Mystery Pollster

Job Approval at Midterm

Earlier in the week, MP posted links to the latest surveys by USAToday/Gallup and CBS News and anticipated an updated trend chart that our friend Charles Franklin delivered as expected later that day.   Unfortunately, a challenging week of travel and day-job distractions prevented me from checking Franklin's site until today - the updated chart appears at the end of this post.  However, I also discovered an amazing analysis that Franklin did earlier in the week comparing the current Bush job rating to the mid-term ratings of other Presidents since 1950.   His conclusion is a bit stunning, even for those of us who obsess over such things: "We have simply never seen a president this unpopular going into a midterm election."

As usual, one of Franklin's pictures is worth a thousand words.  Here is the money graphic following by his description:

President Bush's approval rating is on course to set a record low for mid-term elections.  The magnitude of the problem is greater than commonly perceived. The previous record low approval in the last Gallup poll of October was 41% for President Truman in 1950. Based on approval trends in 2005-06, the President and Congressional Republicans are facing an election day 2006 approval of between 20.4% and 40.8%.

He goes on to explain how he used past trends in the Bush job approval rating to try to project a range of potential ratings conceivable by October, illustrated as the line at the bottom of the chart (above) that extends in both directions from the "GW Bush 2006" point.   But whatever guess one wants to make about what the Bush rating will look like in October, it is hard to argue with his conclusion: 

I was frankly shocked at the above results. Other presidents have suffered low approval ratings, and President Bush still stands above the lows of four of the ten other post-war presidents. But I had not appreciated how much the current approval is below other mid-term approval ratings, even without extrapolating current trends. We have simply never seen a president this unpopular going into a midterm election.

Well, actually, as a commenter on Franklin's site points out, he started with 1950 and thus missed the low of Harry Truman -- 33% in September 1946.  That is not company George Bush wants to share.  In 1946, Republicans won 55% of the congressional vote and took control of both houses of Congress and a majority of state governorships. 

Meanwhile two new polls have been released since my last post earlier this week:  The R-T Strategies poll puts the Bush job approval rating at 36% (results, crosstabulations, presentation) and a new AP-IPSOS survey out today reports it at 33% (AP story, results).   Franklin promises an updated trend chart later today, but meanwhile, here is his version from earlier this week:

Trackback SPAM

Unfortunately, a "housekeeping" item chewed up my blogging time today.  I discovered this morning that MysteryPollster has been receiving several hundred pieces of "trackback" spam every day for since mid-March.   Now, I have quite a mess to clean up. 

A trackback, for those unfamiliar with that particular term of art, refers to the mechanism that allows bloggers to link to and comment on each other's posts simultaneously.  Trackback links appear on this site just above the individual comments left by MP readers.  Unfortunately, the Spammers that have bombarded our email accounts are now hard at work doing the same to the comments and trackback facilities on blogs.

MP has also attracted a smattering of "comment SPAM" -- commercial or otherwise totally off-topic messages left as a comment  -- but it has been so light and sporadic that I was content to simply delete it when it appeared.  However, the trackback spammers posted their deluge of garbage links on MP posts that are more than a few weeks old, so I was unaware of literally thousands of incoming trackback links that mucked up this site. 

As a result of this barrage, I have been experimenting today with the various features offered by Typepad, the service that hosts this blog, to block comment SPAM.  My apologies for any unexpected weirdness to anyone who attempted ot post a comment earlier today.  One change I have put into effect today:   Readers that enter comments will have to "pass a test," as the folks at Typepad put it, "to validate that they are a person and not a machine."  You will be asked to enter some letters and numbers that will appear in a small graphic.  Anonymous comments will remain possible.

Also please note this one minor addition to MP's informal but evolving comment policy:  Comments or trackbacks that refer or link to products or services for sale, that post random news items or include jibberish text unrelated to the topic of the post are all subject to deletion without notice. 

PS:  Readers may have also had difficulty accessing MP on Tuesday night.  That was apparently the result of "sophisticated denial of service attack" aimed at the service that hosts this site.  The Typepad blog has details. 

Fooled Again?

MP received email from several readers over the last few days asking about a rumor that the television networks had decided to abandon exit polls for 2006 elections.   The blogosphere rumor apparently originates with blogger Mark Crispin Miller (author of the book Fooled Again, How the Right Stole the 2004 Election & Why They'll Steal it Again) who provided the following seemingly authoritative report last week:

This year there will be no official exit polls conducted. The media consortium that paid for them in previous elections isn't going to do it any more, ostensibly because such polls have been "exposed" as unreliable; and so we will have no way to determine whether the official vote-counts can be trusted.

He seemed pretty sure, but it was news to me.  So I went to the source and sent an email to the official spokesperson for the National Election Pool (NEP), that news consortion that conducts the exit polls.  I received the following official, on-the record response:

The rumors are false.  NEP will be conducting exit polls during the off year election.

