HTML

ePub

PDF

INTRODUCTION

The idea of state capacity involves the concept of effectiveness, i.e., translating the ability of governments to mobilize multiple resources to achieve organizational goals (Christensen & Gazley, 2008; Gomide and Pires 2016). In this sense, a key question in the agenda is: What state capacities are needed for governments to accomplish their organizational missions and policy proposals? To answer it is not an easy task, especially in a complex context of public governance, which may refer to the redefinition and expansion of the forms of relationship between State and society or between government, private agents, and society. Moreover, the concept of state capacity has evolved to encompass multiple dimensions (Gomide et al., 2017), such as the quality of bureaucracy (Souza, 2016, 2017), structures of governmental coordination (Gomide and Pires, 2016), and networks between actors located inside and outside public administration (Evans, 1993; Grisa, Kato, Flexor, & Zimmermann, 2017).

Although recent studies have pointed out that state capacity is crucial to explain performance in public governance (Centeno et al., 2017; Serikbayeva et al., 2021; Gomide et al., 2021), other issues also emerge. First, which state capacity dimensions or factors do governments mobilize and combine to affect organizational goals? Is the organizational performance among policy sectors in the public administration equally perceived? Do these dimensions or characteristics of state capacity affect policy performance in the same way? To explore these questions, this paper’s main goal is to investigate the effects of key state capacity dimensions on governmental performance and whether they differ among policy sectors. More specifically, we map which dimensions of state capacity are critical for successful governmental outputs. This analysis applies to multiple policy areas of the Brazilian public administration.

The inquiry assumes that policy sectors (government core; infrastructure; productive development; security/citizenship and social/environmental) present different state capacities and, therefore, produce heterogeneous effects on outputs and outcomes. The policy sectors’ classification stems from the official aggregation in the 2016-2019 Pluriannual Plan, established by federal law 13.249 of January 14th of 2016. This assumption is based on previous research that investigated the process of State building in Brazil and the quality of its bureaucracy (Geddes, 1994; Evans, 1993). Historically, public administration in Brazil was organized in the context of the development agenda (which took place from 1930 to 1980). It was based on the idea of “pockets of efficiency”, in other words, semi-autonomous agencies and state enterprises with a high level of professionalization and competence (Nunes, 1997; Schneider, 1991; Sikkink, 1991; McDonnell and Vilaça, 2021). Hence, the Brazilian bureaucracy was built in a very asymmetric pattern. The agencies responsible for the development agenda concentrated state capacity (meaning professionalization and political autonomy) while the majority of the public administration was primarily ineffective and unprofessional. However, in recent decades, an expansion of State capacity happened throughout the Brazilian public administration, although some asymmetries persist, especially in the social area (Souza, 2017).

To investigate this subject, the inquiry uses a survey applied to over three thousand civil servants of the Brazilian federal public administration that covered, in addition to performance questions, the following state and policy capacity dimensions: meritocracy, autonomy, relationship, skills, resources, and accountability (Pereira, Machado, Cavalcante, Gomide, Bersch, Magalhães, Goellner & Pires, 2019). Initially, we employed principal components analysis for creating composite variables, based on the bureaucrats’ perceptions. Then, the paper displays descriptive statistics to explore these synthetic indexes in five different policy areas of government (government core; infrastructure; productive development; security/citizenship, and social/environmental). Lastly, we run a multivariate regression to test the effect of state capacities on policy performance.

In sum, the paper presents interesting findings with theoretical and empirical contributions for this literature, especially addressing a developing country such as Brazil. First, the degree of state capacity development impacts the perception of performance, and, secondly, these effects are quite diverse on the organization’s outputs and outcomes. Although they vary, another unexpected result is that the state capacity dimensions are not so different among the policy sectors, resulting from a more homogenous strategy of strengthening the public administration in Brazil. Regarding the impact of state capacity on performance, the relationship seems indifferent. The standard dimensions of state capacity – resources and meritocracy- present variable effects, while autonomy, accountability, and skills stubbornly affect better performance in public organizations.

The outline of the paper is as follows. The following section discusses the theoretical debates regarding bureaucracy, governance, state capacity, dimensions, and applications in the Brazilian case. The third section explores the descriptive statistics of the performance, and the state capacity dimensions. In the fourth, we detail the multivariate model and discuss the empirical results. Finally, some inquiry conclusions and future research agenda are presented.

Theoretical Debate on State Capacities and Hypotheses Construction

The concept of state capacity is usually related to the idea of performance, meaning the mobilization of resources necessary to achieve governmental effectiveness and to implement official goals (Christensen & Gazley, 2008; Skoc­pol, 1985). Even though the literature on state capacity became quite popular in social sciences and public administration fields over the recent years, its theoretical origin dates to the “statist” movement of the 70s and 80s (Cingolani, Thomsson & Crombrugghe, 2015). At that time, in contrast to society-centered theories (such as Marxism and Pluralism), it was claimed that the state held autonomy to influence political and social processes through their policies. Several studies about economic structural transformations, such as late industrialization, investigated the state’s role in that process; in other words, as the protagonist at explaining social and political outcomes (Evans & Rauch, 1999; Skocpol & Finegold, 1982). Since then, the debates about state capacity have experienced movements of conceptual expansion and new thematic applications (Cingolani, 2013; Fukuyama, 2013; Gomide, Pereira & Machado, 2017; Wu, Ramesh & Howlett, 2015).

Therefore, since its origins, the concept of state capacity has been mobilized to understand state performance. In this sense, while the state capacity agenda from the 1980s and 1990s was mainly concerned with understanding state performance in economic development processes, recently, the concept has been used to explain state performance in several fields – such as public security (Hendrix, 2010); social development (Cingolani et al.2015); environmental sustainability (Abers, Oliveira & Pereira, 2016); levels of corruption (Bersch, Taylor & Praça, 2017); the implementation of infrastructure policies (Gomide and Pereira, 2018); and state responses to Covid-19 crises (Mao, 2021; Serikbayeva et al., 2021).

