Development of a model for the analysis of public policy networks in a federalism and coalitional presidentialism context

Abstract

This is a theoretical essay that proposes a model for analyzing Brazilian public policy networks. The model innovates by incorporating the aspects of federalism and coalitional presidentialism into network analysis, since they are crucial aspects of power relations in Brazil. The literature on networks points out that structural and operational factors, as well as the context in which networks are embedded, are relevant to comprehend their effectiveness. However, the underlying power relations are often neglected by Public Administration analyzes. Thus, in the proposed model, federalism and coalitional presidentialism were included in the contextual characteristics, in order to add elements of power to the analyses of public policy networks in Brazil.

Keywords:
Public policy networks; Federalism; Coalitional presidentialism

Resumo

Este artigo propõe um modelo útil para a análise de redes de políticas públicas brasileiras. Tal modelo inova ao incorporar à análise de redes os aspectos do federalismo e do presidencialismo de coalizão, por considerá-los aspectos cruciais das relações de poder no Brasil. A literatura sobre redes aponta que fatores estruturais e de funcionamento das redes, bem como fatores do contexto no qual as redes estão inseridas, são relevantes para compreender seu grau de efetividade. Porém, muitas vezes, as relações de poder subjacentes são negligenciadas pelas análises realizadas no âmbito da Administração Pública. Diante disso, no modelo proposto, o federalismo e o presidencialismo de coalizão foram incluídos às características de contexto, com o objetivo de agregar elementos de poder às análises de redes de políticas públicas no Brasil.

Palavras-chave:
Redes de políticas públicas; Federalismo; Presidencialismo de coalizão

Resumen

Se trata de un ensayo teórico que propone un modelo útil al análisis de redes de políticas públicas en Brasil. El modelo innova al incorporar al análisis de redes los aspectos del federalismo y del presidencialismo de coalición, por considerarlos aspectos cruciales de las relaciones de poder en este país. La literatura sobre redes apunta que factores estructurales y de funcionamiento de las redes, así como factores del contexto en el cual las redes están insertadas, son relevantes para comprender su grado de efectividad. Sin embargo, las relaciones de poder subyacentes muchas veces son descuidadas por los análisis realizados en el ámbito de la Administración Pública. En el modelo propuesto, el federalismo y el presidencialismo de coalición se incluyeron a las características de contexto, con el objetivo de agregar elementos de poder a los análisis de redes de políticas públicas en Brasil.

Palabras clave:
Redes de políticas públicas; Federalismo; Presidencialismo de coalición

INTRODUCTION

Since the beginning of the 20th century, public managers and researchers have shown growing concern regarding the need for public policies to be more coordinated and efficient in their designs and forms of implementation (MARCH and OLSON, 1983; PRESSMAN and WILDAVSKY, 1984). According to Jennings (1994), the concept of coordination has become one of the most discussed subjects in the literature on organizations and public policy, due to the increasingly complex nature of public sector issues (FERLIE, FITZGERALD, MCGIVERN et al., 2011; O’TOOLE, 1997c; PETERS, 2002; REPETTO, 2009).

For Jennings (1994), coordination is the joint use of resources and processes to obtain specific results. As for Licha and Molina (2006), coordination is a process of multiple dimensions, involving different actors to articulate efforts that favor the achievement of strategic objectives.

It is assumed that a higher level of coordination may contribute to avoiding inconsistencies and overlaps between public policies (PETERS, 2002). Also, coordinated actions reduce the chances of gaps occurring in the provision of public services (PROVAN and MILWARD, 1995).

The traditional way to achieve coordinated action is the Weberian hierarchy, which promotes intra-organizational coordination through clear lines of authority (JENNINGS and EWALT, 1998; LICHA and MOLINA, 2006; O’TOOLE, 1997c; PETERS, 2002). However, the public administration is often organized based on the joint action between several agencies, that deal with complex problems together. The context where they operate is marked by dispersed power, intersections between mandates, and in which the traditional hierarchy has little value (CLINE, 2000; O’TOOLE, 1997c; PETERS, 2002; PROVAN and MILWARD, 1995).

In addition, in several public policies, it is necessary to analyze the relationships established between governmental agencies and private actors and representatives of civil society involved in the provision of public services. The literature calls such interactions as ‘public policies networks.’ According to O’Toole (1997c), networks are structures of interdependence involving multiple organizations, in which the various members are not subordinated hierarchically to the others.

Networks do not replace the role of traditional bureaucratic structures (AGRANOFF and MCGUIRE, 2001; O’TOOLE, 1997b). They add new relationships to hierarchical structures, increasing the complexity of policy management and implementation, and creating a structure of incentives and priorities entirely different from that observed in intra-organizational management (O’TOOLE, 2015).

Networks are the framework in which many policies are implemented. Thus, network analysis is an attempt to “analyze the relationship between the context and process in policy making” (KLIJN, 1997, p. 14). Thus, the context in which public policy networks operate may offer incentives for both cooperation and isolated action (LINDBLOM, 1979; JENNINGS, 1994; CLINE, 2000; REPETTO, 2009).

