Is there a gender pay gap in the Brazilian judiciary?

Abstract

This study tests the hypothesis that there is a clear difference between the average salaries perceived by male judges and female judges in eight Brazilian courts of justice. Using propensity score matching, judges were paired according to position, length of service as a judge, and court. The differences in average salaries persist even after the matching, although to a lesser degree, which can be explained by gender mediators that operate by generating better opportunities for men than for women.

Keywords:
judiciary; gender studies; gender pay gap; inequalities

Resumo

O propósito deste estudo é testar a hipótese de que há clara diferença entre as remunerações médias percebidas por juízes e juízas de 8 tribunais de justiça brasileiros. Por meio da técnica de matching, realizou-se um pareamento entre juízes e juízas condicionado ao cargo, ao tempo de magistratura, ao mês e ao ano da remuneração, bem como ao tribunal. As diferenças nas médias remuneratórias persistem mesmo após o pareamento, o que pode ser explicado pelos mediadores de gênero, que operam gerando melhores oportunidades para homens em desfavor das mulheres.

Palavras-chave:
magistratura; estudos de gênero; diferenças salariais; desigualdades

Resumen

El objetivo de este estudio es probar la hipótesis de que existe una clara diferencia entre las remuneraciones promedio percibidas por los jueces de ocho tribunales de justicia brasileños. Mediante la técnica de matching se realizó un emparejamiento entre jueces y juezas condicionado al cargo, a la duración de la magistratura, al mes y año de remuneración y al tribunal. Las diferencias en los promedios salariales persisten incluso después del emparejamiento, lo que puede explicarse por los mediadores de género, que operan generando mejores oportunidades para los hombres en detrimento de las mujeres.

Palabras clave:
poder judicial; estudios de género; diferencias salariales; desigualdades

1. INTRODUCTION

The increase of women in the composition of the Brazilian judiciary in the last two decades has not been enough to change its sociodemographic profile, especially in the highest hierarchical positions in careers and in courts or commissions presidency. In addition, this increase coexists with several barriers and discrimination that prevent the progression of women and other minority groups in superior positions, resulting in unequal professional relationships (Duarte, Fernando, Gomes, & A. Oliveira, 2014; Fragale, Moreira, & Sciammarella, 2015; Kahwage & Severi, 2019; Rhode, 2001; Schultz & Shaw, 2003; Sommerlad, Webley, Duff, Muzio, & Tomlinson, 2010).

To analyze gender inequalities in the judiciary, the literature has used categories such as vertical segregation (or glass ceilings), horizontal segregation (or glass walls), concentration in areas or functions, and an ideology of professionalism (or the ideology of neutrality of professionalism). Glass ceilings (Bahillo, Saso, & Muñio, 2016; Bergallo, 2007; Bonet, 2014; Centro Regional de Derechos Humanos y Justicia de Género, 2019) refer to invisible barriers and discrimination that prevent women and other minority categories in the judiciary (blacks, indigenous people, people with disabilities, among others) to progress in their careers or to occupy positions of greater power, visibility or prestige. Glass walls involve the difficulties for women to work in certain specialization areas (Kahwage & Severi, 2019; Ribera, Miguel, & Pérez, 2009). The ideology of professionalism works as a barrier to the horizontal progression of women in the judiciary by making gender inequalities in the career invisible or natural (Bonelli, 2011; Gastiazoro, 2016).

This literature has left aside the topic of remuneration. Possibly, this is due to the assumption that, as it is a public career with a pay rate defined in tables and according to objective criteria, there would be no significant difference in the income earned by men and women. Contrary to what is perceived in other legal professions, in the Judiciary, once in a similar position, men and women with the same career length would have equal income. However, when we accept this premise, we fail to consider possible gender differences that result in unequal relations in access to the various benefits and pecuniary advantages arising mainly from the amounts referring to types of payments provided for in the remuneration structure of the bodies of the Brazilian Judiciary, such as bonuses, allowances, representation allowances, aids, and other additional values that may apply.

Therefore, it does not seem unlikely that there will be similar effects on access to remuneration benefits from the Brazilian Courts of Justice. Considering that each court establishes its own remuneration rules, it is possible to assume that, when applying the criteria that result in the payment of benefits to male and female judges, subtle mechanisms based on gender result in more significant advantages for men and disadvantages for women in the paychecks of members of such careers.

Our research sought to investigate differences in the remuneration of judges in state courts of law. The hypothesis of inequality based on gender was developed based on studies on gender in the judiciary and on the analysis of the models for organizing salaries in the Brazilian judiciary, which often result from the sum of a base salary and the values obtained as benefits.

2. GENDER INEQUALITIES IN THE JUDICIARY CAREER AND THE REMUNERATION PROBLEM

According to data made available by the National Council of Justice (CNJ) regarding the sociodemographic profile of the Brazilian judiciary, the increase in women¹ and black people, especially in the last two decades, has not yet been sufficient to change the profile of its members, which is still composed mostly of white men, married or in a stable union, coming from middle and upper social classes, with children, 46 years in average, and Catholics (CNJ, 2018).

The arrival of women and other minority groups to the Brazilian Judiciary took place in the 1990s, reflecting the institutions’ adaptations to the moment of democratizing transformations brought about by the Federal Constitution of 1988 (Sadek, 2006). The expectation at the beginning of the 2000s, nourished by several professional bodies in the legal field, which also saw a significant increase in women in legal and law courses, was that the diversification of the body of the judiciary and their ascension to the final positions would be a matter of time (Bonelli & F. L. Oliveira, 2020).

