It’s Turn of the Skew day here in the PostScript bunker, where we debate who is manipulating polling data to fool whom. More than 5,000 commenters weighed in on Eugene Robinson’s column about this.

Robinson, in his mainstream-media way, wants the reader to keep believing in mainstream-media polling and math, even though there is now some non-mainstream polling (PostScript assumes it would be biased to call it “fringestream”) to disagree with it.

As PostScript understands it, the issue is the number of people sampled who self-describe as Republican or Democrat. Mainstream pollsters, Robinson says, try to predict the voting population using factors other than party identification (age, income, race, location, education). When they do this, they end up with a large number of self-described Democrats in their sample. Those who would “unskew” the polls weight these results assuming closer to 50-50 party identification, trying to replicate the ratios of Republicans to Democrats that voted in the last few elections.

Robinson argues that big media polls are more objective because they don’t assume some ideologically precise numerical balance among voters between Republicans and Democrats. This is more realistic, Robinson says, since it doesn’t try to predict the party ratio. (Adjusting for 50-50 does show Romney doing better.) But, as many commenters contend, Robinson’s mainstreamity and his polls’ mainstreamity are both reasons to think he’s a big liar trying to get Obama reelected.

BayouRod, in fact, thinks Robinson doth protest too much:

Since Obama is definitely going to win, why is Eugene still spending 100% of his time shilling for him?

On the other hand, says JoeT1, it’s dangerous to lump all the polls agreeing with each other as “mainstream,” because you get flagrant errors in calculation:

Or as another blog notes, why does the FOX poll show Obama up by 5? When did Fox join the MSM conspiracy?

SheepCheese follows to a truly mindbending conclusion:

Fox IS part of the mainstream media, as much as they try to pretend otherwise.

While your head’s exploding, try and metabolize this one too. Flyover22 says, PostScript thinks she has this right, that there’s heroism in running a campaign that voters end up not favoring:

There are always two ways to run an election. Run on principle or run on polls. We all know Eugene’s bent. If one has a different opinion about policy, principle, or direction so be it.

To denigrate a candidate or a group because they don’t change their position based on some polls is low. That what the election is for. If the message is wrong, or it’s the right message at the wrong time, the election will sort it out. I prefer that instead of all agreeing on the polls and just voting on who has the best marketing campaign.

Jiji1 argues for the “unskewed” polls, because mainstream polls have always been wrong and nobody’s ever corrected the problem:

Wouldn’t you think it would be smart to tie the percentages in your poll to actual percentages seen in voting in 2008? When Dems had a pretty impressive turnout? Sadly, this is not the case. If you look at some of these polls they are using a +9 or +10 democrat advantage, when that didn’t even exist during the highest Dem turnout in 2008. That is why a lot of us are skeptical of these polls.

Need we go back in time to 1980 or 1984 or even 2000 and 2004? There were significant errors in polling in these elections...always favoring the democrats. I find that really intriguing. Why does it always seem to skew democrat and then turn out to be horridly off-base? That is what you should be questioning.

Guess we’ll all see in November.

And gardyloo is funny:

I asked 100 friends whether they agreed or disagreed with the premise of Mr. Robinson’s article. Seventy-four agreed and 49 disagreed.

PostScript thinks there is indeed something funny about all the mainstream polls sliding toward Democrats all at once, and she’s not the only one. When Paul Krugman and The Fix’s Chris Cillizza both admit they don’t quite know why the electorate is not behaving how they predicted, it is time for some serious chin-stroking.

One lefty commenter’s explanation is about as unsatisfying as the “skewed polls” one:


The main reason that D’s are ahead in most polls in most places is because the citizenship is waking to the facts and the arithmetic of the economic discussion. The secondary reason — the independents are waking to the notion that the GOP has lost its moral compass and in many cases — its mind.

PostScript finds that wishful thinking, too. The Dem surge in the polls reflects a real sea change, but she doesn’t think anyone’s quite figured out why yet, and it leaves both sides scrambling for explanations that make their side feel good.