Synchronous accountability forums: the internal dynamics of budget public hearings

Budget public hearings are spaces for transparency and social participation. These spaces can be considered synchronous accountability forums, where the rendering of accounts and the attribution of responsibilities can co-occur, raising concerns for the responsible parties about possible accountability models. This qualitative and interpretative research analyzed the dynamics of public budget hearings organized by Brazilian municipalities using semi-structured interviews with public managers. The results indicate that those responsible for public hearings adopt strategies minimizing reputational and accountability risks at different times: (i) in the preparation stage, with the definition of participation rules and the use of an informational filter ; (ii) during execution, in the form of the presentation of information and in directing the public’s attention; and, (iii) in the post-hearing stage, through low transparency and document manipulation . The research contributes to understanding hearings as forums of expected accountability, in which the expectation of the authorities can direct the dynamics of public hearings, transforming them into a theater performance. The results have implications for the courts of accounts and the legislative branch, which could exercise greater control and inspection of the hearings; for society, which could act in a more active and organized way; and also for the legislation in force, which could establish clear guidelines for holding hearings.


INTRODUCTION
Public hearings can be considered accountability spaces or forums, which make up one phase of the decision-making process, such as the elaboration and monitoring of the governmental budget process (Fonseca, Rezende, Oliveira, & Pereira, 2014;Lando, 2003).
Even though public hearings are advertised as mechanisms of transparency and participation, they are characterized as invited spaces (Cornwall & Coelho, 2007), whose dynamic rules are generally defined by the state with society just being invited to participate.Within the Brazilian context, this is due to the lack of specific legislation for hearings, which are inexistent even though they are legally required (Fiscal Responsibility Law -Lei Complementar nº 101, de 4 de maio de 2000, art. 48).
Budgetary hearings can be seen as accountability forums, or in other words, a space in which the government (account-giver) presents its accounts to a public forum (account-holders) (Bovens, 2007a), which in this case, can assume various positions and portions of the public, such as representatives of civil society, the Legislative Branch, the media and members of the opposition (Brummel, 2021).The accountability forums discussed by the literature generally occur in an asynchronous format in which the interactions between the account-giver and the accountholder have an interval of time for each interaction, which allows time for reflection and the response.Rocha's study (2013), which analyzes accountability in the Court of Accounts reports, and a study by Patz, Thorvaldsdottir, and Goetz (2022) which analyzes the use of feelings in public reports, are examples of assynchronous accountability.We did not find studies which discuss the synchronicity of accountability forums and how this element can result in different behaviors on the part of those presenting the accounts.
In asynchronous accountability forums, the presenters of the accounts have more time to evaluate the questions, responses, and their possible consequences before they need to react.In public hearings, on the other hand, the mechanism is by its nature synchronous, and since there is the possibility of immediate accountability (through public questioning or reputational effects), there is an attempt to anticipate/preview what can happen (Schillemans, 2016).The anticipation or preview of accountability can favor information use strategies, such as emphasizing good news and omitting/obscuring bad news (Mahler & Regan, 2007), or using information in an opportunistic manner to denigrate the image of a political opponent or strengthen one's own image (Gaber, 2000).
The aspects raised by the literature only focus on the public moments of these hearings.It is not clear in the literature how the internal decisions of organizations -whether they come before or after the hearings -are made and how they can involve strategies related to the way information is produced and utilized, which can influence and shape public hearings.Exceptions to this are Grimmelikhuijsen (2011), Mahler and Regan (2007) and Piotrowski, Grimmelikhuijsen, and Deat (2019), who address this theme in other contexts, such as websites and other means of communication.
This study seeks to analyze how the dynamics of public budgetary hearings are organized by municipal governments.In terms of hearing dynamics, this study considers the preliminary organization process, its realization, and previous activities, such as follow-ups to questions made by the public.We have realized a qualitative interpretivist study with municipal governments in the state of São Paulo as our units of analysis.This study aims to contribute to the literature on public transparency, proposing an analysis which goes beyond the public moments of hearings and focuses on internal decisions and possible strategies for the use of information involved in this process.It also aims to contribute to the accountability literature, discussing a type of forum which has not been explored thoroughly, namely synchronous mechanisms.Our contribution is also directed towards the felt accountability literature (Schillemans et al., 2021), which considers the expectation that the manager will take responsibility before presenting the accounts, which can involve strategies for the conducting and use of information.

