Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=rjpp20
Download by: [Masarykova Univerzita v Brne] Date: 13 October 2016, At: 06:28
Journal of European Public Policy
ISSN: 1350-1763 (Print) 1466-4429 (Online) Journal homepage: http://www.tandfonline.com/loi/rjpp20
POLITICAL GROUP COHESION AND ‘HURRAH’
VOTING IN THE EUROPEAN PARLIAMENT
Shaun Bowler & Gail McElroy
To cite this article: Shaun Bowler & Gail McElroy (2015) POLITICAL GROUP COHESION AND
‘HURRAH’ VOTING IN THE EUROPEAN PARLIAMENT, Journal of European Public Policy, 22:9,
1355-1365, DOI: 10.1080/13501763.2015.1048704
To link to this article: http://dx.doi.org/10.1080/13501763.2015.1048704
Published online: 30 Jun 2015.
Submit your article to this journal
Article views: 410
View related articles
View Crossmark data
Research Agenda Section
Edited by Michael Blauberger
POLITICAL GROUP COHESION AND ‘HURRAH’ VOTING IN THE
EUROPEAN PARLIAMENT
Shaun Bowler and Gail McElroy
ABSTRACT This research note re-examines the evidence for the claim that political
group cohesion has risen over time in the European Parliament. We first demonstrate
that political group cohesion has always been high in the EP and one of the
principal reasons for this is owing to the large number of lopsided votes that take
place. We next demonstrate that on votes in which the main political groups are
opposing, intra-party cohesion is lower than conventional wisdom would have us
believe and, furthermore, cohesion on such votes has not systematically increased
over time.
KEY WORDS Consensus politics; European Parliament; party cohesion; party
discipline; roll call votes.
INTRODUCTION
In this short research note we re-examine the evidence underlying the belief, almost
an article of faith, that political group cohesion has risen over time in the European
Parliament (EP). We first demonstrate that political group cohesion was surprisingly
high even in the Parliament’s earliest days, and that such cohesion is primarily a
function of the number of lopsided votes that occur. We then demonstrate that
on votes in which the main political groups are opposing, intra-party cohesion is
nowhere near as high, nor has cohesion on such votes necessarily increased over time.
We argue that what we can learn about the EP from roll call analysis is heavily
circumscribed and must, as other authors have argued, take account of the processes
by which votes have been generated and the context of individual votes.
Simply treating all roll call votes as equivalent is problematic; it is not necessarily
the case that large datasets of votes are inherently more informative than carefully
chosen sub-samples. We conclude that it is likely time to embrace other
areas of EP research. We outline one broad area in particular – that of consensus
politics – which has received considerable attention in other European Union
(EU) institutions but relatively little attention within the EP. Following on from
this broad topic we suggest some other areas that may offer deeper understanding
of the dynamics of policy making in the European Parliament.
# 2015 Taylor & Francis
Journal of European Public Policy, 2015
Vol. 22, No. 9, 1355–1365, http://dx.doi.org/10.1080/13501763.2015.1048704
ROLL CALL ANALYSIS AND PARTY GROUP COHESION WITHIN
THE EP
The study of party cohesion and its close relative, party discipline, have long
fascinated political scientists (e.g., Bowler and Farrell 1999; Cox and McCubbins
1993; Krehbiel 1993; Ozbudun 1970). Some of the earliest examples of
systematic quantitative work in the field focus on this very issue (e.g., Burton
1936; Rice 1928). Questions that have been addressed include: how party
cohesion arises; why is it higher in some legislatures than others; what electoral
and legislative institutions are most compatible with high levels of discipline
and what are the best measures of cohesion. Scholars of the European
Parliament have been no less captivated by the topic, especially as the challenges
of building cohesive parties seem so demanding in this extremely
diverse environment.
The European Parliament is a particularly testing environment in which to
establish discipline. In addition to the problem of co-ordinating members of
the European Parliament (MEPs) from dozens of distinct national parties
within one group, the structure of the EP and the nature of party competition
in European elections further conspire against cohesive political groups. For
example, there is no government resting on a vote of no confidence in the
EP, thereby eliminating the fear of early elections in the minds of wayward backbenchers.
