Winning elected office for the first time usually takes quite a while – and most people who try never make it. In Brazil, the scale of this competition is striking: in the 2024 elections, about 460,000 people ran for city council, whereas 58,435 were elected. Those who decide to pursue a political career in Brazil typically spend years before securing a seat: they must first join a party, later become candidates, seek a place on an electoral slate, and only then – if everything aligns – be elected (GATTO and THOMÉ, 2024). Many set out on this path, but because it requires sustained resilience over time, a large share ultimately drops out along the way.

In the 2008 elections, which constitute the reference year for our analysis, more than 50,000 city councilors were elected nationwide. Women accounted for about 12.5% of them, while men made up the remaining 87.5%. In the years since, women and men have followed a variety of paths: some have chosen not to run again, others have sought re-election and succeeded, others fallen short, and still others have pursued higher office – with outcomes ranging from success to failure.

Scholarship on gender and politics has consistently shown – both theoretically and empirically – that women’s experiences in political life differ markedly from those of their male counterparts. These differences stem not only from gendered socialization (BIROLI, 2016), but also from the obstacles women face in party nomination processes and informal party practices (GATTO and WYLIE, 2022), in securing resources (SACCHET et al., 2025; SACCHET and SPECK, 2012), in navigating party leadership (GATTO and THOMÉ, 2020; THOMÉ, 2021), as well as from the institutional design of the political system in which competition unfolds (ARAÚJO, 2009, 2001). Scholars working within feminist institutionalism (MACKAY, KENNY, and CHAPPELL, 2010) further argue that the gendered biases embedded in political institutions mean that women’s experiences in politics are frequently shaped by discrimination, which all too often drives them away from arenas of power.

In Brazil, the combined effect of these multiple factors is that, even after the adoption of policies intended to reduce gender inequality, such as candidate quotas, women still held only 17.7% of the seats in the Chamber of Deputies and 18% in the Federal Senate in 2025. At the local level, women accounted for 18% of city councilors and roughly 13% of mayors.

Brazil’s standing relative to other countries underscores the magnitude of this disparity: the country ranks 134th worldwide in women’s political representation – measured by the share of seats in the lower house – and holds the lowest position among Latin American countries (INTER-PARLIAMENTARY UNION, 2024). As a result, a significant body of research has sought to explain the reasons behind women’s persistent underrepresentation in these spaces, particularly through studies focused on descriptive representation¹ at the national level (ARAÚJO, 1998; FEITOSA, 2012; MARQUES, CELINI and SANTOS, 2021; MIGUEL and BIROLI, 2010; REZENDE, 2017).

In this study, we focus on the political trajectories of women candidates, with particular attention to local politics, the level at which barriers to entering and remaining in ‘formal’ politics are generally assumed to be lower (BORCHERT, 2011; MIGUEL, 2003). This article primarily addresses the following questions: What career patterns emerge among women elected to municipal office? How do they differ from those of men? Drawing on these data, we examine a dimension we view as essential to women’s political participation: their capacity to remain resilient throughout their political careers (GATTO and THOMÉ, 2020).

We employ an analytical model that follows the long-term trajectories of these women representatives, in contrast to traditional studies of political ambition, which typically provide a snapshot in time – focusing on career ‘decisions’ rather than the ‘careers’ themselves (CODATO et al., 2022; SAMUELS, 2000; SANTOS, 1999). To that end, we adapt the model developed by Gelape (2022) to identify career patterns among city councilors elected in 2008, taking into account both their prior trajectories and the paths they subsequently pursued.

In doing so, we contribute to the literature on gender and politics while also engaging with scholarship on political elites and political careers (SAINZ, 2024), both in general and at the local level. Within the first body of work, we highlight the factors associated with women’s electoral careers, especially the development of resilience among candidates over time (GATTO and THOMÉ, 2020; MARQUES, 2012; MARQUES, CELINI and SANTOS, 2021; PINTO and SILVEIRA, 2018).

As for the literature on political careers, we offer two main contributions. First, our study is among the few that explicitly foreground the role of time in shaping political careers in the Brazilian context (LIMA, 2017; PINTO and SILVEIRA, 2018). Second, we contribute by analyzing these trajectories at the subnational level, a field that is still developing in Brazil (MESSIAS, 2019; VIEIRA, 2019; WYLIE, 2020) as well as in the comparative literature (FOLKE and RICKNE, 2016; SCHWINDT-BAYER, 2011).

Our results show that women city councilors are disproportionately represented among the less successful career paths in local politics. In regression models that include controls for other predictors of political careers – such as municipality size and prior professional experience – we find that women councilors have a higher predicted likelihood than men of either abandoning a political career or running for re-election and losing. Taken together, these findings provide clear evidence of gender disparities in the career trajectories of city councilors.