Synchronous accountability
The term accountability refers to the obligation to present an organization's accounts, together with justifying and explaining them (Bovens, 2007b).In general, accountability forums occur in an asynchronous manner and are made up of several elements: (i) the information that the accountgiver should give the account-holders; (ii) the existence of a phase of debate and questioning; and (iii) an opportunity for there to be consequences or sanctions issued by the forum, which can even "denounce a policy or publicly condemn a behavior" (Bovens, 2007a, p. 451).
Public hearings for the budgetary process can be understood as expected accountability forums (Bovens, 2007a;Buttny, 2010;Schillemans, 2016), in which an actor presents accounts to a responsible party (Bovens, 2007b;Busuioc & Lodge, 2017) and expects to be judged at that time.Expected accountability appears in hearings in which, at a synchronized moment, governmental representatives can try to foresee or anticipate possible pressure and risks of being held accountable, which are motives for changes in their behavior and the way they present their accounts (Schillemans, 2016).
Given that the information is supplied by agents, the space is shaped by them, and the forum is a meeting that one is invited to, the account-givers can try to foresee and delineate what the dynamics will be like (Schillemans, 2016).Given the synchronous nature of this accountability forum, the state anticipates concerns that may involve being held accountable for its actions or diminishing its reputation, and thus it develops strategies related to the manner in which information is organized and presented.
The psychological and behavioral fields address accountability in a different manner: "What matters is how accountability is perceived by the individual instead of its formal and objective characteristics" (Schillemans et al., 2021, p. 4).Then, the focus is not the moment of accountability, but the perception of the actor in terms of the possibility of accountability and the anticipation of holding someone responsible in the future which results in a behavioral response.In budgetary hearings, the focus can go beyond the public moments of the hearings, like a previous internal moment, in which pressure in an accountability forum is anticipated by a manager, which leads to defensive and strategic behavior.
A perspective which explains the decisions and behavior of public managers is felt accountability (Hall, Frink, & Buckley, 2017;Schillemans et al., 2021).In a manner similar or complementary to expected accountability (Schillemans, 2016), felt accountability refers to expectations or sensibilities which occur when an action can be evaluated and judged, generating a response in terms of the behavior of the actors (Schillemans et al., 2021).This response of the actors in the case of public hearings can involve an anticipatory movement of judgement, or in other words, managers can fear being questioned and held responsible by the public and will thus alter their behavior and/or the published information ahead of time.Busuioc and Lodge (2016) suggest that this anticipation of negative evaluations can generate concerns for their reputations which can shape how agents present their accounts in accountability forums, or in other words, those who are presenting the accounts may not be interested in being true to the facts or may seek to distort information (Busuioc & Lodge, 2016).
These circumstances can also be observed in studies of transparency (Grimmelikhuijsen, 2011;Mahler & Regan, 2007;Piotrowski et al., 2019), such as review procedures and the triage of content that will be published on the government's websites (Mahler & Regan, 2007), as well as various treatments that can be given to positive information in contrast to publicizing information that is more sensitive or complex (Eschenfelder, 2004).The use of this type of transparency mechanism in accountability forums suggest a selection of what content to publish, since there is difficulty in distinguishing to what extent transparency in accountability forums is complete (Grimmelikhuijsen, 2011) and is presenting neutral information, whether it is positive or negative.

