Introduction

1Electoral system reform has been a much-debated topic in Portugal since 1997. However, although several proposals to change from a party list system to a mixed member proportional system were presented and discussed by the major parties in Parliament, and the constitution was even changed to allow it, in the end the reform never passed.

2There is not much literature about the role of voters in electoral reform (for general reviews, see Katz, 2008; Benoit, 2006; Boix, 1999; Rahat, 2004; for papers on the role of citizens on electoral system reform, see Fournier et al., forthcoming; Norris, 2009; Curtice, 2004; Weir, 2005; Banducci & Karp, 1999; Banducci, Donovan & Karp, 1999). However, this is a gap in the literature: ‘the public’s agenda’ is not only a first step in policy making, but also an element of feedback in any reform process that takes place (Norris, 2009).

3Even more scarce is the analysis of deputy-voter congruence and cohesion concerning these topics. This is another gap in the literature, because although voter preferences are a first step in policy-making, legislators have a crucial role in turning voters’ and parties’ preferences into law. Moreover, as we show below, parties are not unitary actors, and the lack of deputy cohesion might help us understand the failure of electoral reform.

4This helps us answer a fundamental question in each and every academic paper: why should we care for the topic being analysed? We should care because while the role of the citizenry in the processes of electoral reform is fundamental it is a neglected subject. We should care because political parties are not unitary actors and both the divisions among the political elite, and the levels of congruence (or incongruence) between the preferences of deputies and voters in terms of electoral reform, can help us to shed some light on reasons for the success or failure of these reforms.

5We also ask ourselves why should pay attention to the Portuguese case in this respect. We believe we should because electoral reform is a much-debated topic in that country and the role of the citizenry in feeding the public agenda on this issue is clear (particularly since 1997), and because there is a large amount of data (a mass survey and an deputy survey, both conducted in 2008) and have a privileged position from which to study voter and deputy preferences on electoral reform, and relate the voters’ attitude to electoral system reform with their evaluation of the political system.

6After a very brief historical overview, we focus on the attitude of voters towards the operation of the political system and their inter-relation with the level of support for electoral reform. Then we focus on the level of deputy cohesion within and across parties, and on the level of deputy-voter congruence around the topics of institutional reform. We seek to understand in what way the level of deputy cohesion and deputy-voter congruence can help us explain why there have been no changes to the electoral system.

7The study relies on two surveys fielded in Portugal in 2008: a survey of 143 of the 230 Portuguese deputies, and of a representative sample of the adult population living in the mainland (N=1350). [2]

8We then seek to address the following five points. First, are there indicators of strong discontent with the functioning of the political system at the voter level that could be said to press political elite towards electoral reform? Second, we try to ascertain if the support of voters for the idea of political reform is related to discontent with political system. Third, we investigate whether, overall and across parties, there is any deputy-voter congruence in terms of preferences concerning the rules of the game? Fourth, do deputies (and electors) from different parties have different perspectives about the major rules of game? Finally, what are the levels of deputy cohesion within and across parties around the topics of institutional reform?

Proposals for electoral reform in Portugal

9In 1974, after 48 years of dictatorship, Portugal began the so-called third-wave of democratisation.

10Since then, apart from the reduction in the number of Deputies in the Assembly of the Republic (AR) from 250 to 230 in 1991, there have been no major changes to the electoral system. Nevertheless, the debate over electoral reform has been present since the transition to democracy (Freire, Meirinho & Moreira, 2008; Cruz, 1998). Before analysing this debate, let us outline the major rules governing election to the lower (and single) chamber.

11Despite having a president who is directly elected, parliamentary elections are the most important for the operation of the political system.

12Parliamentary elections are conducted under a closed-list proportional representation system in 22 multi-member constituencies (D’Hondt formula). The average number of Deputies for each constituency was 11.4 until 1987, then 10.5. However, there is a large imbalance in the size of districts: from 48-38 (Lisbon 48, Oporto 38) to 3-2 deputies (in the interior and in the emigrant constituency).

13Between 1976 and 2008, there were several debates on electoral reform, encompassing a variety of options, ranging from a change to the majoritarian two-ballot system, to maintaining the current system with minor adjustments (Freire, Meirinho & Moreira, 2008; Cruz, 1998).

14When considering the major reasons for electoral system reform, we should first bear in mind criticisms about the operation of the electoral system have focused on three issues: the need to change the system to increase executive stability (during the 1974-87 period of cabinet instability), the allegedly excessive number of deputies (promoted by right-wing parties, particularly the Social Democratic Party), and the unfavourable conditions for a close deputy-voter relationship at the district level.

