Try The Grand Continent. The ideas and analyses you won’t find anywhere else. But you’ll see them everywhere soon. Delivered to your inbox starting at 8 euros per month
The Brazilians will vote this Sunday, October 4, and if no candidate wins outright, in three weeks for a runoff that will inevitably be decisive. This electoral contest is shaping up to be particularly tight: 45.1% for Lula, 40.6% for Flávio Bolsonaro in the first round, according to our poll conducted for The Grand Continent . The runoff appears even more uncertain: 50.4% for Lula against 49.6% for Bolsonaro. The 0.8-point gap in the tested head-to-head is well below the margins of error, signaling a profoundly undecided vote between the incumbent president and the son of Brazil’s former president .
To begin, a few notes on the main players. Luiz Inácio Lula da Silva, known simply as Lula, is by far the best identified figure among them. He is, as is known, a historic figure of the Brazilian left. A former metalworker and union leader, he presided over the country from 2003 to 2010, before returning to power in 2023. As the candidate of the Workers’ Party (PT), he defends the central role of the state in reducing inequalities and in fostering economic development. His program emphasizes raising the minimum wage, strengthening public services, investing in infrastructure, and pursuing a proactive industrial policy. He also proposes a reduction of the weekly working hours to forty hours and pursues the goal of an ecological transition.
Facing Lula, the Bolsonaro name will again be in the race, but not with the same candidate as in 2022. This time it is Flávio Bolsonaro, a senator from Rio de Janeiro and Jair Bolsonaro’s eldest son, who is ineligible to run for his father’s seat. A candidate of the Liberal Party, Flávio Bolsonaro positions himself in the political and ideological continuum of his father. His program combines a reduction of public spending and taxes with a hardening of criminal policy: strengthening high-security prisons, combatting criminal organizations, and tightening penalty reforms for the most serious crimes. On social issues, he opposes expanding abortion rights and proclaims the defense of Christian values. The party name alone would not suffice for a French reader: his program actually blends economic liberalism, moral conservatism, and a tougher approach to criminal justice.
Other candidates listed in the poll register far lower levels of support and are not in a position to qualify for the runoff. With 4.9% of voting intentions, Augusto Cury looks set to be the first of the “minor candidates.” A psychiatrist and writer, he represents Avante and presents his participation as a response to polarization, which already reveals the deep tensions characterizing Brazilian society. Mental health and emotional education occupy a notable place in his program, alongside entrepreneurship and budgetary balance. This discourse clearly distinguishes him from all other candidates
Renan Santos, at 4.0%, is one of the co-founders of the Free Brazil Movement, known notably for its involvement in the mobilizations against the former leftist president Dilma Rousseff. Candidate of the Missão party, he advocates a distinctly autonomous, far-right platform. His program notably includes very high-security prisons inspired by the Salvadoran model and the replacement of Bolsa Família—the emblematic social assistance program introduced during Lula’s first presidency—by a system of work paid for the community. Finally, Ronaldo Caiado, at 3.8%, and Romeu Zema, at 1.4%, emphasize their experience governing a federated state. Caiado, a physician and former governor of Goiás, represents the PSD (Partido Social-Democrático), a center-right party whose name does not correspond to European social democracy. He stresses security, the fight against resources of organized crime, and the need to develop the country’s infrastructure, with a significant emphasis on agriculture. Zema, a businessman and former governor of Minas Gerais, represents the Novo party, a right-liberal group. His program advocates privatizations, reduced public spending, and greater flexibility in labor regulation. Taken together, these two candidacies are right-leaning and, although they share some themes with Flávio Bolsonaro, they differ in background and proposals from the son of the former president.
Understanding the electoral demand through value systems
The concentration of voting intentions around Lula and Flávio Bolsonaro is very strong in the first-round scenario: together they capture 85.7% of expressed choices. And in the second-round matchup, the declared intentions are distributed almost evenly between the two candidates. How can one understand this near-parity in a society described as highly polarized? Which groups back each candidate? What factors explain the two electoral coalitions that will emerge on Sunday?
To attempt an answer, we rely on a study by Cluster17, conducted from January 15 to May 17, 2026, with 3,546 people, focusing on building a typology of the Brazilian electorate. This study helps identify the major systems of values and the groups that carry them. It will thus enable a better understanding of how the two large camps organizing the Brazilian electorate are structured.
