Unionism , the Decision-Making Process and Social Security Reform in Brazil

Despite the vast literature on social security reform there are relatively few studies that analyze the participation of representatives coming from the union movement in the decisionmaking process. This article aims to fill part of that gap in Brazilian academic production on unionism and public policy. In situations in which unionist representatives support the ruling party, does the union bloc tend to defend the specific interests of their base of social representation or to follow the guidance of the party coalition of which they are part? The study addresses this research problemby analyzing the participation of the union bloc in the decision-making process of a social security reform, Proposal of Amendment to the Constitution, Article 40 (PEC 40), during the first term of Luiz Inácio Lula da Silva's government. The main findings of the study confirm the thesis of party predominance. In an institutional political scenario in which party discipline prevails, unionist representatives and senators tend to follow the guidance of the party coalition even under contrary pressures from their electoral base. The research sheds light on the relations involved between the Executive and Legislative branches, in the process of changing public policy, in which a government considered to be allied imposes losses for specific sectors of the union base, notably the active and inactive public servants of the federal government, states and municipalities.

Unionism, the Decision-Making Process and Social Security Reform in Brazil (2016) 10 (2)  e0002 -2/27 n the first years of the redemocratization of the country, the extrapartisan components of the Brazilian political system were identified as being primarily responsible for the problems of the new democratic governments to sustain an agenda consistent with the economic, political and social reforms.
Social heterogeneity, federalism and the power of the governors would be the main variables to explain the chronic instability of the "coalition presidentialism" (ABRANCHES, 1988).
Nevertheless, judging by recent scientific production, even though the term 'coalition presidentialism' is still broadly used, including areas outside of academic circles, its relation to the original form has been vague; especially, regarding the need for forming extra-partisan coalitions in the organization of the government.Limongi (2006) observes that, even though from the descriptive point of view the political system may be defined as a coalition presidentilism, the thesis that the parties are incapable of structuring the necessary coalitions to sustain government policies is weak.The 'Gordian Knot' of coalition presidentialism would 1 In the same line of argument, Limongi (2006) states that, in this new institutional political context, there remains little chance for the minorities to influence the formulation The aim of this study is to investigate how the bloc of unionist representatives behaves when it is part of a partisan political coalition that proposes measures contrary to the interests of an important sector of its electoral base: specifically, those active and inactive public servants of the federal goverment, states and municipalities.
Following this brief introduction, the study is divided into four sections.
First, the legislative commissions are identified as the privileged locus for union action in the decision-making process of social security reform.Second, the nature of the social security reform introduced by Luiz Inácio Lula da Silva's government is discussed.Third, union participation in the decision-making process of social security reform is analyzed.Finally, the substantive conclusions of the article are summarized.
The main findings of this study confirm that unionism is still an important actor in the process of change of public policies, especially in social security reforms (BÉLAND, 2001) 3 .However, it is also shown that union demands are not expressed directly in the legislative debate on social security reform.Much to the contrary, a detachment between the interests of the union entities and the behavior of the unionist representatives is shown.In an institutional political of public policies, "to be part of the governing coalition or to elect the next President of the Republic" (LIMONGI, 2006, p. 41) .
2 It should be noted that this text does not address the so-called "reform of the reform".The Proposal of Amendment to the Constitution Nº 77, October 8, 2003(PEC 77/2003), also known as the PEC Paralela da Reforma da Previdência (PEC Parallel Pension Reform).
3 For recent developments in the relationship between unionism and public policy in Brazil, see the work of Santana and Braga (2009) and Ferraz (2014).
scenario in which party discipline predominates, unionist representatives and senators tend to follow the guidance of the party, even under pressures to the contrary from their electoral base.

The locus of the union minority
In institutional political contexts in which the President of the Republic concentrates legislative prerogatives, there is relative consensus in the literature affirming that the improvement of the work of the commissions is one of the ways in which the legislative branch is strengthened, facing the executive branch (GOMES, 2006;SANTOS, 2002).
The decentralization of the legislative process has, as one of its main objectives, endowing the decision-making process with greater functionality and dynamism.Furthermore, it seeks to offer responses to the multiplicity of topics and issues faced within the political system.From the organizational point of view, the legislative commissions contribute to the solution of coordination issues and to the endowment of greater stability and predictability of the process of framing public policies (ROCHA and BARBOSA, 2008;SANTOS and ALMEIDA, 2011).
From the political point of view, the decentralization of the decisionmaking process allows the minorities to increase their influence in the legislative branch (RICCI and LEMOS, 2004).Participation in commissions constitutes one of the most efficient strategies for minority groups to influence the political game, their main asset being the power to postpone or block the legislative process in matters to be discussed on the floor of the House.
In the case of Brazil, the composition of parliamentary commissions is formally defined by the President of the House.However, in practice, it is the party leaders who centralize the choice of the representatives and select the names for the President4 .Thus, the decision-making power is concentrated in the hands of the President of the House and the party leaders.Individually, the representatives have little ability to influence what will be voted on, how it will be voted on, and when it will be voted on.