15The first issue lost importance after 1987, a period during which the vote was concentrated into support for the two major parties and which led both to a reduction in the fragmentation of the party system and an increase in government stability.

16Since 1992-97, the debate on electoral system reform in Portugal evolved towards two other issues: reducing the number of deputies, and creating more favourable conditions for a closer relationship between constituents and their Deputies (Freire, Meirinho & Moreira, 2008).

17Several suggestions have been put forward in an attempt to overcome the problem of the alleged unfavourable conditions for a close relationship between deputies and their constituents. Since 1997 the preferred solution has been to create a mixed-member proportional system (proposed by the Socialists and the Social Democrats); however, many have complained this would lead to parochialism, clientelism and party political polarisation (Freire, Meirinho & Moreira, 2008). The lack of consensus on the use of single-member constituencies was behind a recently commissioned study by the Socialist Party, which proposed a multiple tiers system with small multimember constituencies in the lower tier and preferential vote (Freire, Meirinho & Moreira, 2008).

18The main reason behind reservations over single-member constituencies was perhaps also that behind the Socialist Party reservations about the preferential vote: they fear losing control of the selection of candidates.

19Nevertheless, the debates have shown that intra-party divisions, and the ‘fear of the unknown’, has contributed towards a failure to introduce reform, and we should remember this when analysing the survey data.

Voters’ attitudes towards the political system

20Although we know from other country studies electoral reform is mainly a top-down process, we also know voters’ discontent with the political system can play a role in the electoral change process. Moreover, reading the proposals for electoral change the two major parties present from time to time, we find references to the need to improve the level of trust the Portuguese have in their political institutions, the progressive decline in identification with the current party system and the ever-increasing level of electoral abstention (Freire, Meirinho & Moreira, 2008). Therefore, we will present a comparative and longitudinal overview of Portuguese attitudes to the political system.

Figure 1

Satisfaction with the way democracy works in Portugal

Satisfaction with the way democracy works in Portugal

21Figure 1 shows Portuguese voter satisfaction with how the democratic system functioned from 1985-2008, and shows there has been an sharp decline since 2002.

22While normally remaining above the levels registered from 2002 on, the reduction in satisfaction with democracy began in 1991 and, despite some ups and downs, has never recovered. The decline in 1991 could be related to the economic crisis that followed the 1986-90 boom; in fact, other downturns might also be connected to economic difficulties and/or austerity packages (e.g. 1995-97, 2000, 2005-08). However, besides the more political-economic reasons, political reasons can also help explain the situation.

Figure 2

How well deputies accurately reflect the views of voters. 2002–08 (% “very well” and “fairly well”)

How well deputies accurately reflect the views of voters. 2002–08 (% “very well” and “fairly well”)

23Figure 2 shows that Portuguese voters don’t feel well represented by their Deputies, and that this feeling has been increasing since 2002. Nevertheless, while it has been shown dissatisfaction with the way Portuguese democracy works and that disaffection with political representatives is rising, that does not mean there is a reduction in the legitimacy of Portugal’s democratic regime (Magalhães, 2004).

24Moreover, neither of the above shown indications tell us about the Portuguese situation in a comparative perspective. Fortunately, both questions were asked in the Comparative Study of Electoral Systems surveys. In Table 1 we see the level of satisfaction with how democracy works in Portugal is well below the average for the 33 countries included in Table 1. The contrast is even greater if we exclude some countries with problematic regimes (e.g. Korea, Peru, Russia, Thailand and Ukraine).

Table 1

Satisfaction with democracy and perceptions of representation in comparative perspective

CSES 1: 1996–2002 CSES 2: 2002–6 CSES 1: 1996–2002 CSES 2: 2002–6 Satisfaction with democracy(a) Deputies know what people think(b) Deputies represent voters’ views(c) Austria 78.0 81.6 14.7 56.3 Belgium 61.7 70.1 28.6 63.2 Bulgaria 47.2 21.3 33.4 52.3 Canada 74.1 70.8 18.6 40.4 Chile – 60.6 32.8 44.4 Czech Republic 61.1 45.9 26.1 28.4 Denmark 89.2 93.4 41.6 79.3 Finland – 69.5 – 47.9 France – 56.1 – 60.6 Germany 67.8 72.2 21.3 49.0 Hong Kong 48.1 46.1 27.8 59.9 Hungary 42.2 47.4 31.1 51.8 Iceland 79.4 69.4 38.0 55.1 Ireland – 80.1 – 62.9 Israel 53.4 35.6 39.2 46.9 Italy – 38.9 – 46.2 Japan 63.3 70.5 15.5 23.9 Korea 41.2 23.0 14.3 – Lithuania 35.7 – 23.7 – Mexico 61.2 33.9 25.6 47.4 New Zealand 69.2 69.4 19.1 55.5 Norway 90.3 78.2 32.4 – Peru 35.3 33.4 33.6 38.3 Poland 63.1 39.0 28.2 48.5 Portugal (date) 53.0 (2002) 47.6 (2005) 32.0 (2002) 37.9 (2002) Romania 43.9 44.4 42.6 37.3 Russia 16.3 31.8 37.5 41.5 Spain 86.0 77.7 31.5 64.4 Sweden 71.0 76.7 23.5 78.4 Thailand 76.9 – 39.5 – Ukraine 9.2 – 53.0 – United Kingdom 75.5 72.6 19.1 49.3 United States 79.9 78.4 23.1 71.5 Mean 1 (all)(d) 59.8 57.9 29.2 51.4 Mean 2(e) 65.7 60.9 27.9 52.0

Satisfaction with democracy and perceptions of representation in comparative perspective

Notes: When there were two elections in one country in each CSES module we always use data from the most recent election. (a) The proportion of people who are “very” or “fairly” satisfied with the democratic regime, (b) proportion of those who believe deputies know what the people think (values 1 and 2 from a scale that goes from 1=deputies know what ordinary people think’ to 5=deputies don’t know what ordinary people think’), (c) those who believe voters’ views are “very” or “fairly” well represented in elections, (d) all countries, (e) countries with strong democratic regimes according to Freedom House ratings (i.e. excluding Korea, Peru, Russia, Thailand and Ukraine).

25The same can be said with respect to voters’ perceptions about “how well their” views are represented in elections (through Deputies)’: the situation in Portugal is again below the average for the stronger democratic regimes. Thus, the comparative perspective reveals reasons for being concerned with Portuguese attitudes towards the political regime, reasons that support the need for reform.

Political reform and voter discontent

26In order to analyse the relationship between voter discontent with the functioning of the political system and attitudes towards political reform, we re-coded several variables included in Table 2 in order for higher values to correspond to both increased criticism and the perception of a need to open new channels for voter participation. Using Principal Component Analysis with Varimax rotation we found there are 3 distinct factors (see Table 2).

Table 2

Factor analysis (PCA) of voter attitudes towards the Portuguese political system

Variables (+ critical positions; – non critical positions) F1: Evaluation of the law-making process F2: Evaluation of the political system F3: Political reform CCS_A3_5 Special interests have too much influence on law making 0.706 CCS_A3_2 The democratic regime is on the verge of losing the trust of the voters 0.651 CCS_A3_6 Lawmaking process is too complicated 0.583 CCS_A3_4 The parties are not the major link between the voters and the state 0.691 CCS_A3_1 There are few opportunities for voter participation 0.603 CCS_A3_3 Legislation reflects the interest of the majority of voters -0.504 0.596 CCS_A3_7 Parliament, not voters, should make final decisions on law and policy 0.508 CCS_A3_8 A certain number of voters should be able to initiate a referendum. 0.794 CCS_A3_9 Quotas are a necessary measure to address the under-representation of women… 0.758 Variance explained 18.99% 17.63% 14.10%

Factor analysis (PCA) of voter attitudes towards the Portuguese political system

Note: 1) In order to analyse the relationship between voter discontent with the functioning of the political system and attitudes towards political reform, we re-coded the variables in order for higher values to correspond to both increased criticism and the perception of a need to open new channels for voter participation; 2) KMO = 0.641; Bartelett’s test – significance = 0.000.

27Those who disagree with the statement “legislation reflects the interest of the majority of voters”, those who think the law-making process is too complicated and too influenced by particular interests, together with those who believe the democratic regime is on the verge of losing the trust of the voters, are all loaded on the same Factor 1.

28Thus, positive values on this factor mean a critical evaluation of the legislative process, and that is why we labelled it “Evaluation of the law-making process”.

29A second factor (F2), labelled “evaluation of the political system”, is composed of those who believe that “the parties are not the major link between the voters and the state”, and also that “there are few opportunities for voter participation”. Those who believe legislation does not reflect the interest of the majority also load positively on this factor. There is yet a fourth variable: those who believe parliament, and not voters, should be the primary political decision maker also load positively in Factor 2.

