STREET-LEVEL BUREAUCRACY AND PUBLIC POLICIES: ANALYZING EDUCATIONAL POLICY IMPLEMENTATION FROM THE PERSPECTIVE OF SCHOOLS AND TEACHERS

ABSTRACT This paper analyzes the implementation of educational policies and the roles of school professionals considering the street-level bureaucracy theory (LIPSKY, 1980). This study assumes that educational reforms elect schools as planning and administration centers, making them and the professionals working there responsible for new attributions which are motivated by improved autonomy, in addition to administrative and educational decentralization. In these contexts, which are marked by schools’ increasing empowerment, the discretionary power exercised by its professionals (teachers, principals, coordinators, among others) is a key element to understanding the availability and implementation of programs and their ability to either influence or change the design of educational policies on a local level. This perspective emphasizes the importance of considering those closer to actions deriving from such policies, that is, the actors who see the bottom-up process because they are in the lower end. According to Lipsky (1980), these are the so-called local policy agents or street-level public agents. As for schools, we understand that to enforce a certain policy, factors such as interpersonal influence, commitments, and informal negotiations are as important as formal processes and regulations. Finally, this study attempts to prove the impacts and contours assumed by changes in school dynamics in terms of translating local educational policies.


INTRODUCTION
The purpose of this paper is to analyze school professionals as agents responsible for implementing public policies. According to Ball (2007), educational reforms "elect" schools as planning and administration centers, making them and the professionals working there responsible for several new attributions which are motivated by improved autonomy, in addition to administrative and educational decentralization and democratization of access to education.
In these contexts, which are marked by schools' increasing empowerment, this study seeks to verify the discretionary power exercised by its professionals (teachers, principals, coordinators, among others) in the availability and implementation of programs and their ability to either influence or change the outline of educational policies on a local level.
Literature offers a variety of definitions for public policies, depending on the theoretical matrix used. For Lynn (1980), public policies are a group of governmental actions targeting specific effects, and which can influence the lives of citizens. This definition points to an arena where conflicts around preferences and interests take place, resulting in different political actors defining cooperation/bargaining/capture strategies motivated by maximizing their goals, with governments as the central axis.
The development of this literature led to a series of discussions on public policies. According to Evans (2010) and Hill and Hupe (2014), one of the most relevant discussions addresses the decision flow in the decision-making process. An analytical school of thought believes that the process can be controlled from top to bottom (top-down), that is, decisions are made by authorities who have a certain level of control over the process and decide what and how policies will be implemented.
According to Hill and Hupe (2014), another perspective emphasizes the importance of considering those closest to actions resulting from policies, that is, the actors who see the bottom-up process, as they are at the lower end. According to this perspective, the population affected by the policy, together with State agents in contact with it (referred to as local policy agents or street-level public agents) could and should influence public policy planning more intensely. In his seminal work, Lipsky (1980), states that these actors know exactly what happens and what the best line of work to achieve the results intended by the policy would be, increasing effectiveness and efficiency gains, and promoting democratic administration in programs.
That being said, we can say that there is a large number of street-level agents involved in the availability and implementation process related to these programs (LIPSKY, 1980;WILSON, 1989;ELLIS, 2011). Teachers, education professionals, social workers and health professionals are some of the main actors in the chain. Each one of them carries out several functions which, to a greater or lesser extent, can lead to changes in the course of actions and/or in the outcomes of programs. Therefore, these agents are actors who have some discretionary power, as they are the ones in direct contact with citizens.
In the case of schools, we understand that to put a certain policy into practice, factors such as interpersonal influence, commitments, and informal negotiations are as important as formal processes and regulations. This means that teachers (and other professionals working at these places) decide how to teach, as well as what and when to teach it. They also decide when and how often they will call and control attendance; they decide which students are more or less dedicated and, as for public policies on education, they may individually or jointly decide to support -or not -programs or actions outlined by educational agencies. The effects of policies are uncertain because the subjectivities of those taking part in school life are crossed by the official discourse, depending on how these policies will be interpreted, on the specific context, and the problems faced by each school.
The following sections will reflect on some aspects related to the expanded decision-making capacity of schools and their professionals, who will be analyzed considering their role as street-level implementing agents of public policies. This study attempts to prove the impacts and contours assumed by changes in school dynamics in terms of translating local educational policies.

PUBLIC POLICY IMPLEMENTATION
The study of public policies and the rules that dictate their development, implementation, and assessment appear as a field of political science that seeks answers to solve economic and social issues. This line of research began in the United States as an academic subject that moved away from patterns adopted by traditional European studies in this field, which used to focus more on analyzing the state and its institutions than on actions carried out by governments. Thus, we can say that the cornerstone of European studies on public policies consists of theories that attempt to explain the role of the State and the government in society. On the other hand, in the USA the focus is not on the role of the State, but rather solely on actions carried out by governments, institutions, and individuals.