No single party controls the legislative agenda; a factor strongly associated
with party unity in Congressional style assemblies (Cox 2000). The absence
of an ‘electoral connection’ (Mayhew 1974), the ‘valuelessness’ of political
group labels and the second-order nature (Reif and Schmitt 1980) of European
elections means that incumbents’ legislative performance is not a top priority for
voters’ at the polls. Finally, given that nomination procedures remain firmly in
the hands of national parties, threats of deselection ring hollow in the ears of
political group rebels. In short, many of the mechanisms identified in the previous
literature as means by which party discipline can be produced are absent
from the EP. Yet, the EP is home to several seemingly highly disciplined party
groups. The puzzle of how this cohesion has been achieved lies at the core of
many studies of the EP and party group cohesion and has long been a topic
of research within the literature (e.g., Attina 1990; Faas 2003; Hix et al
2005, 2007; Kreppel 2002).
Not surprisingly, studies of the EP have employed the standard technique of
roll call analysis in order to explore party group unity within the chamber. The
resulting literature on party group cohesion may be characterized as sharing two
broad conclusions. First, that party groups show ‘surprisingly’ high levels of
cohesion (e.g., Faas 2003; Schmitt and Thomassen 2009) and, second, that
levels of cohesion have been increasing over successive Parliaments (Hix et al
2005; McElroy 2008). There are some finer-grained analyses that embellish
these two broad patterns. A series of papers, for example, has examined the validity
of roll call analysis with the result that there are varying levels of caution
associated with the selection problems underlying roll call analysis
1356 Journal of European Public Policy
(Carrubba et al 2008; Hug 2010). Taking a different focus, Jensen and Spoon
(2010) caution that niche parties will exhibit different patterns of cohesion than
the larger parties and so emphasize inter-party differences in levels of cohesion: a
focus on the European People’s Party (EPP) or Socialists and Democrats (S&D)
may well provide a picture that does not apply EP wide. These other analyses to
one side, it nevertheless seems to be the case that the broad-brush picture of the
EP is that party group cohesion is high and increasing. Consequently the puzzle
provided by the EP, the question at the heart of the much of the literature on the
legislature, is that of trying to understand how party groups have come to thrive
in such a harsh institutional context.
THE MYTH OF RISING GROUP COHESION:
In this section we re-examine all roll call votes in the European Parliament to
demonstrate that political group cohesion was surprisingly high from the very
start. Even in the pre-legislative assembly when parties had few disciplinary
tools at their disposal, MEPs from the three main political groups (Socialists,
Liberals and Christian Democrats) voted with their co-partisans exceptionally
frequently. Party group cohesion does not rise significantly as the EP
becomes more powerful. We do not witness discontinuities in cohesion
scores, as one might expect, with the ‘shocks’ of new legislative powers (e.g.,
the enactment of the Single European Act) or following elections. Cohesion
starts puzzlingly high and remains so across the 35 years we examine.
To demonstrate this, we examined all available roll call votes for the period
September 1979 to May 2014, collated into six-month periods.1
Cohesion
scores are calculated for each of the main political groups that have been in existence
since the first directly elected for each time interval (i.e., there are 10 observations
per parliamentary session, per party, to give a total of 70 observations for
each group). We utilized the two standard measures of internal political group
agreement, the Rice Cohesion Index and the Index of Agreement (AI) favoured
by Hix et al (2007), which takes account of abstentions. Not surprisingly, the
two measures closely track each other, though the AI index inflates cohesion
scores slightly and tends to be flatter in practice.
Figure 1 plots the trend in cohesion scores over time and demonstrates that
both Rice and AI scores for the three main political groups have remained
stable and high throughout the entire 1979–2014 period. The Socialists
reveal a slight upward movement, but high levels of cohesion for all groups
were established by the mid-1980s, from whence they have been rather static.
Despite the huge changes in both the EP and the European Union (EU)
during the time period, the mean AI and Rice scores for the EPP, for instance,
are exactly the same in 2013 as 1983.
While the basic point of surprisingly high cohesion (given the context)
remains, there simply is little support in roll call data for the contention that
party cohesion has risen over time in the EP: the party groups have always
been cohesive – there is no significant trend upwards.