It is not easy to obtain commitment with shared goals, considering that the various organizations have multiple interests (which is true even for the public organizations involved in the public policies processes) (LINDBLOM, 1979; MARCH and OLSON, 1983; OLIVER, 1990; REPETTO, 2009; JENNINGS and KRANE, 1994; PETERS, 2002; CLINE, 2000). Thus, the analysis of the prevalent power relations in the context of a network is essential to allow an understanding of the way it works (O’TOOLE and MEIER, 2004).

Peters (1998) argues that the reason for the lack of coordination may lie in the different preferences of the various political parties, especially when there is a need for coalitions or partisan adjustments. Considering Lindblom’s notion that (1979, p. 522), partisan mutual adjustments “takes the form of fragmented or greatly decentralized political decision making” in which the various participants lobby each other, based on their political resources and influence.

Thus, it is possible to say that two factors are crucial to analyze public policies networks in the Brazilian context: coalitional presidentialism and federalism.

In federalist systems, subnational governments are responsible for implementing their policies as well as the policies established by the central government. Therefore, depending on the distribution of competencies and attributions between the levels of government - which is provided by the legal-institutional framework of the policy -, the public action will require a good vertical coordination.

As for the coalitional presidentialism observed in Brazil (ABRANCHES, 1988), the different parties align with the objectives of obtaining power and guaranteeing governability. This logic shapes the entire structure of political power of the Brazilian government, whether it is the legislative-executive relationship, the federative relations or the relations between the various governmental agencies.

Against this backdrop, this study proposes a model incorporating the elements of coalitional presidentialism and federalism to the analysis of public policy networks. The model is an initial attempt, subject to improvements, and it is a good methodological option for the analysis of Brazilian public policy networks.

It should be noted that public policies take different forms according to the situation (CAIRNEY and WEIBLE, 2017; PETERS, 2002). Therefore, for the analysis of each specific network, it is essential to consider not only the different characteristics of the public policy involved but also the political systems in which they operate. Also, it is crucial to dedicate attention to the distribution of competencies and attributions between levels of government, as well as to the degree of participation of private actors such as companies and civil society organizations in the provision of public services.

This article discusses the network analysis in public administration and the characteristics of federalism and coalitional presidentialism. The next sections present the network effectiveness models used as a reference for this study. After these first sections, we propose a model that synthesizes the ones presented and adds the elements “federalism” and “coalitional presidentialism,” which characterize the political power relations in Brazil.

Network analysis in public administration

Berry, Brower, Choi et al. (2004) identified three different research traditions focusing on networks: the social network analysis, originated in sociology; the analysis of political changes and agenda setting, originated in political science; and the analysis of the effects of the networks’ context, structure, and management on the results of the public policies, originated in public administration.

In the area of public administration, the management of public policies through networks is seen as an attempt to solve the problem of coordination between multiple actors (O’TOOLE, 1997a).

Also, networks are arrangements that operate through bargaining processes, reflecting the different levels of power and influence of the participants (CAMPBEL, 2004; GREENWOOD, OLIVER, SAHLIN et al., 2008; INGOLD, FISCHER and CAIRNEY, 2016). As a result, networks privilege certain groups and harm others, while affecting the future distribution of power and resources of society (GREENWOOD, OLIVER, SAHLIN et al., 2008). Thus, asymmetries of power between networks’ members are crucial to understanding their operation, since there is an intricate link between networks and the results of political processes and delivery of public services (REPETTO, 2009).

However, public administration researchers often neglect political aspects and power games when analyzing networks, which leads the field to a certain “blindness” regarding the distributional consequences of networks (O’TOOLE and MEIER, 2004, p. 683). For O’Toole and Meier (2004), such blindness should be avoided since there is a tendency of exacerbating political dynamics in network environments.

In line with this point of view, the proposed model included aspects of federalism and coalitional presidentialism, which are related to the struggles of political power in the Brazilian context.

Federalism and coalitional presidentialism

This section presents the federalism and the coalitional presidentialism in Brazil. We discuss federalism from its historical constitution in Brazil, observing the behavior of the federal and subnational governments in the system, looking particularly at the barriers toward cooperation. The discussion concludes that federalism in Brazil is something between compartmentalized and cooperative federalism. Finally, this section approaches the coalitional presidentialism, discussing its characteristics and implications as influential elements of public policy networks.

According to Abrucio, Franzese and Sano (2013), federalism is a form of territorial organization in which there are rules for the relationship among the parts of a nation. Arretche (2005) compares federations with marble cakes, where complementarity and interdependence are central features.

In a federation, the autonomy of subnational governments is based on a relationship of complicity and legitimacy regarding decision-making processes. There is more than one government acting to design public policies, and this implies negotiated relations, regardless of the degree of asymmetry between them. When operating based on the agreed rules, the parts of the federation are represented at the central government through the legislative, judiciary, or forums formed by the executive branch. The federation’s primary goal is to reconcile the principle of autonomy with the relation of interdependence between its parts. Nevertheless, the main motivation for establishing a federation is to deal with the diversity within a nation, such as the ethnical differences, and political and economic inequalities. All this can generate problems for collective action that need to be overcome (ABRUCIO, FRANZESE and SANO, 2013).