Throughout the process to consolidate the field, four researchers have made decisive contributions to the development of studies on public sciences, which is why they are considered the founders of this field of study. Laswell (1936), is the first major name in this area of knowledge, as he introduced the expression policy analysis in the study of public policies, and this has, since then, become an essential concept in current studies. The other important names that deserve attention are Simon (1957), Lindblom (1979), andEaston (1965). Simon introduced the concept of bounded rationality for policymakers, arguing that rational knowledge could minimize bounded rationality. In contrast to the rationalist standpoint of Laswell and Lindblom's theories, Lindblom introduced other variables which contributed to the analysis and formulation of public policies, among which are power relationships and the integration between different phases of the decision-making process. Finally, Easton left his mark when he defined public policies as a system that is influenced by macro and microenvironments.
Literature offers several definitions for public policy, of which we highlight those provided by Lynn (1980) and Peters (2001). Lynn (1980), defines public policies as a set of government actions that will produce a particular class of effects. Peters (2001), defines public policies as the sum of government activities, whether the government is acting directly or through agents, which influence the lives of citizens. Thus, we can say that public policies are projects that emerge from the articulation of various groups of interest and which are analyzed once they are put into practice, and, depending on whether this assessment is positive or negative, these projects may lead to changes that will either result in the creation of new public policies or feedback into their design cycle incorporating the new lessons learned.
In other words, when public policies are designed, they seek to translate the purpose of governments by way of programs and actions that lead to positive changes in the real world (HILL, 1997). When these policies are implemented, they are analyzed and may undergo subsequent changes if their course of action so demands.
Several models were created to explain the design and analysis of public policies. Among these models, one that stands out is the policy cycle approach, which sees public policies as a deliberative cycle that divides the policy design process into some distinct phases. Thus, it consists of a series of steps in which distinct functions are attributed and all actions practiced within each phase target the ultimate goal: well-developed public policies. This cycle has six phases: problem identification, agenda setting, policymaking, policy implementation, policy evaluation, and possible correction of the action or feedback.
In the problem identification stage, we can question why some issues are relevant to the policy whereas others are ignored. The purpose of analysts is to define how, among an infinite number of possibilities, some issues are sufficiently relevant to be addressed in the policy while others are left out. Two solutions are proposed. The first one attempts to answer the question by focusing on the collective consciousness. How collective consciousness about the need to face an issue is built? For Souza (2006), the construction of collective consciousness is key in defining problems and can happen through changes -whether these changes take place in the ideological field or in parties in power -which alter the way the process is perceived.
The second answer to this question is to try to focus on when the issues become part of the public agenda. Hill (1997), argues that, from the analytical standpoint, problems only become policy issues from the moment they become relevant action-wise from a political and administrative standpoint. This means that issues only become part of the public agenda with the acknowledgment that something must be done to change them.
The following cycle is the agenda-setting stage. This is when there will finally be a decision whether a theme deemed relevant will truly be pushed on the current political agenda or it will be postponed to a future policy. This decision should not be made without general and specific knowledge on the theme or the current context related to the public policy. All pros and cons of the available options must be analyzed, as well as the possibility and feasibility of the project to succeed in the political arena. If this possibility is confirmed, the project can be carried forward.
The third phase is policymaking, involving the designing of programs and decision-making. In this stage, the means of action to be used in the project are analyzed, and, considering the options available, the one most suitable for that specific policy is chosen.
The policy implementation stage may be considered the most important one for a certain policy to succeed. Because it comes right after policymaking, this stage involves more agents, actors, and information networks. Implementation addresses the execution of actions that allow for the implementation of activities established in the policy-making plan, to reach the goals established in it. This stage has a strong dynamic character because political arenas change over time and this may affect the implementation process and even lead to other policies.
In the evaluation and feedback stages, programs are analyzed based on the effects of implementation and its consequences. Undesired side effects are assessed as well as the reason for such effect if that is the case. Based on that, it is possible to deduce which future action must be taken.
As for implementation, there are three traditional views on the process. In the first model, the process is understood as the execution of activities that allow actions to be implemented targeting the fulfillment of the goals set during the policymaking process. This model does not consider the political cycle itself to be a process. It considers implementation as a "one-round game", in which governmental actions are taken from top to bottom.
