MyDD :: No More Rasmussen for Me

No More Rasmussen for Me

by Jonathan Singer, Fri May 12, 2006 at 02:36:38 PM EST

With the exception of an outlying Fox News Poll, every major survey since the end of April has shown President Bush's approval rating to be 34 percent or lower (source: PollingReport.com). Most recently, polls from USA Today / Gallup and CBS News / New York Times found the President's numbers to be a feeble 31 percent, and the Harris Poll released today by the Wall Street Journal shows a whopping 71 percent of Americans rating George W. Bush unfavorably with only 29 percent rating him positively.

Given the fact that there is near consensus among pollsters that President Bush's job approval numbers are in the low-30s at best, how is it that Rasmussen Reports continues to show President Bush's approval numbers in the low-40s or high-30s? Even more quizzically, how is it possible that Bush's numbers are about 10 points higher in Rasmussen Reports surveys than other such polls even after the company changed its samples to be more Democratic leaning?

Certainly it could be the case that all of the national pollsters are wrong and Rasmussen is right. Perhaps Rasmussen's automated polls are just more accurate than traditional polling operations; after all, Rasmussen performed quite well in 2004. That said, the aforementioned possibility is fairly unlikely.

So what does this mean? For me, until I see some change in the results of the Rasmussen polls -- until their data come closer to the margin of error rather than being 10 points off the general consensus -- I'm not going to post on their results, even from statewide elections. It's quite possible, for instance, that Rasmussen's statewide polls skew as far to the right as do their polls on the President's approval rating. My decision doesn't mean that you won't be seeing the results of Rasmussen polls on this site, as there are more than one of us writing for MyDD. What's more, it's important to note that I'm necessarily correct, statistically speaking, in making this decision. But I'm not the first blogger to make such a decision and might not be the last to do so, either. So unless I see hear a strong argument as to why I'm wrong, it's going to be a while before I post the results of another survey from Rasmussen Reports -- if I ever do.

Tags: Rasmussen Reports, Polls, Pulse (all tags)

Permalink :: 21 Comments :: Post a Comment

Independents (none / 0)

I think I heard a while back (will try to find the source) that Rasmussen's likely voter model includes far fewer Independents than do the models of other major pollsters. Their rationale is that Indies are far less likey to vote than partisans on both sides.

As Chris has pointed out, Independents and Democrats are closely aligned. So cutting the number of Independents in a poll is like removing a bunch of Democrats from the sample.

I think Rasmussen's numbers are valuable, but only if interpreted through this lens and only when compared to the control group of other polls. Their data adds more nuance to Bush's dismal numbers, shedding light on how much those numbers reflect dislike of Bush among Democrats vs. Independents vs. Republicans.

by swatdem on Fri May 12, 2006 at 02:50:25 PM EST

Re: Independents (none / 0)

As much as I agree his polls are off, I think it's a bad decision to choose to ignore it. Fact is he was pretty good last election, and we can't just look at the polls we like to see, we have to look at them all.

by MattyR on Fri May 12, 2006 at 02:58:05 PM EST
[ Parent ]

Re: Independents (none / 0)

This is exactly right and explains the entire discrepancy with the other national polls. There is no need to imagine any conspiracies or to assume Rasmussen is incompetent. This is not complicated.

I happen to believe that Rasmussen is wrong about likely independent turnout this November, that he's being way too harsh. But his is a fairly straightforward assumption that is not without historical antecedent. It's unfair to flame him or suggest his work is without value.

by ColoDem on Fri May 12, 2006 at 04:19:26 PM EST
[ Parent ]

That's not the problem (none / 0)

The problem is that they weigh by party, and lately a lot people have stopped calling themselves Republicans!!!! They finally realized that this was wrong, and changed the weighing system, by adding Democrats and subtracting Republicans (or more precisely, it looks like they are going to anaylze thier long term trends to see if there is a shift).

From thier homepage today:

"One of the major methodological disputes among public opinion pollsters involves weighting data by Political Party. All agree that partisan affiliation is one of the best indicators of voting intentions and perceptions of the President. However, some firms and academic researchers believe that party affiliation changes on a regular basis. At Rasmussen Reports, we do not. We, along with many others, believe party affiliation is generally stable and that people switch their allegiance only rarely.

This view is supported by data and by common sense --how many people do you know that switch political parties on a regular basis?

The challenge for a polling firm is determining the proper partisan mix for a polling sample. What percentage of Americans are Republican, Democrat, or unaffiliated? What are the "right" numbers?