Sidney Jard da Silva (2016) 10 (2) e0002 -5/27 Santos (2002) observes that two basic criteria are considered in the composition of the commissions: 1) party loyalty and 2) expertise.Party loyalty is related to the expectation that, within the commission, the representative will behave according to the guidance of the party in relation to the topic under consideration.Expertise is related to the quality of the information contained in the projects analyzed by the commissions, which may be better evaluated by representatives who act directly or indirectly in the area under consideration.
Thus, the political concerns with loyalty are balanced by the specialized knowledge of the policies considered5 .
In post-1988 Brazil, the majority have had various instruments to overcome possible political-institutional obstacles represented by the commissions.In the Brazilian institutional arrangement, the minority has no way to prevent matters contrary to its interests from being considered and voted on directly in the House.In other words, the system of commissions was not designed to veto (DINIZ, 1999;FIGUEIREDO and LIMONGI, 2001;PEREIRA and MUELLER, 2000).
Nevertheless, the minority may use the few institutional channels available to them, including the strategic location in the commissions to exploit any dissent among the majority and to seek to block or postpone the decisions to be made.As argued by Silveira and Silva (2014), the executive agenda may not be imposed unilaterally on the legislature.
Referring specifically to the object of this article, it has been shown in prior studies that the main opportunity for union participation in the decisionmaking process of social security reform takes place within the legislative commissions (JARD DA SILVA and CORTEZ, 2007;JARD DA SILVA and DINIZ, 2009).This is because, as previously observed, the system of commissions concentrates the main opportunities for the minorities to influence change in public policies.
It is precisely this last aspect of the Brazilian legislative process that will be explored in the following sections, emphasizing the participation of the union bloc in the legislative process of the Proposal of Amendment to the Constitution, Article 40, from 30 April, 2003(PEC 40/2003).
According to Figueiredo and Limongi (1998) and to Melo and Anastasia (2006), social security reforms are political proposals that impose immediate, concentrated costs and generate diffuse long-term benefits.Thus, they are defined as 'politically infeasible' by the academic literature 6 .
Without ignoring the full complexity of the political processes of the retrenchment of social policies, which might minimize these costs both in terms of their concentration as well as their contemporization (PIERSON, 1997), the present study is interested in the fact that this particular feature of social security reforms favors the actions of groups opposed to changes in the status quo, which seek to increase the visibility of unpopular measures and, consequently, the political costs of reform (BÉLAND, 2001;HIROI, 2008).
In summary, in the specific case of this study, it aims to verify the extent to which the unionist representatives exploited the particular characteristics of the Brazilian decision-making process, in defense of the interests of their electoral base, in the legislative debate on social security reform.More specifically, how the union bloc acted in pursuing Amendment 04/2003 to represent the demands of public servants and their representative bodies.

The nature of Brazilian social security reform
The closeness of the rules of Public Social Security System (RPPS -Regime Próprio de Previdência Social) to the precepts practiced under the General Social Security System (RGPS -Regime Geral de Previdência Social) was the main argument of the social security reform of Lula's government.It did not, therefore, deal with the creation of a single social security system for workers in the public and private sectors.Rather, it dealt with the closeness of the regulatory and actuarial principles of the system of public servants to the general system for workers in the private sector.
Sidney Jard da Silva The discussion of the unification of the social security system is not new and it dates back to discussions related to the Constitutional Revision of 1993.
However, although defenders are found in broad segments of society, the effective unification of the RGPS and of the RPPS never got onto the Brazilian social security reform agenda (JARD DA SILVA, 2007SILVA, , 2013)).
In fact, similar to his predecessor, President Fernando Henrique Cardoso (from now on, FHC) (1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002), the reform proposed by President Luiz Inácio Lula da Silva (2003Silva ( -2010) ) was limited to the criteria for eligibility and the rules for granting RPPS and RGPS benefits: It deals with moving in the direction of the convergence of rules between currently existing social security systems, applying to public servants, in any possible way, the requirements and criteria closest to those of workers in the private sector.With this approach, we seek to make Social Security more equitable, socially more just and financially and actuarially viable in the long term (BRASIL, 2003).
The first distortion singled out by the government was the lack of a direct relationship between contributions and benefits in the social security system for public servants.Their pensions had used the final salary of the effective position as the reference; while the calculation of the benefits for workers in the private sector corresponded to the arithmetic mean of their contributions, multiplied by a socialsecurity factor7 .
However, among the public servants and their representative bodies, the idea predominated that the social security system involved the recognition of a pró-labore facto working relationship.That is, the full pension was seen as "a de facto extension of working in the public sector and not because they contributed thereto" (GUERZONI, 1999, p. 59;PACHECO FILHO and WINCKLER, 2005).