The second concept considers policy implementation to be a process and the instabilities and problems identified during implementation are part of the analysis stage. Based on this perspective, these problems either derive from factors related to the agents' ability to determine policies or from political issues related to policy implementation, or, yet, from the resistance faced in specific groups. This is known as the linear perspective, and it assumes that the evaluation and monitoring of policies are means to take corrective actions regarding such policies. Although it is an advance when compared to the classic perspective, the linear perspective is criticized primarily for placing excessive emphasis on public administration as a perfect system that would ensure full and precise implementation of the plan initially proposed during the policymaking stage, dismissing and/or minimizing the impact of variables acting on the process.
For Hill (1997), implementation can be analyzed as a game with successive rounds, in which roles and resources are negotiated and levels of adherence to the political program vary. Unlike the linear theory, this perspective on implementation considers the environment in which political actions take place to be unstable, and full of uncertainties and contradictions. These uncertainties result from the complexity of the social phenomena dealt with in public policies, as well as from limitations related to actual resources and information available. Plans are drawn up in the abstract, yet, in practice, countless unforeseen events take place. Thus, based on this perspective, implementation is perceived as a changeable experiment rather than as something stagnant which is defined on a priori grounds.
This way, programs only partially outline the course of actions and cannot control all steps to be taken. The broad array of possible situations force implementing agents to make decisions and, once such situations cannot always be foreseen, the level of discretion these agents have in the decisionmaking process is very high. When implementation is perceived as a game, discretion becomes essential for a certain policy to succeed or fail. Therefore, implementation is interpreted as an autonomous process in which key decisions are taken, leading up to the creation of new policies. This is why the political cycle cannot be considered to be simple and linear. It must be analyzed as a cycle made up of complex networks involving agents and ourselves, the so-called critics. We, the critics, are at the very point where all unexpected situations come together with the goals set during the policymaking stage. This union creates and harmonizes tensions when decision-makers interferer with the process, leading up to a new direction. These tensions are crucial, as they create learning opportunities, adding new contours to policies. Thus, we can say that the instabilities identified in the implementation of government programs are extremely important as they allow us to understand why governments often fail in their attempt to reach new goals set in the public policy-making plan. This is the point in which implementation theories regarding the political cycle seek to explain why some policies are efficient whereas others are not or, on the other hand, which strategies are used by decisionmakers and how these strategies influence the process.
We can mention two analytical concepts which appeared in the field of Public Policies and helped understand the type of implementation used by each government, based on Hill and Hupe (2014): the top-down and bottom-up models. The question raised by these models is related to the degree of discretionary action implementing agents have. Each model adopts a different perspective on the matter, arguing that successful implementation is influenced by this variable in one of these directions, as well as by the existing connection between the direction chosen, the type of policy, and the environment in which it is introduced.
The top-down model is embedded in the classic and linear perspectives of implementation. It is based on the assumption that this is the stage in which the actions carried out by public agents target the achievement of the goals set during the policy-making stage. The focus of this model is entirely on the policy-making process. Public policy is a series of distinct stages whose concept is made up of political activity, whereas implementation is responsible for administrative activities. Thus, implementation is merely the execution of activities targeting the achievement of goals defined in policymaking. In this respect, the goals set out in the policymaking stage must be strictly adhered to during the subsequent stage, which must bring together all the means necessary to fulfill the plan developed.
According to this concept, there is a clear separation between the policy-making and implementation stages. The assumption is that the policy operates based on a hierarchical flow, from top to bottom. The role of public agents is limited to just meeting the goals set in the policy-making process. In this model, the discretion of street-level agents is not encouraged and the rules must be clear to prevent any individual decision to be made by the agents. Thereby, the top-down perspective does not consider implementation to be a dynamic and transformative process leading up to learning. Policies become static and pre-formed, and unable to adjust to unpredictable situations. This unpredictability occurs for two reasons: first because of the multiplicity and variety of actors that make up the implementation process, and second because the conditions found during the policymaking stage do not remain the same; they undergo changes resulting from various external and internal factors. Therefore, the goals initially set are not always achieved.
The bottom-up model criticizes the previous perspective, which is based on a hierarchical process. In the bottom-up model, the implementation must not be considered separately, but, rather, as part of a continuous process that evolves and changes in each stage. This model proposes that implementation must be considered a game that enables learning and the creation of new policies. Hill (1997), points out that this perspective places minimal importance on goals and plans -what matters are the processes, insofar as results are not predictable.