Public budgetary hearings
Public budgetary hearings have been accepted as mechanisms for transparency, social participation and the exercise of democracy (Ebdon, 2002;Fonseca et al., 2014;Zorzal & Carlos, 2017).Even though they received attention as legal rules after the passage of the Fiscal Responsibility Law in 2000, public hearings are not recent practices, but rather traditional and widely used (Checkoway, 1981).The government normally holds hearings during the implementation process of some public policy to enable citizens to participate in the decision-making process (Yang, 2003;Zorzal & Carlos, 2017).
The literature on social participation and deliberative democracy emphasizes the importance of citizen involvement in decision-making processes for governmental policies (Bohman & Rehg, 1997;Hartz-Karp & Briand, 2009) given that these decisions directly affect the public.Give electors' dissatisfaction in terms of their representatives (Abranches, 2018), this type of participation makes it possible to introduce elements of participative democracy in representative democracy, such as the Brazilian scenario (Costa, 2010;Sacramento & Pinho, 2016).
Even though the positive points and importance of public hearings have already been accepted, previous studies have also emphasized the difficulties involved in the efficient functioning of these spaces, for example, low social participation (Lando, 2003;Zorzal & Carlos, 2017) and the lack of appropriate technical language (Sander & Pedralli, 2013).Low participation may be motivated by a reduced trust in government (Kasymova, 2017), a lack of interest, disinformation about the existence of these spaces, or unviable hours (Bragatti, Nascimento, & Carvalho, 2021).In addition, the participants are not always popular representatives, and sometimes they are defending their politicians, public employees or other groups with their own particular interests in governmental decisions (Checkoway, 1981;Rowe & Frewer, 2000;Zorzal & Carlos, 2017).
In terms of the dynamics of conducting them, these hearings are held in small spaces, normally in the form of an auditorium, whose dynamics generally include a presentation of the information, which has already been prepared and selected by the government with the public just listening (Lando, 2003;Rowe & Frewer, 2000).Then, there is a brief period of participation in a controlled manner.These stages are coordinated by the government, which also uses the space for political speeches and other discussions (Brelàz & Alves, 2013;Corona, 2007).
The invitations to participate are not always timely, and the date and time of these events may not favor participation (Brelàz & Alves, 2013;Checkoway, 1981); the locations normally are municipal government or council buildings where "a citizen can feel in foreign territory" (Checkoway, 1981, p. 567) making the dynamics more convenient and controlled by the government.
There are also critiques in terms of the real impacts of public hearings, given that, even though there may be popular participation, this is not synonymous with having influence on the decisionmaking process.In some cases, hearings consist of the registration of demands (Brelàz & Alves, 2013;Zorzal & Carlos, 2017).This format does not always permit communication and social participation, because "contributions should be a dialogue, not a speech" (Ebdon, 2002, p. 284).For example, public hearings in the city of São Paulo, SP, only allot a brief period (around three minutes) for the public to speak, without the right to reply once or twice, which would make a dialogue/discussion about demands effective (Brelàz & Alves, 2013), with no follow-ups for society (Fonseca et al., 2014).
These public hearings can be seen as practices aimed at acquiring social legitimacy (Topal, 2009), in which the creation of an apparently objective democratic forum demonstrates society's involvement in decision making, if only in a predefined manner.Topal (2009) states that hearings are not held to promote participation, but rather to legitimize the private actions and interests of those who hold them.
These studies of public hearings focus on their informational and public aspects.We did not find studies which address how public hearings are planned and organized and the way in which information is used in these spaces, as well as the dynamics, which should consider the previous preparation for, and the posterior treatment of, the received demands.

Case selection
This study adopts a qualitative and interpretive methodology which focuses in-depth on the selected cases.We are seeking transferability (Power & Gendron, 2015), or in other words, the ability to generalize theoretical propositions about this phenomenon.
Our cases consist of municipal governments in the state of São Paulo (SP).This state was selected due to the discrepancies between the pressure caused by the Court of Accounts (Lino & Aquino, 2018) and previous analysis with information requested by freedom of information act of 26 courts responsible for monitoring city governments.Of the 24 which responded: 50% declared that city governments do not send information about hearings; 46% realize some type of monitoring; and one refused to respond (4%).These responses suggest a divergence in supervision, which generates different pressure on the municipalities in each state, as was verified by the interviews.
In addition to requests for documentation in digital form, the Court of Accounts of the State of São Paulo (TCE-SP) has taken a position on this subject, establishing due dates and overall guidelines (TCE-SP, 2019, 2020), which demonstrates some concern with budgetary hearings.
Table 1 presents information about public hearings in the Effectiveness of Fiscal Management Index (IEGM), utilized by the TCE-SP, which offers us a view of the situation in the municipalities of São Paulo.Practically all of the municipalities in the state declared that they hold public budgetary hearings and document them with minutes.However, when we look at the transparency of information referring to the contents of the hearings, such as the publishing of minutes, and follow-ups to requests, their incidence is lower (36% and 63%, respectively) and lower still (18%) when it concerns permitting monitoring of the inclusion and implementation of proposals in the budget.
We conducted 25 interviews with 19 respondents from 23 municipalities, whose characteristics are presented in Table 2.The difference between the number of respondents and municipalities occurs, because some respondents can act as consultants who deal with more than one municipality.
The municipalities varied in size from small and medium to large and demonstrated, on average, a high degree of transparency.This study did not consider ex-ante the establishment of a fixed number of cases/interviews.What matters is theoretical saturation, or in other words, attaining a depth of knowledge about a phenomenon until the research question is answered, and new interviews will not reveal new insights (Nowell & Albrecht, 2019;Saunders et al., 2018).Court (2016).