The field of studies on the feminization of the judiciary (and other legal professions) also began to gain strength in the country in the same period, in dialogue with works from other countries on the same topic, with an interest in monitoring possible changes in the models of jurisdictional provision and functionalism of the Judiciary derived from changes in the sociodemographic profile of its staff (Bercholc, 2015; Hunter, 2015; Kahwage & Severi, 2019; Menkel-Meadow, 2013). This expectation, however, has not yet been met. The increase in women and other minority groups has been slow and segregated.

Vertical segregation has been analyzed in several studies as a result of the interaction between family life/care tasks and professional performance (Bonelli, 2011, 2016; Fragale et al., 2015; Junqueira, 1988; Marques, 2014). One of the most reiterated elements in such analyzes as responsible for disproportionately affecting the chances and opportunities for women to rise in the judicial career refers to the difficulties in balancing work and family life, especially when these women are responsible for child care or other family members, even when they have other people support (maids and grandparents, for example).

Unlike the official rhetoric that claims that progression is regulated by the criterion of seniority, women, especially when they have children or are primarily responsible for family tasks related to care, find it more difficult to progress in the judiciary career, since many of the opportunities require, for example, spatial mobility between different regions or districts (Marques, 2014). In order to occupy spaces of power, they often need to demonstrate notorious merit, according to masculinized standards of productivity and professional posture (Gastiazoro, 2010; Kahwage, 2017; Sommerlad et al., 2010). Likewise, they need to prove the ability to manage the time dedicated to the role, especially when responsible for care tasks in private life (Fragale et al., 2015).

The need for harmonization between career and family life and care tasks has also been used as an explanatory key for the concentration of women in positions in certain areas or functions in research carried out in Brazil and several other countries. In the analysis made by Bergallo (2007) on the Argentine judiciary, problems with the time to be spent on care tasks and the permanent interaction between work demands at home or family demands at work lead women to opt for career positions that have flexible working hours and guarantee them family stability, which sometimes results in the abandonment of aspirations for ascension (Bahillo et al., 2016; Gastiazoro, 2008, 2016; Schultz & Shaw, 2003).

Bonelli (2010, 2011, 2016) has analyzed the inequalities in the professional progression of the judiciary, articulating gender with the ideology of professionalism as neutrality. According to the author, the construction of the identity of male and female judges linked to an ideology of professionalism such as neutrality was how the traditional forces of the Judiciary’s summits dealt with the expansion in recruitment and the heterogeneity of its composition, which began in the 1990s. With the risk, as a result, of the career losing its prestige and social power. The success of a few women in the entrance exams was converted into evidence that the selection was carried out based on merit, based only on the performance of the most capable people. The issue of gender inequality would then naturally be overcome over time.

In professional practice, neutrality is the parameter for evaluating the good performance of the judiciary - women or men. However, according to Bonelli (2010), neutrality involves less determined technical knowledge and more a form of identity construction of professional practice that generally takes the man and the characteristics attributed to the masculine - white and heterosexual - as points of reference for its construction. In other words, neutrality is a clearly masculine professional ethos - also white, socially favored, and heterosexual. For women, blacks, indigenous people, or those who do not belong to the predominant sociodemographic profile in the judiciary, the chances of advancement, of assuming some prestige or power space in the career, or even of not being subjected to constraints among peers in the exercise of their functions, are related to the varying levels of demonstration of adherence to this dominant frame of reference.

The meaning Bonelli (2010) uses to the term “neutrality” refers to the common sense current in law, or what Hunter (2015) calls judicial ideology, linked to the expectation that judges are capable of putting aside their worldview, beliefs, and personal values in the analysis of the cases and of not committing to the nuances of concrete cases and the subjective aspects of the parties in the decision-making process. This common sense is not only expressed in how they judge but also by a certain professional identity. It is not uncommon for neutrality to be confused with impartiality, which is a duty of the judge arising from the principle that the judge should be the one adequated to decide over the presented case, which expresses the requirement of equidistant action to the parties in the judicial process. If impartiality is a legal duty, neutrality is an ideology, which reinforces the perception that any type of difference made explicit by those who exercise the judiciary (gender, ethnic-racial differences, etc.) is questioned in terms of their qualifications and competence, since who entered the Judiciary under the condition of acting based on neutrality.

Bonelli and F. L. Oliveira (2020, p. 148) discuss the idea of neutrality as an “implicit bias towards women’s work, posing obstacles to their career path” while generating more opportunities for socially favored white men. This bias defines the unequal distribution of privileges and disadvantages in the career in order to favor the path of white and socially favored men to spaces of greater power and prestige while creating difficulties for women, who need to adjust permanently. From the data analyzed by the authors on composition and progression, gender inequalities in the judiciary do not seem to be decreasing over time. They still appear in higher percentages in the starting positions of their careers. The authors also argue that the recent access of some women to positions of direction or presidency in the courts has taken place at a time when such positions are losing “power, resources and autonomy” since they are more regulated by parameters and goals demanded by the CNJ.

The issue of differences in payment was rarely addressed in such studies, as there is a socially shared assumption that the rules that define it in public careers would be more “objective” or less susceptible to some kind of implicit gender bias. When there are differences in remuneration, they would be due to positions or functions held and time in the career. Until 2017, it would be tough to test this hypothesis, given that we did not have access to paycheck data from the Brazilian judiciary.

Private law is one of the legal professions where gender pay gaps are often analyzed. Schultz (2003), considering data from different countries, states that they occur in several places, ranging between 10 and 35%, and these differences are only partially justified by criteria such as specialization, length of the profession, and professional experience. They can also be attributed to gender discrimination. One of the arguments that the author identified in the field of justifications with the category for the salary differences between female and male lawyers is that women do not need, with their salaries, to guarantee the support of their families, but men do.

