Owning the National Question in Quebec
Richard Nadeau
Department of Political Science
Université de Montréal
Paper prepared for the annual conference of Elections, Public Opinion and Parties
(EPOP), Edinburg, Scotland, September 12th-14th.
2
This paper seeks to determine to what extent the “Quebec question” is at the heart of
political dynamics in Quebec. By this, we mean the debates surrounding not only the
political status of the province in the Canadian federation, but also surrounding the
protection and promotion of the identity and culture of the province’s Francophone
majority population. By “political dynamics”, we mean the link between Quebecers’
opinions on the future of their province and the protection of their culture and support for
one of the four big political parties in Quebec.
There has been much work done on the relation between Quebecers’ constitutional
preferences and the political choices they have made over the past few decades (see
Bélanger and Nadeau 2009a for an overview). We intend to contribute to this body of
academic literature by pushing theoretical reflection and empirical analysis a little farther
with regard to the meaning of the “national question” in Quebec and its impact on
partisan dynamics within the province. First, we will present our theoretical framework.
Then, we will examine the link between Quebecers’ constitutional preferences and their
vote choice during the April 7 2014 elections. The results of these analyses clearly show
that the national question is very much at the heart of the political game in Quebec and
that all parties must position themselves with regard to this issue if they wish to be
electorally competitive in the province.
The National Question in Quebec
The first question is to establish whether the national question in Quebec, in all of its
forms, constitutes a niche issue for just one party in particular or a central issue that
3
structures the party system. An essential distinction has to be made from this perspective
concerning the nature of this issue in federal and provincial elections in Quebec. In the
case of federal elections, this issue clearly has the characteristics of a niche issue
(Bélanger and Nadeau 2006, 2009b). One political party, the Bloc québécois, has made
defending the interests of Quebec and promoting the province’s political independence its
central (if not only) goal. The slogans used by this party, which has never run candidates
outside of Quebec, are in and of themselves revealing. In 1993, when it first contested
federal elections, the Bloc québécois asked voters to give themselves “true power” (le
vrai pouvoir). A decade later, thanks to a whimsical play on words in French, this same
party ran in an election dominated by the issue of corruption under the slogan of “a parti
propre au Québec”.
The political positioning of the other parties on the federal scene is revealing. The
Conservative Party, the Liberal Party, the New Democratic Party, and the Green Party all
present themselves as “national” parties, as opposed to the Bloc québécois, which
presents itself as a regional party. The statewide parties field candidates in all provinces
including Quebec and also present themselves as parties with a comprehensive political
platform. This includes, most notably, relations between the federal government and the
provinces.
The parallels between Quebec and the prevailing situation in the United Kingdom are
quite clear. Like the Bloc québécois, the Scottish National Party has also used the
Scottish “national question” as a niche issue during national elections. It has only
4
presented candidates in Scotland and its role has essentially consisted of making
Scotland’s concerns known at Westminster. In this case, as with the Lega Nord in Italy,
the national question, or even the interests of a particular region within a country, is
exploited as a niche issue that allows a party to occupy a specific niche and win votes by
defining itself as the party that defends the interests of a particular nation or region.
The political dynamics play out differently at the sub-national level, where questions of
regional identity and interests are much more important. In these regions, this question,
far from a mere niche issue, often becomes the fundamental political cleavage around
which all political parties position themselves. In moving from the national to the
regional level, this issue evolves from being niche to one of the mainstream pillars of
electoral politics; on the regional level, it is an inevitable question and all parties must
have an official stance on it (Bélanger and Nadeau 2009a; Bechhoffer and McCrone
2010; Hepburn 2014). This situation is clear in the case of Quebec. The Quebec question
is clearly a niche issue when it is debated on the national scene, but at the same time is
also the dominant cleavage in Quebec provincial elections.
A second key question is to ask if this issue has the characteristics of a “valence issue”,
that is, an issue around which public opinion is massively in agreement (Stokes 1963;
Stokes and Dilulio 1993). The question of the economy is often presented as the valence
issue par excellence in that all (or almost all) voters can agree on lowering
unemployment and maintaining price stability. Can we say in the same vein that the
national question in Quebec is a valence issue? In the eyes of Quebec voters, do all
5
provincial political parties have to defend the region’s interests, identity, and language?
Are all provincial parties judged with regard to whether they are working towards these
goals or their ability to do so? Past research has shown that parties that want to
differentiate themselves on a valence issue must show that they are the most determined
and competent party to reach these goals (Bélanger and Meguid 2008; Walgrave,
Lefevere and Tresch 2012). Is this useful to understand Quebec public opinion and the
factors that are the most important in Quebec voters’ political choices?
The answer to this question is complex. In order to understand the answer, we must first
turn to work that has shown that the issue of the “national question” consists of many
dimensions. One of these dimensions has to do with the idea that any government in
power in the regional capital, be it Québec City or Edinburgh, must vigorously defend the
interests of this region. This defense can take several forms. It can consist of simply
defending the prerogatives of the regional parliament against any encroachment by the
national parliament. It can also mean ensuring that the region obtains its fair share of the
national wealth through equalization payments or other mechanisms of fiscal federalism.
Defending regional interests can also mean demanding that some powers be transferred
from the central government to the regional government in order to allow regional-level
elected representatives to better respond to the needs of their population. Finally, it can
consist of defending the region’s distinct economic and social model. It is along this
objective of defending regional interests that the national question truly takes on the
characteristics of a valence issue. All regional-level political parties must have a clear
stance on this question and show that they will vehemently and effectively defend the
6
interests of their region (Nadeau et al 2000; Bélanger and Meguid 2008; Bélanger and
Nadeau 2009b).
A second dimension to the issue of the national question has to do with promoting the
specific identity of a particular region. This question is particularly relevant in Quebec,
due to the precariousness of the French language on the North American continent (Blais,
Martin and Nadeau 1995; Nadeau and Fleury 1995). That Quebecers believe they have a
distinct identity that should be protected and promoted by their regional government is
largely undisputed (Blais and Nadeau 1992; Martin and Nadeau 2001; Bélanger and
Perrella 2008). Thus, the question of protecting and promoting Quebecer identity is
largely consensual and all parties must subscribe to this objective. That said, political
parties in Quebec, while supporting this general goal, can sometimes disagree on the
means necessary to arrive at these ends. This was particularly apparent during past
debates over language laws in Quebec and more recently during the debate over the
adoption of the Quebec Charter of Values, a proposal which would have affirmed the
secular nature of Quebec society.
If it seems that there is a consensus among all parties in the province regarding the
defense of Quebec’s interests and identity, this is not the case for promoting different
constitutional options for Quebec. One could even say that this issue has the
characteristics of a “positional” issue. The case of Quebec is very telling in this regard.
For the past forty years, Quebecers have been divided along two constitutional options.
The first option is political independence and is spearheaded by the Parti québécois, but
7
also endorsed by Québec solidaire. The second option is keeping Quebec in the Canadian
federation and the Quebec Liberal Party has been the most outspoken for this option (the
fourth most important party in Quebec, the Coalition Avenir Québec also supports a
federalist position, but is more nationalist than the Quebec Liberal Party).
The opposition between federalists and sovereignists (or independentists) is the dominant
cleavage in Quebec political life. All political parties in the province must have a clear
position on this question (i.e. federalist or sovereignist) and those who decide to take a
middle-of-the-road position are often accused of either eluding the question or being
confused about their own stance. This dividing line with regard to constitutional options
is often more important than the line between left and right (Nadeau, Guérin and Martin
1995). From this point of view, the dynamics of Quebec politics are unique in the
Canadian federation.
Hypotheses
What can we conclude from this brief overview of the role that the national question
plays on electoral dynamics in Quebec? The first conclusion is that this question is a
“super issue” in Quebec. More so than left-right divides or partisan identification, it is the
main factor that structures political dynamics in Quebec (Bélanger and Nadeau 2009a).