Data collection
The data was collected from semi-structured interviews (Behr, Kaczmirek, Bandilla, & Braun, 2012;Qu & Dumay, 2011), which permit the construction of narratives from each respondent's free-form unstructured responses.We also consulted the transparency portals of these municipalities in order to triangulate the data with published information about public hearings such as minutes and recordings.The interviews were conducted to capture the dynamics of hearings and possible intentions and strategies (Qu & Dumay, 2011) through the narratives of those involved in the process.The protocol was approved ahead of time by an ethics committee.The questions were developed using the Probing Questions approach (McKinnon, 1988), which utilizes general and open questions: how are the hearings prepared?What are the dynamics of the public hearing, with details about the presentation and conducting strategies?Who participates?How is the moment of dialogue/participation organized?What are the follow-up actions after the hearings?
These respondents are representatives of municipalities who are responsible for organizing and presenting public hearings, such as accountants, assistants/consultants, and Secretaries of Planning and/or Finance, among other actors (Table 3).Notes: (1) The indicated respondents/specialists were consulted again to validate our findings; (2) Consultants can act in more than one municipality.Therefore, we calculated the average number of municipalities that they provide services to.
Source: Elaborated based on the information provided by the respondents.
Some respondents were interviewed more than once, due to limitations of time, to deepen our understanding and to validate our analyses.Interviews with specialists were conducted at the initial phase and during the refinement phase (Yeong, R. Ismail, N. H. Ismail, & Hamzah, 2018) until the validation and discussion of our research findings, as in Lino, Azevedo, Aquino, and Steccolini (2022).
This study counted on access to key respondents who have relevant experience with this theme, which made it possible to deepen our knowledge in subjects which require trust and affinity to be addressed (Taylor, 2011).The respondents were selected through the snowball technique in which the interviewee suggested possible new respondents from their network which facilitated introductions to new key informants (Patton, 2015).
The interviews were recorded -with proper authorization -and literal transcriptions were made (verbatim) in a manual manner (Davidson, 2009), respecting ethical principles and protecting the identity of the respondents.

Data analysis
The interview analysis adopted inductive logic through narrative analysis (Esin, Fathi, & Squire, 2014).In this method, research is not derived from preexisting categories to explain the phenomenon; patterns and ideas emerge from the data which is constructed during the study (Kennedy & Thornberg, 2018).In the interpretation process of the narrative analysis, it is not just what is said that is important, but rather the form and perception of the respondent in terms of what is being narrated (Esin et al., 2014).
Analyses are based on a codification process in phases, in which the data is tabulated, analyzed and summarized in search of emerging patterns and the construction of typologies (Watkins, 2017).The first stage of codification generated 92 codes, which initially were grouped without using theoretical constructs so that the data could emerge freely.After the first round of coding, there was a round of second order coding which generated 36 codes.The summarizing of the data consisted in a revision process for the first order codification, which was regrouped based on the theoretical characteristics that various pieces of data had in common.Both rounds were not conducted just a single time, and they were reconstructed to refine the data.Box 1 illustrates this method by passing through the codification process.

BOX 1 FINAL CODIFICATION
1 st Order Cod. 2 nd Order Cod.Response

Selectivity of Negative Information Informational Filter
Mainly when the Secretary is presenting, the bad things often are not theoretically presented with a direct tone, they are not demonstrated, except when there's a fall in revenues that will harm the municipality.

Selectivity of Positive Information
Informational Filter [...] Normally the good things come up in the public hearing.For example, "I will create a water fee for low-income families, I will create a device to help people, I will double the quantity of food baskets …".These subjects are opening themes of public hearings together with the elaboration of the budget.
Source: Elaborated by the authors.
The rounds of codification were realized concomitantly with the interviews to retrofit the data collection process and improve the upcoming interviews, as was discussed by Rapley (2014).We considered that theoretical saturation had been reached when few new codes were added by new interviews.
After analyzing the results, we conducted four validation interviews with respondents and specialists, as proposed by Lino et al. (2021).This phase is intended to verify whether the findings make sense according to the perception of the respondents, in addition to providing additional comments and suggestions (Bogner et al., 2009).