Outside of legal careers, but still talking about public services, there is a set of studies that accumulate evidence on vertical segregation and gender-based payment inequalities in occupations of the three government branches (Lopez & Guedes, 2019). Although access to public positions and careers is through competitive examinations and women constitute an expressive majority in public administration, in 2017, female public servants obtained gross earnings, on average, 24.2% lower than men (Lopez & Guedes, 2019). Even when we consider equivalent positions, the income earned by them is 16.9% lower in relation to the income earned by men (Vaz, 2018). The presumption, therefore, of isonomy in the remuneration of men and women in the public sector does not seem to be confirmed. As for vertical segregation, for example, in the federal public administration, women are less present in higher positions or decision-making levels associated with the job. In general, positions linked to management, advice, and leadership are filled by appointment and are temporary conditions that favor discrimination against women (Vaz, 2010).

These results also suggest a strong heterogeneity in the remuneration practices observed in the public sector, with different factors influencing the formation of wages in each power and sphere of government. The segment with the lowest remuneration is the municipal executive, while the best paid is the federal judiciary. These disparities are not explained only by the observable characteristics of the people who occupy each position (education, professional experience, length of career, etc.), but also by factors such as bargaining power, organizational and representational capacity, and proximity to the central decision-making centers of the State of each category or segment. Financial and administrative autonomy, as well as the type of external and internal control of public finances of each branch, is also a factor that influences the way salaries are set in the public sector (Nakamura & Vaz, 2020).

In the judiciary case, the sum of autonomy to set their own salaries, the great power of negotiation with the Executive (Carvalho & Leitão, 2013), the strategic role with the State and the low external control of their expenses with salaries can explain its position among the highest paying careers. In terms of gender inequality, the judiciary appears, even among the other public careers in the justice system, as the one in which male domination is more evident, that is, the possibilities of ascension and distribution of internal power that the legal professions allow their members are strongly affected by gender-based differences (Almeida, 2010).

3. HOW SALARIES ARE DEFINED IN THE JUDICIARY BRANCH

Transparency regarding the amounts of payments made to judges in the Judiciary is an old discussion. It was only after CNJ (National Justice Council) Resolution No. 215/2015, which regulated the Access to Information Law (Law No. 12,527, November 18, 2011), that the courts began to make data on remuneration available on their websites. Directive No. 63/2017 of the CNJ determined that all Brazilian courts send the CNJ the data related to payments made to members of the judiciary. The information sent by the courts is made available by the CNJ through a single system. Through this system, it is possible to identify the detailed payments, in order to know, for example, how much each judge receives monthly in terms of salaries, aids, compensation, and benefits.

The remuneration of each member of the judiciary can vary greatly - firstly, according to the position in each of the bodies that make up the Judiciary (higher courts, federal, electoral, military, labor, state, and Federal District courts). Each career is made up of hierarchical positions, from lower to higher. Mobility among them (promotion) generally occurs respecting the criterion of seniority among the members.

There is a basic amount (salary) fixed by a specific law for each body and position, respecting the limits imposed by the Federal Constitution (art. 37, item XI), by the Organic Law of the Judiciary (Supplementary Law No. 35, of March 14, 1979), and CNJ resolutions on the subject, especially CNJ Resolution No. 13/2006. One of these limits is the so-called remuneration ceiling, which refers to the prohibition of remuneration of public service personnel higher than the monthly allowance paid to ministers of the Federal Supreme Court (STF). This limit applies only to the amount regularly paid as salary, excluding fringe benefits. Each court is competent to define the types and reference values of the benefits to be granted, considering the role provided for in art. 65 of the Organic Law of the Judiciary: subsistence allowance (moving and transportation allowance, housing allowance, health allowance, daycare allowance, childbirth allowance, etc.), family allowance, per diem, representation, bonus for providing services to the Electoral Justice, bonus for providing services to the Labor Court, bonus for each five-year period of service, teaching bonus for classes given in an official preparation course for the Judiciary or in an Official School for the Improvement of Judges, in addition to bonuses for the effective exercise in a difficult district (based on its location or resources).

This list of benefits is often categorized in the paychecks of judges as follows: personal fringe benefits, which correspond to remuneration paid for the length of service or position exercised within the court; pecuniary aids (or indemnities), corresponding to asset restoration and comprising food allowance, transport allowance, pre-school allowance, health allowance, birth allowance, housing allowance, etc.; occasional benefits, which include payment of rights such as vacations or occasional services, such as replacement; and rewards paid for the exercise of temporary functions (such as presidencies or leadership) or as assistance due to personal expenses resulting from the performance of standard services under atypical conditions.

The list of personal benefits is limited, and it is up to each entity of the federation to establish its own. The pecuniary aids - additional and bonus - are not defined in a technical and systematized way in the legislation, which causes imprecision in the definition of their concepts, but can be identified because they are compensation for common or special services performed by such judges (Meirelles, 2016).

District careers for the judiciary are governed by the organic law of each of the states and the Federal District. Most often in these careers, judges are initially sworn in as substitute judges. When they become permanent, they can continue in promotions organized by the courts in initial, intermediate, and final positions. In addition to the chamber, judges can exercise certain public functions of leadership, advice, or direction, remunerated, normally, through pecuniary additions to the employee’s standard, under the most varied titles, such as gratification. Each court also stipulates other pecuniary advantages, such as additional benefits - for function or length of service - and allowances - such as transportation allowance, housing allowance, daycare allowance, and suit allowance. Thus, in addition to the salaries², which have values fixed by law, district judges can receive fringe benefits, with variable values. The remuneration corresponds to the salary value plus benefits.

