Articles and Analysis


Gallup 'Surge' Epilogue

Topics: 2010 , Gallup , Generic House Vote

For the last 10 days, we've been watching the bouncing ball that is the Gallup weekly tracking of the generic House ballot -- the question that asks voters if they are supporting "the Democratic Party's candidate or the Republican Party's candidate" in their district. It bounced up for the Democrats in Gallup's tracking two weeks ago, and appeared to remain up last week, but I wrote two posts arguing that the apparent "jump" was most likely random noise, especially since other tracking polls did not show a similar pattern.

Well, sure enough, the latest weekly update from Gallup out yesterday shows the numbers bouncing back in the Republican direction. Republicans now have a five-point advantage (48% to 43%), roughly the opposite of the lead indicated for Democrats for the last two weeks.

Having devoted nearly 1,400 words to this subject already, I'll keep this short: The week-to-week variation in the chart above is mostly random noise. In fact, if any real changes in vote preferences are afoot, we can't distinguish them from the random variation built into each poll. That variation, by the way, is what the "margin of error" is all about. The results above are basically a picture of 46%, plus or minus 3%.

I write this not to criticize Gallup: Their results are bouncy in comparison to some other polls because they do not weight their results by party identification, so random variation within the predictable range is inevitable.

That said, the reason we plot results from many different pollsters on one chart, as we have done at Pollster.com for the last four years, is to try to put new poll results into the larger context of all other public polls. Our national generic House ballot can be tricky, because some polls that report frequently -- especially the Rasmussen Reports automated survey -- have large "house effects" that make their results consistently different than other surveys. Sometimes results from one pollster can "fool" the chart.

However, what our chart distills from all of the available public data on the generic ballot is a slight trend in the Republican direction over the last month or so. You can see that trend even if I set our "smoothing tool" to its least sensitive setting (to minimize the impact of individual polls or pollsters):

You see the same trend even if you drop both the Gallup and Rasmussen tracking:

These are relatively small changes -- just a few percentage points movement at most -- but the changes are mostly consistent across polling organizations, which gives me more confidence that they are real than any brief "jump" in an individual pollster's results.

[Cross posted at the Huffington Post.]



Where be the Newsweek poll with Obama at 41%?

Figured Gallup poll last 2 weeks were bunk. They probably had to save face and move to the mean


Emily Swanson:

Hi Scott,

If you're looking for the Gallup/USA Today poll with Obama at 41, it's posted here: /blogs/us_afghanistan_wikileaks_usa_t.html

It's also in our Obama Approval chart: /polls/us/jobapproval-obama.html

If we missed a Newsweek poll please feel free to post a link and we'll get to it right away.



Mark, Rasmussen's large "house effect"?

Almost all other polls in the average are RV's, while Ras is LV. I didn't think "house effect" is even discernible when comparing apples to oranges. But I could be wrong.



Saying that Ras has a large house effect is like saying that the fruit Ras sells at the farmers market is inferior to the fruit sold by all the other vendors. Since Ras' fruit has much thicker skin - an undesirable trait. Ras is selling oranges, and the rest - you guessed it - apples:-)




Yup, you could be wrong [again]...



My Analysis: This is going to be an exciting election! I cannot wait!



Rasmussen's house effect is notable across the board in all their polls, including the party ID poll which is not limited to likely voters. Just because you like what you see doesn't mean that Rasmussen doesn't have a significant pro-GOP lean.


For the umpteenth time, I am writing to ask for regional breakdowns on the generic ballot test. To me, that's the only thing that would make any sense. For example, if you have R's +15 in the South, but you know that, for the most part, the only Southern seats Dems have are in large majority African-American districts, that margin doesn't matter and the R's won't pick up many seats. Winning one district by 75-25 instead of 65-35 is of no relevance.

On the other hand, with Dems +10 in areas like New England where there are currently no Republican-held seats, the same phenomenon occurs. A Dem decline in a district from 60-40 to 53-47 also makes no difference in who the member of Congress is from that district.

So we need to filter these generic results by areas of the country where a difference of +/- 3-6 points could really matter.

I have never seen such an analysis.



Nelcon, that's a reasonable question to ask, but rest assured that had your guesses of GOP regional concentration were true, political analyzerists would be all over that, like flies on honey.

Dems tired to paint GOP as a regional party, but the reverse is actually more accurate. GOP is very well distributed coast to coast, while Dems are more local than even regional. Read this:


Other hints would include looking at some of the Senate races. Even non-controversial, well-liked Dem incumbents are in trouble in some of the deepest blue states. Feingold in WI, Murray in WA. BObama's seat in deep blue Il is a tossup, Biden's seat in DE is a loss, Boxer in CA is not a slam dunk yet. Etc, etc, etc.

Don't you think that's fairly decent evidence of non-regionalness?


@StatyPolly: How anyone could doubt that the Republican Party is totally regionalized is beyond me. High concentrations in the South and mountain west and only a smattering of seats anywhere else. In contrast, Democrats have overwhelming representation in the heavily populated areas of the country such as New England, the Northeast, the Midwest and the Pacific Coast. Republicans simply can't win where there are large concentrations of people in small geographical areas. End of story.

