Wednesday, September 8, 2010

Focusing On An Anomaly

The world of DC politics is all atwitter about the latest Gallup poll on the Congressional Generic Ballot:

A week after leading the charge of polls showing the GOP with a huge lead in generic congressional ballot polls, Gallup is out with new numbers today that dump a bucket of cold water on that very idea.

According to the latest Gallup survey of registered voters, Republicans and Democrats are tied on the congressional ballot question with each drawing 46% of the vote. Last week's poll showed the GOP ahead by 10, leading 51-41.

Narrative-Busting New Gallup Poll Excites Dems Briefly

I made quite a lot of last week's Gallup poll, and I still think it's worthy of the same comments I made. Let's look at this week's Gallup chart for some explanation (click on the graphs to enlarge):
Image credit: Gallup poll

As you can see, this latest data point looks rather anomalous. Last week's, while definitely a departure from the previous weeks', were at least in the ballpark. Such a sudden rise needs more of an explanation than Gallup could manage, which was this:

Last week marked the return of President Barack Obama from his 10-day vacation, and included his national address to announce the official end of combat operations in Iraq. The president's three-day job approval rating rose to 47% for Aug. 29-31 -- a level it had reached only once since mid-July. Last week also brought media commentary in the aftermath of conservative talk-show host Glenn Beck's massive rally in Washington, D.C. It is not clear if these or other factors affected Americans' voting preferences as measured by the generic ballot.

Parties Tied at 46% in Generic Ballot for Congress

I think it's safe to translate that last sentence as "No, we're not buying this, either."

The change is large relative to their margin of error, so it may represent some actual change in the electorate's opinions, but I'd like another data point or two before reaching that conclusion. Here's what Rasmussen's weekly generic congressional ballot has looked like this year, as plotted by my Open Office software:
Image credit: Plot by Cujo359 from Rasmussen data

If there's any recent trend in that data, it's a growing dissatisfaction with both parties - a dissatisfaction I share. [Er, umm, see UPDATE below.] Other than that, I don't see any big gains or losses by either party. Many poll watchers think that Rasmussen tends to favor Republicans more than Gallup, and that seems to be true here. The trend, though, is what's important, and at least among Rasmussen's poll targets, there isn't one that's helpful to Democrats.

Last week's Gallup poll result, even if it turns out to be somewhat anomalous, is important as part of a general trend. It also demonstrates how low the Democrats' support can look in a given poll. This one looks like anomalous data. The Democratic politicians might be making something of this, but there's nothing there that I can see yet that's worth celebrating. That's not to say this can't be the start of a Democratic resurgence, because it could. I just don't see any real reason to think it is, and even if I'm wrong, that Gallup chart shows plenty of transitory ups and downs.

Only a fool would take this seriously on its own.

UPDATE (Sept. 13): I need to be more careful with homemade graphs - the X axis is backward. That means the most recent data points are to the left, where there is a clear swing in favor of Republicans that has occurred in the last few weeks. This tends to support my contention that the most recent Gallup poll was an anomaly. The Rasmussen poll looks like the trend of the previous Gallup poll.

In fact, the Rasmussen poll has the spread as being 12 percent in the 9/05 generic ballot.

Apologies for the oversight. I was so happy to have finally made a graph using a spreadsheet program that I didn't notice that it made it backwards. Sigh.

No comments: