the study of voter turnout: state of the art · 2008-04-15 · democratic participation and...
TRANSCRIPT
FIFTH FRAMEWORK RESEARCH PROGRAMME (1998-2002)
Democratic Participation and Political Communication in Systems of Multi-level Governance
The Study of Voter Turnout: State of the Art
Pat Lyons & Richard Sinnott
Institute for the Study of Social Change Public Opinion & Political Behaviour Research Programme
University College Dublin Ireland
Work in Progress
March 30, 2003
Draft text not to be quoted without permission of the authors
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 1
Introduction This paper provides a critical review of the literature on electoral participation in
systems of multi-level governance. Its purpose is to provide a launch pad for the
project “Democratic participation and political communication in systems of multi-
level governance” [Part-com Multi-Level] being conducted under the auspices of the
Fifth Framework Programme.1 It aims to do this by extracting from the existing
literature the implications and lessons, theoretical and methodological, that can be
applied to the project.
Voter turnout has tended to fall over the last number of decades. This fall accelerated
in the 1990s. The trend is paradoxical because education and access to political
information have generally increased and such factors are seen to conducive to
increased political participation. Both individual and aggregate-level analyses of voter
abstention point to the factors that are responsible for decreasing participation.
However, there has been little work on linking individual perceptions of the political
process with the aggregate level context in which citizens live.
This connection is at the heart of explaining why political participation changes.
Unfortunately previous research has provided a partial picture of this process and how
it works. Understanding why voter turnout is declining requires new data resources,
improved research designs and a more nuanced conceptualisation of both turnout and
abstention. The purpose of the Partcom Multilevel Project is to bring these new
resources and new approaches to bear on the problem and, in this way to substantially
enhance the state of the art in regard to the study of participation with particular
reference to participation in systems of multilevel governance.
Specifying the dependent variables: Who are the voters and the non-voters?
Surprisingly, many studies of political participation do not explicitly delineate the
object of their study – voters versus non-voters. This is most evident in terms of two
aspects, voting population definition and unit of analysis. First, most definitions
1 Details of the programme are available at the following website; http://www.ucd.ie/~dempart/
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 2
reflect the different types of datasets used, whether individual or aggregate, national or
cross-national, single time points or time series. This leads to differing
categorisations of what constitutes voters and non-voters. Whether one uses the
‘voting age population’, ‘registered electorate’, ‘voting eligible population’ can have
fundamental implications for deciding the level of political participation and whether
or not it is changing over time (McDonald & Popkin 2000).
Secondly, post-election surveys, which depend on voter self-reporting have been
characterised by substantial “overreporting”. Response validity research has
attempted to verify, through consultation of public records, whether or not citizens are
registered to vote and have voted as they claim in election surveys (Silver, Anderson
et al 1996; Swaddle & Heath 1989). Such research is important as it leads to an
assessment of whether those who misleadingly claim to have voted are more similar to
persistent abstainers than to regular voters (Abramson & Claggett 1992; Presser &
Traugott 1992). The most recent evidence suggests that infrequent abstainers are
more like regular voters and that their reasons for not voting in specific elections
relate to circumstantial factors such as work, travel or illness (see, Blondel, Sinnott &
Svensson 1998). This highlights the importance of viewing non-voters as a
heterogeneous group. Figure suggests that categorising individuals into five groups,
i.e. core voters, marginal voters, non-voters, non-registered citizens and non-citizens
yields additional insight into the dynamics of changes in voter turnout (see figure 1).
<<< Figure 1, about here >>>
Such a categorisation shows that changes in voter abstention can arise from a variety
of causal mechanisms, some of which relate to specific factors such as interest in a
particular election, personal circumstances and some of which arise from more general
contextual factors such as the socio-demographic context.
Theories of voter turnout Sociological theories of voting behaviour focus on life conditions that have an impact
on individuals’ ability to participate in electoral politics. One key variable often noted
is age, with increasing age being seen to increase political experience and sense of
investment in the community. Sociological perspectives frequently focus on the
concept of ‘social exclusion’. The argument here is that if voters do not have
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 3
adequate skills and resources for electoral participation then turnout will be low,
because socio-economic and educational deprivation transforms social exclusion into
political exclusion.
Psychological or ‘political mobilisation’ theories of voting behaviour focus on
what motivates individuals to go to the polls. The available evidence since the 1980s
suggests that, while interest in politics is strongly related to turnout, turnout has
declined in a context of generally increasing interest in a wider range of issues dealt
with in the media. One explanation advanced for this phenomenon has been
‘negative’ media reporting and the portrayal of politics as a “game” in which there is
great emphasis on the results of opinion polls (Ansolabehere & Iyengar 1995: 112;
Norris 2000; Heath & Taylor 1999). In addition contemporary research on the role of
the media and its impact on turnout, has moved toward an examination of both
individual level and context effects (see, Semetko, Blumler, Gurevitch & Weaver,
1991; Norris, Curtice, Sanders, Scammell & Semetko 1999).
Economic theories of voting behaviour have often argued that it is irrational to
vote because of the very small chance of deciding the outcome and the costs incurred
in going to the polling station, where little benefit is gained. One motivation put
forward for voting has been maintenance of the regime (Riker & Ordeshook 1968;
Enelow & Hinich 1990). Strategic or structural approaches to turnout are related to
economic theories - individual voters are seen to estimate how powerful an electoral
institution is when deciding whether or not to turnout to vote. Consequently, turnout
for second order elections is relatively low, though voting is seen to be effective in
sending signals to where power is concentrated, i.e. national governments. Blondel,
Sinnott & Svensson (1998) argue that the empirical evidence on such strategic
considerations for European Parliament elections overestimates the impact of
individuals’ perceptions of the structures and locus of power and accountability and
that such considerations do not explain differential turnout patterns across countries or
over time. The contribution of “economic” or rational choice approaches to the
explanation of electoral participation is taken up in more detail below.
Situational explanations of voting tend to fit with economic explanations of
voter participation by emphasising empirical evidence that indicates that turnout is
affected by the voting day (weekday or weekend), hours of voting, accessibility of
polling stations, ease of voting registration, facilities such as postal voting, etc. (see,
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 4
Franklin 1996, 1999). Other considerations noted include the density of polling
stations, which may have an impact on the cost of voting as the economic theories of
voting suggest. Empirical research has used elements of these different theories to
explain voter participation, resulting in some general explanations of increasing voter
abstention being put forward.
Eclectic empirical explanations: Recent empirical explanations of electoral
participation tend to use aspects of the four main theories of electoral participation.
These eclectic empirical explanations come in three main forms. The first argues that
individuals’ decisions as to whether or not to vote are based on a individuals'
“resources”. The second perspective highlights the level and degree of “political
mobilisation” while the third strand emphasises the importance of “instrumental
motivation”. The “instrumental motivation” approach is often divided into
“institutional” and “systemic” factors that are of course strongly linked (Franklin
1996; Norris 2000).
Resource explanations look at voter abstention from the individual perspective.
In this approach socio-demographic evidence from survey data is used to illustrate the
importance of variables such as age, level of education, socio-economic status, etc.
Nevitte et al. (2000: 21) using CSES data for eighteen countries found that “SES, and
particularly age, education, attendance at religious services and household income,
continue to influence non-voting, regardless of levels of aggregate economic wealth,
electoral history, electoral rules and number of parties”. Within the literature on voter
participation at European elections there has been debate over whether or not
knowledge and interest in European integration and its institutions affect turnout
(Franklin 1996; Blondel, Sinnott & Svensson 1998).
Mobilisation accounts of electoral participation highlight the importance of
organisations in facilitating greater levels of voting by individuals. These
organisations range from the formal such as political parties, trade unions and
churches to more informal ones such as voluntary groups, families, work colleagues,
etc. (see, Wolfinger & Hansen 1995). A decline in the importance of these formal and
informal organisations has been seen as contributing toward the long-term decline in
voter turnout in the United States (Knack 1992; Putnam 1995: 67-68; Aldrich 1995;
Wattenberg 1998; Putnam 2000). From a policy perspective, Abramson & Aldrich
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 5
(1982) made the important discovery that in the United States changes in attitudinal
factors which worked to reduce turnout had a greater impact than institutional factors
such as extending the franchise and costs of registration which should have boosted
participation.
Institutional explanations have focussed on legal criteria such as voting rights,
registration rules, and type of electoral system. Jackman & Miller (1995) in a cross-
country analysis found that political institutions and electoral laws explained most of
the variation in voter turnout in national elections. Working at the supra-national
level, research into voting in European elections has found that variations in
participation between member states can be largely explained by systemic features
such as proportionality, compulsory voting, closeness of the most recent national
election (see, Franklin et al. 1993; Oppenhuis 1995; van der Eijk et al. 1996).
Relevant systemic features noted in the United States are registration provisions and
the frequency of elections (Wolfinger & Rosenstone 1980).
