@anon7035031, as the TFKâs foremost US political pundit, what do you make of this?
http://www.newyorker.com/news/john-cassidy/why-a-brexit-shocker-is-unlikely-here?mbid=social_twitter
If you listen to Donald Trump or Sean Hannity, the United States is heading for a Brexit-style surprise on Election Day, in which the pollsters and pundits who have largely written off the Republican Presidential candidate get humiliated. âThereâs going to be a lot of Brexit happening in about two weeks,â Trump said in Florida on Tuesday. âRemember that? Everyone went to bed. They said, âOh, gee, I think itâs not going to happen,â and I said itâs going to happen.â
I myself made the mistake of trusting the polls in the run-up to the June 23rd vote in Britain, which makes me wary of entirely dismissing Trumpâs theory. Some of the parallels are hard to ignore, so earlier this week I called Peter Kellner, a veteran British political commentator, and a former president of the polling firm YouGov, which was one of the many survey groups that got the Brexit result wrong, to discuss them. âYour typical Brexit voter was an older white male voter who had not gone to university, and thatâs your typical Trump voter, too,â Kellner told me.
Generally speaking, the U.K. pollsters misjudged such voters, many of whom were alienated from the political process and hadnât voted in recent general elections. Part of the problem was in how the pollsters initially weighted older working-class voters, compared with other population groups. But Kellner said he suspects the main problem was that many Leave voters simply refused to participate in the polls, which, consequently, left them working with biased samples. And while Kellner stressed that he wasnât making a prediction about the U.S. election, he noted that, theoretically at least, the same issue could arise here. âIf there is a correlation between willingness to vote for Trump and distrust of anybody with power or authority, and that bleeds over into the attitude towards pollsters, there could be a problem.â
There could indeed. But there are also a number of reasons to believe that a repeat of what happened in Britain wonât happen in the United States. Here are five of them:
- The polls were much closer in the run-up to Brexit. According to the Huffington Postâs Brexit poll average, which combined the results from a large number of surveys, the Remain side was ahead for most of the two months leading up to the voteâbut its lead was never very big. In fact, it peaked at 4.2 percentage points. And, about two weeks before the referendum, the Leave side took a narrow lead, which it kept until just a few days before June 23rd. Remain then crept ahead again, but not by much. If you averaged the final numbers from each polling organization, and ignored the undecideds, the figures were: Remain, fifty-one per cent; Leave, forty-nine per cent. The actual result was: Remain, forty-eight per cent; Leave, fifty-two per cent.
The polling numbers in the American election matchup look very different. Since the start of August, Hillary Clinton has retained a lead of between roughly four and eight percentage points, according to the Huffington Postâs poll average. The closet Trump has come was in the second week of September, when he reduced the gap to 3.9 points, but it soon widened again. Since the end of September, when the first Presidential debate was held, Clintonâs lead has grown, and in the past couple of weeks it has been in the range of seven to eight points.
Other poll averages, which are constructed a bit differently from the Huffington Post average, indicate the race is a bit tighter. On Thursday evening, the poll averages at Real Clear Politics, the Times, and FiveThirtyEight all estimated the gap between the two candidates at somewhere between 5.5 and 5.9 percentage points. But, looking back a bit, almost all the polls show the same basic picture: Clinton out ahead since the Conventions. If Trump doesnât close the gap much before Election Day, it would take a big forecasting error for him to win the popular vote.
- State polls confirm Trump is struggling. No election anywhere else in the world is subjected to the amount of scrutiny from pollsters that an American Presidential election receives. FiveThirtyEightâs database lists almost a thousand national polls, and that doesnât count all the surveys at the state level, many of them carried out by research organizations with extensive local experience. If the national polls were missing something, you might expect some of the local surveys to pick it up. But the picture at the state level is just as worrying for Trump.
Trump always faced a disadvantage in the Electoral College. To get to the necessary two hundred and seventy votes, he has to carry all the Republican-leaning states; win three toss-up contests in Florida, North Carolina, and Ohio; and then either pull off a surprise in Pennsylvania or win Nevada, New Hampshire, and the 1st Congressional District in Maine, which has one electoral vote.