For Abrucio, Franzese and Sano (2013), coordination among the governments of the federation will be successful when there is a mixture of competitive and cooperative practices. Competitive practices are the ones related to the free participation of the subnational entities in the decision-making process, bargaining and establishing mutual control. As for cooperative practices, they are related to partnerships and integrated arrangements regarding territory and public policy. However, coordination is jeopardized by the interference of subnational governments in national policies to gain autonomy and recognition of their particularities. The need to balance the tasks between the decentralization and the nationalization, and to reconcile managerial efficiency and effectiveness with the accountability of public agents also brings difficulties to the coordination among the governments of the federation.

For Arretche (2005), the governments’ autonomy is related to the authority on tax and fiscal resources. Also, the concentration of political authority varies among policies and depends on existing legal and political incentives (ARRETCHE, 2004; CAVALCANTE, 2011). According to Varanda (2007), an initiative to coordinate efforts leads to a reduction of individual autonomy, and if there are no compensations, few will be willing to cooperate. In this sense, trust and reciprocity are values ​​that facilitate cooperation.

It is necessary to reflect on the origin of the problems of collective action in Brazilian federalism. Abrucio, Franzese and Sano (2013) argue about the importance of what they call constitutive heterogeneities historically observed in Brazil. Among these heterogeneities are the enormous size and geographical diversity of the territory, and the political and social differences between the different regions. The second stood out, implying in the regionalisms characterized by diverse traditions and elites with different forms of reproduction and projects gain power.

According to Arretche (2005), Brazil is marked by cycles of tax centralization and decentralization, according to changes in political regimes. In the period known as Republica Velha [old Republic], it was observed a tax decentralization, whereas in the period of Estado Novo [new state], a centralization. In the democratic period (from 1946 to 1964), there was another moment of decentralization, followed by a centralization led by the military regime after 1964. Finally, the 1998 Brazilian Federal Constitution was the vehicle for a radical tax decentralization. However, the change of the political regime does not entirely explain such variations. Arretche (2005) argues that the national constituent assemblies, as well as other periods of intense legislative activity, formed decision-making arenas that led to changes in the status quo. However, as the Brazilian tax system has evolved, the costs of changing its basic structure (installation of the apparatus to collect taxes and the political costs involved in suppressing revenue sources) have affected legislators’ calculations and strategies.

From the 1988 Brazilian Constitution, the principles of democratization, decentralization, and universalization of public policies gain momentum within intergovernmental relations, and the municipalities gain the full status of federative entities. This fact represented another constitutive heterogeneity of Brazilian federalism, promoting decentralization. It changed the flow of resources and allowed local governments to innovate in public management. On the other hand, it stimulated the foundation of new municipalities and the fragmentation of interests. The new context of autonomy was absent from incentives for coordination and cooperation (ABRUCIO, FRANZESE and SANO, 2013).

These authors describe what they call ‘compartmentalized federalism,’ where a logic predominates based more on the separation and demarcation of competences between the levels of government than on their interweaving. This phenomenon reinforced inequality among municipalities, states, and regions. Nevertheless, the central government has kept essential powers, such as the centralization of the power to issue norms, and the strength derived from the bureaucratic processes, and its bargaining power (ABRUCIO, FRANZESE and SANO, 2013).

Grin and Abrucio (2018) note that municipalities’ low state capacities (administrative, technical, political, and institutional) represent a significant obstacle. It is possible to say that municipalities took on new responsibilities without having the necessary competencies.

For Grin and Abrucio (2018, p. 89), Brazilian federalism is “a political construction to address the relationship between all levels of government around common agendas.” However, they note that the actions of subnational entities are highly interdependent but poorly coordinated. Therefore, it becomes necessary to strengthen state capacities, especially at the municipal level, to promote a less unequal administrative decentralization. According to Arretche (2004), the federal government concentrates authority since it has the role of the main financier, regulator, and coordinator of intergovernmental relations. Thus, it should be responsible for inducing the strengthening of state capacities of subnational authorities, especially local ones (municipalities).

It is important to say that there is no consensus in the literature about the centrality of Brazilian federalism. There is a line of thought that believes that state interests stand out due to the influence of the governors over the federal deputies from their region. Another current says that the power division tends to favor the central government, which has strong legislative power, great tax capacity, and several exclusive powers. Also, there is another point of view advocating that the division favors the municipalities, by the expansion of their competences, their autonomy, and the resources transfer received. Finally, another current says that the division of power in the Brazilian federation is fragmented, with several nodes of power endowed with unequal capacities, and which recurrently compete with each other (SOUZA, 2005; ARRETCHE, 2004; BORGES, 2011; CAVALCANTE, 2011).

In addition to these aspects, the relationship in the federation is associated with the legislative-executive relationship at the federal level (ARRETCHE, 2004). Thus, to understand both the federalism and coalitional presidentialism, they must be analyzed together. Brazil has an institutional model, characterized by federalism and by a presidentialism supported by party alliances, which Abranches (1988) called coalitional presidentialism.

