° 5995, de 19 de dezembro de 2006. Institui o Sistema de Gestão do Projeto de Integração do rio São Francisco com as bacias hidrográficas do Nordeste Setentrional. BrasíliaGoogle Scholar
  15. Brasil (2014) Decreto n° 8207, de 13 de março de 2014. Altera o Decreto n° 5995 que institui o Sistema de Gestão do Projeto de Integração do rio São Francisco com as bacias hidrográficas do Nordeste Setentrional. BrasíliaGoogle Scholar
  16. Brosius JP, Tsing AL, Zerner C (1998) Representing communities - histories and politics of community-based natural resource management. Soc Nat Resour 11:157–168. CrossRefGoogle Scholar
  17. Budds J, Hinojosa LV (2012) Restructuring and rescaling water governance in mining context: the co-production of waterscapes in Peru. Water Alternatives 5:119–137Google Scholar
  18. Castro JD (1975) Sete Palmos de Terra e um Caixão. Seara Nova, LisboaGoogle Scholar
  19. Castro JE (2007) Water governance in the twentieth-first century. Ambiente e Sociedade 10:97–118. CrossRefGoogle Scholar
  20. CBHSF (2016) Plano decenal de recursos hídricos da bacia hidrográfica do rio São Francisco.
  21. Cohen A, Davidson S (2011) The watershed approach: challenges, antecedents, and the transition from technical tool to governance Unity. Water Alternatives 4:1–13Google Scholar
  22. Cox K (1998) Spaces of dependence, spaces of engagement and the politics of scale, or: looking for local politics. Polit Geogr 17:1–23CrossRefGoogle Scholar
  23. De Castro CN (2011) Transposição do Rio São Francisco: análise de oportunidade do projeto. IPEA, Rio de JaneiroGoogle Scholar
  24. De Freitas C (2015) Old Chico’s new tricks: Neoliberalization and water sector reform in Brazil’s São Francisco River basin. Geoforum 64:292–303. CrossRefGoogle Scholar
  25. Delaney D, Leitner H (1997) The political construction of scale. Polit Geogr 16:93–97. CrossRefGoogle Scholar
  26. Demajorovic J, Caruso C, Jacobi PR (2015) Cobrança do uso da água e comportamento dos usuários na bacia hidrográfica do Piracicaba, Capivari e Jundiaí 49:1193–1214.
  27. Empinotti V (2011) E se eu não quiser participar? O caso da não participação nas eleições do comitê de bacia do rio São Francisco. Ambiente e Sociedade 14:195–211. CrossRefGoogle Scholar
  28. Estados (2017) Carta dos Governadores dos Estados do Ceará, Paraíba, Pernambuco e Rio Grande do Norte apresentando ao Governo Federal pleitos e sugestões quanto ao modelo de gestão e ao custo de operação do PISFGoogle Scholar
  29. Fischman RL (2005) Cooperative federalism and natural resources law. NYU Environmental Law Journal 14:179–231Google Scholar
  30. Furlong C, De Silva S, Guthrie L, Considine R (2016) Developing a water infrastructure planning framework for the complex modern planning environment. Util Policy 38:1–10. CrossRefGoogle Scholar
  31. Furtado C (2000) Introdução ao Desenvolvimento: enfoque histórico-estrutural. Editora Paz e Terra, São PauloGoogle Scholar
  32. Garrick DE, De Stefano L (2016) Adaptive capacity in federal rivers: coordination challenges and institutional responses. Curr Opin Environ Sustain 21:78–85. CrossRefGoogle Scholar
  33. Gontijo Jr WC (2013) Uma avaliação da política brasileira de recursos hídricos baseada em dez casos de estudo. Dissertation, Universidade de BrasíliaGoogle Scholar
  34. Gontijo Jr WC, and Trigo AJ (2013) Domínio das águas no Brasil e a gestão integrada por bacia hidrográfica: reflexões sobre o modelo vigente. 4° Encontro Internacional da Governança da Água. São PauloGoogle Scholar
  35. Greenfield G (2001) The realities of images: imperial Brazil and the great drought. American Philosophical Society, PhiladelphiaGoogle Scholar
  36. Hetherington K (2011) Guerilla auditors: the politics of transparency in neoliberal Paraguay. Duke University Press, DurhamCrossRefGoogle Scholar