Fringe benefits are regulated by the principle of impersonality³, which would, in theory, reduce the possibility of differences in remuneration by gender or other social markers. Why imagine that there could be gender differences in the salaries paid to male and female judges? The unequal access to benefits could not occur based on the nature of the judicial or administrative activities to be accumulated since all activities that generate the payment of benefits can be performed by male and female judges.

The interaction between public activities and private life can also result in gender pay inequalities. The progression in the career of judges occurs by merit and seniority (art. 93, III CF/1988). The merit requirement must observe objective criteria and have to be provided for in the Regulations of the Courts of Justice⁴. In general, merit is linked to the level of efficiency (productivity) and participation in internal training. The availability of female judges responsible for the economy of care to attend courses or meet specific productivity demands, for example, is lower than that of male judges who are free from care tasks in the private sphere. As stated at the beginning, the criteria of anteriority still favor the historically larger gender group in the justice system: the masculine.

In some state courts, such as the TJSP (Sao Paulo Court of Justice), seniority is considered as merit, so the only criterion that has been in force for career advancement is this. Even so, as progression often results in moving to other municipalities or regions in which they will act, in practice, many judges who meet the seniority criterion end up delaying promotion, waiting for certain districts, or deciding to remain in state districts - which, as a rule, is the first position in hierarchy - trying not to affect the working conditions of their partners or the schooling of their sons and daughters. Thus, they give up career progression, even fulfilling the seniority criterion, in order to better harmonize professional and personal life (Marques, 2014).

4. METHODOLOGY

The research seeks to test the hypothesis that there is a clear⁵ difference between the average remuneration received by judges in eight Brazilian State Courts of Justice, maintaining control over the time of service in the judiciary and the position held⁶.

The initial proposal was to include all state courts in the research. However, it was not possible to obtain data on the length of practice in the judiciary of all Brazilian state courts - only eight -, implying the limitation of analyzed units. In addition, courts adopt different policies for career advancement. There are courts, such as the TJSP, that opt for a progression according to “entrance” and position, so that a judge joins as a substitute judge, going through the initial and intermediary entrances, or through a special entrance. Even at the second entrance, there are female judges and auxiliary judges. In turn, there are courts that adopt only two progressions: trial judge or justice.

Initially, data were collected from twelve courts: TJBA, TJES, TJGO, TJMG, TJMS, TJPB, TJPE, TJRJ, TJRN, TJSE, TJSP, and TJAL (Bahia, Espirito Santo, Goias, Minas Gerais, Mato Grosso do Sul, Paraiba, Pernambuco, Rio de Janeiro, Rio Grande do Norte, Sergipe, Sao Paulo, and Alagoas). However, four were excluded from the analysis - TJRJ, TJMS, TJAL, and TJGO (Rio de Janeiro, Mato Grosso do Sul, Alagoas, and Goias) - because they do not adopt the same criteria for counting the time spent in the judiciary as the others, so it was not possible to compare them. The analysis was carried out on 92,779 observations on the additional remuneration of 3,499 judges, between the months of September 2017 and December 2020.

Our analyses, delimiting only the courts with similar progression criteria, did not change the estimators for the variable of interest, which is sex⁷, so we prefer to keep a simplified classification: substitute judge, trial judge, justice. Another cut concerns the dependent variable or the response variable, that is, remuneration. Judges perceive identical values for the subsidy, so we prefer to exclude it from the analysis. Our focus was on fringe benefits, namely personal rights, eventual rights, and indemnities, which together we call additional.

As explanatory variables, the position held, the court, and the length of time on the bench are included. The court was included because it ends up absorbing time-invariant factors that are not correlated with the others, such as court remuneration policy, allocative budgetary capacity, and gender-affirming policies.

Among the most used quantitative methods to compare the average salary between two groups, two stand out in the literature. One is the more traditional Blinder-Oaxaca decomposition (Ronald, 1973). The other is matching, which has been increasingly used to compare average salaries between men and women, especially based on the work of Hugo Ñopo (2008) and Markus Frolich (2007).

The Oaxaca-Blinder method seeks to determine the portion of the pay gap between men and women that can be attributed to discrimination and other factors, such as preferences, skills, or gender mediators, that escape the discriminatory aspects. For this purpose, two linear regressions are estimated: one containing only women’s salaries and the other containing men’s salaries. Based on the results of the two regressions, we calculate how much of the average difference is at the level of explanatory variables, the explained difference, and how much is at the level of coefficients, the unexplained difference, which can be attributed to discrimination (Hlavac, 2014).

The limitation of this method can overestimate the importance of discrimination when explanatory variables that affect remuneration are not properly controlled, such as the omission of variables or endogeneity (Djurdjevic & Radyakin, 2007).

In view of the limitations of Oaxaca-Blinder, Ñopo (2008) and Frolich (2007) proposed matching to compare the average salary that men and women receive. It is a non-parametric technique whose objective is to find, in a control group (male judges), members with similar characteristics to those of the treatment group (Silva, 2018), or to balance the distribution of covariates in the control and treatment groups (female judges).

For this study, we used coarsened exact matching (CEM), a matching method of the monotonic imbalance bounding (MIB) type, which means that the balance between the treatment and control groups is determined by the user ex-ante, instead of checking ex post facto, requiring new estimations to find the balance (Iacus, King, & Porro, 2012). Unlike propensity score matching (PSM), CEM allows the user to constrain the degree of model dependence and the error estimator of the mean treatment effect.