Within the statist movement, whereas the state was considered a weighty actor that held autonomy, researchers investigated a new question: what conditions strengthen state capacity? In other words, what factors increase state effectiveness? These questions produced several studies that elaborated specific dimensions of state capacity. Following the Weberian tradition, the first answers understood state capacity as intrinsically linked to public bureaucracies' quality (Cingolani et al., 2015). This perspective led to the administrative approach of state capacity, which is understood as the set of capabilities of state agencies to implement their policies and to produce coordinated actions oriented to the production of results (Gomide & Pires, 2014).

The dimensions of the administrative state capacity were based on the features of the Weberian bureaucracy model, such as professionalization, meritocracy, and autonomy from social influences. Professionalization refers to “intellectual activities of civil administrators engaged in diagnosis societal problems and framing policy alternatives to deal with them” (Skocpol, 1985, p. 11). In other words, it means a form of collective puzzlement on society´s behalf that entails both deciding and knowing. Meritocracy is related to a specific model of recruitment based on impersonal requisites; usually knowledge level proved upon exams and certificates. It is argued that merit-based selection of public employees positively affects bureaucratic output through the increased expertise of administrators, their greater cohesion, and commitment to their organization's goals (Nistotskaya & Cingolani, 2016).

Broadly, autonomy is understood as the formulation of policies by state bureaucracies that are not simply reflective of the demands of social groups or classes (Skocpol, 1985). According to this perspective, administrators can formulate long-term goals transcending partial and short-sighted demands from specific groups (Skocpol, 1985). The lack of state autonomy, on the other hand, is associated with the transformation of public bureaucracies into an arena of bitterly politicized and inconclusive conflicts (Rueschmeyer & Evans, 1985; Skocpol & Finegold, 1982). Whereas state bureaucracies should be independent of social forces, the Weberian approach claims they should be subordinated to political principals, who hold the legitimacy of formulating the political agenda (Fukuyama, 2013). Some concerns highlight that excessive autonomy can test political accountability and primarily negatively affect policy decisions. In this sense, he argues that the relationship between autonomy and government quality is not always linear. Sometimes, it may appear shaped as an inverted U curve. Put differently, a low level of autonomy may provoke slow and inefficient policymaking, while excessive autonomy can allow public servants to bypass the elected officials’ controls. The ideal relationship between bureaucratic autonomy and government quality would happen when agencies provide enough independence for innovation and oversight for accountability (Fukuyama, 2013).

The first applications of this state capacity approach focused on elaborating indexes of bureaucratic quality based on the features of the Weberian model. The classic example of this perspective is Evans and Rauch (1999) analysis of the effects of the Weberian state structures on economic growth, which compared the “weberianess scale” among 35 developing countries for the 1970-1990 period. Even though these studies contributed to the development of scales of quality of bureaucracy, they considered state capacity as homogenous inside the countries.

Recently, a new academic manifesto calls for a more in-depth analysis of states' executive branches and their bureaucracies. In this sense, Fukuyama (2013) claims that although several studies about political institutions limit power, few advances exist in analyzing the institutions that accumulate and use power. Therefore, the past years are marked by a new proliferation of studies that mobilize state capacity concepts to explain state action, culminating in new dimensions of state capabilities. In this sense, recent studies claim that state capacity is the product of the combination between competencies and resources. Ramesh and Howlett (2016) describe three types of competencies – analytical, managerial, and political – that interact with three levels of resources – individual, organizational and systemic. Individual resources refer to the existence of a structure of technical knowledge; organizational resources mean the tools for informational, financial and human resources management; systemic resources refer to conditions located outside the state, such as the levels of social and political thrust on the public bureaucracies (Wu et al., 2015). It is worth mentioning that these authors employ the term policy capacity when addressing to resource and skills dimensions of public bureaucracy and organizations.

Another state capacity approach has shed light on the relationship pattern between state bureaucracies and social actors. For them, it’s also essential to analyze civil society since social groups often contest the policies formulated by state bureaucracies, which might hinder state capacity due to judicialization and interruptions in the implementation process (Abers et al., 2017; Midgal, 2001). Thus, the legitimacy of state action is considered an important factor for state capacity (Mann, 2008). Besides that, the relationship between state bureaucracies and social groups leads to the exchange of information necessary to formulate coherent policies (Evans, 2010; 2011). Also, some studies have emphasized the advantages of social control and accountability to correct policies scope and impacts (Pires & Gomide, 2016). In this context, the initial focus on administrative capacities is complemented by concerns about political capabilities, which are understood as the abilities of state actors to negotiate and process conflicts (Gomide & Pires, 2014). More specifically, the political capacity approach involves understanding the needs and positions of different stakeholders, communication skills, and effective civil service bargain (Wu et al., 2015).

In the Brazilian case, debates about state capacity were first mobilized to study the asymmetric results of the developmentalist agenda from the 1930s to the 1970s. The main inspiration to apply state capacity concepts in the Brazilian case was the Weberian approach. It was believed that the quality of state bureaucracy – meaning especially professionalization and meritocracy – was essential for the developmentalist agenda success. However, in the case of Brazil, the state was not able to modernize the whole public administrative at the same time (Streek & Thelen, 2005). One reason for that is a consequence of the “politician’s dilemma” described by Geddes (1994, p. 281) as a situation in which “the presidents, as well as his coalition partners, faced a wrenching conflict between their own need for immediate political survival and longer-run collective interests in economic performance and regime stability”. While the first need is associated with political support and weakening of state capacity since professionalization and meritocracy are set apart from the requisites to reformulate bureaucracy in a context where appointments in the public administration are used as an exchange for political support, the second is related to the existence of effective bureaucratic organizations.

During the developmentalist Era in Brazil, politicians chose a dual strategy to solve this dilemma: state bureaucracies responsible for the implementation of the economic projects were modernized and professionalized, which led to the creation of “island of excellence” or “pockets of efficiency”; nonpriority bureaucracies for the economic agenda were used for patronage to reach political support (Schneider, 1991; Sikkink, 1991). In this sense, priorities technical bureaucracies were insulated from political influences and pressures from social groups (Nunes, 1997). Therefore, the initial state capacity literature on the Brazilian context shed light on the Weberian approach and emphasized the heterogeneous organization of bureaucratic capabilities.