S. Bowler and G. McElroy: Political group cohesion and ‘hurrah’ voting in the
European Parliament 1357
WHY ARE COHESION SCORES SO HIGH TO BEGIN WITH?
While we have demonstrated that the conventional wisdom that cohesion has
risen over time is inaccurate, the puzzle of why cohesion is (and always was)
so high remains. In this section, we explore this issue further by focusing on
the sample of votes typically used to calculate Rice and AI scores.
When voting is (almost) unanimous in a chamber, by definition, voting
unity scores will be high for any individual subset of representatives (even a
randomly chosen group). If all MEPs across all parties are voting together,
internal group cohesion has little meaning. In the European Parliament, the
number of such lopsided votes is very high and this largely accounts for cohesion
scores that are in excess of those witnessed in, for instance, the US Congress,
where such ‘hurrah’ votes are more rare (Gaines and Sala 2000). In the
European Parliament there are large numbers of roll call votes on issues that
are non-legislative and uncontroversial (e.g., motions for resolution condemning
human rights abuses in country X) and there are many legislative votes in
which consensus has already been achieved, particularly amongst the main
political groups. However, the extent of these lopsided votes is worth high-
lighting.
Taking all votes in each legislative session, we first examine what is the proportion
of total votes for which all (or almost all) MEPs in the chamber vote
together. To arrive at this figure we use a measure of vote closeness from
Figure 1 Cohesion scores over time
1358 Journal of European Public Policy
Carey (2008), where the degree of consensus on an individual vote is defined as
follows:
CLOSEVOTE = 1/Threshold∗|Threshold − %Ayes| (1)
Here, Threshold is the percentage of votes required to pass the bill (e.g., if it is a
simple majority this figure is 50 per cent but it can be higher where absolute or
qualified majorities are needed) and Ayes is simply the percentage voting ‘Yes’.
Using this simple equation reveals that for unanimous votes CLOSEVOTE
equals 0. However, where the wining margin is very small and the chamber is
divided CLOSEVOTE approaches 1.
Figure 2 displays a histogram of this measure of lop-sidedness for all seven
parliamentary sessions (1979–2014). Specifically, it examines how often all,
or nearly all, MEPs from the three main parties vote together. The most striking
element of this figure is how frequently the three main political groups vote
together. These groups not only have high internal levels of cohesion, they
also have surprisingly high levels of agreement with each other. The modal
CLOSEVOTE figure in all seven Parliaments is 0. In EP7, there was absolutely
no division amongst the voting members of the three main groups on 36 per
cent of all roll call votes. Furthermore, for these groups, 87 per cent of roll
call votes had a margin of passage of at least 80 per cent in EP7.
The standard advice in the party cohesion literature is to only include all votes
in which at least 25 per cent of members oppose the winning side (Mainwaring
Figure 2 Closeness and cohesion for the main political groups
S. Bowler and G. McElroy: Political group cohesion and ‘hurrah’ voting in the
European Parliament 1359
and Lin˜a´n 1997) or where the leadership of the two main parties oppose each
other (Cox and McCubbins 1993). But in the case of the EP, this results in
such a small number of the overall votes on which to base cohesion scores
that one has to question the meaning of cohesion itself in this setting.
This vote divisiveness information does, however, offer another intriguing
window onto political cohesion, whether it is real or whether it has changed
over time. If party cohesion is real and rising in the EP, we would expect it
to be strongly related to vote closeness. Specifically we would expect to
witness a U-shaped relationship between the closeness of a vote and the levels
of party cohesion on the vote. By definition, ‘hurrah’ votes will mean all
parties have high cohesion scores, but cohesion should decline as vote closeness
decreases up to a point but rise again where votes are very divisive between
parties, if parties are disciplined. That is, we would expect where the winning
majority is narrow that party members toe the line.
