Abstract This article studies support for upper chamber reforms in multinational countries. In his seminal work on the topic, Lijphart argues that the upper chamber can be used to accommodate minority groups. However, to date, we still know very little about the perception of minority populations of this institution. We explore this topic building upon an original survey experiment conducted in Scotland and Quebec, where there has been a recent debate about reforming the upper chamber. The results demonstrate that the national minorities in the UK and Canada strongly support a reform of the upper chamber. Also, we find that a reform that would increase the democratic legitimacy of the institution is more popular than one that would foster the regional representation of the minority group, even among respondents who report a strong regional identity. 1. Introduction In multinational countries, upper chambers often serve as instruments of intergroup conflict avoidance. In the first, and subsequent, edition of his seminal book Democracies, Lijphart (1984) strongly advocates for such an institutional patch. He argues that the risk of intergroup conflicts diminishes when all the groups have a strong voice in the upper chamber of the parliament, and when this house has an effective power in the decision-making process. Lijphart’s argument is that, in such upper chambers, politicians from different groups have to talk to each other and reach a consensus for a bill to pass. In particular, upper chambers force majority groups to take into consideration minority groups’ preferences. However, we still do not know whether this particularity of upper chambers is perceived as such by minority groups in multinational countries. The UK and Canada are multinational countries currently enthralled in debates seeking to reform their upper chambers. The House of Lords and the Canadian Senate’s unelected nature has fuelled calls to reform these institutions. However, these countries have regions, respectively Scotland and Quebec, with strong nationalist leanings that continuously strive for greater political representation. We are thus placed in front of two important principles: democratic legitimacy and regional representation. Yet, acquiescing these two principles through political institutions might be challenging. Therefore, if forced to choose, which one would engender greater institutional support in national minorities? To answer this question, we conducted an original survey experiment in Scotland and Quebec in which we asked respondents how they feel about various scenarios regarding their upper chamber. We randomised the type of reform (or status quo) to avoid bias in the responses. We then compared support to the status quo, to an upper chamber reform addressing democratic legitimacy and to a reform that tackles regional representation. Considering that Scotland and Quebec are regions with relatively strong nationalist movements, these two cases present a strong test for the potential influence of democratic legitimacy over that of regional representation. The results demonstrate that (i) minority groups in the UK and Canada strongly support a reform of the upper chamber. Also, they show that (ii) a reform that would increase the democratic legitimacy of the institution is slightly more popular than one that would foster the regional representation of the minority group. The findings also demonstrate that (iii) politically interested individuals are more supportive of reforms. 2. Debates on upper chamber reform in the UK and Canada On 28 July 2015, Lord Sewel—the Chairman of Committees of the House of Lords—resigned from the upper chamber after an incident involving drugs and prostitutes was made public. This sordid affair was a dark chapter for the House of Lords but it was not the only scandal to shake the institution. The large expenses of ‘silent peers’—those who make almost no contribution to debates in the House of Lords—and the appointment of Conservative ‘cronies’—including a former MP tarnished by a Commons’ expense claims scandal—also brought serious criticism to the country’s upper chamber. As a former British colony, Canada derives much of its political system from the UK. This is the case for its bicameral parliament with a non-elected upper chamber. As faith would have it, as the House of Lords was being rocked by scandals, the Canadian Senate was also dealing with serious issues of ethics. Notable amongst them was the trial of Senator Mike Duffy on fraud charges, accusations in which staff at the Prime Minister’s Office were allegedly involved. However, Duffy was only the headliner in a much larger scandal in which the Royal Canadian Mounted Police (national police force) investigated almost two-fifths of sitting senators on alleged illegal expense claims. In both cases, the respective scandals made louder the calls to reform these parliamentary bodies. Issues of democratic legitimacy, as both institutions are unelected, and better representation for the diverse populations of the countries are hallmarks of the desired reforms for the respective upper chambers (Franks, 2003; Smith, 2003; Bochel and Defty, 2012). However, democratic legitimacy and political representation are not the same principle and may not be necessarily addressed in the same manner (Barker, 1990; Lipset, 1959b; Kelso, 2006). Besides a parliamentary chamber often described as anachronistic, both countries also share another characteristic that defines their politics: a significant nationalist movement. Scotland and Quebec are regions in which issues of regional representation loom large in political debates. However, democratic legitimacy cannot be said to be unimportant for Scots and Quebecers. These are two principles of democracy that highly impact citizens’ relationship with the political system and its institutions (Schmitt and Thomassen, 1999). 3. Upper chamber reform: democratic legitimacy and regional representation Political institutions hold a special place for the citizenry as they, notably, help forge political attitudes and behaviours. Yet, this influence can be a negative one. Perceived institutional flaws lead to lesser support for these institutions (Anderson et al., 2005). An important manner of imputing the citizenry’s desires into the political machine is through its participation. Research has shown that participation helps to consolidate support for governmental institutions as well as aiding in the acceptance of unsatisfactory electoral outcomes (Nadeau and Blais, 1993). Hence, popular participation in the democratic process can be beneficial to democratic institutions because it allows for a greater consideration of citizens’ interests (Held, 1987). This is one reason that scholarship has paid considerable attention to the participatory role of the citizenry in institutional frameworks (Dalton et al., 2003). By conferring onto political institutions a widespread acceptance for policies as well as greater popular support, popular participation would therefore provide them with political legitimacy (Manin et al., 1999; Zakaria, 2003). Still, political legitimacy is often specifically linked with one type of participation: elections. This is so due to the belief that political legitimacy can only truly be expressed through voting (Goodwin-Gill, 2006; O'Donnell, 2007). The importance of elections has starred in the debate surrounding the legitimacy of upper chambers. The essential role of elections for upper chambers is highlighted