A New Ideological Classification of Brazilian Political Parties

ABSTRACT

Just as democratic politics changes, so does the perception about the parties out of which it is composed. This paper’s main purpose is to provide a new and updated ideological classification of Brazilian political parties. To do so, we applied a survey to political scientists in 2018, asking them to position each party on a left-right continuum and, additionally, to indicate their major goal: to pursue votes, government offices, or policy issues. Our findings indicate a centrifugal force acting upon the party system, pushing most parties to the right. Furthermore, we show a prevalence of patronage and clientelistic parties, which emphasize votes and offices rather than policy.

political parties; political ideology; survey; party models; elections

RESUMO

Assim como a política democrática se modifica, a percepção sobre os partidos que ela compõe também se altera. O objetivo desse trabalho é oferecer uma classificação ideológica nova e atualizada dos partidos políticos brasileiros. Através de um survey aplicado à comunidade de cientistas políticos em 2018, pedimos que classificassem os partidos na dimensão esquerda-direita e também quanto ao seu principal objetivo: a persecução de votos, de posições de governo ou de políticas. Os resultados apontam para um movimento centrífugo do sistema partidário, com a maioria dos partidos caminhando para a direita, e para o predomínio de partidos que podem ser classificados como fisiológicos, priorizando a díade votos-cargos e desprezando a programaticidade.

partidos políticos; ideologia política; survey; modelos de partido; eleições

RÉSUMÉ

À mesure que la politique démocratique change, la perception des partis qu’elle compose change également. L’objectif de ce travail est d’offrir une classification idéologique nouvelle et actualisée des partis politiques brésiliens. À travers une enquête appliquée à la communauté des politologues en 2018, nous leur avons demandé de classer les partis dans la dimension gauche-droite et aussi en fonction de leur objectif principal : la poursuite des votes, des positions gouvernementales ou politiques. Les résultats indiquent un mouvement centrifuge du système de partis, avec une majorité de partis vers la droite, et la prédominance de partis que l’on peut qualifier de physiologiques, privilégiant la dyade votes-positions et faisant fi de la programmaticité.

partis politiques; opinions politiques; survey; modèles de parti; élections

RESUMEN

A medida que la política democrática cambia, también lo hace la percepción de los partidos que la componen. El objetivo de este trabajo es ofrecer una nueva y actualizada clasificación ideológica de los partidos políticos brasileños. A través de una encuesta aplicada a la comunidad de politólogos en 2018, les pedimos que clasificaran a los partidos en la dimensión izquierda-derecha y también en cuanto a su objetivo principal: la búsqueda de votos, posiciones de gobierno o políticas. Los resultados apuntan a un movimiento centrífugo del sistema de partidos, con la mayoría de los partidos moviéndose hacia la derecha, y al predominio de los partidos que se pueden clasificar como fisiológicos, priorizando la díada voto-posición y despreciando la programática.

partidos políticos; ideología política; encuesta; modelos de partidos; elecciones

INTRODUCTION ¹

The ideological classification of political parties undergoes constant changes in democracies. These may stem from changes in party frameworks – with the advent of new parties and the disappearance of old ones, alternations in power, political situations of particular countries, – or shifts in the meaning of ideological positions, as the concepts of left and right are relational and change over time. Although we make no intent to pursue a conceptual discussion as to what each worldview represents, we acknowledge the need for a periodic update about where Brazilian political parties stand in the ideological continuum.

The ideology of Brazilian political parties has always been regarded as a secondary issue in the literature, which has privileged electoral and legislative behavior and relegated the ideological dimension as an insufficient variable for differentiating political parties ( Epstein, 2009 ; Tarouco, Madeira, 2015 ; Zucco Jr., 2011 ). From this perspective, spatial distances between parties would dissipate through electoral personalism and party patronage ( Ames, 2003 ; Mainwaring, 1993 ). While several democracies undergo a process of de-ideologization ( Bolleyer, 2013 ; Scarrow, 1996 ; Van Biezen et al., 2012 ), Brazil has never experienced a ‘golden age’ of enduring, programmatic, and ideologically outspoken political parties, as commonly found in advanced democracies ( Desposato, 2006 ; Janda, Colman, 1998 ; Samuels, 1999 ). In other words, the Brazilian framework has been largely comprised of weakly programmatic parties, regardless of the development stage of the party system.

The paradigm of Brazilian political parties has always been between a practically negligible electoral force – in which personalist campaigns, private electoral funding, high fragmentation, and clientelism prevail – and parties calling the shots in congress – organizing voting agendas, presenting cohesion rates above 50%, and forming coalitions for governability ( Figueiredo, Limongi, 2000 ; Melo, 2015 ; Pereira, Mueller, 2003 ). This behavioral dichotomy of Brazilian parties sends confusing signals to those striving to understand their programmatic bases and provide little clarity about their ideological positions.

On the other hand, experts and politicians acknowledge that ideological cleavages exist which affect the conduct of public affairs and legislative behavior. We may rely on different strategies to comprehend the content of this cleavage. We can measure party ideology based on party programs ( Babireski, 2014 ; Franzmann, Kaiser, 2006 ; Tarouco, Madeira, 2013a ), on the behavior of party representatives ( Scheeffer, 2016 ), on electoral behavior ( Carreirão, 2006 ; Dantas, Praça, 2010 ), on social composition ( Rodrigues, 2002 ), on the perception of party representatives ( Power, Zucco Jr., 2009 ), or on the perception of experts ( Tarouco, Madeira, 2015 ; Wiesehomeier, Benoit, 2007 ). Our classification proposal is in line with the latter, acknowledging its problems and advantages.

For our study, we applied a survey to members of the Brazilian Political Science Association (BPSA) over the course of twenty days in July 2018. Respondents were invited to use a web-based platform to answer about their classification of Brazil’s then thirty-five political parties across the left-right continuum on a spatial scale from zero to ten. The experts did not have access to our numerical index, but only to the spatial point distancing each party from the extreme ends. This avoids the potential problem of them anticipating ideological points and making numbers match positions previously known as extreme left, center-right, left, or right. The line thus encourages classification by proximity, avoiding correspondence between ordinal and ideological positions. As is known, the potential aggregation of average values can be performed after data collection.