Contrary to the linear perspective present in the previous concept, implementation is considered to be a decentralized process whose analysis, monitoring, and understanding are based on street-level agents, especially those in the lower end of the process (MEYERS & VORSANGER, 2007). This model sees implementation as a game and proposes that it must be understood as an experimentation process, which may result in new policies. Moreover, the bottom-up theory supports the idea that the analysis and implementation of public policies occur at the concrete level rather than based on preexisting abstract plans. Based on this perspective, it is not possible to have perfect control over the process, and street-level agents do not necessarily follow what had been defined during the policymaking stage. Lipsky (1980), who is considered to be one of the main researchers of the bottom-up theory, analyzed the difficulty to control the behavior of public agents and believes that this is what makes implementation such a dynamic stage. For him, certain unforeseen situations can only be resolved at the moment they occur and this solution occurs by way of discretionary decisions made by agents. Because agents themselves are the ones directly dealing with citizens and because they are the only ones with true knowledge of the situation, they cannot be required to make decisions based on a solution that had been defined in the abstract. As a result, when a given decision is made by agents, who are influenced by various factors, they may end up creating new means of policy implementation or even setting new goals for the program. Therefore, we can say that policies change as they are implemented and that their success or failure also depends on the performance of agents directly involved in this stage.
As seen earlier, the process of implementing public policies should be understood as a set of interactions between implementing agents and the recipients of those policies. This process enables changes and learning in the field of public policies, which happens through the individual values and ideas of each agent, whose actions end up leading to some transformations in the program. The study of policy implementation shows a high level of agreement among authors in the sense that most of them emphasize the performance of implementing agents and how important their performance is for the failure or success of the program.

THE PERFORMANCE OF SCHOOLS: DILEMMAS AND POSSIBILITIES FOR POLICY IMPLEMENTATION
Schools serve an institutional and, consequently, an organizational role. They lay down, or influence, the rules of the game, creating new ones and/or changing those previously established. Therefore, schools are not "merely" and "solely" concerned about the curricular, educational, and social dimensions. In other words, the role of schools goes beyond these purposes. For Ball (2002;2005;2016), schools have a series of attributions that mark their position as materializers of state actions by way of governmental bureaucracy, especially concerning social policies historically designed for the poor or underprivileged population.
Immunization campaigns, itinerant dental practices, programs to fight malnutrition, actions to promote citizen awareness, traffic education programs, campaigns to raise awareness about the risks of drug use, programs to fight poverty, and programs to promote public safety are among the various examples which, for decades, have elected schools as one of the spheres responsible to mediate the relationship between the State and society.
The great number of public policies somehow related to the school environment has led to changes in this space as well as in the way schools deal with their target audience and the actions that take place in this environment. According to Cavaco (1995), public schools are changing. In these times of change and unpredictability, schools are faced with a great array of different forces -local pressure, the influence of international bodies, in addition to official regulation -which makes it ultimately important to consider the relationship school agents establish among themselves and with the institution, in the sense of promoting articulation and developing spaces for intervention.
Schools' role has more recently been shared with other institutions, which also work in the local sphere (primary care facilities, daycares, community centers, social care centers, churches, among others), leading to changes in regional dynamics. That being said, the autonomy granted to these units has increased, especially due to the decentralizing nature adopted in social policies, which has been spread by the paradigm of New Public Management.
In other words, there is undoubtedly an ongoing process of reform in various central governments globally. This reform imposes a new agenda on the State, resulting from the social, political, and economic dynamic which has become a trademark of o number of national governments (OFFE, 1984;PEREIRA, 1998;OLIVEIRA, 2014;CASTELLS, 2003). Hence, the State's ability to respond to issues has also undergone changes resulting from this ever-changing context.
In the case of educational policies, several authors have shown that the work of teachers, the organizational routine, and the everyday activities of managing teams have changed as a result of the added new attributions they have been given, such as operating bank accounts, filling out spreadsheets, taking part in meetings to assess performance indicators, taking part in regional educational plan meetings, among many other attributions which have expanded the managing scope (PARO, 1992;NÓVOA, 1995;BARROSO, 1996;BALL, 2002).
Considering this scenario of educational policy reform in the past decades in several countries, Ball (2005), highlights the inclusion of the management and performance culture. This has directly affected classrooms and teachers, whose practice is "remodeled" to conform to new rules that are created externally. Autonomy, competition, efficiency, and productivity become key concepts with which new identities and subjectivities emerge.
As for this greater autonomy given to professionals and schools in the decision-making process, it does not represent the abandonment of State control, it means rather the adoption of a less visible control mechanism, as Ball (2002) points out. For Ball, the post-modern search for order, transparency and classification -which characterizes the economic autonomy given to schools -changes not only what teachers do, but also what they think and what they are, that is, it changes their identity (BALL, 2005).
This process, which empowers schools and their professionals, gives rise to fragilities and dilemmas that need to be properly taken into account. The intensification of the work of education professionals, especially teachers (in activities related to teaching, assessment, administration) comes together with countless factors such as lack of identification of teachers with the workplace, a little attention given by schools to "the work of thinking about work", inadequate teacher training and recruiting (as schools and teachers are left out of the process) and limited autonomy outside the classroom.