DYNAMICS AND STRATEGIES OF PUBLIC HEARINGS
This section presents the analysis of how the dynamics of public hearings occur, and it is subdivided into three sections: the preparation of the hearings, their realization, and post-realization, and in each step, we describe some of the strategies identified in the analyses.Figure 1 presents the dynamics of budgetary hearings which emerged from the research analyses.
The striped bars symbolize barriers that impede the flow of information and transparency in the process, or in other words, these are strategies used by those who are presenting the accounts to limit information and public perception.The strategies for each step are described below.

Evaluation of Proposals Policies
Notes: (1) Pre-hearing: Internal moment when only the government has access to decisions about the form and content of the hearings.
There are two groups of information: (i) registered, which occurs as a result of past facts that have already been accounted for; and (ii) latent, which contains unconfirmed facts or those that do not belong to the accounting/budgetary scope.
(2) Hearing: Public moment in which the hearing itself takes place.
(3) Post-hearing: Internal moment in which the proposals are evaluated for the project's finalization and information about the hearing is made available including minutes, videos, and follow-ups to the received proposals.Source: Elaborated by the authors.
Some strategies can be utilized in planning hearings, others can be used in monitoring hearings depending on the scope of the hearing and the situation of the municipality as demonstrated by the following narrative: We have hearings for the elaboration of planning items, and we have public hearings for the results.Normally, during the results the mayor wants to point a finger when there is a bad situation, or he tries to say something to change the result or hide it in the presentation.When it has to do with planning, he points his finger before the item goes to the hearing (Respondent 1, Consultant).
We may observe that the reluctance to be held accountable (Overman & Schillemans, 2022;Schillemans, 2016) generates an effect of anticipation.The differences in the behavior of audiences in planning and execution is because in planning hearings, there is an attempt to avoid accountability or questioning in advance when the piece is being elaborated.In monitoring hearings which have to do with the presentation of results (registered information), the anticipation of being held accountable leads to an attempt to manage the impression given.
Below, the results are presented in three different moments: preparation, realization and postrealization.

Hearing preparation
The hearing preparation stage begins with the organization of budgetary information.However, it is not possible to segregate each item/hearing with a closed cycle, given that during the year, various hearings are held with related themes in which one hearing's dynamics can interfere with others.
In this phase it is common to have an internal meeting beforehand to align the content and presentation, which can involve technical as well as political issues.At this time, the actors adopt an informational filter in relation to the information that can be presented in the hearings (registered and latent information).The desired information is selected, hiding information that is not interesting or sensitive, such as an increase in debt or non-binding payments.
The informational filter can be used in two ways.First it can be used to manage reputation in which the information is used to improve the image of the government, as narrated by the respondent, who is an opposition Council Member in one of the analyzed municipalities: Obviously, the Secretary of the Treasury will emphasize what is politically important.He is nominated by a politician who depends on votes.So, he will want someone technical, just to have some security, but the narrative that he will produce from the data will be his.Thus, it is evident that he will emphasize that which he has accomplished well [...] So he will speak well of his work... he will not talk about anything that is not going well (Respondent 7, Council Member).
The second is an attempt to blame others, which consists of using information in a strategic way to attribute negative facts to others, such as the ample disclosure of unfavorable actions when they can be blamed on the previous administration.
Transparency is exercised in desired moments.The selective attention of those presenting the accounts is sustained by their concerns for their reputation in the view of the public (Busuioc & Lodge, 2016).This concern normally comes from political actors, such as: "At first, it is totally technical, and soon afterward comes the kick in the shins, the famous use of discretionary power.So, it is passed through a filter [...]" (Respondent 1, consultant).In public hearings, agents select information and guide public attention to give a good impression of their actions, or in other words, they focus on aspects that sustain and legitimize the government's conduct (Topal, 2009).
The elaboration of triannual hearings is performed by the Executive Branch, which in general, is handled by the Secretariat of Planning, Finance, or the Treasury unit.These units together with the manager will define the rules and organization of the event, which may involve strategies to limit the public's participation, as a respondent indicates: "The form of hearings inhibits citizen participation [...] people know that the majority of them are held late which limits the public's participation even more" (Specialist 4, accountant).Limitations can go beyond who will participate, because the time the hearing is held during the budget cycle can prohibit the consideration of citizen proposals, because "often there is no time, because the hearing is held soon before the deadline to send it to Congress" (Specialist 4, accountant).The experience of an Ex-Secretary of the Treasury below clearly indicates the existence of anticipatory strategies on behalf of the government: […] often the public hearing is held at a time that will favor the absence of people, during working hours, etc.Who is going to watch a public hearing in the City's Congress at 10 am on a Wednesday?It is designed so that people will not go (Respondent 9, Ex-Sec.Treasury).

