The Hidden Costs of Clean Election Reform

Appendix

197 Appendix Variables and Cases Stated most simply, my goal in this book is to identify the conditions under which clean election reforms cause damage to democracy. In this appendix, a supplement to chapter 1, I provide additional information about the definition of the two key variables: “clean election reform” and “damage to democracy.” I also include a table listing the 122 clean election reforms undertaken from 1991 to 2006 that form the core data set of contemporary reforms I examined for the book. This table, too, summarizes what I could determine about the damage that each of these reforms did, or did not, cause. Clean Election Reform A few words are in order to specify the meaning of both “clean” and “reform.” Clean election reforms are measures designed (in formal appearance at least) to reduce fraud and error in the casting and counting of votes. As I discussed in chapter 1, these reforms aim either to authenticate voter eligibility (to prevent spurious voting), to safeguard voter insulation (to prevent the buying or coercion of voters), or to bolster vote integrity (to prevent the miscounting of votes). Clean election reform also has a temporal dimension. The focus is on reform— on changes to the conduct of elections. It is thus necessary to identify a turning point,a moment when established election procedures are altered to make them less prone to fraud or error. Often such changes are implemented quickly and 198 Appendix entirely within one electoral cycle, making it easy to identify such a turning point. There are, however, three types of reform that are hard to categorize. Pilot tests. Sometimes clean election measures under consideration for implementation are pilot-tested. For this study, I do not count such experiments as instances of reform—unless they subsequently led to plans for full-scale implementation—because of their limited and provisional scope. Thus I exclude from consideration, for instance, the pilot-testing of electronic voting in Panama (1992, 1999), Japan (1999, 2002), and the United Kingdom (2002), since these trials were small in scale and did not lead to a decision to adopt the technology for future use. Aborted reforms. Sometimes during the process of instituting a reform,problems develop that force reformers to abandon implementation midcourse. Unlike a pilot test, which is a limited and nonbinding experiment, such reforms are adopted,but some difficulty,court order,or unforeseen event compels reformers to halt implementation. For instance, the Irish government, after having pilot-tested electronic voting machines in 2002, decided to roll them out across the country. However, a 2004 report revealed major problems ,forcing the government to freeze deployment. I do include such aborted reforms for consideration in this study because they constitute attempts at implementation, and those attempts sometimes have consequences for the conduct of elections. I note in table 2 below which reforms were aborted. Incremental reforms. Sometimes clean election reforms are introduced incrementally over long periods of time and many electoral cycles. In India, for instance,electronic voting was first pilot-tested in 1982 but not fully deployed nationwide until 2004. I also include such incremental reforms for consideration in this study, as long as the first election for which they were partially introduced, or the first election for which they were fully implemented, fell within the 1991 to 2006 time period. For such cases I note in table 2, when applicable, either the initial rollout date or the date of full implementation. For the purposes of this study, in short, I include aborted and incremental reforms but not pilot tests in the category of reform. Damage to Democracy As I discussed in chapter 1, there are three types of “damage” or “harm”—I use these terms synonymously—that clean election reforms might inflict on democracy : vote depression (in the forms of legal disenfranchisement, adminis- Appendix 199 trative exclusion, and partisan demobilization), the proliferation of cheating (in either new or established modes), and alienation (whether in the form of popular protest, voter and civic disengagement, or the disaffection of election losers). Among the greatest difficulties in delineating these forms of damage are specifying magnitude and time frame. Magnitude How many people need to be alienated, how much participation needs to be depressed, or how much cheating needs to take place for democracy to be considered damaged? That is, what threshold of alienation, depression, or proliferation constitutes damage? Certainly we should consider...