Dalia Sussman

The New York Times today touted two polls that supposedly demonstrate support for the Democratic position on unemployment benefits. But a further examination of the poll questions reveals that their findings were inaccurate; the questions misrepresented the issues at play, and the Republican position on the matter.

"Two national polls published last week suggest that most Americans are on [Democrats'] side of this debate," wrote Dalia Sussman. How she knows that fact is a mystery, given that the GOP argument -- that benefits should be extended and paid for with unused stimulus funds -- was never offered as an option to those polled.

Both polls asked, essentially, if respondents thought it was more important to extend unemployment benefits, or to preserve PayGo rules. Majorities said they thought extending benefits is more important. But under the GOP plan, the two are not mutually exclusive. Nowhere in either poll were respondents asked whether they would favor paying for extended benefits with unused stimulus funds. Neither the Times nor anyone else can accurately claim that voters favor one approach over the other since the GOP position was not an option.

New York Times Reporters Jeff Zeleny and Dalia Sussman broke down the latest NYT/CBS News poll Thursday and found Obama's favorable ratings holding steady, but support for his actual policies in the red: "Obama Poll Sees Doubt On Budget And Health Care -- Overall Support High."

Usually the Times team of Adam Nagourney and Megan Thee handle the poll stories, and usually Obama comes off looking great. Perhaps the switch to Zeleny and Sussman helps explains why today's off-lead poll story is less laudatory of Obama, although that might also be a recognition that his numbers aren't quite as favorable this time around.

One thing hasn't changed: The poll's pro-Democrat "weighting" continues. There were complaints in early April, the last time CBS News and the Times teamed up for a poll, that the poll's "weighting" process produced far more self-identified Democrat than Republican respondents, which would certainly tilt the paper's findings to the left.

In that last poll, NYT/CBS managed to turn a eight-point raw Democratic advantage of respondents (35%-27%) into a sixteen-point margin (39%-23%) through its mysterious weighting process. "Weighting" itself is standard polling practice, but the April gap was wide enough to draw questions of pro-Democratic favoritism.

This time around, the gap between the official number and the raw numbers is far wider.