In this context of expanding the autonomy of schools and their professionals, as compared with the incorporation of new routines and attributions, managers become a central player. As Paro (2010), points out, principals and educational managing teams play a strategic role, which is also complex and contradictory. Even though the roles they play have a strong educational connotation, they are also political roles, as these people coordinate individuals who have wills, interests and perform actions which, in addition to being unpredictable, are not limited to the school environment. Moreover, the work they carry out implies knowledge of a complex set of administrative-financial activities, educational work organization, in addition to articulation with local players (families and communities) and with external evaluation and regulatory bodies.
In the case of teachers, in addition to mastering content, they are expected to be able to deal with social and psychological issues presented by their students, as well as to be responsible for their sexual education, and the integration of students with special needs. Besides, they need to be up to date with the use of new technologies, at the risk of seeing their teaching style become obsolete. This all takes place in a broader context related to redefining the role and function of schools which, in the past, were responsible for transmitting the dominant culture to a more homogeneous group of students, and now face the need to cater to students with different social and cultural backgrounds. Thus, in this sense, the role of schools has been considerably expanded, and these places increasingly constitute one of the few spaces where it is possible to promote social interaction among people with distinct social, economic, and cultural backgrounds (BALL, 2002).

SCHOOLS AND LOCAL IMPLEMENTATION: MANAGING UNITS AS PART OF THE SYSTEM
School agents become strategic players in the process of public policy implementation regarding schools considering the governmental perspective of bureaucracy (BALL, MAGUIRE AND BRAUN, 2012). School professionals -principals, educational teams, teachers and administrative teams -are public agents who, to a greater or lesser extent, will translate and implement public policies designed by different governments at the strategic level. This is the very institutional space where schools' roles are expanded and revealed: more than educational and educative spaces, schools are political spaces where an array of state bureaucratic aspects are revealed.
Public schools are one of the many links integrating governmental and state bureaucratic structures and materializing Public Administration, as are hospitals, primary care facilities, universities, social security offices, federal revenue offices, police stations and public offices in general (WILSON, 1989). They are so because they integrate the State's organizational structure, giving visibility to the State in the educational field as State actions cease to exist in isolation. For this reason, public schools can be considered one of several bureaus that are part of public bureaucracy.
Accordingly, we understand that in the process of implementing school-related policies, schools cannot be analyzed as something detached from the system. On the contrary, attention must be given to the bureaucratic dimension of schools, as one of the links that integrate them to public bureaucracy and materialize Public Administration (SELZNICK, 1966). As public schools integrate the State's organizational structure, giving visibility to State's actions in the educational field, these actions cease to exist in isolation and are, by definition, submitted to the principles of Public Administration.
This may be a gray area in the educational field, especially because when schools are linked to this organizational structure and governmental hierarchy, we impose a certain rigor on them, which is often disregarded in studies in the area. The association of this micro-level organizational dimension with the notion of a network or a Public Teaching System cannot be ignored in the everyday actions carried out by agents acting within this system (PETERS, 2001). The very limits and possibilities of school action, in turn, both considering the organization and the institutional perspectives, cannot be disregarded either. This is also the context in which the process of designing and implementing public policies must be analyzed, relativizing the assumption that due to this occasional rigor imposed by the bureaucratic system public agents have little or no room for maneuvering to act with discretion. Meanwhile, when public organizations are analyzed considering the informal social relations established, as it is well-described in sociology and social psychology, the way these policies are perceived by these players may reveal very interesting aspects that will lead to a better understanding of how these institutions work on an everyday basis.
The contributions made by Lipsky (1980), Downs (1999), Vinzant andCrothers (1998) andMaynard-Moody andMuscheno (2000), confirm this thesis in terms of the institutional, organizational and political perspectives. In the same direction, Maguire and Braun (2012), analyzing teachers as political actors, that is, as State representatives -and, why not, as government representatives -in the process of making available public policies, puts the school on the spot as a strategic sphere in this flow.
The analysis made by Tardiff (2002), on the other hand, is closer to the sociology of the professions and educational sociology, as it attempts to draw several routes taken by teachers to prove that teachers themselves are forged dialectically from the alchemy which occurs when multiple exogenous and endogenous variables are combined. This process may explain the "liquidity" of individuals discussed by Bauman (2004), as the articulation between these two dimensions allows one to conceive teachers as liquid and "fluid" actors, who "blend in" their body of knowledge and are far from being solely limited to the educational field.