As noted earlier, this article examines the electoral ‘career trajectories’ of city councilors rather than focusing exclusively on individual ‘career decisions’. Accordingly, we begin by showing how the trajectories of city councilors are classified and how this approach differs from analyses centered on individual career decisions.

Our universe of analysis consisted of 51,478 city councilors elected in 2008⁷. In addition to their electoral results in that year, we identified every office they contested in general and municipal elections between 2002 and 2022, along with the corresponding outcomes. Based on this information, we classified their trajectories before and after 2008 according to the typologies outlined in the research design section.

Figure 01 shows that the largest share of city councilors elected in 2008 (37.3%) falls into the ‘veteran’ category, that is, those who ran for and won at least one city council election in the subsequent 12 years. The next largest group is composed of the ‘resilient’ (23.9%), who continued to run in elections despite not securing additional victories. The third group (20.9%) consists of those who did not run for any office after their 2008 victory, the ‘retirees’. Fewer than 20% of the councilors elected in 2008 went on to run for municipal executive office (either mayor or vice-mayor). Among them, the largest share is made up of the ‘contenders’, who ran but were not successful (9.8% of the total). Finally, the ‘achievers’, who ran for municipal executive office and won, represent the smallest group (8.1%).

Figure 01 Source: Elaborated by the authors using data from the TSE (2024).

In other words, among the city councilors elected in 2008, nearly 40% served at least two terms on the city council up to 2020. Moreover, nearly 80% of them (all except the ‘retirees’) ran in at least two elections. This pattern points to a certain degree of commitment to pursuing a political career at the local level (GELAPE, 2022). In addition, nearly 20% of these city councilors also ran for positions in the local executive branch. These bids for higher office should not be interpreted as mere personal whims, since running for mayor or vice-mayor requires foregoing the opportunity to seek re-election to the city council. Finally, it is worth noting that more than 20% of them did not run for office again. Thus, a substantial share – one in five candidates – chose to leave politics even after serving an elected term.

For comparison, what would these findings look like if, instead of considering the 12-year period following the 2008 election, we focused only on the next electoral cycle (2012) and analyzed the ‘career decisions’⁸ of these city councilors? First, we would observe a substantial shift in the largest category, which would now consist of the ‘retirees’ (38%). They would be followed by councilors who were re-elected (32.2%), corresponding, within our typology, to the ‘veterans’, although they would represent a smaller share of the total than in our career-based analysis. Next, we would observe a similar share of councilors who were unsuccessful in their bids for re-election (23.1%), which corresponds to our ‘resilient’ category. Other notable differences would appear among those who ran for municipal executive office: unsuccessful candidates would outnumber successful ones (3.8% versus 2.8%), but both groups would remain well below the shares observed for the ‘contenders’ and ‘achievers’ categories in our career-based classification.

Figure 02 shows how classifying politicians based on ‘career decisions’ alone can obscure patterns that become visible when their full ‘career trajectories’ are considered. In the figure, each bar represents the percentage of candidates in each trajectory category when classified according to their career decision. For example, if we look specifically at the 2012 decisions of city councilors who, according to our career trajectories typology, are classified as resilient, nearly a quarter (25%) would be classified simply as ‘retirees’. Likewise, when we look at the veterans, who ultimately went on to win another city council election in the long-term analysis, roughly 30% would be classified either as retirees or as unsuccessful candidates (analogous to the resilient category in our typology). Thus, analyzing political careers solely on the basis of career decisions in the subsequent electoral cycle obscures broader patterns in these individuals’ career trajectories (GELAPE, 2022).

Figure 02 Source: Elaborated by the authors using data from the TSE (2024).

Another important point is that focusing on ‘career decisions’ tends to underestimate trajectories toward municipal executive office. In local political careers, the mayoralty represents the highest position (GELAPE, 2022). However, running for this office requires candidates to forgo a potential bid for re-election to the city council and generally depends on the accumulation of local political capital, since it is a single-seat office with far fewer opportunities than the multiple seats contested in city council elections. It is therefore reasonable to expect that candidacies for executive office emerge gradually over time, as political capital accumulates and political careers develop, rather than resulting from a single ‘career decision’.

In summary, we find that a large share of city councilors pursues local political careers (GELAPE, 2022), whether within the city council itself (with or without success) or through bids for municipal executive office. This occurs even though a significant fraction of them (around 20%) did not run for office again after their 2008 victory. We also find that, compared with analyses based solely on career decisions, examining long-term trajectories leads to different inferences, particularly by identifying fewer retirees and a larger number of individuals who run for municipal executive office.