The second conclusion is that the national question is in and of itself multidimensional,
having both “valence” and “positional” characteristics. Defending regional interests and
affirming a distinct identity are dimensions around which there largely exists a consensus
and political parties differentiate themselves not by taking radically different positions,
8
but rather by showing a more or less levels of engagement and determination to achieve
these objectives. The logic that prevails in these two cases brings in issues of spatial
voting. As such, the Quebec Liberal Party, which is the most federalist party in Quebec,
will declare its willingness to defend Quebec and its culture, but will generally do so with
less vigour than the other parties (and less than the Parti québécois, in particular). Thus,
each party should subscribe to the consensus surrounding these questions without
straying too far from the position of the median voter in Quebec, who has nationalist
tendencies. Thus, the political game in Quebec on these dimensions of the national
question is clear. It is in the strategic interest of the most nationalist party in Quebec that
questions of identity and language remain at the forefront during an election and it is in
the interest of the other parties that these same concerns take up less space on the political
agenda.
The political dynamic is entirely different for the third dimension of the national
question: constitutional preferences. For the past few years, Quebec public opinion has
been on the side of the federalists. It is for this reason that the Quebec Liberal Party,
otherwise discreet on the question of defending Quebec’s interests and identity, has put
constitutional issues at the heart of its electoral campaigns. It is also for this same reason
that the Parti québécois, who often brandishes defending Quebec’s interests and identity
during electoral campaigns, has said relatively little about sovereignty. The other parties,
less engaged in the opposition between federalism and sovereignty, would benefit from
this question being removed from the political debate all together and voters’ attention
being drawn to other issues.
9
The preceding section allows us to suggest some hypotheses that will be tested using data
from an online survey carried out in Quebec in the weeks before the April 7, 2014
elections that brought the Quebec Liberal Party into power (details about the survey are
in the Appendix). The hypotheses are the following:
1. The most nationalist party in Quebec, the Parti québécois, is perceived as being the
most adept at defending Quebec’s interests, language, and culture. This advantage is due
to the fact that the majority of the population has nationalist sensibilities and will be
especially marked among Francophone voters.
2. The party most opposed to Quebec independence, the Quebec Liberal Party, will
dominate the constitutional preferences dimension of the national question. This
advantage is due to the fact that the majority of the population favours keeping the
province within the Canadian federation.
3. Support for parties dominating one of the dimensions of the national question, the Parti
québécois or the Quebec Liberal Party, is linked to this issue more so than it is support
for any other party. Thus, support for the Parti québécois and the Quebec Liberal Party is
tied to a) the importance voters themselves give to the national question, b) their feelings
of attachment and identification towards Quebec and Canada, c) their evaluations of the
costs and benefits of the various constitutional options, and d) their own constitutional
preferences.
10
4. Given its importance in Quebec, the national question dominates over all over
cleavages, including that of traditional left-right.
5. The contagion effect of the national question is more marked in Quebec on the left.
This said, the motivations of people voting for right-wing parties are not centered on the
national question.
Quebec Voters and the National Question
The hypotheses will be tested via an online survey carried out during the weeks before
the April 7, 2014 Quebec provincial election. The context of this election lends itself
quite well to a study of the electoral impact of the national question. The Parti québécois
minority government called the elections on the basis of favourable polls amidst the
debate surrounding the adoption of the Quebec Charter of Values, which aimed to affirm
the secular nature of Quebec society. However, the electoral campaign quickly became
centered around the theme of Quebec independence after the addition of notable
businessman Pierre-Karl Péladeau to the ranks of the Parti québécois candidates. Upon
announcing his candidature, Péladeau unequivocally expressed his support for Quebec
sovereignty. This unintentionally derailed the campaign from its focus on issues
favourable for the Parti québécois, such as defending Quebec identity, to ground that was
more favourable for the Quebec Liberal Party, such as constitutional preferences.
11
We will first examine in a descriptive way Quebec’s political landscape before testing our
various hypotheses with the help of bivariate and multivariate analyses.
Nationalist Sensibilities
The first thing we need to establish is the importance of the national question in Quebec.
Quebec is the only province in Canada where a majority of the population is Frenchspeaking. Debates over the place of Quebec within the Canadian federation and the status
of the French language have dotted its history. As a result, partisan divisions are largely a
function of individuals’ attitudes on these questions.
In order to shed light on how central the national question plays into constitutional
preferences, we first asked the following three questions to survey participants: “When
you voted during the last provincial elections on April 7, how important were each of the
following factors in your decision? The ability of the party to defend the current and
future interests of Quebec? The ability of the party to understand Quebec’s history and
culture? The party’s constitutional preferences?”
The figures in Table 1 clearly show that a large majority of voters attribute great
importance to these questions. 93% of respondents see defending Quebec’s interests as
important; on this issue, there is a general consensus and political parties position
themselves accordingly. That a party can defend Quebec’s culture and history is also
important for 3 out of every 4 respondents (75%). A similar level of importance is given
to the constitutional positioning of the party, at 77% of the respondents. In this regard, it
12
is not surprising to see that voters for the two parties most engaged in this debate, the
Quebec Liberal Party and the Parti québécois, give more importance to this question
(87% for the PLQ and 83% for the PQ) than supporters of other parties such as Québec
solidaire (65%) or Coalition Avenir Québec (59%). These third parties seek to move the
political debate in Quebec towards economic and social questions. However, the
importance of the national question, in spite of these differences, is still very evident.
Table 1. Importance attached to various issues in the 2014 Election in Quebec
All
PQ
QS
CAQ
PLQ
Party stands up for Quebec interests
93
97
92
93
90
Party understands Quebec history and culture
75
90
88
66
65
Party constitutional position
77
83
65
59
87
Question: How is it important are the following questions in your decision to vote for this party in the
provincial election of April the 7th in Quebec? Don't knows, refusals excluded. For more details, see the
appendix.
The second important element for understanding the effect of the national question has to
do with the state of public opinion in Quebec. Quebecers’ nationalist sensibilities show
themselves in two ways. The first, as shown in Table 2, is seen in their greater attachment
to Quebec than to Canada. While 91% of Quebecers report being “very” or “somewhat”
attached to their province, this proportion is only 66% for attachment to Canada. This gap
is even wider among Francophones, who make up about 80% of Quebec’s population.
Among Francophones, 94% report being attached to Quebec versus 60% for being
attached to Canada (75% of non-Francophones report being attached to Quebec versus
92% of them report being attached to Canada).
13
Table 2. Level of attachment of Quebec and Canada
Quebec
Canada
Very
59
36
Somewhat
32
30
Not very
8
23
Not at all
2
11
Question: Would you say that you are very attached, somewhat attached, not very
attached or not attached at all to Quebec? Canada? Don’t know and refusals excluded.
See the appendix for details.
A look at the distribution of these feelings along partisan lines is revealing. The Liberal
clientele, largely due to the larger proportion of non-Francophones in its ranks, shows a
higher level attachment to Canada and a lower level of attachment to Quebec than other
partisan groups. The difference between Liberal voters’ degree of attachment to Quebec
and Canada, along with the levels of attachment observed in the population overall,
explain the relative discomfort of the Quebec Liberal Party on language and identity
matters. Stuck between a less nationalist electoral base and a more nationalist general
population, the Quebec Liberal Party, unlike the Parti québécois, tends to avoid these
questions, even if the party subscribes to the general goals of preserving and promoting
Quebec identity.
The third aspect of Quebec public opinion on the national question stems from the
second. To the extent which Quebecers are more attached to Quebec than to Canada, they
also give particular importance to the role that their provincial government plays. For
many Quebecers, their provincial government should have more powers within the
14
Canadian federation. In order to measure the pervasiveness of this attitude, we asked the
following question, “If a referendum were held asking if you want the Quebec National
Assembly to have more powers from the federal government, would you vote YES or
NO?”
