ATTRIBUTES AND CHANCES OF ELECTORAL SUCCESS OF MAYORS IN BRAZIL

Abstract

This article shows the chances of electoral success for local Executive Branch based on the profile of candidates and elected candidates. The results of the municipal elections of 2000, 2004, 2008, 2012 and 2016 were analyzed concentrating on the characteristics sex, age, schooling, and occupation in order to observe the influence of these characteristics in the profile of the elected mayors in the last five elections. The methodology involved documentary and bibliographic research. Data available from the Supreme Electoral Court were analyzed using the chi-square test and logistic regression. The analysis considered Bourdieu’s theory of field and political capital (2011). The results show evidence that the candidate’s sex and occupation are decisive influencing factors, with a decrease in chances of electoral success for male candidates or candidates of a political career in the elections of 2012 and 2016.

Keywords:
chances of electoral success; municipal elections; mayors profile

Resumo

Estudo sobre as chances de sucesso eleitoral do Poder Executivo local com base na evolução do perfil de candidatos e eleitos. Foram analisados os resultados das eleições municipais ocorridas em 2000, 2004, 2008, 2012 e 2016 segundo sexo, idade, escolaridade e ocupação, que revelassem fatores de influência para o perfil dos prefeitos vitoriosos. A metodologia envolveu pesquisa documental e bibliográfica. Tendo como base teoria de campo e capital político de Bourdieu (2011), foram analisados dados disponibilizados pelo Tribunal Superior Eleitoral mediante teste do qui-quadrado e de regressão logística. Os resultados indicaram o sexo e a ocupação declarada como fatores de influência determinantes no perfil dos eleitos. Observou-se diminuição das chances de sucesso eleitoral de candidatos do sexo masculino ou oriundos de carreira política nas eleições ocorridas em 2012 e em 2016.

Palavras-chaves:
chances de sucesso eleitoral; eleições municipais; perfil de prefeitos

Resumen

El artículo estudia las posibilidades de éxito electoral del Poder Ejecutivo local basado en la evolución del perfil de candidatos y elegidos. Se analizaron los resultados de las elecciones municipales ocurridas en 2000, 2004, 2008, 2012 y 2016 según sexo, edad, escolaridad y ocupación que revelasen factores de influencia para el perfil de los alcaldes victoriosos. La metodología incluyó investigación documental y bibliográfica. Con base en la teoría de campo y capital político de Bourdieu (2011), se analizaron datos proporcionados por el Tribunal Superior Electoral mediante la prueba de chi-cuadrado y de regresión logística. Los resultados indicaron el sexo y la ocupación declarada como factores de influencia determinantes en el perfil de los elegidos. Se observó una disminución de las posibilidades de éxito electoral de candidatos del sexo masculino u oriundos de carreras políticas en las elecciones ocurridas en 2012 y 2016.

Palabras clave:
posibilidades de éxito electoral; elecciones municipales; perfil de los alcaldes

1. Introduction

What are the conditions that lead to electoral success in mayoral elections in Brazil? The field literature argues that it is not possible to point to a single determinant factor that has led to campaign success. Various conditions cited by a variety of authors can influence this success like the candidate’s sex, social origin, education, the political socialization to which the candidate is submitted, opportunities provided by the political system, the competitiveness of the candidate’s party, and financial resources (Best & Cotta, 2000; Freire, 2002; Gaxie, 1980; Marvick, 1968; Matthews, 1984; Norris, 1997; Sanbonmatsu, 2006).

In addition, Codato, Cervi e Perissinotto (2013, p. 79) emphasize that the

[…] institutionalization of the political rules and apparatus of representative democracy — parties, parliaments, elections — tends to impose a group of demands which can only be met by those who are professional politicians and not by neophytes — no matter how great their social status or however prominent their educational qualifications.

Seeking a theoretical basis that will help us better understand the factors that influence electoral success, we can examine the concept of political capital and its development based on the Field Theory of Bourdieu (1989). From this perspective emerges the peculiarities and the construction process of the political field, which translates into different forms of capital which is converted in order to construct sufficient symbolic capital to ensure success in the political arena.

Bourdieu (1989), in using the concept of the political field, directs his theoretical approach towards the idea of political capital, which results in factors extrinsic to the political agent examined such as, for example, the candidate’s party, as well as factors that are intrinsic, such as personal characteristics including the candidate’s sex, age, education and personal, professional and family history. From this point of view, the current study seeks to use the line of theory which is based on political capital and statistical analyses to examine the last five Brazilian municipal elections, and identify the elements that can indicate correlations related to the evolution of the dominant emotional profile consisting of the main personal characteristics of all of the candidates, in addition to those who were elected. Therefore this article is focused on the idea of better understanding the influence of some characteristics that have had a greater influence on the political capital attributed to mayors in the elections that have taken place between 2000 and 2016. The current study will observe, for example, how the career of a candidate or personal characteristics such as sex, age and education can influence a candidate’s chances of electoral success. Therefore, this article reflects on the evolving profile of elected mayors in terms of sex, age, education and occupation, to attempt to find a relationship between variables which can give an idea of the influential factors that have determined the winning profile during the last five elections.

2. Theoretical perspectives concerning the conditions that lead to electoral success according to the candidates’ profiles

The understanding of the political-electoral process, based on Norris’s model (1993), is expressed as being dependent on interrelated socioeconomic and cultural aspects as well as aspects of the political-electoral system itself, and establishes two distinct dimensions that influence electoral recruitment by political parties: supply in which parties evaluate, in terms of their standards, potentially eligible competitors and what they can offer society in these electoral contests, and demand in which the parties evaluate the wide array of affiliated citizens of various socioeconomic profiles from which the party seeks to fill its candidate slates with the public’s potential approval in mind.

According to Araújo and Borges (2013, p. 70),

[…] supply and demand occur and are realized in broader scenarios of the political system, and they involve political culture, the party system and legislative races, or in other words, the specific context of each electoral dispute. Beyond these levels, they also include the configurations of the ‘party system,’ in which the authors highlight the relevance of the ‘ideology’ and ‘political culture’ of each party.

The authors also indicate that available campaign resources, the political context of the dispute, and the aspirants’ trajectories, profiles and careers are also important variables in the formation of party slates.