by Lijphart (1999) in arguing that those that are not directly elected lack democratic legitimacy. If democratic legitimacy—or the lack of it—of upper chambers is directly associated with universal suffrage, then most of the polities that use bicameralism are afflicted by this issue as the majority of upper chambers are not directly elected (Russell, 2013). However, democratic legitimacy is not the only critique that might undermine political institutions. The representation of minorities also helps to determine institutional support and acceptance (Anderson et al., 2005; Rothstein, 2009). Yet, minorities cannot simply depend on elections to have their interests expressed or applied (Przeworski, 1991; Rothstein, 2009). Even when their demographic weight provides minorities with a non-negligible political voice, typically when a party that represents their specific interests sends representatives to the national parliament, their interests might still be ignored if no mechanism is in place to warrant regional influence. For example, the 2011 Canadian federal election led to a majority Conservative government, but the party was able to achieve this result by winning only 5 out of Quebec’s 75 seats. A majority government was possible without Quebec, a scenario which until then seemed quite implausible. Thus, not only was the Conservative federal government not dependent on Quebec, the province had very little influence in the cabinet and the party’s caucus; a situation that resulted in arguable many decisions being taken against the interests of Quebec. Such an electoral scenario was also achieved in the UK in 2015. While the Conservatives were able to form a majority government, they did so by winning only one seat in Scotland. Therefore, for minority groups, the principle of democratic legitimacy—deriving from elections—cannot be fully satisfactory. Seeing as political institutions that are perceived to be unfair to minorities can exacerbate social tensions (Birnir, 2007; Karakoç, 2013), alternative standards of democracy may have to be employed (Majone, 1998). Lijphart (1984) strongly advocates for upper chambers to serve a representative role for minority groups and thus diminish the potential of social tensions. Challenging majority rule is actually often an express role of upper chambers (Riker, 1992). According to Russell (2001), upper chambers can be specifically designed to provide political representation to minority groups in multinational countries. In this optic, the political representation of such groups can be assured through nomination, without the need—or desire—for an electoral procedure (Kelso, 2006). Two important principles of democracies therefore potentially compete with each other. Though it is possible to conceive parliamentary institutions to acquiesce both democratic legitimacy and minority representation, achieving satisfaction on both criteria might be a complex ordeal. While some elected upper chambers provide regional representation, it is generally through the concept of equal representation. The upper chambers in, for example, Australia, Switzerland and the USA are directly elected with an equal number of representatives (senators) for each federated state. However, equal representation might be far from the conception of fair regional representation for some minorities. For instance, in Canada, calls for an elected Senate with an equal number of senators for each province have been for decades especially loud in the Western provinces. Yet, Quebec accounts for just under a quarter of the Canadian population and such a reform would thus reduce its regional representation in the Senate and weaken its overall legislative power in the federal parliament. In terms of directly elected upper chambers based on a regional basis, we could only find the example of the upper chamber in Italy (Senato della Repubblica) that accounts for the regions’ share of the national population. But even in this example, proportional regional representation may not satisfy a minority’s desire in terms of regional representation. One could easily imagine the reaction of, for example, Francophones in Belgium or Croats in Bosnia and Herzegovina if proportional regional representation, which would mean a significant loss of representatives for both groups, was proposed for their respective country’s upper chambers. These examples help to illustrate the complexity of acquiescing to both principles of democratic legitimacy and regional representation. Though we accept that such an institutional design is far from impossible, it nevertheless seems quite difficult to achieve an institutional framework to satisfy both of these principles. Thus, if forced to choose, which one of the two principles engenders greater institutional support in national minorities? 4. Scotland and Quebec: similar institutional grievances, divergent secessionist paths Scotland and Quebec offer a tremendous opportunity to explore this question. Both regions have traditionally strived for greater political representation and are in countries currently undergoing important debates regarding the future of their respective upper chambers. Though the previously enumerated scandals have fuelled calls for upper chamber reforms in the UK and Canada, the unelected, anachronistic nature of the British House of Lords and the Canadian Senate has been at the hearth of discussions on reform for decades. In the UK, the combination of Lords Spiritual and Lords Temporal with hereditary and life peerage fails to reflect the modern UK and has led to calls for reform (Russell, 2010; Bochel and Defty, 2012). While the bulk of hereditary peers were removed in a reform in 1999, life peers remain and the House of Lords is thus still seen by many as ‘unreformed’ (Russell, 2010). In Canada, the appointed nature of the Senate has led to its political legitimacy being called into question (Franks, 2003). Although calls for an elected Senate in order to legitimize the upper chamber have been heard for decades, Canada, contrary to the UK, has not managed to reform its upper chamber.1 Yet, the selection of peers and senators has not been the only concern regarding the upper chambers in these countries. The representative role of parliamentary institutions in the two multinational countries has also been of interest, especially for Scotland and Quebec. The former has sought to defend its divergent social interests against England’s dominant political force, whereas the latter has traditionally sought to protect its distinction as a majority Francophone society in an Anglophone-majority Canada (Surridge and McCrone, 1999; Courtney, 2001; Denver, 2002). While regional representation in national politics is of concern for these two regions, it is however their nationalist movements that garner more attention. The constitutional failures of the 1980s and 1990s deepened the political rift between Quebec and the rest of Canada (Russell, 2004).2 The consequence was a dramatic rise in support for independence and a second unsuccessful referendum on secession in 1995. Though the independentist movement came very close of winning this referendum, support for the parties that represent it and the option itself have slowly declined since; sovereignty is no longer the political force it once was (Nadeau and Bélanger, 2013; Durand, 2014). In the last few years, the independentist Scottish National Party (SNP) has been able to win a majority of seats in the Scottish Parliament—even though the electoral system was designed to minimise the risk of an independentist majority (Johns et al., 2013), which led it to hold in 2014 an unsuccessful referendum on secession from the UK—and it came close of sweeping all the seats for the Westminster parliament in the 2015 general election. It would seem that Scotland and Quebec are at different places on the secessionist road. Though independentism in Quebec is less popular than in the past, it is still a structuring factor in provincial and federal politics (Nadeau and Bélanger, 2013; Gauvin et al., 2016). Therefore, in terms of reforming unelected upper chambers, these two cases allow to explore which type of reform would have more of an impact on institutional support in a region with nationalist leanings. 