Table 3. Opinions about transferring more powers to the
Quebec government
All
PQ
QS
CAQ
PLQ
Favorable
64
83
80
57
25
Opposed
36
17
20
43
75
Question: If a referendum was held asking you if you want that Quebec
National Assembly gets more powers from the Federal government,
would you vote YES or NO? See the appendix for more details.
The responses to this question are revealing. Approximately two-thirds of Quebecers
(64%) would vote in favour of this option. Furthermore, the separation of respondents
along partisan lines is quite informative. Support for this option either reaches or
surpasses 80% among Parti québécois and Québec solidaire voters, receives majority
support from the mostly federalist supporters of the Coalition Avenir Québec (57%), and
is backed by only one in four Quebec Liberal Party supporters. Therefore, the question of
repatriating powers is a delicate matter for the Quebec Liberal Party. It is also a touchy
subject for the Parti québécois, as the most hardcore supporters would not be happy that
their party of choice would seemingly abandon its raison d’être to limit itself to merely
pursuing more powers within the federation.
15
The perspective changes completely when one measures support for Quebec
independence. While it is certain that support for independence is now at a low when
compared to decades past, it seems that independence never really had clear and
widespread support within the population (with the exception of a few points in history).
The figures in Table 4 show that during the 2014 provincial elections, barely one in three
voters would support Quebec independence. Interestingly, this choice is also rejected by
about one in four Parti québécois voters (25%) and almost one out of every two Québec
solidaire voters, both of which are sovereignist political parties. Unsurprisingly,
supporters of federalist parties reject independence in even greater proportions. 82% of
Coalition Avenir Québec and no less than 97% of Quebec Liberal Party voters would
vote against Quebec independence.
Table 4. Opinions about the independence of Quebec
All
PQ
QS
CAQ
PLQ
Favorable
34
75
52
18
3
Opposed
66
25
48
82
97
Question: If a referendum was held asking you if you want that Quebec
becomes an independent country, would you vote YES or NO? See the
appendix for more details.
Thus, the picture is clear. Quebec voters have nationalist sensibilities while being in
favour of keeping Quebec within the Canadian federation. These voters want their
provincial government to defend Quebec’s interests and protect its identity and culture.
Along these questions, it would seem natural that voters have confidence in a nationalist
party such as the Parti québécois. However, although nationalists, Quebecers also favored
16
the federalist option. Thus, it would also seem natural that they have confidence in the
most federalist party, the Quebec Liberal Party, in order to keep Quebec within Canada.
To test our hypotheses, we asked respondents which party they thought was the best at
handling certain issues such as defending Quebec’s interests, protecting Quebec’s
language and culture, the economy, taxes, and poverty. The figures in Table 5 confirm
our first hypothesis and show that the most nationalist party, the Parti québécois, has the
advantage when it comes to “valence” dimensions of the national question (i.e. defending
Quebec’s interests and protecting its identity and culture). This advantage is even clearer
among Francophone respondents. Within this group, the Parti québécois is perceived as
being the most adept at protecting Quebec identity (49% of respondents versus 11% for
the Quebec Liberal Party, 10% for Québec solidaire and 9% for Coalition Avenir
Québec). These figures, along with the data showing weak support for Quebec
independence, confirms that the Parti québécois dominates the “valence” dimensions of
the national question in Quebec and that the Quebec Liberal Party has a clear advantage
when it comes to the “positional” aspects of this question (e.g. the choice between
independence or remaining within Canada).
Moreover, it is interesting to note that there is a certain degree of coherence to voters’
perceptions of parties on issues. Québec solidaire, the most leftist party, is most highly
ranked for tackling poverty; Coalition Avenir Québec, the most rightist party, is ranked
most highly for taxes; and the Quebec Liberal Party scores highly on the economy, which
has been the party’s trademark issue for a while. This coherence shows that the data we
17
have on how the parties are viewed by the Quebec electorate paints an adequately
accurate portrait of issue ownership in Quebec, notably on the dimensions of the national
question.
Table 5. Party images in Quebec
Best party:
PQ
QS
CAQ
PLQ
To defend Quebec's identity and culture
43
7
8
16
To defend Quebec' interest
33
7
14
25
To manage the economy
14
3
22
32
To cut taxes
7
5
30
15
To fight poverty
12
26
11
18
Question: Which party is best able to handle the following issues? See the appendix for more details.
Multivariate analyses of support for the four biggest political parties in Quebec will allow
us to test more rigorously and more thoroughly our hypotheses about issue ownership of
the national question in Quebec. They will also allow us to state that the national question
in Quebec is a “super issue” whose influence is greater than that of any other political
cleavage.
A Multivariate Analysis of Support for Quebec Political Parties
We will continue our analysis of the impact of the national question on electoral choices
by looking at the determinants of vote choice in Quebec during provincial elections. To
do this, we will examine the explanatory factors behind voter support for the four main
18
political parties in Quebec: the Parti québécois, the Quebec Liberal Party, the Coalition
Avenir Québec, and Québec solidaire. These four parties won 41.5%, 25%, 23%, and 7%
of the votes during the April 7, 2014 election. In the end, the Quebec Liberal Party won a
majority of seats in the Quebec National Assembly (i.e. 70 seats), followed by the Parti
québécois, who formed the official opposition with 30 seats. Finally, the Coalition Avenir
Québec and Québec solidaire won 22 and 3 seats, respectively.
The dependent variable take a value of 1 when a survey respondent reports having voted
for a party and 0 otherwise (non-responses are excluded). We use a recursive block
approach (Miller and Shanks 1996; Blais et al 2002; Nadeau et al 2012), which means
that the independent variables will be introduced into the model in successive blocks. The
first block contains long-term factors (other than the national question) that weigh on
political choices in Quebec, such as sociodemographic variables (age, gender, language,
education, income) and political attitudes (ideological positioning on a left-right axis,
moral conservatism, political cynicism; a more detailed presentation of the variables and
their coding is presented in the Appendix). The second block consists of just one variable:
the importance given by the respondent to the constitutional position of the parties during
the April 7 elections. The third block includes measures of attachment to both Quebec
and Canada, and also includes a variable on identification (dominant, exclusive, or
shared) with both Quebec and Canada. The fourth block contains two variables. The first
takes its maximum value when the respondent says that Quebec society’s values are
different as those of the rest of Canada. The second measures how respondents view the
tradeoffs between political sovereignty (even if it means belonging to a smaller market)
19
and the benefits of belonging to a bigger market (even if it means a loss of political
independence). Finally, the fifth block measures the effect of constitutional preferences
(independence or more powers) on vote choice in Quebec.
Long-Term Factors
The effect of long-term factors on political choices is presented both in the Appendix and
in Table 6 (we present changes in probabilities in the text, but provide the results of the
regression analyses in the Appendix). The sociodemographic variable that has the biggest
impact on vote choice is that of language. The probability of voting for the Quebec
Liberal Party drops 49 percentage points when a respondent is Francophone, while their
probability of voting for the Parti québécois, Coalition Avenir Québec, or Québec
solidaire rises by 43, 37, and 11 percentage points, respectively. In fact, this linguistic
cleavage translates into constitutional preferences among linguistic communities in
Quebec. Non-Francophones, tending to be federalist, massively support the Quebec
Liberal Party and Francophones, more nationalist, tend to split their votes between the
four big parties in Quebec. The other significant cleavage is age; the Parti québécois
receives more support from older voters and Québec solidaire, an emerging party, draw
upon younger and more educated voters for their support.