30The third factor (F3), labelled “political reform”, is composed of those who believe quotas are a necessary measure to increase the presence of women in parliament and those who claim voters should be able to call for referendums.

31Turning now to the electors’ preferences in terms of electoral system reform, we should emphasise that obviously the common voter’s notion of the electoral systems available is rather crude (Curtice, 2004), and many of the definitions inherent to their study are simply not applicable in a mass survey. However, from the voter survey answers we know that 64 per cent of respondents believe the electoral system should be modified to allow people to vote more for candidates instead of parties (the so-called personalisation of the vote).

32The idea of creating more favourable conditions for a closer deputy-voter relationship (i.e. for the personalisation of the vote) lies at the heart of most proposals for electoral reform in Portugal since 1997. Therefore, we will now analyse in what way support for this idea at the voter level is related with discontent with the political system.

33Using the independent variables listed below Table 3, we see that even after controlling for several important variables (the views about the law-making process: Factor 2, economic perceptions, party identification, gender, religiosity and population, ideology, etc.) [3] there is a highly significant positive correlation between negative evaluations of the political system (Factor 1) and support for greater personalisation of the electoral system (Table 3).

Table 3

Explaining voter attitudes towards “reforming the electoral system to personalise the vote” ordinary least squares regression)

Standardised coefficients (?) Factor 2: Evaluation of the political system 0.185*** Factor 3: Political reform (quotas/referendums) 0.157** Age -0.127* Left-right values (index) -0.158** Party identification Democratic and Social Centre (CDS-PP) 0.138* Libertarian -authoritarian values (index) -0.099* Factor 1: Evaluation of the law-making process Ns Evolution of the economy Ns Party identification Left Bloc Ns Democratic Unity Coalition Ns Social Democratic Party Ns Population (area of residence) Ns Religiosity (church attendance) Ns Gender Ns Adjusted R2 0.154

Explaining voter attitudes towards “reforming the electoral system to personalise the vote” ordinary least squares regression)

Notes: Variable coding: Dependent variable = “the electoral system should be changed so that people can vote for a candidates instead of a party?”, with responses on a five point scale (1=totally disagree, to 5=totally agree). Independent variables: (Perception of the) Evolution of the state of the economy (vis-à-vis the previous year): (1) a lot worst; (5) a lot better.
Factor 1: higher values mean higher criticism towards the lawmaking process. Factor 2: higher values mean higher criticism towards the political system. Factor 3: higher values mean more support for political reform. Party Identitification: 4 dummies: (1) for each party (BE, PCP/CDU, PSD and CDS-PP); reference group (0): PS. Values: higher values mean right wing (left-right) or authoritarian values (libertarian-authoritarian). Age: continuous variable ordered in an ascending fashion. Gender: (0) Male; (1) Female. Church attendance: (1) never; (6) once a week or more. Population: (1) < 2000 inhabitants; (5) ? 300000 inhabitants.
Levels of significance: *p?0.05, **p?0.01, ***p?0.001.
Levels of significance: *p?0.05; **p?0.01; ***p?0.001

34By the same token, “the political reform” Factor 3 (support for quotas and referendums) also has a significant positive impact on the view that the electoral system should be changed to give voters greater opportunity to choose among candidates on the party lists.

35With only 15 per cent of the variance explained, this model is far from impressive. Nevertheless, it provides us with some indications about the impact of several variables, namely those that measure discontent with the political system, in relation to a particular change to the electoral system that was mentioned in the survey, and which the parties have often proposed as part of their pledge for electoral reform: the personalisation of the vote.

Deputies’ and voters’ perspectives on the electoral rules

36Rather less explored in the literature are the levels of deputy-voter congruence (or lack of it) around the topics of institutional reform. However, the lack of such congruence, either on the main objectives for the rules of the game or on the desirable direction for reform, can help explain some of the lack of success in achieving electoral reform.

37Although the literature on electoral reform does focus on partisan attitudes, investigation of the levels of deputy cohesion on reform, particularly using survey data, is relatively under-researched. But can also help us explain some of the lack of success in achieving electoral reform.

38Thus, now we focus on the levels of deputy cohesion within and across parties, and the levels of deputy-voter congruence on matters related to institutional reform (with a greater emphasis on electoral reform) and seek to understand in what way the levels of deputy cohesion and deputy-voter congruence can help us understand why there were no changes in the electoral system, despite there being several proposals since 1997.