In this sense, schools emerge as institutions permeated with "informal" or hidden organizational issues, many of which are disconnected from the field of education. In other words, public schools are also one of the many links in state actions, and they are subject to the different strategic goals of distinct governments. That being said, schools integrate the state's bureaucratic structure, which, to a greater or lesser extent, makes them subject to the guidelines defined by these governments. This aspect reveals an interesting issue: how to characterize schools if, given their well-known specificities, there is not one but various types of schools? Despite having the same social function, countless schools translate this social function with an array of intensities and connotations, yet, in a very unique way.
When analyzing this issue in the school context, Derouet (1996), added the dimension of extraterritoriality concerning central regulatory bodies. For him, public agents who are submitted to longterm internal control in their corporations, lose the sense of service provision and, in the end, the machine runs "free", that is, based on the interest of these professionals themselves. According to the author, the only way to make teachers rediscover their mission is to confront them with social search and, consequently, with the local sphere. In a context of State reform, notably marked by administrative and financial decentralization, we understand that schools must be increasingly empowered, acquiring greater flexibility and autonomy to adapt to the demands imposed by increasing numbers of students as well as by schools' expanded responsibilities.

SCHOOL PROFESSIONALS PERCEIVED AS STREET-LEVEL PUBLIC AGENTS
According to Lipsky (1980), street-level public agents are those who interact directly with citizens in the process of implementing and making public policies available. Teachers and education professionals, social workers, police officers, traffic and healthcare agents, physicians and nurses can be considered classic examples of these players because, according to the literature, they are the actors capable of exerting some influence on the relationship established between citizens and the government.
Thus, within their functional characteristics and considering the positions occupied by these individuals in public organizations, it is they who will mediate, process and follow up on the process. The vast majority of these actors play their roles remotely, that is, away from strategic offices and hierarchical levels. Apparently, their role is not significant, especially when the top-down policy implementation model is used. However, empirical studies have shown the importance of these agents in the implementation process, especially when the bottom-up model is the analytical reference (WINTER, 1993;2003).
It is from this perspective that schools, teachers, and school professionals can be considered typical examples of street-level bureaucracy/bureaucrats, clearly reflecting Lipsky's descriptions. Teachers and education professionals, as well as other professionals analyzed by the authors, have their strategies of action. They are not subject to strict management control and are largely guided by their intuition. Their workplaces, organizational structures, teams, speeds, schedules, routines, and administrative processes are also different.
Schools are conceived, therefore, as dynamic and flexible organizational structures, as their processes and administrative rules take the contours traced by these individuals. It is this process that shapes, in time, the organizational culture of schools, in addition to explaining how public education policies can produce such distinct effects, even when based on uniform general principles.
It can be said, therefore, that these professionals perform a great variety of functions that are not always directly linked to the educational process itself. But because these roles are linked to the school from an institutional, strategic, and sectorial point of view, they eventually bind together all the players that make up the organizational structure of these units, in addition to bringing them closer to other external public agents. These external agents also end up involved and involving the players mentioned before. Once these actors are relied upon for policies to come about, understanding how these roles are played by the actors is key to understanding the road map of policies inside and outside those spaces.
Given the above, we can state that there is a large number of individuals involved in the process of making available and implementing social programs and school-related public policies. Teachers, education professionals, social workers and healthcare professionals are some of the main actors in this chain, each of them performing a set of functions that, to a greater or lesser degree, can promote changes in the course of actions and/or in the outcomes of the programs. We understand that a policy will only be effectively a policy when it becomes available to citizens, and this most often happens through the mediation of street-level agents.
These agents are, therefore, actors with some discretionary power, since they are the ones in direct contact with citizens (MAY and WINTER, 2009). It is precisely this direct contact that allows these agents to add a set of values or representations to the policies, values that can, ultimately, change the course of these policies. From a citizens' point of view, the fact that the implementation agents are recognized as representatives of the government, materializing its very presence, makes them understand public policies from this perspective. This can affect these individuals' behavior towards the relationships that will be established between citizens and government.
In short, these agents' decisions and behaviors are not merely operational. In other words, these individuals are also political agents. Their behaviors and attitudes, supported by the inherent discretion of the function they perform, are also loaded with "political motivations" that may, to a greater or lesser extent, influence how they act daily.
In this sense, based on the contribution of Converse (2004), the school can be seen as an arena where the "belief systems" of these actors meet and merge. According to that author, through a set of values, perceptions, ideas, and interconnected attitudes, individuals will indicate the predictability of their behavior. If we analyze this statement from a cognitive standpoint, that is, from the ability -or lack of it -of individuals to turn/translate values and preferences into behavior, we will notice that the question that is revealed is closely related to the degree of stability of those belief systems, since, to a greater or lesser degree, it will be the cohesion of this system that will allow us to predict the behavior of these actors.