The effect of political attitudes is also significant. The Parti québécois and Québec
solidaire, both centre-left parties, receive more support from respondents who position
themselves on the left and who have more liberal opinions on morality questions. The
opposite trend is observed for rightist parties such as the Coalition Avenir Québec and the
20
Quebec Liberal Party. Interestingly, respondents who are more cynical towards politics
tend to support emergent parties such as Québec solidaire and the Coalition Avenir
Québec rather than traditional parties such as the Quebec Liberal Party and the Parti
québécois.
Table 6. Change in probabilities for voting models in Quebec: Sociodemographic and attitudinal variables
PQ
QS
CAQ
PLQ
0.24***
-0.17***
-0.08
0.07
-0.02
0.00
-0.05**
0.07***
Language
0.43***
0.11***
0.37***
-0.49***
Education
-0.15***
0.12***
-0.08
0.08
Income
-0.13***
-0.01
0.10**
0.05
Left-right
-0.34***
-0.20***
0.15**
0.36***
Moral liberalism
0.15***
0.13***
-0.13***
-0.11***
Cynicism
-0.10**
0.06**
0.15***
-0.13***
1127
1127
1127
1127
Age
Gender
N
***p≤0.01; **p≤0.05; *p≤0.10 (two-tailed tests). Entries are change in probabilities.
The results show that vote choice in Quebec is structured by two important determinants:
voters’ language and their positioning on ideological and moral questions. But as we will
see in the next sections, these cleavages overlap to a large extent with the respondents’
constitutional preferences. Non-Francophones largely support the Quebec Liberal Party
and respondents who are more conservative economically and socially are more favorable
to maintaining the constitutional status quo.
21
The Saliency Effect
The saliency effect can be seen in two ways. The first mechanism is additive and is the
most simple. A party that projects a proactive image with regard to a certain issue will
receive increased support among voters who view this issue as important (Bélanger and
Meguid 2008). The second mechanism is interactive and stipulates that opinion on an
issue will have more weight among people who attach a greater importance to this issue.
We examine the additive effect of voters’ priorities on their electoral choices. To do this,
we have added to the preceding model a variable that measures how important a
respondent believes a party’s constitutional preferences is to their vote choice. The results
of including this variable are shown in the Appendix (Table A2) and in Table 7. The
results are telling. They allow us to see that the two parties that profit largely from the
constitutional cleavage in Quebec are the Quebec Liberal Party and the Parti Québécois.
These are also the two parties that have the most diametrically opposed positions on the
constitutional question. Parties that give less importance to the issue, such as Québec
solidaire and especially the Coalition Avenir Québec, get significantly less support from
voters who view the national question as being very important.
22
Table 7. Change in probabilities for voting models in Quebec: Saliency
PQ
QS
CAQ
PLQ
0.25***
-0.16***
-0.11**
0.08
-0.03
0.00
-0.05*
0.08***
Language
0.44***
0.09**
0.32***
-0.45***
Education
-0.15***
0.13***
-0.11**
0.10*
Income
-0.10**
-0.03
0.08
0.06
Left-right
-0.32***
-0.21***
0.13**
0.35***
Moral liberalism
0.17***
0.12***
-0.15***
-0.12***
-0.08*
0.06**
0.13***
-0.12***
0.19***
-0.08***
-0.27***
0.21***
1101
1101
1101
1101
Age
Gender
Cynicism
IMPCONST
N
***p≤0.01; **p≤0.05; *p≤0.10 (two-tailed tests). Entries are change in probabilities.
These results confirm and nuance the bivariate analyses in the previous section. They
clearly show that attributing the issue of the national question to the most nationalist
parties is a far too simple hypothesis to correspond to reality. In fact, the two parties most
radically opposed to each other, the Parti québécois, the most nationalist party in Quebec,
and the Quebec Liberal Party, the least nationalist party, both benefit from this issue. The
Parti québécois wins votes because it is sovereignist, but also because voters view it as
being the most determined to defend the province’s interests and identity. The Quebec
Liberal Party largely profits from the national question because it is perceived as being
the best protector of the status quo.
23
Thus, the Quebec Liberal Party and the Parti québécois are both protagonists in the
constitutional debate. The more this debate is heated, the more both parties benefit. If the
debate centers more on Quebec’s interests or Quebec identity, the Parti québécois will
come out on top. However, if the debate is more about Quebec independence, such as the
previous election campaign in Quebec was, then the Quebec Liberal Party will come out
ahead. The other parties excluded from this dynamic must somehow insert themselves
into the debate. The leftist party, Québec solidaire, decided to portray itself as a party
favourable to Quebec independence. The Coalition Avenir Québec decided to take a
middle-ground position by claiming to be nationalist party but not in favor of going so far
as sovereignty.
The gains made by the Quebec Liberal Party and the Parti québécois with regard to the
constitutional debate show how both parties, for different reasons, benefit from keeping
this issue at the top of the agenda. The positions adopted by the other parties show a
contagion effect, whereby each political party in a sub-national party system where the
national question is important is forced to take a stance on the issue. As seen in
Hypothesis 5 from the previous section, we argue that the contagion effect is more
evident for the leftist Québec solidaire (favourable to independence) than for the rightist
Coalition Avenir Québec (favourable to federalism). However, the results in Table 6
bring nuances to this characterization. If it is true that the contagion effect can be seen
with regard to the official positions of the parties, this same effect cannot be seen among
voters who choose the parties at the periphery of the national debate, such as Québec
solidaire. In this case, the data show, contrarily, that voters for this party are less
24
interested in the national question (see Table 1) and that their support for Québec
solidaire stems from the desire to see the political debate focus on other issues. Thus, the
contagion effect can be seen at the level of the party leadership, but has not penetrated the
majority of Québec solidaire voters.
Identity and Vote Choice in Quebec
Two variables can help us measure the intensity of respondents’ opinions toward the
national question in Quebec. The first is an indicator of their degree of attachment toward
the two national communities of Quebec and Canada. As we have seen previously (in
Table 2), Quebecers are massively attached to Quebec and more divided in their
attachment to Canada. Another frequently used indicator looks at individuals’ exclusive,
dominant, or shared identification with these national communities. In this case, the
question asks respondents if they view themselves as only Quebecer, first Quebecer then
Canadian, equally Quebecer and Canadian, first Canadian then Quebecer, or only
Canadian. The figures show that Quebecers are split in terms of how they identify with
both Quebec and Canada. A little less than 40% of them identify first with Quebec (10%
exclusively and 27% primarily) and about the same proportion identifies first with
Canada (9% exclusively and 30% primarily). About a quarter equally identifies with both
national communities.
The results of the regression analyses for the model that includes these variables are
presented in the Appendix (Table A3) and in Table 8. Three clear conclusions can be
drawn from these results. First, including these variables considerably weakens the effect
25
of two variables: language of daily use and, especially, ideological positioning. In fact,
the latter variable is only significant for one party: Québec solidaire, a small leftist party
who only won 7% of the votes during the last elections. These results already offer some
kind of confirmation of Hypothesis 4 that the national question in Quebec dominates all
other cleavages, including left-right.
The second conclusion relates to the relative effect of the indicators measuring
attachment to and identification with Quebec and Canada. In most studies on nationalist
movements, it is the variable measuring identification with national communities that is
often used. However, the results clearly show that the variable measuring not
identification, but rather attachment to the national communities performs much better in
the example at hand (the variable measuring identification is not significant in any
model). Thus, it should be seen if this conclusion carries over into other contexts.
The third conclusion has to do more specifically with the effect of feelings of attachment
on support for political parties. From the data, two ideal-types emerge. The first consists
of voters for parties who are clearly implicated in the debate over the national question in
Quebec: the Parti québécois and the Quebec Liberal Party. For them, relative attachment
to Quebec and Canada plays a large role in support for their respective parties. This is not
the case for parties that are on the sidelines of this debate: Québec solidaire and the
Coalition Avenir Québec. Feelings of attachment towards Quebec and Canada for these
two electoral clienteles is much more divided. Taken together, the results give some
degree of confirmation to Hypothesis 3, which stipulated that the two parties most
26
directly engaged in the national debate in Quebec (i.e. the PQ and the PLQ), would
dominate this issue and profit electorally from it.
