Is Voter Fraud Like Littering?: Empirical and Methodological Considerations
The dispute over the extent of voter fraud and voter identification is certain to affect to the 2012 elections. But this debate is muddled in a methodological and evidentiary mess, conducted with little or no reliable data. This article examines the methodological issues in studying voter fraud. The basic argument is that arguments about fraud are often made without reference to a methodology dictated by good social science research. In effect, assertions of voter fraud often invoke untestable claims. Second, inference that the few reported instances of fraud are proof of more extensive occurrences is baseless to the extent that parallels are drawn to speeding or littering. The latter do not pro- vide an analogy to voter fraud. Thus, assertions about voter fraud have generally failed to provide serious social science evidence or testable propositions to test claims.