Table 4.1

Electoral system preferences (I): Voters and deputies (left-wing parties) 2008 (%)

Left Bloc Democratic Unity Coalition Socialist Party Voters Deputies Diff Voters Deputies Diff Voters Deputies Diff What do you prefer, single-party government or coalition government? 1. A government made up of a single party 28.6 (15.4) 0.0 28.6 43.1 (40.0) 0.0 43.1 49.3 (47.1) 95.9 -46.6 2. A coalition government made up of more than one party 71.4 (84.6) 100.0 -28.6* 56.9 (60.0) 100.0 -43.1* 50.7 (52.9) 4.1 46.6* In some countries there are governments in office that rely on less than half of the votes of members of parliament. This form of government is called minority government. What do you think about it? 3. A government needs its own majority in parliament 40.6 (35.7) 25.0 15.6 31.9 (35.7) 0.0 31.9* 57.7 (59.2) 86.1 -28.4* 4. A majority in parliament is less important 59.4 (64.3) 75.0 -15.6 68.1 (62.5) 100.0 -31.9* 42.3 (40.8) 13.9 28.4* Different electoral systems pursue different political goals. If you had to choose between the following political goals, which one would you vote for? 5. Party proportions in votes and seats should closely match 67.6 (73.3) 100.0 -32.4* 51.4 (37.5) 100.0 -48.6* 50.6 (51.4) 86.3 -35.7* 6. The party that obtains most votes should receive more than half of the seats in parliament 32.4 (26.7) 0.0 32.4* 48.6 (62.5) 0.0 48.6* 49.4 (48.6) 13.7 35.7*

Electoral system preferences (I): Voters and deputies (left-wing parties) 2008 (%)

Notes: 1) In the cells under the label “Diff” we present the difference between the percentage of voters minus the percentage of deputies in each cell. Thus, if the “Diff” is positive, voters are more in favour of that option than deputies; if it is negative, it is the other way around; 2) For each “Diff”we tested the respective statistical significance using the “test for the difference of proportions” and considering it to be significant if the value of the test (Z) is lower than -1.96 or higher than +1.96. The significant differences are marked with an asterisk; 3) Some people might argue these topics are too sophisticated for most of the low educated voters and thus they might be making random responses. To control for that possibility, for each cell concerning the voters in parenthesis we show the proportion of more educated voters (secondary education or more) in each particular category.

Table 4.2

Electoral system preferences (I): Voters and deputies (right-wing parties and all) 2008 (%)

Social Democratic Party Democratic and Social Centre All Voters Deputies Diff Voters Deputies Diff Voters Deputies Diff What do you prefer, single-party government or coalition government? 1. A government made up of a single party 42.7 (45.9) 77.8 -35.1* 21.1 (16.7) 12.5 8.6 45.5 (42.4) 80.1 -34.6* 2. A coalition government made up of more than one party 57.3 (54.1) 22.2 35.1* 78.9 (83.3) 87.5 -8.6 54.5 (57.6) 19.9 34.6* In some countries there are governments in office that rely on less than half of the votes of members of parliament. This form of government is called minority government. What do you think about it? 3. A government needs its own majority in parliament 56.3 (59.4) 86.7 -30.4* 30.0 (42.9) 75.0 -45.0 53.9 (54.9) 79.4 -25.5* 4. A majority in parliament is less important 43.7 (40.6) 13.3 30.4* 70.0 (57.1) 25.0 45.0 46.1 (45.1) 20.6 25.5* Different electoral systems pursue different political goals. If you had to choose between the following political goals, which one would you vote for? 5. Party proportions in votes and seats should closely match 58.0 (57.6) 68.9 -10.9 62.5 (40.0) 100.0 -37.5* 56.4 (54.0) 82.6 -26.2* 6. The party that obtains most votes should receive more than half of the seats in parliament 42.0 (42.4) 31.1 10.9 37.5 (60.0) 0.0 37.5 43.6 (46.0) 17.4 26.2*

Electoral system preferences (I): Voters and deputies (right-wing parties and all) 2008 (%)

Notes: See Table 4.1.

39We begin with the levels of deputy-voter congruence and deputy cohesion on the main principles the rules of the game should accomplish (Tables 4.1 and 4.2). In these two tables we consider the response of both voters and deputies to three questions that are closely related to the operation of any electoral system. For each question, we begin with the analysis of deputy-voter congruence and then pass to the question of the level of party cohesion. Please note that to segment voters we used party identification, [4] while to segment deputies we used the party list in which each deputy was elected. [5]

40For the first question, “What do you prefer, single-party government or coalition government?”, the questionnaire suggests two responses: “a government made up of a single party” or “a coalition government made up of more than one party”. Please note that for all the questions, we tested for the statistical significance of the differences between deputies and voters (test for the difference in proportions): the significant differences are marked with an asterisk.

