The forthcoming article “Can Descriptive Representation Help the Right Win Votes from the Poor? Evidence from Brazil” by Zuheir Desai and Anderson Frey is summarized by the author(s) below.
Right-wing parties represent the interests of the wealthy and yet often win elections in many developing democracies that have a preponderance of poor voters. Not surprisingly, existing research discounts the use of descriptive representation, i.e. nominating a representative that mirrors the characteristics of the poor, as a potential explanation for this pattern. If voters value class-based descriptive representation, it is implicitly assumed to be the mainstay of the Left. In this article, we present a novel theory that defies this conventional wisdom, supported by empirical evidence from Brazil.
Our model combines four elements. First, poor voters prefer the ideology of the Left-wing party, while wealthy voters are aligned with the Right. Second, when parties care about policy, any deviation from their ideological brand is not believable in the eyes of the voters. Third, any policy deviation from the party brand may be believable if the party nominates a candidate that is descriptively similar to the opposing class. Fourth, candidates that are descriptively closer to the poor, and thus less educated, are more costly to nominate.
The theory implies the following: when ideological brands are recognized by voters, the Right chooses to reduce policy differentiation with the Left where the electorate is extremely poor. Since voters recognize party brands, the Right cannot simply commit to pro-poor policies. In order to demonstrate such commitment, they nominate candidates that descriptively represent the poor. As the electorate becomes wealthier, the policies offered by the parties diverge, in line with their ideological positions, and candidate profiles converge. It is necessary that both parties and voters know and care about programmatic brands for these results to hold. When programmatic labels have little meaning, both policies and candidate profiles converge everywhere.
We test these predictions using data from Brazilian municipalities. Brazil offers a suitable environment to test our theory given the unequal distribution of income across municipalities, and a consensual Left-Right divide between the main parties. Our measure of pro-poor policy at the municipal level is the share of the municipal budget devoted to health, sanitation, education, and housing, while our measure of descriptive representation of the poor is the education level of the mayoral candidates.
Our results support the theoretical claims. In low-poverty municipalities, Right-wing mayors spend significantly less on pro-poor categories than the Left, as one would expect, and both parties field highly educated mayoral candidates. In high-poverty areas, policy differences disappear, as both Right and Left-wing mayors increase their pro-poor spending to similar levels. However, Right-wing candidates are significantly less educated than their Leftist competitors. Furthermore, these patterns of policy differentiation and candidate selection are stronger in races where local party coalitions are highly aligned with national party brands, and all but disappear in ideologically diffuse races.
Overall, these findings have significant implications for the study of electoral politics in developing democracies. First, we show that descriptive representation of the poor is an important electoral tactic in this context. Furthermore, given the relative paucity of descriptive representation of the poor around the world, we find that it is surprisingly the Right that capitalizes on this shortage. Second, we show that there is more to descriptive representation than its effect on substantive representation, i.e. its effect on policy advocacy by representatives. In particular, we find that parties could also use it as a tool to convey their commitment to off-brand policy shifts.
About the Author(s): Zuheir Desai, Assistant Professor, School of Global and Public Affairs at IE University and Anderson Frey, Assistant Professor, Department of Political Science, University of Rochester. Their research “Can Descriptive Representation Help the Right Win Votes from the Poor? Evidence from Brazil” is now available in Early View and Will appear in a forthcoming issue of the American Journal of Political Science.
Speak Your Mind