In the 1980s, Brazil experienced deep social and political transformations due to the democratization process that culminated in the approval of a new Constitution in 1988. The new institutional framework of the Brazilian state is characterized by territorial decentralization since municipalities became important actors in the supply of social services; the creation of participatory institutions, which led to the inclusion of civil society actors in the decision-making process of public policies; and the strengthened of horizontal accountabilities agencies and mechanisms (Cavalcante, Lotta & Oliveira, 2018; Pires & Gomide, 2016). Besides that, the democratization process also led to the commitment of the Brazilian state to new roles, such as the supply of universal social policies and infrastructure projects (Cavalcante, Lotta and Oliveira, 2018). In the face of these transformations, new research emerged to understand if the Brazilian state could deliver the new policies required within the democratic scenario and investigate which capabilities were necessary for the Brazilian state to act effectively in this new complex context.

Parts of the answers to these questions were still based on the Weberian perspective, focusing on the administrative approach. In this sense, Marenco, Strohschoen, and Joner (2017) mobilized a restricted definition of state capacity as bureaucratic professionalization to investigate urban property tax collection variations. Souza (2017) also resorted to the Weberian approach to examine the process of state capacity building in Brazil between 1995 and 2010, referencing the variables of professionalization and meritocracy of bureaucracy. She claims that the asymmetric capacity of the Brazilian public administration remained until the 1990s since only the agencies responsible for the priority agenda were fully professionalized.

This historical tendency was interrupted during the Working Party administration (2003-2016) when several public contests and the consequent increase of public employees with an undergraduate level of education (Cavalcante & Carvalho, 2017). Souza concludes that today the Brazilian public administration holds the main features of the Weberian bureaucracy. The study of Cavalcante, Lotta and Oliveira (2018) reinforces this conclusion: according to them, from 2003 to 2014, there was an increase in the number of public employees from 480.000 to 615.000 in diversified areas – including, for example, the areas of infrastructure, social policy and regulation. Significant increases followed this in the public service salaries. In this sense, the current research agenda on state capacities surpasses the Weberian perspective by investigating other factors impacting state performance. For instance, Gomide and Pereira (2018) face puzzling data: even though the Brazilian infrastructure bureaucracy is highly professionalized, its performance is very poor in delivering effective policies. To understand that, it was necessary to go beyond the classical Weberian approach by investigating intragovernmental coordination and the relationship pattern between State and society. Similarly, Satyro, Cunha and Campos (2016) present an intriguing conclusion by claiming that the municipalities with Weberian features deliver less social assistance services in Brazil. In contrast, the municipalities that do not have a Weberian bureaucracy can supply these kinds of services. These kinds of conclusions raise the question of what other approaches are necessary to understand state capacity in the case of Brazil.

In this context, Gomide and Pires (2016) claim that, besides the administrative capacity, policy capability is essential to understand the Brazilian performance in implementing public policies in a democratic context. More specifically, while the administrative capacity is associated with implementing the policy goals, policy capability is responsible for innovation and improvements in governmental initiatives. Similarly, when studying the contemporary actions of the Brazilian state in rural development area, Grisa et al. (2017) state that the democratic capacities, i.e., the formal and informal structures of interactions between the state, the market and social organizations – provide legitimacy, and also make it possible to adapt policies to the demands of local stakeholders.

Finally, scholars have investigated if the distribution of state capacity remains heterogeneous in the Brazilian public administration. The main conclusion in this respect is that even though the recent professionalization of the Brazilian bureaucracy overcame the idea of “pockets of efficiency”, it remains some asymmetry. In this sense, Souza (2017) claims that the fields of social policies and infrastructure are less professionalized than the areas responsible for control and accountability. This asymmetry in terms of Weberian features exist even within the same field of public policies: for instance, when studying the infrastructure sector in contemporary Brazil, Cavalcante, Pereira and Gomide (2017) and Gomide and Pereira (2018) conclude that there is some heterogeneity when comparing the transport and energy sectors. Cavalcante et al. (2015) also compare the profile of different Brazilian bureaucracies to verify that infrastructure is the sector with the highest percentage of postgraduate and permanent career servants occupying commissioned positions. In this same group, also known as mid-level bureaucrats, Cavalcante and Lotta (2015) and Cavalcante, Lotta and Kasai (2018a; 2018b) demonstrate that heterogeneity is also the rule among them, which is perceived not only in their profile and background but also in their policy sector performance.

When investigating the Brazilian state capacity to implement large infrastructure projects in the Amazon region, Abers et al. (2016) conclude that the infrastructure agencies concentrate administrative capacity. Still, the environmental bureaucracy presents high levels of political capacity. The uneven patterns of bureaucratic capabilities were also highlighted by the studies of how the core of government is structured and operates in Brazil. In this sector, responsible for conducting the presidential strategic agenda, the units usually are highly professionalized and empowered.

The combination of this literature about state capacities dimensions depicts the idea that these dimensions are distributed asymmetrically along with the public administration agencies. Besides, these dimensions are responsible for different outputs (for instance, administrative capacity is related to policy delivery, political capacity is associated with innovations), which leads to the idea that each state agency, depending on the concentrated capacity dimension, produces different outcomes.

Based on this discussion involving the correlation between state and policy capacity and performance as well as the historical asymmetric pattern in the government, we formulated two hypotheses as follows:

H.: The higher the degree of state capacity, the higher the perception of performance.

H.: The dimensions of state capacity produce heterogeneous effects on outputs and outcomes, according to the policy sector.

