The election that refuses to die

While normal people have moved on from the 2016 presidential election to real life, those in politics and the media have not.

The New York Times’ Nate Cohn writes about pollsters:

Nearly seven months after the presidential election, pollsters are still trying to answer a question that has rattled trust in their profession: Why did pre-election polls show Hillary Clinton leading Donald J. Trump in the battleground states that decided the presidency? Is political polling fundamentally broken? Or were the errors understandable and correctable?

At their annual conference in New Orleans this month, polling experts were inching toward the latter, more optimistic explanation. And there is mounting evidence to support their view.

At least three key types of error have emerged as likely contributors to the pro-Clinton bias in pre-election surveys. Undecided voters broke for Mr. Trump in the final days of the race, or in the voting booth. Turnout among Mr. Trump’s supporters was somewhat higher than expected. And state polls, in particular, understated Mr. Trump’s support in the decisive Rust Belt region, in part because those surveys did not adjust for the educational composition of the electorate — a key to the 2016 race.

Some of these errors will be easier to fix than others. But all of them are good news for pollsters and others who depend on political surveys. …

Errors have happened enough in past elections to know that an upset was well within the realm of possibility in 2016. The Upshot model estimated that a polling misfire was about as likely as a baseball strikeout or a missed midrange field goal in football. It’s not pretty, but it happens and will happen again, and a team wouldn’t release a batter or a kicker because of a strikeout or a missed kick.

Cohn’s summary of post-election polls claims that undecided voters went late to Trump, and Clinton led in polls among likely voters but not among actual voters …

But there are a few loose ends. And those loose ends keep the possibility of a more pessimistic explanation alive.

The loosest end: The state polls weighted by education didn’t fare as well as the national polls. In fact, it’s not clear whether a state poll weighted by education and with the benefit of a perfect likely-voter screen would have shown Mr. Trump ahead in the key states.

CNN/ORC, Quinnipiac and Marquette Law School all conducted live-interview polls weighted by education in Wisconsin and Pennsylvania over the final stretch. Methodologically, they were fairly similar to the best national polls. But Mrs. Clinton still had a considerable lead of four to six percentage points.

The same story seems to hold up among campaign pollsters. The public’s window into private polling isn’t very clear, but presentations by analysts from the Democratic firms Civis Analytics and Global Strategy Group indicated that Mrs. Clinton would have still led in their final surveys in Midwestern states or Pennsylvania, even after weighting by education and correcting the likely electorate.

Perhaps undecided voters could explain the remaining error. After all, these same states had the largest number of voters who switched to Mr. Trump after voting for President Obama; it makes sense that they would have been relatively likely to be undecided.

But there’s a more pessimistic possibility, one that has been mainly promulgated by analysts at the Democratic firm Civis Analytics. Their theory, advanced in postelection interviews and presentations at AAPOR, is that Trump voters weren’t responding to telephone surveys because they had lower levels of civic engagement.

The notion that polls would miss disengaged voters is not new. It’s probably the best-known response bias in polling: Respondents are likelier than people who don’t respond to surveys to be voters, to be trusting of others, and to be volunteers or engaged in their communities. If these voters were also inclined toward Mr. Trump, it would help explain the bias toward Mrs. Clinton in pre-election polls.

This would pose a more serious challenge to survey research. It would suggest that declining response rates have finally taken a toll on the accuracy of political polling, and that would be hard to fix. The 2016 election wouldn’t be like a typical strikeout or a missed field goal — it would be more like an aging player who could no longer muster the leg strength or bat speed he had a decade ago.

Cohn had an analysis that lacked tortured sports analogies last year:

It is entirely possible, as many have argued, that Hillary Clinton would be the president-elect of the United States if the F.B.I. director, James Comey, had not sent a letter to Congress about her emails in the last weeks of the campaign.

But the electoral trends that put Donald J. Trump within striking distance of victory were clear long before Mr. Comey sent his letter. They were clear before WikiLeaks published hacked emails from the Democratic National Committee. They were even clear back in early July, before Mr. Comey excoriated Mrs. Clinton for using a private email server.