There you have it.  I've also heard unofficially that the exit pollsters are hard at work planning for 2006.  So it looks like we'll have network exit polls to kick around again in November 2006 (although they promise to hold back the results, even from reporters, until 5:00 p.m. -- so maybe no mid-day leaks this time).

This rumor episode tempts me to remind readers about placing too much faith in similarly authoritative theories about how the 2004 exit polls are evidence of a stolen election.  But that would be awfully snarky of me, wouldn't it?

PS:  While the networks will conduct exit polls in November, as far as I know they will not  conduct exit polls in connection with any of the primary elections, including tonight's Ohio primary.   

However, those concerned -- with good cause -- about the problems with electronic voting and the integrity of our voting system might want to check out this Cleveland Plain Dealer story that mentions what sounds like an innovative exit poll being conducted today in Ohio's Cuyahoga County by the Election Science Institute (ESI).   This sort of work has far more potential to help us monitor and improve the conduct of our elections than pouring over the remnants of the 2004 exit polls, which were simply not designed to check for fraud.  Hopefully, we will hear more about the ESI project soon.   

Et tu, Gallup?

Two new surveys suggest the slow but steady decline in the Bush job rating continues.  The 34% job approval rating on the USA Today/Gallup poll (see USAToday story, results and Gallup analysis) and 33% on the CBS survey (story, results) both represent "new lows" for each organization.  Although the difference from the previous low on both polls is small (and thus, the usual caution about "new lows" applies), the downward trend across all public polls remains for the most part steady and consistent.  The results of both surveys appear to fall below the trend line that Professor Franklin updated a week ago.

MP recommends reading the analysis by Gallup's Frank Newport and Joe Carroll today (before it disappears behind Gallup's subscription wall tomorrow). They roll together all of Gallup's data to confirm the slow steady decline in Bush's rating in 2006.  This analysis represents a subtle change.  A month ago, Newport looked at results from three successive polls conducted in February and early March and concluded that Bush's job performance had "not undergone a dramatic free fall" but rather appeared to be entering "another period of stability" following a "small shift."  Franklin, who plotted and analyzed the data from all the public polls, disagreed. 

In this latest analysis, Newport and Carroll roll together their data and look at monthly and quarterly averages and conclude that "Bush is losing about one point per month because his monthly job approval averages have fallen 16 points in 16 months since January 2005."  In the first four months of 2006, they show the job rating falling from steadily:  43% in January, 40% in February, 37% in March and 36% in April. 

Source: Gallup Organization

Rasmussen Adopts Dynamic Party Weighting

As promised here on MP, pollster Scott Rasmussen yesterday announced a new party weighting procedure for his automated tracking survey:

Beginning today (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. Today's reading [of 38% job approval] would have been 39% using the old approach.

Thus, visitors to Rasmussen Reports will now benefit from the following new features:  Regular disclosure of the party weight target and the knowledge that Rasmussen is weighting to reflect a recent overall average, not some relatively arbitrary and (previously) hidden target.  That's a big improvement. 

Another added plus:  Rasmussen has added a table providing monthly averages for the Bush job rating on his surveys dating back to February 2005.  That table will remain, with a monthly update, in addition to the usual 2-3 week window of 3-day rolling average results. 

For all the controversy over Rasmussen's automated methodology, the biggest limitation of his free updates from my perspective has been that he only provides a 3-day rolling average.  The day-to-day variation usually looks like mostly sampling "noise" to me, yet the occasional random tick up or down inevitably leads to fruitless speculation somewhere in the blogosphere.  One suggestion for Scott Rasmussen:  Add a regularly updating last-7-day or last-30-day average, as those occasional ticks up or down in the 3-day results will look even more deceptively dramatic in comparison to the very stable 30-day averages. 

Finally, one quibble with Rasmussen's characterization of the debate over party weighting:

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?

To be fair, most that oppose weighting by party would also agree that "party affiliation is generally stable and that people switch their allegiance only rarely."  The issue is that a small percentage of Americans will switch their "allegiance" more easily, usually moving from a one party to the independent category and back again but rarely all the way from Democrat to Republican (or vice versa). I have written previously about evidence of this sort of movement, gathered by surveys that interview the same respondents more than once over a period of a few weeks or months.  Other research has shown evidence that the wording of the party identification question, the answer categories offered (see Franklin) or the questions that come before it may also alter the answer respondents give. 

Yes, when we aggregate large amounts of data over the course of a month or year, the day-to-day variation disappears and what trends do remain (such as those seen since early 2005) tend to be glacial (also documented exhaustively in this highly regarded book).  But some real yet fleeting variation may still occur on a day-to-day basis.  We have also seen some evidence of short term jolts (such at the aftermath of 9/11 or the 2004 Republican convention).  The problem is that when looking at individual surveys, we usually cannot tell how much of hte short term variation in party ID is real and how must just ordinary sampling "noise."  That's really the nub of the debate, which will certainly continue.

But let's give Rasmussen his due for moving toward greater disclosure and adopting a more sensible approach to party weighting.  That's progress.