There is a consensus among various authors in the field literature (Miguel, 2004; Bordignon, 2013; Campos, 2015) that intra-party dynamics seem to heavily influence the political underrepresentation of groups, depending on the filters employed. In other studies, Codato, Cervi e Perissinotto (2013) investigate and measure the explanatory potential of economic capital and the postulated attributes, political strategies and resources of Brazilian mayoral candidates in municipal elections in 2012. These authors infer that

[…] the variables that revealed themselves to be the most important in explaining electoral success were, as demonstrated, of a political nature. It quickly becomes apparent that the accumulation of an individual’s own political capital qualifies that individual to compete for political positions in which he or she has a reasonable chance of victory (Codato, Cervi, & Perissinotto, 2013, p. 79).

Moreover,

[…] for candidates up for reelection, being part of a party coalition and being in a party characterized by strong performance are the main explanatory principles for the acquisition of a municipal mandate. They are followed by the variable that measures the availability of campaign resources. The more revenues the candidate has, the greater the chance that the candidate will be elected (Codato, Cervi, & Perissinotto, 2013, p. 790).

Political capital, in this context, proves to be an alternative in understanding the analyzed scenario. Once the concepts of political capital and field are related, to better understand the idea of political capital, it is of interest to first contemplate the limits of this field and the agents that make up the field theory and political capital perspective of Bourdieu (1989).

2.1 POLITICAL CAPITAL AND THE PROFESSIONAL CAREERS OF CANDIDATES AND ELECTED OFFICIALS

Characterized, within Bourdieu’s construct (2011), as social relationship space that assumes specific activities, the field concept is tied to professional, religious, ideological, and emotional aspects, among others. For example, the religious field or professional fields like medicine and law possess rites that enable new individuals to enter these fields. The rites of initiation into ecclesiastic life or receiving a degree in some profession or passing exams that confer specific professional licensing, such as law exams in Brazil, are clear and objective ways of entering these respective fields. The political field, on the other hand, is not endowed with such clear limits.

A field’s limit is what defines who is recognized as a part of it; who belongs to it. According to Bourdieu (2011, p. 202), “[…] it recognizes the presence or existence of an agent in a field by the fact that he or she transforms the state of the field (or that, if removed, things change significantly).” To the author, the limits of the field extend until the field effects cease, or in other words, the limit of a field is the limit of its effects. In the case of the political field, this limit is difficult to define and cannot be easily perceived. It’s possible to belong to it in many forms, for example, through social movements, civil organizations, and political parties, among other things. Its agents also assume various functions, and can be journalists or public opinion specialists as well as political party members, whether they occupy public office or not. In this way, field agents will be characterized by the extent to which they are capable of being affected and producing effects in a field (Bourdieu, 2011).

Based on this, it should be understood that political capital is formed within a field. According to Bourdieu (2011, p. 204), “[…] there are, in the political field, symbolic fights in which adversaries have arms, capital and symbolic powers that differ in strength.” Political capital, to Bourdieu (2011, p. 204), is characterized within this context by “[…] a species of reputation capital, symbolic capital that is linked to the manner in which one is known.”

Within the political field, the battles that are fought seek to strengthen ideas that are supported by various interest groups that are consolidated there. The arms used by field agents are political capital, the recognition by people, groups or institutions that this agent is potentially capable of mobilizing society and producing change in his or her own field.

The development of political parties institutionalizes roles and influences the capital of their agents. According to Bourdieu (2011, p. 204) “The political capital of a political agent mainly depends on the political weight of the agent’s party and of the agent within that party.” The authors further points out that,

Currently, the party is a type of bank of specific political capital, and the party secretary is a type of banker (perhaps it is not a coincidence that all of our past and future presidents are former party secretaries...) who controls access to political, bureaucratized, and bureaucratic capital, guaranteed and bureaucratically certified by the party bureaucracy. To the extent that the political field is bureaucratized, being admitted to the institution supposes rights of access, and today these rights are more and more granted by the parties (and by important schools, notably the National School of Administration, the ENA). The most tenacious conservatives of a party are the ones who are most dependent on it (Bourdieu, 2011, p. 204)¹.

Political capital is, therefore, acquired by field agents in an unequal manner. This is because it depends on the agent’s personal and social conditions in terms of access to the political field. This deals with, for example, financial resources, free time to dedicate to the political issues in play, education, professional career, and other personal characteristics of the agent, such as age, sex, and family history, among others.

Those who have better financial conditions and free time tend to have greater ease in acquiring political capital as compared to competitors who lack the same conditions. In the same way, those that have higher levels of education will have greater ease in being admitted to and remaining in the political field. This is due to their greater acceptance by a portion of those who are already in the field, because they give people greater recognition, broaden their relationship networks and make them better able to present themselves within this field that they depend on and are held accountable to.

It is from this perspective, that we encounter the political capital categories proposed by Bourdieu (1989). The first, based on the assumption of influence linked to political parties, identifies delegated capital, which is qualified by and comes from the political capital that historically has been accumulated by institutions, parties and unions, which have symbolic capital in terms of recognition and loyalty. Delegated capital is, therefore, invested in those who stand out in institutions, and is normally temporary, but is frequently perpetually renewed, and is more significant to the extent that these institutions have been strongly established and bureaucratized, and therefore possess greater political capital. Within this category, the capital attributed to the agent belongs, in truth, to the organization and public office, in general.

In the second category, we have personal capital, which is based on the fact that people “[…] have a certain number of specific qualifications which are a condition for acquiring and maintaining a good reputation” (Bourdieu, 1989, p. 191).

In another form, Miguel (2003) divides Bourdieu’s categories into three subgroups of political capital: delegated capital, with a definition similar to that of Bourdieu; converted capital; and heroic capital. Converted capital consists of the transfer of capital accumulated over a lifetime in other professional, family, religious, cultural, and economic domains which enable an agent to be considered notable in the political field. This is the case, for example, with intellectuals, artists and athletes who enter the world of politics. In this case, according to Miguel (2003), it’s important to identify the “conversion rate,” which varies according to the type of capital due to battles in the political field. Other subcategories emerge in this category which promote a conversion to political capital. They are: cultural capital, artistic capital and economic capital, among others, as well as heroic or prophetic capital, considered to be a subspecies of converted capital which has been acquired by singular actions taken during times of crisis.

Miguel (2003), in establishing these new subcategories, seeks to create, from Bourdieu’s perspective, a theoretical framework more aligned with the Brazilian situation, which is characterized by fragile party institutions, and is endowed with little political capital.

Miguel understands delegated capital “[…] to be the original capital of the political and/or state field. That is, notoriety comes from previous electoral mandates, occupying public positions of trust within the executive branch and the party leadership” (2003, p. 121). Converted capital comes from factors such as “[…] popularity gained from other fields and transferred to politics, in addition to economic capital, through expensive electoral campaigns” (p. 121).