In summary, CEM works by stratifying the covariates, categorizing the continuous ones and/or regrouping the categories in order to reduce the granularity of the data. Then, exact matching is applied to keep only the observations whose strata contain at least 1 treatment (female judge) and 1 control (male judge). In strata where there is much more than 1 control for each treatment or vice versa, weights are assigned to indicate the proportion between treatment and control in each stratum.

In the present study, five-year strata were established for the time spent in the judiciary, except for the first strata, which was two years, as it corresponds to the initial probationary stage, to which all incoming judges are submitted.

5. RESULTS AND DISCUSSION

At the time of data collection, it was possible to obtain information from nine courts corresponding to the years 2017 to 2019.

The table below contains data on the average salary of judges by court and by position:

TABLE 1
AVERAGE SALARY BY COURT AND BY POSITION

Position	Sex	TJBA	TJES	TJMG	TJPB	TJPE	TJRN	TJSE	TJSP
Substitute	Female	22.349 (n = 288)	18.925 (n = 159)	16.724 (n = 151)	8.691 (n = 34)	12.168 (n = 472)	15.210 (n = 99)	11.364 (n = 60)	23.320 (n = 967)
	Male	22.679 (n = 392)	18.591 (n = 366)	15.742 (n = 494)	9.392 (n = 50)	14.714 (n = 369)	15.952 (n = 143)	11.361 (n = 288)	26.320 (n = 2211)
Trial Judge	Female	18.707 (n = 4415)	21.369 (n = 2491)	23.057 (n = 1616)	7.345 (n = 1091)	13.651 (n = 1666)	14.026 (n = 2122)	21.788 (n = 690)	18.965 (n = 19275)
	Male	18.913 (n = 6620)	22.619 (n = 4656)	25.643 (n = 3667)	7.553 (n = 1990)	15.288 (n = 2932)	14.764 (n = 2490)	19.061 (n = 719)	21.675 (n = 27650)
Justice	Female	27.954 (n = 682)	33.649 (n = 78)			1.068 (n = 1)	3.749 (n = 6)
	Male	26.374 (n = 735)	31.815 (n = 642)			1.068 (n = 1)	3.068 (n = 1)

• Source: Elaborated by the authors.

From Table 1, it appears that female judges are not always paid less than male judges. For example, in Bahia (TJBA) and Espírito Santo (TJES), female justices have a higher average salary than male judges. This difference is sometimes apparently not significant in some courts, such as the one in Paraíba (TJPB).

TABLE 2
AVERAGE LENGTH OF SERVICE IN MONTHS PER COURT AND PER POSITION

position	sex	TJBA	TJES	TJMG	TJPB	TJPE	TJRN	TJSE	TJSP
Substitute	Female	404 (n = 288)	46 (n = 159)	35 (n = 151)	13 (n = 34)	121 (n = 472)	25 (n = 99)	54 (n = 60)	198 (n = 967)
	Male	404 (n = 392)	43 (n = 366)	41 (n = 494)	12 (n = 50)	111 (n = 369)	29 (n = 143)	55 (n = 288)	240 (n = 2211)
Trial Judge	Female	388 (n = 4415)	215 (n = 2491)	172 (n = 1616)	173 (n = 1091)	164 (n = 1666)	228 (n = 2122)	154 (n = 690)	193 (n = 19275)
	Male	384 (n = 6620)	199 (n = 4656)	200 (n = 3667)	220 (n = 1990)	187 (n = 2932)	235 (n = 2490)	142 (n = 719)	204 (n = 27650)
Justice	Female	440 (n = 682)	363 (n = 78)			158 (n = 1)	229 (n = 6)
	Male	440 (n = 735)	333 (n = 642)			160 (n = 1)	274 (n = 1)

• Source: Elaborated by the authors.

The average time spent on the bench does not seem consistent in the TJBA(Bahia) and the TJSP (Sao Paulo). In the others, there is an indication that the positions correspond to the time in the bench. However, this is not necessarily associated with the position since merit has an important weight.

Before matching through CEM, on an unbalanced basis, it was found that judges receive 12% less than judges - which is equivalent to R$ 2,025.00. After balancing, this percentage dropped to 8.4%, or R$1,218.00.

The variable length of service was grouped in intervals of five years to perform the pairing, except for the first, which was two years. That is because of the two years probationary period mentioned.

Following the pairing, 308 observations corresponding to male judges and 122 corresponding to female judges were removed. The balanced base was submitted to simple linear regression using the weights generated by the CEM. Likewise, the standard error was corrected using clustered standard errors per judge. Below, is the love plot with the covariance before and after the match.

GRAPHIC 1
DIFFERENCE IN NORMALIZED AVERAGE BETWEEN THE BALANCED BASIS AND THE UNBALANCED BASIS

The linear regression result indicates that there is a significant difference between men and women (b = -0.084, t = -5.37, p < .01) (log) or (-1,183, t = -6, p < .01) in reais (R$).

Since they are both public servants with similar positions and time on the bench, it cannot be said that there is systematic pay discrimination to the detriment of women, but rather that these differences stem to a greater extent from gender mediators, such as availability, on the part of men, to take on internal positions that require more hours of work or to carry out training courses at times that are not favorable for women.

This difference sheds light, for example, on the discussion about the type of activity to be performed in functions for which bonuses are paid. The expectation is that activities are carried out in conjunction with regular career work, that people have time availability or locomotion conditions (to travel or attend meetings), which are often incompatible with reconciling life tasks most common to women.