We sent 963 invitations and had a significantly high return rate (59.3%), which shows the relevance of the topic and the collaborative nature of the political science community. However, when considering submission errors (duplicate submissions) and incomplete responses, the final return rate reached 53.9%, totaling a universe of 519 individuals. Most respondents had a PhD degree (48.4%), while master’s graduates accounted for 32.8%, and 18.8% had an undergraduate degree. Despite the survey’s long and strenuous character, respondents answered it in average in seventeen minutes. The result is an estimate of the average ideological position of each political party by placing parties further or closer to the left or right poles.

The text is divided as follows. The next section presents the debate on party ideology, its theoretical relevance, and the way the literature has addressed this topic. We then present the result of the ideological classification of Brazilian political parties and discuss possible problems through a comparison with previous classifications. Lastly, we suggest two possible explanations for the changes in classification of Brazilian political parties.

PARTY IDEOLOGY: HOW AND FOR WHAT?

If we understand political parties as carriers of a worldview and a corresponding program, ideology becomes a ‘cognitive shortcut’ for decision-making at the time of voting and a guiding thread for decision-making when in government office ( Rose, 1984 ). Ideology structures communication and amalgamates interests dispersed over time and space. It follows that political parties are the organizations responsible for performing this movement in democracies through electoral differentiation.

Furthermore, ideology enables comparisons over time and space, allowing us to notice movements in individual preferences as well as in the dynamics of party systems, their (mis)alignments, polarizations, and directions in the competition among peers. The differentiation of political parties between left and right is often based on their stance on the unidimensional role of state intervention in the economy ( Downs, 1999 ) and the defense of social equality as a natural or constructed phenomenon ( Bobbio, 1995 ). However, the characterization of the meaning behind these ideological positions and their corresponding ideals have changed over time and space. The definition that divides right and left through the struggle for equality has become opaque as specific agendas have entered the scene and social demands have become increasingly fragmented with the decline of class societies in Western Europe since the 1960s ( Arce, 2010 ; Botelho, 2018 ).

In other words, while an ideology-based classification of political parties serves as an analytical tool for understanding party organizations themselves, it is also important to observe the way these parties are perceived differently over time. Although political parties may be classified according to electoral behavior ( Carreirão, 2006 ; Dantas, Praça, 2010 ), legislative behavior ( Scheeffer, 2016 ), self-reported ideology ( Power, Zucco Jr., 2009 ; Zucco Jr., 2011 ), social composition ( Rodrigues, 2002 ), public policy choices (Dias, 2012), or the positions expressed by the parties themselves ( Franzmann, Kaiser, 2006 ; Tarouco, Madeira, 2013b ), all methods have advantages and drawbacks. Likewise, classification through expert surveys, such as the one we present here, also has its pros and cons.

The first advantage is that expert surveys yield valid and reliable measures, as shown by Benoit and Laver (2006 ; 2001). When compared to other indicators, such as electoral party behavior, experts’ classifications converge. Secondly, consulting experts makes the classification more likely to avoid the typical contaminations of political arenas, such as the logic of electoral and legislative alliances, the burden of making difficult decisions when in government, or the general disregard for party manifestos, especially in Brazil.². Furthermore, mass survey classifications such as Latinobarómetro echo the typical problems of Brazil’s party system, since the quantity and diversity of parties hinders efforts to differentiate them. While this problem also affects experts’ capacities, relying on them at this moment may also be an advantage.

On the other hand, Benoit and Laver (2006) argue that expert-survey methods have a considerable advantage over other ideological measurements. First, any attempt to classify political parties will involve potential error sources or biases. Second, political ideology is a concept that entails some level of abstraction and complexity, and is therefore not easily accessible. In this regard, experts would be better able to provide precise answers about ideology when compared to other social groups. Furthermore, placing political positions in a spatial line demands further precision, as respondents must reduce what is commonly known as differential item functioning (DIF), i.e., every respondent’s ability to understand questions the same way ( King et al., 2004 ).

Nonetheless, the categorization of parties according to experts also entails problems. First, as in all survey-based research, results are sensitive to the time and place in which data was collected. Important events, impactful political episodes, and respondents’ surrounding context may influence results. Second, expert surveys may show polarizing tendencies, given that experts tend to occupy the ‘entire ideological spectrum’ when trying to differentiate political parties. This is especially sensitive for highly fragmented systems ( Mair, 2001 ; Wiesehomeier, Benoit, 2007 ) like the Brazilian one, in which attempts to differentiate one party from another by avoiding empty spaces may potentially lead to an artificial polarization.

AN UPDATED IDEOLOGICAL CLASSIFICATION OF BRAZILIAN POLITICAL PARTIES

The first notable point of survey results was the absence of answers for some parties and highly varied answers for other parties ( Table 1 ). The correlation between these two items is intuitive. The greater the number of responses classifying a given political party, the greater the coefficient of variation. The correlation presented (r .437) indicates that a greater number of individuals judging themselves able to ideologically classify a specific party implies an increase in the heterogeneity of the classification and, therefore, entails further variations in the assigned positions. This is especially true for parties classified to the left of the ideological line (PSTU, PCO, PCB, PSOL, and PCdoB), which have values well above one hundred points in their coefficients of variation.