This conceptual framework is sufficiently practical to analyze the career of countless Brazilian politicians. Those who sustain themselves mainly on delegated capital are those who seek to occupy positions of less national prominence, and then move on to greater quantities of capital that are accumulated slowly and gradually. It’s the case of politicians who first succeed in being elected councilmen, then mayors of small cities, then state representatives, federal representatives, senators, etc. In another manner, politicians who can better take advantage of converted capital are allowed to make “jumps” in their career, to find out if they can take on positions with higher profiles, like Secretary of State, government minister, senator or mayor of a large city (Miguel, 2003). Within this configuration, a “political career” comes about through the accumulation of political capital which implies the guarantee of permanence in the political field.

2.2 THE CHANCES OF BEING ELECTED MAYOR IN BRAZIL: THE MAIN CONDITIONS NEEDED FOR A CAREER AND THE ACCUMULATION OF POLITICAL CAPITAL

The city is the environment of everyday life, and constitutes the space of property and democratic activity. The Constitution of 1988 ensures that there is a municipal aspect to the context of its reform of the State which is mainly based on political, fiscal and administrative decentralization. The recognition of the municipality as a federal entity changes its institutional status, increasing its political autonomy (Arretche, 2000).

It is precisely due to the importance given to the role of local government by Arretche (2000) and other authors that there is no lack of challenges in the work that faces mayors and councilmen. This scenario, public life, involves the acting of political agents in the municipal context, as Bordignon (2013) points out by saying that few politicians begin their elected career on the state or federal level (within the executive or legislative branches) without first holding positions on a municipal level. Considered the main route of access to elective political office, the municipal level is a space where identity is diffused, electoral bases are formed, and intraparty approval is bestowed.

2.2.1 INFLUENCE OF PROFESSIONAL CAREERS ON ELECTORAL SUCCESS AND AN EVOLUTIONARY PORTRAIT OF THE INEQUALITY BETWEEN THE SEXES IN POLITICS

The influence of factors associated with political capital is also revealed, in addition to the political trajectories of individuals, by party decisions related to candidates in terms of their careers and sexes. The studies of Bourdieu (1996) effectively accentuate the role of social origins in the production and reproduction of inequalities in all spheres of life, including professional life.

Political capital “indicates the social recognition that permits some individuals, more than others, to be accepted as political actors and, therefore, to act politically” (Miguel, 2003, p. 115). This driving force has assumed a braking effect on the careers of politicians who have accumulated little of this capital, including women, for whom gender socialization attributes roles more associated with the care of public issues (Miguel, 2003; Pinheiro, 2007). The force of this finding has not been sufficient to transform this reality, and the growth of the participation of women in the job market during the second half of the twentieth century and the beginning of the twenty-first century has not corresponded to a proportionate increase in gains, and in general, women need to demonstrate greater expertise to assume leadership positions. As a result, despite recent advances, the scenario of female participation in the workplace has been characterized by great difficulties and challenges, especially in terms of equal remuneration, the conditions required to reconcile work with raising a family (a double work day), and a concentration of women in positions with lower levels of remuneration (Pinheiro, 2007).

In the public sector, however, the participation of women in the workforce is one of its greatest paradoxes, characterized by a proportionately inverse relationship between the number of women who occupy non-elective positions where they represent the majority (of public exam candidates), and elective positions (in terms of participating in elections) where there are relatively few women, despite quotas for women in politics (Pinheiro, 2007).

Again the notion of Bourdieu’s political capital (2011) offers an interesting key to explaining this phenomenon. Given that political activity occurs within a context of a dispute for power, political capital implies that the occupation of dominant positions requires an accumulation of this capital, whose obtainment cannot rationally be related only to cultural and educational capital, or in other words, based on meritocracy (Pinheiro, 2007). In this way, Bourdieu understands that there exists another principle of differentiation, another type of capital, whose unequal distribution is based on the verified differences: political capital. In relation to this, Bourdieu (1996, p. 31) asserts:

[political capital] assures that those who possess it have a way to privately acquire public goods and services […] we therefore see that social capital of a political nature that is acquired through the apparatuses of unions and parties is transmitted through networks of family relationships which lead to the construction of true political dynasties.

In line with Bourdieu’s understanding, Miguel (2003, p. 121) concludes:

Political capital is, to a large extent, a species of symbolic capital; the recognition of the legitimacy of this individual to act in politics. It is based on portions of cultural capital (cognitive training for political action), social capital (networks of established relationships) and economic capital (those who have the necessary free time for politics). Like every form of capital, political capital is unequally distributed in society. At the base of the pyramid we have voters [...] on the top, the leaders who [...] are recognized as representatives of various social strata.

We conclude this section by alluding to the concepts of the glass ceiling, the glass wall, and the crystal labyrinth, coined by various authors. These metaphoric images, each with its own particular characteristics, refer to the invisible barriers that limit the choices and opportunities that women have in advancing their careers.

3. Methodology

In order to comparatively analyze the conditions that have determined success in mayoral elections in Brazil from 2000 to 2016, this study analyzes the universe of candidates and mayors based on the methodology detailed here. This study is exploratory in nature based on bibliographic, documentary and field research. We have consulted various political science authors, especially Bourdieu (1989, 1996, 2011), Miguel (2003), Urbinati (2006) and Codato, Cervi and Perissinotto (2013). We have performed documentary research based on the data made available by the Supreme Electoral Court (SEC), in which we have sought to develop a methodology of statistical analysis that can establish an association between the variables. The repository of electoral data, made available in the form of data files, has made it possible to select and analyze the variables of interest in this study. Based on this data, we present, initially, the profiles of the mayors elected between 2000 and 2016 in terms of age, education, sex and occupation.

In relation to the statistical study, this study seeks, at first, to determine whether social and professional attributes favor electoral success based on the cross-referencing of independent variables (sex, age, education and occupation) with dependent variables (resulting in election: elected/not elected). The chi-squared test (χ²) indicates the level of dependence between variable pairs, or in other words, it is a test of hypotheses which intend to evaluate the association that exists between qualitative variables, seeking a dispersion value for two nominal variables which is used to verify the dependence or independence of the variables considered.