In dialogue with the literature review carried out here, we can say that women generally do not occupy career positions that generate such remuneration. When elections among career members define access to these roles, it is more common that they are not elected by their peers - most of whom are men. In addition, many refuse to participate in electoral processes or to accept positions that require them to work beyond their regular working hours. For them, special roles may have a negative impact on how they balance their professional duties with their roles as mother, wife, and homemaker.

Therefore applying the rules that define the payment of personal benefits ends up interacting with the social norms of gender that regulate the concrete conditions for men and women to obtain them. According to the gender literature and the legal professions, these rules are part of the so-called glass ceilings - so that, in this case, women have the same opportunities for remuneration.

6. CONCLUSION

Gender inequalities in the remuneration of legal professions are most often addressed in private legal positions. In the judiciary, this issue is still little explored, either because of the assumption that the legal regime of public law would prevent the reproduction of discrimination that would result in this type of inequality, or because, up to three years ago, research on the subject, in the model proposed here, would find it difficult to collect data on the remuneration of the Brazilian judiciary.

We seek to test the hypothesis of the gender pay gap in the judiciary based on findings from the literature on gender and legal professions and on recent studies that identified gender pay gaps in the various Brazilian public services. With the help of CEM, we carried out a pairing between male judges and female judges conditioned to the position, the time served, and the court. Differences in the average salaries persist even after pairing, although to a lesser extent, which can be explained by gender mediators, who operate by generating better opportunities for men at the expense of women.

The results of our research corroborate the qualitative analyzes carried out on gender inequalities in the Brazilian judiciary, which appoint to an unequal distribution of privileges and disadvantages in the career in terms of gender, in order to favor the path of white and socially favored men, creating obstacles for women in accessing positions of greater power or resources. We already had cumulated evidence on the low permeability of female participation in the judiciary, especially in the highest positions in the career. Now, the present research pointed to unequal access in terms of gender to the remuneration items related to benefits. It is worth noting here that these items are the ones in which the courts have more autonomy to define the amount paid and the rules.

In terms of future research, disaggregated and qualitative studies could further investigate the functioning of the rules relating to the remuneration structure per court, considering the combination of gender and other factors of discrimination that result in inequalities, such as race and ethnicity, in the participation in bonuses and remuneration benefits.

ACKNOWLEDGEMENTS

The authors are grateful for the support of the Foundation for the Development of Teaching and Research in Law (FADEP).

APPENDIX

TABLE 1A
BASE SUMMARY BEFORE PAIRING

term	means_treated	means_control	std_mean_diff	var_ratio	e_cdf_mean	e_cdf_max	std_pair_dist
time_year8	18,355	18,961	-0,065	0,87	0,03	0,077
month_year9	17854,737	17849,45	0,02	1,016	0,006	0,01
Substitute position	0,061	0,076	-0,063		0,015	0,015
Trial Judge position	0,918	0,899	0,067		0,018	0,018
Justice position	0,021	0,024	-0,023		0,003	0,003
TJBA	0,148	0,137	0,03		0,011	0,011
TJES	0,075	0,1	-0,096		0,025	0,025
TJMG	0,049	0,074	-0,117		0,025	0,025
TJPB	0,031	0,036	-0,03		0,005	0,005
TJPE	0,059	0,059	0,001		0	0
TJRN	0,061	0,047	0,061		0,015	0,015
TJSE	0,021	0,018	0,02		0,003	0,003
TJSP	0,557	0,529	0,055		0,027	0,027

• Source: Elaborated by the authors.

TABLE 2A
BASE SUMMARY BEFORE PAIRING

term	means_treated	means_control	std_mean_diff	var_ratio	e_cdf_mean	e_cdf_max	std_pair_dist
time_year	18,378	18,507	-0,014	0,978	0,004	0,015	0,208
month_year	17854,813	17854,813	0	1	0	0	0
Substitute position	0,059	0,059	0		0	0	0
Trial Judge position	0,92	0,92	0		0	0	0
Justice position	0,021	0,021	0		0	0	0
TJBA	0,149	0,149	0		0	0	0
TJES	0,075	0,075	0		0	0	0
TJMG	0,049	0,049	0		0	0	0
TJPB	0,031	0,031	0		0	0	0
TJPE	0,057	0,057	0		0	0	0
TJRN	0,061	0,061	0		0	0	0
TJSE	0,02	0,02	0		0	0	0
TJSP	0,559	0,559	0		0	0	0

• Source: Elaborated by the authors.