The concept was questioned in the academic and political debate on the subject: In fact, this critical point largely dates back to the promulgation of the Federal Constitution because, prior to the institution of the Regime Jurídico Único dos Servidores Públicos Civis da União -RJU (Single Unionism, the Decision-Making Process and Social Security Reform in Brazil (2016) 10 (2)  e0002 -8/27 Legal System of Public Servants of the Union), as well as of the states and municipalities, the public workers who were already held in statutory status, contributed only to the pensions that were to be paid to their spouses and dependents (and not to their own pensions) (BRASIL, 2003, p.09).
The second distortion is in regard to the lack of a salary cap for public employees.While workers in the private sector were limited to the RGPS cap, in the public sector there was no salary cap for benefits.According to the government, this situation promoted a large inequity in the granting of social security benefits, not only between workers from the public and private sectors but among the public servants themselves working on different branches (executive, judicial and legislative).
The social security reform proposal of Lula's government also identified distortions in regard to the death benefit.In the public system, the benefit corresponded to 100% (one hundred per cent) of the salary of active workers or 100% of the benefit of inactive workers, with no maximum limit.For workers in the private sector, though, the same benefit was calculated by the mean of the contributions of the active workers or 100% of the benefits of the inactive workers, both limited to the RGPS ceiling.
In terms of social equity, the differences in the criteria for granting benefits offered by the RPPS and the RGPS were shown by the government as a redistribution of income from workers in the private sector to workers in the public sector: Under the current rules, a regressive situation can be seen according to which public servants who, notoriously, have a higher average income than that of workers in the private sector, end up being rewarded with the granting of benefits without the necessary contributions, an authentic distribution of income in reverse, wherein the poorest collaborate to sustain the benefits of the richest (BRASIL, 2003, p. 12).
In general terms, the social security reform of Lula's government may be classified into two large blocs: 1) measures for bringing the rules of the RPPS closer to the norms of the RGPS; and, 2) corrective measures for the inequalities of the RPPS.It is not possible to draw a firm line between these two groups of Sidney Jard da Silva (2016) 10 (2)  e0002 -9/27 measures, but the proposed classification allows clearer visualization of the meaning of each group of changes: Setting a ceiling on public pension benefits Setting a reduction on the benefits of public sector early retirement pension (5%) Changing how public pension benefits are adjusted Abolishing public sector proportional retirement pensions

Raising the ceiling of private pension benefits
Setting a contribution for pensioners (11%) Among the steps for bringing the rules of the systems of public servants closer to the precepts of the social security system for private sector workers, the following stand out: 1) change in the basis for calculating pensions; 2) change in the calculation of readjustment of benefits; 3) establishment of a ceiling for social security benefits; 4) raising the RGPS ceiling.
Among the specific adjustments to the RPPS rules, the following stand out: 1) to reduce the entire anticipated pension (5% of the benefit value for each year of retirement before the minimum age); 2) to set a ceiling for the pension benefit; 3) to end the early proportional pension; and 4) to set a contribution by public retirees.
The emphasis on the inequity between the public and private social The increased political cost of the break with an important sector of their base of union representation was felt by the directors of the CUT: Why was the Minister of Social Security not from the PMDB?If this was to be corrected, the burden should have been put on the shoulders of the PMDB, which was the largest party of the allied base and which was responsible for all the damage caused to social security: you will correct the damage that you did in the past!But the burden was put on the shoulders of a militant [Ricardo Berzoini], of a disciplined militant who defended the project in general, and who had to face the consequences.That is what I am saying: we are smart to attack, but sometimes we are not so smart to defend ourselves.Why does the Minister of Social Security need to be a union director?(CAVALCANTE, 2013, Interview).
However, Melo and Anastasia (2006) identify this "change in contextual character" (MELO and ANASTASIA, 2006, p. 305) as one of the main explanatory factors for the relative success of the social security reform of Lula's government vis-à-vis the social security reform of FHC's government.For these authors, "the change of places between situation and opposition" (MELO and ANASTASIA, 2006, p. 327) interfered in the distribution of the preferences and the resources among the actors involved in the reform debate.In this case, the conversion of the PT and of the PC do B (Partido Comunista do Brasil) to the precepts of social security reform, which they opposed during FHC's government, is particularly noteworthy.Also, according to the same authors, the change of position of the main leaders of PT and CUT in relation to social security reform would have isolated the public servants in the decision-making process of social security policy.
Lula, on the contrary, focused on how to reform the public social security, thus isolating public servants, who had already lost an old ally -the PT itself -and drawing national labor confederations to his side, in the fight against 'privileges' and in favor of equity.Hence while Cardoso united several attentive publics against his proposal, Lula sought to divide them (MELO and ANASTASIA, 2006, p. 11).
