Democracy is surely one of the most important concepts in political science, but it is also one of the most elusive. Two and a half millennia after its Athenian incarnation, scholars continue to debate what constitutes its essence in the modern world. Is democracy only about elections? Are civil rights, a market economy, the rule of law, or human development necessary components? Is “democracy” the same thing across different cultural contexts? As one can imagine, the challenges are still greater for those who want to reduce the concept to numbers that can facilitate the systematic study of patterns across time and space.
With this remarkable little book, Gerardo Munck succeeds in adding clarity to a muddled discussion, presenting a distinct conceptualization of democracy and putting the effort to quantify it on much more solid logical ground. Informed by both theory and practice, Munck's effort is important reading for those in both academia and the policymaking community who wish either to use or to create data on democracy or the lack of it.
After laying out the different uses to which data on democracy are put in today's world, Munck launches into a critique of existing measures and how they are used. While all are found wanting, Freedom House's Freedom in the World indices come under particularly strong criticism, primarily for their lack of theoretical grounding (e.g., why should the index be additive instead of multiplicative?), their murky coding rules that complicate replication, and their methodology's tendency to change from year to year without adjustment of prior scores to ensure consistency.
At the same time, Munck develops some important principles for how indices of democracy should be constructed. Most fundamentally, they should be theory driven. Analysts must explicitly disaggregate the concept of democracy, paying special attention to different levels of disaggregation. If democracy, for example, consists of the two attributes of contestation and participation, then each of these attributes break down into several components. Freedom of the press and the right to form parties, for example, are components of contestation. Good indices must carefully distinguish between these different levels, avoiding redundancy or the conflating of levels. In addition, theory must guide how different components and attributes are aggregated in the index. For example, Freedom House treats aggregation as an additive task, counting and averaging point scores. But Munck ultimately argues for a multiplicative approach, meaning that if a country scores a zero on an essential component of democracy, it scores a zero on the larger index.
The central contribution of Measuring Democracy is to propose an actual measure, the Electoral Democracy Index (EDI), which the author together with Jay Verkuilen developed to evaluate democracy in Latin America for the United Nations Development Programme (UNDP). The index itself is elegant and commonsensical, breaking “democracy” down to the following four attributes: the “right to vote,” “clean elections,” “free elections,” and the filling of the most important state offices (legislative and executive) by elections (p. 55). Importantly, “each attribute is held to be a necessary condition” (p. 57). Thus where an additive index would rate a country with universal suffrage and clean elections for all important state offices, but with only half-free elections, as still seven eighths democratic (i.e., pretty good), Munck's index would rate that country only half democratic. Munck also develops an admirably clear scale that expert coders can use to assign the corresponding numeric values, a scale developed according to principles he helpfully lays out. He presents specific data from Latin America that reveal the index to be highly reasonable.
While compelling, the EDI does have some weaknesses, at least as described in the book. For one thing, the multiplicative nature of the index can magnify any problems in the definition of the attributes or the coding rules. Take, for example, the attribute of clean elections. In essence, on a three-point scale, a country scores a zero if there are major irregularities that determine the outcome of an election, and a one if irregularities exist but do not have a major impact on the outcome (p. 58). But where an election is close, a small amount of fraud can determine the outcome. Thus if we assume a small amount of fraud is a constant, it could be the case that a country moves from a one to a zero on the “clean elections” criteria not because fraud increased, but because the election got more competitive, meaning that the fraud became decisive. The consequences would be severe, as the zero would multiply with the other attribute measures to produce a zero for the whole democracy rating, equating such a country with North Korea on the EDI.
The book is also not clear on exactly how what many (e.g., Steven Levitsky and Lucan Way, “Why Democracy Needs a Level Playing Field,” Journal of Democracy 21, no.1 (2010): 57–68) call “the level playing field” factors in to the EDI. This would seem to fit most naturally under “free elections,” but Munck explicitly notes that the “free elections” attribute “does not include factors that affect the ability of parties and candidates to compete in equality of conditions, such as public financing, access to the mass media, and the use of public resources” (p. 58). It is also not included in the other categories as far as I could tell. This is puzzling, because later in the book, Munck develops a very promising method and associated index (the Index of Democratic Elections) for evaluating whether elections are democratic that explicitly includes a level playing field for candidates as an indicator of whether elections are competitive (p. 90). The apparent omission of the level playing field from the EDI renders it unable to make crucial distinctions among degrees of democracy in highly clientelistic social contexts, where regimes can be very sophisticated in manipulating mass media and economic levers to disadvantage opposition.
These particular problems are, of course, eminently correctable within the framework of the index, and indeed part of the value of the EDI is that it lends itself to relatively easy adjustment. And I do not rule out that these issues may simply be the result of a lack of clarity in the book that would easily be sorted out in practice. But because the multiplicative nature of the index can magnify certain minor problems, great care must be taken when employing it. And the need for such care, including the need to avoid relying solely on an index for one's evaluations, is also a point the author himself makes.
Munck concludes the book with a discussion of the meaning of “democracy” more generally and how to develop measures that go beyond the EDI, which captures only “electoral democracy.” The author rules the rule of law out of the definition of full democracy while arguing that “the attainment of social integration” (p. 127) is a necessary component. Not everyone will agree, of course, but the book breaks important new ground and is an excellent example of social science fruitfully applied to significant real-world problems.
