Sunday, June 26, 2016


Do the result of the Brexit referendum mean that Donald Trump is going to be our next president? I agree with Jamelle Bouie that differences in the two electorates mean that the outcomes probably won't be the same:
The chief reason is that, unlike the U.K., the U.S. has a large voting population of nonwhites: Latinos, black Americans, Asian Americans, etc. In Britain, “black and minority ethnic” people make up about 8 percent of the electorate. By contrast, people of color account for nearly 1 in 3 American voters. In practice, this means that in the past two national elections, there has been an electoral penalty for embracing the most reactionary elements of national life. And we see this in the polling between Trump and Clinton. If the United States were largely white -- if its electorate were as monochromatic as Britain’s -- then Trump might have the advantage.
"Might" is an understatemnt, as I'll explain below.

There's far less polling in the U.K. of the kind we're used to, but the Lord Ashcroft Poll tells us this about the ethnic breakdown of the Brexit vote:
White voters voted to leave the EU by 53% to 47%. Two thirds (67%) of those describing themselves as Asian voted to remain, as did three quarters (73%) of black voters.
"Leave" was the winner in the Brexit referendum by approximately 4 points; according to this poll, whites backed "Leave" by 6.

By contrast, a new Washington Post/ABC poll has Hillary Clinton leading Donald Trump by 12 in a two-person race, and by 10 in a four-person race (with Gary Johnson and Jill Stein) -- but this massive Clinton lead comes despite the fact that Trump leads by 10 among white voters in a two-person race, and by 12 in a four-person race. So Trump is doing much better among whites than Brexit did -- yet he's getting clobbered in the overall race.

And Trump isn't even able to take full advantage of the fact that the Republican Party is the party of Team White People: Exit polls showed in 2012 that Mitt Romney beat Barack Obama by 20 points among whites, in an election he lost overall by 4 points. In 2008, John McCain beat Obama among whites by 12 points, in an election he lost overall by 7. So Trump is underperforming those guys among whites -- he's even underperforming McCain, who ran in a year when Obama had the second-best Democratic vote percentage in a presidential election since 1944.

Does this mean Trump can't win? No -- the Post/ABC poll is somewhat of an outlier; a new NBC/Wall Street Journal poll shows Clinton up by only 5 in a two-person race, and up by 1 in a four-person race. It's a contest, according to this poll -- though he's still losing.

But what if the polls are underestimating support for Tump, as they did for Brexit? Note that the final Brexit poll average, according to Pollster, showed "Remain" leading by 0.5; "Leave" won by 3.8% But in the presidential race, poll averages suggest that Clinton's lead is solid -- Real Clear Politics says she's now up by an average of 6.7 in a two-person race and 5.6 in a four-person race. That means that even if Trump's support is being underestimated by 4.3 points, as Brexit's was, he'll still lose.

But in that case it'll be a tight race -- and yes, that will be thanks to "embattled whites," with non-whites saving us from ourselves once again.


Ernest Lamonica said...

"Does this mean Trump can't win? No -- the Post/ABC poll is somewhat of an outlier". Excuse me Steve but Reuters/IPSOS has Hillary up 13. Does that mean may be WaPo is the "real"outlier. And as far as undercounting you know as well as I do that "Zip code prejudice" in polls undercounts minorities. Which means Hillary is probably being undercounted.

rclz said...

my only concern about polling is if they've figured out that they need to poll cell phone users who no longer have land lines. That becomes a bigger chunk every year and I worry there is a surprise in that box.

KenRight said...

One can certainly make the argument that, considering the past rapid demo- graphical retreat since circa 1950, America's embattled whites will not summon enough fight until they are the literal minority and then separatism will be the only viable political choice. Not segregation, separatism and formal surrender of certain geographical areas.

Feud Turgidson said...

relz - Who is this "they"?

If by "they" you mean pollsters per se, IMO its not useful to discuss pollsters categorically. IMO pollsters fall into several categories:

1) the saallest group of objective academics,
2) the largest group of custom piece-work technicians,
3) a tweener group of propagandists good at deceptive practices
4) another tweener group of propagandists transparently bad at deception.

The first 3 of those groups know it's not possibly to capture cell-phone only CPO voters directly like land line only LLO or types who have and use both. Instead, you have to be smart to ingenious in resorting to indirect measures and be preapred to exercise heightened perspicacity, because it requires resorting to consciously manipulated poll bases and to online polls.

There's a group related more to #2 thaa any other, poll aggregators, who take what I like to call a Richard Feynman-like Sums Over Infinities approach, basically saying, fuck all that noise I'm just going to stack them all up and measure how tall. It's obviously more sophisticated than that, but among if not the actual best is Sam Wang at PEC so you'd be better off checking with him than expecting me to explain and defend that approach. Mainly what Sam and Nate Silver and the other aggregators have going for them is ridiculously reliable results corrected with high numbers of polls, like you get with a presidential race.

If "they" is GOP/phone and Interet pollsters, they either don't want to hear about it now because they believe that reality enervates their base, or from a technical POV want to know how and where to concentrate their vote theft.

Steve M., there's a very respectable argument that the WaPo poll isn't an outlier at all, but simply the product of a deliberate guess as to what the general electorate will end up looking like this cycle.

I'm actually way more leery of the NBC poll, which to me looks like a by-product of combining overly self-conscious centrist backwash (something NBC and MSNBC seem to favor especially, given the prominence there of Chuck Todd, Andrea Mitchell, Chris Mathews and Brian Williams) and possibly a bit of reverse push polling (something else Todd in particular is known for). At this point, I wouldn't trust any NBC polling partnership with any pollster outside the first group.

Josh Fennell said...

Well, Ken, that could be something certain sub-morons call for at some point in the distant future, but even should that come to pass the response of the overwhelming majority of all Americans is and always will be "STFU," not "GTFO".

Never Ben Better said...

Hey, Ken, there's a lot of unclaimed real estate in Antarctica, and it's really, really white there. You and your buddies should look into it.

Erik C. said...

Yo, Ken

THIS white doesn't feel the least bit "embattled" over the US's changing demographics. Pretty sure I'm not the only one who feels that way, too.

rclz said...


I meant the Nate Silver's of the world.

I liked your post. It was an interesting take on pollsters.

petrilli said...

Ken, If any petulant "embattled" white guys really want to go down that road, kindly remind them that since the days of general George Washington, our federal central government knows how to handle them. This is not a new idea. Tougher, better men than them have tried it and died, or lived to regret it. I'm not even talking about the civil war.

Ten Bears said...

I do get a chuckle at the thought of these petulant white dogs and their backyard militias pointing their pop-guns at the US Army.

KenRight said...

Lincoln was a anti-egalitarian deportationist, though also a tyrant and Washington an ethnic supremacist. Both would probably object to the historicist slant in any regard.