Recent decades have seen growing polarization amongst political parties. This growing polarization is contrary to the expectations of the classic Downsian model which suggests parties, particularly in a two-party system, will converge on the center. This study seeks to build a model explaining increasing party polarization and public extremism. It seeks to combine the traditional Downsian model with the saliency/issue competition model of electoral politics, which assumes multiple issue dimensions in a polity. Specifically, we argue that, unlike party systems in the 1960s and earlier, modern party systems have multiple issue dimensions across which electoral competition occurs. Further, we relax the assumption that voters are normally distributed across these dimensions. Flowing from these conceptions we arrive at the logic that a political party stands to see greater net vote gain by emphasizing those issues in which the voter distribution is asymmetrical with at least two peaks and the party is alone near the modal position. In essence, we argue that parties have incentives to emphasize those issues in which there is greater polarization both among the public and among the parties. Moreover, we draw on the literature suggesting that political parties are actually capable of moving public opinion, arguing that political parties attempt to drive public opinion to the extremes so as to create issue dimensions that promise greater net vote gain. Importantly, it is also posited that parties continue to push public opinion to the extremes even after an issue dimension is asymmetrically distributed with at least two peaks. This study provides a theoretical explanation of both political party and public polarization and extremism, thus helping us to build a stronger understanding of parties, party systems, and the public in modern politics.