The Rational Choice approach and empirical research In general, the most influential theoretical approach to voter turnout during the 1970s
and 1980s concentrated on individual-level rational choice explanations. However,
most empirical research on voter participation has tended to deal with aggregate level
data and macro-level explanations based on sociological and institutional factors. (see
tables 1 & 2). This cross-level division resulted in confusion. Consequently, by the
early 1990s the voter turnout literature had two sets of hypotheses concerning voter
participation that were not only mutually exclusive, but also in some cases
inconsistent (Lane & Ersson 1990).
<<< Tables 1 & 2, about here >>>
For example, empirical analysis of aggregate level data suggested a number of
conflicting mechanisms affecting electoral participation. Such inconsistent
hypotheses derived from the failure to develop a theoretical conception of voters’
motivations. In contrast, individual-level hypotheses based on rational choice theory
were criticised for being rather simplistic and accounting for changes in turnout and
not the more fundamental question of why voters turn out to vote in the first place. In
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 6
addition, the rational choice approach was criticised for being unrealistic (i.e. large
numbers of citizens do vote) and for ignoring the fact that the costs of voting are small
and real benefits may derive from voting (Crepaz 1990, see also, Grofman 1993: 94).
However, rational choice attempts to introduce the idea of “side payments”,
where group leaders encourage turnout in a group, such as a trade union, in order to
have more influence over candidates or the idea of voting on the basis of patronage
have been rejected as being empirically unproven (Uhlaner 1989, Schwartz 1987,
Green & Shapiro 1994: 53). Because of the theoretical, empirical and methodological
difficulties involved recent rational choice research has modified its theoretical
assumptions and used alternative empirical techniques such as simulation.
The “paradox of voting” issue has also been examined using game theory.
Using a game theoretic, rather than decision theoretic approach as espoused by
Downs, has one very important implication. Game theoretic solutions are based on
the idea that for all of the electorate to abstain cannot be an equilibrium point because
each voter could then decide the election by opting to vote. By using mixed strategies,
game theory models indicate the existence of many Nash-equilibria with high voter
turnout. However, most of these (non-trivial) asymmetric equilibria involve accepting
that there is a high level of coordination among voters. Not surprisingly such game
theoretic results are seen to be unrealistic.
As a result, the modelling of coordination in ‘voting games’ has become an
important area of research. One method of solving the coordination problem has been
to argue that the common knowledge held by voters is limited. Consequently, voters
are uncertain about either the payoffs of voting or the number of players (Palfrey &
Rosenthal 1985; Myerson 1998). In such modified games many of the remaining
asymmetric equilibria result in predictions of low voter participation just as the
Downsian decision theoretic approach predicted. Thus, the “paradox of voting”
returns - with positive costs of voting, the expected number of people who vote will
be a small proportion of the total electorate (Palfrey & Rosenthal 1985: 64).
One attempt at finding an answer as to why so many citizens do vote in elections
is based on changing the payoff assumptions. The most influential expression of this
view (Riker & Ordeshook 1968) was that the costs of voting are negative because
individuals are socialised to see voting as a responsibility of citizenship.
Unfortunately, using the idea of negative costs to answer the paradox of voting
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 7
question fails to solve the problem of why voters vote in the first place. More
recently, two other alternative approaches have been suggested.
In the first, voting is seen to be costly and the payoffs of voting are seen to be
positive as in the classic decision theoretic approach. However, the assumptions of
rationality are changed. Voters are seen to use adaptive rationality where “they learn
to vote or stay at home” by a process of trial and error. This mechanism is based on
reinforced learning where individuals decide whether or not to vote on the basis of an
‘aspiration level’ which defines whether or not election outcomes and payoffs are
positive or negative. The individuals’ aspiration level is itself determined by prior
experience, i.e. the payoffs from previous elections. This complex model based on
neo-institutionalist and behavioralist theories was tested on simulated data. Solving
the voter coordination problem through adaptive rationality suggests that while the
costs of electoral participation may not be inconsequential, turnout can still be high.
In addition, high participation will exist even if there are large numbers of voters and
the probability of casting the deciding vote is very small (Bendor, Diermeier & Ting
2000).
Recent work by Sanders (1996; 2001) has shown that if one wishes to relate
information and uncertainty to voter turnout, this should be undertaken by modelling
turnout and vote choice together. Registered voters evaluate electoral choices
differently to those citizens who decide not to register. Furthermore, while
information will increase the probability of registration, its effect on turnout depends
on the intensity of individual preferences. With more information the precision of
evaluations of electoral options increases, as does the probability of voting, to a point
where small perceived differences between electoral choices results in abstention.
Such recent research suggests that there has been a growing tendency for
empirical work to develop stronger theoretical foundations and move away from
highlighting empirical observations that are explained on a post-hoc basis. At the
same time, theoretical work has attempted increasingly to formulate testable
hypotheses, which may be investigated using survey or aggregate level data.
However, all empirical work based on statistical analysis has had to deal with a
number of methodological issues depending on the type of data used.
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 8
Methodological Issues Most published work on voter participation in elections has adopted one of the
following four main approaches to data analysis. The choice of method depends on
the research question being investigated and also on the data available. Given the
inconsistency of some of the published empirical results, there is a growing concern
that the statistical methods used on some datasets may have been inappropriate. One
can distinguish five methodological approaches.
1. Cross-sectional analysis of several independent variables, some of which
are post hoc measures, and turnout across different countries for national [lower chamber] parliament elections.
2. Longitudinal analysis of several independent variables and turnout across
time for a specific country's [lower chamber] parliamentary elections. 3. Correlation analysis of electoral participation where statistical techniques
such as regression, factor analysis and some forms of ecological inference explain variation in participation, the dependent variable, in terms of variation in independent variables such as age, education, closeness of competition etc.
4. Survey data analysis of several independent variables and turnout for
national [lower chamber] parliament elections. 5. Ecological inference analysis, which attempts to make inferences about
individual behaviour and motivations using aggregate data.
By and large the main approach has been to use a regression model where the
dependent variable (turnout) is related to a number of independent variables such as
age, institutional factors, closeness of electoral competition, etc. This raises two
fundamental issues. Firstly, with aggregate analyses (often associated with
approaches 1-3 above) specific independent variables such as, for example, closeness
of electoral competition should be examined carefully since actual election closeness
may not be a good surrogate measure of perceived election closeness (see, Wolfinger
1993).
Secondly, the assumption that census records or the electoral register provide an
accurate reflection of the true voting age population may not always be legitimate
(see, McDonald & Popkin 2000). The administrative mechanisms for the inclusion
and removal of names from the voting register may be influenced by exogenous
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 9
factors such as lifestyle, e.g. working in an urban area but voting in a rural ‘home’
constituency; jury service for the court system – this is often a cost voters wish to
avoid, and emigration. Factors that could deflate turnout might include multiple
registration(s) with a mobile working population and the administration’s inefficiency
at removing the names of deceased electors or emigrants from the register. These
factors could result in a register that inflates or deflates voter turnout for exogenous
reasons. In addition it implies that the basis for cross-sectional survey work may be
systematically biased. More specifically the following problems can be identified.
Methodological problems with cross-sectional analyses: Firstly, if country is
used as the unit of analysis inferences about the relationship between turnout and
various independent variables, which reflect individual voter characteristic, may be
vulnerable to the ecological fallacy. Secondly, turnout research that focuses on one
particular type of election may fail to take account of the fact that there may be a
variety of other elections or referendums taking place simultaneously. Consequently,
within such cross-sectional analysis, some of the time series pattern or deviations are
really “intra-election effects” (see, Foster 1984). In other words, there is the danger of
confusing cross-level and cross-time effects and thus making inferences about turnout
problematic.
Methodological problems with longitudinal analyses: One important limitation
to current longitudinal studies is that they do not cover a sufficiently long time period.
Links between important variables such as political competition are more likely to be
found if a sufficient degree of variance, i.e. history, is built into a study. Thus, for
example if a country experiences less competition between parties due to an emerging
consensus on the main issues, it is plausible to expect a decrease in turnout relative to
previous levels, assuming control is made for a secular decline in international trends.
Methodological problems with correlation analyses: Some of the turnout
literature based on formal approaches such as rational choice, social choice and game
theory highlight the implications of differences between countries in the size of each
country’s mean vote, implying that the probability that a single voter will decide an
election differs from the probability in other countries. Consequently, voters will have
differential benefits from voting and candidates / parties will have a differential
incentive to influence voting in some countries more than others. In practice, this
might suggest that higher turnout is the result of a candidate(s) or party(s) intensive
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 10
campaign strategies. Therefore, turnout and election closeness, institutional factors,
age, etc. will be correlated because of the candidate or party strategy - but this
correlation may have nothing to do with actual turnout and closeness of competition,
institutional factors, age, etc.
Methodological problems with survey analyses: In general survey data results
may be misleading if individual voters expectations are correlated with their electoral
preferences. Two approaches have been adopted with survey data. Firstly, one may
test whether on average turnout is influenced by specific factors. Secondly, the
attitudes or attributes of a respondent may be correlated with such respondents’
intention to vote. Both of these research strategies have some problems. Voter
participation and variables such as expected closeness of the election, institutional
features, age, etc. may be correlated with one another for reasons that have little direct
influence on motivations to vote.