In a seminal article comparing the characteristics of political systems in Brazil and developed countries, Abranches (1988) concluded that Brazil shared the main institutional characteristics of developed democracies, such as the proportional representation regime, and the level of party fragmentation. However, Brazil was the only country that combined the models of bicameralism, proportional representation, multi-parties, and that simultaneously governed by forming broad coalitions. For the author, in unequal societies, governability and institutional stability depend on the formation of broad alliances, which imply a high capacity for negotiation (ABRANCHES, 1988). The need for negotiation often results in the deconstruction of the content of public policies, with relevant and not always positive redistributive impacts for the country (BITTENCOURT, 2012; NICOLETTI, 2014).

According to Rocha Neto and Borges (2016, p. 440 and 471), the Brazilian coalitional presidentialism promotes “fragmentation and competition between government agencies,” which means that “transversality [has] little practical effectiveness in Brazil, prevailing only in theory.” For the authors, each agency operates based on the interests of the various political parties that influence the different ministries (ROCHA NETO and BORGES, 2016). According to Nicoletti (2014, p. 7), forming coalitions that guarantee the majority in Congress implies in bargains involving the appointment of ministers and positions in the upper and middle echelon, besides granting leadership and influence in the main committees established in the legislative houses. Often, these bargains embrace a broad ideological spectrum, which leads to the challenge of being able to govern by conciliating very diverse interests and achieving a minimum of political coherence.

In addition, another layer of complexity has yet to be added to this scenario. Abranches (1988) points out that the formation of coalitions in Brazil is oriented both by the partisan and the regional (state) perspectives, which explains the fact that the country has frequently formed broader coalitions than it would be necessary to obtain majorities in Congress, also serving the need to gain support from governors.

On the other hand, although it is necessary to consider the centrifugal effects of the 1998 Constitution (among them increasing the magnitude of the districts, increasing the number of political parties, strengthening the Senate and governors), it is possible to see an expansion of the federal executive’s control in the process of agenda setting, as well as a greater control of the president and party leaders over the legislative process. Thus, there is some centralization of the decision-making process, which affects the policies’ outcomes (FIGUEIREDO and LIMONGI, 2006). For the authors, the agenda-setting mechanisms may increase the president’s ability to pursue cooperation, reinforcing cohesion in the governing coalition and overcoming dissent from ideological differences and disagreements over public policy. This centralized decision-making process gives the executive the “power to propose,” and the Internal Congressional rules offer party leaders significant control over the law-making process, facilitating the effectiveness of the governing coalition’s joint action, neutralizing individual initiatives, and inhibiting partisan adjustments (FIGUEIREDO and LIMONGI, 2006).

Notwithstanding, Bittencourt (2012) argues that bargaining and maintaining the governing coalition may be costly. Keeping the system’s stability may force the executive to rely on a set of tools such as the distribution of political favors and positions, concessions in the content of proposed policies, and even the adoption of corruption practices, such as the ones of the “Mensalao” vote-buying scandal, which occurred in Brazil in 2005.

Thus, the ‘political game’ has a significant effect on the functioning of Brazilian public policy networks. It is necessary to obtain support from the various levels of government and coalition members related to a policy so that everyone is convinced of its importance and has the political will to invest resources in it (REPETTO, 2009). The government needs to work, therefore, on the institutionalization of relations among coalition members, adopting operational tools that enable the coordination process (ABRANCHES, 1988; REPETTO, 2009; ROCHA NETO and BORGES, 2016).

Considering the multidisciplinarity of social issues that reach the government agenda, the inclusion of different agents from the three levels of government in the formulation and implementation of public policies is natural. This phenomenon implies a greater diversity of interests and conceptions about the policies’ scope and structuring. In Brazilian history, the institutional context has been marked by federalism that sometimes favors cooperation and sometimes stimulates compartmentalization, making fluid the balance between autonomy and commitment of the actors. From this perspective, analyzing public policies using the lens of the literature on networks seems a coherent choice, considering the elements found in the theoretical framework.

Models of effectivity of public policy networks

Networks can be analyzed either as dependent variables, insofar as the conditions of their context determine their configuration or as independent variables that impact the effectiveness of public action (O’TOOLE, 2015). This study presents a model to allow the analysis of networks as independent variables, highlighting which of their aspects can have a significant impact on the achievement of public policy results, without disregarding the effects of the context on the networks’ operation and structure (networks as dependent variables).

Based on five network analysis models, selected for their relevance and comprehensiveness, a new model adapted to the Brazilian public policy networks was developed. The models strongly relate to each other. They are presented below, highlighting their main characteristics and particularities.

Provan and Milward (1995) created the first model, selected because it is a mark in the study on network effectiveness. Also, it is considered as the best starting point for new theories on the subject (RAAB, MANNAK, and CAMBRÉ, 2013; TURRINI, CRISTOFOLI, FROSINI et al., 2010). The second model was Mitchell and Shortell’s (2000). It represented a significant advance by adding to the previous model aspects of the networks’ governance and management, as relevant elements to understand their operation and outcomes. The model by Provan and Kenis (2007) is the third model observed, showing criteria that allow categorizing the different types of network governance. The fourth model is developed by Turrini, Cristofoli, Frosini et al. (2010), who present a broad model developed from an extensive and well-documented literature review, using as a starting point the model by Provan and Milward (1995). Finally, the last model selected as a reference for this work was Raab, Mannak and Cambré’s (2013), who state that their model tests and expands the seminal model of Provan and Milward (1995).