Beginning this past Sunday, Rasmussen Reports Job Approval updates are based upon data using a slight modification to our targeting and weighting process. From this point forward, we will set our partisan affiliation weighting targets based upon survey results obtained during the previous three months. These shift only modestly month-to-month, but the change could be significant over a long period of time.

Based upon the past three months, the current targets are 36.6% Democrat, 33.5% Republican, and 29.9% Unaffiliated. These targets will be updated monthly. Previously, our weighting targets assumed an equal number of Republicans and Democrats.

We have adopted this system because we believe it allows us to maintain the day-to-day stability needed to follow trends while adjusting periodically for any substantive shifts in partisan affiliation (see trends in party affiliation).

The practical impact of this revision is modest in the current environment. The new approach will result in the President's reported ratings being a point or two lower than they would have been under the old system.

Data in the table to the right shows month-by-month Job Approval averages using the new weighting methodology. Each month's number is based upon interviews with approximately 15,000 American adults. The full month average for April will be released tomorrow (Monday). At that time, we will also update our month-by-month summary of shifting party affiliation among the general public."

by Geotpf on Fri May 12, 2006 at 09:14:34 PM EST
[ Parent ]

Robo calls? (none / 0)

Seriously who listens to these? Whenver I have heard that second of silence and the computerized voice I hang up. I am assuming that Rasmussen's are not much different.

9 year and lonely 79 year shut ins are probably the only people responding to Rasmussen.

by dpANDREWS on Fri May 12, 2006 at 03:01:54 PM EST

Re: No More Rasmussen for Me (none / 0)

I don't think anyone noticed this, but just after the beginning of May, Rassmussen had been showing President CokeSpoon below 40% approval for several days in a row, they then decided that they needed to change their methodology for analysis, after the switch, no more results under 40% and it's been that way all month. They even appear to have recalculated past analysis to eliminate all indications that even their rightwing skewed polling had Chimpy dropping like a stone. Anyone else have info on this - next time I see that kind of thing I'll remember to grab a screen shot, and not just rely on the rest of the reality-based community to do it for me.

by bluebear on Fri May 12, 2006 at 03:13:11 PM EST

Re: No More Rasmussen for Me (none / 0)

Rasmussen leans right. It makes no bones about being a Republican poll.

That said, the only good a current Rasmussen poll is comparing it only to other Rasmussen Polls. After their adjustment, they restarted at 40%, following a couple of days at all-time lows of 39%.

In total, today's rating is an all-time LOW approval rating, and is accompanied by an all-time HIGH disapproval rating of 60%. This poll is also useful to measure the loss of his Republican base, which is commented on almost daily.

by ocdemocrat on Fri May 12, 2006 at 06:29:32 PM EST
[ Parent ]

Then I should be really encouraged... (none / 0)

...by Rasmussen's last round (April 19) of Ohio polls, which gave Ted Strickland a 52-35 lead over Ken Blackwell in the gubernatorial race and had Sherrod Brown down only 43-41 against Mike DeWine in the Senate race.

by KTinOhio on Fri May 12, 2006 at 10:01:34 PM EST
[ Parent ]

Re: Then I should be really encouraged... (none / 0)

The current Rasmussen polls are even better news for our side, as Brown has pulled ahead by 3. Not that it matters...

by KTinOhio on Sat May 13, 2006 at 07:59:33 AM EST
[ Parent ]

Re: No More Rasmussen for Me (3.00 / 2)

I don't believe that the point is that Rasmussen's
voter sample is different, but rather the
question that Rasmussen asks is different.

For example, Harris Interactive reported that
29% say President Bush's performance is ``excellent or pretty good'' and 71% say it is ``only fair or poor.''

I think this wording makes it easy for many people who think President Bush's performance is about average to put themselves in the second group (and
say he is ``only fair.'')

Other polling companies ask a somewhat different
question, and this leads to somewhat different results. Rasmussen, for example, asks people if
they ``strongly approve, somewhat approve, somewhat disapprove, or strongly disapprove'' of
President Bush's performance. Probably some
people who answer one question by saying that President Bush's performance is ``only fair'' answer the other by saying that they ``somewhat approve'' of his performance. The difference in the choice of question may well be the biggest
factor in the different results that the different companies get.

by strings on Fri May 12, 2006 at 03:37:43 PM EST

Re: No More Rasmussen for Me (3.00 / 1)

Isn't this sort of silly? Why don't you just preface any post about Rasmussen with the point that they're consistently show 10% higher approval for Bush. Maybe they're more accurate, maybe they're less accurate but neither would justify keeping information from readers.

by js noble on Fri May 12, 2006 at 03:43:45 PM EST

Re: No More Rasmussen for Me (3.00 / 1)

I dunno, this sounds like something the Bush Administration would do -- we don't like their results, so we won't even acknowledge they exist. Doesn't seem very honest.