In addition, all survey research is based on two key features. Data are derived
from respondents self-reporting of their own future or past behaviour – actual voting
behaviour is not measured. Also, mass representative surveys are not primarily
designed to measure the impact of geographical factors (King 1996: 159).2
Unfortunately, most studies of voter turnout do not use a coherent and consistent
research design. In methodological terms, this has resulted in under specified models
of voter turnout. Parsimonious models with less than five explanatory variables have
strong explanatory power but are unrealistic (Cassel & Luskin 1988). In contrast,
more realistic complex models yield more ambiguous results with weak explanatory
power. Such problems indicate the need for analyses based on different types of data
and theoretical mechanisms to overcome ‘identification’ problems (Manski 1996). In
essence, what is required is a multi-level research strategy.
Voter turnout at European Parliament Elections Individual level explanations of voter turnout for European elections since 1979 have
resulted in two distinct positions. These two positions take different views on
whether voting in European elections is related to the level of information or apathy
2 King (1996: 160) states that “geographic variation is usually quite large to begin with, but after we control for what we have learned about voters, there isn’t much left for contextual effects. So in this narrow sense geography matters but contextual effects do not” (italics in original).
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 11
among Europeans about the Union, the salience of Union issues to citizens and
whether or not ordinary Europeans support the integration project and its various
institutions.3
Firstly, investigation of voter turnout in the European elections of 1989 & 1994
found that citizens’ attitudes toward the European Union (EU) and European
Parliament (EP) did not have a significant impact in explaining whether or not
Europeans voted (Schmitt & Mannheimer 1991; Oppenhuis 1995: 65; Franklin et al.
1996). Secondly, using data from a post-election survey contained in Eurobarometer
41.1, Blondel, Sinnott & Svensson (1998) found that attitudes toward the EU and EP
did have a significant impact on explaining whether or not citizens turned out to vote.
Aggregate level explanations have found that compulsory voting, simultaneous
elections, timing in the electoral cycle, Sunday voting, proportionality of the member
states electoral system all increased voter turnout ceteris paribus (Oppenhuis 1995:
75; Franklin et al. 1996; Blondel, Sinnott & Svensson 1998: 245; Smith 1999: 118-
120).4 Such results are perhaps more correctly seen as the interaction between
individual and aggregate country level effects.
The most influential theory of EP elections is the “second-order elections”
model of Reif & Schmitt (1980). The key idea in this theory is that voter turnout is
lower in EP elections because they are not as important as national elections as there
is no election of a European government. However, one implication of the theory is
that as the EP gains more powers, as was the case with the (Maastricht) Treaty on
European Union (1993), voter turnout should increase. In fact, the empirical evidence
indicates that voter turnout has decreased in succeeding EP elections, despite the fact
that the EP has gained more powers. One other shortcoming of the “second-order
theory” is that it does not explain why there is such variation in voter turnout between
countries within the EU.
Blondel, Sinnott & Svensson (1998: 13-19, 243) argue more generally that the
survey evidence from the 1994 EP elections does not support the “second-order
election model” explanation of voter abstention. It was found that EP election turnout
3 This has been one of the key themes in EP electoral research, see Blumler & Fox 1982; Blumler 1984; Reif 1985b; Schmitt & Mannheimer 1991a; van der Eijk et al. 1995; Blondel, Sinnott & Svensson 1998; Mattila 2002. 4 Finally, contrary to previous research they found that electoral proportionality and timing in the electoral cycle did not have a significant impact on explaining voter turnout in European elections. See also, Mattila 2002.
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 12
is not “mainly influenced by national political attitudes”. Attitudes toward the EU, the
current level of integration and future prospects were significant in explaining turnout.
Also, citizens’ perceptions of the power to the EP and of differences between the
power of the national parliament and of the EP were not significant in explaining
turning out to vote.5
Recent criticisms of the “second-order theory” suggest that one of the central
tenets of this model (that EP elections are viewed in terms of national elections) may
not be always true. Mattila (2002) suggests that the evidence from the 1999 European
elections is that turnout at EP elections is affected by general factors affecting turnout
as such along with some specific EU factors, such as public support for the EU,
country is a net contributor/beneficiary within the EU – the latter effect being
somewhat smaller. For a summary of explanations of turnout in EP elections see table
5.
<<< Table 3, about here >>>
Consequently, there is the implication that the “less at stake” assumption in the
“second-order-theory” is not what European voters have in mind when deciding to
vote. The cost-benefit analysis seems to be more closely connected to being a net
contributor or receiver of subsidies and having more MEPs.
Electoral reforms & voter turnout Research on voter turnout has important policy implications as it attempts to provide
an explanation as to why citizens decide to vote or not vote on election day. Most
reforms relating to the administration of electoral contests aim at increasing the
absolute numbers of citizens voting, thus improving political transparency and
additionally making the voting population as representative as possible of society.
Reforms which have been implemented may be summarised under three headings;
1. Lowering the costs of registration for citizens not registered;
• Election day registration • Mail registration
5 An alternative reason for observing “apparent second-order election effects (e.g. low turnout) may be due to strategic calculations in the minds of political elites and the actions that follow from them rather than to any complex perceptions and calculations regarding the relative power being allocated in different electoral arenas that the second-order election model assumes are made by the electorate” (Blondel, Sinnott & Svensson 1998: 24).
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 13
• Motor voter registration
2. Lowering the costs of voting for registered voters;
• Early voting • Voting by mail • Internet voting
3. Mobilising registered voters to go and vote on election day;
• Canvassing door-to-door • Telephone campaigning • Direct mail campaigns • Leaflet campaigns
Lowering the costs of registration: Early analyses of the costs of registration and
voter turnout rates using aggregate data for some US cities and states found that
participation in the 1960s was higher where the costs of registration were lower
(Kelley, Ayres & Bowen 1967; Kim, Petrocik et al. 1973). Kim, Petrocik et al.’s
(1973) study was innovative in using both aggregate and individual level data where
individual level factors such as age, education and race explained over the half the
variance and systemic factors such as administrative and institutional provisions
explained almost a third of the variance.6
Later work by Rosenstone & Wolfinger (1978, 1980) found that extending the
date for registration, increasing the number of voting hours on election day could
increase national turnout by 9.1%. However, level of education seemed to have little
impact on this process (Nagler 1991). This influential line of research indicated that
lowering the costs of registration was not likely to make the voting population more
representative of society.
Lowering the costs of voting for registered voters: One of the methods most
often implemented to increase access to voting has been to allow ‘absentee balloting’,
i.e. allow voters to vote in advance of election day and submit their ballot by mail.
Research indicates that voter participation increases by 2% and that this is mainly a
result of political parties ensuring that their supporters have received a ballot (Oliver
1996).
6 Both of these studies expressed the view that greater understanding could only be obtained by gathering data and doing analysis at lower levels of aggregation.
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 14
Early voting is another method that has been used to increase turnout where
voters are allowed to vote in person in advance of election day in convenient locations
such as shopping malls. This reform was seen to increase participation by 5% on
previous elections and the effect was larger for primary rather than congressional
elections (Richardson & Neeley 1996). Other research found that early voting was not
based on partisanship, though voters who had lower incomes, were older or had strong
party attachment tended to use the voting early option. In general, early voting is seen
to be more strongly correlated with voters attitudes than their socio-demographic
background (Stein 1998).
Postal voting: Recently in Britain with the ‘Representation of People Act
(2000)’ there has been a concerted effort to undertake research to design new electoral
arrangements that increase electoral participation. Consequently, pilot studies were
undertaken during the local government elections in 2000 and 2002. One of the key
findings from these local election studies was that “only postal voting appeared to
have the potential to significantly increase local election turnout”. It was found that
with postal voting turnout in many local authority areas increased by 50 percent and
turnout more than doubled where “no declaration of identity was required”. Such
evidence suggests that electoral facilitation policies can have significant effects.
Furthermore, postal voting did not appear to have any partisan effects favouring one
political party or another. However, using postal voting only can is more costly than
conventional voting-in-person polling. This is because extra security is required to
ensure against abuses of secrecy and integrity of the ballot (Rallings & Thrasher 2000:
5-6; The Electoral Commission 2002: 2-3).
Mobilising registered voters: One of the key reasons noted for the increase in
electoral abstention since the 1960s has been the decline in face-to-face canvassing of
voters by parties. While political mobilisation in campaigns still reaches many
citizens the nature of this contact has become increasingly impersonal (see, Gerber &
Green 2000; Abramson, Aldrich & Rohde 1998).
In general, policy innovations that have attempted to increase citizen
participation in elections by reducing the costs of registration and of casting a ballot
have had only small effects. The absolute number of electors voting increases by less
than 10 percentage points while efforts to make voters more representative of all
citizens do not appear to work. However, little research has been undertaken on the
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 15
combined impact of a number of reforms and the effect of policy innovations over
long time periods. Preliminary research does however highlight two key things.
Firstly, initial model specification should deal either with voter registration or turnout.
It is important theoretically and empirically not to conflate both of these phenomena.
Secondly, electoral participation and electoral choice should be modelled together
rather than be seen as separate processes.
Policy implications of variations in turnout From a normative perspective whether or not a citizen casts a ballot at election time is
potentially important in establishing political transparency and democratic legitimacy.