Table 1
Descriptive by party: ideology 3

Political Party	Average	Median	Mode	Standard deviation	N	Coefficient of variation
PSTU	0,51	0	0	1,13	514	220,23
PCO	0,61	0	0	1,19	509	196,03
PCB	0,91	0,9	0	1,27	512	139,91
PSOL	1,28	1	1	1,18	513	92,45
PCdoB	1,92	2	2	1,50	512	78,26
PT	2,97	3	3	1,34	504	44,99
PDT	3,92	4	4	1,48	483	37,84
PSB	4,05	4	4	1,59	468	39,33
Rede	4,77	4	4	1,72	451	35,99
PPS	4,92	5	4	2,36	472	47,83
PV	5,29	5	6	1,79	435	33,83
PTB	6,1	6	7	2,18	481	35,74
Avante	6,32	6	6	2,27	451	35,88
SDD	6,5	7	6	1,98	469	30,51
PMN	6,88	7	7	2,10	444	30,48
PMB	6,9	7	7	2,03	431	29,46
PHS	6,96	7	7	1,94	443	27,82
MDB	7,01	7	6	1,66	453	23,71
PSD	7,09	7	6	1,72	479	24,22
PSDB	7,11	7	6	1,69	496	23,81
Podemos	7,24	7	8	1,76	469	24,27
PPL	7,27	8	10	2,68	465	36,87
PRTB	7,45	8	7	1,98	462	26,52
Pros	7,47	7	8	1,58	472	24,47
PRP	7,59	7	7	1,57	472	20,66
PRB	7,78	8	8	1,48	495	19,97
PR	7,78	8	8	1,53	488	19,61
PTC	7,86	8	8	1,65	480	20,97
DC	8,11	8	8	1,55	503	19,09
PSL	8,11	8	10	1,69	489	20,79
Novo	8,13	8	10	1,72	474	21,17
Progressistas	8,20	8	8	1,53	496	18,69
PSC	8,33	8	10	1,52	505	18,23
Patriota	8,55	9	10	1,58	490	18,54
DEM	8,57	9	10	1,40	511	16,31

survey

The explanation for this phenomenon may lie in the way experts classify political parties and they ‘divide’ the spatial line. Thus, if we use Partido Verde (PV) as the center party which divides the line between left and right, the segment on the left has ten parties, while the segment on the right has exactly twice as many. That is, parties considered to be on the right of the ideological spectrum are more squeezed along the line, forcing respondents to scale one segment more accurately than the other. Consequently, undifferentiation occurs more often on the much more populous segment than in the other. We must remember, of course, that we are dealing with the political science community, with experienced respondents who have enough training and knowledge to establish comparable relationships. However, given the myriad of political parties, even the most trained eyes will find it difficult to establish a memory while following the same instruction several times.

On the other hand, at the other end of the table we find that traditional right-wing parties are highly consistent, with a low coefficient of variation. Progressistas (formerly PP, formerly PPB), PSC, Patriota (formerly PEN), and DEM (formerly PFL) are parties widely associated with the right. A notable exception is Patriota, which ranks as the second ‘furthest right-wing’ party, likely due to Jair Bolsonaro’s political convergence with the party. This alliance was subsequently frustrated in January 2018, when the party had already changed its name to accommodate the then future candidate. All other parties comprise the core of the long-standing Brazilian right. Progressistas and DEM, heirs of the military dictatorship (1964-1985), are responsible for representing agribusiness and embracing a neoliberal agenda in Brazil. PSC, a party with evangelical roots and links to the more conservative wing of the Catholic Church, closes the bloc. In short: the expert either has background information about where to place these parties or a specific context which enables them to place them unhesitatingly in a tight space when they deem it necessary to differentiate the parties from each other.

Graph 1
Coefficient of variation of classification on the ideological scale by party

The lack of responses about parties such as PMN, PMB, Avante (formerly PTdoB), PHS, PRTB, and PPL is usually linked to the low salience of their ideological stances, recent name changes, or short time of existence. On the other hand, parties such as the PV, Rede, and MDB have little cohesion in the left-right dimension. They were sometimes regarded as programmatic, such as Rede, or conveying a clear agenda, such as PV’s environmentalism, but lack markedly distinct positions within the criteria for the left or right categories. MDB is a classic case of a catch-all party, with many colors in different places and times. MDB’s nature as an amorphous political party dates back to its foundation, and its past makes it difficult to place it in a definitive position ( Kinzo, 1988 ).

As we searched for content on how political scientists have classified Brazilian political parties, we were able to pinpoint the difficulty behind the attempts to categorize the myriad parties comprising Brazil’s party system. We also asked experts how they would classify each party based on behavior – or party objectives – as theorized by Müller and Strøm (1999) and applied to party models by Wolinetz (2002) . Party objectives are a proxy for what parties consider to be their raison d’être , reflecting what their party organizations work for. For example, two different parties may have a large membership base, but while one party pursues public policies and mobilizes its members to convince voters, the other party uses electoral campaigning and adapts its program to garner as many votes as possible. The former would be closer to the policy-seeking party model, while the latter would be closer to the vote-seeking model. On the other hand, political parties that prioritize coalitions by taking advantage of ticket holders or surrender their leverage in a certain region striving towards political offices and patronage in future elected governments are commonly classified as office-seeking parties.

According to Wolinetz (2002) , the model that classifies political parties by behavior tends to be more advantageous insofar as it is more flexible and allows for a more precise analysis of variations over time and space. At the same time, the party behavior indicator should not be interpreted through isolated objectives. Most of the time political parties perform all three or use one behavior to achieve another. A party will unlikely implement its political program without receiving an expressive number of votes, whether to legitimize the party’s policies or secure enough power to govern. Nonetheless, parties’ behaviors prioritize different strategies with different intensities as they react to political and institutional constraints.

Thus, different party objectives coexist in (im)balance as these institutions form complex units, distant from the reductionist schemes of rational choice ( Downs, 1999 ; Smith, 2004 ). The question thus becomes whether we can overlap the ideological and behavioral dimensions of parties and whether we can find some level of consistency in how experts understood these two classifications in our survey.