41The majority of deputies (80.1 per cent) opted for single-party government; however, the majority of voters (54.5 per cent) said they would prefer a coalition. This significant mismatch is due to the fact the two major parties, which control more than 70 per cent of the seats in parliament, prefer single party government, and only the deputies of the smaller parties prefer coalition government (between 87.5 and 100 per cent in favour).

42However, among the voters, supporters of the major parties are more divided than their representatives (with 50.7 per cent of Socialist Party and 57.3 per cent of Social Democratic Party sympathisers favouring coalition), which explains why there is a majority of voters supporting coalition.

43We know from previous studies that voters’ attitudes and knowledge about electoral systems is not only rather limited, but also dependent on the format of the question (Fournier et al., forthcoming; Curtice, 2004). For this reason, Tables 4.1 and 4.2 show in parenthesis the percentages for the more educated strata of the population (have completed more than the compulsory nine years of formal education). For the previous question, we can see that, except for those who identify with the Social Democratic Party, the results add strength to the idea the majority of the population (in general and across parties) prefer coalition to single-party government.

44The answers to the second question about principles (“In some countries there are governments in office that rely on less than half of the votes of members of parliament. This form of government is called minority government. What do you think about it?”) point in a similar direction.

45Overall, Deputies (79.4 per cent) are much more in favour of majority government than voters (53.9 per cent), although both are now more in favour (significant difference). At the elite level, a majority of Deputies from the Socialist Party and the parties to its right prefer majority government (remember the Democratic and Social Centre-Popular Party is a small party that was in government from 2002–05,while the parties to the Socialist Party’s left have not been in power). Only Deputies from the parties to the left of the socialists prefer minority government.

46At the voter level, there are large divisions within each party, although the divisions are greater (and more often significant) between those who identify with the major parties (with Socialist and Social Democrat party supporters being more in favour of majority government), while the supporters of the smaller parties (Left Bloc, Democratic Unity Coalition and Democratic and Social Centre-Popular Party) favour coalitions. Controlling for the voters’ level of education makes no real difference to the results, indicating their validity.

47The third question on major principles (“Different electoral systems pursue different political goals. If you had to choose between the following political goals, which one would you vote for?”) offers an alternative between a proportional (“Party proportions in votes and seats should closely match”) and a majoritarian electoral system (“The party that obtains most votes should receive more than half of the seats in parliament”).

48Both deputies (82.6 per cent) and voters (56.4 per cent) are more in favour of proportionality, although paradoxically the former are more so than the latter (significant differences). At the deputy level, this can be explained by the fact the Portuguese system of proportional representation is not only disproportional (Freire, Meirinho & Moreira, 2008, p. 28), but most of all it has returned three single-party majorities and two quasi-majorities. Recall also that proportional representation is a constitutional requirement.

49Although deputies from the minor parties are more in favour of proportionality than their peers in the larger parties, the significant aspect is that among the latter there is a great deal of support for proportional representation (around 86 per cent for both parties). Among the voters, and especially among those sympathetic to the smaller parties, the level of support for majoritarian rule (Left Bloc: 32.4 %, Democratic Unity Coalition: 48.6 %, Democratic and Social Centre: 37.5 %) may be the result of a lack of information.

50However, apart from those who sympathise with the Democratic Unity Coalition and the Democratic and Social Centre (the more educated voters from these parties turned out to be more in favour of majoritarian rule – 62.5 and 60 % – than the less educated: 48.6 and 37.5, respectively), all the other results remain unchanged after controlling for education.

51The results from supporters of the Democratic Unity Coalition and the Democratic and Social Centre are not easy to explain, but we should bear in mind these breakdowns are calculated using a limited number of cases.

Deputies’ perspectives over desirable reformist solutions

52Due to its inherent difficulty for the average voter, the last question concerning the desirable directions of electoral reform was only presented to deputies, meaning we are only able to analyse the issue of party cohesion (Table 5).