Empirical research indicates that different groups in society vote at different rates and
prefer different policies. Thus, if for example those with high incomes vote at a
higher rate than those with low incomes, the policies favoured by the higher income
group should hold more influence over parties and candidates seeking election.
In such a situation declining rates of voter turnout undermine the normative
principles of transparency and legitimacy in a liberal democratic polity. This kind of
analysis has led to a proposal to make voting compulsory being put forward in an
American Political Science Association presidential address (Lijphart 1997: 4).
Voting determines who are elected and those elected determine public policy.
Marginalised groups can only have political influence if they vote. Therefore, voter
turnout is important because it is a key factor in explaining electoral outcomes and
subsequent government policy making.
If one accepts the assumption that voters and non-voters have different
characteristic what would happen if there was universal turnout? Very little
counterfactual research of this kind has been undertaken outside of the United States.
DeNardo (1980; 1986) argued that higher turnout helped the Democratic Party in
presidential, congressional, senatorial and gubernatorial elections – but only some of
the time. A second line of research based on the assumption that all citizens vote and
hence making comparisons with mass surveys found that in presidential elections after
1972 there was no significant partisan effect. Voters were seen to represent citizens
with some variation between elections (Wolfinger & Rosenstone 1980: 109ff;
Teixeira 1992: 87ff; Abramson, Aldrich & Rohde 1998: 86-89; Schneider 1985). This
variation between elections could of course be accounted for by net differences in the
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 16
profiles of marginal voters deciding to vote or abstain in a series of elections for a
variety of voluntary and circumstantial reasons.
Criticism of these two lines of counterfactual research may be summarised in
the argument that if non-voters (and marginal voters) were to decide to vote rather
than abstain, they would be motivated to express (different) preferences on issues and
candidates. The implicit assumption is that with partisan alienation, whole issue
domains that would be of concern to non-voters if they voted remain untapped
potential sources of political mobilisation. With universal turnout, issue domain
lacunae become apparent and form the basis for a change in the overall pattern of
political mobilisation.
The belief that voter turnout has important electoral and policy consequences
may be summarised on the basis of Lijphart’s (1997) arguments in favour of making
voter participation compulsory in order to reduce inequality. Lijphart’s Unequal
Participation contentions are listed below in the form of six hypotheses.7
H1. Core voters as a group tend to be more conservative than all
citizens on issues relating to morality, economic redistribution and individual freedom.
H2. Core voters, marginal voters and non-voters all express preferences on the basis of partisanship, ideology and specific issues. Marginal and non-voters tend to have preferences of lower intensity than core voters.
H3. Core voters tend to have higher incomes than marginal and non-voters.
H4. Non-voters and marginal voters have an underdeveloped sense of class-consciousness.
H5. Non-voters and marginal voters are not politically mobilised during election campaigns, i.e. canvassed by telephone or contacted by party activists.
H6. A majority of non-voters and marginal voters share similar demographic characteristics, e.g. young age, residentially mobile and low education.
All of these hypotheses have been tested in the United States by Highton & Wolfinger
(2001: 118) who found that “Non-voters do differ from voters in the directions
predicted [by Lijphart (1997)]. They are less likely to have opinions on issues,
7 If analysis is undertaken at a single time point then dividing voters into ‘marginal voters’ and ‘non-voter’ groups is impossible. Marginal voters who do not vote are classified as ‘non-voters’. Only with a dynamic analysis with at least two time points can the important groups of ‘marginal voters’ be analysed and modelled in a systematic manner.
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 17
identify with a party, express an ideological preference, be class conscious, or be
stimulated by campaign activists. Yet, except for the last point, the differences are
generally trivial”. No similar exercise has been undertaken in Europe.
Partcom-multilevel project – new data & new research design The principal reason, for adopting a multi-level strategy is that the problem of
electoral participation varies across levels of governance. (The concept of “levels” of
governance involves a horizontal and a vertical dimension – the horizontal dimension
comprises sub-national, national and supranational governance; the vertical dimension
distinguishes between presidential, legislative, parliamentary and plebiscitary poles of
governance). Previous research has tended to concentrate on the national level or, at
most, to extend one step beyond this (either in a sub-national or in a supranational
direction) (see figure 2).
<<< Figure 2, about here >>>
These piecemeal approaches have failed to exploit the veritable laboratory for the
analysis of electoral participation constituted by the full range of multi-level
governance and, in doing so, have lost major analytical leverage. This has meant, for
example, that the dominant analytical model in the field (the second-order election
model) has only been subjected to very partial testing. Finally, the principle of
subsidiarity reinforces the decision to pursue a multi-level research strategy.
Implicitly or explicitly guided by this principle, the European integration project has
for quite some time been strongly committed to the participation and involvement of
citizens at the sub-national level and has done much to nurture institutions and
processes of devolved decision-making at such levels. This aspect of European
integration underlines the need for research on democratic participation that goes
beyond the national level to include both the supranational and subnational levels.
The main objective of the multi-level project is to explain differences in voter
turnout cross-nationally, at different levels of governance, over time. Quite often
research on voter turnout has focussed on a single country at a single time point or
over time. Due mainly to data constraints, there is considerably less research looking
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 18
at many countries comparatively over time. Moreover, such cross-country and cross-
time research as exists has almost always focussed on specific types or levels of
elections, i.e. national or European. Inevitably, such research has focused on national
elections as the most important type of election for democratic representation and
there is as a result little or no comparative research that looks at different types of
elections.
In fact, the main explanation of supra-national European elections, the so-called
“second-order theory” builds on the idea that elections at the European level are in
fact determined by considerations at the (more important) national level. One of the
alleged consequences is that European elections have lower turnout.
The assumption that differential rates of voter abstention are connected to
voters’ evaluations of an election's importance places a relatively high informational
demand on voters in terms of awareness of institutional provisions, of the current
political situation and of the perceived signals that political elites will receive from the
outcome of non-national elections. However, if citizens are simply not aware of these
factors, as much of the empirical evidence indicates they are not, then the decision of
whether or not to vote is based on a mix of general factors relating to voting
behaviour in all elections and specific factors relating to particular features of sub- and
supra-national elections.
Analyses of lower turnout at mid-term congressional elections, and at European,
local, regional and municipal elections suggest that voters do not necessarily rank
different election types on the basis of perceived importance, nor do they necessarily
treat each election type separately. On the other hand different election types are
connected to each other in the minds of voters. The published research on turnout has
never fully explored or explained this relationship. Analysing citizens' voting
behaviour across a wide span of countries, time-points and types of election will make
it possible to get a more realistic picture of the decisions that citizens face on election
day, starting with the decision on whether or not to go to the polls. In pursuing this
objective the Partcom-multilevel projects employs both aggregate and individual level
data. In practical terms, aggregate voting data is matched with census data at the
lowest sub-national political unit feasible in seven countries. With such data it is
possible to use not only regression techniques but also undertake ecological inference
estimations. Individual level data will be based on a number of mass surveys
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 19
undertaken for various types of elections where a variety of regression techniques will
be used to test a variety of hypotheses taken from the literature reviewed in this report.
Select Bibliography of Voter Turnout Literature General Abramson P.R. & Aldrich J.H.: The decline of electoral participation in the United States. American
Political Science Review, 1982, Vol. 73, No. 3, 502-521. Abramson P.R., Aldrich J.H. & Rohde D.W.: Change and continuity in the 1996 elections.
Washington, DC: Congressional Quarterly, 1998. Abramson P.R. & Claggett W.: The quality of record keeping and racial differences in validated
turnout. Journal of Politics, August 1992, Vol. 54, 871-880. Aldrich J.H.: Rational Choice and Turnout. American Journal of Political Science, 1993, Vol. 37(1),
246-278. Alvarez M.R. & Nagler J.: The likely consequences of internet voting for political representation.
Paper presented at the Internet Voting and Democracy Symposium, Loyola Law School, Los Angeles, California, October 26, 2000.
Belli R.F. & Traugott M.W.: What leads to voter overreports? Constrast of overreporters to voters and admitted nonvoters. Ann Arbor: 2000, Unpublished paper.
Bendor J. Diermeier D. & Ting M.: A behavioural model of turnout. Unpublished paper. September 2000, Kellogg Graduate School of Management, Northwestern University.
Blais A. & Carty R.K.: Does proportional representation foster voter turnout? European Journal of Political Research, 1990, Vol. 18, 167-181.
Blais A. & Dobrzynska A.: Turnout in electoral democracies. European Journal of Political Research, 1998, Vol. 33, 239-261.
Blais A.: To Vote or Not to Vote: The Merits and Limits of Rational Choice Theory. Pittsburgh, PA.: University of Pittsburg Press, 2000.
Brians, C.L. & Grofman. B.: When registration barriers fall, who votes? An empirical test of a rational choice model. Public Choice, 1999, Vol.99, 161-176.
Brians C. & Grofman B.: The longitudinal effects of election day registration and closing date on voter turnout. Social Science Quarterly, 2000 forthcoming.
Burden B.C.: Voter turnout and the National Election Studies. Political Analysis, Vol. 8, No. 4, 2000, 389-398.
