Kings-Hants Favourite Son: Scott Brison s Personal Popularity and Local Liberal Strength Help Overcome Some Misgivings about Gay Marriage COMPAS Inc. Public Opinion and Customer Research June 9, 2004
Liberal Landslide Scott Brison, the prominent Conservative-turned-Liberal, is set to win a landslide. He would also have won a landslide had he run again as a Conservative even though this historically Conservative riding is transforming itself into an intrinsically Liberal area, according to emerging evidence to be discussed. In the present contest, running as a Liberal Brison has a comfortable 49% lead over Conservative candidate Bob Mullan (29%) and NDP candidate Skip Hambling (16%), according to a COMPAS riding poll (n=300) conducted for the National Post, Global Television, and other CanWest newspapers (see table 1). The riding has become an intrinsically Liberal constituency as evidenced by the voting intentions expressed by respondents in a hypothetical election in which they had no information about the candidates. In such a hypothetical contest, the Liberals would garner 40% to 29% for the Conservatives, as shown in table 2. Brison s move to the Liberals has provided a boost to the party, raising Liberal support by 9 percentage points over what an un-named Liberal candidate might receive (table 2). Brison augments Liberal support by a successful appeal to many who say that they voted for the Tories or Alliance in the last election. Indeed, 28% of his current support is drawn from former Tory or Alliance voters, as shown in table 3. Yet another indication of Brison s popularity is that he would win a landslide as a Conservative candidate. He would earn 47% support in the hypothetical situation that Scott Brison had not switched to the Liberals and were running for the Conservatives this time, as shown in table 4. The responses to the hypothetical scenario suggest that Brison would win essentially the same amount of support as a Conservative (47%) as he will running as a Liberal (49%). Brison s popularity in the electorate is reflected in the open-ended responses such as: I will vote for Scott Brison because of his personality and determination. He is a good person. Scott Brison is the best candidate for the riding and I am not satisfied with the job that the other candidates have done. Despite his popularity, Brison is a bit of a polarizing figure. His public stance in support of gay marriage may be a factor. He may offend some voters for other reasons as well, notably his crossing of party lines. As one 2
detractor put it, I do not trust Scott Brison. He lies and betrayed his own party. While Brison appears to have a comfortable lead, it should be noted that there was an especially high proportion of do not know/refused responses (36%). The undecided group is made up primarily of females aged 40+ with high school education. This group would be more apt to vote for the Liberals (44%) than the Conservatives (35%) in the hypothetical situation that Brison had stayed with the Conservatives. A high turnout among this group could attenuate but not remove Brison s lead. But such a turnout would not take away victory except in the unlikely event that the undecideds harboured a strong, silent opposition to him over his party switching or advocacy of gay marriage rights. Table 1: Vote Intentions, DNKs Excluded In your present thinking, will you vote for [ROTATE IF UNDECIDED/REFUSED TO ANSWERS] Whom are you then leaning to? [PROMPT ONLY IF NECESSARY SAME ROTATION AS ABOVE] TOTAL Scott Brison, the Liberal candidate 49 Bob Mullan, the Conservative candidate 29 Skip Hambling, the NDP candidate 16 The Green candidate 5 Other Party 1 Would not vote (excl. from calc.)* DNK/ Undecided/ Refused (excl. from calc.)** *Percent Would not vote excluded from final percent calculation **Percent DNK/ Undecided/ Refused excluded from final percent calculation 3 33 3
