Ever since Aristotle, conventional wisdom has been that a robust middle class is a sine qua non for stable democracy. Put simply: no middle class, no democracy.
For decades, modernization and democratization theorists believed the prospects for stable democracy were grim in Latin America since there was “no middle class to speak of.”1 Conversely, others found evidence of a growing middle class, but warned about the potential for political destabilization in the face of middle-class mobilization2 and the breakdown of cross-class alliances.3 And more recently, multilateral banks and the media have hailed the growth of the middle class in Latin America, attributing it to a felicitous mix of economic stability, economic growth and innovative social programs.
These competing views raise two questions: 1) How large is the group of individuals who consider themselves to be middle class in Latin America and the Caribbean today? 2) Will this group be a force for democratic stability or political decay?
Our analyses below suggest that Aristotle was not always right. Using a preliminary dataset from the 2012 AmericasBarometer survey by the Latin American Public Opinion Project (LAPOP) at Vanderbilt University, which included over 41,000 interviews in 26 countries, we found that the self-defined middle class has become the largest class grouping in the Americas today. But their political and democratic attitudes and practices are not that different from those who identify with the lower classes, and, in one case, even less democratic.
How Large? The data show that a plurality of citizens of the Americas (41.0%) perceive themselves as belonging to the middle class. When we combine those who identify themselves as “lower-middle” class (29.2%), we see that a large portion of Latin American societies (70.2%) consider themselves broadly in the middle class.
These figures are based on a question from the 2012 AmericasBarometer survey, which asked respondents to identify themselves as belonging to one of five class categories: lower class, lower-middle class, middle class, upper-middle class, and upper class.4 This is the measurement of class used for the analyses that follow.
We find evidence that the self-identified middle class is actually less politically tolerant than those who report belonging to the lower class, measured by the responses to questions about support for dissenting political views. As the figure here shows, this is true even controlling for place of residence, age, gender, education, and material wealth (number of household items owned). While the self-identified middle classes have a negative relationship to political tolerance, material wealth itself is a positive predictor, as is education and urban residence. We also see that women in Latin America and the Caribbean are less politically tolerant of dissent; age has no statistically significant effect.
The level of political participation of the new middle class is also not significantly greater than other classes. There is a difference, however, between protest-oriented behavior and electoral participation. The self-identified middle class tends to participate less in protests, once we control for standard socioeconomic and demographic charecteristics. But, when we turn to conventional participation (voting), both the lower-middle and middle classes are comparatively more likely to participate than the lower and upper-middle classes. In sum, those who identify with the middle class are marginally more likely to vote as opposed to taking to the streets or gathering signatures.
The self-identified middle classes are only somewhat more interested in politics than the lower class. In responses to a question asking how much interest the individual has in politics— none, little, some, or a lot—the average level of political interest, on a 0 (no interest) to 100 (a lot of interest) scale, among those who report belonging to the lower class is 31.1. Among those who consider themselves part of the upper-middle class, this mean number is 40.4. The values for those in the lower-middle and middle classes fall in between, at 35.6 and 38.7, respectively.
Despite assertions about the limited size of the middle class and its relationship to dramatic political change (positive or negative), the fact is that the middle class in the Americas is a broad and a diverse group. And despite research and theories on the importance of the middle class for democratic development and stability differences with other classes seem small—and in the case of tolerance, the opposite of what had been expected.