It was clear from the start that Mrs. Clinton was struggling to reassemble the Obama coalition.

At every point of the race, Mr. Trump was doing better among white voters without a college degree than Mitt Romney did in 2012 — by a wide margin. Mrs. Clinton was also not matching Mr. Obama’s support among black voters.

This was the core of the Obama coalition: an alliance between black voters and Northern white voters, from Mr. Obama’s first win in the 2008 Iowa caucuses to his final sprint across the so-called Midwestern Firewall states where he staked his 2012 re-election bid.

In 2016, the Obama coalition crumbled and so did the Midwestern Firewall. …

Campaign lore has it that President Obama won thanks to a young, diverse, well-educated and metropolitan “coalition of the ascendant” — an emerging Democratic majority anchored in the new economy. Hispanic voters, in particular, were credited with Mr. Obama’s victory.

But Mr. Obama would have won re-election even if he hadn’t won the Hispanic vote at all. He would have won even if the electorate had been as old and as white as it had been in 2004.

Largely overlooked, his key support often came in the places where you would least expect it. He did better than John Kerry and Al Gore among white voters across the Northern United States, despite exit poll results to the contrary. Over all, 34 percent of Mr. Obama’s voters were whites without a college degree — larger in number than black voters, Hispanic voters or well-educated whites.

He excelled in a nearly continuous swath from the Pacific Coast of Oregon and Washington to the Red River Valley in Minnesota, along the Great Lakes to the coast of Maine. In these places, Mr. Obama often ran as strong or stronger than any Democrat in history.

In 2016, Mr. Trump made huge gains among white working-class voters. It wasn’t just in the places where Democratic strength had been eroding for a long time, like western Pennsylvania. It was often in the places where Democrats had seemed resilient or even strong, like Scranton, Pa., and eastern Iowa.

It was a decisive break from recent trends. White voters without college degrees, for the first time, deviated from the national trend and swung decidedly toward the Republicans. No bastion of white, working-class Democratic strength was immune to the trend.

For the first time in the history of the two parties, the Republican candidate did better among low-income whites than among affluent whites, according to exit poll data and a compilation of New York Times/CBS News surveys.

And whose fault is this? John Fund reports:

Hillary has clearly bought into a victimization model of her loss. She told New York magazine:

I would have won had I not been subjected to the unprecedented attacks by Comey and the Russians, aided and abetted by the suppression of the vote, particularly in Wisconsin. . . . Whoever comes next, this is not going to end. Republicans learned that if you suppress votes you win.

What Hillary is talking about is a liberal theme spread by The Nation magazine and Priorities USA, a Clinton super PAC, that laws requiring voter ID and “other suppression rules” prevented many people from voting. Representative Keith Ellison of Minnesota, who came within a hair of becoming the new head of the Democratic National Committee, got out in front on the issue shortly after the election. Just this month, he tweeted that “Wisconsin’s Voter-ID Law Suppressed 200,000 votes in 2016 (Trump Won by 22,748) — via @AriBerman @thenation.”

But honest liberals haven’t let their brethren get away with such reasoning. Slate — in a story headlined “Did a Voter ID Law Really Cost Clinton a Victory in Wisconsin?” — quoted several election experts who poured “a big bucket of cold water” on the idea. The reliably liberal fact-checker Snopes ruled the claim “unproven,” noting that even if some people lacked the ID required to vote and didn’t bother to fill out provisional ballots, it didn’t mean they wanted to vote.

The liberal website Vox went further and pointed out that 1) Clinton lost in key states that didn’t have new voter-ID laws and 2) her margin of defeat was too big to be explained by any suppression. Even the New York Times filed a report from Wisconsin that found that black voters were far less excited about Hillary as a candidate than they had been about Obama.

Hillary’s loss was Hillary’s fault. It was also Hillary’s campaign’s fault, and the candidate is ultimately responsible for the campaign. We’ve known for decades that the Clintons have no shame, but Hillary’s nationwide woe-is-me tour, in addition to her faulty history (your husband got impeached, not Richard Nixon, remember?) is unseemly.



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s