5. Hypotheses The theoretical considerations about democratic legitimacy and regional representation as well as the presentation of the two cases we examine allow to put forward four hypotheses related to upper chamber support among citizens. We focus on the citizenry for two main reasons. Firstly, popular support towards democratic institutions is related to socio-political stability as well as the effectiveness of the institutions (Anderson and Guillory, 1997; Rohrschneider, 2002). Secondly, major reforms to legislative institutions are prone to be put to referendums; this would especially be the case for Canada. Our first hypothesis expects democratic legitimacy to actually have a greater influence. Though it might be expected for regional representation to hold precedence on institutional support in nationalist leaning regions, this principle has been less prominent in the upper chamber debates that we have described than democratic legitimacy. In the UK, the bulk of the reform proposals have stressed elections and the examples of Scottish dissatisfaction towards these reforms seem to be scarce. It is actually difficult to find examples of Scottish concerns—whether expressed by elites or citizens—towards the region’s representation in the House of Lords. As for Quebec, the provincial government actually agreed to an elected and equal—same number of senators for all provinces—Senate in a failed 1992 constitutional accord. Furthermore, though Quebec contested the legality of Ottawa acting alone on Senate reform, it was arguably more in an effort to preserve its effective constitutional veto rather than its share of senators. The emphasis on democratic legitimacy in Scotland and Quebec corresponds to the manner in which electoral democracy has become a norm in Western societies. Even in strong nationalist regions, the narrative has been focused on acquiescing the norm of democracy rather than getting more political power for the region. We therefore expect the principle of democratic legitimacy to be stronger, even in these two nationalist contexts, and propose the following general hypothesis: H1: Support for a democratic reform of the upper chamber is greater than support for a reform to increase the regional representation of the upper chamber. Seeing as Scotland and Quebec are regions with relatively strong nationalist movements, we see these two cases as a strong test for this hypothesis. If the democratic legitimacy principle trumps regional representation in these cases, it would be reasonable to expect that this finding would have widespread applications. Moreover, research has highlighted the importance of inter-individual heterogeneity in the development of political attitudes, including in conceptions of democracy (Sniderman et al., 1991; Miller and Krosnick, 2000; Dalton et al., 2007;). Political interest stands out in the variation of political attitudes between individuals and in their sensitivity to political stimuli as individuals with high levels of political interest have been shown to be more likely to understand political messages (Zaller, 1992). In relation to policy and institutional reforms, high political sophistication is related to abstract political notions, such as democracy and fair representation, whereas less sophisticated individuals tend to focus more on concrete and materialistic issues such as, for example, security and the state of the economy (Lipset, 1959a; Delli Carpini and Keeter, 1996; Inglehart, 1977; Luskin, 2002). Furthermore, research has also demonstrated that individuals with greater political interest also tend to have higher levels of political sophistication (Luskin, 1990). Therefore, it is reasonable to think that citizens who are interested and informed about politics are more sensitive to abstract principles such as legitimacy and representation than those who are not. We thus propose a second hypothesis: H2: Support for upper chamber reform, regardless of its type, is greater among politically interested individuals. We also propose hypotheses that are more specific to the two cases we explore. The Canadian Senate was designed with a stern focus on regional representation; this has not been the case in the UK. Scots, unlike Quebecers, have not had a formal representative mechanism in the House of Lords. Furthermore, as was illustrated, independentism is currently more popular in Scotland than in Quebec; and secessionism has been linked to institutional dissatisfaction (Anderson, 2007). Finally, recent studies have highlighted the important role that regional identity has on support for secession (Burg, 2015; Munoz and Tormos, 2015). The current prominence of the independence debate in Scotland emphasises the distinctiveness of the region compared to the UK and consequently renders the Scottish identity salient. Thus, the greater support for secession in Scotland should also be linked to stronger regional identification compared to Quebec. These facts lead to the proposal of two further hypotheses: H3a: Support for upper chamber reform, regardless of its type, is greater in Scotland than in Quebec H3b: The influence of regional identity on support for upper chamber reform, regardless of its type, is greater in Scotland than in Quebec. 6. Data and results To test these hypotheses, we conducted two online surveys, one in Quebec—from 24 to 30 November 2015—and one in Scotland—from 27 November to 7 December 2015. The samples are composed of 752 adults in Scotland and 770 in Quebec. We used pre-existing survey panels from Survation and Survey Sampling International, respectively for the Scottish and Quebec samples; as well as socio-demographics quotas to ensure their diversity. The questionnaire lasted about five minutes. Table 1 describes the two samples in terms of socio-demographics (age, gender, and education). It shows that in both regions most respondents are between 30 and 64 years old. Also, the proportion of female respondents is 43% for Scotland and 60% for Quebec; whereas the one for respondents with a university degree is 39% for Scotland and 34% for Quebec. Although the samples are not strictly speaking representative, given that we used pre-existing survey panels, they offer enough diversity for us to be confident about the validity of our analysis. Table 1. Description of the sample and key variables Quebec (%) Scotland (%) Socio-demographics <20y old 2.26 1.75 20–29y old 13.83 8.37 30–39y old 22.87 16.19 40–49y old 20.35 22.40 50–64y old 24.47 33.47 >64y old 16.22 17.81 Female 59.71 43.05 University degree 34.31 38.60 Dependent variable Institutional approbation (0–100) 58.03 (24.59) 60.78 (25.49) Independent variables Not interested in politics at all 6.52 5.53 Not interested in politics 8.11 4.72 Neither not