Deep Ambivalence About Polls, Despite Increased Consumption & Use
HIGHLIGHTS:
-> Polls are increasingly used for strategic voting (by as many as 1 in 5)
-> But, growing concern about potential corrosive impact on democracy
[OTTAWA – October 2, 2008] – Many Canadians are prepared to change their vote, depending on how they think the parties are doing in this election. How do we know? The polls tell us so.
So-called “strategic voting” is common, perhaps even increasingly common, in modern Canada. NDPers voting Liberal to stop the Conservatives; Liberals voting Conservative to make sure they have a winning MP for their area; Greens voting for another party because they think their local candidate can’t win.
None of this would be possible without polls, of course, and Canadians say they are watching the numbers ebb and flow. Two-thirds of Canadians say they are paying attention to the polls in this election campaign; 30% say they are paying a great deal of attention. Attention to the polls is higher amongst seniors and in places where the races are tighter.
Twenty-two percent of Canadians are prepared to say they will take the party standings into account before they cast their ballot. Unsurprisingly, poll consumption (and strategic voting) is lowest in Alberta where using polls to predict the outcome is akin to measuring a hot dog with a micrometer.
But most of us are uneasy – downright uncomfortable – with the role polls play in our elections.
Half of Canadians say they think the polls tend to reduce elections to horse races, and in that way weaken the democratic process. Thirty-seven percent say the polls give them useful information that helps them make choices.
So the people have spoken! And how do we know that? We did a poll. (Which means there might be a selection bias where the staunchest poll-a-phobics are missing from the sample altogether.)
Detailed Tables:
Attention to Polls
Q. How much attention have you been paying to recent public opinion polls?
|
CANADA |
|
|
|
|
|
|
|
BASE: Canadians |
Election |
Current |
BC |
AB |
MB/SK |
ON |
QC |
ATL |
n= |
1306 |
1496 |
244 |
124 |
108 |
569 |
298 |
153 |
Margin of error (+/-) |
2.7 |
2.5 |
6.2 |
8.8 |
9.4 |
4.1 |
5.7 |
7.9 |
No attention whatsoever |
41 |
34 |
40 |
38 |
34 |
25 |
39 |
41 |
A moderate amount of attention |
25 |
36 |
28 |
32 |
37 |
38 |
40 |
32 |
A great deal of attention |
33 |
30 |
31 |
30 |
28 |
37 |
19 |
28 |
Polls’ Influence On Vote Intention
Q. Before you make your final decision, how likely are you to take party standings as reported in the public opinion polls into consideration?
|
CANADA |
|
|
|
|
|
|
|
BASE: Canadians |
Election |
Current |
BC |
AB |
MB/SK |
ON |
QC |
ATL |
n= |
1306 |
1496 |
244 |
124 |
108 |
569 |
298 |
153 |
Margin of error (+/-) |
2.7 |
2.5 |
6.2 |
8.8 |
9.4 |
4.1 |
5.7 |
7.9 |
Not very likely (1-3) |
62 |
77 |
77 |
84 |
76 |
76 |
77 |
72 |
Somewhat likely (4) |
22 |
13 |
12 |
8 |
14 |
13 |
14 |
18 |
Likely (5-7) |
15 |
9 |
9 |
7 |
7 |
11 |
7 |
5 |
Polls – Weaken the Democratic Process or Help Make Choices?
Q. Which of the following statements is closer to your own point of view? 1) The current focus on polls during the election campaign reduces elections to horse races and weakens the democratic process. OR 2) Polls are reflections of what people are thinking and provide useful information to help make better democratic choices.
|
CANADA |
|
|
|
|
|
|
|
BASE: Canadians |
Election |
Current |
BC |
AB |
MB/SK |
ON |
QC |
ATL |
n= |
1306 |
1496 |
244 |
124 |
108 |
569 |
298 |
153 |
Margin of error (+/-) |
2.7 |
2.5 |
6.2 |
8.8 |
9.4 |
4.1 |
5.7 |
7.9 |
Weaken the democratic process |
47 |
50 |
50 |
47 |
46 |
54 |
48 |
43 |
Help make better choices |
47 |
37 |
35 |
33 |
40 |
37 |
39 |
38 |
Do not know / no response |
6 |
13 |
15 |
20 |
14 |
9 |
13 |
19 |
Methodology:
Today’s poll was conducted using EKOS’ unique hybrid internet-telephone research panel, Probit©. This panel is randomly recruited from the general population, meaning that, the only way to be included in Probit© is through random selection. Unlike opt-in internet-only research panels, Probit© supports confidence intervals and error testing.
The field dates for this survey are September 29 to October 1, 2008. In total, a random sample of 1496 Canadians aged 18 and over responded to the survey. A sample of this size provides a margin of error of /- 2.5 percentage points, 19 times out of 20. The margin of error increases when the results are sub-divided (i.e., error margins for sub-groups such as regions).
All the data were statistically weighted to ensure the samples composition reflects that of the actual population of Canada according to Census data.
Click here to download PDF: election-08-polls-oct2