There's been a good bit of buzz this week, especially on the Left, concerning the results of the latest Washington Post/ABC News survey, which found an apparent spike in President Obama's public approval rating, to 54 percent. The 54 percent figure represents a five-point increase over the results obtained in December 2010.
The survey also found what appears to be the first big slide in public perception of the Tea Party, with a net disaproval rating of 52 percent and 35 percent approval. The disapproval number represents a seven-point increase, compared to the results in August 2010.
I suggested yesterday on the Chris Jansing show on MSNBC that the Post/ABC survey sounded like one of those occasional outliers that crops up as public opinion is measured over time, but noted that I hadn't had an opportunity to look into the demographics of the survey sample.
Now that I have had a chance to look at the demographics, I still think there is something of an outlier feel to the survey. I say that mainly because of the self-identification numbers of liberals, moderates and conservatives. With conservatives making up 40 percent of the electorate and liberals 20 percent, according to Gallup, the Post/ABC respondents appear to line up fairly well, with 22 percent of them calling themselves liberal and 36 percent conservative.
But the survey doesn't account for how many of those who choose to call themselves moderates, 38 percent, are actually liberals, just as it doesn't account for how many of those who declare as conservatives are actually libertarians.
Gallup puts the three at 21 percent liberal, 35 percent moderate and 40 percent conservative. If we assume the three-point differential in moderate identification to represent liberals "hiding" behind the moderate label, it would boost the liberal percentage in the Post/ABC poll from 22 to 25 percent.
Combined with the four-point differential on conservatives between Gallup and Post/ABC, you might have a sufficient demographic over-representation in the Post/ABC sample to account for at least some proportion of the marked changes in Obama and Tea Party approval numbers.
Note that I said might. There is no way that I am aware of to account statistically for how many liberals or conservatives might be hiding in the Post/ABC sample, so this is purely speculation on my part. But politics is both science and art, and I still have a gut feeling this survey will look rather outlier-ish as 2011 progresses.
I say this because the fundamental alignments that shaped the 2010 congressional election have not changed in any measurable degree since the election and, while transitory events like the Tucson Massacre can move the numbers temporarily one way or another, the overall trend is almost certainly going to remain as before.
You, of course, should judge for yourself and to do that, go here for the Post/ABC demographics, which are found at questions 901-908.