interested nor interested 22.87 13.23 Interested in politics 43.75 46.83 Very interested politics 18.75 29.69 Regionalist identity 41.36 54.12 N 752 741 Quebec (%) Scotland (%) Socio-demographics <20y old 2.26 1.75 20–29y old 13.83 8.37 30–39y old 22.87 16.19 40–49y old 20.35 22.40 50–64y old 24.47 33.47 >64y old 16.22 17.81 Female 59.71 43.05 University degree 34.31 38.60 Dependent variable Institutional approbation (0–100) 58.03 (24.59) 60.78 (25.49) Independent variables Not interested in politics at all 6.52 5.53 Not interested in politics 8.11 4.72 Neither not interested nor interested 22.87 13.23 Interested in politics 43.75 46.83 Very interested politics 18.75 29.69 Regionalist identity 41.36 54.12 N 752 741 Note: Entries are proportions for categorical variables, and means and standard deviations (in parentheses) for continuous variables. Table 1. Description of the sample and key variables Quebec (%) Scotland (%) Socio-demographics <20y old 2.26 1.75 20–29y old 13.83 8.37 30–39y old 22.87 16.19 40–49y old 20.35 22.40 50–64y old 24.47 33.47 >64y old 16.22 17.81 Female 59.71 43.05 University degree 34.31 38.60 Dependent variable Institutional approbation (0–100) 58.03 (24.59) 60.78 (25.49) Independent variables Not interested in politics at all 6.52 5.53 Not interested in politics 8.11 4.72 Neither not interested nor interested 22.87 13.23 Interested in politics 43.75 46.83 Very interested politics 18.75 29.69 Regionalist identity 41.36 54.12 N 752 741 Quebec (%) Scotland (%) Socio-demographics <20y old 2.26 1.75 20–29y old 13.83 8.37 30–39y old 22.87 16.19 40–49y old 20.35 22.40 50–64y old 24.47 33.47 >64y old 16.22 17.81 Female 59.71 43.05 University degree 34.31 38.60 Dependent variable Institutional approbation (0–100) 58.03 (24.59) 60.78 (25.49) Independent variables Not interested in politics at all 6.52 5.53 Not interested in politics 8.11 4.72 Neither not interested nor interested 22.87 13.23 Interested in politics 43.75 46.83 Very interested politics 18.75 29.69 Regionalist identity 41.36 54.12 N 752 741 Note: Entries are proportions for categorical variables, and means and standard deviations (in parentheses) for continuous variables. The dependent variables are based on thermometer feeling questions in which we asked respondents how much they like, on a scale from 0 (strongly dislike) to 100 (strongly like), the upper chamber of their country and possible upper chamber reform. The default position of the thermometer was set at 50. In order to make sure that all participants knew what we were describing, we presented them a short text on the upper chamber of their country as a preamble of the question. We randomly assigned each respondent to one of three preambles. In the first one, we presented a text that simply described the current function and composition of the upper chamber in the respondent’s country (status quo group, N in Scotland = 238, N in Quebec = 222). In the second preamble, we added to the description of the status quo a description of a reform proposal consisting in directly electing the upper chamber (democratic reform group, N in Scotland = 257, N in Quebec = 252). In the third preamble, we added to the description of the status quo a description of a reform proposal consisting in ensuring the regional representation of the respondent’s region in the upper chamber (regionalist reform group, N in Scotland = 273, N in Quebec = 251). The exact preambles are in Supplementary Figure S1.3Table 1 shows that the mean and standard deviation of the institutional approbation variable are relatively similar in Quebec and Scotland (mean = 58–61, standard deviation = 25). The skewness is –0.45 and the kurtosis is 2.84. It is important to note that we could have presented the three preambles to the respondents and asked them to report their feeling about each one, instead of randomising them. However, we expected Scottish and Quebec citizens to have some bias against the status quo and in favour of democratic arguments if we presented the three one after the other. Democracy as imposed itself as a norm in Western societies, and the population of the two regions of interest, as previously mentioned, have a negative leaning towards the current form of the state. For example, we expected that respondents would have been reluctant to give a higher score to the ‘non-democratic’ status quo than to democratic legitimacy reform. Our randomisation is thus a strong test of our hypotheses. Figure 1 reports the degree of approval of the status quo and reform proposals in Scotland and Quebec. Firstly, it shows that, in both regions, the mean support for the status quo (47) is much lower than the mean support for the reforms (between 61 and 71). This difference is statistically significant at a level of p < 0.01. Secondly, the mean approval tends to be larger for the democratic legitimacy reform than for the regional representation one, but the difference is small (around 5 points in both regions). Thirdly, the approval ratings are remarkably similar in the two regions. The biggest difference concerns the democratic reform, which is slightly more popular in Scotland than in Quebec (mean of 71 against median of 65). Figure 1 View largeDownload slide Mean institutional approbation Figure 1 View largeDownload slide Mean institutional approbation Furthermore, Figure 1 reveals that Scottish respondents are more supportive of a reform of the upper chamber in general. Their mean approbation score is of 71 for the democracy legitimacy reform and 64 for the regional representation reform. The approbation of both reforms is lower in Quebec: 65 and 61, respectively. In both instances, the difference between two regions are statistically significant at least at a level of p < 0.01. In order to thoroughly test our hypotheses, we estimate an ordinary least squares (OLS) regression model predicting the support for each of the institution/reform presented, in which we add three-way interactions between (i) the randomised treatment (status quo, democratic reform, and regionalist reform), (ii) the respondent’s region (Quebec or Scotland), and (iii) several independent and control variables.4 In line with our hypotheses, we have two key independent variables: reported levels of political interest (in five categories: not interested at all, not interested, neither interested nor not interested, interested, very interested), and regional identity. For this variable, we use the classic subjective identity—also known as the Moreno question—(in five categories: only [country], more [country] than [region], equally both, more [region] than [country], only [region]). We then dichotomise this variable in distinguishing respondents that feel only [region], or more [region] than [country], from the other options. We also add several control variables: age (in six categories), education (in two categories: holding a university degree or not), and gender. The distribution of all the independent variables is presented in Table 1. They are relatively similar across samples, and show enough variations to conduct a meaningful analysis. The full regression table is in the Supplementary Table S2. In order to make the results of this regression with three-way interactions more readable, we created two different figures reporting the predicted values of the institutional support as the randomized treatment, the respondent’s region, and the two key independent variables (political interest and regionalist identity) vary. Figure 2 shows that support for