Logistic regressions were then employed as a way to evaluate the effects and the force of all of the explanatory variables as a group to understand which independent variables are more relevant to the victory of a mayoral candidate. This methodology consists of using statistical analysis and employing dummy variables (binary variables that assume a value of 0 = no occurrence and 1 = occurrence). The function of a logistic regression is to indicate the probability that a given fact will occur given a change in an independent characteristic. Here, the fact is whether the individual was or was not elected mayor, and the characteristics are the explanatory variables: sex, age, education and occupation. Thus, the results of a logistic regression express the probability of predicted values of a dichotomous variable (yes/no). In this case, the values are 0 = not elected or 1 = elected. For example, taking the “sex of the candidate” variable in relation to the dependent variable, the result will be interpreted in terms of intensity: how much the chances that the candidate is elected (or in other words, go from 0 to 1 in this variable) increase or decrease when the candidates are of the female or male sex.

The explanatory variables considered were: a) Age: verifying whether the age of the candidate has any impact on the chances of being elected. The binary variable will represent the counterpoint between the “mature” group, defined in the present study as the group from 40 to 54 years of age, and the sum of the other groups, namely: the “young” group of 20 to 39 years of age and those belonging to the “older” group with ages of 55 or more. This variable assumes a value of 1 (one) for the “mature” group; b) Education: measuring the greater or lesser impact caused by educational level in election results. The binary variable was constructed taking into account whether the candidate has a college degree or not, with 1 (one) indicating the affirmative and 0 (zero) indicating other levels of education; c) Occupation: measuring the chances of the candidate who has a previous political career (reelection or previous political trajectory). We include candidates who declare occupations classified in the subgroup “government officials and leaders” (Horta et al., 2017). It may be noted that most of these candidates are mayors (a few others are councilmen, representatives, etc.). The binary variable assumes a value of 1 (one) for those candidates who belong to the referenced subgroup, and 0 (zero) for other occupations. d) Sex: to measure whether men have greater or lesser chances of being elected mayor. In regard to the last subgroup, it determines how much being a man influences or does not influence the election of a mayor. How great a weight does being a man have in the results of Brazilian mayoral contests? The binary variable uses 1 (one) for men and 0 (zero) for women.

4. The profile of mayors elected between 2000 and 2016

Our analysis of the profile of mayors elected during the period from 2000 to 2016 is presented below, organized by the following characteristics: a) age and education and b) occupation and sex.

4.1 AGE AND EDUCATION

The results of the municipal elections from 2000 to 2016 presented in table 1 demonstrate a characteristic pattern in terms of the age of the elected mayors and show a concentration among the age groups of 40 to 44, 45 to 49 and 50 to 54, which make up more than half of the elected mayors from 2000 to 2012, and roughly half in the election of 2016. However, in the last three elections, the representation of these three age groups has diminished. The oldest individuals, namely those 55 or over, have represented a slightly larger percentage than the youngest age group, from ages 20 to 39. In the election of 2016, 49.8% of the elected mayors were between 40 and 54 years of age. In terms of the others, 21.0% were between the ages of 20 and 39, while 29.2% were 55 or over. The changes observed over time led to the systematic increase of the average age of mayors in the country. This age was roughly 48.0 years old in the 2000 election and 48.8 years old in the election of 2016.

Table 1
Relative distribution of elected mayors in terms of age groups (2000, 2004, 2008, 2012 and 2016)

Year	Age groups
	20 to 24	25 to 29	30 to 34	35 to 39	40 to 44	45 to 49	50 to 54	55 to 59	60 to 64	65 to 69	70 to 74	75 to 79	80 to 84	85 to 89	90 to 94	Total
2000	0.2	1.1	4.4	13.1	19.6	21.8	16.6	11.2	6.7	3.3	1.5	0.5	0.1	0.0	0.0	100.0
2004	0.3	1.8	4.9	13.0	20.0	19.6	17.5	11.1	6.5	3.4	1.3	0.4	0.1	0.1	0.0	100.0
2008	0.3	1.9	4.7	10.2	19.1	21.8	18.0	12.2	6.8	3.1	1.2	0.3	0.2	0.0	0.0	100.0
2012	0.5	2.2	6.8	10.4	15.3	20.9	17.7	12.5	7.8	3.6	1.7	0.4	0.1	0.0	0.0	100.0
2016	0.4	2.3	6.6	11.7	14.6	17.0	18.2	12.7	9.0	4.4	2.3	0.6	0.2	0.1	0.0	100.0

• Source: Elaborated by the authors.

In terms of education, this study demonstrates the undeniable growth in the level of education of elected mayors, verifying that there are still candidates with little education, even though their percentage has been decreasing over time. Two points ate noteworthy in the distribution of elected mayors in terms of education: the predominance of those with degrees in higher education and the gradual increase in education between the elections of 2000 and 2016. Table 2 shows that of the mayors elected in 2000, 38.7% declared themselves as having a college degree, which grew to 52.5% by 2016. Thus, we can see the elevated proportion of elected mayors with college degrees, which supposes that people with less education have fewer chances of being elected. While in the election of 2000, 15.5% of all mayors had not completed elementary and middle school,² by 2016, this percentage had fallen to 7.0% of all elected mayors, the fruit of significant and successive declines during the four election periods. It should also be emphasized that in the elections of 2000, 23.6% of elected mayors had a high school degree. In the election of 2016, mayors with this level of education represented 25.7% of the total.

Table 2
Relative distribution of elected mayors in terms of education (2000, 2004, 2008, 2012 and 2016)

Year	Level of education
	Read and Write	Incomplete Elementary and Middle School	Complete Elementary and Middle School	Incomplete High School Education	High School Degree	Incomplete College Education	College Degree	Total
2000	2.0	13.5	10.0	5.1	23.6	7.0	38.7	100.0
2004	1.7	11.9	7.5	4.3	26.4	7.0	41.2	100.0
2008	1.0	9.4	7.3	3.9	27.0	6.8	44.6	100.0
2012	0.8	7.1	6.5	2.9	27.5	6.0	49.2	100.0
2016	0.8	6.2	6.5	2.5	25.7	5.7	52.6	100.0

• Source: Elaborated by the authors.

The results obtained in the age variable analysis present significant changes over the last five elections, but they demonstrate that the quantity of political capital necessary in winning a mayoral election requires candidates with notoriety acquired through long periods of maturation. Thus, mayors elected before they complete 40 years of age are less frequent, which signifies that the political capital, whether it is delegated or personal, necessary to win mayoral elections produces fruit over the long term.

In terms of education, the data shows that a college education is valued in Brazil. In the last 5 municipal elections, each time more mayors have been elected with college degrees, which could be associated with the expansion of higher education during this period or the recognition that higher education confers on people, representing a symbol of an elevated social position. Moreover, higher education could also be related to the economic capital of these candidates — citizens who can dedicate themselves to their studies — and the expansion of their professional relationship networks, while many others need to dedicate themselves to work as soon as they become adults.