Your Senate examples are not germane to the discussion of the generic Congressional vote, since that is relevant only to House seat analyses. However, you conveniently overlook strong Democratic/independent challenges in normal GOP states like Ohio, Kentucky, North Carolina, & Florida because of the foolishness of GOP voters nominating fringe/extremist candidates.

And by the way, if I was a betting man, I would take any bets that come Jan. 1, 2011, newly sworn in Senators will indeed include Feingold, Murray, Boxer & Reid. As for open seats such as IL & DE & NH (which you didn't mention because it's a GOP seat currently), they're indeed open for discussion as are the 3 other GOP seats I mentioned above. There's actually better than a 50-50 chance that if things break right for the Democrats on the Senate side, they could actually pick up seats this year, even if they lose IL & DE.

But the Senate is a totally separate discussion. I still think only a regional breakdown of the generic vote for Congress is of any value. These national polls, to me, have no relevance at all given the regional splits between the two parties.


The Vorlon:

Gallup certainly bounces around a great deal.

My question is to Mark. Gallup (and likely Pew) are just about the only true disciples of "The Church of the Random Sample" (To use Mark's phrase) left.

Given the huge practical difficulties in getting a semi-true random sample these days, do you think Gallup will be forced to move to some form of sample stratification along party ID lines as many others have done?



"High concentrations in the South and mountain west and only a smattering of seats anywhere else."

Dude, even with a heavy 3 to 2 Dem advantage in the House, the map looks like this.


Rest assured it will look wayyyyy redder come after this election.

Even with a huge BOBO's win, the map looks like this.


A minor scattering of blue here and there drowning in a sea of red.




Part II

Look at the fourth map on this page:


That's what a huge Dem majority looks like.

This is what a 50.7 GOP win looks like:


Dem party can't even be called regional. It's big in a handful of locations and that's it.


@StatyPolly: Yep. The Republicans win all the districts without people, while the Democrats have the public on their side. Your map shows that perfectly. I love the first map you showed (innovation). Acres and acres of no people are the red areas (do wheat or corn stalks vote? or maybe cattle and sheep?), while the blue areas happen to be where the vast majority of US citizens live.

If that wasn't true, how do you explain the fact that Obama won a massive landslide victory (thanks to real people voting) in both the popular vote and the electoral college?



Nelcon, I don't dispute anything in your last post. There are more registered Dems than Repubs in the country. I did dispute your original assertion "that the Republican Party is totally regionalized"

Region is a geographical term. Repubs are well distributed throughout the country. Dems are primarily huddled in a handful of small areas. You can walk or drive from sea to shiny sea in a variety of directions and easily avoid a Dem dominated area. Can't do it and avoid GOP strongholds. You'd have to be a heck of leaper.

Also, don't forget that the South has become the most populous region of the country, with a steady stream of North-Easterners and Mid-Westerners exodusing for decades now. Census estimates that nine more congressional seats will relocate from the North and Mid-West to the south, after this year's count. Perhaps those people are just global warming refugees:-)



Well said. The thing that is not being accepted, among others, is that any time your party gets its butt whipped in elections by 10% on the average, you will look like a regional party until you whip the other side. Its a statistical thing. Basically, duh, you will be reduced to a core of areas where you are a majority even though you have 40% or more nearly everywhere.

After November, if Republicans do as well as current polls show, dems could easily be a minority whose seats are mostly in urban conclaves, NE, and CA. They then will APPEAR to be a regional party, just as the reps do now.

I also think many of our liberal friends (I will exempt nelcon) are emotionally and politically invested in the meme of a southern Republican party because, as some point out so often, white southerners are nearly all racists (with the noble exception of folks like Morris Dees and his Southern Poverty Law Center). Southern support therefore invalidates Republicans in their eyes. Or, more precisely, they already have little or no respect for Republicans but cries of racism are their weapon of choice to invalidate reps in the eyes of others.

Careful though, note that liberal MSM is now finding that enlightened Conn apparently is full of racist whites, as demonstrated by their getting shot by a black driver who went "postal."

Note: Drudge recently published photos of all of the officers of the Southern Poverty Law Center. Every single one is white and made a minimum of $147k per year. Morris Dees also allowed some magazine to publish 60 photos of his crib, which is a multi-million dollar monument to poor white trash bling. You have to be see those photos. Vulgar taste or not, nice to be able to make that kind of money just by accusing other southern whites of being racists.

To date, charges of racism have been today's version of McCarthyism. As the accusers know so well, just to be accused hurts you, no matter how undeserved the charge. Look at the relatives of those whites killed by the black truck driver in Conn. While trying to cope with unexpected loss and grief they find themselves reduced to saying, "No, no, no, he [or she] was not a racist!", knowing very well that their truthful denials will not stop the defamation of someone they loved.

The driver shot everyone he could put in his gun sights. Amazing that they all were racists. Everywhere he aimed he found a racist to shoot. That only goes to prove that all whites in Conn are racists, obviously. Probably all immigrated from the south.


Post a comment

Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.