Representational Roles

Agnieszka

Dudzińska,

Corentin

Poyet,

Olivier

Costa,

and

Bernhard

Weßels,

“Representational Roles” in Kris Deschouwer and Sam Depauw (ed.), Representing the People: A Survey Among Members of Statewide and Substate Parliaments, Oxford University Press, 2014, p. 19-38.

Abstract Political representation is a multifaceted concept. The literature has made a distinction between the focus and the style of representation. The first refers to the groups in society that MPs claim to represent. They can refer to the people of the country as a whole, or rather focus on specific groups and specific interests. The style of representation captures the degree in which MPs act as a delegate with a clear mandate or rather as a trustee who defines more independently the lines to be followed. In this chapter we assess the way in which and the degree in which the focus and the style of representation are affected by the political system. More in particular the level of governance and the electoral system appear to have an effect on the representational roles. The party organization and especially the way in which parties select their candidates also defines both style and focus of members of parliament.

http://ukcatalogue.oup.com/product/9780199684533.do

1

DRAFT VERSION

Political representation has often been defined as an activity of making citizens’ voices, opinions, and perspectives “present” in the public policy making processes1. Anyway, for some theorists, this definition constitutes only a part of the concept. Pitkin (1967) differentiates four different forms: formalistic representation (institutional arrangements), symbolic representation (a symbol brings the symbolized/corresponding object to mind even if it is not similar to it), descriptive representation (reflecting the whole picture by its part, e.g. reflecting the socio-demographic profile of citizens or their opinions), and substantive representation (the activity of representing). Hence, representation is a multiple concept with several dimensions (Vieira and Runciman 2008) that are very difficult to address simultaneously by both politicians (how should MPs behave according to norms and expectations? How do they actually behave?) and academics (which dimension of representation should be studied? Should it be analyzed from the perspective of citizens or the one of MPs?). Representation needs to be studied as a complex and dynamic political process, and not as a purely technical mechanism of delegation of power. Due to the complexity of this concept, it is impossible to take into account all its dimensions at the same time. The objective of this chapter is to study representation from the MPs’ point of view by asking them about their vision of their job and by explaining their choices. This is in line with the general political science approach towards legislative roles, which is more focused on MPs’ selfperception and less on social expectations (Blomgren and Rozenberg 2012b, 9). Of course, as said by De Winter (1997), the MPs’ considerations are numerous and difficult to reconcile even if general trends can be drawn. In other words, our questioning is how MPs view their role and which factors are influencing them. We try to explain the inter-country differences in the representative role perception by exploring institutional factors: those concerning the political system and political parties. Many authors have shown that representatives are submitted to multiple sources of pressure and often have several principals (e.g. citizens in general, their constituencies, those who voted for them, their parties, their parliamentary groups, public opinion leaders, the media, or interest groups; see Bradbury 2007; Costa 2002; Farrell and Scully 2010; Kerrouche 1

Stanford Encyclopedia of Philosophy, http://plato.stanford.edu/entries/political-representation/ [last consultation: 07/04/12].

2

2004; Weßels and Giebler 2011). These findings challenge the legal requirement and cultural norm that MPs are to represent all people in the country, as well as the legal and normative idea of the independent mandate, according to which MPs should follow their own opinions rather than someone else’s instructions. Some scholars point to a populist paradigm, which reduces representation to a mere substitute of direct decision making (see Essaiasson 2001). Others argue that, due to a selection, representation may be better than direct democracy (Brennan and Hammlin 1999, see also arguments of Mansbridge 1999 and Urbinati 2000). In the realm of actual political representation, MPs sometimes face conflicting interests: they have to choose between following the idea of independency and following someone else’s instructions, and further to determine whether they find it more important to promote the interest of the whole (nation, region) or the interest of just a part of it. There is empirical evidence that MPs undertake different roles depending on the situation and the problem to be solved (Andeweg 2012, 81). The seminal study of Miller and Stokes shows that, for example, models of representation and thus role structures differ between issue areas (Miller and Stokes 1963). Costa and Kerrouche (2007, 2009) have highlighted a dual conception of the mandate in France: all MPs claim to represent the Nation but, at the same time, a vast majority of them consider(s) that it is also their duty to advocate for their constituency and constituents, and that there is no contradiction in these terms. The same tension between the universal norm of representing all people and actual trade-offs between conflicting foci of representation has also been found at the regional and the local level in Poland (Dudzińska 2011, Dudzińska and Post 2009). The way MPs perceive their role as representatives has received much attention in political science since the famous speech of Edmund Burke to the Electors of Bristol in 17792. His role definition was both universalistic (one national interest) and independent (following one’s own judgment). 20th century’s theorists of political representation (Eulau et. al. 1959) distinguished two different dimensions of Burke’s view on the representational role: they called them focus and style. Although they are presented separately, it is the common view 2