the status quo does not vary with political interest, neither in Quebec, nor in Scotland. However, in both regions, the predicted support for the democratic reform goes from a value below 50 for respondents who are not interested at all in politics to almost 80 for those who are very interested in politics. As for the effect of political interest on support for the regionalist reform, the situation differs between Scotland and Quebec. The variable has a smaller effect in the latter (+10 points), whereas Scots who are very interested in politics are more supportive of the regionalist reform than those who are not at all interested in politics by about 20 points. Figure 2 View largeDownload slide Effect of political interest on institutional approval Figure 2 View largeDownload slide Effect of political interest on institutional approval Figure 3 shows that the regionalist identity of the respondents does not seem to affect institutional support in Quebec. By contrast, in Scotland, those with a regionalist identity are less supportive of the status quo by 10 points, and more supportive of the democratic reform by the same margin. Scots with a regionalist identity are also slightly more supportive of the regionalist reform but the difference between the two predicted values is of only 5 points. We believe that these findings concord with the fact that the House of Lords has not had a formal representative mechanism for Scots, contrary to the Canadian Senate that was conceived to somewhat account for regional distinction in Canada, and therefore regionalist Scots, being more sensitive to such issues, have the most negative feelings towards the status quo. Figure 3 View largeDownload slide Effect of regionalist identity on institutional approval Figure 3 View largeDownload slide Effect of regionalist identity on institutional approval In Table 2, we offer a strict test of our four hypotheses. For each of them, we report the related marginal effects of interest, still based on the same OLS regression, the standard error and the p-value. Testing H1, we see that our regression indicates that the democratic and regional reforms are more supported than the status quo. The difference is of +21 points (democratic reform) and +16 points (regionalist reform). Also, we see that the democratic reform is more popular than the regionalist reform (+4 points). All these differences are statistically significant at a level of p < 0.05. Furthermore, we see that politically interested respondents are more supportive of the upper chambers than those who are not interested in politics. The marginal effect of increasing of one unit in the political interest scale presented above increases support for the democratic reform by 7 points, and support for the regionalist reform by 4 points. These effects seem small, but the political interest scale has 5 categories. As reported in Figure 2, the full effect (i.e. ‘going from not interested in politics at all’ to ‘very interested in politics’) on support for the democratic reform is larger than 30 points in Quebec and 20 points in Scotland. What is more, both effects are statistically significant at a level of p < 0.01. We thus find strong evidence for H1 and H2 in our regression. Table 2. Test of hypotheses Hypothesis Test OLS Logit H1 Support for democratic reform v/s status-quo +20.83*** (1.59) +32.34*** (3.35) Support for regionalist reform v/s status-quo +15.92*** (1.58) +24.71*** (3.35) Support for democratic reform v/s support for regionalist reform +4.91*** (1.43) +8.63** (3.16) H2 Effect of political interest on support for democratic reform +7.20*** (.99) +11.30*** (2.00) Effect of political interest on support for regionalist reform +3.54*** (0.99) +9.15*** (2.15) H3a Support for democratic reform in Scotland v/s Quebec +1.63 (2.14) +3.68 (4.49) Support for regionalist reform in Scotland v/s Quebec +1.05 (2.05) +3.40 (4.44) H3b Effect of regionalist identity on support for democratic reform in Scotland v/s Quebec +7.38* (4.10) +8.70 (7.58) Effect of regionalist identity on support for regionalist reform in Scotland v/s Quebec +2.88 (4.00) +9.02 (8.15) Hypothesis Test OLS Logit H1 Support for democratic reform v/s status-quo +20.83*** (1.59) +32.34*** (3.35) Support for regionalist reform v/s status-quo +15.92*** (1.58) +24.71*** (3.35) Support for democratic reform v/s support for regionalist reform +4.91*** (1.43) +8.63** (3.16) H2 Effect of political interest on support for democratic reform +7.20*** (.99) +11.30*** (2.00) Effect of political interest on support for regionalist reform +3.54*** (0.99) +9.15*** (2.15) H3a Support for democratic reform in Scotland v/s Quebec +1.63 (2.14) +3.68 (4.49) Support for regionalist reform in Scotland v/s Quebec +1.05 (2.05) +3.40 (4.44) H3b Effect of regionalist identity on support for democratic reform in Scotland v/s Quebec +7.38* (4.10) +8.70 (7.58) Effect of regionalist identity on support for regionalist reform in Scotland v/s Quebec +2.88 (4.00) +9.02 (8.15) Note: Entries are marginal effects computed from the OLS and Logit regressions presented in the Table S2 in the Supplementary Material. For the OLS regression, the marginal effects are for changes in the value of the thermometer feeling (0–100). For the logit regression, they are for changes in the probability of entering into the approval category of the thermometer feeling, in percentage-points. Standard errors are in parentheses. * p < 0.1, ** p < 0.05, *** p < 0.01 Table 2. Test of hypotheses Hypothesis Test OLS Logit H1 Support for democratic reform v/s status-quo +20.83*** (1.59) +32.34*** (3.35) Support for regionalist reform v/s status-quo +15.92*** (1.58) +24.71*** (3.35) Support for democratic reform v/s support for regionalist reform +4.91*** (1.43) +8.63** (3.16) H2 Effect of political interest on support for democratic reform +7.20*** (.99) +11.30*** (2.00) Effect of political interest on support for regionalist reform +3.54*** (0.99) +9.15*** (2.15) H3a Support for democratic reform in Scotland v/s Quebec +1.63 (2.14) +3.68 (4.49) Support for regionalist reform in Scotland v/s Quebec +1.05 (2.05) +3.40 (4.44) H3b Effect of regionalist identity on support for democratic reform in Scotland v/s Quebec +7.38* (4.10) +8.70 (7.58) Effect of regionalist identity on support for regionalist reform in Scotland v/s Quebec +2.88 (4.00) +9.02 (8.15) Hypothesis Test OLS Logit H1 Support for democratic reform v/s status-quo +20.83*** (1.59) +32.34*** (3.35) Support for regionalist reform v/s status-quo +15.92*** (1.58) +24.71*** (3.35) Support for democratic reform v/s support for regionalist reform +4.91*** (1.43) +8.63** (3.16) H2 Effect of political interest on support for democratic reform +7.20*** (.99) +11.30*** (2.00) Effect of political interest on support for regionalist reform +3.54*** (0.99) +9.15*** (2.15) H3a Support for democratic reform in Scotland v/s Quebec +1.63 (2.14) +3.68 (4.49) Support for regionalist reform in Scotland v/s Quebec +1.05 (2.05) +3.40 (4.44) H3b Effect of regionalist identity on support for democratic reform in Scotland v/s Quebec +7.38* (4.10) +8.70 (7.58) Effect of regionalist identity on support for regionalist reform in Scotland v/s Quebec +2.88 (4.00) +9.02 (8.15) Note: Entries are marginal effects computed from the OLS and Logit regressions presented in the Table S2 in the Supplementary Material. For the OLS regression, the marginal effects are for changes in the value of the thermometer feeling (0–100). For the logit regression, they are for changes