PDF Ebook Registration and Voting under Rational Expectations: The Econometric Implications
Alone among modern democracies, the United States makes voter registration a personal responsibility rather than a governmental function. In almost all states, registration dead( lines occur well before elections. Failure to register by the deadline makes the probability of voting exactly zero. This sequential feature of the registration and voting decisions has been skipped over by most researchers, who simply ignore registration. Others, notably Timpone (1998), have used the seemingly appropriate Heckmansstyle selection model, but have arrived at findings diffi cult to believe. This paper investigates the appropriate choice of a registration model under a rational expectations assumption about the desire to vote, showing that, rather surprisingly, conventional selection models will generally perform less well than ignoring the selection effect of registration entirely. However, neither is quite correct. Finally then, the paper proposes and tests a flexible model for registration as a step toward substantively appropriate joint modeling of registration and voting.
Voter turnout is central to the theory of democratic government. Without equal par( ticipation, elections lose legitimacy. Many believe that uneven participation also biases government policy (Lijphart 1997, Griffi n & Newman 2005). Thus uneven turnout across the population has concerned political scientists for a long time (Merriam and Gosnell 1924, Arneson 1925, Gosnell 1927, Tingsten 1937), and prominent political scientists continue to study it empirically and to test our theoretical notions against the evidence (e.g., Kelley et al. 1967, Wolfinger & Rosenstone 1980, Verba et al. 1995, Blais 2000).
Turnout cannot be understood without prior study of the voter registration process. Every democracy maintains lists of eligible voters. In rural areas where everyone is known to the neighbors, the lists may be informal. Far more commonly, though, the electoral register is a written or electronic document maintained by government authorities. Those not on the list at election day cannot vote. In every country, the lists are fallible to a greater or lesser degree.
In the United States, voter rolls are a state responsibility by national constitutional provision. Each state sets its own rules for registering voters, with most requiring that citizens register themselves at least one month before the election at which they wish to vote. Becoming registered is the citizenps responsibility, typically requiring a trip to city hall, a phone request that an application be mailed, or the downloading of a registration form. In principle, most states offer a chance to register when a driverps license is obtained, but in practice, enforcement of the rule may be uneven. Only three states have elections day registration or no registration at all. The result is that many American citizens are not registered, amounting to perhaps 25% of the eligible adult population.
Statistical studies of voter registration and turnout are far too numerous to cite. (A recent review is Highton 2004.) But with few exceptions (for example, Uhlaner 1989, Jackson 1996), they have ignored the registration decision and focused directly on the decision to vote. The many studies showing that age, partisanship, and interest in the campaign make people more likely to vote arebased on such models. But in the United States, no such specification can be correct. Growing older or becoming a partisan has no effect at all on the turnout of those who are not registered. The marginal effect is exactly zero. In spite of many methodological studies attempting to improve our statistical analyses of turnout (for example, Nagler 1994, Achen and Sinnott 2009), this aspect of the turnout problem has received little attention.
download
PDF Ebook Registration and Voting under Rational Expectations: The Econometric Implications
Posted in :