Table 5

Deputies’ electoral system preferences (II) (%)

Left Bloc Democratic Unity Coalition Socialist Party Social Democratic Party Democratic and Social Centre All A personalised proportional system with two votes (as used in Germany) 0,0 0,0 6,6 36,4 0,0 14,7 A single member district system (as used in Britain) 0,0 0,0 10,5 9,1 0,0 8,5 A single member district system with two rounds (as used in France) 0,0 0,0 2,6 0,0 0,0 1,1 A pure proportional system (as used in the Netherlands) 40,0 22,2 1,3 4,5 50,0 7,3 Maintain the actual system in country as it is, perhaps with some adjustments 40,0 55,6 35,5 27,3 25,0 34,2 Maintain the fundamental aspects of the actual system in country, but modify some components to create more favourable conditions towards a more personalised representation (as used in Sweden and Denmark) 0,0 0,0 38,2 18,2 12,5 27,4 Another system 20,0 22,2 5,3 4,5 12,5 6,7

Deputies’ electoral system preferences (II) (%)

53Overall, deputies appear divided, providing answers spread across three main options. The first, “Maintain the actual system in country as it is, although perhaps with some adjustments”, receives 34.2 %. The second, which is the proposal that was most recently discussed (Freire, Meirinho & Moreira, 2008), “Maintain the fundamental aspects of the actual system in country, but modify some components to create more favourable conditions towards a more personalised representation (as used in Sweden and Denmark)”, receives 27.4 %. Thirdly, the proposal for a mixed member proportional system (“A personalised proportional system with two votes [as used in Germany]”), received the support of only 14.7 % of the deputies.

54The remaining 23.7 % were split among four options. The lack of cohesion concerning the most preferred form of reform is perhaps one reason why it has been so difficult to change the electoral system (since 1997). Of course, we cannot use these results to explain what happened before 2008, but we know both from the press and from interviews with party leaders that the lack of cohesion within the party elites in relation to electoral reform is not new (Freire, Meirinho & Moreira, 2008).

55Moreover, there is a huge lack of cohesion both within and across the two major parties. For the Socialists, the top three solutions are: “maintain the fundamental aspects of the actual system in country, but modify some components to create more favourable conditions for more personalised representation (as in Sweden and Denmark)” (38.2 %), maintain the present system with perhaps minor changes (35.5 %), and a plurality single-member constituency system (10.5 %).

56As for the centre-right Social Democrats, the top preferences are a mixed-member system like that used in Germany (36.4 %), to maintain the current system with minor changes (27.3 %) and a two-tier system along Scandinavian lines (18.2 %).

57Thus, we see that the major parties are not only divided within (with a strong preference for maintaining the actual system), they are also divided between themselves. Clearly, this lack of consensus on the direction electoral reform should take goes far towards explaining the failures of the reform process—especially if we assume this is not new.

58Understandably, deputies from the smaller parties are divided between support for retaining the existing system and changing to a pure proportional system. Deputies from the Left Bloc split evenly (around 40 per cent), while Democratic Unity Coalition Deputies split 55.6 % - 22.2 % for the status quo and pure proportional, respectively, and among Deputies of the Democratic and Social Centre-Popular Party the split was 25 and 50 %, respectively.

59Of course, deputies representing these parties fear whatever change there may be (apart from a pure proportional system) might hurt them, so they tend to vote to keep things as they are. While they do not have the power to change the rules, unless they have the support of one of the major parties, their position reinforces the view that there is a lack of cohesion among party deputies concerning the desirable forms of electoral reform.

60Of course, divisions within parties can be overcome by the decision of a majority of party officials; however, deputies retain an important role in this type of decision, and while they are so deeply divided amongst themselves electoral reform will remain problematic.

Concluding remarks

61There is not much literature concerning the role of voters in processes of electoral reform, and this is gap is relevant because voters’ preferences are usually the first step in policy-making. Also scarce is the analysis of deputy-voter congruence and cohesion on this topic—a lacuna that is all the more significant because legislators have a crucial role in turning the preferences of voters and parties into law.

62Thus, building on unique survey data, we analysed the role of voters in the process of electoral reform in Portugal, and the level of deputy-voter congruence and cohesion in respect of changes to the rules of the electoral game.

63The paper has mainly addressed electoral reform in Portugal. This is an issue that has been much debated since the transition to democracy in 1974. In the first cycle of these debates, 1978-89, political actors presented three main reasons for reforming the electoral system: increase executive stability; improve parliamentary efficiency by reducing the number of deputies; create more favourable conditions for a closer relationship between voters and their representatives within the constituencies.

64The first reason has lost much of its relevance since the 1990s, when the vote began to concentrate towards the two main parties with a concomitant increase in the level of cabinet stability. However, the other reasons retained their relevance during the second cycle of those debates from 1990 to the present, and reflect the major criticisms that have always been made in respect of some characteristics of the present electoral system—particularly the closed list system and the existence of some districts that return a very large number of deputies.