In short, we argue that the dimensions of state capacity have experienced an intense enlargement since the “statist” movement. Nowadays, the administrative capacity encompasses the classical features of the Weberian bureaucracy model and specific competencies and resources held by individuals and organizations, drawing from the newer literature of governance and policy capacity (Ramesh & Howlett, 2016; Wu et al., 2015). Besides that, the emphasis on the bureaucratic autonomy from social forces has lost ground to the political dimension, which claims for a close relationship between state actors and social groups and demands. In the case of Brazil, this enlargement has been mobilized to understand the contemporary Brazilian state capacity. The peculiarity in the Brazilian case is the focus on the asymmetry related to state capacity. The following section presents data regarding these state capacities dimensions and advances in the analyses of their relation to performance, in the different sectors of the Brazilian federal government.

STATE CAPACITY AND PERFORMANCE

Methodology

The database used for the paper’s analysis stems from the survey government quality and state capacity that is part of joint research called Governance Project, between the Institute of Applied Economic Research (Ipea), the Center on Democracy, Development, and the Rule of Law (CDDRL) of the University of Stanford with other international research centers.

The survey target population was composed of civil servants (permanent employees and those with commissioned positions) who formulate and implement public policies in the ministries, executive agencies (autarchies/foundations) and regulatory agencies. Due to the global feature of the joint research project and, consequently, the need for similarities among the countries' databases, the Brazilian survey sample excluded positions and careers of the street-level bureaucracy from the state-owned enterprises and mixed-capital companies and the military. The layers that are part of this research were delimited considering three criteria: positions and careers, which relate to the type of bond established between the civil servant and the federal public administration; having or not commissioned post; and the kind of organization (ministry, executive agency or regulatory agency).

According to these layers, the sampling frame was assembled based on the available data collected between May 15 and July 17, 2018, totaling 3,226 respondents, almost 70% of the sample required. In order to expand sample results to the population, the sample weights of each layer were calculated, then, we got the total of 263,468 servants, used as a basis for the selection of the sample. Its confidence interval was 95%, which means that the estimates contained in the survey are statistically reliable for the set of selected respondents. The survey also covered sociodemographic and professional characteristics, such as gender, race, year of birth, level of schooling, length of service in the public sector, among others included in the survey report (Pereira et al., 2019).

Based on the survey questionnaire and responses and theoretical grounded in the literature discussed in the previous section, we formulated synthetic performance indicators and state capacities dimensions, such as meritocracy, autonomy, relationship, skills, resources, and accountability. It is worth mentioning that the proper fit of the question in each of these dimensions is not an easy task because their complexity level reflects interrelated and overlapping concepts. To cope with this issue, we followed the pattern used by other studies in the Governance Project (Boittin, Distelhorst & Fukuyama, 2016).

The research employs principal component analysis (PCA) to formulate these variables (the specific breakdown of how the indexes were formed is detailed in the appendix). Generally, composite indexes aim to summarize complex and multidimensional subjects helping to interpret, classify and rank units of analysis in a particular case. After running the principal component analysis, the index scores ranged from 0 to 100; the more significant the index, the higher degree of each dimension.

In order to demonstrate the variety of the public servants’ perceptions about their organizations’ capacities and performance, the paper employs descriptive statistics to explore these synthetic indexes and compare their means by policy sectors, also using analysis of variance (ANOVA) to test whether the differences between groups of data are statistically significant. It is worth mentioning that the sectors’ means must be compared in each dimension since they were based on different questions and response scales. Lastly, we run a multivariate regression to analyze the effect of state capacities on policy performance.

State Capacity Dimensions

The following figures illustrate the distribution of these indicators and reinforce the assumption of heterogeneity of state capacity among the bureaucracy. To begin with, it seems that in all dimensions, the sectors’ indexes vary from each other with different intensity and their ranks oscillate as well. The variances, in general, are relatively slight than expected, which is proved by the Anova tests that showed statistical significance in three of six among the dimension’s averages.

Figure 1 encompass the means of meritocracy (Anova Test: F = 3.53, p-value = .007) and autonomy (Anova Test: F = 2.16, p-value = .07) indexes. The graphs show differences among the sectors in both cases, and the tests confirmed their statistical variance. However, in the first, they are higher. Another aspect that draws attention is the variation of the positions among the dimensions as, theoretically, it is expected that meritocracy and autonomy would be quite convergent as new Weberian features of bureaucracy. In this sense, the result is convergent with the Fukuyama (2013) argument that not always extensive autonomy is positive; in fact, it is reasonable to expect that it varies according to the particularities of different areas and careers in the public sector. For instance, security/citizenship has the greatest average in meritocracy and the lowest in autonomy. Most of them are from military and police forces, well-known for their hierarchical promotion process between ranks. Still, at the same time, their decisions and actions tend to be very procedural and regulated. In this sense, one reasonable explanation for the low autonomy of military and police forces stems from the risks of high-level independence and discretion that may harm government quality and even democracy.

Regarding resources and skills (Figure 2), the indexes’ means vary among the policy sectors, especially in the resources indexes. The differences are the only one statistically significant (F = 9.71, p-value = 0.000). In either case, once again, ministries of CoG show the highest score average as observed in autonomy. Unsurprisingly, productive development and infrastructure are in second and third places, depending on the dimension. These findings are quite in line with the literature (Cavalcante & Lotta, 2015) that demonstrates that most of the ministries and agencies from these sectors have professionalized careers and resourceful programs, such as the finance and planning ministries and general attorney’s office that are part of the core of government. Security and citizenship, once again, drop to the lowest rank, followed by social and environmental ministries in the skills and resources indexes.

The last state capacity described involves the complexity of the bureaucrat’s networking and the level their organizations are held accountable by society. In either case, the differences among policy sectors are quite reduced than the previous dimensions, and, in both cases, the Anova tests are not significant. The CoG also stands out and confirms the expectation since this sector has crucial functions of communication and coordination of the Executive branch, which naturally demands such capacities (Cavalcante & Gomide, 2018). Moreover, security/citizenship ministries continue presenting weaker capacities, under the averages. In contrast, social and environment are among the best in relationship and accountability, making sense considering their policymaking features, the first more restrictive. The second is based on institutions that foster open and participative processes. In this dimension, the results indicate that the analysis of the public organization’s connections seems to be more productive, focusing on the relation’s types than on its overall intensity.