“Your Representative owes you, not his industry only, but his judgement; and he betrays, instead of serving you, if he sacrifices it to your opinion. (...) Parliament is not a Congress of Ambassadors from different and hostile interests; which interests each must maintain, as an Agent and Advocate, against other Agents and Advocates; but Parliament is a deliberative Assembly of one Nation, with one Interest, that of the whole; where, not local Purposes, not local Prejudices ought to guide, but the general Good, resulting from the general Reason of the whole. You chuse a Member indeed; but when you have chosen him, he is not Member of Bristol, but he is a Member of Parliament.” (Mr. Edmund Burke. Speech to the Electors of Bristol, The Online Library of Liberty: http://oll.libertyfund.org/index.php?option=com_staticxt&staticfile=show.php%3Ftitle=659&layout=html)

3

that these dimensions are strongly interrelated. The focus of representation defines whom MPs represent or should represent: the whole electorate or a part of it. In this discussion, Eulau and Karps (1977) point to three categories: interests defined territorially (geographical representation), interests defined functionally (group representation), and interests defined individually (representation of the own voters).3 The style describes the way the MP comes to his decisions: following his own judgment of the general interest or following others’ instructions; this second dimension is usually described by the labels of “trustee” (own judgment) and “delegate” (instructions of a principal)4. Based on empirical evidence from France, Converse and Pierce (1986) further distinguished between “party’s delegate” and “voters’ delegate”. The investigators of representative roles have developed standardized questions to highlight MPs’ preferences for a specific style (eg. Gross 1978, Weßels and Giebler 2010). The trustee/delegate model has been much challenged and discussed, anyway it seems to remain useful (Blomgren and Rozenberg 2012a, 4). The strongest criticism has come from the defenders of the responsible party model (Mair 2008; Thomassen and Schmitt 1997; Weßels 2007), for which the main actor is the party, and not an individual MP. In this regard, who stands for an election is less important than what his party stands for, and parties are central to explain MPs’ behavior, beyond their own views and those of their constituents. This criticism does not withstand empirical evidence: even if parties/groups are playing a key role in elections and in parliamentary work, MPs can still influence public policy without their support and generally enjoy more autonomy than predicted by the theory (Thomassen and Andeweg 2004). Further, some answers to the critics of the representative role concept have been delivered by the theory of delegation (Braun and Gilardi 2006), which incorporates parties into the delegation chain (see also Müller 2000). According to this, political parties help parliamentary democracies to deal with adverse selection (parties reduce the information problems of citizens), although they do not solve the problem of moral hazard which concerns elected representatives. Still another way of dealing with parties is to argue that instructions from voters also comprise those from political parties. In this view, elections are a multistage process and parties are important selectors: MPs feel obliged to be responsive to both the 3

Esaiasson (2001) points to the lack of party interests in this approach, but it seems legitimate to count them as group interest (a functional representation). 4 Eulau et al. distinguished also a third style, called „politico”, a MP who takes the role of a trustee or a delegate depending on the situation. However, the “politico” style has not found much response among later analysts of the representative role.