Figure 2 shows all Brazilian political parties according to the assigned party objectives in line with Wolinetz (2002) . More precisely, the question asked to respondents included the following vignette: ‘One way to understand political parties is to classify them according to the main objective they pursue and the subsequent organizational consequences. According to the literature, we may classify parties under three categories as summarized below ( Müller, Strøm, 1999 ; Strøm, 1990 ; Wolinetz, 2002 ):

Figure 2
Policy, office, or vote-seeking classification of Brazilian political parties in 2018

• - vote-seeking parties seek to maximize votes regardless of the competitive electoral context or their programmatic limitations or alliances. Political positions are flexible. Party organization intensifies during elections and goes unnoticed in between. They are flexible in building alliances, adapting their worldview according to each electoral race or political opponent in an attempt to gather as many voters as possible;

• - office-seeking parties maximize participation in governments and political alliances, even if this means not maximizing electoral votes or ignoring their political program. In order to survive, they would rather enjoy the spoils of office from their winning electoral partner than be left out of government. Party activities are restricted to electoral periods and the party’s structure depends on resources from holding offices;

• - policy-seeking parties prioritize occupying the state to ensure the pursuit of a political issue agenda. They tend to be permanently mobilized, even between elections, and involve a myriad of party spheres. Ultimately, they seek to convince voters instead of adapting their program to the preferences of constituencies. They ultimately prefer alliances with parties that share their worldview, even if this means making some programmatic concessions.

Figure 2 also presents the ideological coloring of each party according to the average position in the left-right continuum. The leftmost point coincides with the value zero and the rightmost point with the value ten. On the one hand, the high non-response rate ( Table 2 ) may be explained by the difficulty in identifying political parties that pursue public policies outside the left spectrum. With the exception of Rede, Novo, and PSDB, all other parties either lack a clear objective (PSB, PDT, PPL⁵, PV, and DEM) or resemble the pair ‘office–vote seeking’, thus suggesting that physiologism/clientelism seems to be the tonic of Brazil’s national party system.

Table 2
Non-response rate by party

Party	%
PMB	16,95
PV	16,18
PHS	14,64
PMN	14,45
Avante	13,10
Rede	13,10
MDB	12,72
PRTB	10,98
PPL	10,40
PSB	9,83
Podemos	9,63
SDD	9,63
Pros	9,05
PRP	9,05
PPS/CDD	9,05
Novo	8,67
PSD	7,71
PTC	7,51
PTB	7,32
PDT	6,94
PR	5,97
PSL	5,78
Patriota	5,59
PRB	4,62
PSDB	4,43
Progressistas	4,43
DC	3,08
PT	2,89
PSC	2,70
PCO	1,93
DEM	1,54
PCdoB	1,35
PCB	1,35
PSOL	1,16
PSTU	0,96

survey

Looking more closely at the ideological positions represented by the colors on the line, some constants stand out in segments within the same amplitude. In our scale, parties with an ideological average between 0 and 1.5 were classified as far-left; parties that scored 1.51 to 3 were classified as left; from 3.01 to 4.49 as left-center parties; between 4.5 and 5.5 as center; values between 5.51 and 7 as center-right; 7.01 to 8.5 were classified as right-wing and; lastly, parties that scored an average between 8.51 and 10 were classified as extreme right. The furthest right-wing parties on the spectrum are DEM, which had an average ideology of 8.57, and Patriota, with 8.55, both occupying the extreme right. On the other end, we find PSTU, with 0.51, and PCO with 0.61.

We may classify the extreme left parties (PSTU, PCO, and PCB), as anti-system⁶( Ribeiro, 2003 ; Sartori, 1980 ). Their sole objective is to disseminate political ideas, without concern for gathering votes. They usually advocate against the democratic system, proposing its general overhaul by means other than the electoral route. As we move towards the center of the classification, we find the emergence of systemic behaviors and the left expands its ‘triangle’, with further focus on the pursuit of votes and government offices. Both PCdoB and PT are closer to programmatic parties, as shown in Figure 1 . Although programmatic at their core, they strive to advance their policies by winning over voters and implementing policies when occupying state offices.

Figure 1
Party objectives and party models 4

The center-left, on the other hand, is closer to the center than to the actual left. These parties seem to lack any clarity about their objectives and their polygons do not point to any specific direction. With the exception of Rede, the other center-left and center parties suggest an incoherent behavior, as even experts seem at a loss to understand their proposals. PPS/Cidadania is perhaps understood as more resolute in seeking government offices or nominations when compared to its peers, but still lacks any cohesive behavior in either direction. The center-right (light blue) encompasses parties that seek votes and patronage, and are always in government whether via elections or patronage. This ideological slice is characterized by an absolute dominance of physiological parties

The center-right, which begins with PTB and ends with PHS, and the right, which ranges from MDB⁷to PSC, is also a predominantly physiological domain, despite exceptions. The first of them is PSDB, whose geometry grants reasonable weight to the vote and policy-seeking dimensions. In fact, the PSDB graph is akin to the one found in another electoral behavior study which uses our same model ( Rebello, Giora, Pereira, 2020: 94). The second exception is Novo, generally understood as a political party more concerned with pursuing programmatic goals, thus located in parallel with left-wing parties. Lastly, respondents considered DEM, on the extreme right, to have a similar behavior to center-left and center parties, i.e., lacking a clear objective. The particular case of DEM also involved references to being a satellite party at the service of PSDB in government administrations and electoral races ( Bolognesi, 2013 ; Carreirão, 2006 ). However, this pattern is not verified when the ‘satellite’ party belongs to the left – namely PCdoB, which has always been an auxiliary force of PT ( Braga, 2007 , 2010 ) and was not perceived as a springboard to government offices.

The extreme right, despite comprising only two parties, incorporates programmatic behavior among its goals, which is not the mode among its neighboring ideological peers. While distant from the policy-seeking behavior of the left and the extreme left, this element is consistent with political parties in polarized positions, as both DEM and Patriota present a manifest ideology.

As a comparative criterion, Rebello, Giora, and Pereira (2020: 92) use the triad policy, office, and vote-seeking to categorize the objectives of some political parties based on their behavior in state-level elections. Similar to our findings regarding the perception of political scientists about this behavior, the authors show that most of the analyzed parties (such as PSDB, PDT, DEM, PSB, and Progressistas) prioritize holding federal and state offices. MDB has a different behavior and does not prioritize any specific objective, while PT is the only party that places policy ahead of other dimensions.

Nonetheless, ideological positions change over time. Patriota party, for example, relatively undistinguished, until then known as PEN, and operating as the political arm of Neo-Pentecostal evangelical churches ( Cassotta, 2016 ), gained prominence when it changed its name in 2018 to host the candidacy of Brazil’s current president. This may have prompted experts to place Patriota on the extreme right – rather than the expected physiologism.

To observe the flow of these changes, we compared our classification to Tarouco and Madeira’s, which also surveyed the political science community (2015). As stated above, there are of course many other classifications and methods for measuring party ideology. However, preserving the community of respondents – even though this is not a panel study – seems to us to be an important criterion for a comparative control strategy ( Harmel, Janda, 1978 ; Ragin, 1987 ; Sartori, 1991 ).⁸

The first significant finding is the party system’s centrifugal tendency towards the right. Graph 2 points in this direction: from 2010 (when the first survey with BAPS associates was performed) until 2018, more parties began to occupy right-wing positions and, additionally, more parties began to express more extreme positions. The left, on the other hand, has not grown in size, but its positions have escalated, with the exception of PT.

Graph 2
Comparison of ideological positionings 2010 vs. 2018

One hypothesis for this phenomenon lies precisely in the pivotal importance of PT for understanding and classifying other parties. The party’s immutability in two classification rounds, after two national elections (2010 and 2014), in addition to a highly significant episode in its recent history – the impeachment of Dilma Rousseff in April 2016 – points to PT’s stability as a reference party when ranking other parties. Brazilian political scientists use PT as an ideological beacon to express their worldview about Brazilian political parties, thus suggesting, as observed in other countries, that large parties with programmatic salience tend to serve as a reference point for structuring perceptions about other political parties ( Mair, 2001: 12-13).

Another party that has undergone little change is PSOL. In spite of being much younger than PT, it also seems to serve as a reference when positioning other parties. Interestingly, in both classifications, PCdoB was placed between PT and PSOL. PCdoB was initially closer to PT, when the latter governed the country and the former belonged to its government coalition. Later they grew distant in 2018, following a general trend of left-wing organizations seeking to dissociate themselves from PT. Furthermore, PT also serves as a reference for the centrifugal momentum towards the right. Without any exception, all parties classified to the right of PT intensified their ideological positions in the opposite direction. Even parties that were once close and shared almost the same midpoint as PT in 2010, such as PSB, have moved sharply to the right.

As discussed in our introduction, changes in party classification may also stem from a rearrangement of the party system. The growing number of parties tends to promote a race from the center towards the fringes, as experts attempt to fill out the entire ideological spectrum. Mair (2001) suggests that fragmentation can lead to polarization, and thus the center of the line becomes less crowded than the poles. However, in this case we find a fragmentation that has not effectively led to polarization, but only further complicates efforts to position center-right and right-wing parties. This is one of the possible interpretations for the centrifugal trend shown in Graph 2 , since no less than ten parties were added to the survey from 2010 to 2018, most of which placed to the right of the ideological spectrum. Lastly, the classification is based on individual judgments that may weigh different dimensions (legislative behavior, public policy proposals, electoral behavior, public opinion etc.) as valid dimensions when ranking political parties. This is especially significant for experts, insofar as their academic background trains them to understand a phenomenon – or a party – as a complex organization and they tend to detach particular experiences from their evaluations ( Tarouco, Madeira, 2015: 26).

There is yet another fundamental problem within the party dynamics itself. Political parties are not organizations frozen in time. Perceptions about parties change as they participate in government administrations, hold public offices, support certain personalities, and adapt their programs to electoral races ( Wills-Otero, 2016 ). Such variations over time may also affect experts’ perceptions about the ideological coloring of parties. This problem has also been linked to the relationship between Western European parties and their voters and the image projected on public opinion. The history of political parties – particularly their origins – thus becomes a fundamental dimension for experts to classify them ideologically in Europe, as they tend to seek a party’s most precise spatial position ( Kligemann et al., 2006 ) in party families ( Mair, Mudde, 1998 ). In Brazil, this may be the answer for the steadfastness of PT, a party distinguished in the party scenario for its origins and history (Amaral, Power, 2015). At the same time, the exact opposite happens with the rest of the party system, with its fluid and inconstant historical ties ( Carreirão, 2014 ).

CONCLUDING REMARKS

Through a descriptive approach, we sought to present an updated classification of Brazil’s party constellation. Neither the technique nor the object are novel. Our contribution to the debate on political parties and their programmatic positions stems from a complete overview of the party system alongside a classification that seeks to move beyond the left-right dyad.

Our decision to include all political parties in our survey entails benefits and losses. The first gain is a classification of the entire Brazilian party system without having to resort to low-validity procedures, such as the analysis of manifestos or coalitions. Small and non-ideological parties have swarmed Brazil’s democracy, i.e., parties with empty programs and pragmatic electoral behavior, which makes it difficult to classify them. By asking experts to undertake such classification, even if by comparing them against programmatic parties, we may better understand the organization of the representative system. On the other hand, by adding the objectives of political parties, the ideological classification was clearly more salient for some categories than others. Notably, left and far-left political parties have more programmatic salience, while center, center-right, and right parties are more commonly associated with physiological and catch-all parties ( Wolinetz, 1991 ).

Furthermore, when comparing our data with previous classifications, we find general trends in the overall set of parties. In our analytical case, we observed a centrifugal trend pushing most parties to the right, with some exceptions to the left of PT in the first wave. We observed a similar movement when analyzing parties based on their representatives ( Zucco, Power, 2021 ). We attribute this phenomenon both to the impeachment of a PT president in 2016, which pushed allied and centrist parties towards the opposition, as well as the rise of the extreme right, which has shrunken the space of the right and center-right.