Caldeira G.A., Clausen A.R. & Patterson S.C.: Partisan mobilization and electoral participation, Electoral Studies, 1990, Vol. 3, 191-204.
Cassel C.A. & Luskin R.C.: Simple explanations of turnout decline. American Political Science Review, 1988, Vol. 84, No. 2, 1321-1330.
Cox G.W.: Electoral Rules and the Calculus of Mobilization. Legislative Studies Quarterly, August 1999, Vol. 24: No. 3, 387-419.
Crepaz M.M.L.: The impact of party polarization and postmaterialism on voter turnout, a comparative study of 16 industrial democracies. European Journal of Political Research, 1990, Vol. 18, 183-205.
Diermeier D. & Mieghem van, J.A.: Coordination in turnout games. Unpublished paper. November 17, 2000, Kellogg Graduate School of Management, Northwestern University.
Downs A.: An Economic Theory of Democracy. New York: Harper & Row, 1957. Electoral Commission: Modernising elections: A strategic evaluation of the 2002 electoral pilot
schemes. London: The Electoral Commission, August 2002. Elster J.: A plea for mechanisms. In; Hedstrøm & Swedberg R. (eds.): Social Mechanisms: An
analytical approach to social theory. Cambridge: Cambridge University Press, 1998, pp 45-73. Enelow J. & Hinich M. (eds.): Advances in Spatial Theory of Voting. Cambridge: Cambridge
University Press, 1990. Foster C.B.: The Performance of Rational Voter Models in Recent Presidential Elections. American
Political Science Review, 1984, Vol. 78, 678-690. Franklin M.N.: Notes and Comments: Electoral Engineering and Cross-National Turnout Differences:
What Role for Compulsory Voting? British Journal of Political Science, 1999, Vol. 29, 205-224.
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 20
Gerber A.S. & Green D.P.: Does canvassing increase voter turnout? A field experiment. Proceedings of the National Academy of Science, USA, Vol. 96, 10939 – 10942, September 1999.
Gerber A.S. & Green D.P.: The effect of a nonpartisan get-out-the-vote drive: An experimental study of leafleting. The Journal of Politics, Vol. 62, No.3, August 2000a, 846-857.
Gerber A.S. & Green D.P.: The effects of canvassing, telephone calls, and direct mail on voter turnout: A field experiment. American Political Science Review, Vol. 94, No.3, September 2000b, 653-663.
Glaser W.A.: Fluctuations in turnout. In; McPhee W.N. & Glaser W.A.: Public Opinion & Congressional Elections. Westport, Conn.: Greenwood Press, Publishers, 1962, 19-51.
Gordon S.B. & Segura G.M.: Cross-National Variation in the Political Sophistication of Individuals: Capacity or Choice? The Journal of Politics, 1997, Vol. 59(1), 126-147.
Gosnell H.F.: Getting out the vote: an experiment in the simulation of voting. Chicago: University of Chicago Press, 1927.
Gray M. & Caul M.: Declining voter turnout in advanced industrial democracies, 1950 to 1997: the effects of declining group mobilization. Comparative Political Studies, Nov. 2000, Vol.33, No.9, 1091-1122.
Green D. & Shapiro I.: Pathologies of Rational Choice Theory. New Haven: Yale University Press, 1994.
Green D.P. & Shacar R.: Habit formation and political behaviour: Evidence of consuetude in voter turnout. British Journal of Political Science, 2001, Vol. 30, 561-573.
Grofman B.: Is Turnout the Paradox that Ate Rational Choice Theory? In, Grofman B. (ed.): Information, Participation, and Choice: An Economic Theory of Democracy in Perspective. Ann Arbor, MI: The University of Michigan Press, 1993, pp 93-103.
Grofman B., Collet C. & Griffin R.: Analyzing the turnout-competition link with aggregate cross-sectional data. Public Choice, 1998, Vol. 95, 233-246.
Grofman, B., Owen G. & Christian C.: Rethinking the Partisan Effects of Higher Turnout: So What’s the Question? Public Choice, 1999, Vol. 99: 357-376.
Highton B. Wolfinger R.E.: The political implications of higher turnout. British Journal of Political Science, 2001, Vol. 31, 179-223.
Horiuchi Y.: The “Turnout Twist” in Japanese Elections. Paper delivered at the annual meeting of the Public Choice Society, San Antonio, Texas, March 9-11, 2001.
Knack S.: Social altruism and voter turnout: Evidence from the 1991 NES pilot study. George Mason University, Center for Study of Public Choice, January 1992.
Knack S.: Does ‘Motor Voter” work? Evidence from State-level data. Journal of Politics, 1990, Vol. 57, 796-811.
Lane J.E. & Ersson S.: Macro and micro understanding in political science: What explains political participation? European Journal of Political Research, 1990, Vol. 18, 457-465.
Leighley J. & Nagler J. Individual and Systemic Influences on Turnout: Who Votes? 1984. The Journal of Politics, 1992, Vol. 54, No. 3 (August): 718-739.
Leighley J. & Nagler J.: Individual and Systemic Influences on Turnout: Who Votes? Journal of Politics, August 1992, Vol. 54, 718-740.
Leighley, Jan. 1995. Attitudes, Opportunities and Incentives: A Field Essay on Political Participation. Political Research Quarterly 48: 181-192.
Lijphart A.: Unequal Participation, Democracy's Unresolved Dilemma. American Political Science Review, 1997, Vol. 91(1), 1-14.
Marsh M.: ‘Surge and decline’ in European parliament elections: A new challenge for a classic theory of electoral change. Paper presented at the American Political Science Association Meeting, Washington DC, 2000.
Manski, C.: Identification Problems in the Social Sciences. Cambridge & London: Harvard University Press, 1995.
Matsusaka J.G.: Explaining voter patterns: an information theory. Public Choice, 1995, Vol. 84, 91-117.
McDonald M.P. & Popkin S.L.: The myth of the vanishing voter. Paper presented at the annual meeting of the American Political Science Association meeting, Washington, August 30 – September 3, 2000.
Norris P.: Stays home? Political mobilisation. In; Norris P. (ed.): A virtuous circle: Political Communications in post-industrial societies. New York: Cambridge University Press, 2000, chapter 12.
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 21
Overbye E.: Making a case for the rational, self regarding ‘ethical’ voter … and solving the “Paradox of not voting” in the process. Public Choice, January 1995, Vol. 27, 369-396.
Popkin S.L. The Reasoning Voter: Communication and Persuasion in Presidential Campaigns. Chicago: University of Chicago Press, 1991.
Presser S. & Traugott M.W.: Little white lies and social science models: correlated response errors in a panel study of voting. Public Opinion Quarterly, 1995, Vol. 23, No. 4, 409-426.
Rallings C. & Thrasher M.: Elections – the 21st model: An evaluation of May 2000 local electoral records. London: Local Government Association, 2000.
Ranney A.: Nonvoting is Not a Social Disease. Public Opinion, 1983, Vol. 6 (Oct./Nov.): 16-19. Riker W. & Ordeshook P.: A Theory of the Calculus of Voting. American Political Science Review,
1968, Vol. 62, 25-42. Rose, Richard. 1983. Elections and Electoral Systems: Choices and Alternatives. In; Bogdanor V. &
Butler D. (eds.): Democracy and Elections: Electoral Systems and Their Political Consequences. Cambridge: Cambridge University Press, 1983.
Sanders M.: Information, registration, indifference and turnout. Paper prepared for Southern Political Science Association, Atlanta, 1996, November 6-9, 1996.
Sanders M.: Uncertainty and turnout. Political Analysis, Vol. 9, No. 1, January 2001, 45-57. Schlozman K.H., Verba S. & Brady H.E.: Participation’s not a paradox: the view of American activists.
British Journal of Political Science, 1995, Vol. 25, 1-36. Sniderman P.M., Brody R.A. & Tetlock P.E.: Reasoning and Choice: Explorations in Political
Psychology. Cambridge: Cambridge University Press, 1991. Straits, B.: The social context of voter turnout, Public Opinion Quarterly, 1990, Vol. 54, 64-73. Swaddle K. & Heath A.: ‘Official and reported turnout in the British general elections of 1987’, British
Journal of Political Science, 19, 1989, 537-551. Teixeira, R.A.: The Disappearing American Voter. Washington, DC: Brookings, 1992. Tam Cho W.K.: Naturalization, socialization, participation: Immigrants and (non-) voting. Journal of
Politics, November 1999, Vol. 64, No. 4, 1140-1155. Traugott M.W.: Why electoral reform has failed: if you build it, will they come? Paper presented at
the Political Participation: Building a Research Agenda Conference, Princeton University, October 12-14, 2000.
Uhlaner, C.: Electoral Participation: Summing Up a Decade. Society, 1991, (July/ August): 35-40. Verba S., Schlozman K.L. & Nie N.H.: Citizen activity: who participated and what do they say?