4.2 SEX AND DECLARED OCCUPATION

The results of the balloting in the last five elections indicate an underrepresentation of women as a proportion of Brazilian mayors, despite the fact that the percentage of them elected grew continually between 2000 and 2012, and remained stable during the election of 2016. Census information from 2000 and 2010 shows that a bit over half of the Brazilian population is made up of women, while the representation of women in Brazilian municipalities has been much less significant. In the year 2000, 5.7% of elected mayors were women. As table 3 shows, this percentage gradually grew during the next three elections reaching 11.8% in 2012. Or in other words, it more than doubled over four election periods, but as mentioned above, it remains at a reduced level. In the next election, this percentage receded slightly to 11.6%.

Table 3
Relative distribution of elected mayors in terms of sex in 2000, 2004, 2008, 2012 and 2016

	Sex
Election	Female	Male
2000	5.7	94.3
2004	7.4	92.6
2008	9.4	90.6
2012	11.8	88.2
2016	11.6	88.4

• Source: Elaborated by the authors.

The slow inclusion of women in the political field demonstrates the low recognition of women as political agents, as well as the invisible barriers mentioned above.

An analysis of the occupations declared by elected mayors in the elections from 2000 to 2016 indicates important changes in terms of the occupational categories or groups.³ In terms of declared occupation, “arts and sciences professionals” presented the largest participation, but this has declined over time. According to the sample in table 4, which shows the study’s main findings,⁴ in the year 2000, this group represented the largest percentage of elected mayors (28.4%) surpassing the total of 25.6% in 2016. This last election represents another large percentage; however, it ceased to be the leading group of occupations. The subgroup “biological sciences and health services professionals” stands out within this group in all of the elections analyzed, and also presented a decline in its percentage.

This study shows that starting with the election of 2008, “arts and sciences professionals” ceased to be the first occupational group, ceding its place to “government officials, leaders of political organizations and companies, and managers.” This fact is relevant due to the significant growth of the participation of this last occupational group over the past five elections in terms of relative importance. Its percentage practically doubled, going from 22.0% in 2000 to 42.0% in 2016. Within this group, the results point to the elevated and growing participation of the subgroup “government officials and leaders” in the composition of elected mayors. If in 2000 it represented only 5.9 % of all elected mayors, it reached 21.0% by 2012. Almost all of those in this occupational group declared themselves to be mayors or councilmen, and could classify themselves as “career politicians” or “reelected politicians.” Besides this subgroup, the “business and organization leaders” subgroup has presented a growing participation even though it previously presented low levels of participation. In 2000 it represented 2.3%, rising to 15.0% in 2016. It should be emphasized that the two groups highlighted above represent more than half the occupations declared by mayors — 50.4% in 2000 and 67.7% in 2016.

It should be noted that in this data analysis it is not possible to split “business and organization leaders” in terms of attributing which portions are represented by private companies, organizations and public companies. This group has demonstrated significant growth during the years under examination, increasing from 2.25% of all elected mayors in 2000 to 14.96% in 2016.

Table 4
Relative distribution of elected mayors in terms of declared occupation during the years 2000, 2004, 2008, 2012 and 2016

Description	2000	2004	2008	2012	2016
GROUP 0: Armed forces and military policemen and firemen	0.27	0.22	0.13	0.13	0.18
GROUP 1: Government officials, leaders of political organizations and companies, and managers	22.00	26.44	43.93	41.54	42.02
Government officials and leaders	5.94	7.35	26.62	22.27	20.98
Leaders of organizations and companies	2.25	7.32	9.03	12.52	14.96
Leaders and managers of health, educational, cultural, social and personal services	0.07	0.13	0.16	0.13	0.13
Managers	13.74	11.65	8.12	6.63	5.95
GROUP 2: Arts and sciences professionals	28.44	30.34	25.61	25.50	25.63
Biological sciences and health related professionals	11.74	11.56	8.93	7.90	8.31
GROUP 3: Midlevel technicians	3.47	3.93	3.44	2.84	2.20
GROUP 4: Administrative service workers	0.25	0.49	0.74	0.88	0.79
GROUP 5: Service workers, commercial salespeople in shops and markets	3.26	1.29	1.44	1.02	1.21
GROUP 6: Agricultural, forestry, hunting and fishing workers	16.94	16.05	12.60	11.24	11.21
GROUP 7: Industrial goods and services production workers (steelworking and metalworking)	0.04	0.04	0.05	0.05	0.18
GROUP 8: Industrial goods and services production workers	0.02	0.00	0.04	0.00	0.05
GROUP 9: Maintenance and repair workers	0.00	0.00	0.00	0.00	0.22
GROUP 10: Public servants	6.32	7.99	5.81	8.51	7.59
GROUP 11: Retirees	3.10	2.59	1.76	2.46	3.15
GROUP 12: Other Categories	0.63	1.96	1.66	1.40	0.96
GROUP 13: Not declared	4.50	2.35	0.00	0.00	0.00
GROUP 999: Others	10.77	6.32	2.81	4.42	4.61
Total	100.00	100.00	100.00	100.00	100.00

• Source: Elaborated by the authors.

These results demonstrate that a professional career is one of the most significant sources for accumulating political capital for political actors in a local context. In sum, having occupied strategic posts in governmental institutions is one way to acquire delegated capital, and it has been the most frequent activity of elected mayors since the election of 2008. This can be understood as an increase in the bureaucratization of the Brazilian political field in recent years, in which public institutions have developed the ability to stock significant amounts of political capital that can be transferred to field agents, unlike in the past in which personal and converted capital were of the greatest relevance in defining the strongest candidacies.

5. Chances of being elected mayor in Brazil: electoral success in terms of social and professional attributes and probabilities

Table 5 presents the relative distribution of elected and non-elected candidates during the elections of 2000, 2004, 2008, 2012 and 2016 in terms of the constructed binary variables: sex, age between 40 and 54, having a college degree, and listing their occupation as “political career,” as specified above. The results of the chi-squared (χ²) test appear right after the cross-references, and the level of significance is presented within parentheses.