What's wrong with reporting all the results, and pointing out that Rasmussen appears to be an outlier because its results are consistently farther to the right than all of the other polling operations? Honest and informative.

by steveh2 on Fri May 12, 2006 at 03:44:48 PM EST

Going cold turkey on Rasmussen?? (3.00 / 1)

Then I'll have to post this, from Rasmussen via Political Wire:

OH-SEN
Brown 44%
DeWine 41%

Matt - DemConWatch

by msn1 on Fri May 12, 2006 at 03:48:03 PM EST

Re: No More Rasmussen for Me (3.00 / 1)

Speaking as just one reader, I would rather see the Rasmussen polls, and discount it as appropriate given that they seem to be an outlier on presidential approval, rather than simply pretend they don't exist.

"Another problem we have...is that in election years we behave somewhat as primitive peoples do at the time of the full moon." --Harry Truman

by Steve M on Fri May 12, 2006 at 03:57:06 PM EST

Silly, R seems honest to me (3.00 / 1)

The wording of his question is harsher than the others, so Bush wiill have slightly highr numbers.

Preference polls of this sort aren't voting/candidate polls at the end of the line, however. Someone who thinks the president is only fair might be more likely to vote for him than someone who slightly disapproves. Who is to say which wording is more accurate. I think R's, but I'm no expert.

I have also had the impression (maybe incorrect) that Rassmussen pushes the respondents a bit to make a decision. This might increase bushes number a point or so also.

Relax. Don't take it personally or be too paranoid. It's information for us to use with other information.

A point or two one way or another with a president well below 50, however approval is defined, isn't going to affect fund raising or arganising very much.

by Reptile on Fri May 12, 2006 at 04:02:22 PM EST

Re: No More Rasmussen for Me (none / 0)

While I am sometimes annoyed at the partisan leaning commentary on the Rasmussen site, and I agree that their numbers do tend to be more Bush-leaning than they should, I find the fact that Rasmussen polls "every F-ing Day" to be an amazing thing. If you visit them every day you might not get the accurate numbers, but you are most likely to understand the trends.

That said, I also am annoyed when Zogby makes partisan statements too. Political opinions have no place in Statistics.

by surrendering on Fri May 12, 2006 at 04:14:21 PM EST

Re: No More Rasmussen for Me (none / 0)

Link to and discuss Rasmussen polls, but only in the context of other polls. They're still useful when it comes to confirming trends... really, all polls should be looked at in context with the other polls released on a race.

by Newsie8200 on Fri May 12, 2006 at 04:30:29 PM EST

Re: he hides the internals (none / 0)

IMO, if Rasmussen didn't hide the poll internals behind a subscription firewall, he might have more
credibility as one could analyze the numbers, the questions and make comparisons with other polls.

by phillydem on Fri May 12, 2006 at 04:53:10 PM EST

Off Topic: PJM (none / 0)

Jonathan: since I don't know how to email you, I'm posting this here.

Pajamas Media has started putting up some of our posts in such a way that they look as though we post at PJM. They did this to one of yours, and I thought you should know, in case you weren't OK with this. (I'm not. I've written them; no response as yet.) Your post on their site is here.

by hilzoy on Fri May 12, 2006 at 05:19:20 PM EST

Re: No More Rasmussen for Me (none / 0)

Well, they're not always bad news...

http://politicalwire.com/archives/2006/0 5/12/in_ohio_brown_pulls_ahead_of_dewine .html

"In Ohio's U.S. Senate race, Rep. Sherrod Brown (D) holds a three-point lead over Sen. Mike DeWine (R), 44% to 41%, according to a new Rasmussen Reports poll. "

by musa on Fri May 12, 2006 at 07:26:39 PM EST

Rasmussen Is A Tool (none / 0)

A semi-useful tool, but a tool none the less --

Whether it is by skewed sample, based on self identification or voter registration or whatever -- Rasmussen represents the GOP outlier trend at the moment.

Given the Democrats ineptitude, isn't it possible that Party ID will firm up in this direction by November?

Not that our team doesn't have the wind in our sails -- but do our team leaders know how to take advantage of it?

by ck on Sat May 13, 2006 at 01:10:30 AM EST

You are not logged in.

In order to post a comment, you must be logged in. If you have a member account, please log in to comment.

If not, you can make an account right here. It's quick and free.