Performance by sectors

Usually, every public organization aims to improve its performance in delivery services and reach its goals with efficiency and quality. Performance can be understood as the mobilization of resources needed to achieve governmental effectiveness and implement strategic objectives (Christensen & Gazley, 2008). However, as we all know, it can be a broad and ambiguous definition in the public sector that, despite its simple, common-sense concept, involves complex and not trivial theoretical and methodological difficulties. Policy and public organizations’ performance depend mainly on various state capacity dimensions and, above all, on the effectiveness of their actions, meaning the actual benefits to the targeted population, which can be expensive and difficult to precisely measure. It becomes even more complicated when we analyze policy sectors, considering they may involve several ministries or agencies with dozens of programs that often do not have suitable performance indicators or management systems. Other assessment strategies focus on public opinion, citizens or elites, and civil servants' perception.

Based on this last option, for this analysis, we formulated a composite index based on four questions that encompass a broader approach of this concept, including questions regarding the organization’s production of expected results and if the unit has achieved more or fewer outputs compared to five years ago. Moreover, the performance score also covers the bureaucrat’s perception of whether the organization is well evaluated by society and if it is creative and innovative.

The index formulation followed the same procedure employed to the state capacity scores discussed above. The average of the indexes grouped by policy sectors is different, with an overall mean of 55. However, the standard deviations are relatively high, revolving around 21%, i.e., almost forty percent of the index mean. Figure 4 shows the indexes distribution in the box plot graphs.

The standard deviation in performance is higher than the other synthetic indexes, but, mainly, the mean and median (shown in the box plots) are the greatest among the policy sectors. After employing an analysis of variance (ANOVA) test, the results confirmed that these differences are also statistically significant (F = 12.82, p-value = 0.000). Surprisingly, the best scores are found in the security and citizenship ministries, even though they presented the worse indexes in the majority of the capacity dimensions. So, in this sector, the bureaucrats have, on average, a negative view of their capabilities but a positive perception of how they perform. With the best scores in the previous analysis, the core of government shows the second-highest means and medians of their employees’ view of performance in the Executive branch. Conversely, the productive development units present the worse perspectives about achieving their goals, improving effectiveness over the last five years, and being creative and innovative.

Performance Determinants

To analyze if the bureaucrats’ perceptions of state capacity affect their views about organization performance, in this subsection, we empirically test this possible correlation using multivariate models for all survey data and, specifically, by each policy sector. On the model’s left side, the dependent variable is the synthetic index of performance, above described, while the right side is composed of state capacity indexes. Therefore, the basic statistic model is defined as follows:

Performance. . β_{0 +} β. Meritocracy_{i +} β. Autonomy_{i +} β. Resources_{i +} β.Skills_{i +} β. Relationship_{i +} β. Accountability_{i +} u. (1)

The models’ results from Ordinary Least Square regression (OLS) using cross-sectional data are interesting in different ways. Table 1 presents the estimated coefficients, standard errors in parentheses, and the models’ coefficients of determination for all the sample (general model) and each policy sector analyzed.

Table 1: The Performance Determinants

Importantly, due also to the large sample, T-test and F-test are valid asymptotically. Even though a few variables are not statistically significant, overall, the significance of the regressions is confirmed (Wooldridge, 2006). After the regression, a check for multicollinearity was carried out, and the results proved that the degree of collinearity among the independent variables is not worrisome.

To begin with, the coefficients of determination (R.) in all models are relatively expressive, considering that the independent variables together explain from 40% to 50% of the performance in federal government organizations. Secondly, it is also noticeable that most independent variables affect the performance index, however, with different patterns and intensities.

Overall, the regression results allow us to confirm both hypotheses of this paper. In the general model and in every policy sector, as most capacity dimensions positively affect the bureaucrat’s perception of performance (H1). Moreover, while most of the variables present a significant and substantial impact on the dependent variable in all models, the other coefficients estimated or affect only some policy areas’ performance, primarily, relationship. Based on this empirical evidence, we can confirm hypothesis 2, state capacity dimensions produce heterogeneous effects on the organization’s results, according to the policy sector.

The first dimension is meritocracy, meaning an administrative environment that values recruitment and promotion based on impersonal requisites of skills, technical competence, and expertise, in contrast to political party and personal relations criteria. In the general model and almost all policy areas, except for infrastructure ministries, the estimates confirm the assumption that merit-based organizations have a positive effect on their performance (Evans & Rauch, 1999; Rueschmeyer & Skocpol, 1985), although in these cases, the coefficients were substantially the lowest ones in the models. Some recent findings from the Brazilian literature may help us understand these minor effects on the performance perception of the bureaucrats.

A reasonable explanation may be that most public service selection and promotion (i.e., occupation of commissioned post) are highly regulated and mostly restricted to permanent career servants. Regarding the appointment, since the Federal Constitution of 1988, all permanent staff had to be approved in transparent, open, and nondiscriminatory public tenders, which reinforces the merit as the basis. Nonetheless, on the promotion side, the legislation sets many limits to appointing outsiders (professionals that are not part of the public sector careers), leaving approximately 10% of the positions in the federal government for them. Notwithstanding, Cavalcante e Carvalho (2017) have already shown that, since 1995, the majority of this percentage is also occupied by a permanent civil servant. Moreover, studies on the profile and performance of mid-level bureaucrats in Brazil (Cavalcante et al., 2018a; Cavalcante & Lotta, 2015) have proven that promotion in the civil service is a consequence of meritocratic mechanisms since the professional background and formal education have a positive relation with commissioned position appointments.

The literature has also diminished the political partisan’s relevance in these cases; in other words, the bulk of the posts (low and mid-level bureaucracy) are much less affected by partisanization than the top officials (D’Araujo & Petek, 2018; Lopez, 2015), which does not mean that other forms of politicization would not be frequent inside government. However, nonpartisan political networks tend to be more challenging to measure (Praca & Lopez, 2019).

