Our book and that by James Loxton approach important questions in Latin American politics with very different research strategies. We are glad that Loxton saw the value of our work and how it might reshape scholars’ understanding of who participates in politics and how. We also appreciate this conversation that highlights the contributions of quantitative and qualitative research.
Loxton’s main critique focuses on how measurement choices related to our Political Participation Index might undermine the book’s main finding—that poor people are overall more politically active in Latin America than the nonpoor. Luckily, this is an easy matter to resolve empirically. During our analysis, we estimated models using a version of the index that gives contacting the same weight as voting and protesting, and we can report that the substantive findings are very similar to the ones in the book. In making his critique, however, Loxton is pointing to a bigger question that we thought a lot about: How should we best measure the dependent variable and ensure that our findings do not depend on how we operationalize political participation?
In the end, the findings in the book are very robust to different specifications, but we faced some difficult trade-offs. For example, we initially created an index of political participation that included eight different modes of political participation asked in LAPOP surveys up until 2012. This index was more robust and less susceptible to bias from one kind of political act alone. It also produced a stronger negative relationship between wealth and political participation than the 4-point index we eventually used. One of the most difficult decisions we had to make was whether to include our full index of political participation valid only up until 2012 or to use the 4-point index that allowed us to bring the analysis closer to the present. Given that the core substantive results were essentially the same regardless of how we constructed the index, we opted for using the most up-to-date information available to us at the time.
Perhaps more importantly, readers should not worry that our results depend on how we constructed the Political Participation Index. As we describe at length in chapter 3, we sometimes use the additive index, but the bulk of the analysis is based on separate statistical models for voting, protesting, and contacting government officials. In this way readers can see for themselves how each political act contributes to or detracts from political equality.
We agree with a second point raised by Loxton: the book would have benefited from the inclusion of case studies to bring the numbers to life and help tease out causal mechanisms, exactly the kind of rich case-study material that makes Loxton’s book so compelling. That said, we drew heavily from our extensive fieldwork experience in several countries and from excellent qualitative studies by other scholars that enriched our understanding of political participation, partisan mobilization, and community organizing. In the end, this book came together as a complex quantitative story, one that we believed was important to make available.
