Political parties and candidates who focus on the poor tend to be found on the left of the political spectrum. Yet, right-wing parties frequently win elections in developing countries despite the fact that a vast majority of the electorate lives in poverty. Why do so many poor voters favor candidates on the right?
A team of University of Rochester political scientists explains the apparent contradiction in a recent paper published in the American Journal of Political Science.
Anderson Frey, an assistant professor of political science, and Zuheir Desai, who earned his PhD from Rochester in 2020 and is now an assistant professor of political science at Madrid-based IE University’s School of Global and Public Affairs, find that right-wing parties rely successfully on so-called “descriptive representation” to win elections. Descriptive representation—in which candidates present themselves as sharing certain qualities with large segments of their electorates—is based on the idea that a group is more likely to elect a candidate whose characteristics mirror some of the more typical experiences and outward markers of the group.
In other words, candidates themselves need to appear similar. In the case of municipal elections in Brazil the candidates need to appear less wealthy, less educated, less privileged—to convince a majority of prospective working- or lower-class voters.
“Candidates generally—on the left and right—emphasize their resemblance with the targeted voting population—in this case, the poor—in any characteristic that might lead voters to see the candidate as one of them,” explains Frey. In the case of Brazil, candidates emphasize their lack of formal education to connect to these voters.
‘Descriptive representation’ works—but only when the policies match
One would expect left-wing parties to jump on the idea of class-based descriptive representation. Indeed, former Brazilian president Luiz Inácio Lula da Silva (better known as “Lula”), a union leader during the 1970s who served as the nation’s president from 2003 to 2010, often points to his lack of education to emphasize his ability to succeed as a politician.
Yet, the Rochester political scientists uncover an empirical pattern in Brazil’s municipal elections that defies this conventional wisdom: it is the political right that really capitalizes on descriptive representation in the poorest areas, not the left.
“To credibly shift its position, the right-wing party nominates candidates who are descriptively closer to the poor,” says Frey.
Frey and Desai develop a theory of candidate selection and policy choice and test it by examining Brazilian municipal elections from 2004 to 2016. They discover that candidates on the right are successful in poorer regions only when they can match the pro-poor policies espoused by candidates on the left and manage to appear cut from the same cloth as their prospective voters in high-poverty areas.
For example, the team finds that right-wing mayors in Brazil can afford to spend less on the poor than left-wing mayors in better-off municipalities. But in high-poverty municipalities, the right not only has to match the left’s policies, it also nominates less educated candidates.
Why not go for ‘the real McCoy?’
What then prevents parties on the right from simply nominating candidates with more genuine connections to the broad—and less privileged—populace?
First, candidate selection does not happen in a vacuum and depends on the available pool. In Brazil, as in most democracies around the world, the candidate pool doesn’t really reflect the population. For example, while more than 60 percent of Brazilian voters did not graduate from high school, fewer than 20 percent of Brazilian mayoral candidates did not. This pattern has been well-established in even richer democracies, such as Sweden.
“People with higher human capital, such as higher levels of education, are simply more likely to choose to go into politics,” says Desai. “At a basic level, it’s more costly for political parties to nominate candidates who are less educated and look more like the poor simply because it’s too hard to find them.”
Second, there are a few aspects of policy and mayors’ electoral performance where parties and voters prefer candidates with so-called “higher human capital.” Desai and Frey find suggestive evidence that those municipalities that are governed by less educated mayors receive fewer resources from the federal government to invest in infrastructure projects, including hospitals and schools. What is more, these municipalities have lower school enrollment rates and show fewer doctors’ visits.
These candidates are also not as adept at attracting votes for their parties in subsequent federal elections: “Voters could punish parties for nominating less educated candidates because of their worse administrative performance, despite a favorable redistributive policy position towards the poor,” says Desai.
That’s why parties may be wary of nominating less educated candidates. “While this strategy may win them this election,” Desai says, “the parties might have to sacrifice too much in the long run.”
Do the findings translate to the US and elsewhere?
Right-wing municipal candidates in Brazil won because they employed descriptive representation matched by pro-poor policies. Yet, Frey and Desai caution that they can’t speak to whether the same evidence-based causal relationship exists for—say the 2016 general US general election. Frey, however, observes that despite his elite background, former President Trump was able to connect with working-class voters by emphasizing his position as a “Washington outsider” and by “speaking their language,” adding that President Biden also appeals to his working-class roots in an attempt to win over those same voters.
The same is true for Viktor Orbán, the right-wing leader of Hungary, who has been this European country’s Prime Minister since 2010, and prior from 1998 to 2002. According to Frey, Orbán demonstrates precisely the behavior they study—describing himself publicly as a “village boy,” with an “uncultured” background. “Orbán is trying to identify himself with his targeted voters in order to signal that he’ll implement their desired policies, even though he is none of those things anymore,” says Frey.
Yet, in order to know if those observations really confirm a true empirical pattern, researchers would need to establish the same causal connection that Frey and Desai found in Brazil. Until then, all we know is that more research is needed, cautions the Rochester team.