The models were compared and synthesized to support the development of a model adapted for the Brazilian case, presented in this article.

Provan and Milward’s (1995 ) model

For Provan and Milward (1995), public policy networks are formed by organizations connected for the provision of services to a target audience, even if these organizations are not particularly specialized or utterly dedicated to the provision of such services. The study that originated this model was the first to systematically examine which factors play a decisive role in the effectiveness of public policy networks.

Provan and Milward (1995) empirically verified the extent to which network effectiveness can be explained by structural and context factors, particularly the network centralization and integration, external control, system stability, and stability in obtaining resources for the network. The authors proposed a preliminary network effectiveness model and argued that networks are effective under structural conditions of centralized management and direct and non-fragmented external control. Also, effectiveness is favored when the system in which the network is inserted is stable, and there is resource munificence.

Environment stability is considered a necessary condition, but not sufficient for a network to be effective. In cases of significant turbulence, network effectiveness tends to improve only when the resulting changes have been fully absorbed, and a new level of stability has been constructed (PROVAN and MILWARD, 1995; O’TOOLE, 1997a).

The model by Provan and Milward (1995), despite being classified by the authors as a preliminary model, is still considered robust and relevant.

Mitchell and Shortell’s (2000 ) model

Mitchell and Shortell (2000) proposed a model for assessing governance and the management of public policy networks. For the authors, governance actions involve defining priorities, strategic goals, network members, and accountability measures, as well as obtaining the necessary resources for the network. According to this understanding, governance is related to the convergence between the network’s members and the complexity of the problems it addresses. In this sense, complex problems demand networks with more heterogeneous members, whereas homogeneous networks can manage simpler problems.

Management activities involve maintaining the commitment of the various participants in the common mission, implementing appropriate coordination mechanisms, promoting constructive conflicts and managing destructive conflicts. It also involves implementing information and monitoring systems, and making changes to leadership or network composition, whenever necessary (MITCHELL and SHORTELL, 2000). For the authors, management activities have to do with the convergence between the number of services provided by the network, its decision-making processes, and its coordination mechanisms. In addition, the actors consider a third aspect in their model, which they call centrality. For Mitchell and Shortell (2000), centrality is the importance of the network in the power structure of its environment, before other networks and organizations, and is related to the sustainability of the network over time.

Provan and Kenis’ (2007 ) model

The model by Provan and Kenis (2007) establishes categories for different types of network governance based on two criteria: whether a specific organization manages the network; and whether this organization is included among the network’s members. The authors argue that networks with different configurations and contexts demand different governance. Thus, they point out three forms of governance: shared governance; governance by a leading organization; and governance through a Network Administrative Organization (NAO). Provan and Lemaire (2012) clarify that these are ideal types of governance, and hybrid types may exist in the real world.

Networks governed by their members, or ‘shared governance,’ are the simplest and most common networks. In these cases, there are no organizations with an established role regarding governance. The governance activities may be carried out either formally, through regular meetings, or informally, through “ongoing, but typically uncoordinated efforts” (Provan and Kenis, 2007, p. 234). In such cases, the commitment of members is the only guarantee that the network will operate properly. Power relations in these networks are to some extent symmetrical, as there is no organization with the prerogative to give the final word.

The second form of governance described in the model is through a leading organization. It happens when one of the members of the network plays the role of governance, usually because of its central position in the flow of customers or resources. The leading organization may be chosen by members, appointed by an external stakeholder, or provided by regulation. This is a centralized type of network, marked by asymmetric power relations (PROVAN and KENIS, 2007).

Finally, the model describes a form of network governance through a Network Administrative Organization (NAO). The difference between governance by a leading organization and governance by a NAO is that the second is not a network member, i.e., it does not provide end-to-end services since it is created for the sole purpose of governing the network. Because it does not have to focus on operational aspects, the NAO can be in a better condition to control and monitor the network, and also to coordinate the various agencies for common results and services (JENNINGS, 1994; RAAB, MANNAK and CAMBRÉ, 2013).

Regarding the relationships within the networks, central actors must be cautious when exercising power, because too much direct imposition can be counterproductive and undermine the members’ commitment to the network (PETERS, 2002; RHODES, 1997; PROVAN and KENIS, 2007). This applies both to governance through the leading organization and through a NAO (PROVAN and KENIS, 2007). For the authors, there is no form of governance necessarily superior to another. On the contrary, aspects such as the density and distribution of trust relationships; the number of participants; the level of consensus on the network objectives; and the need for managerial skills at the network level need to be analyzed to define the best form of governance for each case (PROVAN and KENIS, 2007).

The model by Turrini, Cristofoli, Frosini et al. (2010)

Turrini, Cristofoli, Frosini et al. (2010), carried out an extensive literature review and elaborated a new model that points to three broad groups of factors that affect the effectiveness of public policy networks: their structural characteristics; the network’s functioning characteristics; and the characteristics of the context in which it is inserted. This model is very comprehensive and highlights the following elements as structural characteristics of a network: instruments of external control and accountability; integration mechanisms and tools; the size and composition; the degree of formalization; and the level of stability of the network’s internal relationships.