Effectively, once in the office of the Presidency of the Republic, Lula used the closeness of the rules of the social security systems as one of the main arguments in favor of reform, placing the 'privileges' of the RPPS against the 'rights' of the RGPS.In other words, he exploited the differences between the interests of the public and private sector workers that formerly, under the leadership of his own party (PT) and of his political arm in the union movement (CUT), together resisted the social security reform of FHC's government.

Unionism and the decision-making process
A few months before the election, facing the clearly worsening financial crisis of the country, then-candidate for the Presidency of the Republic Luiz Inácio Lula da Silva revealed, in a meeting of the Partido dos Trabalhadores (Workers' Party), his 'Letter to the Brazilian People' 9 .In that letter, the future president committed himself to not promoting sudden and unilateral changes, to respecting the contracts and external obligations, to preserving the budget surplus, to promoting fiscal balance and to controlling public finances (LULA DA SILVA, 2002).
Its popular appeal notwithstanding, the main targets of the letter were the national and international financial markets.In it, the structuring of a tenuous balance of the class coalitions involving sectors of the industrial bourgeoisie, the organized working class, the financial bourgeoisie and the Brazilian middle class was begun (SINGER, 2009;2012;2015)    The absolute majority of unionist representatives and senators integrated the parliamentary base of the government (see Table 2).There were 44 unionist representatives in the House of Representatives from PT, seven from PCdoB, one from PDT, one from PPS, one from PPB and one from PFL.In the federal Senate, all five senators were from the Partido dos Trabalhadores (Workers' Party).Adding up the union representatives (deputies and senators) from the PT, PC do B and PDT, it can seen that 95% of the union bloc belonged to the governing coalition13 .In the Committee on Constitution and Justice and Citizenship (CCJC), the governing coalition had 36 parliamentarians (63.16%) among the 57 that made up the commission.Of these, only one was a member of the union bloc, representative Luiz Carlos da Silva, Professor Luizinho (PT-SP).In the vote on the constitutionality of the matter, the project was approved by 44 representatives (77%) and rejected by 13 (23%) (see Table 03).This was a first sign that the executive could count on the opposition sectors to offset any dissent in its support base.In the Special Committee on Social Security (CESP) (see Table 3), the Executive was able to ensure greater discipline from its support base, while the opposition parties, PSDB and PFL, continued to be divided in relation to PEC 15 It should be observed that the PPB kept the same pattern of parliamentary action showed at the beginning of the legislative process of social security reform of the President Fenando Henrique Cardoso, that is, to position themselves formally as an independent party, but vote according to the orientation of the government (FIGUEIREDO and LIMONGI, 1998;JARD DA SILVA, 2007). 16It is worth remembering that in FHC's government the presidency and the rapporteur of the first CESP were given to two critics of the Executive's proposal, representatives Euler Ribeiro (PMDB-AM) and Jair Soares (PFL-RS), facilitating the obstruction of the matter by the opposition and forcing the Chamber President, Luis Eduardo Magalhães (PFL-BA), to dissolve the Commission (FIGUEIREDO and LIMONGI, 1998;MELO, 2002).The five CESP unionist representatives (four from PT and one from PC do B) followed the government's guidance and voted in favor of the project.Once again, in the governing coalition, the party coalition was ahead of the corporate interests of the union base.Even representative Jandira Feghali (PC do B-RJ), known as a fierce advocate of public servants, voted with the government17 .
During FHC's government, with the support of the public servants, the workers party representatives played an important role in exploiting the weaknesses of the governing coalition and postponing the legislative process of the social security reform (FIGUEIREDO and LIMONGI, 1998;JARD DA SILVA and CORTEZ, 2007).In Lula's government, however, the PT, in addition to favoring reform, had the support of former adversaries, supporters of deeper changes in the social security program of public servants, as was the case with the representatives of the PSDB and PFL.
This change in the institutional political context was also observed by important former leadership of the Workers Party: At the time that I was the leader of the opposition to Fernando Henrique, Fernando Henrique's leader was Romero Jucá, who became the leader of Lula's government.And the leader of Lula's government in the entire National Congress was Senator Roseana Sarney.... no wonder the PSDB bloc voted mostly with Lula's government in social security reform (HELENA, 2013, Interview).
In fact, the government not only had the disciplined support of its base, but it also had the opposition votes.Notably, it had the support of the representatives from PSDB and PFL from states where these parties were in power and that, therefore, had direct interest in extending the measures adopted in the federal employees' system to state social security systems (NAKAHODO and SAVOIA, 2008).
At this point, all state governors supported the social security reform of Lula's government, which would promote an improvement in public state funds regarding the expenses with retirees (FLEURY and ALVES, 2004).Thus, representatives linked to their respective state governments voted along with the federal government, even if they were from the opposition bloc (MELO and ANASTASIA, 2006).
