The purpose of Latin American Elections is to advance an all-encompassing theory of voting behaviour in Latin America. It uses evidence from developed democracies to identify the set of characteristics that make up the Latin American voter. It combines demographic variables with socio-economic determinants, political anchors and topical issues to explore voting patterns across 18 countries in the region. It uses the funnel of causality, commonly known as the Michigan model, to sequentially add variables into a single model of vote choice. It looks at an extraordinarily large amount of data stemming from three surveys to quantitatively describe similarities and differences in voting patterns across the region.

The book makes ground-breaking progress in identifying the applicability of the model in Latin America. Its major contribution is to propose an unprecedented single overarching scientifically unified model of regional voting. Indeed, despite some recent contributions, and a long list of individual country case studies, few efforts have been made in this direction. In many ways, it is the first of its kind. One of the elements that separate this book from previous literature is its use of data. For the first time there is enough information to fully represent the variables needed to test the theory. It presents the most complete model tested yet, with data stemming from over 50,000 individuals polled between 2008 and 2012.

It is divided into six chapters. The first chapter covers the evolution of public opinion research in Latin America and describes the empirical basis of the book. The second chapter looks at demographic determinants related to age, gender, religion, region and race. The third chapter follows in the same vein and looks at socio-economic variables, like class, education and unemployment. Similarly, the fourth chapter looks at the incidence of social psychological anchoring forces, including party identification and ideological orientation. The fifth chapter turns to the influence of short-term issues, such as crime and the economy. The sixth chapter summarises the comparative takeaway.

The book shows how some variables are more salient than others when it comes to vote choice. For example, within the demographic determinants, race and religion emerge as the most important predictors, while gender and age appear to be less meaningful. Among the socio-economic factors, material wealth is surprisingly found to be the most relevant, relegating education, income and employment to a secondary position. Notably, the social psychological set of variables seems to have the largest impact, with partisan identification overshadowing ideological orientation as the principal driving force. Finally, among the short-term issue variables, corruption, democratic rights and the economy are found to be the most relevant.

Now, when considering the model as a whole, the demographic and socio-economic cleavages account for only a small portion of the variance in vote choice. In comparison, the social psychological anchors have a significantly larger effect. Both partisan identification and ideological orientation seem to be the fundamental driving forces behind vote choice in the region. Their step-wise inclusion significantly drives up the explanatory capacity of the model. Remarkably, after they are incorporated into the model, following the demographic and socioeconomic variables, the predictive capacity increases to about 40 per cent from about 10 per cent.

Naturally, the model works better in countries in which parties have stronger linkages with their citizens, such as Uruguay, the Dominican Republic, Venezuela and Nicaragua. In contrast, it performs relatively worse in democracies in which the link is weaker, such as Paraguay, Colombia and Brazil. However, the trend is neither neat nor uniform. There are some significant deviations from the pattern – most notably, the case of Chile, in which ideological identification is a substantially better predictor than partisan identification. The authors attribute this statistical anomaly to the legacy of the country's particular authoritarian experience and subsequent transition to democracy.

It is important to note that the explanatory capacity of the Michigan model is weaker in Latin America than it is in established democracies. Indeed, the findings in the book suggest that the Latin American voter is much more similar to the Eastern European voter than to the United States or Western European voter. Indeed, what Latin America and Eastern Europe have in common is their less pronounced development of political anchors. In comparison to advanced democracies, party and ideological identification is weaker in developing countries. The authors predict that as democracy consolidates, so will the applicability of the model. In the meantime, voter choice will remain related to circumstantial factors.

The findings in this book will be useful to students of Latin America. As the first attempt to create an overarching theory of voters it will naturally consolidate as a stepping stone for future research. While some readers will find it too broad, others will find that it provides the perfect point of departure to embark on further research. The book will also be of value to experts in electoral studies. As a leading theory in the regional subfield, it will provide a crucial point of comparison for those who study elections across the globe. It will be especially useful for scholars of developing democracies, such as those in Eastern Europe. Ultimately, it will contribute to the development of an integral theory of voting behaviour.

Some questions on the applicability of the model remain. For example, recent changes in the political direction of the region raise doubts as to how useful it will be in the near future. The recent trend of deconsolidation casts a shadow of doubt on its validity. At the time being, there is some evidence that the region is reversing into a populist and authoritarian place, which alters the premises that ground the theory. Also, some questions of omitted variable bias remain. Since the book suggests that short-term issues are relevant predictors, the large differences across the region present difficulty for the consolidation of a parsimonious theory. The variety of individual country cleavages contrasts starkly with the notion of a uniform Latin American voter.

Overall, the book shows that the funnel of causality does a good job at explaining individual-level vote intention in Latin America. It shows that the Michigan model is both applicable and useful to explain how Latin Americans vote. It shows that long-term determinants are especially important for explaining electoral behaviour, with the particularly strong influence of partisan identification. These findings allow for positioning the region on the map and comparing its voters to those elsewhere. As the first of its type it will naturally consolidate as the main avenue towards further investigations into the behaviour of Latin American voters.