Table 8. Change in probabilities for voting models in Quebec:
attachment and identification variables
PQ
QS
CAQ
PLQ
0.18***
-0.14***
-0.08
0.09*
-0.00
0.01
-0.05**
0.04*
Language
0.15***
0.06
0.33***
-0.20***
Education
-0.19***
0.11***
-0.10*
0.12**
Income
-0.04
-0.01
0.05
0.01
Left-right
-0.04
-0.19***
0.05
0.10*
Moral liberalism
0.07*
0.10***
-0.12***
-0.01
Cynicism
-0.10**
0.05
0.12***
-0.09**
IMPCONST
0.13***
-0.09***
-0.26***
0.18***
-0.02
-0.00
0.04
-0.08*
AttachQueb
0.41***
-0.06
-0.12**
-0.16***
AttachCAN
-0.45***
-0.09***
0.12***
0.60***
1076
1076
1076
1076
Age
Gender
Identity
N
***p≤0.01; **p≤0.05; *p≤0.10 (two-tailed tests). Entries are change in probabilities.
Cost-Benefit Analyses
The debate over the national question in Quebec tends to revolve around two questions.
The first, most often put forward by independentists, says that a sovereign Quebec would
be better able to correspond to Quebecers’ values and aspirations. This argument assumes
27
that Quebecers’ values differ from those of other Canadians and that nothing short of full
control of a nation-state would allow Quebec to act upon their supposedly different policy
preferences. The second argument, often mentioned by federalists, states that in the
modern era, it is preferable to be part of a larger economic entity, even if that means
giving up a little bit of national sovereignty.
We tested the impact of these two arguments central to the national debate in Quebec by
including two variables into our models. The first sees whether Quebecers think their
values and aspirations differ from those of other Canadians. It is a scale variable where 0
means that the respondent believes that Quebecers have very distinct priorities and 10
means that they believe Quebecers have exactly the same values as other Canadians. The
other variable is also a scale variable where 0 means that the respondent believes it is
better to belong to a larger market, even if this means a loss of sovereignty, and a 10
means that the respondent has a preference for sovereignty, even if this means belonging
to a smaller market.
The distribution of these variables is interesting. With regard to values, Quebecers are
rather split. Slightly more of them believe that they share the same values as other
Canadians (44% answered between 0 and 4 on the scale) than those who believe that they
do not (40% answered between 6 and 10 on the scale; 12% chose the middle value and
4% refused to answer). With regard to the other question, Quebecers show a rather clear
preference for a bigger market, even at the cost of autonomy: 42% of the respondents
expressed their support for this view (by responding from 0 to 4 on the scale), which
28
contrasts with only 24% who chose the opposite perspective (by responding from 6 to 10
on the scale; 19% chose the median score and 15% refused to respond).
The effect of these variables on our multivariate model is presented in the Appendix
(Table A4) and in Table 9. The results are clear. Belonging to a larger market, even if it
translates into a loss of sovereignty, has a larger impact than perceived common shared
values on Quebecers’ electoral choices. The impact of these variables related to the
national question is once again much larger for the parties at the centre of this debate, the
PQ and the PLQ, than for peripheral parties. Preferences are quite clearly divided along
partisan lines. Voters for the Parti québécois opt for more sovereignty and those for the
Quebec Liberal Party value belonging to a larger market. The data confirm once again
that the PQ and the PLQ dominate the debate over the national question and their
electoral support depends on it more so than for Québec solidaire or Coalition Avenir
Québec.
29
Table 9. Change in probabilities for voting models in Quebec: Values
and Market variables
PQ
QS
CAQ
PLQ
0.20***
-0.13***
-0.09*
0.07
0.00
0.01
-0.05**
0.03
Language
0.14***
0.05
0.34***
-0.18***
Education
-0.20***
0.10***
-0.10*
0.13***
Income
-0.03
-0.01
0.05
-0.00
Left-right
-0.02
-0.19***
0.05
0.08
Moral liberalism
0.06
0.10***
-0.12***
-0.00
MDEM
-0.10***
0.04
0.13***
-0.08**
IMPCONST
0.11***
-0.10***
-0.25***
0.18***
-0.02
-0.00
0.04
-0.07
AttachQueb
0.32***
-0.08*
-0.10
-0.09*
AttachCAN
-0.31***
-0.07**
0.08
0.47***
Values
0.06
-0.01
-0.01
-0.05
Market
0.22***
0.07*
-0.09
-0.18***
1076
1076
1076
1076
Age
Gender
Identity
N
***p≤0.01; **p≤0.05; *p≤0.10 (two-tailed tests). Entries are change in probabilities.
Constitutional Preferences and Vote Choice
The final dimension of the debate over the national question is the parties’ constitutional
positioning. For the past forty years, two positions have dominated the scene: remaining
within Canada (i.e. federalism) and accession the status of an independent country (i.e.
30
sovereignty). As stated before, the sovereignist stance is advocated for by the Parti
Québécois and that of the federalists by the Quebec Liberal Party. After some intense
internal debate, Québec solidaire decided to join the sovereignist cause, even if this
question was not at the centre of its political discourse. The position of the Coalition
Avenir Québec is more ambiguous. This party wishes to put the constitutional debate
aside and instead focus on strengthening the province’s political and economic position.
That said, the party’s position falls more into the federalist camp, even if the party has a
takes a more nationalist posture than the Quebec Liberal Party does when it comes to
defending Quebec’s interests and promoting its identity. In this sense, the CAQ’s position
could be associated with the quest for more powers for Quebec. However, this middle-ofthe-road position, which has long had support in Quebec, has never before been explicitly
articulated by a political party. Therefore, it is a middle-of-the-road position that is
popular, but still an orphan on the political scene.
Two variables are introduced into the model to measure the impact of constitutional
positions: support (or not) for the independence of Quebec and the support (or not) for
transferring more powers from the federal to the Quebec government (see Tables 3 and
4). The results for the models including these variables are presented in Appendix A5 and
in Table 10. Support for independence and for more powers for Quebec contribute to
increasing the chances of supporting the Parti québécois. Rejecting the independence
option contributes significantly to increasing the chances of supporting the Coalition
Avenir Québec (by 15 percentage points) but even more for the Quebec Liberal Party (by
20 percentage points). The superficial way in which Québec solidaire supports
31
independence can be seen quite clearly. First, this party is the only one for which the leftright dimension is more important than the federalism-sovereignty dimension. In fact, if a
respondent positions themselves on the right, their probability of voting for this party
decreases by 20 percentage points). What is more is that being in support of
independence does not have a positive, but rather a negative, effect on the probability of
voting for this party (decrease of 4 percentage points). Also, the data show that the
Quebec Liberal Party receives support from not only opponents of independence, but also
those who oppose further transfer of powers to the Quebec government. This result
suggests that the Quebec Liberal Party is not only a federalist party, but also one for the
constitutional status quo.
Two additional results are worth of mention. We should first note that the most popular
position in Quebec, repatriation of more powers, does not seem to be associated with any
party. For PQ voters, this is a fallback position that is less interesting than that of
independence. The Quebec Liberal Party clientele rejects this position. Voters of
Coalition Avenir Québec are rather divided on this question (see Table 3; the effect of
this variable on party support is not statistically significant). Finally, Québec solidaire
voters seem favourable to this option, but it does not figure very high into their decision
to support this party. Thus, the position of more powers for Quebec is the most popular
position, but also one that seems to be the hardest for which to find a political home.
The last result to highlight is that constitutional preferences are more linked to support for
the PQ and the PLQ than for other parties. This result underscores the domination of
32
these parties on this issue, along with the superficial nature of QS’s support for the
sovereignist option.