For the authors, the network operates based on the capacity of its managers to mobilize partners, create commitment, cope with instabilities, promote the coordination of internal processes with related goals, cultivate relationships and establish networking standards, as well as using traditional management strategies, inside and outside organizations.

Also, the model by Turrini, Cristofoli, Frosini et al. (2010) highlights the characteristics related to the context of the network. These characteristics are the network’s environment external stability, the resources munificence, and the support that the network finds in society. These components impact the level of uncertainty that the network must face, facilitating or making difficult the operation of its tasks.

The model by Raab, Mannak and Cambré (2013)

Raab, Mannak and Cambré (2013) define networks as systems composed of a structure, forms of governance, strategies, and goals, people and management processes. They argue that the network’s effectiveness depends on the alignment among these components, to fit the context in which the network is inserted.

The study by Raab, Mannak and Cambré (2013) evaluated how the network’s time of existence, structure (regarding the density of the members’ relations and the centrality of the decision-making processes), the context (especially issues of resource availability and system stability), and the form of governance mode - in line with Provan and Kenis (2007) -, affect the effectiveness of public policy networks.

According to Raab, Mannak and Cambré (2013), the study of networks must be carried out through a configurational methodological approach, in which the various explanatory factors are not examined in isolation or by their additive effects, but are identified as conditions that collectively lead to specific objectives.

The authors found robust evidence that effectiveness is impaired by 1) short time of existence; 2) lack of centralized integration; and 3) lack of resources (when there is no NAO - according to the authors, the lack of resources can be offset by the existence of a neutral and effective organization playing a central role in governance).

Proposal of a model of public policy networks’ effectiveness for the Brazilian context

In the process of developing a model adapted to the Brazilian context, we decided to organize the various elements of the five models studied before in the categories “structural characteristics,” “functioning characteristics,” and “contextual characteristics,” as proposed by Turrini, Cristofoli, Frosini et al. (2010). This procedure facilitated the categorization of all relevant elements.

The comparison among the models (Box 1) allows observing the contributions and limitations of each one before proposing the new model, which combines the most relevant aspects from the previous contributions, given the particularities of the Brazilian case.

Box 1
First definition of the model, based on the literature

As shown in the last line of Box 1, the variables selected for the proposed model are originated in categories that were considered significant for the effectiveness of public policy networks by at least one of the reference models. The nomenclature of the variables in the proposed model may be slightly different in comparison to the terms used in the original models but maintain the same meaning.

In addition to the elements obtained from the models used as a reference, two new elements of context were added to the proposed model: federalism and coalitional presidentialism. Adding these elements, the expectation is that the proposed model becomes a valid methodological alternative to support the analysis of Brazilian public policy networks. Figure 1 schematically represents the developed model, showing its elements and their relations.

Figure 1
Model of analysis of public policies networks proposed to the Brazilian context

In comparison to the models adopted as a reference, the proposed model emphasizes the characteristics of the context. When incorporating the political aspects - federalism and coalitional presidentialism - in the model, a series of factors emerge and may hinder the coordination of public policy networks. These factors are, among others, the search for autonomy, power asymmetries, low interest of local government, legislators, and ministers, centralization of political authority, state’s low capacities (particularly in subnational governments), divergent interests within the governing coalition, partisan and ideological fragmentation, competition among governmental agencies, and low level of identification between the constituents and the political parties.

On the other hand, the same political aspects are recognized as important factors that can strengthen the networks’ coordination, for example: the federal executive branch’s high influence in agenda setting, the control of the legislative process by party leaders, centralization of the decision-making process, the federal government’s high state’s capacities, the institutional status of the central government as financer and regulator, and the territorial power of association.

In the proposed model, the characteristics of the context directly influence the effectiveness of public policy networks as well as their structural and functioning characteristics, as observed in dynamics represented by the arrows in Figure 1. On the other hand, the networks’ structure and functioning elements can also protect them from unfavorable contexts.

In addition, we emphasize that the elements obtained from the international literature and incorporated in the model were considered coherent with the particularities of the Brazilian case.

CONCLUSION

Public policy networks are arrangements characterized by relational aspects and are increasingly frequent in the Brazilian public administration. After all, social issues raised on the government’s agenda are increasingly perceived as complex problems involving different actors, interests, and beliefs. The issue of childhood obesity, for example, involves health, sports, education, and leisure. Therefore, it seems natural that public policies should be formulated and implemented in a cross-cutting way, structured in the form of a network. Thus, the public administration needs to balance the principles of autonomy and cooperation, which is a great challenge.

Therefore, the management of networked policies entails specific conditions and capabilities in order to achieve the intended results. The governance and management activities in network environments are not limited to those that are necessary for intra-organizational environments. Similarly, one should not evaluate the effectiveness of network performance in the same way that the effectiveness of intra-organizational action is evaluated.

Thus, models to analyze the operation of public policy networks are relevant and need to be elaborated, tested and improved in the light of concrete cases. This article seeks to contribute to this process by proposing a theoretical model that emphasizes the importance of the characteristics of the political context, specifically the Brazilian one. The next step for this research is to test the model empirically, based on the proposed variables and their relationships.