REFERÊNCIAS

• Almeida, F. (2010). A nobreza togada: as elites jurídicas e a política da justiça no Brasil (Tese de Doutorado). Universidade de São Paulo, São Paulo, SP.
• Bahillo, C. G; Saso, C. E; & Muñio, C. M. (2016). Feminización de la judicatura española. Convergencia - Revista de Ciencias Sociales, 23(70), 199-226.
• Baltagi, B. (2008). Econometric analysis of panel data Hoboken, NJ: John Wiley & Sons.
• Bercholc, J. (2015). Mujeres supremas: la cuestión de género en cortes supremas y tribunales constitucionales. Datos comparados sobre la perfomance de las juezas. Revista Científica Codex, 1(1), 211-226.
• Bergallo, P. (2007). ¿Un techo de cristal en el poder judicial? la selección de los jueces federales y nacionales en Buenos Aires. In L. Cabal; & C. Motta (Eds.), Más allá del derecho: justicia y género en América Latina Bogota, Colombia: Siglo del Hombre.
• Bergoglio, M. I. (2009). Diversidad y desigualdad en la profesión jurídica: consecuencias sobre el papel del derecho en América Latina. Via Iuris, 6, 12-28.
• Blinder, A. S. (1974). Toward an economic theory of income distribution Cambridge, UK: MIT Press.
• Bonelli, M. G. (2011). Profissionalismo, gênero e significados da diferença entre juízes e juízas estaduais e federais. Revista Contemporânea, 1, 103-123.
• Bonelli, M. G. (2016). Carreiras jurídicas e vida privada: intersecções entre trabalho e família. Cadernos Pagu, 46, 245-277.
• Bonelli, M. G; & Oliveira, F. L. (2020). Mulheres magistradas e a construção de gênero na carreira judicial. Novos Estudos Cebrap, 39(1), 143-163.
• Bonet, M. A. R. (2014) Desigualdades de género en el mercado laboral: un problema actual Islas Baleares, España: Universitat de les Illes Balears.
• Cahuc, P; Carcillo, S; & Zylberberg, A. (2014). Labor Economics Cambridge, UK: MIT Press .
• Carvalho, E; & Leitão, N. (2013). O poder dos juízes: Supremo Tribunal Federal e o desenho institucional do Conselho Nacional de Justiça. Revista de Sociologia Política, 21(45), 13-27.
• Centro Regional de Derechos Humanos y Justicia de Género. (2007). Informe regional: iguales en méritos, desiguales en oportunidades - acceso de mujeres a los sistemas de administración de justicia Bogota, Colombia: Autor
• Conselho Nacional de Justiça. (2014). Censo do Poder Judiciário. Vide - Vetores Iniciais e Dados Estatísticos Brasília, DF: Autor.
• Constituição da República Federativa do Brasil de 1988 (1988). Brasília, DF. Recuperado de http://www.planalto.gov.br/ccivil_03/constituicao/constituicao.htm
  » http://www.planalto.gov.br/ccivil_03/constituicao/constituicao.htm
• Croissant, Y; & Millo, G. (2008, julho). Panel data econometrics with R: the plm Package. Journal of Statistical Software, 27(2), 1-43.
• Duarte, M; Fernando, P; Gomes, C; & Oliveira, A. (2014) The feminization of the judiciary in Portugal: dilemmas and paradoxes. Utrecht Law Review, 10(1), 29-43.
• Djurdjevic, D; & Radyakin, S. (2007). Decomposition of the gender wage gap using matching: an application for Switzerland. Swiss Journal of Economics and Statistics, 143(4), 365-396.
• Dushoff, J; Kain, M. P; & Bolker, B. M. (2019). I can see clearly now: reinterpreting statistical significance. Methods in Ecology and Evolution, 10(6), 756-59.
• Fragale, R Filho Moreira, R. S; & Sciammarella, A. P. (2015). Magistratura e gênero: um olhar sobre as mulheres nas cúpulas do judiciário brasileiro. E-cadernos CES, 24, 57-77.
• Frölich, M. (2007). Propensity score matching without conditional independence assumption - with an application to the gender wage gap in the United Kingdom. The Econometrics Journal, 10(2), 359-407.
• Gastiazoro, M. E. (2008). La concentración de las mujeres en la administración de justicia. Anuario del CIJS, 11, 705-721.
• Gastiazoro, M. E. (2010). Construcción de la identidad profesional y de género en la administración de justicia Argentina. Via Iuris, 9, 11-32.
• Gastiazoro, M. E. (2013). Género y trabajo: mujeres en el poder judicial Córdoba, Argentina: Centro de Estudios Avanzados.
• Gastiazoro, M. E. (2016). Inserción de las mujeres en las grandes firmas jurídicas de Argentina. Trabajo y Sociedad, 27, 139-153.
• Hlavac, M. (2014, novembro). Oaxaca: Blinder-Oaxaca decomposition in R. SSRN Recuperado de http://dx.doi.org/10.2139/ssrn.2528391
  » https://doi.org/10.2139/ssrn.2528391
• Hunter, R. (2015). More than just a different face? Judicial diversity and decision-making. Current Legal Problems, 68(1), 119-141.
• Iacus, S. M; King, G; & Porro, G. (2012). Causal inference without balance checking: Coarsened exact matching, Political Analysis, 20(1), 1-24.
• Instituto de Pesquisa Econômica Aplicada. (2020). Boletim Mercado de Trabalho: conjuntura e análise Brasília, DF: Autor .
• Junqueira, E. B. (1998). A mulher juíza e a juíza mulher. In C. Bruschini, & H. B. Holanda (Eds.), Horizontes plurais: novos estudos de gênero no Brasil São Paulo, SP: Fundação Carlos Chagas.
• Kahwage, T; & Severi, F. C. (2019). Para além de números: uma análise dos estudos sobre a feminização da magistratura. Revista de Informação Legislativa, 56(222), 51-73.
• Lei complementar nº 35, de 14 de março de 1979 (1979). Dispõe sobre a Lei Orgânica da Magistratura Nacional. Brasília, DF. Recuperado de http://www.planalto.gov.br/ccivil_03/leis/lcp/lcp35.htm
  » http://www.planalto.gov.br/ccivil_03/leis/lcp/lcp35.htm