Realization -the theatrical moment
The hearing's realization normally occurs in the Congressional building or an auditorium.Frequently resources such as a computer, a projector, videos, and printed documents are used in the presentation.As a consequence of the coronavirus pandemic in 2020, many city governments migrated to a remote model, accompanying the recommendations of the Court of Accounts (TCE-SP, 2020), which facilitates participation but does not guarantee it.
Low participation may be due to a lack of trust and frustration on the part of the population in terms of unaddressed demands: "During our last hearing, we had good participation with many people asking questions, but I did not hear any answers" (Specialist 4, accountant).In other words, even the presence of the population is not synonymous with participation and influence in the decision-making process, because the government decides which questions/suggestions to respond to or consider in the budget.
The poor performance of the Legislative Branch is also highlighted by the respondents: Normally they only say two words, "I approve" or "I'm against"; they do not analyze any bill that is sent to them: the hearings concerning their results are only attended when things are going badly; when they are going well, they do not even show up (Respondent 1, consultant).
The participation of council members has been described as seasonal, depending on the moment during the administrative cycle (such as the beginning of the mandate, when the public pays more attention); the political-party relationship between the mayor and the council member, which could influence the pressure inside and outside of the hearings; and the context of the municipality and the government, such as cases of corruption or sensitive subjects.
In this phase, the strategies used by the government involve the form of presentation of the information and the directing of the attention of the public.In the form of presentation, camouflage is used for information which has passed through an informational filter, but it is presented in a strategic order or intonation.Meanwhile, language can harm or facilitate the public's understanding, such as the use of technical terms for less positive information and more easily understood language for subjects that are more favorable to the government.The following excerpt clearly indicates the deliberate use of strategies in how the information is presented, not treating it as a "spontaneous" event, but in a certain way as a "rehearsed" event that has already been reflected on: Normally, bad things, are not theoretically presented with a direct tone, you know?They are not demonstrated [...], but normally good things always appear in public hearings, don't they?For example, "Ah, I will create a special water fee for low-income families, I will create a device to help people, I will double the number of food baskets […]".These subjects are themes for the openings of public hearings together with the elaboration of the budget (Respondent 1, consultant).
In terms of directing attention, the choice of the actors matters in the hearing scenario, to the extent that the presenter of the accounts is an actor who can assume different roles, knows the information and can choose how to present it, even utilizing other strategies such as language to try to minimize the risks of being held accountable, as the following narrative indicates: [...] then comes the Secretary of Social Assistance, for example, talking about the practices of the Secretariat, taking the focus away from the pure budget numbers.[...] I want to better demonstrate the social assistance situation, perhaps because it is a situation in which social assistance is good; if my social assistance were bad, I would certainly talk about another subject […] so, there is directing (Specialist 1, accountant).
The Secretaries, as public figures, are also guided by reputational concerns.In this narrative, the presence of the Secretary of Social Assistance at the hearing is thought to distract attention from the numbers or other less favorable subjects.
The utilization of public hearings as political stages is a strategy that is also related to selfrepresentation and the reputational concerns of the political actors, as the following passage suggests: [...] we have had great controversies here about non-binding payments, so various million nonbinding payments were made, even more here … thus, they gained a political connotation because they are the fruit of previous administrations, so there is that game of "ah, this administration is paying for the errors of the previous one [...]" (Specialist 1, accountant).
The public moment can have a political connotation when it is used as a tool even by the mayor in a "war of words" (Respondent 10, consultant).This and other strategies (such as the presence of political actors to distract attention) demonstrate that the moment that the accounts are presented is not neutral.The forum is dependent on the information that is presented at that moment, given that the information is not available beforehand.
In addition, the process of defining the rules of the hearing and the norms for the moment of participation can be used strategically by the actors to minimize questioning or pressure.The following excerpt exemplifies this strategy.
You hold a hearing but restrict the participation of the public.How: "Other questions we will answer at the end".You know that very political language?And we choose the person as well, don't we?We have someone talking about all sorts of things and then "look, we only have time for one more question.Send us your questions and we'll respond", and we never respond (Respondent 9, Ex-Sec. of the Treasury).