65Although several learned studies and press articles have shown that the number of deputies in Portugal’s Assembly of the Republic is not great in comparison with other nations, the Social Democratic Party has always made a reduction in the number of deputies a central aspect of its electoral reform proposals.

66With respect to the need to create more favourable conditions for a closer deputy-voter relationship at the constituency level, there is some consensus between the major parties. The constitution was even changed in 1997 to allow for a mixed-member proportional system (with single-member districts and a national constituency as compensation). However, according to several accounts, past failures to reform the electoral system have been due to divisions within and between the main parties—not only with respect to reducing the number of deputies (which was proposed by the Social Democratic Party, but rejected by all others with parliamentary representation), but also in terms of precise solutions for achieving the target of a closer relationship between the voters and their representatives (a mixed-member system with some single member districts or a multiple tiers system with small multimember constituencies—not single-member constituencies—at the lower tier). Since at least 1997, several criticisms have been raised from within both the Socialist and the Social Democratic parties in relation to single-member constituencies.

67Since one of the reasons for electoral reform presented by political actors was the growing problem of the relationship between voters and the political system (falling levels of turnout, declining levels of confidence in the political system and political actors, weaker identification with political parties, etc.) and one of the reasons for the failure to reform has been the lack of cohesion within and between (especially the two main) parties, this paper has analysed levels of deputy-voter congruence and deputy cohesion in terms of the major principles of the electoral rules.

68However—based not only on several voter surveys across time, but also on the academic literature—we have also revised some of trends in terms of the growing problem in the relationship between the voters and the political system. We found there is an enormous amount of voter dissatisfaction with the way Portugal’s democracy works as well as with the representative function performed by deputies. Comparative survey data strengthens our conclusion.

69Moreover, even after introducing several control variables (socio-demographic and political), we found that voters’ “evaluations of the political system” and their “support for political reform” (via quotas and increased referendums) also helps explain support for greater personalisation of the vote: more criticism, more support for this reform.

70We know from some case studies that the attitudes of voters towards the political system and the voters’ support for electoral reform can play an important role in explaining why some reforms were implemented; however, despite the large amount of dissatisfaction with the performance of the political system, and the support for opening new political participation channels (through reform of the electoral system) no reform has taken place in Portugal.

71Of course, this reveals that while voters’ attitudes are important they may not be crucial—at least while (the two main) parties have almost complete control of the process and while referenda on constitutional matters remain impossible—both of which are characteristic of Portugal.

72There is a significant lack of cohesion within and across parties in respect of the path electoral reform should follow. We also discovered that around 30 % of the deputies of the two main parties prefer the current system—albeit with minor adjustments—to any type of reform.

73As for the smaller parties that are afraid of being punished by electoral change, maintenance of the present rules or change to a full proportional system are their two preferences. Of course, we cannot explain past failures with current survey data, but those divisions are in line with what we know from previous accounts, and therefore we can reasonably assume they are probably related to the these failures.

74The lack of cohesion between and within the two main parties is a major obstacle to electoral reform. While we found that both deputies (of all parties) and voters are mainly in favour of (constitutionally required) proportional representation, we found a significant mismatch between voters (who preferred coalition government) and the deputies of the main parties (who favour single-party majority government) on some fundamental rules of the political game. This reveals that while the main parties might favour some seat maximisation with any future reform (even if they never state it publicly)— increasing the possibility of manufactured single-party majorities through the winner receiving a larger seat bonus—our data indicates it is not popular with the voters. Reform in that particular direction could result in increased criticism of the political system from the voters.

75We have shown that the role of Portuguese voters in the process of electoral reform is both important and limited. We also show that within and between the two main parties, deputies are strongly divided over the direction reform of the electoral system should take, which might helps us understand the difficulty in turning the preferences of the voters and the parties into changes in electoral law.

76Finally, we should stress that it is not only electoral system reform that can help solve the voters’ discontent with the political system. There are several other problems behind voter dissatisfaction, and these can be solved with other type of reforms. For example, parties and politicians are not well connected with voters’ organisations, and politicians —once elected—often forget about their constituencies or their electors, particularly in respect of electoral promises. Thus, the solution to these problems is not only electoral reform, but through making politics more attractive, strengthening the ties with individuals and organisations, and helping bring political institutions to the voters. In this respect, electoral reform is only one aspect towards a resolution of these problems.