Archive for the ‘Polling’ Category


Obama’s Approval Rating & The Big Picture

September 7, 2012

Pres. Obama illustrates which way his approval is headed.

Today’s Gallup poll on President Obama’s approval rating was interesting for a couple of reasons.  First, and most obviously, the current results (using a three-day rolling average) show the President with a 52% approval versus a 43% disapproval.  You have to go back to  June of 2011 to find a better margin.

Second, however, the results seem to validate a criticism of the GOP Convetion — that it fell flat in terms of reaching out and persuading people outside the Republican base.  Yes, there was some pandering to women, but the bulk of what went on was either self-serving (Chris Christie) or hyperbolic nonsense (Clint Eastwood), the overwhelming majority of which was aimed at the kind of people who were already going to vote for Mitt Romney.

Consider the following daily ratings (again, using a three-day rolling average).  The dates listed are the last day included in the polling sample:

8/28: 43-47
8/29: 44-47
8/30: 45-46
8/31: 45-46
9/01: 43-48
9/02: 45-48
9/03: 45-48
9/04: 47-47
9/05: 49-45
9/06: 52-43

Of course, 8/28 was the scheduled start date for the GOP Convention, and it ended on 8/30.  The Democratic Convention ran 9/04 through 9/06.  So, even accounting for some statistical noise (MOE +/- 3%), the impact of the Democratic Convention on how people perceive President Obama was much higher than the GOP’s impact.  I’m curious to see if today’s job report has a downward effect on the poll over the next few days, but, as it stands, this poll cannot be seen as anthing but bad news for a Romney campaign that was already facing a highly likely defeat.



Well, When You Put It THAT Way . . .

February 29, 2012

I feel the same way about internal polls released by a candidate (or a candidate’s consultant) as I do about the Kardashians—they don’t matter, and I never understand why so many people seem to care about them at all.  They (both the polls and the Kardashians) are, in a word, absurd.  But, then, we expect these polls to be absurd.  After all, no candidate or consultant is going to release a poll that shows the candidate losing, and the polls are almost always worded to elicit the answer that the candidate wants.

That’s why your typical, flawed internal poll will contain questions similar to this one from the one commissioned by Rep. Tiffany Rogers (D-Stuttgart):

Question: As you make your decision to vote for State Senate are you more likely to support?
1) A member who has worked with Governor Beebe to pass the largest tax cut in state history and pass budgets that have kept our state’s budget balanced while maintaining essential services – 56%
2) A member of the opposition party to the Governor that recently attempted to hold up his budget proposals and has tried to implement multiple policies in opposition to the Governor’s agenda – 12%
3) Undecided – 31%

Wow.  That question could only be worse if the second option ended with “and also likes to kick stupid puppies in their stupid puppy faces.”  But, to repeat myself, we expect this kind of silliness from the candidate.

What we don’t expect, however, is for someone who styles himself as a legitimate news source to buy into this kind of internal poll and write a few hundred words about the poll’s validity.  Someone like, say, Michael Cook.  But, there in that link, that’s exactly what you get, including a statement about how the question of who respondents were more likely to vote for “was one of the first ones asked in the poll.” That’s the type of comment that only someone who knows a poll is flawed, but who still wants to use that poll to prove a point, would make.

Obviously, I hope that Cook is right when he says (in a forced, folksy way) that Arkansas Senate Dist. 28 “is currently as tight as a tick and it should be viewed as a complete toss-up.” I am no fan of Jonathan Dismang — because he sucks as a legislator (see, e.g.,  here and here) — and I’d like a Democrat to take that seat.  That said, I’m no more or less convinced of Rogers’ chances based on this poll, and Michael Cook should not be, either.