When Does Electoral Volatility Occur? Evidence from Elections around the World

• In Event: Parties & Legislatures

Abstract

When does electoral volatility occur? Electoral volatility—the total change in party vote shares occurring between consecutive elections—is a central concept in the literature on political party systems. In recent decades, political scientists have learned much about the cross-national and even cross-election correlates of electoral volatility from hundreds of democratic elections around the world, finding that weak and young party systems tend to have more volatility than institutionalized ones (Roberts and Wibbels 1999). Scholars have advanced knowledge of electoral volatility by parsing it into within-system change—vote shifts among existing parties—and extra-system change—vote shifts from traditional to new parties (Mainwaring, Gervasoni, and España-Najera 2017). Still, far less is known about when volatility occurs. Do voters, when volatility is moderate or high, shift across partisan lines during campaigns or before them? Our central research question asks when electoral volatility occurs and what causes differences in the timing of electoral volatility.

For this project, we will parse electoral volatility into its pre-campaign component and its campaign component. The pre-campaign component captures how much volatility occurs between one election day and the beginning of the subsequent campaign, usually a period of several years. Volatility that occurs during this period results from slow-moving factors (such as the economy and policy-making) that shift voters in the years after an election occurs. The campaign component is volatility that occurs between the start and finish of the campaign. This is typically just a period of a few months, but these are months when citizens are most attentive to politics. Parsing electoral volatility in this new way—along the temporal dimension—can inform what drives voting behavior and election outcomes. If a country’s electoral volatility primarily occurs during the pre-campaign period, then volatility is probably a result of economic voting or dissatisfaction with the current government. If a country’s electoral volatility instead occurs during the campaign period, then shorter-term campaign-related factors influence its electorate.

Determining when electoral volatility occurs—before or during the campaign—for elections around the world requires new data collection. The existing literature on electoral volatility is rich because the data are easy to collect: the raw materials are simply election returns. To answer our question, by contrast, the researcher needs aggregate survey results from “trial heats” (“If the election were held today, for whom would you vote?”) that are administered over the course of a campaign. The researcher also needs the de facto or de jure starting dates of campaigns—dates that are not always straightforward to determine, since their designation varies dramatically across space and time. Nonetheless, we have collected such data for over three hundred elections around the world. We will analyze this new cross-national dataset to reach novel conclusions about the correlates of when electoral volatility occurs.

Authors

• Andy Baker, University of Colorado, Boulder

• Dalton Cunha Dorr

• Trenton Gregory Marlar, University of Colorado Boulder

• Samantha Moya, University of Colorado, Boulder

• Rachel O'Neal