The results show that almost all of the explanatory variables have an individual relationship with the dependent variable, at a 5% level of significance. The exceptions refer to the age from 40 to 54 variable for the 2004 and 2016 elections and the college degree variable for the 2008 and 2016 elections. It can be inferred that, in terms of the sex of the candidates, there was an important relationship with whether the candidate was elected or not. Being a man contributed significantly to being elected during the election of 2000 (χ² of 45.331), which means that, in other words, even if we consider the high number of male candidates compared to female candidates, the former were proportionately more likely to be elected. This was the variable which presented the strongest relationship that year. Men represented 92.4% of the candidates, while representing 94.3% of those elected. The relationship continued during the following three elections with a lower chi-squared value, but one that still had strong statistical significance. In other words, the sex of the candidate was responsible for greater or lesser success in elections, but this variable has gradually decreased in importance between the elections of 2000 and 2016.

Table 5
Brazil: Relative distribution of elected and non-elected candidates in mayoral elections in terms of sex, age, education and political career in 2000, 2004, 2008, 2012 and 2016

2000 Election			2004 Election			2008 Election			2012 Election			2016 Election
Elected	Sex		Elected	Sex		Elected	Sex		Elected	Sex		Elected	Sex
	Female	Male		Female	Male		Female	Male		Female	Male		Female	Male
No	8.7	91.3	No	10.6	89.4	No	12.6	87.4	No	14.3	85.7	No	13.9	86.1
Yes	5.7	94.3	Yes	7.4	92.6	Yes	9.4	90.6	Yes	11.8	88.2	Yes	11.6	88.4
Total	7.6	92.4	Total	9.5	90.5	Total	11.4	88.6	Total	13.4	88.6	Total	13.1	86.9
Chi-squared: 45.3 (0.000)			Chi-squared: 43.1 (0.000)			Chi-squared: 36.0 (0.000)			Chi-squared: 18.2 (0.000)			Chi-squared: 17.1 (0.000)
Elected	Age from 40 to 54		Elected	Age from 40 to 54		Elected	Age from 40 to 54		Elected	Age from 40 to 54		Elected	Age from 40 to 54
	No	Yes		No	Yes		No	Yes		No	Yes		No	Yes
No	46.6	53.4	No	44.9	55.1	No	47.2	52.8	No	48.4	51.6	No	50.7	49.3
Yes	42.7	573.	Yes	43.5	56.5	Yes	40.9	59.1	Yes	46.1	53.9	Yes	50.2	49.8
Total	45.1	54.9	Total	44.4	55.6	Total	44.9	55.1	Total	47.6	52.4	Total	50.5	49.5
Chi-squared: 21.5 (0.000)			Chi-squared: 2.7 (0.103)			Chi-squared: 55.1 (0.000)			Chi-squared: 7.5 (0.006)			Chi-squared: 0.3 (0.562)
Elected	Education		Elected	Education		Elected	Education		Elected	Education		Elected	Education
	College degree	Other		College degree	Other		College degree	Other		College degree	Other		College degree	Other
No	59.1	40.9	No	56.5	43.5	No	55.1	44.9	No	50.9	49.1	No	47.8	52.2
Yes	62.0	38.0	Yes	58.8	41.2	Yes	55.5	44.5	Yes	50.8	49.2	Yes	47.4	52.6
Total	60.1	39.9	Total	57.3	42.7	Total	55.2	44.8	Total	50.9	49.1	Total	47.7	52.3
Chi-squared: 12.6 (0.000)			Chi-squared: 7.6 (0.006)			Chi-squared: 0.2 (0.652)			Chi-squared: 0.03 (0.853)			Chi-squared: 0.2 (0.647)
Elected	Political career		Elected	Political career		Elected	Political career		Elected	Political career		Elected	Political career
	No	Yes		No	Yes		No	Yes		No	Yes		No	Yes
No	96.0	4.0	No	96.0	4.0	No	96.0	4.0	No	84.7	15.3	No	82.3	17.7
Yes	94.2	5.8	Yes	94.2	5.8	Yes	94.2	5.8	Yes	77.7	22.3	Yes	79.0	21.0
Total	95.3	4.7	Total	95.3	4.7	Total	95.3	4.7	Total	82.2	17.8	Total	81.5	18.5
Chi-squared: 25.6 (0.000)			Chi-squared: 25.6 (0.000)			Chi-squared: 25.6 (0.000)			Chi-squared: 119.7 (0.000)			Chi-squared: 35.0 (0.000)

• Source: Elaborated by the authors.

The age variable also presented, individually, a relationship with the electoral result. Candidates between the ages of 40 and 54 had a greater chance of success compared to those who are younger (20 to 39 years old) and older (55 years old and over) in the elections analyzed, with the exception of 2004 and 2016 as pointed out above.

The education of the candidates demonstrated a relationship with electoral performance only during the elections of 2000 and 2004, but those favored were those who did not have college degrees as compared with those who did. Among the study’s dependent variables, the one that indicates the weight of the “political career” variable in terms of success in mayoral elections stands out. In the same way as the sex of the candidates, having a “political career” contributed substantially to the candidate’s chances of being elected, establishing elevated values for χ². In other words, having a “political career” clearly increases the chances of winning mayoral elections.⁵ In the election of 2016, 18.5% of the candidates classified themselves as having a “political career,” and the percentage of them elected mayor was larger, namely 21.0%.

With the objective of evaluating the effects and force of the explanatory variables together, and seeking to verify their relevance to the victory of the mayoral candidate, we next will present the results of the logistic regressions. Only the variables presenting statistical significance were considered, as indicated by the p-value (less or equal to 0.05). The analysis will be based on the chance ratios calculated based on exp(β). The value of this statistic indicates the intensity of the probability of being elected mayor and the sign indicates its direction, or in other words, whether there exists a greater or lesser chance of election due to each of the characteristics analyzed.

In the election of 2000, being a man, or being between the ages of 40 and 54, or having a “political career” significantly increased the chances of being elected mayor. The results of table 6 show that a male candidate had a 57.9% greater chance of being elected than a female candidate. Also of significance is the advantage that candidates with “political careers” had. They had a 47.9% greater chance of victory than other candidates. Being between the ages of 40 and 54 also led to a greater chance of being elected, even though at a lower level (19.3% greater than the other candidates). Finally, in terms of the education of the candidates, having a college degree diminished the chances of being elected.

Table 6
Logistic regression coefficients for 2000

Description	B	Wald	Sig.	Exp(B)	Chance
Sex	0.457	43.806	0.000	1.58	57.89
Age from 40 to 54	0.176	26.105	0.000	1.19	19.28
College degree	-0.124	12.467	0.000	0.88	-11.67
Political career	0.391	24.924	0.000	1.48	47.90
Constant	-1.025	205.488	0.000	0.36	-64.12

• Source: Elaborated by the authors.