Individuals’ positions on the major divisions that cut through a society are indeed among the most determining factors of their political choices. Opinions on highly divisive issues—such as women’s rights and the rights of sexual minorities, religion, security, social welfare, wealth distribution, the role of the state, the environment, and the functioning of institutions—are strongly associated with electoral preferences. Moreover, these attitudes tend to be relatively stable over the medium term. During an electoral campaign, one rarely moves from a firm opposition to abortion to full acceptance, nor from a strong stance in favor of redistribution to positions against welfare on economic grounds. These attitudes fit into broader opinion systems, themselves linked to social and cultural identities characterized by a high degree of inertia.
Starting from this principle, the method used in this study aims to identify the major cleavages that divide Brazilian society, then to position individuals relative to these cleavages. To do so, a test comprising 30 propositions was administered to a representative sample of 3,546 Brazilian voters. These propositions were selected for their ability to generate meaningful divisions and to cover the main issues likely to structure society, such as mores, religion, security, authority, redistribution, economic liberalism, the environment, the Amazon, indigenous peoples, institutions, and foreign policy.
The statistical processing of these responses makes it possible to identify groups of individuals, called “clusters,” who share the same system of opinions, or, in other words, the same attitudinal configuration. This common politico-ideological sensibility does not imply that all members of a cluster think exactly the same, but that their responses organize themselves along sufficiently close logics for them to be distinguishable from other groups. Based on these value systems, it then becomes possible to understand why some segments move en masse toward a single political offering, but also to identify the contested zones between candidates and the formation of electoral coalitions.
This latter dimension is essential. No electorate is homogeneous, whether sociologically or politically. Gathering tens of millions of voters requires assembling groups that share certain fundamental preferences, yet may diverge strongly on other topics. The cluster-based approach precisely identifies what holds an electoral coalition together, but also the tensions that run through it and the groups that might drift away if certain themes and issues became central during the campaign.
A multipartite democracy structured by a strong presidential bipolarization
To understand the current structure of the Brazilian electorate, one must place it in the context of the end of the military dictatorship. The democratic transition of 1985, followed by the adoption of the 1988 Constitution, restored political and civil liberties, established new social rights, and laid the institutional foundations of the “New Republic.”
The system that emerged from this process cannot be described as perfectly bipartite, even if the presidential election is typically framed as a duel between two broadly dominant candidacies. Brazil is characterized by a very strong party fragmentation and the need to form broad parliamentary coalitions. Yet this fragmentation coexists with a presidential competition that is far more concentrated around two major political forces. From 1994 to 2014, candidates from the Workers’ Party (PT) and the Brazilian Social Democracy Party (PSDB) captured between 70% and 90% of the votes in presidential elections, while legislative competition remained largely fragmented.
This bipolarization gradually took on a social and territorial content of increasing clarity. The work of Amory Gethin and Marc Morgan shows, in particular, the transformation of the PT after Lula’s 2002 election and its ascent to power in 2003: initially more supported by younger, educated, relatively well-off populations in the south, the party gradually strengthened its foothold among poorer, less educated groups and in the Northeast, as Lula’s redistribution policies, including the Bolsa Família program, shaped a new electoral cleavage.
The 2018 election upended the other half of the system. The PSDB ceased to be PT’s principal adversary, while Jair Bolsonaro organized and brought together a new right-wing coalition. Works analyzing candidates’ programs indeed identify a sharp increase in ideological polarization that year. This political reconfiguration reflects and accompanies deep transformations in Brazilian society.
The first is religious. Brazil remains predominantly Catholic, but to a lesser extent than before. Between 2010 and 2022, the share of Catholics fell from 65.1% to 56.7%, while the share of evangelicals rose from 21.6% to 26.9%. At the same time, the proportion of people of no religion also increased slightly, from 7.9% to 9.3%.
Brazil thus experiences both a rise in evangelical influence and a slight secularization. The relationship between religious practice and electoral behavior has become particularly visible since 2018: several studies highlight greater support for Bolsonaro among evangelical voters, somewhat analogous to observations in the United States with Donald Trump, as well as the role of regular attendance at worship venues and the statements of religious leaders in mobilizing voters.