4

party structures and the party voters. This is partly due to the electoral system, as we will demonstrate later. Finally, some scholars adapt the Miller-Stokes diamond for political parties (e.g. Schmitt 2001). Our analysis concentrates on the representative role, an approach closely related to the substantive representation as acting. Sharing the view of Wahlke (1962), the argument is that an MP’s role influences his legislative activity. In other words, delegate’s activities differ from those of a trustee. In this chapter, we explore how the members of national and regional parliaments perceive their role. Specifically, we ask MPs how important it is for them to promote the views and interests of certain groups of people (their perception of the focus of representation), and how MPs should decide facing conflicting opinions in voting situations (their perception of the norm concerning the style of representation). We then compare the relative importance of different foci and look for possible determinants of both focus and style. The most recent studies on this topic stem from neo-institutionalist perspectives and emphasize the influence of the institutional context on representation (Müller and Saalfeld 1997; Searing 1994; Weßels 1999a; Strøm 2012). Since the seminal works by Miller and Stokes (1963), Mayhew (1974), and Fenno (1978), a large body of literature has studied the effect of electoral institutions on MPs’ legislative activities in detail. These authors have demonstrated how electoral institutions affect MPs and how they shape their vision of representation. According to this literature, these institutions are decisive in explaining MPs’ attitudes whereas individual factors and background traits have little impact on them (Jewell 1970; Weßels 1999a). In this chapter we will show that institutions, including the political system and political parties, matter. We will go beyond the demonstration of their effects by making the more specific claim that institutions matter as they should. Our assumption is that institutional arrangements are designed for particular forms of representation and thus should produce specific representational roles. Furthermore, we assume that parties are producers of particular representational roles when their organizational power, namely their power over the selection and nomination of candidates, allows this. If this claim is correct, the relevance of this study on role orientations thus results from the demonstration of the link between norms embedded in the political system and organizations.

5

I. STATE OF RESEARCH, APPROACH AND HYPOTHESES In the field of representation studies, the majority of research is country-specific (Bradbury 2007) and comparative studies are rather rare (see for example Pilet, Freire and Costa 2012a; Miller et al. 1999; Schmitt and Thomassen 1999). Some works have a comparative ambition, but they often suffer from a lack of sufficiently standardized data to allow reliable comparisons or they rely only on a limited number of cases. In order to determine which factors impact representational roles, it is necessary to adopt a comparative perspective and to use standardized data from a large number of countries presenting various institutional designs. The PARTIREP dataset is perfectly suited for this purpose. In our study, we analyze the influence of three sets of factors on the perception of representative role defined by its two dimensions, focus and style5. The factors that we expect to influence role perception are primarily the characteristics of the political system and party factors, including selection and nomination procedures. We also consider some individual characteristics. We will begin by looking at the relationship between focus and style: although interrelated, they concern different aspects of the representative role. Thus, in search for its determinants , we will analyze each of them separately in the next sections. In the multivariate analysis, we first investigate the influence of political system’s factors: (1) multi-level versus unitary system, (2) regional versus national parliament, (3) electoral formula (4) type of the list, (5) preferential vote, (6) whether ballot papers mention candidates’ names, (7) the influence of district magnitude. Many of these factors have already been studied for some countries, although the results are not consistent. In their analysis of the possible influence of ballot structure on MPs attitudes, Pilet, Freire and Costa (2012b) have not found any relation between ballot structure and constituency representation: district magnitude appears to be more important than ballot structure. Other studies have shown a positive effect of open-list systems on the MPs’ focus on their constituency (Bowler and Farrell 1993; Carey and Shugart 1995). Literature has also highlighted the effect of electoral formula: according to Thomassen and Andeweg (2004) and Weßels (1999b), proportional systems have negative effects on the importance of the constituency. However, the effect of this factor must be put into perspective because it is hard 5

There are also other typologies of representative roles. Many of them develop the typology of Searing (1994). For their review and comparison see Andeweg (2012)