In addition to examining subsequent trajectories – that is, from 2008 onward – the literature on political careers also emphasizes the importance of candidates’ prior trajectories. How are the city councilors elected in 2008 distributed across both their prior and subsequent trajectories? Figure 03 presents these findings.

Figure 03

Note: The Executive category (n = 8; 0.02% of observations) was removed for legibility.

To assess the effect of gender on the subsequent trajectories of city councilors elected in 2008, we estimated ordered logistic regression models. We selected this approach because our dependent variable – the subsequent trajectory of these councilors – is ordinal, meaning that its categories follow a ranked order from lowest to highest (in this case, from ‘retirees’ to ‘achievers’).

In addition, we tested this association using three different models. The first tested the relationship between subsequent trajectory and gender, without including any additional control variables. The second added two control variables: the politician’s prior trajectory and municipality size. The third model built on Model 02 by incorporating additional controls: the councilor’s level of education, marital status, age, and party ideology – the latter two categorized according to the criteria established by Silva (2014) and Codato, Berlatto, and Bolognesi (2018), respectively – as well as fixed effects for the state where the councilor was elected.

Interpreting the results of logistic regression models can be challenging. To facilitate interpretation, rather than presenting a table of regression coefficients⁹, we present a figure that shows the predicted probability of each category of the dependent variable by gender. In other words, the figure displays, for each individual in the dataset, the average predicted probability of being classified into each subsequent career trajectory category under each estimated model. These results are shown in Figure 06.

Figure 06

Note: In Model 03, standard errors are robust and clustered at the municipality level (this variable is not included in Models 01 and 02).

By examining the political careers of city councilors, this study sought to address gaps in the literature on gender and politics and on local politics by providing evidence on the ‘career trajectories’ of women in municipal politics beyond isolated ‘career decisions’. As the literature review shows, few studies have examined women’s presence at the subnational level together with their political careers through longitudinal analyses covering extended periods. This remains a developing field that offers ample opportunities for further research (SAINZ, 2024).

By examining the trajectories of city councilors elected in 2008, we find that women are disproportionately represented in the less successful trajectory groups when predicting their subsequent electoral trajectories. They have a higher probability of withdrawing from electoral competition (‘retirees’) or failing to secure another term on the city council (‘resilient’). Men, in turn, are more likely to follow successful trajectories, either by winning another term as city councilors (‘veterans’) or by running for municipal executive office, whether unsuccessfully (‘contenders’) or successfully (‘achievers’).

Because city council seats are the most available and accessible elective offices in the Brazilian political system (BORCHERT, 2011), the presence of gender inequality at this stage points to bottlenecks in women’s participation in Brazilian politics, even at a level where such barriers might, in principle, be easier to overcome. It is important to note that career choices at the local level are not merely individual decisions but are also constrained by political parties (GATTO and THOMÉ, 2024; GELAPE, 2022). For women candidates, this challenge is even greater, given their often difficult relationship with political parties (THOMÉ, 2021). In the Brazilian context – where gender quotas make it relatively easy to secure a place on a party ticket – decisions to withdraw from electoral competition may nevertheless be closely tied to candidates’ awareness of this hostile environment (PISCOPO, 2019).

As is well known, Brazil still has relatively low levels of women’s political representation compared with other countries in the region. We hope this study contributes to improving the diagnosis of this problem and to identifying ways to mitigate these inequalities. In addition, we emphasize how these inequalities unfold over time, suggesting that researchers should also consider this temporal dimension.

Beyond the subfields of political careers and gender, we believe our approach also offers a methodological contribution to the study of political careers in Brazil and elsewhere. When we speak of ‘career trajectories’, it is important to consider a sequence of political decisions rather than focusing on a single point in time, as is common in many studies of ‘political ambition’. We therefore hope that scholars consider this factor when designing research to examine other cases.

Finally, we acknowledge several limitations in our analysis, including constraints related to data availability. For example, because we chose to maximize the temporal window of the analysis, we are unable to examine gender from an intersectional perspective, as the race variable has only been reported by the Superior Electoral Court (TSE) since 2014¹⁰. In addition, studying political careers presents inherent challenges, as they constitute a highly endogenous phenomenon.

Thus, while we find evidence of gender inequality, we do not provide a causal analysis of the factors that may produce or mitigate such inequality. Another promising direction for future research would also be to explore potential heterogeneous effects across variables such as municipality size and candidates’ sociodemographic characteristics, such as race and ethnicity, among others.