Post-realization
At the end of the hearing, the participants sign an attendance list that is sent to the court.As the literature has indicated and as was confirmed in these cases, social participation is low.To meet the supervisory demands of the TCE-SP, which requires the sending of the minutes and the attendance list, employees and other agents are often invited to participate and sign the list (or just sign the list) to constitute a quorum, as stated in the following report: "To have a sufficient number of people on the attendance list, you collect the signatures of people who were not even there" (Respondent 1, consultant).In other words, there is a manipulation of documents which are difficult to verify.
Even though the Court of Accounts of the State of São Paulo exercises some type of supervision in terms of hearings, this does not guarantee participation.We have identified a low perception of coercion by agents in terms of supervision by the court, as related: So, it's become a joke that we also have in terms of the Court of Accounts, that you can put a cake recipe there and nothing will come of it, because they don't check (Respondent 1, consultant).
None of the respondents related any situation in which the court demanded more explanations about the hearings.Even though it is not possible (or desirable) to generalize these findings, despite the greater concern when compared with other courts, the results indicate little attention is being paid to supervising the effectiveness of budgetary hearings.
Low transparency after hearings can also be used strategically, through the absence of the disclosure of information and restrictions in terms of monitoring, which is described below: I think it is important that we think of these [follow-ups], as a possible strategy to discourage participation.You go to a [hearing], there is no follow-up, you go to two and there still isn't, you say "I'm not going again, this is useless".Understand?So, this can often be used with this intention in mind (Respondent 9, Ex-Secretary).
Even though minutes and other documents are sent to the court, they are not always disclosed to citizens.Restricting monitoring of public demands and contributions can be used as a way of discouraging social participation (Kasymova, 2017), given that the public in general is not aware of later follow-ups or mechanisms to accompany proposals.In analyzing the municipal portals, the minutes of the 2022 Annual Budget Bill and the Budget Guidelines Bill were only located in one of the 23 municipalities.
The described strategies are summarized in Box 2, in accordance with the logic of each moment of the hearing and the respective actions realized.It should be noted that these strategies are not necessarily used in every hearing, because they are motivated by the reputational concerns of the managers.In contexts where there was less pressure on the government, it was not possible to identify strategies such as cases where the mayor had support in the Legislature.-Reputation management -Among the information regarding the management's actions, only the desirable information is publicized, omitting negative/ sensitive information and emphasizing positive information.
-Blaming others -Emphasis on negative information when related to the opposition, shifting blame.

Definition of the rules
-Restricting participation -The choice of date and time matters, because it can restrict the participation of the public, such as holding a hearing during work hours with untimely disclosure.
-Limiting the time available for proposals -The date of the hearing matters.If the period between the hearing and the deadline for sending a bill is short, it is unlikely that suggestions and contributions will be accepted (or even evaluated), due to the lack of time argument.

Form of presentation -Camouflage
-Some information cannot be avoided and passes through the informational filter, in which case presentation strategies are employed such as "give the bad news first and then the good news", selecting the sequence in which they are presented.
-Language -Some subjects can be treated in a more technical manner than others, depending on the willingness to convey them in an understandable form.
Directing attention -Choice of actors -The participation of political actors or those who are not part of the planning team to distract attention to other subjects avoiding a focus on the numbers.
-Political stage -Actors use the space for a "war of words" for political subjects.
Post-hearing/ Internal moment Manipulation of documents -Adjustments in documents subject to supervision -Given low participation, some measures are taken to meet the requirements of the supervision of the Court of Accounts asking employees and even people who were not there to sign the attendance list.
Little transparency -Absence of public disclosure of minutes -Even though the mandatory documents are sent to the supervising body, this is not a guarantee that the documents will be available to citizens.
-Restricting monitoring -The absence of transparency in the supervision and monitoring of follow-up decisions diminishes confidence in the process as well as its effectiveness.
Source: Elaborated by the authors.

