In my article recapping the election as it unfolded over the summer, I took some time to break down why the mainstream media making fun of ‘Poll Truthers’ could potentially backfire. More specifically, I explained why the comparison between Trump supporters’ distrust of the polls in 2016 and that of Romney supporters in 2012 was off base:
Back in 2012, Republicans were convinced that polling which showed President Obama leading were heavily skewed in favor of Democrats. They were made to pay for that optimism on Election Day, and in pointing to that experience, the likes of Perino and others have been quick to point out that the polls were right in 2012, and as such they are Gospel now.
The issue with that analysis starts with the fact that in 2012, Republicans had the Tea Party shake up of 2010 midterm election fresh in their memories. This, to many, marked a repudiation of President Obama’s agenda, and as such a torrent of Republicanism would sweep the nation in 2012. Romney was assured to ride on in that wave, and thus, the fact that the polls said otherwise meant they were faulty.
The disconnect was that Romney as a candidate was nowhere near the Tea Party ideal, to put it mildly. Thus, he didn’t come close to capturing the momentum the Tea Party had generated. In the end, he marginally improved on McCain’s effort in 2008, which meant that Obama won comfortably, despite having a lower turnout than he had four years earlier.
Those who took the polls as Gospel earlier in this summer are making the inverse error as Republicans made in 2012. Whereas Republicans then thought the electorate had shifted from where it was in 2008, those giddy about Clinton +12 polls in early August might prove to have been erroneous in thinking that the electorate hadn’t shifted from 2012.
Right on cue, this article from Politico confirms part of my thinking – that mainstream polls are skewed towards the 2012 electorate:
MSNBC ‘unskewed’ a CNN national poll on Tuesday that showed Donald Trump leading Hillary Clinton by two points, re-weighting the results to match the 2012 electorate and showing a four-point lead for the former secretary of state.
The poll of likely voters, released Tuesday by CNN/ORC, showed Trump ahead of Clinton nationwide in a four-way contest, 45 percent to 43 percent. But MSNBC host Chuck Todd explained that the poll, in his network’s estimation, may have oversampled white voters without a college degree, one of Trump’s strongest groups.
“Whites without a college degree appear to make up nearly half of their sample. In 2012, by the way, whites without a college degree was slightly more than a third of all voters,” Todd said. “The point is, your numbers may not be wrong but your weighting may be, your assumptions. So the CNN folks assumed an electorate that is not an impossible scenario for Trump, but it would be an historic shift if it occurred.”
With the numbers adjusted to reflect how the electorate shook out four years ago, Clinton’s two-point deficit shifted to a four-point lead, 46 percent to 42 percent.
It is my view that thinking the 2016 electorate will end up looking like the 2012 electorate is a recipe for failure, yet this is what most mainstream polls do. I believe that the 2016 electorate, owing to the large number of new/disaffected voters Trump is bringing to the table, is far more amenable to the Republican candidate than it was for Mitt Romney.
Data from that CNN poll bears this out:
According to that, the electorate as a whole is less enthusiastic about going to vote than it was four years ago. But given the choice between Trump and Clinton, there is more enthusiasm for the former.
Despite that major flaw in mainstream polls, they have tightened, to the point where many of them show at best a Clinton lead within the margin of error, if not a dead heat. Given that those polls refelct the 2012 Obama friendly electorate, the Clinton campaign should have some serious concerns. Because once the relatively large, and more enthusiastic Pro-Trump portion of the electorate is truly reflected, those polls might look a lot different. Indeed, we might not find out the true extent of that until election night.