in the probability of entering into the approval category of the thermometer feeling, in percentage-points. Standard errors are in parentheses. * p < 0.1, ** p < 0.05, *** p < 0.01 However, we find little evidence of H3a and H3b. Firstly, the democratic and regionalist reforms are only slightly more supported in Scotland than in Quebec (+2 and +1 point, respectively). The difference is not statistically significant at a level of p < 0.1. Secondly, the variable regionalist identity only has a slightly larger effect on support for reforms in Scotland than in Quebec. This difference is of +7 points for the democratic reform (p < 0.01) and of +3 points for the regionalist reform (not statistically significant). There is thus little evidence to support our hypotheses H3a and H3b. As a robustness test, we re-estimate the models in changing the dependent variable. Instead of a continuous dependent variable, we create a dummy dependent variable differentiating respondents who express some support for the status quo/reform (approval of more than 50) versus the others (approval lower than 50). By default, the thermometer was set at 50. The analysis is thus a logit regression. The reason why we include this robustness test is that we envision the possibility that in thermometer feeling questions—like the one we use to measure our dependent variable—the precise value might not always be meaningful. Some of the respondents might be sloppy in their manipulation of the cursor that appears on their screen. Therefore, we separate respondents between those pushing the cursor to the right (approving the institution) and those pushing it to the left (opposing the institution). Table 2 also reports the marginal effects of interest to test our hypotheses as derived by the logit regression. The results are remarkably similar for all four of our hypotheses, which increases the trust we have in our findings. The full results of this regression are also reported in the Supplementary Table S2. Finally, it is worth adding a few words about the results regarding the control variables. Firstly, in Quebec, only age appears as a significant predictor of the status quo: the older people are, the less they support the status quo. While based on past findings we would expect older individuals to be more risk-averse and supportive of the status quo (Pålsson, 1996; Duchesne et al., 2003), our result is contrary to this expectation. The somewhat counterintuitive finding might have to do with the fact that the issue of Senate reform has been relatively salient for decades in Canada; and thus older Quebecers might have grown less accepting of the continued status quo. Seeing as the debate over House of Lords reform is a relatively recent one, the results show no age difference in Scotland. Secondly, the results point to a gender difference: women are significantly more negative towards the regional representation reform than men. Furthermore, this is the only significant gender difference. We do not believe that the classic gender difference in terms of risk-aversion, specifically towards nationalists politics would account for this gender difference (Johns et al., 2012; Verge et al., 2015). We are thus not certain why female Quebecers are more negative towards regional representation than their male counterparts. Overall, the results of our analyses demonstrate that a dissatisfaction with the current upper chambers in the UK and Canada amongst Scots and Quebecers. It is thus not surprising that the two reforms presented to the respondents garner more support than the status quo. Our findings also underline that acquiescing the principle of democratic legitimacy leads to greater support than proposing a reform to increase regional representation. While this might be a surprising result to some, considering the two cases with strong regional nationalist leanings, we believe that it testifies to the influence of the democratic norm in Western society. 7. Conclusion Lijphart (1984) argues that the upper chamber in multinational countries can be designed to accommodate minority groups, and thus assure a representative role for minorities. Institutional accommodation towards minorities tend to place great importance on the principle of representation. However, research had not attempted to explore if representation was in fact more important for minority group members than other important democratic principles. The present paper took up this challenge through an original survey experiment conducted in Scotland and Quebec that explores support for upper chamber reforms in multinational countries. The scandals that have recently rocked the British House of Lords and the Canadian Senate have made louder the calls to reform these important parliamentary institutions. Yet, the seemingly anachronistic character of these upper chambers has been a contentious issue in both countries well before the latest indignities. At the heart of reform discussions has been the issue of political legitimacy caused by Lords in the UK and Senators in Canada being appointed rather than elected. However, the UK and Canada are diverse multinational countries. Both have to deal with significant nationalist movements from, respectively, Scotland and Quebec. For these regions, regional representation in the national political institutions has been of concern. These are therefore two cases in which two democratic principles—democratic legitimacy and regional representation—are competing for influence in an environment in which upper chamber reform is quite salient. Three findings are derived from the results of the present study. Firstly, we observe that upper chamber reforms are largely supported by Scots and Quebecers. The difference of support compared to the status quo is significant and substantial. Secondly, Scots and Quebecers tend to show greater support for a democratic legitimacy reform of the upper chamber than for a regional representation reform. We consider our study as a strong test in this respect as the political landscape of Scotland and Quebec has been structured along issues that emphasize regional political representation. Yet, even in these two cases, support for a democratic legitimacy reform trumps that of regional representation. While our study specifically sought to gauge the preference of regional minorities for institutional reforms that respond to democratic legitimacy or reforms that increase regional representation, the findings can also be of interest in another manner. Institutional arrangements can at once acquiesce democratic legitimacy and regional representation. Though, as we have presented, the examples of elected upper chambers that also satisfy regional representation are limited, democratic legitimacy and regional representation can both be characteristics of an upper chamber; these are not mutually exclusive principles. In the case that a proposed institutional reform seeks to acquiesce both democratic legitimacy and regional representation, our results demonstrate that it would be more rewarding to sell such a reform to the population by focusing on the democratic legitimacy aspect. In other words, by framing such an institutional change as a democratic legitimacy reform rather than one which increases regional representation. However, our study does not permit to test the influence of a reform that tackles both democratic legitimacy and regional representation, but we can presume that highlighting both principles would be a powerful communication