Table 2: Pure Hypothetical Party Vote by Actual Intended Vote (Vertical %) Let s imagine for a moment that all of the current candidates in your particular riding resigned and pulled out of the election for personal, family, or health reasons. Let s suppose that the [ROTATE] Liberal, Conservative, NDP and Green parties all nominated new candidates whom you knew nothing about. In this hypothetical situation, which party would you vote for? [PROMPT ONLY IF NECESSARY] TOTAL BRISON LIBERAL MULLAN CONSERVATIVE HAMBLING NDP GREEN OTHER UNDECIDED/ DNK/REF The Liberal candidate 40 73 13 9 0 0 37 The Conservative candidate 29 10 78 13 0 0 26 The NDP candidate 21 15 4 74 11 0 22 The Green candidate 9 3 4 4 89 100 11 Other Party 1 0 0 0 0 0 4 Would not vote (excl. from calc.)* DNK/ Undecided/ Refused (excl. from calc.)** 7 4 5 6 0 0 10 34 19 13 22 10 0 65 *Percent Would not vote excluded from final percent calculation **Percent DNK/ Undecided/ Refused excluded from final percent calculation Like some other candidates, Brison outperforms his party rather than earning less voter support than his party could normally expect. A combined Appendix for all 12 ridings studies, containing COMPAS' proprietary electability-traction index scores for each significant candidate in every riding, is provided as a separate, stand-alone document. 4
TOTAL Table 3: Voting History, DNKs Excluded In the last federal election, did you vote [SAME ROTATION AS IN THE VOTE QUESTION] BRISON LIBERAL MULLAN CONSERVATIVE HAMBLING NDP GREEN OTHER UNDECIDED/ DNK/REF PC/Alliance 40 28 71 26 0 0 34 Liberal 37 62 10 13 33 0 34 NDP 23 10 14 61 67 100 32 Other Party Did not vote (excl. from calc.)* DNK/ Don t Recall/ Refused (excl. from calc.)** 1 0 4 0 0 0 0 13 8 5 19 30 0 19 23 10 7 9 40 0 46 *Percent Would not vote excluded from final percent calculation **Percent DNK/ Undecided/ Refused excluded from final percent calculation Table 4: Pure Hypothetical Candidate Vote by Actual Intended Vote (Vertical %) Suppose Scott Brison had not switched to the Liberals and were running for the Conservatives this time. In this hypothetical situation, which party would you vote for? [SAME ROTATION AS IN THE VOTE QUESTION TOTAL BRISON LIBERAL MULLAN CONSERVATIVE HAMBLING NDP GREEN OTHER UNDECIDED/ DNK/REF Brison as hypothetical Conservative candidate 47 36 93 21 0 0 35 Liberal 34 58 2 11 0 0 44 NDP 17 6 6 68 40 0 18 Other Party 2 0 0 0 60 100 3 Did not vote (excl. from calc.)* DNK/ Don t Recall/ Refused (excl. from calc.)** 6 0 4 6 10 0 12 25 7 2 6 40 0 56 5
*Percent Would not vote excluded from final percent calculation **Percent DNK/ Undecided/ Refused excluded from final percent calculation Health and Leadership are Top Drivers To the extent that they are influenced by issues, voting decisions among residents of Kings-Hants appear to be driven above all by healthcare and leadership. When asked to volunteer what factor most affects how they will probably vote, respondents are most apt to volunteer issues related to healthcare (25%) or leadership (20%), as shown in table 5. Healthcare is especially important to those who have not decided how they will vote (36%). This may suggest that undecided voters are more provincially focused than federally focused. In COMPAS national polling for CanWest conducted in late May, the national party segments tended to identify healthcare in equal shares as drivers of the vote. In some individual riding studies, Liberal voters tended to identify healthcare more frequently than Conservatives, as if explaining their vote in terms of the cues provided by their party leader. In Kings Hants, Conservative voters are pointing to healthcare more frequently 27% vs. 16% among Liberals and 13% among NDPers. The emphasis of Conservative voters on healthcare may reflect their high regard for their candidate, a respected local physician, as well as the older character of the Conservative voter. Among the many volunteered responses about healthcare are the following: Health is number one. There is nothing else. More money needs to be invested in healthcare. I would like to see more federal money go to the provinces to help with healthcare. Waiting times are a big problem. Our current health care system has a lack of doctors and nurses. We should cut down waiting times. At my age what concerns me is my health. I am concerned about healthcare. It is really number one in my books. Healthcare is my main issue. There is not enough funding for nursing schools, not enough healthcare professionals available, and a lack of incentives to make them want to stay here. It does not really matter what happens or who gets in though because they all 6
do their own thing and follow their own plan once elected. As in other ridings, Conservative voters play slightly more emphasis than others on corruption or integrity themes but these are not as much of focus for Conservative voters in Kings Hants as elsewhere. 7
Table 5: Unprompted, Volunteered Vote Drivers Ballot Issues Thinking of how you re likely to vote, please tell me what factor most affects how you ll probably vote? [USE STANDARD CLASSIFICATION SCHEME FOR ISSUES] TOTAL BRISON LIBERAL MULLAN CONSERVATIVE HAMBLING NDP GREEN OTHER UNDECIDED/ DNK/REF Health 25 16 27 13 10 100 36 Leaders or leadership in general, integrity 20 24 20 28 10 0 16 Taxes and economy 9 7 9 13 0 0 11 Miscellaneous party policies Depends on local candidate Education in general improve or protect public education 7 9 7 3 10 0 6 4 9 2 3 0 0 2 2 2 2 0 10 0 2 Environment 2 1 0 3 30 0 2 Gay, women s and other rights; social issues Adscam, Liberal Government corruption 2 1 0 6 10 0 1 2 0 5 3 0 0 2 Pro-Liberal 4 10 2 0 0 0 0 Anti-Liberal 2 0 4 9 0 0 0 Pro-Tory 1 0 4 0 0 0 1 Anti-Tory 1 0 0 6 0 0 0 Pro-NDP 1 1 0 9 0 0 0 Anti-NDP 1 1 2 0 0 0 0 Martin like 1 2 0 0 0 0 0 Martin dislike * 0 2 0 0 0 0 Harper dislike * 1 0 0 0 0 0 Other 2 4 2 0 10 0 0 DNK 15 10 14 3 10 0 22 8
Local Candidate More Important than National Party Leader and National Party After being asked open-ended questions about the factors driving their votes, respondents were then asked to score the importance of party, party leaders, and local candidates in their voting decisions. Given the fact that Brison outpolls his new party, one might expect local candidates to matter to Kings Hants voters, especially to Liberal voters. That is what the findings show. Local candidates appear to be more important (mean 5.4; 52% score a 7 or 6) than national party leaders (mean 5.1; 43%) and national parties (mean 4.9; 38%), as shown in table 6. Liberal voters are significantly more apt to say that the local candidate is important (mean 5.9; 65%) than are Conservative (mean 5.6; 61%) and undecided (mean 5.0; 44%) voters. Liberal voters (mean 5.5; 51%) are also significantly more apt than undecided voters (mean 4.7; 33%) to deem important the national party leaders. Table 6: Importance of Leaders, Parties, and Candidates as Stated Drivers of the Vote Using a 7 point scale where 7 means very important and 1, the opposite, please tell me the importance of each of the following when deciding how you ll vote [ROTATE] MEAN 7 6 5 4 3 2 1 DNK The local candidates in your riding Overall 5.4 35 17 19 9 4 4 5 6 Brison Liberal Voters 5.9 47 18 18 9 2 3 1 2 Mullan Conservative 5.6 45 16 20 7 4 5 4 0 Voters Hambling 4.7 22 13 28 6 6 16 6 3 NDP Voters Green Party Voters 5.0 10 20 20 20 10 0 0 20 Other Party Voters N/A 100 0 0 0 0 0 0 0 Undecided/ DNK/ REF 5.0 25 19 16 10 5 2 10 13 The national party leaders Overall 5.1 29 14 23 11 7 3 7 7 Brison Liberal Voters Mullan Conservative Voters Hambling NDP Voters 5.5 33 18 24 14 2 1 4 4 5.5 30 18 34 5 7 0 4 2 4.6 28 13 13 13 16 13 6 0 9