In Figure 3 we plot Rice scores against closeness scores for the three main
groups for all votes by Parliament (using a smoothed quadratic fit plot). By
and large, Rice scores decline as voting divisiveness increases, but then begin
to rise as the votes become very close, but the patterns that emerge are very
varied and it is not the case that on close votes the political groups are highly
cohesive. Interestingly, there is no evidence that close votes are more cohesive
in EP7 compared with earlier legislative sessions. If anything, very divisive
votes show lower degrees of intra-party cohesion in EP7. Also, from Figure 3,
it is clear that the Liberals struggle more than the other two parties to get
Figure 3 Cohesion and vote divisiveness
1360 Journal of European Public Policy
their members to vote together on divisive votes, and this has not changed with
the passage of time.
The big takeaway from Figures 2 and 3 is surely that the divisions and partisan
battles amongst the main three political groups have been settled long before
bills get to the floor of the house in Strasbourg. How the three main parties are
able to manage this degree of consensus is clearly a question of some interest,
especially given they have managed to do so for a long period of time. Even
the large exogenous shocks of successive enlargements and several major revisions
to the Treaties are not seen to have had much effect on party control.
Also, it is worth emphasizing that, for the handful of bills which cannot be
agreed upon beforehand, the main political groups are not as cohesive as
global Rice and AI figures would have us believe.
DISCUSSION: SOME SUGGESTIONS FOR FUTURE WORK ON
THE EP
Roll call data themselves show the limitations of roll call analysis in the context
of the European Parliament. This in turn means that future work on the EP
should pay less attention to roll calls themselves and more attention to other
factors.
Plainly, we believe that greater examination of how consensus may be built is
a key topic of research. These patterns of consensus have been well documented
in other EU institutions (e.g., Hayes-Renshaw and Wallace [2006], Heisenberg
[2005] and Mattila and Lane [2001] find this in the Council), but have less frequently
been address in the EP (for notable exceptions see Ringe [2010] and
Settembri and Neuhold [2009]). We currently have little understanding of
how those high levels of consensus within or between groups may be built or
maintained within the EP, yet consensus is clearly a hallmark of the EP (and
the EU generally). The organizations, norms or other institutions that
support such a process are not well-understood. Part of this may stem from
the Anglo-American skew in studies of legislatures. Actors familiar from European
experience – notably rapporteurs – do not feature in Westminster or
Washington life and, hence, do not feature in United Kingdom (UK) or
United States (US) models. But institutions such as the rapporteur would
seem to be one mechanism by which consensus is built and maintained.
Other features of EP life – the Bureau and party group co-ordinators – are
also ones that either do not appear in the US/UK or appear in different
forms. Arguably, it is the Bundestag and not Parliament or Congress that provides
a model for the EP, as it does in terms of institutional structure (Bowler
and Farrell 1999). An alternative, of course, is to move upstream in the policymaking
process to examine the very early stages of policy formulation rather
than the downstream final point of roll call.
If we take as our starting point the stylized fact that the EP is dominated by
consensus then this does generate a number of quite specific research topics.
S. Bowler and G. McElroy: Political group cohesion and ‘hurrah’ voting in the
European Parliament 1361
First, and most narrowly, if the ‘Big Three’ have indeed decided most things
in the chamber between them, then it should not be surprising that the smaller
groups within the chamber have to purse different strategies. Jensen and Spoon’s
(2010) finding, then, makes sense, and we should see other distinct ways in
which the smaller party groups operate. In some sense, the lower case ‘o’ opposition
of the EP is found within these smaller party groups. The smaller groups
seem not to be invited to the party, as the Big Three make deals. How these
groups respond in terms of parliamentary behaviour to what seems to be a
near permanent exclusion from decisions is plainly a topic of interest. These
groups therefore offer a case of minority political action within a legislative
setting.
Second, this lack of scope for MEPs may, in part, explain the high turnover
rates, especially in the smaller groups. On the face of it this can be added to the
laundry list of factors that are stacked against the EP institutionalizing and
developing strong party groups. Why and how politicians choose the EP as a
career path remains something of a mystery. There would seem to be scope
for greater study of legislative careers, possibly along the lines of the earlier
work of legislative scholars. Such studies would, in turn, inform models of rep-
resentation.
Related to the question of turnover is abstention. How MEPs do not vote may
be as consequential as how they vote when they do – especially given that most
roll calls inside the Chamber seem more consistent with an interpretation of
being stitch ups rather than being exercises in individual legislative will.