Teachers are, therefore, "liquid agents" whose existence and action depend on their ability to master, integrate and mobilize this belief system -or types of knowledge, according to Tardif (2002) -as conditions for their practice. In other words, these types of knowledge are not limited to the institutional and academic production, historically coordinated by universities and shaped by teaching systems. Although this articulated network of "formal knowledge" is indispensable for the educational process -and the implementation of school-related policies or programs -it does not suffice to explain the processes of social knowledge production in the universe of modern and contemporary culture.
Finally, there is a set of more "porous" and subjective -or liquid -types of knowledge, as per Bauman's (2004), formulation -built through the everyday practice and as a result of the environment where the teacher-individual is. This knowledge, which Tardiff called experiential knowledge, "springs from experience and is validated by it". According to Tardiff (2002), studies indicate that, for teachers [and other players working in schools], the knowledge acquired through professional experience is the basis of their competence. It is based on that that teachers judge the relevance or realism of the reforms introduced into programs or methods. Finally, it is based on the experiential knowledge that teachers come up with everyday strategies for their profession.
A simple and direct conclusion emerges from this characterization: teachers are the individuals who articulate these types of knowledge. It is from this perspective that teachers can also be considered technicians, bureaucrats, or public agents of implementation at the local level, as defined by Lipsky (1980). Their practical and everyday knowledge forms a set of representations and values that are interpreted and understood differently and individually by teachers. This individual decoding ultimately explains countless decisions that are put into action every day, in the various school organizations, by each of the agents working there.
It is this complex network of representations and decisions that shape and make up the socalled "teaching culture". This happens because these professionals work in an environment limited by multiple variables related to concrete -but not necessarily objective -situations that cannot be addressed by ready-to-use solutions and linear recipes/methods of cause and consequence. These often require a large amount of improvising and personal skill, which can only be acquired through the intense experiential process that results from years of work.
Unlike other professions, these individuals do not work alone, although, ultimately, many decisions are made exclusively by them. However, this does not mean that their work is devoid of interaction. After achieving some consensus and a minimum level of information on a given issue, in a scenario of constant integration with other actors (other school professionals, students, families and the school community), it is up to them to make the revealing decision of this network of implications.
Regardless of working inside or outside classrooms, according to Tardiff (2002), these actors do not act upon an object, a phenomenon to be understood or a work to be produced. On the contrary, their actions take place in a network of interactions with other people, in a scenario largely dependent on the human factor, filled with values, feelings, and attitudes that are amenable to interpretation and decision, usually in a context of urgency.
Thus, the knowledge of teachers -and other school professionals -covers a diversity of objects, issues, and problems that are all related to their work. For "professional teachers", their work experience seems to be the main source of knowledge and teaching skills. This combination of knowledge and teaching skills influences how they perform their work, including factors of cognitive and even administrative nature. In other words, their personality, talents, empathy, prejudices, love of their job, emotional and psychological structure, political preferences, representations of education and teaching, relationship with the school, with their colleagues and with families, all together, also bring to light individual knowledge and attitudes that will somehow produce the aforementioned teaching culture.

SCHOOL PROFESSIONALS AND LOCAL IMPLEMENTATION: TRADE-OFFS AND RELATIONAL SUPPORTS
The work done by public implementation agents materializes the State's action in the interaction networks that are set up locally, not only among themselves but, obviously, between them and citizens. Depending on the discretion inherent in these professionals' behavior, local-level bureaucrats can distribute and allocate the available resources to provide more or less access to the policies, both from the perspective of citizens and the public sector. This means that these individuals can be strong -or weak -links in the construction of these interaction patterns, driving users closer to -or away from -the public arena. The border between State and society can, therefore, be more or less porous, depending on how these actors establish their relations with the population.
According to Hill and Hupe (2014), studies on policy implementation have suggested that the attitudes, values, expectations, and beliefs of street-level agents can have a substantial impact on the outcome of policies. Likewise, the author argues that the way the implementing agents understand, interpret and perceive the design, objectives and operative rules of policies influences how such actions will be performed by them. Since they constantly interact inside and outside the spaces where they work, the relationships built over time are also decisive for us to understand how implementation can be constrained by these individuals.
Research on networks applied to social policies has advanced and multiplied in the last decades, especially after sociology became interested in the micro-processes involved in the relationships among social groups, entities and individuals. Marques and Moya (2012), makes a very brief compilation of the publications in the area, indicating, among others, studies on processes of migration, poverty and segregation in metropolitan areas; building and breaking social relationships; spheres of deliberation in urban spaces, networks in the business world, the search for employment and the impact of public policies on community networks.