Table 10. Change in probabilities for voting models in Quebec:
constitutional preferences
PQ
QS
CAQ
Age
0.18***
-0.13***
-0.10*
PLQ
0.10**
Gender
0.02
0.01
-0.06**
0.03
Language
0.10*
0.03
0.32***
-0.15***
Education
-0.15***
0.10***
-0.12**
0.12**
Income
-0.02
-0.01
0.04
-0.03
Left-right
-0.01
-0.20***
0.04
0.06
Moral liberalism
0.03
0.10***
-0.12***
0.01
Cynicism
-0.09**
0.04
0.12***
-0.07**
IMPCONST
0.09**
-0.08***
-0.23***
0.20***
-0.03
-0.00
0.04
-0.05
AttachQueb
0.20***
-0.07
-0.06
-0.03
AttachCAN
-0.17***
-0.08**
0.00
0.33***
Values
0.02
-0.02
0.00
0.00
Market
0.12**
0.09**
-0.04
-0.11**
Independence
0.16***
-0.04*
-0.15***
-0.20***
More powers
0.09***
0.04*
0.04
-0.09***
1076
1076
1076
1076
Identity
N
***p≤0.01; **p≤0.05; *p≤0.10 (two-tailed tests). Entries are change in probabilities.
33
Conclusion
In this paper, we have examine examined the effect of the national question on support
for the main provincial political parties in Quebec. To this effect, we discern three aspects
of this issue: defending Quebec’s interests, affirming Quebec identity, and parties’
constitutional options. We have concluded that there is a large consensus within the
province over the first two dimensions. This consensus reflects the nationalist
sensibilities of the province, thereby giving a valence issue character to these aspects of
the national question. However, constitutional options, and particularly the opposition
between federalists and sovereignists, polarize Quebec voters. Unlike the first two
dimensions, this aspect of the national question has the typical characteristics of a
positional issue.
Previous work on the effect of autonomist demands in regional and national legislatures,
notably focusing on Quebec, led us to formulate five hypotheses. The first three relate to
the notion of issue ownership. Contrary to existing work, we hypothesize that the national
question does not necessarily belong to the most nationalist party in the sub-state unit of a
country. We believe that the more consensual aspects of this question will indeed be
dominated by the most nationalist party, but that the more conflictual aspects (such as
asking for independence) actually also benefit the party who opposes this stance with the
most vigour.
The data show that this characterization adequately describes the political dynamic in
Quebec on the national question. The issue of the national question is dominated both by
34
the most nationalist party, the Parti québécois, who seem the most determined and able to
defend Quebec’s interests, and by the least nationalist party, the Quebec Liberal Party,
which represents the main opposition to the sovereignist project. Both parties benefit
from the national question being at the centre of the political debate for the past few
decades.
The key position occupied by the PQ and PLQ is explained by the importance of the
national question in Quebec. Here, the national question is a “super issue” that supersedes
all other cleavages in the province, including left and right. In fact, the only party in
Quebec able to distinguish itself on the left-right issue is Québec solidaire, a small leftist
party who won 7% of the popular vote during the last elections. In this context, it clearly
seems that the national question is the dominant cleavage in Quebec and that questions
related to the left-right division represent a niche issue instead.
The dominance of the national question in Quebec forces parties to adopt a stance
regarding this question, while taking into account public opinion. Two parties are directly
involved in the debate. The Parti québécois focuses on Quebecers’ nationalist sensibilities
in order to maximize its votes and the Quebec Liberal Party uses its federalist inclinations
among these same voters to build winning electoral coalitions. Any other party must also
find a way to insert themselves into the debate if they want to remain relevant. The first
option is to take a position in the debate. This is the route chosen by Québec solidaire,
which has joined the sovereignist camp. This behavior seems to confirm the hypothesis
that a contagion effect of the national question would affect leftist parties more than
35
rightist ones. However, one important nuance should be added. If it is true that Québec
solidaire is “officially” independentist, this contagion has not really reached its voters yet,
who mostly support this party for its leftist policies.
The other possible option when faced with such a dominant issue is to take a middleground position or, at least for a short period of time, offer voters a short of truce in the
debate in order to focus on other issues. This was the strategy of the Coalition Avenir
Québec, which has a nationalist, but non-sovereignist profile. This party wants the
national question to be put aside so that the province can work on consolidating its
finances and jumpstarting its economy. This nationalist position has its fair share of
ambiguity built into it. It reflects the needs for all parties in Quebec, under threat of being
perceived by the public as not sensitive to Quebec’s aspirations, to take a position on the
national question, even if this positioning can be superficial, as in the case of Québec
solidaire, or ambiguous, as in the case of Coalition Avenir Québec.
We still need to look deeper into the impact of the national question in Quebec. However,
it is clear that the question of autonomist demands in national sub-entities cannot be
reduced to the simple category of “niche issues”. On the contrary, this question is a
mainstream issue in Quebec whose effect on political dynamics is complex. Unraveling
this complexity is useful and important for better understanding Quebec politics.
However, the questions explored of this chapter can be extended beyond the borders of
the province. Work on this theme is important not only to better measure the effect of
36
autonomist debates in general, but also (and more widely) to enrich our understand of the
effect of issues on electoral behavior.
37
References
Bechhofer, Frank and David McCrone. 2010. « Chosing National Identiy », Sociological
Research Online (http://www.socresonline.org.uk/15/3/3.html>10.5153/sro.2191).
Bélanger, Éric and Bonnie McGuid. 2008. « Issue Salience, Issue Ownership and Issuebased Vote Choice », Electoral Studies 27 : 477-91.
Bélanger, Éric and Richard Nadeau. 2009a. Le comportement électoral des Québécois.
Montreal : Presses de l’Université de Montréal.
Bélanger, Éric and Richard Nadeau. 2006. « The Bloc Québécois: A Sour Taste Victory
». In The Canadian Federal Election of 2006, Jon H. Pammett and Christopher Dornan
(eds.), Toronto, Dundurn: 122-142.
Bélanger, Éric and Richard Nadeau. 2009b. « The Bloc Québécois: A Victory by Default
». In The Canadian Federal Election of 2008, Jon H. Pammett and Christopher Dornan
(eds.), Toronto: Dundurn: 136-161.
Bélanger, Éric and Andrea M. L. Perrella. 2008. « Facteurs d’appui à la souveraineté du
Québec chez les jeunes: une comparaison entre francophones, anglophones et allophones
», Politiques et Sociétés 27: 13-40.
Blais, André and Richard Nadeau. 1992. « To Be or not Be Sovereignist: Quebeckers’
Perennial Dilemma », Canadian Public Policy XVIII: 89-103.
Blais, André, Pierre Martin and Richard Nadeau. 1995. « Attentes économiques et
linguistiques et appui à la souveraineté du Québec », Revue canadienne de science
politique 28: 637-657.
Blais, André, Elisabeth Gidengil, Richard Nadeau and Neil Nevitte. 2002. Anatomy of a
Liberal Victory: Making Sense of the Vote in the 2000 Canadian Election. Peterborough:
Broadview Press.
Hepburn, Eve. 2014. « Spreading demands for self-determination: the ‘contagion effect’
of sub-state nationalist and regionalist parties on the territorial stances of other parties ».
Paper presented at the international conference Self-determination in Europe:
contemporary challenges of minorities in multinational states. University of Bern, 9-11
October, 2014.
38
Martin, Pierre and Richard Nadeau. 2001. « Understanding Opinion Formation on
Quebec Sovereignty », In Citizen Politics: Research and Theory in Canadian Political
Behaviour, Joanna Everitt et Brenda O’Neill (eds.), Don Mills: Oxford University Press,
142-158.
Miller, Warren E. and J. Merril Shanks. 1996. The New American Voter. Cambridge,
Mass.: Harvard University Press.