strategy. Thirdly, the role of political interest is also highlighted. As it was expected, politically interested individuals are more supportive of upper chamber reform in general. Seeing as political interest is intertwined with political sophistication (Luskin, 1990), individuals with higher levels of political interest can more easily understand the arguments, and their benefits, in favour of the reforms. Yet, even among individuals with high political interests in nationalist regions, support for democratic legitimacy reform tend to be greater than support for regional representation reform. One reason for the persistence of democratic legitimacy to trump regional representation in our results might be that regional political voice is not solely reliant on the national parliament. In our two cases, strong regional assemblies serve as bulwarks to regional issues and render representation at the national-level less important (Lodge et al., 2004; Taucar, 2004). We also speculate on a second potential reason for this result. Democratic legitimacy would permit the entry of independentist politicians into the upper chambers of both countries. Seeing as institutional designs permit, or limit, citizens’ democratic preferences to be registered (Elhauge et al., 1997), democratically legitimate upper chambers in the UK and Canada would permit a greater political voice for independentism in Scotland and Quebec. Currently there is no British Lord nor Canadian Senator that is from an independentist party. Thus, there is an absence of an independentist voice in the upper chambers of the two countries, an omission that greater regional representation by itself would not be guaranteed to correct. But, ultimately, exploring nationalist cases in which there is an absence of an effective regional parliament would provide greater insight into the influence of both principles. The results of this study underscore the importance of democratic legitimacy on democratic approval. Even in societies that give great importance to regional representation, legitimacy still stands out as a force of institutional support. Though distinct in many ways, in terms of democratic principles, Scotland and Quebec appear to be less so. However, we also observe that minority groups in multinational societies are largely favourable to a reform towards an elected upper chamber. Supplementary material Supplementary data for this article are available at Parliamentary Affairs online. Funding This article was made possible by a grant from the Centre for the Study of Democratic Citizenship. Conflict of Interest The authors report no conflicts of interest. Footnotes 1 Though many years—and much political capital—have been spent on debating Senate reform in Canada, no real changes have been implemented due arguably to the nature of the constitutional amendments needed: approval from both chambers of the federal parliament and all 10 provincial legislatures. 2 Though the 1980 referendum on Quebec independence was unsuccessful for the independentist side, it forced a re-exploration of Quebec’s place in Canada. The result was a new constitution in 1982, but one that the Quebec provincial government has never agreed to sign. This constitutional dissatisfaction led to a decade of major constitutional debates aimed at integrating Quebec into the constitutional fold and resulted in two separate reform packages: the Meech Lake Accord in 1987 and the Charlottetown Accord in 1992. Senate reforms that would provide powers to provinces in the selection of Senators were a hallmark of both proposals. However, both reforms ultimately failed to be implemented. 3 The Quebec respondents were presented with French texts whereas those in Scotland received texts in English. In the Supplementary Table S1, we show a balance test, in which we estimate the association between the randomised treatment and the other independent variables. The results reveal that there is no statistically significant association between them (at a level of p < 0.1), suggesting that the randomisation was conducted properly. 4 We conducted several diagnostics tests for our main OLS regression. Firstly, we conducted a Breush Pagan test. The results indicate the presence of heteroscedasticity (χ2 = 9.14, p < 0.01). We thus use robust standard errors in our analysis. Secondly, we constructed a scatter plot of the residuals against the fitted values. The plot does not reveal any particular non-linear trend in the data (see Supplementary Figure S2). Thirdly, we conducted a Ramsey RESET test. The results indicate that the conditional meet assumption might be violated and that there might be some omitted variables (χ2 = 3.41, p < 0.05). We are aware of this limit, that is unfortunately very common in social sciences. It is important to note that one of our key independent variables is a randomised treatment, which partially addresses this issue. References Anderson C. J. , Blais A. , Bowler S. , Donovan T. , Listhaug O. ( 2005 ). Losers' Consent: Elections and Democratic Legitimacy: Elections and Democratic Legitimacy , Oxford , Oxford University Press . Google Scholar CrossRef Search ADS Anderson C. J. , Guillory C. A. ( 1997 ) ‘Political Institutions and Satisfaction with Democracy: A Cross-National Analysis of Consensus and Majoritarian Systems’ , American Political Science Review , 91 , 66 – 81 . Google Scholar CrossRef Search ADS Anderson L. ( 2007 ) ‘Federalism and Secessionism: Institutional Influences on Nationalist Politics in Quebec’ , Nationalism and Ethnic Politics , 13 , 187 – 211 . Google Scholar CrossRef Search ADS Barker R. ( 1990 ) Political Legitimacy and the State , Oxford , Oxford University Press . Google Scholar CrossRef Search ADS Birnir J. K. ( 2007 ) Ethnicity and Electoral Politics , Cambridge , Cambridge University Press . Bochel H. , Defty A. ( 2012 ) ‘A More Representative Chamber’: Representation and the House of Lords’ , Journal of Legislative Studies , 18 , 82 – 97 . Google Scholar CrossRef Search ADS Burg S. L. ( 2015 ) ‘Identity, Grievances, and Popular Mobilization for Independence in Catalonia’ , Nationalism and Ethnic Politics , 21 , 289 – 312 . Google Scholar CrossRef Search ADS Courtney J. C. ( 2001 ) Commissioned Ridings: Designing Canada's Electoral Districts , Montreal , McGill-Queen's University Press . Dalton R. J. , Scarrow S. E. , Cain B. E. ( 2003 ) ‘New Forms of Democracy? Reform and Transformation of Democratic Institutions’. In Cain B. E. , Dalton R. J. , Scarrow S. E. (eds), Democracy Transformed? Expanding Political Opportunities in Advanced Industrial Democracies , Oxford , Oxford Univesity Press . Google Scholar CrossRef Search ADS Dalton R. J. , Shin D. C. , Jou W. ( 2007 ) ‘Understanding Democracy: Data from Unlikely Places’ , Journal of Democracy , 18 , 142 – 156 . Delli Carpini M. X. , Keeter S. ( 1996 ) What Americans Know about Politics and Why It Matters , New Haven , Yale University Press . Denver D. ( 2002 ) ‘Voting in the 1997 Scottish and Welsh Devolution Referendums: Information, Interests and Opinions’ , European Journal of Political Research , 41 , 827 – 843 . Google Scholar CrossRef Search ADS Duchesne E. , Eagles M. , Erfle S. ( 2003 ) ‘Constituency Homogeneity, Economic Risk and Support for Quebec Sovereignty: A Research Note’ , Canadian Journal of Political Science , 36 , 643 – 656 . Google Scholar CrossRef Search ADS Durand C. ( 2014 ) ‘Générations, Souveraineté, Parti Québécois’, accessed at http://ahlessondages.blogspot.ca/2014/06/generations-souverainete-parti-quebecois.html on 8 August 2014. Elhauge E. , Lott J. R. Jr , Manning R. L. ( 1997 ) ‘How Term Limits Enhance the Expression of Democratic Preferences’ , Supreme Court Economic Review , 5 , 59 – 80 . Google Scholar CrossRef Search ADS Franks C. ( 2003 ) ‘The Canadian Senate in Modern Times’. In Joyal S. (ed.), Protecting Canadian Democracy: The Senate You Never Knew , Montreal , Mcgill-Queen's University Press . Gauvin J.