MEAN 7 6 5 4 3 2 1 DNK Green Party Voters 5.1 20 20 30 20 0 10 0 0 Other Party Voters N/A 100 0 0 0 0 0 0 0 Undecided/ DNK/ REF 4.7 25 8 20 10 8 2 13 14 The national parties Overall 4.9 23 15 23 12 8 4 9 8 Brison Liberal Voters Mullan Conservative Voters 5.1 24 19 29 14 4 1 7 2 5.3 27 14 30 14 5 2 2 5 Hambling NDP Voters 4.6 22 9 22 16 13 9 6 3 Green Party Voters 4.7 30 20 0 10 0 20 10 10 Other Party Voters N/A 100 0 0 0 0 0 0 0 Undecided/ DNK/ REF 4.4 18 13 16 8 12 4 14 15 Majority Opposes Gay Marriage But the Issue Is Not an Election Driver and Brison Is Unaffected As shown in table 7, 40% support gay marriage rights while 48% support only civil union or neither union nor marriage Liberal (48%), NDP (41%), and undecided (37%) voters are more apt than Conservative voters (25%) to support rights for gay marriage. A similar pattern of less pro-gay attitudes among Conservative/Alliance voters occurs when relating attitude toward gay marriage with reported vote in the last federal election. The pattern would not emerge in the hypothetical situation of Brison s running as a Conservative. In this scenario, 40% of hypothetical Conservative/Brison voters would support gay marriage compared to 45% of hypothetical Liberal voters. It is as if voters on the right end of the political spectrum are more opposed than the public as a whole to gay marriage rights but would not let their difference of opinion with Brison get in the way of voting for him if he were a Conservative. 1 For the media, the lesson of this finding is that gay marriage is an issue of very moderate importance in the election, even of moderate importance in Kings Hants, which boasts one of Canada s most high 1 A competing explanation is that pro-gay marriage Liberals would follow Brison to the Tories while anti-gay marriage Tories would vacate for the Liberals. But our detailed examination of the attitudes of the small numbers of switchers in this hypothetical scenario lends no support to this interpretation. 10
profile gay advocates. The issue is significant enough to distinguish Conservatives from other voters in the absence of a gay Conservative candidate. But gay marriage s significance dissipates in the real life situation of a popular gay advocate running as a Conservative. Table 7: Vote Intentions vs. Views on Same Sex Marriage Turning to a specific social policy issue, which of the following views about homosexual or gay marriage is closest to your own? [ROTATE] Homosexuals should be allowed the same marriage rights as a man and a woman Homosexuals should be allowed the right to a civil union but not marriage Homosexuals should not be allowed either civil unions or marriage DNK/ Undecided/ Refused TOTAL BRISON LIBERAL MULLAN CONSERVATIVE HAMBLING NDP GREEN OTHER UNDECIDED/ DNK/REF 40 48 25 41 80 0 37 33 31 38 44 10 0 30 15 9 21 13 10 100 18 12 12 16 3 0 0 15 Demographic Communities Age and Education Minor Factors, Gender Potentially a Major Factor Ethnicity is not a factor while age and education play a minor role. The oldest cohort, those 60 and over, are disproportionately supporters of Mullan. This group accounts for about 30% of all voters and traditionally 11
has a high turnout. Hence, Mullan s ultimate share of the vote may be higher than estimated earlier in this report. Brison does especially well among university graduates. An intriguing issue is why the Conservative candidate earns at least twice as much support among men as women. It is possible that women in a riding somewhat pre-occupied with healthcare may be ambivalent about losing a local family physician to the House of Commons. A related issue is whether the high undecided rate among women reflects a common reticence among elderly women respondents or whether they are taking time to reflect on whom they will support. Methodology A representative sample of n=300 residents entitled to vote in the coming election were interviewed by professional interviewers using computer-assisted telephone interviewing technology June 4-7, 2004. Samples of this size are deemed accurate to within 5.8 percentage points 19 times out of 20. The principal investigator on this study was Dr. Conrad Winn. 12