These religious transformations occurred alongside major social upheavals. In the 2000s and early 2010s, Brazil saw a sharp reduction in poverty, rising incomes, and significant social mobility. According to the World Bank, the Gini coefficient fell from about 0.58 to 0.53 between 2001 and 2013, while tens of millions of Brazilians improved their economic standing. Yet these gains did not erase the historically stark inequalities that characterize the country. Brazil is thus today more urban, more religiously diverse, and socially transformed, but still deeply unequal.
A macro attitudinal cleavage that orders Brazilian society around morality, authority, security, and wealth redistribution
The statistical analysis of Brazilians’ positions on the 30 tested measures allows the identification of five major axes of division. However, their relative importance is highly uneven. In the decomposition used, the first axis accounts for 74% of the total contribution, versus 11% for the second, 7% for the third, 5% for the fourth, and 3% for the fifth. In other words, a single axis structures a large portion of attitudes: questions about mores, authority, the economy, or attitudes toward institutions tend to align to produce a single line of fracture within a society that itself is composed of a plurality of competing stakes.
On one side of this cleaving axis lie individuals favorable to minority sexual rights, to substantial state intervention in the economy, to wealth redistribution, to strengthening certain democratic institutions, to protecting the Amazon and indigenous lands, and to closer cooperation with the BRICS. On the other end cluster, there is a concentration of moral conservatism, defense of authority and increased policing prerogatives, the right to bear arms, privatizations, shrinking the public sector, limiting welfare benefits, and a foreign orientation increasingly oriented toward the United States.
The result is striking: in Brazil, major economic, cultural, security, and institutional divisions tend to overlay on the same axis. The political consequences are spectacular. This organization yields two sets that are almost perfectly balanced numerically and starkly opposed. The nine clusters located to the left of the main axis comprise about 51% of the population, and the seven clusters on the right roughly 49%.
In total, sixteen clusters span the axis almost perfectly. At one extreme lie the Social-Progressives, the Eco-Solidaries, and the Libertarians, who gather the most progressive, most secularized, and most supportive of redistribution, minority rights, and Amazon protection. At the other pole, the Traditionalists and the Authoritarians concentrate the most conservative mores, security-minded attitudes, and support for economic liberalism. Between these two poles lie more intermediate profiles, often more distant from politics, more distrustful of institutions, or traversed by more composite value systems. It is noteworthy that these latter clusters sit slightly left of the axis’s median. Whereas in other countries studied, the “disengaged” and apolitical groups lean toward authoritarianism and social conservatism, the Brazilian case shows more nuance, with two clusters, the Defiants and the Moderates, lying more on the progressive side, although, as we will see, that does not automatically guarantee PT votes.
The size of the clusters also confirms the highly polarized nature of Brazilian society. The groups at the two ends of the axis are the largest: the Social-Progressives account for 9% of the population alone, while Traditionalists and Authoritarians together number 9% and 10% respectively. At the progressive end, Social-Progressives, Eco-Solidaries, and Libertarians together make up 21% of the population; at the other end, Liberals, Traditionalists, and Authoritarians comprise 26%. Between these two poles that together gather almost half of the electorate, most clusters weigh between 4% and 7%, painting a landscape not of a patchwork of small sensitivities but of a society organized around two powerful poles that tend to attract swing or more moderate groups into their orbit.
This strong overlay of the cleavages is an important characteristic of Brazil that is not found in most societies studied to date. In many democracies, recent transformations have instead favored an increasing dissociation between economic and cultural conflict, or the emergence of other cleavages around attitudes toward “elites” and institutions. In our survey, Brazil appears less ideologically fragmented and, for this reason, strongly bipolar: being conservative on mores tends to correlate with economic liberalism, support for arms, and a more repressive view of public order; conversely, being progressive culturally is often associated with higher demands for redistribution, minority protection, and environmental concerns.
The land rights and environment issue: a transversal stake
A second cleavage emerges, adding a degree of complexity to the structuring of attitudes and thus to electoral orientations. However, this second axis is markedly weaker, which explains why it cannot disrupt the overall ordering of the Brazilian electorate. It is particularly structured by the relationship to the Amazon, to indigenous populations, and to environmental constraints. The prohibition of deforestation and the expansion of protected lands for indigenous peoples are among the variables contributing most to this axis. Yet, on these issues, the left-right distinction becomes considerably less mechanical. This complexity does not pertain to the progressive space, which remains globally homogeneous on these issues. Ecology tends to unite the left. It is mostly to the right that it begins to blur the grand system of opposition described above.