Indeed, from the vote on the constitutionality of the matter in the House, until the second round vote in the Senate, the government counted several times on favorable votes from the opposition to compensate for unfavorable votes in the governing coalition.In other words, while government sectors voted with the opposition, opposition sectors voted with the government.This institutional-political context reduced considerably the power of unionism pressure on the legislative debate of social security reform.As we have seen, in the first procedural phase of PEC 40/2003, representatives from the leftist parties, who once defended the interests of public servants in the Congress, voted in favor of the project submitted by the Executive, along with representatives linked to center and right-wing parties, fierce advocates for social security reform.
In previous studies we demonstrated that the unions do not have institutional veto power on the decision-making process of social security reform but can, through the bloc of unionist representatives, influence the decisionmaking game of reform (JARD DA SILVA, 2007; JARD DA SILVA and CORTEZ, 2007).This strategy was used by Brazilian unions to exploit the lack of unity in the coalition bloc during the first term of Fernando Henrique Cardoso's government.With regard to the mathematics of the votes, considering the major voting of PEC 40/2003, Lula's government did not have any difficulties in getting the number of votes required for the approval of the proposal in the different instances of the legislative process.On average, over the course of the decisionmaking process, the government got 72% of the votes to approve the social security reform project, more than the minimum required (60%) for the approval of constitutional amendments.As occurred in the legislative committees, in the plenary of the House and the Senate, the government counted on the disciplined vote of the union bloc in favor of the social security reform project.On average, 80% of the unionist representatives voted in favor of the executive proposal.At this point, we must say that at the key moments of the voting, in the second round in both Houses, the loyalty of the union bloc overcame that of the allied base itself, in which it was included (see Table 04).Throughout the legislative process of social security reform, the government had, on average, 72.6% of the votes of the representatives in the plenary session of the House and the Senate.In the first and second rounds in the Senate, the government had its best (85.2%) and its most difficult victories (63.0%), respectively.The governing coalition held together in the voting, averaging 82.0% of votes in favor.The union bloc, in turn, was equally faithful to government guidance with favorable voting averaging 81.5%.
The expected resistance of unionist representatives, pressured by the public servant representative bodies and by social movements against the reform, did not take place.The party orientation weighed more than the corporate pressure on the decision of unionist representatives.In other words, between the specific interests of an important sector of Brazilian unionism and the orientation of the party leadership, the latter prevailed.
This detachment of unionist representatives in relation to their union base, was also noticed by CUT union leaders: Representatives from the union movement do not entirely follow the ideas of the union base.Much to the contrary, they are disciplined and their activities in the Congress will only and exclusively be guided and directed by the bloc's decisions […].They go up to a limit: they follow, strengthen, go to demonstrations, call the offices, make visits, make petitions; but, when the bloc leader organizes and gives the guidelines, then the action is of the party (DINIZ, 2013, Interview).
In the first-round vote on the social security reform, of the 50 unionist representatives present, only 04 voted against and 05 abstained.The few dissenting votes (against and abstentions) came mainly from representatives who had their professional careers linked to workers in the public sector (public servants and workers of state-owned enterprises): 03 physicians, 03 professors, 01 public servant, 01 telephone employee and 01 employee from the electrical sector.However, even in the broad category of 'public sector workers' the government was victorious in having the support of 17 of the 26 union representatives (65%).
The government reached a similar result in the second-round vote.The same representatives that had voted against or abstained in the first-round vote repeated their votes in the second round.So, the government once again had the majority of the votes (65%) from the public sector union bloc.
In the Senate, the situation encoutered by the executive was even calmer.
Only Senator Heloisa Helena (PT-AL, professor) opposed systematically the government's proposal (an absence and a vote against).The other unionist senators voted for the social security reform in the two rounds: Ana Julia Carepa (PT-PA, bank clerk), Fátima Cleide (PT-RO, teacher), Ideli Salvati (PT-SC, teacher) and Paulo Paim (PT-RS, steel worker).
It should be stressed that, once again, even among teachers, one of the most important professional categories of the public sector and most resistant to social security reform, the favorable vote of the unionist senators won: 02 against 01.On the other hand, among the senators from the professional categories linked to the private sector (banking and steelworkers), government loyalty was absolute.
At this point, it should be observed that part of the negotiation for the Based on such principles, we appeal to Your Excellency, in the capacity of guardian of the State and reviewer of the House of Representatives, to reject the idea of presenting a "PEC Paralela", at the initiative of the Senator Tião Viana (PT-AC), or any other attempt to destroy the solid structures of this Senate, historically constructed by democratic ideals (ASSOCIAÇÕES DE SERVIDORES PÚBLICOS, 2003, p. 13).
Despite resistance from the public servants and their representative bodies (associations and unions), in little more than eight months of legislative proceedings with the support of the union bloc and parliamentary sectors of the opposition, Lula's government approved the main points of the social security reform in Congress.
Regarding the controversy over the contribution from retirees, the executive agreed to increase the exemption limit from R$1,200.00 to R$1,440.00.