Continuing in the Weberian approach, the next dimension is autonomy, which means the relative independence of the bureaucrats from social and political groups to decide their way of work and make decisions grounded in technical considerations and with a certain level of discretion. This state capacity would allow the separation of policy implementation from instabilities stemming from competition from the political system, making the management environment more predictable and policies more resilient (Lewis, 2003; Miller, 2000). As a result, it could avoid the process of capture in the State, which would negatively affect national projects and the long-term policies listed in a rational (Beazer, 2012; Cingolani, 2013).

In this case, regression models show that the autonomy indexes are more influential on an organization’s performance than meritocracy, including every policy area. In the whole sample, the coefficients are statistically significant. On average, a change in the autonomy index would affect .16 in the performance scores, ceteris paribus. In this sense, we can state that the higher the bureaucrat’s sense of autonomy, the better their perception of performance, which is allied with the theoretical assumptions (Cingolani et al., 2015; Evans & Rauch, 1999; Fukuyama, 2013)

The third dimension of state capacity assumes that the employment of resources achieves performance to reach governmental effectiveness and implement official goals (Christensen & Gazley, 2008; Ramesh & Howlett; 2016; Skocpol, 1985; Wu et al., 2015). The synthetic index encompasses a variety of factors that depict a range of management factors that may hamper the conditions for the organization to achieve its goals, such as human resources, budgeting, technology, auditing, among others. In this case, the estimates are significant in four models, indicating, as expected, the positive correlation between organization resources and performance, keeping other variables constant. The exception is the security and citizenship ministries, which presented the lowest average on resources and didn’t have a statistically significant coefficient in the regression model. It is a finding that deserves further analysis.

Advancing to a more comprehensive approach of administrative capacity, beyond the conventional dimensions discussed above, the regression shows different results concerning relationship, skills, and accountability.

The bureaucrat ability to interact with different stakeholders has become increasingly important in a context of complex governance arrangements (Evans, 2010; 2011; Wu et al., 2015), in which policy implementation demands constant coordination inside the Executive branch, with other branches, private sector, subnational governments, civil society, and international agencies. It is essential to underline that the variable measures frequency of interaction, a proxy for coordination, but doesn’t encompass every aspect of it. Unpredictably, the relationship has the lowest coefficients, and the variable shows statistical significance only in one model (infrastructure), nonetheless, with a negative coefficient. This finding is corroborated in the recent study by Gomide et al. (2021) on the Brazilian federal bureaucracy. What can we draw from it? A reasonable explanation can assume that a public servant's frequency and variety of interactions do not reflect barriers or facilitators to policymaking. The relationship patterns in each policy sector naturally vary according to the sector’s features, independently of the impact on performance.

On the contrary, the last two state capacity dimensions confirm the previous assumption that organizations’ skills and the degree of accountability affect their performance. The first variable contained different aspects of civil servant capabilities, including analytical, interpersonal, and managerial competencies (Wu et al., 2015). At the same time, the accountability index covers factors related to preventing corruption, social participation in the policymaking, and holding the organization accountable for better results. To illustrate some of these effects, Figure 5 presents graphs with predicted values from the general model that depicts all independent variables on performance.

Clearly, both coefficients estimated demonstrate considerable impacts of skills and accountability indexes in the dependent variable. In objective terms, an increase of a point in skills, on average, represents a growth of approximately .32 in the performance score (varying from .26 in productive development and .37 in social and environmental), ceteris paribus. While the effect of accountability is even higher, e.g., in the general model, it positively affects performance indexes in .37, on average. In summary, organizations that are more skillful and held accountable tend to perform better, in line with previous studies (Mann, 2008; Pires & Gomide, 2016; Ramesh & Howlett, 2016).

CONCLUSIONS

An essential subject in the public governance agenda is how governments can achieve their organizational missions and policy goals. So, this paper aimed to investigate the effects of crucial state capacity dimensions on governmental performance and whether they differ among policy sectors.

This analytical approach has become more relevant for three main reasons. First, due to the recognition of the increasing complexity of the public sector, reflected in new agencies and state responsibilities worldwide. Secondly, the field of research has evolved and, more recently, presented a comprehensive perspective on the concept of state capacities and their impacts on governance and government effectiveness. Lastly, due to the historical asymmetry among public organizations and their bureaucracy professionalization, it is worth studying if this still produces different impacts on policy sector performance, considering that decades of democratization may have diminished this heterogeneity. However, the most striking finding is that the difference in the exact state capacity dimensions among policy areas is not as expressive as supposed. This result might be explained by an even evolution of agency structures and bureaucracies strengthening in the past years in Brazil, which have created a more homogenous public administration compared to the historical asymmetry related to state capacity. This finding reinforces recent studies that claim that governmental investment in state capacity became broader and more inclusive as it incorporated several agencies that do not belong to the classical “pocket of efficiency” (Cavalcante & Silva, 2020; Souza, 2017).

Another exciting finding is that the policy sector may be well ranked in some state capacity dimensions but might be poorly ranked in another dimension. It is worth highlighting; however, that core of government units leads almost all index averages, except for meritocracy. Finally, regarding the regression results, just the relationship dimension does not substantially affect the performance’s perception in contrast to the literature that mobilizes the ties between social actors and bureaucrats to explain state performance, such as Evans (1994). In contrast, the traditional dimensions of state capacity – resources and meritocracy, impact performance differently according to the models. On the contrary, the effects of autonomy, accountability, and skills indexes, these latter two encompass a broader governance approach, are strongly associated with better performance in the public administration, including all policy sectors.

In short, the paper shows instigating insights at advancing in the study of state capacity, governance, and public sector performance, especially because it deepens comparative analysis on these critical relations for the public administration that are mostly theoretical and less empirically explored (Wu et al., 2015)Moreover, the research contribution is also original as it highlights the case of policymaking in a developing country, different from the prevailing literature from nations with mature democratic and capitalistic institutions.

