The merits of Neo-Downsian modeling of the alternative vote: A reply to Horowitz

In Professor Horowitz’s rejoinders (2004, 2006) to Fraenkel and Grofman (2004, 2006a), he mischaracterizes our formal results, retreats from previous claims about the conditions for the alternative vote electoral system to generate centripetal outcomes, renders explicit his dubious assumptions about voter behavior in divided societies, and greatly exaggerates the global evidence in support of pro-moderation outcomes under the alternative vote. Here we respond to Horowitz's (2004), criticism in this journal of the formal model of Fraenkel and Grofman (2004) and to the broader defense in Horowitz (2006) of majoritarian vote pooling arrangements as means of mitigating ethnic conflict in deeply divided societies.