The results of the 2004 election, presented in table 7, show that a candidate’s sex, education and occupation contributed in an important manner to the election of mayors. In terms of the chance ratios, the importance of occupation was the determinant characteristic, and having a “political career” led to a 62.4% greater chance of victory. This was followed by the sex of the candidate as the determinant characteristic, because male candidates had a 49.9% greater chance of being elected than female candidates. Once again, in this election, it was observed that having a college degree did not increase the chances of being elected mayor, and in fact resulted in a 7.2% lower likelihood of being elected.

Table 7
Logistic regression coefficients for 2004

Description	B	Wald	Sig.	Exp(B)	Chance
Sex	0.405	41.904	0.000	1.499	49.87
Age from 40 to 54	0.056	2.634	0.105	1.057	5.75
College degree	-0.075	4.666	0.031	0.928	-7.23
Political career	0.485	47.294	0.000	1.624	62.37
Constant	-0.985	221.188	0.000	0.373	-62.66

• Source: Elaborated by the authors.

In the 2008 election, the importance of a “political career” stands out in terms of being elected mayor. The candidates classified in this category had a 139.8% greater chance of being elected as compared to all the other candidates. Once again it was followed by being a member of the male sex which increased the chances of being elected mayor by 41.0% when compared to the female sex. That year, the age variable proved to be statistically significant, with the candidates with ages from 40 to 54 having a 29.1% greater chance of being elected than the other candidates. Beginning with this election, the college degree variable ceased to have an explanatory impact on the election of mayors.⁶

Table 8
Logistic regression coefficients for 2008

Description	B	Wald	Sig.	Exp(B)	Chance
Sex	0.344	36.616	0.000	1.410	41.00
Age from 40 to 54	0.255	53.528	0.000	1.291	29.10
College degree	0.034	0.937	0.333	1.034	3.40
Political career	0.875	405.351	0.000	2.398	139.80
Constant	-1.141	338.178	0.000	0.320	-68.00

• Source: Elaborated by the authors.

The same three variables: sex, age and “political career,” continued to represent the greatest chances of being elected in 2012. Once again those candidates with a “political career” stood out. Belonging to this category increased the chances of being elected by 58.5% as compared to other candidates. The sex of the candidate was also significant and a determining factor in the election. Once again being male increased the chances of being elected, this time by 25.0% in relation to female candidates. Finally, the age of the candidate also provided advantages. Those who were between the ages of 40 to 54 had a 9.0% greater chance of being elected.

Table 9
Logistic regression coefficients for 2012

Description	B	Wald	Sig.	Exp(B)	Chance
Sex	0.223	19.120	0.000	1.250	25.02
Age from 40 to 54	0.086	6.480	0.011	1.090	8.97
College degree	0.040	1.374	0.241	1.041	4.06
Political career	0.461	116.515	0.000	1.585	58.52
Constant	-0.925	266.982	0.000	0.397	-60.33

• Source: Elaborated by the authors.

In the last election for mayor, we may observe the consolidation of the sex and occupation attributes in terms of the electoral chances of the candidates. That year, just two of the variables were statistically significant. Candidates with a “political career” had a 28.1% greater chance in the election than the other candidates, and being male increased the chances of the candidates by 24.5%.

Table 10
Logistic regression coefficients for 2016

Description	B	Wald	Sig.	Exp(B)	Chance
Sex	0.219	18.302	0.000	1.245	24.49
Age from 40 to 54	0.018	0.306	0.580	1.019	1.86
College degree	0.043	1.600	0.206	1.044	4.36
Political career	0.248	34.903	0.000	1.281	28.12
Constant	-0.911	254.270	0.000	0.402	-59.79

• Source: Elaborated by the authors.

6. Conclusions

In terms of the sex variable, the data analysis confirms the theories that we have addressed, reinforcing a negative trait of Brazilian society: the belief that there are limitations to the extent that women should occupy leadership positions. It is very clear from this study that they have greater difficulty in gaining access to the political field, since they have to deal with social restrictions in terms of entering parties and other public organizations that can make them strong mayoral candidates. Even though there has been improvement in the participation of female candidates among the group of victorious candidates from 2000 to 2012, increasing from 5.7% to 11.8%, the slight decrease in this number in 2016 to 11.6% is a piece of data that is worthy of concern.

Thus, the perceived differences in the chances of being elected due to the sex of the candidates and the stagnation in the advance of feminine participation make it clear that a quota policy for candidates in Brazilian elections should be examined in depth, because these elections are far from promoting equal opportunities for men and women. We consider that the consistent underrepresentation of women justifies affirmative policies that will stimulate feminine participation in the country’s political life, especially gender quotas for the candidates. In this sense, the quotas defined by Law n. 12,034/2009⁷ are applicable only to proportional legislative positions, and even so the effects of this type of incentive are limited in an open list system, as Miguel (2004) points out.

In terms of female chances in these elections, it should also be noted that there has been a reduction, over these five elections, in the chances of being elected as a male candidate. This reduction can be observed between 2000 and 2008, followed by a significant reduction in 2012. This could be related to the fact that Dilma Rousseff won the 2010 election, becoming the first woman to occupy the presidency of the republic the following year. Once again, however, we can see that there was a reduction in this inequality between 2012 and 2016.

In terms of a college degree, there’s one surprising result. Even though it was expected that this variable would have a strong relationship with the dependent variable analyzed, it can be seen that in some years it was discarded by applied statistical techniques or presented a negative influence, or in other words, a university diploma has not, in thesis, influenced the electoral success of candidates. This result is surprising and opens the possibility of studying this research area in greater depth considering the fact that the level of education of the victorious candidates in the elections from 2000 to 2016 presented a constant increase, with 52.6% of the elected mayors in 2016 having a college degree. From this it could be inferred that there is a supposed relationship between the personal capital that people use to enter politics and their economic conditions and level of education. Based on the assumption that higher education in Brazil is still only accessible to wealthier members of the population, education could be related not directly to a candidate’s chances of being elected, but rather to the conditions that are capable of enabling the more substantial accumulation of political capital, which could come in the form of a greater availability of time to dedicate to political issues in their surroundings, having better access to information, having more extensive relationship networks, having abundant economic capital, belonging to more influential families, etc. In this sense, the growing education of candidates could be an indicator that the political field has become more closed to new entrants, with greater barriers to achieving electoral success. This signifies that there is an increasingly larger demand for education in order to be elected to higher posts in party structures and public organizations, with this demand being even greater for women than men, which can be seen in the more recent indices of college degrees for mayors in general and mayors elected in 2016.