Future research may carry out instrumental case-studies to test the model and to collect evidence on the need to consider other aspects that can improve the proposal. These case-studies would necessarily be cross-cutting policies or programs involving different actors, at different levels of government, as well as actors from outside the government, such as is observed in the case of the policy on family farming in Brazil.

Future research may also focus on causal analyzes to verify relationships between the networks’ characteristics and the results of public policies. This focus would undoubtedly contribute to understand how Brazilian public policy networks have operated and their impact on the results delivered.

Also, there are interesting questions to explore, which can expand or refine the scope of the model proposed in this research:

• Which elements related to the behavior of non-state actors may interfere in the effectiveness of public policy networks?

• To what extent can the performance of bureaucracy (high, medium, and street level) influence the structure and operation of these networks?

• Apart from the political elements considered here, are there others that should be considered?

And, finally:

• Recognizing the particularities of the different policies in the various areas (such as health, sanitation, education, and security), which variables of the proposed model that would be more (or less) influential?

• Are there specific areas where managing by networks would be most effective?

These questions point to interesting and worthwhile research paths.

• ABRANCHES, S. H. H. Presidencialismo de coalizão: o dilema institucional brasileiro. Dados, v. 31, n. 1, p. 5-34, 1988.
• ABRUCIO, F. L.; FRANZESE, C.; SANO, H. Trajetória recente da cooperação e coordenação no federalismo brasileiro: avanços e desafios. In: CARDOSO JR., J. C.; BERCOVICI, G. (Ed.). República, democracia e desenvolvimento: contribuições ao Estado brasileiro contemporâneo. Brasília, DF: Instituto de Pesquisa Econômica Aplicada, 2013. v. 10, p. 129-164.
• AGRANOFF, R.; MCGUIRE, M. Big questions in public network management research. Journal of Public Administration Research and Theory, v. 11, n. 3, p. 295-326, 2001.
• ARRETCHE, M. Federalismo e políticas sociais no Brasil: problemas de coordenação e autonomia. São Paulo em Perspectiva, v. 18, n. 2, p. 17-26, 2004.
• ARRETCHE, M. Quem taxa e quem gasta: a barganha federativa na federação brasileira. Revista de Sociologia e Política, v. 24, p. 69-85, 2005.
• BERRY, F. S. et al. Three traditions of network research: what the public management research agenda can learn from other research communities. Public Administration Review, v. 64, n. 5, p. 539-552, 2004.
• BITTENCOURT, F. M. R. Relações Executivo-Legislativo no presidencialismo de coalizão: um quadro de referência para estudos de orçamento e controle. Brasília, DF: Núcleo de Estudos e Pesquisas do Senado, 2012. (Textos Para Discussão n. 112).
• BORGES, A. The political consequences of center-led redistribution in Brazilian federalism: the fall of subnational party machines. Latin American Research Review, v. 46, n. 3, p. 21-45, 2011.
• CAIRNEY, P.; WEIBLE, C. M. The new policy sciences: combining the cognitive science of choice, multiple theories of context, and basic and applied analysis. Policy Sciences, v. 50, n. 4, p. 619-627, 2017.
• CAMPBEL, J. L. Institutional change and globalization. Princeton, NJ: Princeton University Press, 2004.
• CAVALCANTE, P. Descentralização de políticas públicas sob a ótica neoinstitucional: uma revisão de literatura. Revista de Administração Pública, Rio de Janeiro, v. 45, n. 6, p. 1781-1804, 2011.
• CLINE, K. D. Defining the implementation problem: organizational management versus cooperation. Journal of Public Administration Research and Theory, v. 10, n. 3, p. 551-571, 2000.
• FERLIE, E. et al. Public policy networks and “wicked problems”: a nascent solution?Public Administration, v. 89, n. 2, p. 307-324, 2011.
• FIGUEIREDO, A. C.; LIMONGI, F. Poder de agenda na democracia brasileira: desempenho do governo no presidencialismo pluripartidário. In: SOARES G.; RENNÓ, L. (Ed.). Reforma política: lições da história recente. Rio de Janeiro: Ed. FGV, 2006. p. 249-280.
• GREENWOOD, R. et al. The SAGE handbook of organizational institutionalism. Thousand Oaks, CA: SAGE, 2008.
• GRIN, E. J.; ABRUCIO, F. Quando nem todas as rotas de cooperação intergovernamental levam ao mesmo caminho: arranjos federativos no Brasil para promover capacidades estatais municipais. Revista do Serviço Público, v. 69, p. 85-122, 2018. Edição especial.
• INGOLD, K.; FISCHER, M.; CAIRNEY, P. Drivers for policy agreement in nascent subsystems: an application of the advocacy coalition framework to fracking policy in Switzerland and the UK. Policy Studies Journal, v. 00, n. 00, p. 1-22, 2016.
• JENNINGS, E. T. Building bridges in the intergovernmental arena: coordinating employment and training programs in the American States. Public Administration Review, v. 54, n. 1, p. 52-60, 1994.