In his studies on networks applied to social policies, the author says that social networks make up the fabric of relations among individuals, groups and entities in our society, structuring the fields where social phenomena come about. On a micro-social scale, the relationships built by social actors every day influence the sociability of these agents, affecting the way they solve their problems and seek help, where and how they find work and where they build and transform social bonds.
In another dimension, the web formed by the bonds among individuals, groups and organizations determines the relationships of power and the conflicts present in the State and its surroundings, involving all the stakeholders. The author also says that these relationships can influence governments in the process of providing public facilities and services, affecting welfare levels.
Just like in the research carried out by Marques and Moya (2012), we can infer that school professionals build relationships that can also be based on a trade economy, with positive and negative incentives to the cooperation between individuals who work in the same territory, whether it is in the same school, different schools or in social facilities somehow related to the school, like health centers, community centers, daycare centers or similar bodies.
According to Ball (2016), interactions indicate that agents forge links that bind their social group to others, regardless of their location. In practice, agents build bridges with colleagues in their territory (other schools, health centers, churches, daycare centers and other reference spaces), but they are not limited to them. On the contrary, they relate to other agents in different sectors of the city, state and third sector.
It is interesting to notice that the more intense and broad this network is, the greater the possibility of creating new circles. In other words, the agents who build these bridges tend to be at the border of their respective social groups, playing a strategic role. In this sense, the quality and depth of these links may indicate the stock of relational capital (or social-political power, according to Marques and Moya (2012) of each of these professionals, reinforcing the thesis that such arrangements are very heterogeneous and dynamic.
These notions are in line with Lipsky's (1980) thesis that social relations, even in spaces with a high degree of institutionalization, continue to mediate the access between public organizations as well as between these organizations and citizens. Therefore, we can say that public policies eventually induce and encourage this type of interaction at the local level, allowing successive exchanges to take place among the agents involved. Policies will then be the result of these various combinations.

FINAL CONSIDERATIONS
Schools and other public facilities have their dynamics, which go beyond what organizational planning manuals prescribe. According to Ball (2007), teachers and principals have been facing several types of changes that have affected their routines and the scope of their activities, which have been expanded (involving technology, growing violence, unemployment, different family setups, new social, political and economic contexts, among other countless changes).
Nevertheless, the fact that schools will adapt to this situation does not guarantee that their professionals will unconditionally support such pervasive social policies. On the contrary, strategies of sabotage, resistance, apathy, or indifference are powerful mechanisms that explicit the political positions adopted by these agents, which is why they should be better decoded by policymakers, as suggested by Evangelista and Shiroma (2007) and Ball (2007).
It is precisely this direct contact with users that enables agents to add a set of values or "belief system" to policies. These values and beliefs can, ultimately, change the course of these very policies (ELLIS, 2011). From a citizens' point of view, the fact that those individuals are recognized as representatives of the government, materializing its presence, makes them understand public policies from this perspective. This can affect these individuals' behavior towards the relationships that will be established between citizens and government (SICILIANO, 2017).
We can assume that this type of (re)signification results from the social interactions that permeate institutions or public spaces through the various informal and primary groups. That is why it is very difficult for the State to keep total control over how educational policies and/or programs are implemented. These spaces of action will inevitably be affected by a network of new and real meanings, translations, attitudes and perceptions concerning official proposals. This process is so informal and so subtly integrated into the organizational culture of these institutions that it is difficult to see it in the everyday relations that characterize these spaces. On the other hand, it is quite possible to say that these different "belief systems" are incorporated into the decisions, positions, and attitudes of these agents in their daily routine and that, for this reason, they must be considered by policymakers.
It is in this sense that teachers-individuals -and not only them but all the professionals working in schools -are living in a world that is filled with rules, prescriptions and obligations that must be respected and followed. These "rules of the game" apply to a set of routines that include the administrative, pedagogical, relational, political and educational spheres.
If, on the one hand, schools and their professionals cannot escape certain rules, on the other hand, it is they who will decide the contours that a given policy will take in each school unit. In this sense, every educational policy involving schools may be distorted or modified by teachers or other actors involved in these spaces. This re-signification carried out by school actors, at the other end of the formulation of educational policies, is a constituent of these very policies, and cannot be naively ignored by program managers. This can only be observed if we shift our gaze from bureaucratic aspects to the micro-political processes and the relations established between the groups and actors within schools. Given the growing autonomy and accountability of education professionals in the context of the New Public Management, it becomes mandatory to recognize the political action -porous and unpredictable -carried out by these local agents to better understand the implementation of educational policies in school spaces.