Elections and Democracy: To Cheat or to Change?

Abstract

The Forensics of Election Fraud: Russia and Ukraine. By Mikhail Myagkov, Peter C. Ordeshook, Dimitri Shakin. Cambridge: Cambridge University Press, 2009. 289 pp., $30.99 paperback (IBSN-13: 978-0-521-74836-0). The Politics of Electoral Reform: Changing the Rules of Democracy. By Alan Renwick. Cambridge: Cambridge University Press, 2010. 312 pp., $90.00 hardcover (IBSN-13: 978-0-521-76530-05). Cambridge University Press has released two monographs recently that deal with the complex relationship of elections and democracy. Though substantively distinct, these books both address a core concern regarding the ways political actors attempt to influence electoral outcomes and the consequences of such attempts for democracy. For Myagkov, Ordeshook, and Shakin, the question is: how can we detect election fraud where we suspect it has occurred? Renwick's questions center on the dynamics of electoral system reform. Both works bring an arguably under-utilized methodological approach to bear on their questions, and both provide empirical analysis of a small number of cases. As indicated by the title, The Forensics of Election Fraud , Myagkov et al. investigate election fraud in Russia and Ukraine, while Renwick's The Politics of Electoral Reform traces the process of electoral system reform in France, Japan, Italy, and New Zealand. Myagkov et al. attribute their motivation for Forensics to the work of one Dr. Alexander Sobyanin, a scholar committed to exposing electoral fraud in Russia. In many ways, the book maintains this tight focus throughout, though one can certainly imagine theoretical extensions and methodological applications. The authors’ fraud detection efforts focus very specifically on those illegal activities that directly distort election results: ballot box stuffing and the deliberate manipulation of vote counts. They introduce three forensic indicators of election fraud, based on the logic of what fraud-free elections should look like, in terms of turnout, candidate vote shares, and the relationship between the two. Deviations from any of these three expectations might indicate fraud. First, they argue that as turnout increases so should the votes for each candidate, at a rate proportional to their support in the electorate. Second, they argue that turnout should be normally distributed across districts, with small proportions of high and low turnout districts …