-P. , Chhim C. , Medeiros M. ( 2016 ) ‘Did They Mind the Gap? Voter/Party Ideological Proximity between the BQ, the NDP and Quebec Voters, 2006-2011’ , Canadian Journal of Political Science , 49 , 289 – 310 . Google Scholar CrossRef Search ADS Goodwin-Gill G. S. ( 2006 ) Free and Fair Elections , Geneva , Inter-Parliamentary Union . Held D. ( 1987 ) Models of Participation , Cambridge , Polity Press . Inglehart R. ( 1977 ) The Silent Revolution: Changing Values and Political Styles among Western Publics , Princeton, NJ , Princeton University Press . Johns R. , Bennie L. , Mitchell J. ( 2012 ) ‘Gendered Nationalism: The Gender Gap in Support for the Scottish National Party’ , Party Politics , 18 , 581 – 601 . Google Scholar CrossRef Search ADS Johns R. , Mitchell J. , Carman C. J. ( 2013 ) ‘Constitution or Competence? the SNP's Re-Election in 2011’ , Political Studies , 61 , 158 – 178 . Google Scholar CrossRef Search ADS Karakoç E. ( 2013 ) ‘Ethnicity and Trust in National and International Institutions: Kurdish Attitudes toward Political Institutions in Turkey’ , Turkish Studies , 14 , 92 – 114 . Google Scholar CrossRef Search ADS Kelso A. ( 2006 ) ‘Reforming the House of Lords: Navigating Representation, Democracy and Legitimacy at Westminster’ , Parliamentary Affairs , 59 , 563 – 581 . Google Scholar CrossRef Search ADS Lijphart A. ( 1984 ) Democracies: Patterns of Majoritarian and Consensus Government in Twenty-One Countries , New Haven , Yale University Press . Google Scholar CrossRef Search ADS Lijphart A. ( 1999 ) Patterns of Democracy: Government Forms and Performance in Thirty-Six Democracies , New Haven , Yale University Press . Lipset S. M. ( 1959a ) ‘Democracy and Working-Class Authoritarianism’ , American Sociological Review , 24 , 482 – 501 . Google Scholar CrossRef Search ADS Lipset S. M. ( 1959b ) ‘Some Social Requisites of Democracy: Economic Development and Political Legitimacy’ , American Political Science Review , 53 , 69 – 105 . Google Scholar CrossRef Search ADS Lodge G. , Russell M. , Gay O. ( 2004 ) ‘The Impact of Devolution on Westminster: If Not Now, When?’. In Trench A. (ed.), What Difference Has Devolution Made , Exeter, Imprint Academic . Luskin R. C. ( 1990 ) ‘Explaining Political Sophistication’ , Political Behavior , 12 , 331 – 361 . Google Scholar CrossRef Search ADS Luskin R. C. ( 2002 ) ‘Political Psychology, Political Behavior, and Politics: Questions of Aggregation, Causal Distance, and Taste’. In Kuklinski J. H. (ed.), Thinking about Political Psychology , Cambridge , Cambridge University Press . Majone G. ( 1998 ) ‘Europe’s “Democratic Deficit”: The Question of Standards’ , European Law Journal , 4 , 5 – 28 . Google Scholar CrossRef Search ADS Manin B. , Przeworski A. , Stokes S. ( 1999 ) ‘Elections and Representation’. In Przeworski A. , Stokes S. , Manin B. (eds), Democracy, Accountability, and Representation , Cambridge , Cambridge University Press , pp. 29 – 54 . Google Scholar CrossRef Search ADS Miller J. M. , Krosnick J. A. ( 2000 ) ‘News Media Impact on the Ingredients of Presidential Evaluations: Politically Knowledgeable Citizens Are Guided by a Trusted Source’ , American Journal of Political Science , 44 , 301 – 315 . Google Scholar CrossRef Search ADS Munoz J. , Tormos R. ( 2015 ) ‘Economic Expectations and Support for Secession in Catalonia: Between Causality and Rationalization’ , European Political Science Review , 7 , 315 – 341 . Google Scholar CrossRef Search ADS Nadeau R. , Bélanger É. ( 2013 ) ‘Un modèle général d'explication du vote des Québécois’. In Bastien F. , Bélanger E. , Gélineau F. (eds), Les Québécois aux urnes: les partis, les médias et les citoyens en campagne , Montreal , Les presses de l'Université de Montréal . Nadeau R. , Blais A. ( 1993 ) ‘Accepting the Election Outcome: The Effect of Participation on Losers' Consent’ , British Journal of Political Science , 23 , 553 – 563 . Google Scholar CrossRef Search ADS O'Donnell G. A. ( 2007 ) ‘The Perpetual Crises of Democracy’ , Journal of Democracy , 18 , 5 – 11 . Google Scholar CrossRef Search ADS Pålsson A.-M. ( 1996 ) ‘Does the Degree of Relative Risk Aversion Vary with Household Characteristics?’ , Journal of Economic Psychology , 17 , 771 – 787 . Google Scholar CrossRef Search ADS Przeworski A. ( 1991 ) Democracy and the Market: Political and Economic Reforms in Eastern Europe and Latin America , Cambridge , Cambridge University Press . Google Scholar CrossRef Search ADS Riker W. H. ( 1992 ) ‘The Justification of Bicameralism’ , International Political Science Review , 13 , 101 – 116 . Google Scholar CrossRef Search ADS Rohrschneider R. ( 2002 ) ‘The Democracy Deficit and Mass Support for an EU-Wide Government’ , American Journal of Political Science , 46 , 463 – 475 . Google Scholar CrossRef Search ADS Rothstein B. ( 2009 ) ‘Creating Political Legitimacy Electoral Democracy versus Quality of Government’ , American Behavioral Scientist , 53 , 311 – 330 . Google Scholar CrossRef Search ADS Russell M. ( 2001 ) ‘What Are Second Chambers for?’ , Parliamentary Affairs , 54 , 442 – 458 . Google Scholar CrossRef Search ADS Russell M. ( 2010 ) ‘A Stronger Second Chamber? Assessing the Impact of House of Lords Reform in 1999 and the Lessons for Bicameralism’ , Political Studies , 58 , 866 – 885 . Google Scholar CrossRef Search ADS Russell M. ( 2013 ) ‘Rethinking Bicameral Strength: A Three-Dimensional Approach’ , Journal of Legislative Studies , 19 , 370 – 391 . Google Scholar CrossRef Search ADS Russell P. H. ( 2004 ) Constitutional Odyssey: Can Canadians Become a Sovereign People ? Toronto , University of Toronto Press . Schmitt H. , Thomassen J. J. A. ( 1999 ) Political Representation and Legitimacy in the European Union , Oxford , Oxford University Press . Google Scholar CrossRef Search ADS Smith D. E. ( 2003 ) The Canadian Senate in Bicameral Perspective , Toronto , University of Toronto Press . Google Scholar CrossRef Search ADS Sniderman P. M. , Brody R. A. , Tetlock P. E. ( 1991 ) Reasoning and Choice: Explorations in Political Psychology , Cambridge , Cambridge University Press . Google Scholar CrossRef Search ADS Surridge P. , McCrone D. ( 1999 ) ‘The 1997 Scottish Referendum Vote’. In Taylor B. , Thomson K. (eds), Scotland and Wales: Nations Again , Cardiff , University of Wales Press , pp. 41 – 64 . Taucar C. E. ( 2004 ) Canadian Federalism and Quebec Sovereignty , New York , Peter Lang . Verge T. , Guinjoan M. , Rodon T. ( 2015 ) ‘Risk Aversion, Gender, and Constitutional Change’ , Politics & Gender , 11 , 499 – 521 . Google Scholar CrossRef Search ADS Zakaria F. ( 2003 ) The Future of Freedom: Illiberal Democracy at Home and Abroad , New York , W.W. Norton . Zaller J. ( 1992 ) The Nature and Origins of Mass Opinion , Cambridge , Cambridge University Press . Google Scholar CrossRef Search ADS © The Author(s) 2018. Published by Oxford University Press on behalf of the Hansard Society; all rights reserved. For Permissions, please e-mail: email@example.com
Parliamentary Affairs – Oxford University Press
Published: Feb 9, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera