Sunday, November 03, 2013

How popular is Rob Ford? An Update

On November 2 Forum Research released another poll taken just after the dramatic news about the Ford video on Thursday. That poll is now being reported, for example, by CBC as "Ford Approval goes up despite video scandal".  This is simply not an accurate portrayal of the state of public opinion.

Here are six months worth of results from Forum's approval question:

Poll date
Approve
Disapprove
Oct. 31st
44
56
Oct 28-29th
39
61
Sep 23rd
49
51
Aug 29th
44
56
July 29th
46
54
June 25th
47
53
May 24th
42
58
May 10th
44
56
April 11th
43
57

Average

44

56

As you can see the Oct. 28-29th poll number is the exception here not the rule.  Forum's own headline on its release is closer to the facts in saying "Ford's approvals unaffected by controversy".  However, the same poll also stated that a huge majority (a margin of 60 to 36 per cent) say he should resign. This fact is further down in the CBC story but deserves equal emphasis. What looks like some degree of confidence really isn't.
There is an additional problem with this poll and all other instant polls.  Public opinion takes at least some time to form. It may well be that the approval rating won't change but a survey on that should not be conducted until a number of days have elapsed. Media constantly seek dramatic headlines and therefore make mistakes like the one cited above. They ought to give more reflection to how they report all polling.

And as I discussed in my previous post the Approve/Disapprove in this poll does not distinguish between those who strongly or just somewhat approve or disapprove. Those who make use of public opinion put much more emphasis on those who express strong views.  The other series of questions Forum asked in the Oct. 28-29th survey gives us better insight on that and it is not favourable to Mr. Ford. It is clear that a far greater share of citizens in Toronto strongly dislike him than greatly admire him.

ONE MORE COMMENT: the poll that gave us reports on Ford's approval rating that caused the media to say it was "going up" was taken on Halloween.  That is not a good night to conduct a public opinion poll.  I also noticed some of the internal crosstabs don't make sense. For example, Ford had more support among younger respondents, less among older, which is completely inconsistent with earlier surveys. This particular poll should be discounted.