Conceptualizing and measuring democracy have never been straightforward. In fact, it is probably one of the most difficult research variables to quantify in terms of political and institutional change. The multivalent nature of democracy and the dynamic trends it displays at the global level make it difficult to locate what falls within its purview.

Varieties of Democracy: Measuring Two Centuries of Political Change represents one of the most systematic attempts to bring clarity to problems of validity and reliability surrounding the existing indicators and indices of democracy by dominant measures such as Freedom House and Polity IV. In so doing, the volume serves as an invaluable resource for evaluating issues of democracy across political and institutional contexts. In essence, the book is meant to serve as a guideline for the Varieties of Democracy (V-Dem) project, an online democracy dataset developed by a network of researchers from American and European universities. The project came about as a result of the “2006-7 National Research Council (NRC) consultations that concluded that no democracy indicators existed that were sufficiently fine grained and reliable to assess the impact of democracy-promotion programs” (5).

The core argument of the book is that the existing indicators and indices of democracy have largely failed to account for small changes and differences in institutional and political developments witnessed across the globe. As a result, measuring the quality of democracy and its versatility across political and institutional contexts has become virtually impossible. The book is organized around seven chapters. Chapter 1, which also serves as the introduction, provides an overview and “story” of the V-Dem project. Chapter 2 clarifies the core concepts and dimensions of democracy, arguing that the existing standards of measurement lack precision. This chapter draws on Robert Dahl's concept of “polyarchy” to provide clarity to the varying dimensions of democracy (30–39). Chapter 3 centres on data-collection procedures, including coding types. Chapter 4 introduces the measurement model of the V-Dem project and grounds for its reliability. Chapter 5 operationalizes five principles of democracy (electoral, liberal, participatory, deliberative and egalitarian) and how they are aggregated into high-level indices. Chapter 6 focuses on data validation and assessment. Finally, chapter 7 discusses the potentials of the V-Dem dataset and how to account for measurement challenges, among others.

This book has several merits, four of which stand out in my perspective. First, the V-Dem dataset is exhaustive, spanning the years between 1789 and 2018. This historical backdrop allows readers to trace the evolution of democracy indicators and indices across time and space. The rationale for this broad historical scope is premised on the assumption that the predictions of future political systems should be rooted in a sound understanding of past political institutions and trends. Second, the book introduces several indicators and indices (450+) for a systematic comparison and measurement of institutional and political developments (see chaps. 2 and 6). This material draws on more than 27 million observations collected from 202 countries between 1789 and 2018, which is important because it increases the levels of precision with regard to standards for measuring democracy. The book also introduces the V-Dem model of measurement, drawing from item response theory (IRT) modelling (see chap. 4). Third, the V-Dem dataset showcases impressive country coverage and cross-national data comparability. This dataset shows what collaborative efforts can yield over time. The project comprised 6 principal investigators, 19 project managers, 19 postdoctoral fellows, 37 regional managers, 160+ country coordinators, several graduate and undergraduate assistants and over 3,200 country experts who participated in coding the V-Dem dataset (2–4)—a level of participation that underscores how high the levels of data disaggregation are, making V-Dem an inclusive and useful dataset across the globe (see chap. 3). Lastly, the arguments presented in this book draw on a multidimensional approach to conceptualize democracy within well-defined political units. The book also integrates cross-sectional methodological procedures for collecting, coding and analyzing more than 27 million cross-country data entries.

While this is an excellent book, much credence is given to quantitative data and statistical inferences, making the arguments and discussions presented primarily accessible to scholars with a preference for numerical data. This is not to suggest that overreliance on statistical inferences represents a weakness; instead, it underscores the need for the reader to be familiar with advanced statistical tools such as Stan probabilistic programming language and Bayesian item response theory in order to fully appreciate the V-Dem model. Without this prior knowledge, the main chapters of the book (for example, chaps. 4, 5, 6 and 7) will be difficult to read. Perhaps the book targets readers and scholars with statistical preferences. Finally, while chapter 4 discusses the viability of the V-Dem measurement model, it is not clear whether the model experiences estimation challenges with certain category of indicators. This information could be useful for readers who wish to know how well the model performs across the categories of indicators discussed.

Overall, this is an outstanding contribution to democratization research and certainly a very useful read in comparative politics. The book provides important and fine-grained metrics on democratization that are likely to change how we engage with political and institutional changes across time and space. Finally, the book is likely to provide compelling evidence-based insights to the ongoing discussions on whether there is an increase or decrease in autocracies across the world and how this increase/decrease knits together (or not) with pro- and anti-democracy claims.