I recently ran across this opinion piece by John Pohoretz on polling. Trump’s in the lead — but these polls are junk
It has to do with why Donald Trump’s lead in the polls shouldn’t be trusted because polls these days can’t be trusted. He cites some odd facts which I don’t recall, and wanted to see if I am misremembering things or if this guy is misrepresenting the facts. He claims that the polls missed the Republican takeover of the Senate in 2014. I remember the polls predicting just about the results that actually occurred for the 2014 Senate races. He also claims that the 2012 national presidential polls had Romney in the lead right up to Election Day. I don’t remember any of those polls. I thought the thing with polls at the time was about the “skewed” polls that some Republicans thought showed Romney in the lead, but none that actually did so. Is this guy just making up facts for his opinion piece?
Polls are polls. They are an indication of public opinion, but by no means entirely accurate. I would never underestimate the publics willingness to “troll” pollsters. By trolling I don’t necessarily mean telling an untruth to pollsters, but a willingness of respondents to change views in the days after an opinion poll is taken. People tend to be more conservative (with a small c) in the voting booth than when replying to opinion pollsters.
Every election cycle has it’s early frontrunners. Many of them barely make it to the end of the Primaries. Opinion polling is not a precise science. When they are wrong they are rarely all wrong for the same reasons. Im reminded of the criticism of the military that they are always fighting the last war. I get the same impression with opinion pollsters.
I am not sure how much my conviction that Trump will never get the GOP nomination (nor Sanders the Democratic one) is based on the question pollsters tend not to ask - “Are you out of your fucking mind?”
The first primary is what, five months away? It is still mostly name recognition at this point.
Regards,
Shodan
Most poll aggregators/modelers (but not all) picked the Republicans to take the Senate in 2014, but by a small margin and with a relatively low degree of confidence. Republicans significantly exceeded the poll expectations in that election.
As I recall the 2012 election, there was a split between the national polls, many of which showed Romney in the lead (most notably Gallup) and the state polls which tended to favor Obama. There was some question as to how to reconcile this and which polls counted more, but most modelers went with the state polls and desperate Republicans pinned their hopes on the national polls. In the end, the state polls turned out to be correct, but Obama also won the popular vote by a decent margin so the national polls turned out to be wrong.
I’ve not read the linked article. I don’t think the polls showing Trump in the lead count for much, but not because they’re inacurate; rather because of the early stage of the race (with lack of focus and crowded field) and their national focus.
IIRC, the national polling aggregates still showed Obama leading through most of the 2012 race (and most importantly, at the end) – it was just a few outliers (like Gallup) that had Romney ahead.
He’s just wrong about 2012. Here’s the RCP average for the 2012 election. Scroll down a bit to see it graphed over time. National polls here and there showed a Romney lead, but that is expected in a race that ultimately falls within the margin of error of a typical poll.
On looking a bit further, it looks like the national polls showed Romney with a slight lead until swinging to a slight Obama advantage very close to the election.
At any rate, Obama significantly exceeded his margin in the national polls.
A guy whose resume includes work at the New York Post, Insight on the News, Commentary Magazine, the Weekly Standard, and the Washington Times warning people not to believe what they read in other, less-reputable media?
What a country!
Just nitpicking here, but from this link and the RCP link above it seems that Romney only had a lead in the polling aggregates in the immediate aftermath of the first debate – before and after that, Obama had a consistent lead (if I’m reading the cites correctly), if smaller than his ultimate margin of victory.
Yeah, there’s the root problem: taking John Podhoretz seriously in the first place.
The sad thing is that there *are *less-reputable media.