CONCLUSIONS
This study concludes that public hearings, because they are considered synchronous spaces of accountability, raise reputational concerns among the presenters of accounts, which causes a felt accountability effect in which the anticipation of pressure in this forum favors strategies to avoid risks to one's reputation and accountability.In the specific case of hearings, this effect is facilitated by the government since the organization is totally centered in the hands of those responsible for presenting the accounts in that they define the hearing's rules and dynamics, whose context is favored by the absence of a legal definition of its scope.
Even though, a priori, it would be an important instrument of transparency, public hearings can be more advantageous for the government than society, because they have served to legitimize its actions, often simulating giving voice to citizens, even though they do not listen to them.In practice, hearings -and maybe other participative forums -have greater symbolic value as opposed to real vale within the budget process.
This study has some implications.First of all, in terms of the Court of Accounts, it could exercise greater control and supervision of the participation mechanisms, and it may not be fulfilling its role.Even in the analyzed state (São Paulo), in which the Court of Accounts maintains some type of supervision in this area, the actors perceive little coercion on its part, and they relate that they could send "a cake recipe" to the Court, because to them, the content is not analyzed and there have never been any type of demands or clarifications in this regard.The Court could make other types of additional requirements, such as recordings of hearings and direct clarifications/demands made to the municipalities.Second, the Legislature could contribute, since by knowing the mechanisms and strategies adopted in hearings, it could demand through local legislation, changes in their dynamics and a greater transparency of information and the scope of hearings before and after their realization, which would give greater support for citizens and the public to prepare and participate.For example, they could make the information that will be presented and discussed available beforehand and require a public report covering all of the requests received.Third, this paper has implications for the public at large, which could act in a more active and organized fashion to demand information from the government on a regular basis and not just during hearings, requiring the public disclosure of the questioning/proposals and follow-ups, which could also be performed with support from other control bodies, such as the Prosecutor's Office, which could supply support to get around the government's strategies.
This study also has implications for the accountability literature, given that in general the discussed forums do not consider the existence of two types, synchronous and asynchronous, which are subject to different effects in terms of the behavior of those presenting the government's accounts.
Finally, this research has implications for current legislation which has not defined the scope of public hearings and has not even delineated the rules for holding them.The current format gives the government great power and freedom to make decisions about what, how and when it presents information.Legislation could establish clear guidelines and times for their realization, together with broad and timely disclosure of their rules and content, minimum frequency, and period of anticipation, considering the maximum timeframe to send bills to planning hearings, as well as normalizing the procedures to be adopted in the post-hearing phase to guarantee the possibility of monitoring demands discussed in transparency portals, for example.This study has some limitations.First of all, to understand how the moments which are public function, this study has been limited to the perceptions and narratives of the respondents.Second, the selection of cases focused on a single state has methodological benefits by reducing the effects of different Courts of Accounts, but it also limits the results to the analyzed cases and similar ones.Finally, this study has analyzed just the side of those responsible for these hearings, without observing how citizens perceive them, which could be a limitation as well as an opportunity for future studies.
New approaches could consider how different roles and responsibilities can interfere in the realization of public hearings in municipalities, considering the perspective of society in terms of these hearings, and identifying whether these strategies are being perceived by the public.In addition, we can observe that in this country's current context, there has been a movement through which the Legislature has conquered significant portions of the budget in the form of direct expenditure allocations, through mandatory budget amendments (in the federal case, through Emenda Constitucional nº 86 de 17 de março de 2015), which is a relationship that has already been found between the proportion of favorable votes for Executive Branch bills and the value executed in these amendments (Graton, Bonacim, & Sakurai, 2020).This could be a mechanism to legitimize the government and legislature's interests to the detriment of society, which may suggest future studies of transparency mechanisms.
Ordered by transparency judged by the Federal Prosecutor's Office; (2) Dependence on revenues: revenues received through taxes ÷ total revenues.Source of Information: (a) Brazilian Institute of Geography and Statistics (2019); (b) Federal Prosecutor's Office (2016); (c) Institute of Applied Economic Research (2019); (d) Electoral Supreme FIGURE 1 DYNAMICS OF PUBLIC BUDGETARY HEARINGS