Election Frequency and Voter Turnout

• Published date: 01 December 2023
• DOI: http://doi.org/10.1177/00104140231169020
• Author: Filip Kostelka,Eva Krejcova,Nicolas Sauger,Alexander Wuttke
• Date: 01 December 2023
• Subject Matter: Articles

Article

Comparative Political Studies

2023, Vol. 56(14) 2231–2268

© The Author(s) 2023

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/00104140231169020

journals.sagepub.com/home/cps

Election Frequency and

Voter Turnout

Filip Kostelka

1,2

, Eva Krejcova

3

,

Nicolas Sauger

4

, and

Alexander Wuttke

5



Abstract

In recent decades, liberal democracies have considerably expanded the scope

for citizen participation, calling their citizens to vote in a growing number of

popular votes. This research investigates the effects of the rising election

frequency on electoral participation. It expands on the voting calculus and

theorizes which, when, and how past votes affect current voter turnout. We

argue that all election types contribute to a common factor of election

frequency, whose high values depress turnout and reduce the effectiveness of

party mobilization even in the most important elections. We f‌ind support for

the new theory using an original database of all signif‌icant elections and

referendums held in 22 European democracies between 1939 and 2019, two

natural experiments, and survey data from the Comparative Study of Electoral

Systems. Our f‌indings shed light on contemporary participation trends and

have major implications for democratic citizenship and democratic institu-

tional engineering.

1

European University Institute, San Domenico di Fiesole, Italy

2

University of Essex, Colchester, UK

3

University of Cambridge, Cambridge, UK

4

Sciences Po, Paris, France

5

Ludwig Maximilian University of Munich, Munich, Germany

Corresponding Author:

Filip Kostelka, Department of Political and Social Sciences, European University Institute, Via dei

Roccettini 9, San Domenico di Fiesole (FI), Italy.

Email: f‌ilip.kostelka@eui.eu

Keywords

voter turnout, election frequency, political participation, elections, public

opinion and voting behavior, European politics, representation and electoral

systems, institutional reform

Introduction

The expansion of opportunities for citizen participation has been perhaps the

most striking transformation of liberal democracy in recent decades. De-

centralization, regional integration, frequent use of referendums, and other

institutional reforms have reshaped the functioning of democratic states

(Hooghe & Marks, 2003) and, simultaneously, considerably broadened the

potential scope for public involvement in political decision-making (Scarrow,

2001). But how have these reforms affected participation levels? Have they

attenuated or worsened the declining trend in voter turnout that has been

observed around the world since the 1960s (Kostelka & Blais, 2021)?

Answering these questions would help us understand contemporary par-

ticipation dynamics and contribute to scientif‌ic debates on democratic citi-

zenship and citizen involvement. Moreover, as theorists (della Porta, 2019;

Pateman, 2012), politicians (Bowler et al., 2017), and citizens (Bowler &

Donovan, 1998) call for further participatory expansion, it would also yield

critical insights for decision-makers. If the participatory scope (i.e., number of

popular votes) exerts an effect on citizen behavior in every single vote

(i.e., voting rates), evidence of such unintended consequences would help

inform future decisions about the design of electoral institutions (Fabre, 2010)

and prevent the undesirable consequences of low participation (Blais et al.,

2020).

The political science literature advances conf‌licting theoretical perspec-

tives and presents only a few fully relevant empirical f‌indings. Participatory

theorists claim that opportunities for participation should boost overall par-

ticipation levels (della Porta, 2019;Pateman, 1970;2012). By contrast,

comparativists argue that high election frequency is the reason for historically

low voter turnout in countries like the United States or Switzerland (Blais,

2014;Lijphart, 1997, p. 8). Although a number of existing empirical studies

corroborate the negative effect of frequent elections on participation in various

contexts (Fauvelle-Aymar & Stegmaier, 2008;Garmann, 2017;Schakel &

Dandoy, 2014), they typically focus on elections of the same type and status,

or on turnout in less salient, second-order elections (SOEs).

1

To answer the

key question of whether the recent transformation of democracy harms or

benef‌its participation, it is important to analyze the effect of second-order

contests, which have proliferated, on turnout in f‌irst-order contests such as

legislative elections, which pre-existed the reforms. As we review in detail

2232 Comparative Political Studies 56(14)

below, such studies are rare and the available results point to null effects,

which questions the views of both participatory theorists and comparativists.

The present manuscript conducts the most comprehensive investigation of

the effect of election frequency

2

on turnout in legislative elections to date. It

bridges the competing theoretical perspectives and formulates an explicit

theory on how, which, and why past electoral contests matter for legislative

turnout. The theory provides a novel and systematic conceptualization of

election frequency and, to describe its effects, it expands on the voting

calculus (Downs, 1957). The new theory broadens the calculus perspective

from a single election to the whole electoral cycle and explains how election

frequency affects the individual decision to vote. We hypothesize that all

separately held elections contribute to a single cumulative factor of election

frequency whose high values depress individuals’propensity to vote and

reduce the effectiveness of party mobilization. Importantly, we argue that f‌irst-

order elections (FOEs) are not immune to these detrimental effects. Drawing

on the new theory, we design a new index of election frequency,which ref‌lects

countries’recent electoral history and decays non-linearly through time.

We successively analyze three types of data: an original database of all

signif‌icant elections that were held in 22 European democracies between

1939 and 2019, two natural experiments, and survey data from the Com-

parative Study of Electoral System. The results lend consistent and robust

support to the hypothesized negative effect of all election types on legislative

turnout. The new index of election frequency (but also alternative, more naive

indicators) yields statistically and substantively signif‌icant results across the

three analyses. Our f‌indings demonstrate that participation even in FOEs

suffers from the proliferation of second-order contests and that a key

mechanism is the reduction of citizens’responsiveness to mobilization. This

means that the more often citizens are given the choice to participate in

elections, the less often they actually turn out to vote. This has important

implications for the debate between participatory theorists and comparativists;

democratic institutional engineering; and our understanding of participation

trends in the 21st century.

Literature Review

The theory of participatory democracy argues that participation is self-

sustaining and that more opportunities for participation lead to higher par-

ticipation levels. By participating in different arenas, participatory theorists

claim, citizens develop attitudes and skills that are necessary for participation

at all levels, including the most important, national level (Barber, 2003, p. 152;

della Porta, 2019, p. 612; Pateman, 1970, pp. 42–43). Recent democratic

reforms empowered citizens and extended their say at all levels of govern-

ment, often allowing them not only to elect their representatives but also to

Kostelka et al. 2233