Certain clusters situated in the conservative space prove to be strongly attached to environmental protection. The Eco-Liberals, for example, combine a marked economic liberalism—with 91% of them in favor of privatizing Petrobras and 98% in favor of cutting corporate taxes—with a pronounced ecological sensitivity: 90% want to ban any deforestation in the Amazon and 64% would expand protected lands for indigenous peoples. Conversely, the Restorers, who also occupy the conservative space, stand out for a strong opposition to these measures: only 6% of them want to ban deforestation and none support expanding lands protected for indigenous peoples. The environment thus becomes one of the main differentiating factors within the same ideological blocs. It less mechanically pits left against right than it does issues such as arms, privatization, policing, or certain moral questions.
A broadly shared political mistrust that expresses itself differently depending on the targeted institutions
If the cluster method identifies cleavages, it also logically identifies areas of consensus. While in many Western countries we observe a significant split regarding trust in elected officials and institutions—with one side favoring rupture and distrust, and the other favoring continuity and institutional order—Brazil stands out for a generalized mistrust of political personnel. Indeed, 90% of respondents support the idea of paying parliamentarians the average salary of a Brazilian worker. This proposal is majority-supported in each of the sixteen groups and exceeds 90% in many of them. Likewise, 75% of Brazilians want to give voters the option to recall an elected official through a referendum. There exists, therefore, a transversally anti-elite or anti-political-disturbance sentiment that transcends the classic left-right divide and tends to create a broad consensus in the country.
But this consensus ends when one moves from judging the political class in general to analyzing institutions that themselves have become targets of political confrontation. This is especially striking for the Supreme Court. While 56% of respondents want to empower it to defend democracy, this proportion rises to 92% among Social-Progressives, 82% among Eco-Solidaries and Libertarians, and 98% among the Discontents. It drops to 19% among Conservatives, 17% among anti-welfare advocates, 13% among Liberals, and 0% among Restorers and Authoritarians—a very low level within the most conservative groups.
The distrust of elites is thus a fairly cross-cutting sentiment. However, the relationship to the institutions that organize and arbitrate political conflict is highly polarized.
Gender, age, and religion partially structure value universes
The value systems described above are not independent of individuals’ social characteristics. They tend to be anchored in distinct sociological universes. Gender, education level, religious affiliation and practice all contribute to increasing the probability of having internalized certain value systems and, more generally, to positioning oneself on one side of the macro-attitudinal cleavage that structures Brazilian society.
The gender composition within the clusters reveals a fairly clear gender gap, albeit not perfectly linear. At the progressive end of the spectrum, the most feminized groups are the Social-Conservatives (69% women), the Humanists (68%), and the Revolted (67%). The Social-Progressives and Eco-Solidaries are also predominantly female, at 59% and 57% respectively. Moving toward the conservative pole, the composition becomes much more masculine: men account for 61% of Anti-Welfare advocates and Traditionalists, 62% of Restorers, 64% of Authoritarians, and up to 76% of Liberals.
This result aligns with what comparative literature terms the “modern gender gap.” While women were historically more conservative than men in Western democracies, this differential has gradually reversed in many countries: women are now more often found on the left and adopt more progressive positions on several cultural and social issues. The work of Ronald Inglehart and Pippa Norris links this evolution to generational changes, rising education levels, greater female autonomy, and shifting gender norms. However, more recent research cautions against universalization; the gap varies strongly by country and generation, though it is visible in a large portion of democracies studied.
Brazil follows this trend. In the 2018 election, Matthew Layton, Amy Erica Smith, Mason Moseley, and Mollie Cohen already established that, for comparable characteristics, women were less likely than men to vote for Jair Bolsonaro. Their analysis found gender, alongside race and religion, to be among the particularly salient dimensions in that election. Studies of the 2022 election confirmed this gap in Lula vs. Bolsonaro support and showed that it was closely linked to ideological orientation and attitudes toward feminism.
Religion is one of the sociological dimensions most strongly associated with clustering, but in a country with a relatively small share of “no religion” individuals, merely identifying as Catholic does not by itself create a clear fracture line. Catholics are present in vastly different value universes: they account for 89% of the Humanists, but also 65% of the Traditionalists, two groups positioned on opposite sides of Brazil’s grand cleavage. The case of the Humanists is particularly revealing: overwhelmingly Catholic and practicing, they nonetheless stand out for open positions on several social issues. Indeed, 94% of them support same-sex couple adoption and 92% oppose bans on gender changes. Catholic affiliation thus partially transcends the border between progressivism and conservatism.
