Polls have become simply a way to package an editorial position as front-page news. And news organizations have sunk even lower than that. Not only are the polls deeply flawed, but the reporting on the polls is false and misleading.

Case in point: the WSJ/NBC poll (pdf) produced on the eve of the filibuster showdown to “frame” the nightly news coverage. (The poll is also page one in the WSJ, and while their editorial page leans right, the news staff leans left.) Here’s the headline from the NBC News website:

Voters dissatisfied with Bush, Congress

NBC/WSJ poll reveals ‘angry electorate’

The first problem is that the poll referred to in the headline is not, for the most part, a poll of voters at all; it is 1005 adults, for almost all its questions. Moreover, the poll dsiplays the reddest of red flags: it does not reveal its demographic mix. We won’t be surprised if the reason the poll’s demographics have not been released along with the poll results is that they have been cooked to weight the poll to support the “news” story. (We’ll report back, one way or the other, once the full demographics are released.) Here’s the way that NBC’s website is framing the issue:

Senate Majority Leader Bill Frist, R-Tenn., speaks to reporters in the Capitol about judicial nominees on Tuesday. The latest NBC-Wall Street Journal Poll indicates that a majority of voters believes the Senate should make its own decision about the president’s judicial nominees, rather than just confirming them.

Ah, but that is untrue; the poll says no such thing. That question about judicial nominees (#11) has no result for registered voters, only for adults. Thus, we have no idea what voters think at all from this poll question, or indeed from this poll, which will be used to scare up all sorts of hobgoblins. Further examples: anti-social security reform 56%, Congress should be investigating Tom DeLay 52%, Congress is has different priorities than you 65%. Oh my, the people are angry. That’s certainly what Democratic pollster Peter Hart, whose firm participated in the poll, wants you to think, via NBC:

Perhaps the most revealing finding in the poll is the attitude toward Congress. Just 33 percent of the respondents approve of Congress’ job. That’s down 6 points since a poll in April and 8 points since January.

“The public is exceptionally displeased with the Congress,” Hart said. “It is [its] lowest set of numbers since May of 1994,” the year when congressional Republicans defeated their Democratic counterparts in the midterm elections to take control of both the House and Senate. According to this poll, by 47 percent to 40 percent the public says it would prefer Democrats controlling Congress after the 2006 elections.

And that, dear children, is the single greatest lie in polling. As Scott Rasmussen told us last year, the generic congress question typically favors a Democratic Congress. Indeed, the Peter Hart poll immediately prior (10/29-31) to the 2004 election showed a 44/43 plurality in favor of a Democrat-controlled Congress. You would think greater modesty would be forthcoming from a polling organization that can’t get the numbers right four days before the election.

Maybe the electorate is angry. We don’t think so, but we claim no special knowledge. Based on the flawed and deliberately misleading reporting above, we wish the pollsters and the news organizations adopt a more appropriate modesty and respect for truth.

This entry was posted on Thursday, May 19th, 2005 at 7:36 am and is filed under Democrats, General, MSM, New Media, Polling. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.