Nadeau, Richard and Chistopher Fleury. 1995. « Gains linguistiques anticipés et appui à
la souveraineté du Québec », Revue canadienne de science politique 28: 35-50.
Nadeau, Richard, Daniel Guérin et Pierre Martin. 1995. « L’effondrement du Parti
progressiste-conservateur à l’élection fédérale de 1993 », Revue québécoise de science
politique 27: 123-148.
Nadeau, Richard, André Blais, Elisabeth Gidengil and Neil Nevitte. 2000. « Perceptions
of Party Competence in the 1997 Election ». In Party Politics in Canada 8e ed. Hugh
Thorburn and Alan Whitehorn (eds.), Toronto: Prentice-Hall, 413-430.
Nadeau, Richard, Éric Bélanger, Michael Lewis-Beck, Bruno Cautrès and Martial
Foucault. 2012. Les élections présidentielles. Le vote des Français de Mitterrand à
Sarkozy. Paris : Presses de Sciences Po.
Stokes, Donald E. 1963. « Spatial Models of Party Competition ». American Political
Science Review 57(3): 368-377.
Stokes, Donald E. and John J. Dilulio. 1993. « The Setting: Valence Politics in Modern
Elections. » In The Elections of 1992, Michael Nelson (ed). Washington
DC:Congressional Quarterly Press.
Walgrave, Stefann, Jonas Lefevere and Anka Tresch. 2012. « The Associative Dimension
of Issue Ownership », Public Opinion Quarterly 76: 771-782.
39
Appendix
A. Tables for logistic regression
Table A1. Logistic regression for vote (Model 1)
PQ
QS
CAQ
1.38***
-2.65***
-0.47
Age
(0.31)
(0.55)
(0.32)
-0.14
0.06
-0.31**
Gender
(0.15)
(0.24)
(0.15)
2.40***
1.69***
2.24***
Language
(0.40)
(0.55)
(0.36)
-0.86***
1.88***
-0.48
Education
(0.33)
(0.57)
(0.34)
-0.75***
-0.15
0.62**
Income
(0.29)
(0.45)
(0.30)
-1.89***
-3.16***
0.90**
GcDr
(0.36)
(0.63)
(0.36)
0.83***
2.12***
-0.81***
CMORAL
(0.25)
(0.52)
(0.25)
-0.59**
1.01**
0.91***
MDEM
(0.25)
(0.43)
(0.26)
-1.93***
-4.85***
-3.37***
Constant
(0.57)
(0.93)
(0.57)
0.21
0.29
0.14
Nagelkerke pseudo-R
72.1%
92.7%
76.7%
% correctly predicted
1127
1127
1127
N
PLQ
0.47
(0.33)
0.45***
(0.16)
-3.08***
(0.25)
0.50
(0.34)
0.30
(0.30)
2.22***
(0.38)
-0.67***
(0.25)
-0.83***
(0.26)
0.65
(0.50)
0.37
77%
1127
***p≤0.01; **p≤0.05; *p≤0.10 (two-tailed tests). Entries represent unstandardized logistic
regression coefficients (with standard deviations in parentheses).
40
Table A2. Logistic regression for vote (Model 2)
PQ
QS
1.48***
-2.61***
Age
(0.32)
(0.56)
-0.16
0.04
Gender
(0.15)
(0.25)
2.57***
1.41**
Language
(0.40)
(0.56)
-0.88***
2.04***
Education
(0.34)
(0.60)
-0.58**
-0.43
Income
(0.30)
(0.47)
-1.84***
-3.36***
GcDr
(0.36)
(0.66)
1.00***
1.98***
CMORAL
(0.26)
(0.52)
-0.48*
0.89**
MDEM
(0.25)
(0.43)
1.12***
-1.29***
IMPCONST
(0.26)
(0.40)
-3.21***
-3.47***
Constant
(0.64)
(1.01)
0.24
0.30
Nagelkerke pseudo-R
74.5%
92.7%
% correctly predicted
1101
1101
N
CAQ
-0.68**
(0.34)
-0.30*
(0.16)
2.03***
(0.37)
-0.72**
(0.36)
0.50
(0.31)
0.82**
(0.38)
-0.97***
(0.26)
0.85***
(0.26)
-1.73***
(0.26)
-1.56**
(0.63)
0.21
77.1%
1101
PLQ
0.48
(0.34)
0.53***
(0.16)
-2.90***
(0.25)
0.64*
(0.36)
0.41
(0.32)
2.28***
(0.38)
-0.76***
(0.26)
-0.80***
(0.26)
1.33***
(0.28)
-0.63
(0.57)
0.40
78.3%
1101
***p≤0.01; **p≤0.05; *p≤0.10 (two-tailed tests). Entries represent unstandardized logistic
regression coefficients (with standard deviations in parentheses).
41
Table A3. Logistic regression for vote (Model 3)
PQ
QS
1.40***
-2.23***
Age
(0.39)
(0.59)
0.00
0.17
Gender
(0.18)
(0.25)
1.18***
0.90
Language
(0.43)
(0.62)
-1.46***
1.74***
Education
(0.40)
(0.62)
-0.34
-0.20
Income
(0.35)
(0.49)
-0.33
-3.02***
GcDr
(0.43)
(0.73)
0.55*
1.66***
CMORAL
(0.30)
(0.54)
-0.77**
0.73
MDEM
(0.31)
(0.45)
1.01***
-1.47***
IMPCONST
(0.31)
(0.42)
-0.18
-0.03
Identity
(0.30)
(0.39)
3.16***
-0.95
AttachQueb
(0.58)
(0.66)
-3.43***
-1.54***
AttachCAN
(0.31)
(0.47)
-2.81***
-1.28
Constant
(0.90)
(1.22)
0.46
0.32
Nagelkerke pseudo-R
81.6%
92.8%
% correctly predicted
1076
1076
N
CAQ
-0.52
(0.35)
-0.32**
(0.16)
2.15***
(0.37)
-0.64*
(0.37)
0.35
(0.32)
0.35
(0.41)
-0.79***
(0.27)
0.80***
(0.27)
-1.69***
(0.27)
0.24
(0.30)
-0.80**
(0.38)
0.77***
(0.28)
-1.50**
(0.75)
0.21
78.0%
1076
PLQ
0.80*
(0.41)
0.32
(0.19)
-1.76***
(0.28)
1.03**
(0.44)
0.13
(0.38)
0.87*
(0.48)
-0.10
(0.32)
-0.75**
(0.31)
1.56***
(0.32)
-0.68*
(0.40)
-1.38***
(0.42)
5.21***
(0.42)
-3.84***
(0.83)
0.61
84.5%
1076
***p≤0.01; **p≤0.05; *p≤0.10 (two-tailed tests). Entries represent unstandardized logistic
regression coefficients (with standard deviations in parentheses).