REFERÊNCIAS

• Amaral, Oswaldo Estanislau do; Power, Timothy J. (2016), “The PT at 35: Revisiting Scholarly Interpretations of the Brazilian Workers’ Party”. Journal of Latin American Studies , v. 48, n. 1, pp. 147-171.
• Ames, Barry. (2003), Os entraves da democracia no Brasil . Rio de Janeiro, FGV.
• Arce, Moisés. (2010), “Parties and Social Protest in Latin America’s Neoliberal Era”. Party Politics , v. 16, n. 5, pp. 669-686. Disponível em: http://ppq.sagepub.com/cgi/doi/10.1177/1354068809346005
  » http://ppq.sagepub.com/cgi/doi/10.1177/1354068809346005
• Babireski, Flávia. (2014), “As Diferenças Entre a direita do Brasil, Chile e Uruguai: Análise dos Programas e Manifestos Partidários. Paraná Eleitoral , v. 3, n. 1, pp. 171-198.
• Benoit, Kenneth; Laver, Michael. (2006), Party Policy in Modern Democracies . London, Routledge.
• Bobbio, Norberto. (1995), Direita e esquerda: razões e significados de uma distinção política . São Paulo, Editora UNESP.
• Bolleyer, Nicole. (2013), New Parties in Old Party Systems: Persistence and Decline in Seventeen Democracies . Oxford, Oxford University Press.
• Bolognesi, Bruno. (2013), “A Seleção de Candidaturas no DEM, PMDB, PSDB e PT nas Eleições Legislativas Federais Brasileiras de 2010: Percepções dos Candidatos Sobre a Formação das Listas”. Revista de Sociologia e Política , v. 21, n. 46, pp. 45-68.
• Bolognesi, Bruno; Graça, Luis Felipe Guedes da; Bonifácio, Robert; Nunes, Wellington. (2020), “Resenha de Routledge Handbook of Brazilian Politics” . Brazilian Political Science Review , v. 14, n. 1, pp. 1-17.
• Botelho, João Carlos Amoroso. (2018), “Da Perda de Legitimidade à Polarização: Os Partidos e os Sistemas Partidários de Argentina e Venezuela”. Cadernos CRH , v. 31, n. 83, p. 407-426.
• Braga, Maria do Socorro Sousa. (2010), Eleições e Democracia no Brasil: a Caminho de Partidos e Sistemas Partidários Institucionalizados. Revista Brasileira de Ciência Política , n. 4, p. 43-73.
• Braga, Maria do Socorro Sousa. (2007), “O Processo Político-partidário Brasileiro e as Eleições de 2006”. Política & Sociedade , v. 6, n. 10, pp. 53-89.
• Carreirão, Yan de Souza. (2006), “Ideologia e Partidos Políticos: Um Estudo sobre Coligações em Santa Catarina”. Opinião Pública , v. 12, n. 1, pp. 136-163.
• Carreirão, Yan de Souza. (2014), “O Sistema Partidário Brasileiro: Um Debate com a Literatura Recente”. Revista Brasileira de Ciência Política , n. 14, pp. 255-295.
• Cassotta, Priscilla Leine. (2016), “Uma Análise do Comportamento dos Deputados Evangélicos no Legislativo Brasileiro”. E-Legis , v. 9, n. 20, pp. 75-101.
• Dantas, Humberto; Praça, Sérgio. (2010), “Pequenos Partidos no Brasil: Uma Análise do Posicionamento Ideológico com Base nas Coligações Municipais de 2000 a 2008”, in Krause, S.; Dantas, H.; Miguel, L. (orgs.). Coligações partidárias na nova democracia brasileira: perfis e tendências . Rio de Janeiro; São Paulo, Konrad-Adenauer-Stiftung; Ed. Enesp, pp. 99-133.
• Desposato, Scott W. (2006), “Parties for Rent? Ambition, Ideology, and Party Switching in Brazil’s Chamber of Deputies”. American Journal of Political Science , v. 50, n. 1, pp. 62–80. Disponível em: http://doi.wiley.com/10.1111/j.1540-5907.2006.00170.x
  » http://doi.wiley.com/10.1111/j.1540-5907.2006.00170.x
• Dias, Marcia Ribeiro; Menezes, Daiane Boelhouwer; Ferreira, Geison da Cunha. (2012), “‘A Quem Serve o Graal?’ Um Estudo Sobre a Classificação Ideológica dos Partidos Políticos Através de Seus Projetos de Lei”. Civitas – Revista de Ciências Sociais , v. 12, n. 2, p p. 209-235.
• Downs, Anthony. (1999), Um teoria econômica da democracia . São Paulo, Edusp.
• Epstein, David. J. (2009), “Clientelism Versus Ideology: Problems of Party Development in Brazil”. Party Politics , v. 15, n. 3, pp. 335-355.
• Figueiredo, Argelina; Limongi, Fernando. (2000), “Presidential Power, Legislative Organization, and Party Behavior in Brazil”. Comparative Politics , v. 32, n. 2, pp. 151-170.
• Franzmann, Scott; Kaiser, André. (2006), “Locating Political Parties in Policy Space: A Reanalysis of Party Manifesto Data”. Party Politics , v. 12, n. 2, pp. 163-188.
• Harmel, Robert; Janda, Kenneth. (1978), Comparing Political Parties . Washington, The American Political Science Association.
• Janda, Kenneth; Colman, Tyler. (1998), “Effects of Party Organization on Performance During the ‘Golden Age’ of Parties”. Political Studies , v. 46, n. 3, pp. 611-632.
• King, Gary; Murray, Christopher; Solomon, Joshua; Tandon, Ajay. (2004), “Enhancing the Validity and Measurement in Survey Research”. American Political Science Review , v. 98, n. 1, pp. 191-297.
• Kinzo, Maria D’ Alva Gil. (1988), Oposição e autoritarismo: gênese e trajetória do MDB (1966-1979) . São Paulo, Vértice.
• Kligemann, Hans-Dieter; Volkens, Andrea; Bara, Judith; Budge, Ian. (2006), Mapping Policy Preferences II: Estimates for Parties, Electors, and Governments in Eastern Europe, European Union and OECD 1990-2003 . Oxford: Oxford University Press.
• Laver, Michael. (2001), “Why Should We Estimate the Policy Positions of Political Actors?”, in Laver, Michael (ed.). Estimating the Policy Position of Political Actors . New York, Routledge, pp. 3-9.
• Mainwaring, Scott. (1993), “Democracia Presidencialista Multipartidária: O Caso do Brasil. Lua Nova: Revista de Cultura e Política , n. 28-29, pp. 21-74.
• Mair, Peter. (2001), “Searching for the Positions of Political Actors: A Review of Approaches and a Critical Evaluation of Expert Surveys”, in Laver, Michael (ed.). Estimating the Policy Position of Political Actors. New York: Routledge, p. 10–30.