American Political Science Review, June 1993, Vol. 87, No. 2, 303-318. Verba S., Schlozman K.L & Brady H.E. (2000) ‘Rational Action and Political Activity’. Journal of
Theoretical Politics, 2000, Vol. 12: 3, 243-268. Wolfinger, R. E. 1991. Voter Turnout. Society (July/August): 23-26. Wolfinger R.E.: The rational citizen faces election day or what rational choice theorists don't tell you
about American elections. In; Kent Jennings M. & Mann T.E.: Elections at Home and Abroad. Essays in Honour of Warren E. Miller. Ann Arbor, University of Michigan Press, 1994, pp 71-89.
Zimmer, T.: Political competition, voter turnout and party vote, The Social Science Journal, 22 (3), 1985, p. 103-109.
Cross-National Research on Voter Turnout Aguilar E.E. & Pacek A.C.: Macroeconomic conditions, voter turnout, and the working-
class/economically disadvantaged party vote in developing countries. Comparative Political Studies, October 2000, Vol. 33, No. 8, 995-1017.
Barnes S. & Kaase M. et al.: Political Action: Mass Participation in Five Western Democracies. Beverly Hills, CA: Sage, 1979.
Budge I. & Fairlie D.: Explaining and Predicting Elections: Issue Effects and Party Strategies in Twenty-Three Democracies. New York: Allen & Unwin, 1983.
Crepaz M.L.: Constitutional structures and regime performance in 18 industrialized democracies: A test of Olson's hypothesis. European Journal of Political Research, 1996, Vol. 29, 87-104.
Crewe I.: Electoral Participation. Butler D. (ed.): Democracy at the Polls: A Comparative Study of Competitive National Elections. Washington: American Enterprise Institute, 1981.
Crewe, Ivor. 1981. Electoral Participation. In Democracy at the Polls: A Comparative Study of Competitive National Elections, eds. David Butler, Howard Penniman, and Austin Ranney. Washington: American Enterprise Institute for Public Policy Research.
Dittrich, K. & Johansen, L.N.: Voting turnout in Europe, 1945-1978: In; Daalder H. & Mair P. (ed): Myths and realities, in Western European Party Systems. Sage, 1983, p. 95-114.
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 22
Erikson R.S.: Pooling and Statistical Control, a rejoinder to Radcliff. American Politics Quarterly, 1995, Vol. 23(4), 404-408.
Franklin M., Mackie T. & Valen H. et al.: Electoral Change, Responses to Evolving Social and Attitudinal Structures in Western Democracies. Cambridge: Cambridge University Press, 1992.
Franklin M.: Electoral Participation. In; LeDuc L, Niemi R.G. & Norris P. (eds): Comparing Democracies, Elections and Voting in Global Perspective. London: Sage, 1996, pp 216-235.
Glass, David, Squire P. & Wolfinger R.: Voter Turnout: An International Comparison. Public Opinion 1984, 6 (Dec./Jan.): 49-57.
Gray M. & Caul M.: Declining voter turnout in advanced industrial democracies, 1950 to 1997: the effects of declining group mobilization. Paper presented at the annual meeting of the American Political Science Association meeting, Washington, August 30 – September 3, 2000.
Hempel C.: Aspects of Scientific Explanation. New York: Free Press, 1965. Hirczy W.: Explaining near-universal turnout: The case of Malta. European Journal of Political
Research, 1995, Vol. 27, 255-272. Høst V. & Paldam M.: An international element in the vote? A comparative study of 17 OECD
countries 1948-85. European Journal of Political Research, 1990, Vol. 18, 221-239. Inglehart R.: Culture Shift in Advanced Society Industrial Society. Princeton, NJ: Princeton University
Press, 1990. International Institute for Democracy & Electoral Assistance (IDEA): Voter Turnout from 1945 to
1997: A Global Report on Political Participation, Second Edition. Stockholm, Sweden: IDEA Publications, 1998. http://www.int-idea.se
Jackman R.W. & Miller R.A.: Turnout and the Economy Reply. Comparative Political Studies, 1996, Vol. 29(6), 724-733.
Jackman R.W. & Miller R.A.: Voter turnout in the industrial democracies during the 1980's. Comparative Political Studies, 1995, Vol. 27(4), 467-492.
Jackman R.W.: Political Institutions and Voter Turnout in the Industrial Democracies. American Political Science Review, 1987, Vol. 81, 405-423.
Jackman R.W.: Rationality and Political Participation. American Journal of Political Science, 1993, Vol. 37(1), 279-290.
Jennings M. Kent & van Deth J. et al.: Continuities in Political Action. Berlin: de Gruyter, 1990. Marsh M., Franklin M. & Lyons P.: The Tally of Turnout: Understanding Cross-National Turnout
Decline Since 1945 Paper presented at the annual meeting of the American Political Science Association meeting, Washington, August 30 – September 3, 2000.
Miller W.: The Puzzled Transformed: Explaining Turnout Declines. Political Behaviour, 1992, Vol. 14, 1-43.
Narud H.M. & Valen H.: Decline of electoral turnout: The case of Norway. European Journal of Political Research, 1996, Vol. 29, 235-256.
Nevitte N. Blais A., Gidengil E. & Nadeau R.: Socio-economic status and non-voting: A cross-national comparative analysis. Paper presented at the XVIIIth World Congress of the International Political Science Association, Quebec, August 1-5, 2000.
Norris P., Curtice J., Sanders D., Scammell M. & Semetko H.: On Message: Communicating the Campaign. London: Sage. 1999.
Powell G. Bingham: American Turnout in Comparative Perspective. American Political Science Review, March 1986, Vol. 80(1), 17-43.
Powell G. Bingham: Voting Turnout in Thirty Democracies. In, Rose R. (ed.): Electoral Participation. Beverly Hills, Calif.: Sage, 1980.
Powell G. Bingham: Contemporary Democracies: Participation, Stability and Violence. Cambridge MA: Harvard University Press, 1982.
Powell, B.G. American Voter Turnout in Comparative Perspective. American Political Science Review 1986, Vol 80, No. 1 (March): 17-39.
Putnam R.D.: Making Democracy Work, Civic Traditions in Modern Italy. Princeton, NJ: Princeton University Press, 1993.
Radcliff B.: The welfare state, turnout, and the economy: A comparative analysis. American Political Science Review, 1992, Vol. 86(2), 444-454.
Radcliff B.: Turnout and the Economy: A comment on Jackman and Miller's Voter turnout in the industrial democracies in the 1980’s. Comparative Political Studies, 1996, Vol. 29(6), 718-723.
Radcliff B.: Turnout and the vote revisited, a reply to Erikson. American Politics Quarterly, 1995, Vol. 23(4), 397-403.
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 23
Rhine S.L.: Registration Reform and Turnout Change in the American States. American Politics Quarterly, 1995, Vol. 23(4), 409-426.
Rosenstone S. & Hansen J.: Mobilization, Participation and American Democracy. New York: Macmillan, 1993.
Schlozman K.H., Verba S. & Brady H.E.: Participation’s not a paradox: the view for American activists. British Journal of Political Science, 1995, Vol. 25, 1-36.
Semetko H., Blumler J.G., Gurevitch M. & Weaver D.H.: The Formation of Campaign Agendas: A Comparative Analysis of Party and Media Roles in Recent American and British Elections. Hillsdale, NJ: Lawrence Erlbaum, 1991.
Teixeira R.: The Disappearing American Voter. Washington, DC: Brookings Institution, 1992. Topf R.: Electoral Participation. In; Klingemann H.D. & Fuchs D. (eds): Citizens and the State.
Oxford: Oxford University Press, 199, pp 27-51. Uhlaner C.J.: Rational Turnout: the Neglected Role of Groups. American Journal of Political Science,
1989, 33(2), 390-422. Verba S. & Nie N.: Participation and Political Equality: A Seven Nation Comparison. Cambridge:
Cambridge University Press, 1978. Verba S., Schlozman K.L. & Nie N.H.: Citizen Activity: Who participates and what do they say?
American Political Science Review, June 1993, 87(2), 303-318. Wattenberg M.P.: The Decline of Party Mobilization. Paper presented at the 1998 Annual meeting of
the Japanese Political Science Association, Kyoto, Japan. Wolfinger R.E & Rosenstone S.J.: Who Votes? New Haven: Yale University Press, 1980. Wolfinger R.E., Glass D., & Squire P.: Predictors of Electoral Turnout: An International Comparison.
Policy Studies Review 1990, 9:3, 551-574. European Parliament Elections Blondel J., Svensson P. & Sinnott R.: People and Parliament in the European Union: Democracy,
Participation and Legitimacy. Oxford: Oxford University Press, 1998. Blumler, J.G. & Fox A. D.: The European Voter: Popular Responses to the First European Community
Elections. (London: Policy Studies Institute), 1982. Blumler, J.G.: “European Voters' Response to the First Community Elections, ” In; Reif, K. (ed.):
European Elections 1979/81 and 1984: Conclusion and Perspectives from Empirical Research. Berlin: Quorum, 1984.
Blumler, J.G.: Communicating to Voters: Television in the First European Parliament Elections. London: Sage, 1983.
Cayrol, R. (1983) “Media Use and Campaign Evaluations: Social and Political Stratification of the European Electorate”. In; Blumler, J.G. (ed.): Communicating to Voters. London: Sage, 1983.