42
Table A4. Logistic regression for vote (Model 4)
PQ
QS
1.56***
-2.16***
Age
(0.40)
(0.59)
0.02
0.15
Gender
(0.18)
(0.26)
1.11**
0.76
Language
(0.46)
(0.63)
-1.56***
1.69***
Education
(0.40)
(0.62)
-0.27
-0.17
Income
(0.35)
(0.49)
-0.17
-3.10***
GcDr
(0.44)
(0.75)
0.45
1.63***
CMORAL
(0.31)
(0.54)
-0.81***
0.71
MDEM
(0.31)
(0.45)
0.86***
-1.57***
IMPCONST
(0.31)
(0.43)
-0.19
-0.04
Identity
(0.30)
(0.39)
2.51***
-1.26*
AttachQueb
(0.61)
(0.71)
-2.49***
-1.22**
AttachCAN
(0.37)
(0.57)
0.50
-0.20
Values
(0.34)
(0.59)
1.78***
1.22*
Market
(0.42)
(0.63)
-3.79***
-1.40
Constant
(0.95)
(1.26)
0.49
0.33
Nagelkerke pseudo-R
82.6%
93.0%
% correctly predicted
1076
1076
N
CAQ
-0.60*
(0.36)
-0.33**
(0.17)
2.20***
(0.38)
-0.62*
(0.38)
0.33
(0.32)
0.31
(0.41)
-0.77***
(0.27)
0.83***
(0.27)
-1.67***
(0.27)
0.24
(0.30)
-0.64
(0.40)
0.51
(0.34)
-0.04
(0.31)
-0.58
(0.38)
-1.24
(0.78)
0.21
78.2%
1076
PLQ
0.63
(0.42)
0.31
(0.20)
-1.64***
(0.29)
1.19***
(0.45)
-0.01
(0.38)
0.69
(0.49)
-0.04
(0.33)
-0.72**
(0.32)
1.64***
(0.33)
-0.63
(0.40)
-0.83*
(0.44)
4.29***
(0.47)
-0.42
(0.33)
-1.64***
(0.42)
-2.91***
(0.87)
0.62
84.5%
1076
***p≤0.01; **p≤0.05; *p≤0.10 (two-tailed tests). Entries represent unstandardized logistic
regression coefficients (with standard deviations in parentheses).
43
Table A5. Logistic regression for vote (Model 5)
PQ
QS
1.59***
-2.23***
Age
(0.41)
(0.60)
0.14
0.08
Gender
(0.19)
(0.26)
0.90*
0.55
Language
(0.47)
(0.64)
-1.31***
1.66***
Education
(0.42)
(0.62)
-0.13
-0.23
Income
(0.36)
(0.49)
-0.12
-3.31***
GcDr
(0.46)
(0.77)
0.26
1.60***
CMORAL
(0.32)
(0.55)
-0.81**
0.74
MDEM
(0.32)
(0.45)
0.74**
-1.41***
IMPCONST
(0.34)
(0.44)
-0.25
-0.04
Identity
(0.31)
(0.40)
1.69***
-1.12
AttachQueb
(0.64)
(0.75)
-1.45***
-1.39**
AttachCAN
(0.41)
(0.59)
0.16
-0.30
Values
(0.37)
(0.61)
1.01**
1.42**
Market
(0.45)
(0.66)
1.41***
-0.67*
Independence
(0.25)
(0.37)
0.76***
0.63*
Power
(0.24)
(0.37)
-4.14***
-1.32
Constant
(0.98)
(1.27)
0.53
0.34
Nagelkerke pseudo-R
83.2
92.9%
% correctly predicted
1076
1076
N
CAQ
-0.67*
(0.36)
-0.40**
(0.17)
2.13***
(0.38)
-0.78**
(0.38)
0.29
(0.33)
0.27
(0.42)
-0.77***
(0.28)
0.82***
(0.28)
-1.51***
(0.27)
0.26
(0.31)
-0.42
(0.40)
0.00
(0.38)
0.01
(0.32)
-0.30
(0.40)
-0.98***
(0.25)
0.24
(0.20)
-0.91
(0.78)
0.23
77.9%
1076
PLQ
0.96**
(0.45)
0.28
(0.20)
-1.46***
(0.30)
1.12**
(0.46)
-0.27
(0.40)
0.62
(0.51)
0.10
(0.33)
-0.71**
(0.33)
1.88***
(0.34)
-0.45
(0.43)
-0.24
(0.45)
3.20***
(0.51)
0.02
(0.35)
-1.09**
(0.45)
-1.88***
(0.43)
-0.90***
(0.22)
-2.68***
(0.90)
0.65
85.0%
1076
***p≤0.01; **p≤0.05; *p≤0.10 (two-tailed tests). Entries represent unstandardized logistic
regression coefficients (with standard deviations in parentheses).
44
Data and Coding
The study was conducted among the general adult population in Quebec. The fieldwork,
which started immediately after the Quebec general election of April the 7th 2014, was
performed between April the 9th to April the 16th 2014. Respondents were selected from
the panel LégerWeb wich includes more than 400,000 Canadian households. The weight
variable for the survey is based on the following variables: SEX, AGE, REGION and
LANGUAGE, according to the latest Statistics Canada census.
Variables
Dependent Variables
PQ
Dummy
1= Vote for “Parti Québécois” in Quebec elections
2014
0= Vote for another party
Abstentions, blank votes or missing observations are
coded as missing
QS
Dummy
1= Vote for “Quebec Solidaire” in Quebec elections
2014
0= Vote for another party
Abstentions, blank votes or missing observations are
coded as missing
CAQ
Dummy
1= Vote for “Coalition Avenir Québec” in Quebec
elections 2014
0= Vote for another party
Abstentions, blank votes or missing observations are
coded as missing
PLQ
Dummy
1= Vote for the liberal party in Quebec elections 2014
0= Vote for another party
Abstentions, blank votes or missing observations are
coded as missing
Sociodemographic variables
Age
Scale
Age of respondents, rescaled from (0) to (1)
Gender
Dummy
1=Female; 0=Male
Language
Dummy
1=Respondents who speak French at home;
0=Non-French
Socioeconomic variables
45
Education
11 point scale
Highest level of education that has been completed by
the respondents. Scores are rescaled from 0 (no
education) to 1 (master or doctoral).
Missing observations are coded as missing.
Income
9 point scale
Respondents’ household income of a 9-point level
scale, rescaled from 0 (less than $8,000) to 1
($104,000 and more).
Missing observations are coded as missing.
Political Attitude
GcDr
11 point scale
Respondents’ ideology of a 11-point level scale,
rescaled from 0 (left) to1 (right).
Missing observations are coded as 0.5
CMORAL
Scale
Scores are means of two following variables:
Are you for or against marriage between people of the
same sex? 0 (Against); 0.5 (missing); 1 (For)
Are you for or against death penalty? 0 (For); 0.5
(missing); 1 (Against)
MDEM
Scale
Scores are means of three following variables:
The National Assembly of Quebec does not care much
what people think: 0 (Disagree); 0.5 (missing); 1
(Agree)
People like me have no say in what the provincial
government in Quebec do: 0 (Disagree); 0.5 (missing);
1 (Agree)
Overall, are you satisfied (e) the way democracy
works in Quebec? 0 (Satisfied); 0.5 (missing); 1
(Unsatisfied)
Identity
Identity
5 point scale
Respondents’ opinion on their identity varying from 1
(Only Canadian, not Quebec people) to 5 (Only
Quebec people, not Canadian). Scores are rescaled
from 0 to 1.
Missing observations are coded as missing.
AttachQUEB
4 point scale
Respondents’ level of attachment of Quebec varying
from 1 (not at all attached) to 4 (very attached),
rescaled from 0 to 1.
Missing observations are coded as missing.
46
AttachCAN
4 point scale
Respondents’ level of attachment of Canada varying
from 1 (not at all) to 4 (very attached), rescaled from 0
to 1.
Missing observations are coded as missing.
Values
Values
11 point scale
Respondents’ opinion on their distinct values varying
from 1 (Same values) to 1 (Disntinct values). Scores
are rescaled from 0 to 1.
Missing observations are coded as 0.5
Market
11 point scale
Respondents’ opinion on the importance of market
varying from 1 (larger market with less sovereignty) to
11 (More sovereignty with smaller market). Scores are
rescaled from 0 to 1.
Missing observations are coded as 0.5
Political issues
IMPCONST
4 point scale
Respondents’ opinion on the constitutional preferences
of the party varying from 1 (not at all important) to 4
(very important). Scores are rescaled from 0 to 1.
Missing observations are coded as missing.
Independence
Dummy
0=Otherwise
1=Opinion in favor of the independence of Quebec
Missing observations are coded as missing.
Power
Dummy
0=Otherwise
1=Opinion in favor of more power for the National
Assembly of Quebec
Missing observations are coded as missing.