There are different reasons for examining the two types of ‘non-voting’ –
abstention and absenteeism. Abstention could quite easily be folded into a
model of individual MEP level rationality: especially when abstentions can be
counted as Nays (Hix et al.’s [2005] rationale for counting abstentions within
the AI speaks exactly to this point). Absenteeism is more interesting, since it
speaks to the institutionalizations of the legislature as a whole and, while
there probably is an individual level model of motivations to absent oneself
from the chamber, it is not clear that the makings of such a model are
present in existing literatures. Broader still is the general question of nondecision
and the power to keep divisive issues off the agenda. The point of
work by Carrubba et al (2008), Hug (2010) and others is to highlight the endogeneity
of roll calls in general. Some roll calls in the EP are essentially triggered
by the actions of the Council and Commission, but to the extent that the EP
leadership of the Big Three, or perhaps more precisely the Big Two, can
shape the agenda of the Parliament, then there is scope for them to squash controversial
or divisive roll calls early on in the process. Again, this pushes the
point of study of the EP further ‘upstream’ from roll call itself.
Third, the patterns of consensus do raise the question of the role of the representational
model of party groups in the EP. Again, the consensus suggests
that this is not a model of representation founded in ‘delegate’ terms. There
is little here to suppose that party groups are representing electoral constituencies
in the sense of Mayhew’s (1974) model. Otherwise, we would have to
1362 Journal of European Public Policy
believe that the three main party groups have faced the same constituency pressures
Europe wide for a generation. Who or what the representational goals of
party groups are is therefore worth some more thought. Similarly, the persistently
high degree of consensus suggests a shift away from a level of analysis
focusing on the individual level MEP. Quite bluntly, the individual MEP
seems to have little freedom of manoeuvre if the votes are ‘hurrah’ votes.
What it means is that the goals of party groups would seem to trump the
goals of individual MEPs and that an individual level analysis of vote behaviour
is consequently limited.
In short, once the central ‘fact’ to be explained becomes the overwhelming
consensus with the EP, then a number of consequences follow. Consensus
becomes a topic of interest in itself. The mode of analysis shifts away from
roll calls and towards some different kinds of data – possibly to more interview
and case study analysis. Not only that, but it generates a series of quite specific
newer research questions – turnover, career choice, minority political action
and a new ‘ideal type’ in the Bundestag.
To be sure, not all of these topics and approaches are absent from the literature
on the EP. After the US Congress and UK Parliament, the EP probably
constitutes the third biggest literature in political science on a single legislative
body. It would be surprising indeed if the topics we have identified were absent
from such a large literature. Indeed, we have identified some of the examples
where scholars have explored some of these issues. In a way, too, there is not
an immediate rallying cry here to a single broad-brush theme other than
‘enough already with roll call’. Still, moving away from roll call analysis may
well imply there is a value in some different tools of analysis as much as different
topics of study.
Biographical notes: Shaun Bowler is distinguished professor of political science
at the University of California, Riverside. Gail McElroy is professor of political
science at Trinity College, Dublin.
Addresses for correspondence: Gail McElroy, Department of Political
Science, Trinity College, Dublin 2 Ireland. email: mcelroy@tcd.ie / Shaun
Bowler, Department of Political Science, University of California, Riverside,
CA 92521, USA. email: shaun.bowler@ucr.edu
ACKNOWLEDGEMENTS
We would like to thank David Beatty for very assiduous research assistance, as
well as the three JEPP referees for their insightful comments.
NOTES
1 We also analysed different subsets of votes, e.g., collated in three months, one month,
and the results remain unchanged.
S. Bowler and G. McElroy: Political group cohesion and ‘hurrah’ voting in the
European Parliament 1363
REFERENCES
Attina, F. (1990) ‘The voting behaviour of the European Parliament members and the
problem of the Europarties’, European Journal of Political Research 18(5): 557–79.
Bowler, S. and Farrell, D.M. (1999). ‘Parties and party discipline within the European
Parliament: a norms-based approach’, in S. Bowler, D.M. Farrell and R.S. Katz
(eds), Party Discipline and Parliamentary Government, Columbus, OH: Ohio State
University Press, pp. 208–22.