• Mair, Peter; Mudde, Cas. (1998), “The Party Family and Its Study”. Annual Review of Political Science , v. 1, n. 1, p. 211–228.
• Melo, Carlos Ranulfo. (2015), “Avaliando Vínculos Entre Partidos e Deputados nas Assembleias Legislativas Brasileiras”. Opinião Pública , v. 21, n. 2, pp. 365-392. Disponível em: http://www.scielo.br/scielo.php?script=sci_arttext&pid=S0104-62762015000200365&lng=pt&nrm=iso&tlng=en
  » http://www.scielo.br/scielo.php?script=sci_arttext&pid=S0104-62762015000200365&lng=pt&nrm=iso&tlng=en
• Müller, Wolfgang C.; Strøm, Kaare. (1999), Policy, Office, or Votes? How Political Parties in Western Europe Make Hard Decisions . Cambridge, Cambridge University Press.
• Neumann, Sigmund. (1955), Modern Political Parties. Chicago, University of Chicago Press.
• Pereira, Carlos; Mueller, Bernardo. (2003), “Partidos Fracos na Arena Eleitoral e Partidos Fortes na Arena Legislativa: A Conexão Eleitoral no Brasil”. Dados , v. 46, n. 4, pp. 735-771. Disponível em: http://www.scielo.br/pdf/dados/v46n4/a04v46n4.pdf
  » http://www.scielo.br/pdf/dados/v46n4/a04v46n4.pdf
• Power, Timothy J.; Rodrigues-Silveira, Rodrigo. (2019), “The Political Right and Party Politics”, in Ames, B. Routledge Handbook of Brazilian Politics . New York, Routledge.
• Power, Timothy J.; Zucco JR. (2009), Cesar. “Estimating Ideology of Brazilian Legislative Parties, 1990–2005”. Latin American Research Review , v. 44, n. 1, pp. 218-246.
• Ragin, Charles C. (1987), The Comparative Method. Moving Beyond Qualitative and Quantitative Strategies . Berkley: University of California Press.
• Rebello, Maurício Michel; Giora, Gustavo; Pereira, Matheus Henrique Gomes. (2020), “Os Objetivos dos Partidos no Brasil (1982-2018): Construindo o IOP”. Revista Debates , v. 14, n. 3, pp. 70-98.
• Ribeiro, Pedro Floriano. (2003), “O PT Sob uma Perspectiva Sartoriana: de Partido Anti-Sistema a Legitimador do Sistema”. Política & Sociedade , n. 3, pp. 45-70.
• Rodrigues, Leôncio Martins. (2002), “Partidos, ideologia e Composição Social: Um Estudo das Bancadas Partidárias na Câmara dos Deputados”. São Paulo, Edusp.
• Rose, Richard. (1984), Do Parties Make a Difference? London; Basingstoke, The Macmillan Press.
• Samuels, David J. (1999), “Incentives to Cultivate a Party Vote in Candidate-centric Electoral Systems: Evidence from Brazil”. Comparative Political Studies , v. 32, n. 4, pp. 487-518.
• Sartori, Giovanni. (1991), “Comparing and Miscomparing”. Journal of Theoretical Politics , n. 3, pp. 243–257.
• Sartori, Giovanni. (1980), Partidos y sistemas de partidos . Madrid, Alianza Editorial.
• Scarrow, Susan E. (1996), Parties and Their Members . Oxford, Oxford University Press.
• Scheeffer, Fernando. (2016), Ideologia e comportamento parlamentar na Câmara dos Deputados: faz sentido ainda falar em esquerda e direita? Tese de Doutorado em Sociologia Política, Universidade Federal de Santa Catarina, Florianópolis, Brasil.
• Singer, Matthew; Kitschelt, Herbert. (2011), “ ‘Do Everything’ (DoE) Parties: When Can Politicians Combine Clientelistic and Programmatic Appeals?”, in Workshop on Democratic Accountability Strategies , Durham, pp. 1-31.
• Smith, Claire M. (2004), “Money to Burn: Party Finance and Party Organization in Germany and Austria, in Midwest Political Science Association , Ohio, pp. 1-46.
• Strøm, Kaare. (1990), “A Behavioral Theory of Competitive Political Parties”. American Journal of Political Science , v. 34, n. 2, pp. 565-598.
• Tarouco, Gabriela da Silva; Madeira, Rafael Machado. (2015), “Os Partidos Brasileiros Segundo Seus Estudiosos: Análise de um Expert Survey” . Civitas , v. 15, n. 1, pp. 24-39.
• Tarouco, Gabriela da Silva; Madeira, Rafael Machado. (2013a), “Partidos, Programas e o Debate Sobre Esquerda e Direita no Brasil . Revista de Sociologia e Política , v. 21, n. 45, pp. 149-165.
• Tarouco, Gabriela da Silva; Madeira, Rafael Machado. (2013b), “Esquerda e Direita no Sistema Partidário Brasileiro: Análise de Conteúdo de Documentos Programáticos”. Revista Debates , v. 7, n. 2, pp. 93-114.
• Van Biezen, Ingrid; Mair, Peter; Poguntke, Thomas. (2012), “Going, Going, . . . Gone? The Decline of Party Membership in Contemporary Europe”. European Journal of Political Research , v. 51, n. 1, pp. 24-56.
• Wiesehomeier, Nina; Benoit, Kenneth. (2007), Parties and Presidents in Latin America: Data from Expert Surveys in 18 Latin American Countries, 2006-2007 . Konstanz; Dublin: University of Konstanz, Trinity College Dublin.
• Wills-Otero, Laura. (2016), “The Electoral Performance of Latin American Traditional Parties, 1978–2006: Does the Internal Structure Matter?”. Party Politics , v. 22, n. 6, pp. 758-772.
• Wolinetz, Steven B. (2002), “Beyond the Catch-All Party: Approaches to the Study of Parties and Party Organization in Contemporary Democracies”, in Ghunter, R.; Montero, J.; Linz, J. (eds.). Political Parties: Old Concepts, New Challenges . Oxford, Oxford University Press, pp. 136-165.
• Wolinetz, Steven B. (1991), “Party System Change : The Catch-all Thesis Revisited. West European Politics , v. 14, n. 1, pp. 113-128.
• Zucco Jr., Cesar; Power, Timothy J. (2021), “Fragmentation Without Cleavages? Endogenous Fractionalization in the Brazilian Party System”. Comparative Politics , v. 53, n. 3, pp. 477-500.
• Zucco Jr., Cesar. (2011), “Esquerda, Direita e Governo: A Ideologia dos Partidos Políticos Brasileiros”, in Power, T.; Zucco, Cesar (orgs.). O Congresso por ele mesmo: autopercepçoes da classe política brasileira. Belo Horizonte, UFMG, pp. 37-60.