6

to determine whether it is a result of the electoral formula or rather of the number of MPs by district (single- or multi-member district). More empirically, the majority system with single member districts seems to encourage MPs’ focus on the party (Pilet, Freire and Costa 2012b). The impact of district magnitude is emphasized by Mayhew (1974) and Fenno (1978). However, its effect is controversially discussed in the literature. For Weßels (1999b: 222), large districts bring MPs to define their focus of representation less in terms of their district, because large districts are not, in Rehfeld’s terms (2005), communities of interests. Empirically, this negative effect has been demonstrated by Pilet, Freire and Costa (2012b), who compared the cases of Belgium, France and Portugal. These results corroborate partially those of Carey and Shugart (1995): for them, the effect of district magnitude depends on electoral rules and particularly on the possibility for voters to modify the lists or not. They showed that large district magnitude negatively affects the focus on constituency only in closed-lists systems. A specific set of institutional arrangements related to the MPs’ representative roles stems from political parties. Party factors are: (1) government participation, (2) party ideology, (3) the degree of centralization of the nomination procedure, (4) nomination procedure by the party (party agency or primaries) or by party leaders, (5) change in the vote share for the party from the previous election, (6) position of a party on the left-right scale. As far as we know, internal party selection rules have rarely been analyzed directly in connection with representative roles. However, Weßels and Giebler (2010) have shown some limited effects of party selection rules. For example, they found a higher relevance of party representation than representation of people in a region if national party leaders have a high impact on candidates’ nomination. Several other studies, which do not concern representation as such, hint to the impact of other aspects of nomination and selection on role orientations. According to Morgenstern (2004), selection procedures influence the level of party cohesion and, hence, MPs’ attitudes and behaviors. Coppedge (1994) has proved that centralized candidate selection enhances party cohesion. Costa and Kerrouche (2007) have found similar results in France, where party leaders have considerable power over candidates’ selection. Hence, we can expect MPs from parties with strong leaders and centralized selection procedures to be very focused on their party. However, in order to favor the personal vote and to maximize their chance to win the election, MPs are also encouraged to take care of their reputation in the constituency and, thus, to pay much attention to this level of representation. This, obviously, only makes sense when a ballot paper mentions candidates. 7

We also have checked the influence of some individual factors on the role perception of MPs. Here, we have taken into account MPs’ socio-demographic profiles: (1) sex, (2) age, and (3) education, as well as their political background: (4) political experience at the local level, (5) holding a prominent position in the parliament (speaker, party group leader, chair of a permanent committee), and (6) parliamentary experience.

Our general hypothesis concerning foci of representation (the represented) is that, despite the normative expectation of promoting the universal interests of the political entity rather than some more specific ones, MPs often tend to choose a particularistic or a group approach to political representation (for a classification of group specificity and regional scope of representational foci see Weßels 1999b). In this view, the importance of the universal focus6 will not be determined by institutional factors, but anyway they will incline MPs to appreciate their constituencies, parties, voters, or specific groups. A focus on the party will be connected with party concerning factors, especially internal selection rules: the more centralized and exclusive the nomination procedure, the stronger the focus on the party. Focus on voters should be stronger in single-member districts. Finally, the focus on the geographical constituency should be stronger for regional MPs than for national ones. Taking a particular style orientation will depend mainly on party factors and some institutional ones. The style of a party delegate will be enhanced by a centralized nomination procedure and a lack of inclusive primaries, as well as by closed lists under PR system. The style of a trustee will be connected with open lists, and a lower level of nomination. The style of a voters’ delegate will be more frequent under majority/plurality formula and in small districts. The general mechanisms, which produce different priorities in roles, both style and focus can be understood as incentives set out by the electoral system and the combination with rules of nomination. From a stylized perspective, confronting the two general poles of electoral systems, the argument for a fully proportional list system would be the following: because candidates are running on lists (electoral system), and parties decide who gets where on the list (nomination), a candidate's success in terms of getting a mandate is fully dependent on where the party places him. Parties as organizations have an inherent interest in coherence, demanding to fit the party. Thus, the appropriate role for a candidate to take in order to make 6

Operationalized as “all people in the country” for national MPs or “all people in the region” for regional ones.