Burton, R. (1936) ‘The French Chamber of Deputies: a study of party allegiance,
attitudes, and cohesion’, The American Political Science Review 3(30): 549–56.
Carey, J.M. (2008) Legislative Voting and Accountability, Cambridge: Cambridge
University Press.
Carrubba, C.J., Gabel, M. and Hug, S. (2008) ‘Legislative voting behavior, seen and
unseen: adjusting for selection effects in roll call vote analysis’, Legislative Studies
Quarterly 33(4): 543–72.
Cox, G.W. (2000) ‘On the effects of legislative rules’, Legislative Studies Quarterly 25(2):
169–92.
Cox, G.W. and McCubbins, M.D. (1993) Legislative Leviathan: Party Government in
the House, Berkeley, CA: University of California Press.
Faas, T. (2003) ‘To defect or not to defect? National, institutional and party group
pressures on MEPs and their consequences for party group cohesion in the European
Parliament’, European Journal of Political Research 42(6): 841–66.
Gaines, B.J. and Sala, B. R. (2000) ‘A further look at universalism and partisanship in
congressional roll-call voting’, Political Analysis, 8(4): 399–407.
Hayes-Renshaw, F. and Wallace, H. (2006) The Council of Ministers, Basingstoke:
Macmillan.
Heisenberg, D. (2005) ‘The institution of “consensus” in the European Union: formal
versus informal decision-making in the Council’, European Journal of Political
Research 44(1): 65–90.
Hix, S., Noury, A. and Roland, G. (2005) ‘Power to the parties: cohesion and competition
in the European Parliament, 1979–2001’, British Journal of Political Science
35(2): 209–34.
Hix, S., Noury, A.G. and Roland, G. (2007) Democratic Politics in the European Parliament,
Cambridge: Cambridge University Press.
Hug, S. (2010) ‘Selection effects in roll call votes’, British Journal of Political Science
40(1): 225–35.
Jensen, C.B. and Spoon, J. J. (2010) ‘Thinking locally, acting supranationally: niche
party behaviour in the European Parliament’, European Journal of Political Research
49(2): 174–201
Krehbiel, K. (1993) ‘Where’s the party?’, British Journal of Political Science 23(2):
235–66.
Kreppel, A. (2002) The European Parliament and Supranational Party System: A Study in
Institutional Development, Cambridge: Cambridge University Press.
Mainwaring, S. and Lin˜a´n, A.P. (1997) ‘Party discipline in the Brazilian constitutional
congress’, Legislative Studies Quarterly 22(4): 453–83.
McElroy, G. (2008) ‘Committees and party cohesion in the European Parliament’,
O´ sterreichische Zeitschrift fuer Politikwissenschaft 37(3): 357–73.
Mattila, M. and Lane, J.E. (2001) ‘Why unanimity in the Council? A roll call analysis of
Council voting’, European Union Politics 2(1): 31–52.
Mayhew, D.R. (1974) Congress: The Electoral Connection, New Haven, CT: Yale University
Press.
Ozbudun, E. (1970) Party Cohesion in Western Democracies, a Causal Analysis, London:
Sage.
1364 Journal of European Public Policy
Reif, K. and Schmit, H. (1980) ‘Nine second order national elections: a conceptual
framework for the analysis of European election results’, European Journal of Political
Science 8(1): 3–45.
Rice, S.A. (1928) Quantitative Methods in Politics, Oxford: Knopf.
Ringe, N. (2010) Who Decides, and How? Preferences, Uncertainty, and Policy Choice in
the European Parliament, Oxford: Oxford University Press.
Schmitt, H. and Thomassen, J. (2009) ‘The European party system after enlargement’,
in J. Thomassen (ed.), The Legitimacy of the European Union after Enlargement,
Oxford: Oxford University Press, pp. 23–43.
Settembri, P. and Neuhold, C. (2009) ‘Achieving consensus through committees:
does the European Parliament manage?’, Journal of Common Market Studies 47(1):
127–51.
S. Bowler and G. McElroy: Political group cohesion and ‘hurrah’ voting in the
European Parliament 1365