Cayrol, R.: European Elections and the Pre-Electoral Period: Media Use and Campaign Evaluations. European Journal of Political Research (19:1) 1991, pp. 17-30.
Eijk, Cees van der and Hermann Schmitt (1991) 'The Role of the Eurobarometer in the Study of European Elections and the Development of Comparative Electoral Research'. In Karlheinz Reif and Ronald Inglehart (eds.), Eurobarometer: The Dynamics of European Opinion, London: Macmillan, pp. 257-274
Eijk, Cees van der and Mark N. Franklin, Community Politics and Electoral Representation: Evidence from the 1989 European Election Study. European Journal of Political Research, 1991 Vol. 19:1, 105-128.
Eijk, Cees van der, Erik Oppenhuis, Hermann Schmitt et al. The European Elections Study 1989: Design, Implementation and Results (Computer file and code-book). Amsterdam: Steinmetz Archives, 1993.
Eijk, Cees van der, Manfred Kuechler and Hermann Schmitt 'The European Voters Study 1989'. IASSIST Quarterly, 1990, Vol. 14:2, 9-13.
Eijk, Cees van der, Mark N. Franklin and Michael Marsh (1996) 'What Voters Teach Us About Europe-Wide Elections; What Europe-Wide Elections Teach Us About Voters'. Electoral Studies 1996, Vol. 15:2, 149-166.
Eijk, Cees van der, Mark N. Franklin and Michael Marsh (1996) 'What Voters Teach Us About Europe-Wide Elections; What Europe-Wide Elections Teach Us About Voters'. Electoral Studies (15:2), pp. 149-166. Reprinted in Pippa Norris (ed.) Elections and Voting Behavior. Aldershott: Dartmouth, 1988.
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 24
Eijk, Cees van der, Mark N. Franklin, et al. (1996) Choosing Europe? The European Electorate and National Politics in the Face of Union. Ann Arbor MI: University of Michigan Press.
Eijk, Cees van der, Mark N. Franklin, et al.: Choosing Europe? The European Electorate and National Politics in the Face of Union. Ann Arbor: University of Michigan Press, 1996.
Eijk, van der C. & Oppenhuis E.: European Parties Performance in Electoral Competition. European Journal of Political Research, 1991, Vol. 19:1 55-80.
Eijk, van der C. & Oppenhuis E.: Turnout and Second Order Effects in the European Elections of June 1989 Evidence from the Netherlands. Acta Politica, 1990/1, pp.76-94.
Eijk, van der C.: 'Ongerustheid over kiezersopkomst Terecht of niet?' Namens, Vol. 5, October 1990, 18-22.
European Communities: Voter turnout at regional and local elections in the European Union, 1990-1999. Luxembourg: Office for Official Publications of the European Communities, 1999. - (CdR ; I-3/99 rev 2 ; 250.. – 9282879151
Franklin M.N.: (2001) How structural factors cause turnout variations at European Parliament elections. European Union Politics, Vol. 2 No. 3, 309-328.
Franklin, M.N. (1991) 'Getting Out the Vote: Social Structure and the Mobilization of Partisanship in the 1989 European Elections'. European Journal of Political Research, 1991, 19:1, pp. 129-148.
Franklin, M.N. (1996) Electoral Participation, in Laurence Leduc, Richard Niemi and Pippa Norris (eds.) Elections and Voting in Global Perspective. Thousand Oaks CA: Sage.
Franklin, M.N. and Wolfgang Rüdig (1992) 'The Green Voter in the 1989 European Elections', Environmental Politics (1) 129-159. Reprinted in David Judge, ed. (1993). A Green Dimension for the European Community. London: Frank Cass.
Franklin, M.N. and Wolfgang Rüdig 'On the Durability of Green Politics: Evidence from the 1989 European Election Study'. Comparative Political Studies, 1995, Vol. 28, 409-439.
Franklin, M.N.: “European Elections and the European Voter”. In; Richardson J. (ed.) European Union: Power and Policy-Making. London: Routledge, 1996.
Franklin, Mark N., Cees van der Eijk and Erik Oppenhuis (1995) Turnout Decline and the Motivational Basis of British Participation in the European Elections of 1994, in Colin Rallings, David Farrell, David Broughton and David Denver (eds), British Elections and Parties Yearbook 1995. Ann Arbor: University of Michigan Press.
Grunberg G., Perrineau P. & Ysmal C. (eds.): Le vote des quinze, les elections européennes du 13 juin 1999. Paris, Presses de la Fondation Nationale des Sciences Politiques, 2000.
Hermann Schmitt, Franz U. Pappi, Cees van der Eijk, Evi Scholz et al. (1997) European Elections Study 1994: Design, Implementation and Results. (Computer file and codebook). Koeln: Zentralarchiv fuer Empirische Sozialforschung.
Inglehart, R. & Rabier J.R.: Europe Elects a Parliament. Cognitive Mobilization and Pro-European Attitudes as Influences on Voter Turnout. Government and Opposition, 1979, Vol. 14: 479-507.
Katz, Richard (1997) Role perceptions of MEPs and styles of representation. European Journal of Political Research (24:4).
Keating, P. & Marsh M.: “The European Parliament election”. In; Gallagher M. & Sinnott R.: (eds.), How Ireland voted 1989. Galway: Centre for the Study of Irish Elections, 1990, pp. 131-147.
Kuechler, M.: Issues and Voting in the European Elections 1989. European Journal of Political Research, 1991, (19:1), 81-104.
Marsh, Michael (1998) Testing the Second-Order Election Model after Four European Elections, British Journal of Political Science (28) 591-607.
Marsh, Michael and Bernhard Wessels (1997) Territorial Representation. European Journal of Political Research (24:4).
Marsh, Michael and Pippa Norris, eds. 1997 Political Representation in the European Parliament. Special Issue of the European Journal of Political Research, Vol. 24:4
Mattila M.: Why bother? Determinants of turnout in the European elections. Electoral Studies, forthcoming, 2002.
Norris, Pippa and Mark N. Franklin (1997) Social Representation. European Journal of Political Research (24:4)
Oppenhuis, E.: Voting Behaviour in Europe : a comparative analysis of electoral participation and party choice. Amsterdam : Het Spinhuis, 1995.
Perrineau, Pascal and Colette Ysmal (1995) Le Vote des Douze: Les elections europeenes de juin 1994. Paris: Presses de la Fondation Nationale des Sciences Politiques.
Reif, K. (ed): European Elections 1979/81 and 1984. Conclusion and Perspectives from Empirical Research. Berlin: Quorum, 1984.
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 25
Reif, K. (ed.): Ten European Elections: Campaigns and Results of the 1979/81 First Direct Elections to the European Parliament. Aldershot: Gower, 1985b.
Reif, K. “Ten Second-Order National Elections.” In; Reif, K. (ed.): Ten European Elections. Aldershot, 1985a.
Reif, K., & Schmitt H.: 1980. Nine Second-order National Elections: A Conceptual Framework for the Analysis of European Election Results. European Journal of Political Research, 1980, Vol. 8: 3-4.
Reif, K., ed. (1984b) European Elections 1979/81 and 1984. Conclusion and Perspectives from Empirical Research. Berlin: Quorum.
Reif, K.: “Democratie communautaire: Le jour o l'on pourra renverser un gouvernement Europeen ...” In; Whyte F. (ed.): L’Europe visage humain. Paris: L'Harmattan, 1994.
Reif, K.: National Electoral Cycles and European Elections 1979 and 1984. Electoral Studies. 1984a, Vol. 3, 244-255.
Reif, K.: Survey Research on European Elections. ESOMAR News No. 8, September 1994. Rüdig, W. & Franklin M.N.: “Green Prospects: The Future of the Greens in Britain, France and
Germany”. In Wolfgang R. (ed.): Green Politics. Edinburgh: Edinburgh University Press. Schmitt, H. & Renato Mannheimer, R.: About Voting and Non-Voting in the European Elections of
June 1989. European Journal of Political Research (19:1), 1991, 31-54. Schmitt, H. & Scheuer A.: ‘Region - Nation - Europa. Drei Ebenen politischer Steuerung und
Gestaltung in der Wahrnehmung der Brger’. In Thomas Knig, Elmar Rieger and Hermann Schmitt (eds.), Das Europische Mehrebenensystem (Mannheimer Jahrbuch fr Europische Sozialforschung Bd. 1), Frankfurt: Campus, 1996.
Schmitt, H. & Thomassen J. et al.: Political Representation and Legitimacy in the European Union. Oxford: Oxford University Press, 1999.
Schmitt, H.: “Parteibindungen und Issuekompetenz der Parteien als Determinanten der Wahlentscheidung”. In; Niedermayer O. & Hermann Schmitt H. (eds.): Wahlen und Europische Einigung, Opladen: Westdeutscher Verlag, 1994.
Schmitt, H.: Party Attachment and Party Choice in the European Elections of June 1989. A cross-national comparative analysis of the post-electoral surveys of the European Voters Study 1989. International Journal of Public Opinion Research (2: 2) 1990, 169-184.