  37. IBGE – Instituto Brasileiro de Geografia e Estatística (2010) Censo demográfico de 2010.
  38. Jacobi PR, Cibim J, Leão RS (2015) Crise hídrica na Macrometrópole Paulista e respostas da Sociedade Civil. Estudos Avançados 29:27–42. CrossRefGoogle Scholar
  39. Kaika M (2003) Constructing scarcity and sensationalising water politics: 170 days that shook Athens. Antipode 35:919–954. CrossRefGoogle Scholar
  40. Kenny ML (2002) Drought, clientelismo, fatalismo and fear in the Northeast Brazil. Ethics, Place and Environment 5:123–134. CrossRefGoogle Scholar
  41. Kincaid J (1990) From cooperative to coercive federalism. ANNALS, AAPSS 509:139–152. CrossRefGoogle Scholar
  42. Lebel L, Garden P, Imamura M (2005) The politics of scale, position, and place in the governance of water resources in the Mekong Region. Ecol Soc 10:18–36CrossRefGoogle Scholar
  43. Lemos MC, Oliveira JLFD (2004) Can water reform survive politics? Institutional change and river basin management in Ceará, Northeast Brazil. World Dev 32:2121–2137. CrossRefGoogle Scholar
  44. Lima LC (2005) Além das águas, a discussão no Nordeste do rio São Francisco. Revista do Departamento de Geografia 17:94–100Google Scholar
  45. Marston S (2000) The social construction of scale. Prog Hum Geogr 24:219–242. CrossRefGoogle Scholar
  46. Mello DL (1976) O controle dos governos municipais. Revista de Administração Municipal, Rio de Janeiro, 36:27–42Google Scholar
  47. Molle F, Mollinga PP, Wester P (2009) Hydraulic bureaucracies and the hydraulic mission: flows of water, flows of power. Water Alternatives 2:328–349Google Scholar
  48. Moreira MMMA (2010) A Atuação dos governos estaduais nos comitês de bacia. In: Abers RA (ed) Água e Política. Annablume, São Paulo, pp 137–158Google Scholar
  49. Movik S, Mehta L, Manzungu E (2016) The flow of IWRM in SADC: the role of regional dynamics, advocacy networks and external actors. Water Alternatives 9:434–455Google Scholar
  50. Nelson DR, Finan TJ (2009) Praying for drought: persistent vulnerability and the politics of patronage in Ceará, Northeast Brazil. Am Anthropol 111:302–316. CrossRefGoogle Scholar
  51. Norman ES, Bakker K, Cook C (2012) Introduction to the themed section: water governance and the politics of scale. Water Alternatives 5:52–61Google Scholar
  52. Norman ES, Cook C, Cohen A (2015) Negotiating water governance: why the politics of scale matter, Ashgate, SurreyGoogle Scholar
  53. Obertreis J, Moss T, Mollinga PP, Bischsel C (2016) Water, infrastructure and political rule: introduction to the special issue. Water Alternatives 9:168–181Google Scholar
  54. PHBSF - Plano Decenal de Recursos Hídricos da Bacia Hidrográfica do rio São Francisco (2004) Comitê da Bacia Hidrográfica do rio São Francisco (CBHSF), Salvador.
  55. Pierre J, Stoker G (2000) Towards multilevel governance. In: Dunleavy P, Gamble A, Holliday I, Peele G (eds) Developments in British politics, 6th edn. Macmillan, London, pp 29–46Google Scholar
  56. Rebouças AC (1997) Água na região Nordeste: desperdício e escassez. Estudos Avançados 11:127–154. CrossRefGoogle Scholar
  57. Ribot J (1996) Participation without representation: chiefs, councils, and forestry law in the West African Sahel. Cultural Survival Quarterly 30:40–44Google Scholar
  58. Ribot J, Agrawal A (2006) Recentralizing while decentralizing: how national governments reappropriate forest resources. World Dev 34:1864–1886. CrossRefGoogle Scholar
  59. Roman P (2017) The São Francisco interbasin water transfer in Brazil: tribulations of a megaprojects through constraints and controversy. Water Alternatives 10:395–419Google Scholar
  60. Scharpf FW (1988) The joint-decision trap: lessons from German federalism and European integration. Public Adm 66:239–278. CrossRefGoogle Scholar
  61. Selcher W (1989) A new start toward a more decentralized federalism in Brazil? Publius 19:167–183Google Scholar
  62. Smith G (1995) Federalism: the multiethnic challenge. Longman, LondonGoogle Scholar
  63. Schiefler FR (2010) Ecos de federalismo: centralização e federalismo no Brasil (1820-l841). Revista Três, Belo HorizonteGoogle Scholar
  64. Soares E (2013) Seca no Nordeste e a transposição do rio São Francisco. Geografia 9:75–86Google Scholar
  65. Souza C (2005) Federalismo, Desenho Constitucional e Instituições Federativas no Brasil pós -1988. Rev Sociol Polit 24:105–121CrossRefGoogle Scholar
  66. Souza C (2002) Brazil: the prospects of a centre-constraining federation in a fragmented polity. Publius 32:23–48CrossRefGoogle Scholar
  67. Stepan A (2004) Toward a new comparative politics of federalism, multinationalism, and democracy: beyond Rikerian federalism. In: Gibson E (ed) Federalism and democracy in Latin America. Johns Hopkins University Press, Baltimore, pp 29–84Google Scholar
  68. Swyngedouw E, Page B, Kaïka M (2002) Sustainability and policy innovation in a multi-level context: crosscutting issues in the water sector. In: Heinelt H, Getimis P, Smith R, Swyngedouw E (eds) Participatory governance in multi-level context: concepts and experience. Springer, London, pp 107–131CrossRefGoogle Scholar
  69. Swyngedouw E, Heynen N (2003) Urban political ecology, justice and the politics of scale. Antipode 35:898–918. CrossRefGoogle Scholar
  70. Tendler J (1997) Good government in the tropics. Johns Hopkins U.P, BaltimoreGoogle Scholar
  71. Warner J, Wester P, Bolding A (2008) Going with the flow: river basins as the natural units for water management? Water Policy 10:121–138. CrossRefGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Federal University of ABCSão PauloBrazil
  2. 2.National Water Agency – ANABrasíliaBrazil

Personalised recommendations