In relation to pensions, he agreed to decrease the reducer index from 50% to 30% for benefits in excess of R$2,400.00.Finally, the minimum amount of time in a position required for full retirement of public servants was reduced from ten to five years.
On the other hand, the goverment promoted deep modifications in the RPPS, including: 1) definition of the minimum requirements for obtaining full retirement; 2) end of the full pension and establishment of a R$2,400.00 ceiling for retirement benefits of new public servants; 3) maintenance of the parity between active workers and retirees only for those who obtained full retirement and differentiation of benefit raises for future workers; 4) end of proportional retirement and implementation of a reducer index of 05% a year for those who retire before the minimum age; 5) collection of a contribution from retirees and pensioners; 6) 30% discount in the amount of pensions; 7) definition of the highest remuneration of the Federal Supreme Court (STF) as salary cap for the federal public service (FLEURY and ALVES, 2004;MELO and ANASTASIA, 2006;NAKAHODO and SAVOIA, 2008).
The data analyzed in this study corroborate the thesis of party dominance even in the processing of legislative proposals that negatively affect interest groups with parliamentary representation (ARRETCHE, 2012;FIGUEIREDO and LIMONGI, 2001).The result of the vote and the content of the approved reform indicate that, even among the representatives and senators linked to professional categories of the public sector, party bloc orientation prevailed at the expense of the specific demands of the public servants.
In short, the social security reform of Lula's government fostered important changes in the system specific to the public servants.Many of these changes were a resumption of the reformist agenda of FHC's government, which suffered, at the time, great resistance from the opposition parties -PT, PDT and PC do B -and the main Brazilian unions -CUT, FS and CGT.
Nevertheless, the bloc of unionist representatives voted in a disciplined way in favor of the proposal of the executive, following the guidance of the party leadership.Paradoxically, they had former political opponents on their side -from PSDB and PFL -that were once unable to get the union support for their social security reform bill.

Conclusions
In situations in which unionist representatives support the ruling party, does the union bloc tend to defend the specific interests of its representation base or to follow the guidance of the party coalition of which it is part?This was the fundamental question raised at the beginning of this study.
Regarding this issue, there are at least two possible reactions: 1) the representatives elected with the support of the unions oppose initiatives that At its core, the issue that guided this study refers to the relations involved between the Executive and Legislative branches in a process of change of the public policies in which a coalition government imposes losses on specific sectors of the union base (MURILLO, 2001(MURILLO, , 2000(MURILLO, , 1997)), notably the active workers and retirees from the public sector.
The issue was addressed here based on the analysis of the participation of representatives from Central Única dos Trabalhadores (CUT), in the decisionmaking process of the social security reform led by the Workers' Party (PT), during the first term of the government of Luiz Inacio Lula da Silva.
The peculiar characteristics of the Brazilian political system defined the opportunities for the participation of interest groups in the legislative process of the Proposal of Constitutional Amendment 40, April 30, 2003(PEC 40/2003).In the case of union participation, the main opportunities arise at the time that the matter is analyzed by the legislative committees (JARD DA SILVA, 2007; JARD DA SILVA and CORTEZ, 2007).
In Brazil, the Executive, by party majority, has at its disposal several institutional political instruments to overcome the resistance of minorities within the legislative committees (FIGUEIREDO and LIMONGI, 1998;RICCI and LEMOS, 2004;SANTOS, 2002).This fact limits the range of strategies available to representatives to oppose the processes of change in public policies.
This article argues that the rules of the Brazilian coalition presidentialism and the loyalty of unionist representatives to party orientation have limited the power of public servants and their representative bodies to resist the decisionmaking process of social security reform.
Contrary to what occurred in Fernando Henrique Cardoso's mandate (1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002), in Luiz Inácio Lula da Silva's mandate (2003Silva's mandate ( -2010) )  The 'contextual change' was the main asset of Lula's government in the social security reform (MELO and ANASTASIA, 2006).The leftist parties that, when they were in the opposition, rejected the social security reform, once in government, came to support it.The opposition, in turn, whether for political cohesiveness or pragmatic interests regarding the financial stability of the subnational social security programs in states and municipalities where they ruled, also endorsed the federal government project (FLEURY and ALVES, 2004;NAKAHODO and SAVOIA, 2008).