• JENNINGS, E. T.; EWALT, J. A. G. Interorganization coordination, administrative consolidation, and policy performance. Public Administration Review, v. 58, n. 5, p. 417-428, 1998.
• JENNINGS, E. T.; KRANE, D. Coordination and welfare reform: the quest for the philosopher’s stone. Public Administration Review, v. 54, n. 4, p. 341-348, 1994.
• KLIJN, E.-H. Policy networks: an overview. In: WALTER, J. M. et al. (Ed.). Managing complex networks: strategies for the public sector. London: SAGE, 1997. p. 14-34.
• LICHA, I.; MOLINA, C. G. Coordinación de la política social: criterios para avanzar. Washington, DC: Instituto Interamericano para el Desarrollo Económico y Social, 2006. (Serie de Documentos de Trabajo I-58).
• LINDBLOM, C. Studies still muddling, not yet through. Public Administration Review, v. 39, n. 6, p. 517-526, 1979.
• MARCH, J. G.; OLSON, J. P. Organizing political life: what administrative reorganization tells us about government. The American Political Science Review, v. 77, n. 2, p. 281-296, 1983.
• MITCHELL, S. M.; SHORTELL, S. M. The governance and management of effective community health partnerships: a typology for research, policy, and practice. The Milbank Quarterly, v. 78, n. 2, p. 241-289, 2000.
• NICOLETTI, A. S. O presidencialismo brasileiro e sua relação com o Poder Legislativo: discutindo o governo de coalizão. In: VSEMINÁRIO NACIONAL SOCIOLOGIA E POLÍTICA, 2014, Curitiba. Anais...Curitiba: Ed. UFPR, 2014.
• OLIVER, C. Determinants of interorganizational relationships: integration and future directions. The Academy of Management Review, v. 15, n. 2, p. 241-265, 1990.
• O’TOOLE, L. J. Implementing public innovations in network settings. Administration and Society, v. 29, n. 2, p. 115-138, 1997a.
• O’TOOLE, L. J. The Implications for democracy in a networked bureaucratic world. Journal of Public Administration Research and Theory, v. 7, n. 3, p. 443-459, 1997b.
• O’TOOLE, L. Treating networks seriously: practical and research-based agendas in public administration. Public Administration Review, v. 57, n. 1, p. 45-52, 1997c.
• O’TOOLE, L. J. J. Networks and networking: the public administrative agendas. Public Administration Review, v. 75, n. 3, p. 361-371, 2015.
• O’TOOLE, L. J.; MEIER, K. J. Desperately seeking Selznick: cooptation and the dark side of public management in networks. Public Administration Review, v. 64, n. 6, p. 681-693, 2004.
• PETERS, B. G. Managing horizontal government: the politics of coordination. 2002. Disponível em: <Disponível em: https://doi.org/10.1111/1467-9299.00102 >. Acesso em: 22 fev. 2019.
  » https://doi.org/10.1111/1467-9299.00102
• PRESSMAN, J. L.; WILDAVSKY, A. Implementation: how great expectations in Washington are dashed in Oakland. 3. ed. Berkeley/Los Angeles: University of California Press, 1984.
• PROVAN, K. G.; KENIS, P. Modes of network governance: structure, management, and effectiveness. Journal of Public Administration Research and Theory, v. 18, n. 2, p. 229-252, 2007.
• PROVAN, K. G.; LEMAIRE, R. H. Core concepts and key ideas for understanding public sector organizational networks: using research to inform scholarship and practice. Public Administration Review, v. 72, n. 5, p. 638-648, 2012.
• PROVAN, K. G.; MILWARD, H. B. A preliminary theory of interorganizational network effectiveness: a comparative study of four community mental health systems. Administrative Science Quarterly, n. 40, p. 1-33, 1995.
• RAAB, J.; MANNAK, R. S.; CAMBRÉ, B. Combining structure, governance, and context: a configurational approach to network effectiveness. Journal of Public Administration Research and Theory, v. 25, n. 2, p. 479-511, 2013.
• RHODES, R. From marketisation to diplomacy: it’s the mix that matters. Australian Journal of Public Administration, v. 56, n.2, p. 40-53, 1997.
• REPETTO, F. Coordinación y gestión pública integral en América Latina: notas conceptuales. In: REPETTO, F. et al. (Ed.). Reflexões para Ibero-America: planejamento estratégico. Brasília, DF: Enap, 2009. p. 11-20.
• ROCHA NETO, J. M.; BORGES, D. F. Políticas públicas coordenadas e presidencialismo de coalizão: o caso do Promeso. Cadernos EBAPE.BR, Rio de Janeiro, v. 14, p. 440-472, 2016. Edição especial.
• SOUZA, C. Federalismo, desenho constitucional e instituições federativas no Brasil pós-1988. Revista de Sociologia e Política, v. 24, p. 105-121, 2005.
• TURRINI, A. et al. Networking literature about determinants of network effectiveness. Public Administration, v. 88, n. 2, p. 528-550, 2010.
• VARANDA, M. P. Acção colectiva entre pequenos empresários: uma análise de redes sociais. Análise Social, v. 42, n. 182, p. 207-230, 2007.

• [Translated version] Note: All quotes in English translated by this article’s translator.

Publication Dates

• Publication in this collection
  17 Apr 2020
• Date of issue
  Jan-Mar 2020

History

• Received
  20 Sept 2018
• Accepted
  31 Jan 2019