8

it to a promising place on the list is to present himself as a party representative, focusing on party voters. In contrast, the success of a candidate in a single member district depends on his individual ability to mobilize votes. Every party in such a system would be well-advised to nominate candidates according to this ability. In this case, the incentive comes from the constituency and demands the representation of its interests. Thus, in a majority or plurality system, the dependency of candidate's success changes from party to voters. This very sketchy summarized mechanism only provides a very general claim of how the much more complex mechanisms can work. It is a very simple image of how institutions may matter. The empirical results demonstrate that the real world is not that simple. At the same time they suggest that there is something to the thesis not only that they matter but also that they matter how they should.

9

II. DATA AND RESULTS II.1. Relevance of Foci and Styles of Representation For the measurement of role orientation, we have used five foci, based on the questions on importance of different foci from the PARTIREP questionnaire. With the aid of a 7-point scale (1= of no importance, 7= of great importance), the importance for the following foci has been measured: all people in the country (or all people in the region for regional MPs), the party, one’s voters7, the constituency, and a specific group8. We have also analyzed tradeoffs between the importance of foci in pairs what has required to build up 10 derivative variables, one for each pair of foci. Three styles of representation have been distinguished, based on the transitive patterns of the answers to three questions, concerning decisions in case of certain pairs of conflicting opinions: party opinion vs. MP’s own opinion, MP’s own opinion vs. voters’ opinion, and voters’ opinion vs. party opinion9. The transitive styles are: ‘trustee’, ‘party delegate’, and ‘voters’ delegate’. Those who do not demonstrate any transitive pattern fall into the category of ‘undecided’. The mean importance of the five foci under study, illustrated in table 1, is almost the same, around a mean value of 5.5 for most of them (except for the focus on a specific group which is rated significantly lower than other foci). There is thus seemingly not much tension between different foci of representation.

Table 1. Mean importance of 5 foci of representation PARLIAMENT National, unitary system National/federal, multi-level system National in general Regional In general

All people 5,5 5,2 5,3 5,6 5,4

Party 5,7 5,4 5,5 5,7 5,6

FOCI Voters 5,3 5,5 5,4 5,6 5,5

Specific group 4,6 4,3 4,4 4,8 4,6

Constituency 5,1 5,7 5,5 5,9 5,7

Source: PARTIREP MPs survey.

7

Asked only where the electoral system allowed for candidate preference. We decided to drop “party voters” from our foci because its importance of this category is strongly correlated with “one’s own voters” (r=0,643) and “the party” (r=0,642). 9 The distribution of answers for each single question was: 61% the party’s opinion rather than own opinion (39%), 67% one’s own opinion rather than voters’ opinion (33%), and 65% party’s opinion rather than voters’ opinion. 8

10

Some differences arise if we compare the two geographical levels of representation (regional and national) and if we split national MPs into two groups (unitary and multi-level systems). The universal focus on ‘all people’ is the most important for regional MPs, then for MPs from national parliaments in unitary systems, and the least important (with still the high value of 5.2, however) for national MPs in multi-level systems. ‘Constituency’, although quite important for all three groups of MPs, is the least important at the national level in unitary systems (5.1 compared with 5.7 and 5.9 for national MPs in multi-level systems and for regional MPs, respectively). This suggests a relative acceptance for the principle of a unitary system, not only in terms of governance, but also in terms of political culture. The nation as a whole is there considered as more important than the constituency. Table 2. Comparison of importance in pairs of foci (percentages in rows) More important All people 36 All people 41 All people 21 All people 35 Party 34 Party 14 Party 33 Voters 17 Voters 28 Specific group 55

Equally important Party 31 Voters 29 Specific group 51 Constituency 24 Voters 27 Specific group 55 Constituency 26 Specific group 56 Constituency 26 Constituency 14

33 30 28 41 39 31 41 27 46 31

Source: PARTIREP MPs survey.

