The discussion about party preferences in Brazil shows their low rates. Though, research reveals
that the influence of the Workers’ Party (PT) on the partisanship of Brazilians elevates the
country to the world average of party identification, ahead of countries like Germany, Spain,
and Mexico. Historically, this party has been linked to an impoverished social stratum, the
working class, social movements, and trade unions, as a result of its organization and partisan
brand. Nevertheless, electoral strategies and pragmatic changes in the PT in recent years have
brought it closer to different social classes. Analyses of social classes and their party
preferences in Brazil have often associated them with economic income and this has resulted in
a high aggregation of social class locations. In this way, I use an alternative measurement,
through occupation in the labour market and education. I adapt a typology of hybrid social
classes, with neo-weberian aspects, based on the EGP model, and, neo-marxist, based on the
research of Santos (2005; 2010). Thus, I analyse petismo - the party preference for the PT - and
its role in social classes, with a historical series from 2002 to 2018. I use data from CESOP/
Datafolha surveys in a table for four categories of social classes. In this typology, I elaborate
descriptive data, molding the so-called “Thomsen index” in the differences in levels of
partisanship between social classes, in the likelihood of a social class position showing party
sympathy to the PT, against the likelihood of another social class position doing the same,
standardizing from the natural logarithm of the odds ratio. I use a logistic regression model with
the response variable to petismo. In addition to the social class, in this model I use the
covariables of sex, age, education, region, and government evaluation. The descriptive results
denote a greater party preference to the PT for the category of individuals in the service class,
with 36.10% in 2002 and a minimum of 8.9% among entrepreneurs and professionals in 2018.
Such values fluctuate, being the self-employed and precarious exposing greater petismo
throughout the series. However, the alternative measurement of social classes to the usual of
Brazilian political science to petismo, the application of the Thomsen index, and logistic
regression confirmed the limits of the social class variable for party preferences to the PT. This
research concludes that petismo was not a social class phenomenon, but with a moderate and
relative social class partisanship, mainly for precarious sectors of society.
|