Nevertheless, the composite indexes were formulated based on the bureaucrats’ perceptions about the effectiveness of their actions; as such, they obviously must be analyzed cautiously. As with every analytical strategy, the use of survey data has limitations. One of them is that the likelihood of positive perception of state capacity is also influenced by performance. However, it does not harm the scientific validity of this inquiry. On the contrary, recognizing its shortcomings helps to emphasize the need for complementary approaches to complex and dynamic phenomena. In this broader research agenda, the paper results can become even more relevant for the field of study if complemented by comparative cross-nationally analysis or qualitative detailed case studies. For instance, it focused on how policy sectors employ different coordination instruments and their effects on performance in a governance context. The future research paths are diverse and may encompass different perspectives of policymaking. Baekgaard et al. (2018) pointed out a gap in an investigation of how bureaucratic characteristics influence the policy agenda-setting, which could be explored based on the quality of bureaucracy and the dimensions of state capacity mobilized in this paper. Specifically, future research on state capacity should focus on earlier stages of the policy process, in addition to the implementation phase. Another avenue is indicated by Filgueiras et al. (2020), who claims that further investigations incorporating individual factors in state capacity studies are necessary. According to them, capacity analysis is usually “based on macro-structural factors as determinants of results” (p. 21)—however, individuals matter since they may not perform the functions they are assigned to.

References

Abers, R., Oliveira, M., Pereira, A. (2017). Inclusive development and the asymmetric state: big projects and local communities in the Brazilian Amazon. The Journal of Development Studies, v. 53, n. 6, pp. 857-872.

Bækgaard, M., Mortensen, P. B., Seeberg, H. B (2016). The Bureaucracy and the Policy Agenda. Journal of Public Administration Research and Theory 28(2), 239-253.

Beazer, Q. H. (2012). Bureaucratic Discretion, Business Investment, and Uncertainty. The Journal of Politics, 74(3), 637-652.

Bersch, K.; Praca, S., Taylor, M. (2017). State Capacity, Bureaucratic Politicization, and Corruption in the Brazilian State. Governance: an international journal of policy, administration, and institutions, 30(1), 105-124.

Boittin, M.; Distelhorst, G., Fukuyama, F. (2016). Reassessing the quality of government in China. Osgoode Legal Studies Research Paper Series (n. 197). Toronto: Osgoode.

Cavalcante, P., Lotta, G. (Eds.) (2015). Burocracia de Médio Escalão: perfil, trajetória e atuação. (1a ed., v. 1, pp. 13-21). Brasília: Escola Nacional de Administração Pública, Enap.

Cavalcante, P.; Camoes, M., Knopp, M. (2015). Burocracia de médio escalão nos setores governamentais: semelhanças e diferenças. In P. Cavalcante & G. Lotta (Eds.). Burocracia de Médio Escalão: perfil, trajetória e atuação (1a ed., v. 1, pp. 57-90). Brasília: Escola Nacional de Administração Pública, Enap.

Cavalcante, P., Carvalho, P. (2017). Profissionalização da Burocracia Federal Brasileira (1995 - 2014): avanços e dilemas. Revista de Administração Pública, 51, 1 - 26.

Cavalcante, P. L. C., Gomide, A. A. G. (2018). O Presidente e seu núcleo de governo: a coordenação do Poder Executivo (1a ed.). Brasília: Instituto de Pesquisa Econômica Aplicada, Ipea.

Cavalcante, P., Pereira, A., Gomide, A. (2017). Capacidade estatal e burocracia: uma abordagem dos setores de energia e transporte. In J. De Paula, P. Palotti, P. Cavalcante, & P. Alves. Burocracia Federal de Infraestrutura Econômica: Reflexões sobre Capacidades Institucionais, Brasília: Ipea and Enap.

Cavalcante, P., Lotta, G., Kasai, E. (2018a). The mid-level bureaucrats’ performance: determinants of their activities and work relationships. Cadernos Ebape.Br., 16, 14 - 34.

Cavalcante, P., Lotta, G., Kasai, E. (2018b). Exploring mid-level bureaucracy: a tentative typology. Revista Brasileira de Ciência Política, 26, 187-222.

Cavalcante, P. Lotta, G., Oliveira, V. (2018). Do Insulamento Burocrático à Governança Democrática: as transformações institucionais e a burocracia no Brasil. In R. Pires, G. Lotta, & V. Oliveira (Eds). Burocracia e Políticas Públicas no Brasil (cap. 2). Brasília: Ipea, Enap.

Cavalcante, P., Silva, M. S. (Eds.). (2019). Reformas do Estado no Brasil: trajetórias, innovações e desafios (1a ed., v. 1). Brasília: Ipea/Eclac.

Centeno, M. A. et al. (2017). Unpacking states in the developing world: Capacity, performance, and politics. In: States in the developing world: 1-34. UK: Cambridge University Press.

Christensen, R., Gazley, B. (2008). Capacity for public administration: Analysis of meaning and measurement. Public Administration and Development: The International Journal of Management Research and Practice, 28(4), 265-279.

Cingolani, L., Thomsson, K., Crombrugghe, D. (2015). Minding Weber More than ever? The impacts of state capacity and bureaucratic autonomy on development goals. World Development, 72, 191-207.

Cingolani, L. (2013). The state of state capacity: a review of concepts, evidence and measures. Maastricht: Maastricht University, Working Paper.

D’Araújo, C., Petek, J. (2018). Recrutamento e perfil dos dirigentes públicos nas áreas econômicas e sociais (1995-2002), Revista de Administração Pública, 52(5), 840-862.

Evans, P. (1993). O Estado como problema e solução. Nova Lua, 1(28-29), 1-29.

Evans, P. (2010). Constructing the 21st Century Developmental State: Potentialities and Pitfalls. In O. E. Edigheji (ed). Constructing a Democratic Developmental State in South Africa: Potentials and Challenges. Cape Town: Human Sciences Research Council.

Evans, P. (2011). The Capability Enhancing Developmental State: Concepts and National Trajectories. Center for Studies on Inequality and Development, Discussion Paper No. 63 – March.