Schmitt, H.: Was war europaeisch am Europawahlverhalten der Deutschen? Eine Analyse der Europawahl 1989 in der Bundesrepublik Deutschland. Aus Politik und Zeitgeschichte (B43/89), 1989, 39-51.
Schmitt, Hermann (1995) 'National Party Systems and the Policies of the European Union'. First Results from the 1994 European Elections Study. Il Politico 90:1 (1995), pp. 93-107.
Schultz, W.: “One Campaign or Nine”. In Blumler J.G. (ed.): Communicating to Voters: Television in the First European Parliament Elections. London: Sage, 1983.
Wessels, Bernhard and Richard Katz, et al. (1999) European Parliament and European Integration. Oxford: Oxford University Press.
Democratic participation in systems of multi-level governance
PARTCOM26
Figure 1, ‘Concentric circle’ typology of electoral participation
Note, derived from Sigelman & Jewell (1986). ‘Regular voters’ always infrequently; ‘Non-voters’ refuse to vote although registered; ‘Unregistervote; ‘Non-citizens’ are most often not entitled to vote.
Irregular voters
s Unregistered-
voters
Regular voters
Non-voter
Non-citizens
– MULTI-LEVEL HPSE-CT-1999-00029
vote; ‘Irregular voters’ vote ed voters’ are not allowed to
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 27
Figure 2, Typology of areas electoral participation in systems of multi-level governance
Level of governance
Pole of governance – where decision making power is located
Indirectly elected executive answerable to a parliamentary assembly
Directly elected executive
Legislative assembly
Electorate as a whole
Sub-national [1] Municipal, local, state, regional governments
[2] City mayors & state governor elections
[3] Municipal, local, state, regional elections
[4] Local and state referendums
National [5] Parliamentary elections
[6] Elections of heads of government, e.g. President
[7] Legislatures such as US congress & upper chamber assemblies
[8] National referendums
Supra-national [9] European Parliament (EP) elections
[10] Series of EU related based referendums
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 28
Table 1, Factors used to explain voter turnout as a proportion of the voting age population at the aggregate level in comparative perspective over time
Literature source (see notes below)
Independent variable A B C D E F G H I J K L
Mobilisation Frequency of changes in executive x Party-group linkages X Polarisation (1) X Postmaterialist party X Vote-seat difference of
two largest parties x X X
Change in labour Party vote share X Change in Union density X N parties in legislature (4) X X X x (log)X X Decisiveness (3) X First party-second party x One party majority government X Life satisfaction x Levels of trust x Political discussion rates x New voters (11) X Facilitation Compulsory voting X X X (X) X X x Disproportionality [10] X X Disproportionlity * PR X Electoral system (dummies) (2) X Voting age x x X Suffrage change x National competition, districts (5) X X X (X) Uncameralism (6) X X X x Postal voting X Sunday voting X Number polling days X Voluntary voter registration [X] x Resources Life expectancy x Illiteracy X Pop density X GNP % (log)X Income % [change] X Inflation x Social security spending as %GDP x Social security*income X Spending on education x GDP-change on previous year X x x x GDP change on previous election x Socio-demographic Age [% over 34] [X] Change in VAP X Change in VAP (30-69yrs) X
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 29
Sources:
A. Powell (1986) B. Jackman (1987) C. Radcliff (1992) D. Blais & Carty (1990) E. Crepaz (1990) F. Jackman & Miller (1995) G. Radcliff (1996) H. Jackman & Miller (1996) I. Franklin et al. (1996) J. Blais (1998) K. Cox (1999) L. Gray & Caul (2000)
Legend: X – statistically significant factor; x – not a statistically significant factor Notes: 1. Castles & Mair (EJPR: 1994) expert survey data: distance between the extreme parties. 2. PR, plurality, majority, mixed systems. 3. Is election highest, and if not, was election coterminous with highest. 4. Most to the literature uses the effective number of parties. However, Blais (1998) found after considerable analysis that this
formulation works best. 5. Really ordered and grouped, constant size 6. Lijphart (1984: 212-213): allows for weak bicameralism and strong bicameralism. 7. Also show results from fixed effects and simple pooled model, following suggestion of Radcliff (1992). Include lagged
turnout in his model. 8. This analysis and the article it is a rejoinder to the use of lagged turnout as an independent variable, which can attenuate the
effects of the other explanatory factors (see, Achen 2000). Jackman concludes time series data raises real questions about the importance of institutions.
9. This deals with all elections in Mackie & Rose- but only investigated only institutional variables. Results superseded by Blais (1998).
10. Found disproportionality only significant post 1945, not before. 11. Increase in enfranchisement
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 30
Table 2, Overview of factors affecting voter turnout in national or parliamentary elections
Factor affecting electoral participation Increase turnout
Decrease turnout
Voted in previous election (Grofman 1993; Green & Shachar 2000) X Civic & political involvement (Putnam 2000) X Compulsory voting (Lijphart 1997; Franklin 1999) X Low barriers to voter registration (Kim, Petrocik et al. 1973; Wolfinger & Rosenstone
1978, 1980) X
Motor Voter [NVRA] registration (Knack 1995, 1999; Highton & Wolfinger 1998; Hanmer 2000)
X
Election day registration (Rhine 1995) X Absentee balloting allowed (Oliver 1996) X Early voting (Stein 1998) X Postal voting (Magleby 1987; Berinsky, Burns & Traugott 2000) X Internet voting (Solop 2000; Alvarez & Nagler 2000; Norris 2000) X Holding elections at the weekend / holidays (Franklin 1996) X Face-to-canvassing (Gerber & Green 1999, 2000b) X Direct mail canvassing (Gerber & Green 2000b) X Leaflet campaign (Gerber & Green 2000a) X Proportional electoral system (Blais & Carty 1990) X Linkages between social groups & political parties in addition to party competition
(excluding compulsory voting) (Powell 1980, 1987) X
Political competition or high intensity contests resulting from social or ideological cleavages between parties increase turnout (Powell 1980)
X
Increased volume of political mobilisation (Rosenstone & Hansen 1993) X Party polarisation and an increasingly postmaterialist electorate (Markus & Crepaz
1990) X
Expected majority of the front running candidate is small (Downs 1957; Ferejohn & Fiorina 1975; Grofman 1983)
X
Citizen participation has increased in more policy oriented and less organised arenas such as referenda (Butler & Ranney 1981).
X
The activity of citizen groups, single-issue groups, etc. has increased leading to increased voter participation (Barnes & Kaase 1979; Jennings & van Deth 1990; Ginsberg & Shefter 1990).
X
Impact of campaign contact (Rosenstone & Hansen 1993) X Increasing rate of voter overreporting in NES (Burden 2000) X Political mobilisation (Gosnell 1927; Caldeira, Patterson et al. 1985) X X Structural incentives, e.g. number of party choices and type of legislative power
influence turnout (Jackman 1987). X X
Voter turnout variation is cyclical (Nardulli, Dlager et al. 1996) X X Contextual factors for specific elections & variations in personal circumstances
(Sigelman & Jewell 1986; Berinsky, Burns & Traugott 2000) X X
Nature of prevailing economic conditions (Southwell 1996) X X Increased level of information to voters (Sanders 1996, 2001; Diermeier & van
Mieghem 2000) X X
Bad weather on election day (Laver 1997) X The changing nature of the relationship between the individual and the mass media
which has altered the nature of political campaigns and political participation (Semetko et al. 1991; Miller 1990).
X
Frequent voting in municipal, local, regional, state elections & referenda (Powell 1982; see also,Crewe 1981).
X
The increasing heterogeneity of public issues – leading it has been suggested in some cases to distinct issue publics (note, the influence here of schema theory within public opinion research (Budge & Fairlie 1983; Franklin 1992).
X
Increasingly inaccurate measure of voting age population (McDonald & Popkin 2000) X Decline in partisanship and perceptions of government responsiveness (Abramson &
Aldrich 1982) X
Generational replacement (Miller & Shanks 1996) X
Democratic participation in systems of multi-level governance
PARTCOM – MULTI-LEVEL HPSE-CT-1999-00029 31
Table 3, Factors used to explain voter turnout in European Parliament elections
Effect on
turnout
Oppenhuis (1995)
Van der Eijk et al. (1996)
Blondel et al. (1998)
Mattila (2002)
Type of analysis Aggregate analysis X X X X Individual level analysis
X X X
Institutional factors Compulsory voting + X X* X X Simultaneous elections + X X X Weekend voting + X X X X Country net contributor - X Country net beneficiary + X Strict party lists - X First elections in EU + X Share of seats in EP + X Multiple constituencies + X Proportionality + X X Election timing +/- X Demographics Age < 25 - X X X Age 26 - 35 - X X X Religion + X X Education + X General attitudes Party attachment + X X X Interest in politics + X X X Informed on national politics
+ X
Left-Right + X Importance of election + X† X Attitudes toward EU Public support for EU + X Pro-EU + X EU involvement + X EU knowledge + X EP reliability + X Euro party difference + X Euro candidate difference
+ X
Notes: * Used as a interaction variable with political interest, campaign mobilisation and appeal of best choice. † The importance of the European election was judged in terms of timing, i.e. the interval between two adjacent national elections.