A Fast, Easy, and Efficient Estimator for Multiparty Electoral Data

Katz and King have previously developed a model for predictingor explaining aggregate electoral results in multiparty democracies.Their model is, in principle, analogous to what least-squaresregression provides American political researchers in that two-partysystem. Katz and King applied their model to three-party electionsin England and revealed a variety of new features of incumbencyadvantage and sources of party support. Although the mathematicsof their statistical model covers any number of political parties,it is computationally demanding, and hence slow and numericallyimprecise, with more than three parties. In this paper we producean approximate method that works in practice with many partieswithout making too many theoretical compromises. Our approachis to treat the problem as one of missing data. This allowsus to use a modification of the fast EMis algorithm of King,Honaker, Joseph, and Scheve and to provide easy-to-use software,while retaining the attractive features of the Katz and Kingmodel, such as the t distribution and explicit models for uncontestedseats.