The level of religious practice, however, seems far more tightly linked to value universes. At the progressive extreme, only 11% of Social-Progressives, 19% of Libertarians, and 22% of Eco-Solidaries and Revolts declare themselves believers and active practitioners. At the opposite end of the spectrum, this proportion rises to 54% among Conservatives, 55% among Restorers, 69% among Traditionalists, and 51% among Authoritarians. Based on such findings, evangelicals and conservative Catholics appear more exposed to direct church influence and maintain a more intense relationship with faith. By contrast, non-belief presents a generation-related dimension: as in many other countries, younger generations are more exposed to secularization and to a distancing from religious institutions. Research on Brazil also shows a close association between secularization and more liberal attitudes on moral and social issues. A recent survey of young Brazilian adults reveals more liberal positions among non-believers concerning sexuality and intimacy. This dynamic fits into a broader phenomenon often interpreted as a generational renewal effect: secularization accompanies, without mechanically determining, a broader diffusion of values of individual autonomy and tolerance toward different lifestyles. In our typology, this relationship between secularization and cultural liberalism appears particularly clear: non-believers represent 56% of Social-Progressives and 55% of Libertarians, versus only 6% of Conservatives and Restorers, 0% of Traditionalists, and 10% of Authoritarians.
The Evangelical movement draws an even sharper dividing line. Evangelicals represent only 7% of Social-Progressives, 8% of Eco-Solidaries, and 9% of Libertarians, while they reach 43% of Conservatives and, especially, 58% of Restorers. The left-right divide is therefore not limited to believers versus non-believers; it hinges more on the intensity of religious practice and on the nature of the religious universe (Catholic and Evangelical) to which individuals belong.
Amy Erica Smith and Taylor Boas’ work shows that in Latin America, evangelical voters have long supported left-wing forces, notably due to their roots in the working classes. It is the rise of “culture wars” around abortion, family, LGBT rights, and gender theory that gradually pushed a significant portion of this electorate toward conservative forces. The growing salience of cultural issues thus seems to have produced, in Brazil as in many European countries and the United States, a tipping of a portion of the working class toward conservative parties.
The 2018 election marks a crucial turning point in this respect. According to Layton, Smith, Moseley, and Cohen, in their model that accounts for sociodemographic characteristics, evangelical affiliation in 2018 was associated with a Bolsonaro vote in the second round that was 36 percentage points more likely than for non-religious individuals.
Age also appears to contribute to value polarization, though in a less pronounced manner than gender or religion. On average, those under 35 are more represented on the progressive side, but the gaps among the leading clusters remain relatively modest, and the relationship is not linear. The overrepresentation of those over 50 is, however, much more pronounced on the right: they account for 55% of Conservatives, 70% of Eco-Liberals, 72% of Restorers, 74% of Traditionalists, and 67% of Authoritarians, compared with a 52% average. Thus, in Brazil as in most democracies, there is a correlation between age and political orientation, with older people overrepresented in the more conservative groups.
From value systems to electoral choices
The value universes described above are reflected in electoral choices. Lula’s and Bolsonaro’s electorates split very clearly along the main axis of cleavage identified earlier. In the memories of the first round of the 2022 presidential election, the vote for Lula is reported at between 79% and 96% in the seven groups most distinctly located on the progressive side of the cleavage, from Social-Progressives to Defiants. The vote for Jair Bolsonaro ranges from 83% to 98% in the seven clusters spanning Conservatives to Authoritarians.
The positions on the two sides of this macro cleavage are thus strongly correlated with electoral orientations. Among the six clusters most distinctly situated to the left of this main axis, only the Humanists include a somewhat significant share of Jair Bolsonaro voters: 20%, versus 0–4% in most groups on the same side of the cleavage. As a group that is both very religious and practicing, this particularity (still relative) illustrates the specific role of religiosity: it can bring some individuals, who on many other dimensions remain to the left of the value space, toward a conservative electoral offering.
The central-space clusters logically display more mixed choices. In the memories of the first round of 2022, Lula’s vote stands at 79% among the Defiants, but only 42% among the Moderates and 51% among the Refractors, compared with 6%, 30%, and 39% respectively for Jair Bolsonaro. The relationship between values and vote is therefore not identical across these three groups. Their intermediate position on the axis does not mean they identify with a centrist party offering or that they stand at equal distance from the two candidates.
The core of the first-round contest, and perhaps even more so the second, thus rests on the ability of the two protagonists to mobilize nearly the entire ideological space while penetrating as deeply as possible into the three groups that appear open and thus potentially swing from one side to the other.
The comparison of the 2022 vote with September 2026 intentions also allows us to compare the candidates’ positions within the different value universes.
The scenario of the first round tested in September 2026 confirms first the solidity of Lula’s electoral core. In the five most progressive clusters, he collects between 85% and 93% of the voting intentions, levels close to those of 2022. The contrast is, however, much clearer and more problematic for the PT candidate at the periphery of this bloc: Lula remains a majority among the Humanists and the Defiants, but collects only about 15% among Moderates and Refractors. In these intermediate groups, voting intentions are more dispersed between Flávio Bolsonaro and several alternative candidacies, notably Renan Santos, Ronaldo Caiado, and Augusto Cury.
On the conservative side, Flávio Bolsonaro also benefits from a very solid base. In September, he gathers between two-thirds and nearly all voting intentions in the seven groups located to the right of the value-space. This foothold largely overlaps Jair Bolsonaro’s 2022 position, without reproducing it exactly: the son appears less dominant than his father in several liberal or “anti-welfare” groups, but he is more supported among Restorers.
However, his electorate still transcends this conservative nucleus. Flávio Bolsonaro attracts about half of the Moderates and Refractors, as well as a non-negligible share of Humanists and Defiants. These comparisons between the 2022 vote and September 2026 intentions show that the intermediate groups, along with the Humanists, constitute the main spaces where the confrontation between the two blocs remains the most open.
The second-round scenario further intensifies the concentration of choices within the two poles of the electoral space. In September’s poll, Lula collects between 96% and 100% of voting intentions in the five most progressive groups, while Flávio Bolsonaro reaches 100% among Conservatives, Eco-Liberals, anti-welfare advocates, Restorers, Liberals, Traditionalists, and Authoritarians. Twelve groups out of sixteen thus show voting intentions almost unanimously oriented toward one or the other candidate. This association, which rests on no electoral variable, underscores the strength of the link between value systems and declared choices.
Choices are relatively shared in four groups that are thus more contested and should therefore act as the election’s arbiters. Lula gathers 63% of voting intentions among the Humanists and 75% among the Defiants. Flávio Bolsonaro reaches, for his part, 75% among Moderates and 81% among Refractors. The degree of openness of these groups is thus, in reality, quite limited, but it is within this gap that the ballot could be decided.
One can complement this reading by looking at clusters’ second-round preferences according to the first-round candidate chosen.
Among those who chose Augusto Cury in the first round, 57% state they prefer Lula in the tested duel, and 43% Flávio Bolsonaro. Supporters of Renan Santos and Ronaldo Caiado lean more toward Flávio Bolsonaro, at 70% and 75%, respectively, which is consistent with these candidates’ orientation.
The main takeaway of our study is to show that the clash between the two main Brazilian political offerings rests on deep and powerful social and cultural logics. Electoral polarization is thus not solely, or even primarily, the product of a confrontation between two personalities or two political machines, even if those dynamics contribute through the issues they emphasize. It is rooted in coherent value systems, associated with distinct social universes. Positions on mores, authority, security, and redistribution tend to overlay, producing intense polarization that translates into a distribution of electoral choices.
At the same time, two large blocs do not mean two perfectly homogeneous electorates. Religion, the relationship to the environment, or the state’s role also differentiate groups within each bloc. Humanists remind us that intense Catholic practice can coexist with open positions on social issues. Eco-Liberals show that support for privatizations can go hand in hand with environmental protection of the Amazon. A mid-range position within the value space does not automatically make a voter a centrist, especially in a context where the logic of a “useful vote” pushes many voters to choose between the two candidates vying for final victory.
The strength of belonging and values thus plays a decisively shaping role in this polarized society that is Brazil in 2026, one of the most polarized we have studied, as the near-perfect homogeneity of 12 out of 16 clusters reveals, and the relatively closed character of the four least polarized clusters. But the election will be played precisely in these four middle groups, those least positioned on the grand macro-attitudinal cleavage that divides the country.