• Lei nº 12.527, de 18 de novembro de 2011 (2011). Regula o acesso a informações previsto no inciso XXXIII do art. 5º , no inciso II do § 3º do art. 37 e no § 2º do art. 216 da Constituição Federal; altera a Lei nº 8.112, de 11 de dezembro de 1990; revoga a Lei nº 11.111, de 5 de maio de 2005, e dispositivos da Lei nº 8.159, de 8 de janeiro de 1991; e dá outras providências. Brasília, DF. Recuperado de http://www.planalto.gov.br/ccivil_03/_ato2011-2014/2011/lei/l12527.htm
  » http://www.planalto.gov.br/ccivil_03/_ato2011-2014/2011/lei/l12527.htm
• Lopez, F; & Guedes, E. (2019). Três décadas de funcionalismo brasileiro (1986-2017): atlas do Estado brasileiro Rio de Janeiro, RJ: Ipea.
• Malleson, K. (2003). La justificación de la igualdad de género en la magistratura: por qué la diferencia no funciona. Revista Jurídica de Universidad de Palermo, 1(8), 35-56.
• Marques, G., Jr . (2014). Espaço, profissão e gênero: mobilidade e carreira entre juízes e juízas no estado de São Paulo. Cadernos Pagu, 43, 265-297.
• Meadow, C. M. (2013). A sociologia comparada das advogadas: a feminização da profissão jurídica. Panóptica, 8(25), 67-96.
• Meirelles, H. L. (2016). Direito administrativo brasileiro São Paulo, SP: Malheiros.
• Nakamura, L. M., & Vaz, D. V. (2020). Quão heterogêneo é o setor público no Brasil? uma análise das diferenças salariais entre os poderes Executivo, Legislativo e Judiciário. Administração Pública e Gestão Social, 12(4), 1-24.
• Ñopo, H. (2008). Matching as a tool to decompose wage gaps. The Review of Economics and Statistics, 90(2), 290-299.
• Oaxaca, R. (1973). Male-female wage differentials in urban labor markets. International Economic Review, 14(3), 693-709.
• Portaria nº 63, de 17 de agosto de 2017 (2017). Impõe dever de apresentar dados sobre estrutura e pagamento remuneratório de magistrados pelos Tribunais do País ao Conselho Nacional de Justiça para cumprimento da Lei n. 12.527, de 18 de novembro de 2011 (Lei de Acesso à Informação) e a Resolução n. 215, de 16 de dezembro de 2015. Brasília, DF. Recuperado de https://atos.cnj.jus.br/atos/detalhar/2456
  » https://atos.cnj.jus.br/atos/detalhar/2456
• R Core Team. (2020). A language and environment for statistical computing Vienna, Austria: R Foundation for Statistical Computing.
• Resolução nº 13, de 21 de março de 2006 (2006). Dispõe sobre a aplicação do teto remuneratório constitucional e do subsídio mensal dos membros da magistratura. Brasília, DF. Recuperado de https://atos.cnj.jus.br/atos/detalhar/177
  » https://atos.cnj.jus.br/atos/detalhar/177
• Resolução nº 215, de 16 de dezembro de 2015 (2015). Dispõe, no âmbito do Poder Judiciário, sobre o acesso à informação e a aplicação da Lei 12.527, de 18 de novembro de 2011. Brasília, DF. Recuperado de https://atos.cnj.jus.br/atos/detalhar/2236
  » https://atos.cnj.jus.br/atos/detalhar/2236
• Rhode, D. L. (2001). The unfinished agenda: women and the legal profession Chicago, IL: American Bar Association.
• Ribera, T. B., Miguel, S. E., & Pérez, C. M. D. (2009). Obstáculos en la promoción de las mujeres: el “techo de cristal”. Congreso de Ingenieria de Organización, 12, 133-142.
• Sadek, M. T. (2006). Magistrados: uma imagem em movimento Rio de Janeiro, RJ: FGV Direito Rio.
• Schultz, U. (2003). Women lawyers in Germany: perception and construction of femininity. In U. Schultz; & G. Shaw (Eds.), Women in the world’s legal professions Oxford, UK: Hart Publishing.
• Silva, G. P. (2018). Desenho de pesquisa Brasília, DF: Enap.
• Sommerlad, H., Webley, L., Duff, L., Muzio, D., & Tomlinson, J. (2010). Diversity in the legal profession in England and Wales: a qualitative study of barriers and individual choices London, UK: University of Westminster.
• Vaz, D. V. (2010). Segregação hierárquica de gênero no setor público brasileiro (Boletim Mercado de Trabalho, nº 42). Brasília, DF: Ipea.
• Vaz, D. V. (2011). Evolução da participação feminina no setor público brasileiro no período 1992-2008. In M. A. Abreu (Org.), Redistribuição, reconhecimento e representação: diálogos sobre igualdade de gênero Brasília, DF: Ipea .
• Vaz, D. V. (2013). O teto de vidro nas organizações públicas: evidências para o Brasil. Economia e Sociedade, 22(3), 765-790.

• 1
  Compared to the 1990s, 25% were women and 75% were men. In 2018, 37% were women and 63% were men.
• 2
  According to art. 39, §8 of the CF/1988, there is the possibility, for public servants organized in career, as is the case of judges, for the State to opt for remuneration through subsidy (single installment established by law). However, as a rule, the states opt for the remuneration of judges through salary (fixed installment established by law) and pecuniary advantages (variable installment).
• 3
  See art. 37 - Federal Constitution (1988).
• 4
  See art. 80, II of Complementary Law No. 35/1979, which provides for the Organic Law of the National Judiciary (Loman).
• 5
  We follow the terminology adopted by Kain and Bolker (2019) of using “evident” or “clearly” rather than “statistical significance”.
• 6
  Code and data can be viewed in the following public repository: https://github.com/jjesusfilho/remuneracao
• 7
  We used the term “sex” because it is the category used in the data that served our analyses.

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

• 8
  Time in the judiciary service in years.
• 9
  Time in the judiciary service in months.

Publication Dates

• Publication in this collection
  18 May 2022
• Date of issue
  Mar-Apr 2022

History

• Received
  03 May 2021
• Accepted
  08 Apr 2022