New Pew Poll puts Romney in the lead. WaPo questions poll’s validity
Funny stuff. All of us out here who have been questioning the accuracy of various presidential polls and being called “poll truthers” (lord help you if you question the establishment or authority) now see a poll that favors the GOP candidate being called into question by none other than the Washington Post.
The Pew Research Center for the Public and the Press released a poll that puts Mitt Romney in the lead for the first time in their polling (Rasmussen also released a poll with Romney in the lead).
The Pew poll keys off the first debate, Romney’s big win and says:
In turn, Romney has drawn even with Obama in the presidential race among registered voters (46% to 46%) after trailing by nine points (42% to 51%) in September. Among likely voters, Romney holds a slight 49% to 45% edge over Obama. He trailed by eight points among likely voters last month.
The latest national survey by the Pew Research Center for the People & the Press, conducted Oct. 4-7 among 1,511 adults, including 1,201 registered voters (1,112 likely voters), finds that 67% of Romney’s backers support him strongly, up from 56% last month. For the first time in the campaign, Romney draws as much strong support as does Obama.
“Likely voters”, as we’ve mentioned in the past, is the key demographic. Forget registered voters. Among likely voters, Pew is recording a 12 point swing. That’s pretty significant. You can hit the Pew link to go through all the particulars.
So what’s WaPo’s disagreement with the poll? Well they don’t really “disagree” so much as imply there might be a problem with the poll’s makeup (you know, the same thing we “poll truthers” have been talking about for months):
That pesky party ID question: The Pew sample for this poll was 36 percent Republican, 31 percent Democratic and 30 percent independent. That’s a major shift from the organization’s September poll which was 29 percent Republican, 39 percent Democratic and 30 percent independent. In the 2010 election, the electorate was 36 percent Republican, 36 percent Democratic and 27 percent independent, according to exit polling. In 2008, 39 percent of the electorate identified as Democrats while 32 percent said they were Republicans and 29 percent said they were independents.
So in this poll, Pew was R+5. That’s different than the D+8 they ran in September. There’s you 12 point shift. Of course in 2010, they were completely wrong calling the D/R split even. Republicans ran a historic blow out during that election taking 60 seats in the House. In 2008, they were probably slightly undercounting self-identified Democrats. But not this time as Pew points out in their survey. Enthusiasm among GOP voters is up. It isn’t up among Democrats. And that, one supposes, is the Pew justification for plussing the GOP on this poll.
September polls are notoriously inaccurate. Polling companies, at least those who want to continue to be taken seriously, refine their models as they approach an election. This poll appears to be an example of that. As should be obvious to anyone who pays attention, the excitement, support and enthusiasm Obama enjoyed in 2008 doesn’t exist in this election, at least not anywhere to the degree it did then. That means the ratios have changed. Whether or not R+5 is the correct weight polls should give the GOP vote remains to be seen, but it certainly makes a lot more sense than D+8, the number we “poll truthers” were questioning all along.