Finally, the political career variable, as treated, demonstrates that the political field in Brazil was more closed from 2000 to 2008. The results indicate that the chances of a candidate who belongs to this group rose significantly during these three elections, reaching a peak of 139.8%. During this period, the political field in Brazil was very strong, driven by a very favorable and optimistic economic scenario — remembering here that the crisis of 2008 was set off by the bankruptcy of Lehman Brothers on September 15 that year, but this did not generate direct repercussions in the elections held on October 5 and 26 in Brazil. During this period, there was a growing participation of “professional politicians” among the victorious candidates, those who were reelected or already occupied public offices, such as councilmen, municipal secretaries, labor leaders or offices in other organizations of a public nature.

However, after 2008, there was a notable reduction in the chances of candidates with political careers, reaching a value of 28.1% in 2016. Here we can perceive a greater acceptance of new entrants, with some candidates in large capitals, like João Doria in São Paulo and Alexandre Kalil in Belo Horizonte, having benefitted from discourses that denied that they were members of the political class.

Even so, the numbers demonstrate that the Brazilian political field still has some level of sedimentation and stability, which is demonstrated by the elevated chances of victory for candidates who have a political career. This means that the entrance of new actors capable of producing meaningful change in the political field depends on overcoming countless barriers, which makes the inequalities necessary for accumulating political capital even more evident. The great issue is whether to wait for these barriers to become more republican, or in other words, for them to be related to the quality of the proposals and actions of people who are interested in entering politics, rather than the shameful practices of politics based on patrimonialism, the perpetuation of powerful family groups, and the private appropriation of the State.

REFERÊNCIAS

• Araújo, C; & Borges, D. (2013, junho). Trajetórias políticas e chances eleitorais: analisando o “gênero” das candidaturas em 2010. Revista de Sociologia Política, Curitiba, 21(46), 69-91.
• Arretche, M. T. S. (2000). Estado federativo e políticas sociais: determinantes da descentralização Rio de Janeiro: Revan; São Paulo: Fapesp.
• Best, H; & Cotta, M. (Ed.). (2000). Parliamentary representatives in Europe 1848-2000 Legislative Recruitment and Careers in Eleven European Countries Oxford: Oxford University Press.
• Bordignon, R. R. (2013). Candidatos e eleitos: recrutamento e chances de sucesso nas eleições de 1998 e 2010. Revista Paraná Eleitoral, 2, 35-62.
• Bourdieu, P. (1989). O poder simbólico Lisboa: Difel.
• Bourdieu, P. (1996). Razões práticas: sobre a teoria da ação (5^a ed.). Campinas: Papirus.
• Bourdieu, P. (2011). O campo político. Revista Brasileira de Ciência Política, 5, 193-216.
• Campos, L. A. (2015). Socialismo moreno, conservadorismo pálido? Cor e recrutamento partidário em São Paulo e Rio de Janeiro nas eleições de 2012. DADOS - Revista de Ciências Sociais, Rio de Janeiro, 58(3), 689-719.
• Codato, A; Cervi, E; & Perissinotto, R. (2013). Quem se elege prefeito no Brasil? Condicionantes do sucesso eleitoral em 2012. Cadernos Adenauer, Rio de Janeiro, 14(2), 61-84.
• Freire, A. (2002). A profissionalização política dos deputados portugueses. Revista de Sociologia da Universidade do Porto, 1(12), 27-56.
• Gaxie, D. (1980). Les logiques du recrutement politique. Revue Française de Science Politique, 30(1), 5-45.
• Horta, C; Silva, M; Rocha, M; & Dufloth, S. (2017). Perfil dos prefeitos eleitos nos municípios brasileiros: estudo evolutivo das eleições realizadas no período entre 2000 e 2012 2012 (Textos para Discussão, n. 4). Belo Horizonte: Fundação João Pinheiro. Recuperado de http://www.fjp.mg.gov.br/index.php/docman/texto-para-discussao/703-td-no4-claudia-horta/file
  » http://www.fjp.mg.gov.br/index.php/docman/texto-para-discussao/703-td-no4-claudia-horta/file
• Marvick, D. (1968). Political recruitment and careers. In D. L. Sills, International encyclopedia of the social sciences (pp. 273-282). Nova York: Macmillan.
• Matthews, D. R. (1984, November). Legislative recruitment and legislative careers. Legislative Studies Quarterly, 9(4), 547-585.
• Miguel, L. F. (2003, junho). Capital político e carreira eleitoral: algumas variáveis na eleição para o Congresso Brasileiro. Revista de Sociologia e Política, Curitiba, 20, 115-134.
• Miguel, L. F. (2004). Participação eleitoral e gênero no Brasil: as cotas para mulheres e seu impacto. In Anais do 2o Congresso da Associação Latino-Americana de Ciência Política (pp. 1-19), Cidade do México: Alacip.
• Norris, P. (1993). Conclusions: comparing legislative recruitment. In Norris, P; & Lovenduski, J. (Ed.), Gender and party politics (pp. 309-330). Londres: Sage.
• Norris, P. (1997). Passages to power: legislative recruitment in advanced democracies Nova York: Cambridge University Press.
• Pinheiro, L. S. (2007). Vozes femininas na política: uma análise sobre mulheres parlamentares no pós-Constituinte Brasília: Secretaria Especial de Políticas para Mulheres.
• Sanbonmatsu, K. (2006). The legislative party and candidate recruitment in the American states. Party Politics, 12(2), 233-256.
• Urbinati, N. (2006). O que torna a representação democrática? Lua Nova, São Paulo, 67, 191-228.

• 1
  Bourdieu (2011) is referring to French presidents, who often occupy the position of secretary general in their respective parties. In Brazil, parties are led by presidents which are the equivalents of secretary generals.
• 2
  Includes those who declare “read and write” and “incomplete primary education.”
• 3
  The methodological procedures used to construct these occupational groups and subgroups can be found in Horta et al. (2017).
• 4
  The relative distribution in terms of occupational groups and subgroups can be found in Horta et al. (2017).
• 5
  Remember that most of the candidates classified in this subgroup declared their occupation as “mayor.”
• 6
  In the 2012 and 2016 elections the college degree variable did not have statistical significance.
• 7
  The mini-electoral reform of 2009 (Law n. 12,034) established in Paragraph 3 of Article 10 of the Election Law (Law n. 9,504/1997) states that “each party or coalition has to fill at least 30% and at most 70% for candidates of each sex.”

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

Publication Dates

• Publication in this collection
  Jan-Feb 2019

History

• Received
  28 July 2017
• Accepted
  22 Mar 2018