Right now, according to the Real Clear Politics polling averages, the three toss-up states are still toss-ups, as is Nevada. But Clinton is leading by 6.5 percentage points in New Hampshire, and by five points in Pennsylvania. If you want to dismiss these state polls, as some Trump supporters do, you have to argue that the pollsters are systematically underestimating Trump support at the local level as well as the national level. But, as the next points make clear, that seems unlikely.
- Voter-registration figures donât support the idea of a Trump surge. In the Brexit vote, one of the big stories was turnout. At the May, 2015, British general election, the turnout was sixty-six per cent; in the June, 2016, referendum, it was seventy-two per cent. Some three million unanticipated voters showed up at the polls, and most analysts believe almost all of them voted for Leave.
Could something similar happen in the U.S.? In a recent article at FiveThirtyEight, the political analyst David Wasserman estimated that about forty-seven million white voters without a college degree did not vote in 2012. If Trump could get just one in eight of these voters to turn out on his behalf, Wasserman wrote, he would âwipe out Obamaâs 2012 margins in three states â Florida, Ohio and Pennsylvania â and win both the Electoral College and the popular vote.â
That sounds encouraging for Trump. The problem is that about three-quarters of these white voters werenât registered to vote before this year, and thereâs little evidence that they have been enrolling en masse. Wasserman looked at the figures for new-voter registrations from Florida, North Carolina, Ohio, and Virginia. In all four of these states, he found that registrations had increased in white working-class areas, but that those increases lagged the ones seen in areas with a lot of minorities and highly educated voters. This isnât particularly surprising. Trump doesnât really have a ground game: heâs been relying on the efforts of the Republican National Committee. The Clinton campaign, by contrast, has invested heavily in voter-registration drives.
- Independent and third-party voters donât trust Trump. If the Republican candidate isnât going to get a big boost from voting groups that the pollsters are missing, or undercounting, the only way for him to close the gap is to win over independents and people currently supporting third-party candidates. But this wonât be easy (I am ruling out completely the possibility that, at this late stage, he will be able to convert Clinton supporters in significant numbers.)
Trumpâs problem here is one he has had all along. While he boasts a large number of enthusiastic supporters, people who arenât committed to him tend to view him with a lot of mistrust. You can see this even in polls that show him within hailing distance of Clinton, such as a Fox News survey that was released on Wednesday and showed Trump just three points behind in a four-way matchup and five points behind in a head-to-head contest.
In responding to the pollsters, sixty-four per cent of self-identified independents said they didnât think Trump had the temperament to be President, and fifty-seven per cent said they were ânot at allâ or ânot veryâ confident in his ability to handle a crisis. Clintonâs ratings among independents werenât stellar, either. But on these key questions they were better than those of her opponent. âTrump needs a solid majority of undecided voters and wavering supporters of third-party candidates,â Chris Anderson, one of the pollsters who carried out the survey, commented, âand thatâs extremely unlikely since most of them think he lacks the judgment, temperament, and qualifications to be president.â
- A referendum isnât a Presidential election. In voting to leave the European Union, the disaffected masses in the U.K. expressed support for the nationalistic (and outmoded) ideal of Great Britain alone and independent. But one thing the Leave voters didnât have to do was express support for any particular individualânot Michael Gove, the Conservative cabinet minister who served as the official leader of the Leave campaign; or Boris Johnson, the former mayor of London, who was the campaignâs unofficial leader; or Nigel Farage, the longtime leader of the anti-E.U. U.K. Independence Party. The public symbol of the campaign was not a person but a flag: the Union Jack.
A Presidential election is different. As Alec Phillips, an economist at Goldman Sachs, wrote this week in a research circular distributed to the firmâs clients, âwhile both situations represented an opportunity for voters to endorse a change in the status quo, voters in the UK were asked to decide on an idea whereas in the U.S. they are being asked to decide on a person.â If Americans want to register a call for change, rather than a continuation of the past eight years, the person they have to vote for is Trump.
That changes things. Since the Brexit vote, it has become evident that, beyond communicating a huge âscrew youâ to Britainâs political, media, and business establishmentâwhich was largely on the Remain sideâmany Leave voters had little concrete idea what they were voting for. The consequences of a vote for Trump are a lot more clear: come January, he will be in the Oval Office. And ever since he won the Republican nomination, that hasnât seemed like a prospect that a majority of Americans can unite behind.