When we compare the importance of the five foci of representation in pairs, we find a relatively large group of MPs who rate both foci equally important (ca. 30%). Conflicts between representative roles have already been identified by the literature, which also highlighted conflicts between legislative roles finding additional evidence for the seminal work of Elau et al. (1959). Our findings show that the resolution of potential conflict is not present in about 30% of the cases. This assertion is particularly significant in a situation of high importance given for each focus.

11

In turn, when we look at the style dimension, we can observe the dominance of the party delegate model (48%) and the trustee model (31%) over the minority of MPs who present the style of voters’ delegate (16%). The figures for regional MPs do not differ much from those for national MPs, although among them the voters’ delegate is slightly more frequent and the trustee less frequent. This result is important when we compare it with citizens’ expectations: in the cases of Finland and the United Kingdom, voters do not expect an MP to be a party delegate (Bengtsson and Wass 2011; Carman 2006). Hence, there is some indication for a gap between citizens’ expectations and the reality. Some striking differences are found between MPs from national parliaments, depending on the unitary or multi-level system of governance. The unitary system, with a figure of 61%, seems to promote the style of a party delegate the most. This is obviously at cost of the role of a trustee which is twice less appealing among national MPs from unitary countries than among those from multi-level countries. Some institutional arrangements may be partly responsible for these differences what we will explore later in multivariate analysis.

Table 3. Style distribution (percentages) PARLIAMENT National, unitary system National/federal, multi-level system National in general Regional In general

Trustee 20 40 33 28 31

Party delegate 61 42 48 48 48

Voters’ delegate 12 14 13 18 16

Undecided 7 5 6 5 5

Source: PARTIREP MPs survey. Percentages sum up in rows.

To determine if the two role orientations are empirically interrelated, we can compare the mean values on the scale of importance of the foci for different styles. Although most of the differences seem small, they are statistically significant for every focus except the universal one. These results are in line with the traditional view of two interrelated dimensions. The most noticeable one is the positive relationship between being a party delegate and having higher scores for the focus on the party10. Also, for voters’ delegates, 10

This result is confirmed by an additional analysis of the importance of representing one’s own party in relation to other foci of representation, i.e. by comparing the scores for party with all other foci in pairs. Party is always more or equally important than other foci for 19% of the MPs in general (?), out of which it is more important for 3%. The respective figures for party delegates are higher: 21% and 5%, while the lowest scores are among voters’ delegates (12% and 1%, respectively).

12

voters are, as expected, a much more important focus than for the other style orientations. Another particularistic category that receives its highest score in this group is the constituency. Voters’ delegates also find it more important to promote specific groups than others do.

Table 4: Link between focus and style

Trustee Party delegate Voters delegate In general

All people

Party

Voters

32 5,4 5,5 5,4

5,2 5,8 5,5 5,6

5,5 5,4 6,1 5,5

Specific group 4,4 4,6 4,9 4,6

Constituency 5,6 5,7 6,0 5,7

Source: PARTIREP MPs survey. Coefficients represent the means on each focus scale for each style.

Before we go into the multivariate analysis, we present the two measures of representative role country by country.

Table 5 illustrates significant differences in the importance of the five foci among the countries under study. The mean importance of the universal focus of representation, measured on a 7-point scale ranges from 4.8 in the United Kingdom to 6.5 in Israel, of the party focus – from 4.8 in Ireland to 6.4 in Norway, of the voters’ focus – from 4.9 in Ireland to 6.0 in Hungary, of the specific group – from 3.5 in France to 5.4 in Austria, and of the constituency focus – from 3.9 in Israel (where it was defined as ‘area of residence’) to 6.2 in Austria. For the style, the range of percentages for specific styles is also very broad: from 7% (Norway) to 68% (Switzerland) for trustee, from 14% (Switzerland) to 86% (Netherlands) for party delegate, and from 5% (Norway) to 26% (Portugal) for voters’ delegate. We suppose that these differences result from a unique combination of institutional factors in each of the countries rather than from a specific political culture.