Evans, P., Rauch, J. E. (1999). Bureaucracy and growth: A cross-national analysis of the effects of" Weberian" state structures on economic growth. American sociological review, 64(5), 748-765.

Filgueiras, F, Koga, N., Viana, R. (2020). State Capacities and Policy Work in Brazilian Civil Service. Rev. Sociol. Polit., 28(74), 23-45.

Fukuyama, F. (2013). What is governance? Governance, 26(3), 347-368.

Geddes, B. (1994). Politician's Dilemma: Building State Capacity in Latin America. Berkeley: University of California Press.

Gomide, A. (2016). Capacidades estatais para políticas públicas em países emergentes: (des)vantagens comparativas do brasil. In A. Gomide, R. Boschi, (org.). Capacidades estatais em países emergentes: o brasil em perspectiva comparada (1a ed., v. 1, pp. 15-47). Rio de Janeiro: Ipea.

Gomide, A., Pereira, A. K, Machado, R. (2017). Capacidade Estatal e a Pesquisa Científica: contribuições do dossiê temático. Sociedade e Cultura, 20(1), 3-12.

Gomide, A., Pereira, A. K. (2018). Capacidades estatais para políticas de infraestrutura no Brasil contemporâneo. Revista de Administração Pública, Rio de Janeiro 52(5), 935-955.

Gomide, A., Machado, R., Albuquerque, P. M. (2021). Capacidade estatal e desempenho na percepção dos burocratas brasileiros: desenvolvimento e validação de um modelo de equações estruturais. Cadernos Ebape, 19(especial), 1-16.

Grisa, C. Kato, K., Flexor, G., Zimmermann, S. (2017). Capacidades Estatais para o Desenvolvimento Rural no Brasil: análise das políticas públicas para a agricultura familiar. Sociedade e Cultura, 20(1), 13-38.

Howlett, M., Ramesh, M. (2016). The Achilles Heel of Governance: Critical Capacity Deficits and Their Role in Governance Failures. Regulation and Governance, 10(4), 301-313.

Hendrix, C. S. (2010). Measuring state capacity: Theoretical and empirical implications for the study of civil conflict. Journal of Peace Research, 47(3), 273-285.

Mann, M. (2008). Infrastructural Power Revisited. Studies in Comparative International Development (SCID) 43(3), 355–65.

Mao, Yexin (2021). Political institutions, state capacity, and crisis management: a comparison of China and South Korea. International Political Science Review, 1-17.

Marenco, A., Strohschoen, M. T., Joner, W. (2017). Capacidade estatal, burocracia e tributação nos municípios brasileiros. Rev. Sociol. Polit., 25(64), 3-21.

Midgal, J. (2001). State in Society: studying how states and societies transform and constitute one another. New York: Cambridge University Press, 2001.

Nistotskaya, M., Cingolani, L. (2016). Bureaucratic structure, regulatory quality, and entrepreneurshio in a comparative perspective: cross-sectional and panel data evidence. Journal of Public Administration Research and Theory, 26(3), 519-534.

Nunes, E. (1997) A gramática política do Brasil: clientelismo e insulamento burocrático (cap. 2). Rio de Janeiro: Jorge Zahar.

Pereira, A. K., Machado, R., Cavalcante, P., Gomide, A., Bersch, K., Magalhães, A, Goellner, I., Pires, R. (2019). Qualidade do Governo e Capacidades Estatais: Resultados do Survey sobre Governança Aplicado no Brasil. (Relatório de Pesquisa), Ipea.

Pires, R., Gomide, A. (2016). Governança e capacidades estatais: uma análise comparativa de programas federais. Revista de Sociologia e Política, 24, 121-143.

Praca, S., & Lopez, F. (2019). Political appointments, political parties and bureaucracy. In B. Ames (ed.) Handbook of Brazilian Politics (pp. 358-370). New York: Routledge.

Rueschemeyer, D., Evans, P. (1985). The state and economic transformation: Toward an analysis of the conditions underlying effective intervention. In P. Evans, D. Rueschemeyer, & T. Skocpol. Bringing the State Back In (pp. 44-77).

Satyro, N., Cunha, E., Campos, J. (2016) Análise espacial da burocracia da assistência social nos municípios brasileiros: notas para uma reflexão sobre a capacidade de implementação dessa política. Opinião Pública, 22(2), 286-317.

Schneider, B. (1991). Politics within the state: elite bureaucrats and industrial policy in authoritarian Brazil. Pittsburgh: University of Pittsburgh Press. Sikkink, K. Ideas and institutions: developmentalism in Brazil and Argentina. Ithaca: Cornell University Press.

Serikbayeva, B. et al. (2021). State capacity in responding to COVID-19. International Journal of Public Administration, 44(11-12), 902-920.

Skocpol, T. (1985). Bringing the State Back In: Strategies of Analysis in Current Research. In P. Evans, D. Rueschemeyer, T. Skocpol. Bringing the State back in (pp. 3-43). Cambridge: Cambridge University Press.

Skocpol, T., Finegold, K. (1982). State capacity and economic intervention in the early New Deal. Political science quarterly, 97(2), 255-278.

Souza, Celina. (2016). Capacidade Burocrática no Brasil e na Argentina: quando a política faz diferença. In A. A. Gomide, R. R. BOSCHI. Capacidades Estatais em Países Emergentes: o Brasil em perspectiva comparada (pp. 51-105). Rio de Janeiro: Ipea.

Souza, C. (2017) Modernização do Estado e construção de capacidade burocrática para a implementação de políticas federalizadas. Revista de Administração Pública, 51(1).

Streek, W., Thelen, K. (2005). Institutional change in advanced political economies. In W. Streek, & K. Thelen (ed.). Beyond Continuity: institutional change in advanced political economies. Nova York: Oxford University Press.

Wu, X., Ramesh, M., Howlett, M. (2015). Policy capacity: A conceptual framework for understanding policy competences and capabilities. Policy and Society, 34(3-4), 165-171.

IR