How popular has the Occupy Movement made itself over the last few weeks?  At first, it enjoyed a plurality of support, thanks to generous media coverage and lingering mainstream frustration over bailouts and the state of the economy.  After the demonstrations turned violent in ways large and small, that support has begun to erode.  In the latest national poll from Quinnipiac, the movement has a plurality disapproving of the demonstrations and demonstrators:

By a 39 – 30 percent margin, American voters have an unfavorable opinion of the Occupy Wall Street movement, with 30 percent who don’t know enough about it for an opinion.

The Tea Party movement gets a slightly more negative 45 – 31 percent unfavorable rating, with 24 percent who don’t know enough about it for an opinion.

I’ll return to the comparison in a moment. First, National Journal notes the significance of the shift:

A sign that the Occupy Wall Street movement isn’t the best long-term vehicle for Democrats to connect themselves with: A new Quinnipiac poll, showing a plurality of voters viewing the group unfavorably.

The poll, released today, show 30 percent of voters surveyed view the movement favorably, 39 percent unfavorably, with an additional 30 percent not hearing enough to have an opinion.  It’s one of the first national polls to suggest voters are growing skeptical of Occupy Wall Street- and it comes as police have clashed with protesters in several cities.  Previous national polls have shown a plurality of adults supporting the movement.

It’s probably worse than NJ imagines.  Yesterday, both NJ and I noted that the latest Q-poll had oddly boosted Barack Obama’s approval rating even though he hadn’t improved in any of the partisan demographics.  I suspected that Quinnipiac’s sample had shifted significantly to Democrats, and  Steven Shephard confirmed it later:

The sample of voters to whom Quinnipiac talked in the new poll is significantly more Democratic — and less Republican — than the early October survey. In the current poll, 35 percent of respondents identified as Democrats, 22 percent as Republicans and 36 percent as independents, according to data provided to National JournalWednesday morning. In the early October poll, 31 percent of respondents were Democrats, 28 percent were Republicans and 33 percent were independents.

(For reference: In 2008, according to exit polls, 39 percent of voters identified as Democrats, 32 percent identified as Republicans and 29 percent identified as independents. In the 2010 midterm elections; the percentages of Democrats and Republicans were equal; midterm elections typically feature higher Republican turnout.)

In a phone interview Wednesday morning, Quinnipiac Poll director Doug Schwartz said his organization does not weight by party identification, and the poll’s large sample size guards against over-sampling or under-sampling a certain subgroup. A review of other demographic data from the two polls shows that there were not significant discrepancies in the breakdowns by age, gender, geographical region or race across the surveys.

This poll was conducted with the same sample — that features a D+13 shift, and unweighted.  Not only is that almost twice as large as the 2008 turnout split from exit polling, it’s also almost twice as much as Barack Obama’s margin of victory in the 2008 election.  Unless Quinnipiac wants to argue that the country has become significantly more Democratic over the last three years of Obama’s term, this sample is ridiculously unrepresentative of the electorate.

In response, I took Quinnipiac’s sample and reweighted it for both the 2008 and 2010 exit polling percentages for partisan split.  In the 2008 model, the approval rating for the Occupy movement goes to 29/41; with the 2010 model, it goes to 28/42.  The Tea Party, on the other hand, fares a bit better in both models, although it’s still underwater in both. The 2008 model has a 32/44 approval rating, while the 2010 model is 34/41.

And what happens to Obama’s approval rating?  It doesn’t stay at 47/49.  In the 2008 model it falls to 46/50, and in the 2010 model to 43/52 — which turns out to be pretty close to his previous Q-poll rating of 41/55.

Update (Allahpundit): Philip Klein looks at Quinnipiac’s sample and makes a nice catch about how OWS is viewed by people with different income levels.