In summary, the data presented in this article corroborate the thesis of party discipline, even in situations in which losses are imposed on organized interest groups, such as the case of the public servants.Between the pressure of the base against the reform and the pro-reform government orientation, the absolute majority of unionist representatives, including those originally from the public sector, followed the party guidance and voted with the government.
have been cut by the constitution of 1988.Contrary to what occurred in the Magna Carta of 1946, the Constitution of 1988 gave the President of the Republic the necessary instruments to govern through the majority party.These are not ad hoc negotiations with representatives of extra-partisan interests in the legislature, which support the political agenda of the executive in the national congress, but rather a coalition of the ruling parties.The concentration of legislative prerogatives in the executive branch made the partisan activity of the governing base to be the best strategy for representatives to influence the legislative process.Contrary to what happened in the period from 1946-1964, when representatives counted individually on institutional resources to favor their clientels, in the post-constituent (1988), "the collaboration with parliamentary parties is the best strategy for representatives to strengthen their bargaining power with the Chief Executive" (SANTOS, 2002, p. 246) 1 .
political scenario places an interesting issue before us regarding union participation in social security reform: in situations in which unionist representatives support the ruling party, does the union bloc tend to defend the specific interests of its representative base or to follow the guidance of the partisan coalition of which it is part?The present study approaches this research problem by analyzing the participation of unionist representatives in the legislative process of the Proposal of Amendment to the Constitution, Article 40, on April 30, 2003 (PEC 40/2003), from its introduction in the National Congress to its transformation into Constitutional Amendment no.41, on December 19, 2003 2 .
security systems shows the first large rupture in Lula's government with an important part of the constituency of the Partido dos Trabalhadores (PT) (Workers' Party) and the union base of the Central Única dos Trabalhadores (CUT) (Central Workers' Union).During the two terms of FHC's government, with the support of the PT and the CUT, the public servants were the main protagonists of the resistance to social security reform (JARD DA SILVA, 2007, 2013).In addition to Lula himself, the main leader of the opposition at the time, important authorities from the executive and legislative branches, originating from Unionism, the Decision-Making Process and Social Security Reform in Brazil (2016) 10 (2) e0002 -10/27 the union environment, had supported the resistance of the public servants to the social security reform of FHC's government.This was the case of the President of the House of Representatives, João Paulo Cunha (metallurgy); the Minister of Social Security, Ricardo Berzoini (banking); and, the Minister of Communication, Luiz Gushiken (banking) 8 .
10 .The promotion of the structural reforms was an integral part of the commitment to a "new social contract" for Brazil (LULA DA SILVA, 2002).The social security reform was the first of the reforms announced in the inaugural speech of President Luiz Inácio Lula da Silva on January 1, 2003.

9
As recorded byMachado (2007): "The 'Letter' was read by Lula during the opening of the 'National Conference: Government Program', when leaders and PT activists and representatives of society were gathered in a hotel in São Paulo, whether or not they were supporters of the candidate's proposals"(MACHADO, 2007, p. 97).10For an interesting critique of the argument of Lula's electoral realignment, see Boito Jr. (2013).
of the same year, accompanying the 27 governors of the federation and the 82 members of the newly-created Council of Economic and Social Development (CDES), Lula delivered his proposal for social security reform to the Congress.As Singer (2012), ex-spokesperson for the Presidency of the Republic in the first term of Lula's government, describes: (...) on April 30, 2003, the president descended the ramp from Planalto ahead of a large entourage to deliver personally the conservative Social Security reform bill to the Congress.Among other things, PEC (Proposta de Emenda à Constituição [Proposal of Amendment to the Constitution]) 40 ended the full pension for future public servants (SINGER, 2012, p. 10).The bill revived several points defeated in the social security reform proposal of Fernando Henrique Cardoso's government(1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002), many of which, it is important to point out, had been harshly criticized by the Partido Trabalhadores (PT) and by its main ally in the union movement: the Central Única dos Trabalhadores (CUT).Comparing the two reform bills, asMelo and Anastasia (2006) andNakahodo and Savoia (2008) point out, Lula achieved a greater degree of success in the decision-making process of social security reform than did FHC.Comparatively, Lula's government suffered few defeats in the legislative process of the Proposal of Amendement to the Constitution no.40 (PEC 40/2003), contrary to FHC's government which suffered several defeats in the approval of Proposal of Amendement to the Constitution no.21 (PAC 21/95) 11 .The first step to social security reform within the House of Representatives was the evaluation of the constitutionality of the matter.This was the first time, since the proclamation of the Constitution of 1988, that representatives from PT, including the project rapporteur himself, Representative Maurício Rands (PT-PE), were among those opposed to the social security rights of 11 PEC 21/95 was split into four distinct proposals: 30/1995, 31/1995, 32/1995 and 33/1995.The latter, with several changes, focused on social security reform.the publications of the Departamento Intersindical de Assessoria Parlamentar (DIAP) (Inter-Union Parliamentary Advisory Department) 12 .
Workers' Party; 4. Party of the Brazilian Democratic Movement; 5. Liberal Party; 6. Brazilian Labor Party; 7. Brazilian Socialist Party; 8. Democratic Labor Party; 9. Popular Socialist Party; 10.Communist Party of Brazil; 11. Green Party; 12. Liberal Front Party; 13.Party of Brazilian Social Democracy; 14. Party for the Reconstruction of National Order; 15.Brazilian Progressive Party.* Without party.The government was supported by the disciplined vote of 33 (91.7%) of its 36 members in the CCJC, with only 02 votes against in the PMDB and 01 in the PDT.The opposition, on the other hand, was divided: 05 representatives (50%) of the PFL and 05 (71%) of the PSDB followed the vote of the rapporteur.The independent representatives of the PPB also split with 01 vote in favor of the constitutionality of the matter and 01 against 14 .The DIAP registered 10 PT votes among the 44 considered opposed to the public servants in assessing the constitutionality of PEC No. 40/2003 in the CCJC.Conversely, among the 13 votes favorable to public servants, there were none from any PT representative.The only unionist representative in the commission, Prof. Luizinho (PT-SP), was loyal to party orientation and voted for the constitutionality (2016) 10 (2) e0002 -15/27 of the matter.It was the first indication of how the union bloc would behave during the legislative process of the reform.
40/2003.PPB representatives, despite declaring independence from the governmental ruling base, also voted in favor of the reform 15 .As regards the institutional political game, the ruling majority has been extremely judicious on the indication of the presidency and the rapporteur of the CESP.The first went to representative Roberto Brant (PFL-MG), former minister of social security in Fernando Henrique Cardoso's government; and the latter went to representative José Pimentel (PT-MG), loyal to the government.These were two political leaderships, one from the opposition and the other from the government, both strongly committed to the success of the reform 16 .In the CESP, the governing coalition had 23 representatives (62.16%) of the 37 in the Commission.The opposition had 11 (29.72%) and the PPB caucus, who declared themselves independent, had 03 (8.10%) representatives.The Commission still had the participation of five unionist representatives, four from PT and one from the PC do B. The absolute majority, 22 (95.65%) of the 23 representatives of the governing coalition, voted in favor of the social security reform project.Only 01 representative from the PDT voted against the proposal, but the government still had the vote of 3 PPB 'independent' representatives.The opposition parties (PFL, PSDB and Prona) also showed greater unity, only 03 (27.27%) of the 11 representatives of CESP voted in favor of the government proposal.
However, during Lula's government, this indirect participation in the decision-making process of reform was hampered due to the PT's change of position in relation to the reform of the social security program for public servants.The positioning of the union bloc within the governing coalition changed completely the reform game in relation to what had occurred in FHC's government, when the unionist representatives were the opposition.
favorable vote for PEC 40/2003 involved a commitment to submit a new proposal for a constitutional amendment.On the initiative of Senator Paulo Paim (PT-RS), member of the union bloc, the proposed PEC Paralela was presented to the Senate.This Proposal for Constitutional Amendment Nº 77 (PEC 77/2003) aimed to ensure that the original project did not suffer changes in the Review Committee Unionism, the Decision-Making Process and Social Security Reform in Brazil (2016) 10 (2) e0002 -20/27 and that it was again sent to the House of Representatives 18 .The actions of Senator Ideli Salvatti (PT-SC), also a member of the union bloc, and of Senator Tião Viana (PT-AC), rapporteur of the social security reform, were equally important for the negotiation between the Legislative and the Executive in relation to the terms of the new PEC.However, once again, while the unionist senators supported the initiative, the public servants rejected the proposal of 'the reform of the reform'.At that time, the main representative bodies of the public servants in the country signed the Letter to the Senators about the 'PEC Paralela', with harsh criticism of this initiative: their electoral stronghold; 2) unionist representatives follow party orientation to the detriment of the demands of their representation base.
the union bloc was part of the government support base in the Congress.Between opposing pressures from the public servants and party orientation for the approval of social security reform, unionist representatives followed the latter.Only a minority of representatives linked to the public sector unions resisted the government proposal, facing threats of retaliation from the party Sidney Jard da Silva (2016) 10 (2) e0002 -23/27 leadership.It is noteworthy that, as a result of the vote against the social security reform, Senator Luiza Helena and federal representatives Luciana Genro, João Barbosa (Babá) and João Fontes were expelled from the PT 19 .

Finally, it should
be observed that in the process of the approval of PEC 40/2003, the Senate proposed a 'reform of the reform', the so-called PEC Paralela of the Social Security Reform, which was under consideration in Congress for over a year and a half.However, again, while unionist representatives supported the negotiation of PEC 77/2003 with the Executive, the public servant representative bodies rejected the proposal of a new constitutional amendment on the same topic.The legislative process of the Proposal of Constitutional Amendment Nº 77, October 8, 2003, will be the subject of further studies on the topic.Judging by the results found in this study, one expects to identify the same pattern of behavior of the union bloc in relation to the demands of its social base and the guidelines of the party coalition of which it is part, i.e. the prevalence of the latter.Translated by ViaMundi Submitted in February 2015Accepted in August 2015

Table 01 .
Main measures for approaching the rules of the social security systems

Table 03 .
Voting behavior by party and legislative commissions -2003 Note: 1. Committee on Constitution and Justice; 2. Special Committee on Social Security; 3.

Table 04 .
Party discipline by bloc and chamber (%)