13

Table 5. Mean importance of 5 foci of representation in 15 countries All people

Party

Voters

Specific group

Constituency

Austria Belgium France Germany Hungary Ireland Israel Italy Netherlands Norway Poland Portugal Spain Switzerland United Kingdom

5,0 4,9 5,9 5,2 5,3 5,1 6,5 5,4 5,2 5,8 5,7 6,0 5,9 5,0 4,8

5,8 5,3 5,1 5,5 5,8 4,8 5,6 5,1 5,8 6,4 4,9 5,5 6,1 5,4 5,4

6,0 5,2 5,3 5,7 6,0 4,9

5,6 5,1

5,4 4,4 3,5 4,6 5,1 4,0 5,3 4,0 5,2 3,7 3,8 4,1 5,2 4,6 3,8

6,2 5,1 5,5 6,1 5,6 5,5 3,9 5,3 4,1 5,5 5,3 5,8 6,2 5,2 5,9

In general

5,4

5,6

5,5

4,6

5,7

5,3 4,9 5,1

‘Constituency’ was replaced by ‘area of residence’ in Israel and the Netherlands. Source: PARTIREP MPs survey.

Table 6. Style distribution in 15 countries (percentages)

Austria Belgium France Germany Hungary Ireland Israel Italy Netherlands Norway Poland Portugal Spain Switzerland United Kingdom

In general

Trustee

Party delegate

Voters’ delegate

Undecided

26 26 47 34 25 10 20 38 8 7 39 32 9 68 40 31

48 52 43 55 47 77 53 28 86 81 45 37 65 14 44 48

22 13 10 6 18 10 17 22 6 5 12 26 24 15 10 16

4 9 0 5 10 3 10 12 0 7 4 5 2 3 6 5

Source: PARTIREP MPs survey. Percentages sum up in rows.

14

II.2. Determinants of the Focus of Representation

Table 7 All people

Party

Constituency

Voters

Specific group

B

B

B

B

B

-,186***

-,076*

,080**

,187***

,074*

,131***

-,095**

,093**

,060*

-,150***

-,021

-,144***

,034

-,160***

-,129***

Political system factors multi-level system of governance Regional level of governance paper mentions candidates voters can vote for (a) candidate(s) Open list Majority system district magnitude logged R2 Party factors governing party Socialists Christian democrats Conservatives

-,021 -,085

,008 ,085** ,061* ,036

,016

-,129***

-,012

,033

,053*

-,041

-,035

-,247***

-,001

.054 B

.033 B

.144 B

.036 B

B

-,003

-,004

-,047

-,059

-,019

-,057

-,057

,094

,020

-,041

-,036

-,117*

,158***

,078

,040

,081

,043

,172***

,137

,020

-,116*

-,020

-,040

-,050

,036

,057

-,057

-,125***

,004

-,090*

-,001

,049

-,053

,041

,089*

,083

-,007

-,066

,055

,196***

,136***

-,024

,027

,041

,088**

-,008

,018

,022

-,069*

-,007

,096*

-,020

-,017

-,034

-,100*

,024

-,138***

.044 B

.026 B

.056 B

.047 B

.025 B

-,015

,134***

-,008

-,027

,132***

-,059

-,043

-,016

-,032

-,111***

,015

,058*

,005

-,022

-,050

,057

,047

,016

,024

-,019

,069*

,003

,010

,024

-,029

-,033

-,061

-,089**

-,106*

-,019

.013

.036

.009

.016

.042

Liberals Green

,073

far right party low level of nomination Selection by party leaders index of volatility of party voters (continuous) position of the party on a L-R scale (from 0 left to 10 right) R2 Individual factors Woman age (in years) university eduaction local political experience Prominent functions parliamentary experience (in years from the first entrance) R2

Source: PARTIREP MPs survey*** P