NOTES

• 1
  . This research would not have been possible without the collaboration of the Brazilian Political Science Association ( www.cienciapolitica.org.br ), which kindly provided us with a list of its members willing to collaborate with our scientific data survey. We also thank Ana Paula Maciel for her support in our data collection, Gabriela Tarouco, André Castro, Maria Cecília Eduardo, and Simone Dourado for collaborating in the pre-test survey and giving important feedback for improving our analytical tool. We only contacted associated political scientists who authorized the use of their registration data for academic and scientific purposes. Additionally, respondents expressed their consent to participate in the survey through a filter question in which they could choose whether or not to fill in the questionnaire. Furthermore, we did not store or publish any personal identification, thus ensuring the anonymity of those who collaborated with our research.
• 2
  . See Benoit and Laver (2006) for classifications based on political behavior and Franzmann and Kaiser (2006) for how manifestos may present analytical problems.
• 3
  . PSTU – Partido Socialista dos Trabalhadores Unificado(United Socialist Workers’ Party), PCO – Partido da Causa Operária (Workers’ Cause Party), PCB – Partido Comunista Brasileiro (Brazilian Communist Party), PSOL – Partido Socialismo e Liberdade (Socialism and Liberty Party), PCdoB – Partido Comunista do Brasil (Communist Party of Brazil), PT – Partido dos Trabalhadores (Workers’ Party), PDT – Partido Democrático Trabalhista (Democratic Labor Party), PSB – Partido Socialista Brasileiro (Brazilian Socialist Party), Rede – Rede Sustentabilidade (Sustainability Network), PPS – Partido Popular Socialista (Popular Socialist Party), PV – Partido Verde (Green Party), PTB – Partido Trabalhista Brasileiro (Brazilian Labor Party), SDD – Solidariedade (Solidarity), PMN – Partido da Mobilização Nacional (Party of National Mobilization), PMB – Partido da Mulher Brasileira (Party of the Brazilian Woman), PHS – Partido Humanista da Solidariedade (Humanist Party of Solidarity), MDB – Movimento Democrático Brasileiro (Brazilian Democratic Movement), PSD – Partido Social Democrático (Social Democratic Party), PSDB – Partido da Social Democracia Brasileira (Brazilian Social Democracy Party), PPL – Partido Pátria Livre (Free Fatherland Party), PRTB – Partido Renovador Trabalhista Brasileiro (Brazilian Labor Renewal Party), PROS – Partido Republicano da Ordem Social (Republican Party of the Social Order), PRP – Partido Republicano Progressista (Progressive Republican Party), PR – Partido da República (Republican Party), PTC – Partido Trabalhista Cristão (Christian Labor Party), DC – Democracia Cristã (Christian Democracy), PSL – Partido Social Liberal (Social Liberal Party), Partido Social Cristão (Social Christian Party), DEM – Democratas (Democrats).
• 4
  . The concept of ‘physiological parties’ is not commonly found in the international literature on political parties. Other concepts such as ‘personalistic parties’ or ‘clientelistic parties’ are more common. However, both definitions do not quite convey the same meaning as the Brazilian physiologism, in which candidates and elected officials are party attributes – rather than the party being a ‘good’ of the candidate. The party operates in the wake of its members, who maintain parochial practices based on the individuals’ electoral connections. These are unlike the personalist party, grounded on the figure of a charismatic leader, or the clientelistic party, underpinned by a patronage network comprised of notable locals and coordinated by the centralized party ( Singer, Kitschelt, 2011 ).
• 5
  . Incorporated into PCdoB in 2018.
• 6
  . Wolinetz attributes the position of anti-system parties to ‘mass integration parties’, a term coined by Neumann (1955) . But the central idea remains the same: parties more concerned with political activism than electoral competition.
• 7
  . The (P)MDB is usually positioned in the ideological center in most of the works that address this theme. Recently, there has been a relocation of the PMDB towards the right, especially after the party’s stance in the impeachment of former President Dilma Rousseff (PT). The first data we have with the PMDB on the right can be found in Power and Rodrigues-Silveira (2019) , who used the method of positioning parliamentarians, a position that Power criticized in Bolognesi et. al. (2020) , but which seems to have been proven right since the further accumulation of ideological classifications.
• 8
  . To proceed with the comparison, we normalized values on the 0 to 10 scale.

• ERRATUM
  In the article “A New Ideological Classification of Brazilian Political Parties”, DOI number: 10.1590/dados.2023.66.2.303x, published in the journal DADOS – Revista de Ciências Sociais, 66(2):1-29, on page 13:
  WHERE IT WAS WRITTEN:
  

  
  Figure 2 Policy, office, or vote-seeking classification of Brazilian political parties in 2018
  
  SHOULD READ:
  

  
  Figure 2 Policy, office, or vote-seeking classification of Brazilian political parties in 2018
  

Publication Dates

• Publication in this collection
  09 Sept 2022
• Date of issue
  2023

History

• Received
  6 July 2021
• Accepted
  16 Jan 2022