As Donald Trump slides further in national and battleground polling -- with the Clinton campaign making an eleventh-hour play for a bevy of traditionally Republican states -- the Republican nominee is making increasingly frequent references to the election being "rigged." His surrogates insist that he's referring to biased media coverage (the bravado about Trump's preternatural ability to bend the media narrative to his will, and to uniquely fight back, seems to have quieted), but this tweet also makes reference to voter fraud:
The election is absolutely being rigged by the dishonest and distorted media pushing Crooked Hillary - but also at many polling places - SAD— Donald J. Trump (@realDonaldTrump) October 16, 2016
Fraud is a real phenomenon that liberals conveniently deny and dismiss, but there is no evidence that it's widespread enough to seriously alter multiple statewide election outcomes. Trump supporter Peter King, a Congressman from New York, is tamping down the "rigged" talk -- as is Trump's running mate, who says the ticket will accept the vote tallies as legitimate in November. Ridiculous hoaxes like this don't help matters. In any case, less nihilistic and misinformed Trump fans are returning to a line of argument that we addressed months ago. Given its resurgence, let's take a second look. Does the 'Brexit' polling upset foretell a possible surprise Trump victory? Politico raises the issue:
Recent elections outside the United States should check too much complacency in the Clinton camp, especially when the side that is perceived to be losing is preaching nativist populism to voters who have been economically left behind and feel culturally under threat from ethnic change. Voters, in other words, who are especially motivated to vote for change. Less than four months ago the United Kingdom held a national referendum on whether it should exit the European Union, known as Brexit. Ahead of that contest, the betting markets, pundits and media were united in predicting a comfortable win for the pro-EU side, who wanted the U.K. to remain in the EU. Most of the polls, too, put “Remain” ahead (especially polls conducted by telephone), while the few online polls that suggested a Brexit victory were dismissed as rogue outliers riddled with sampling errors. Pundits pointed to the unfavorable ratings of leading Brexiteers like Nigel Farage who, they argued, were too divisive for Brexit to win. Others pointed to how even most voters accepted there did not seem to be much of a plan for life after Brexit. The Remain camp, we were also told, had the superior ground game—it seemed to be knocking on more doors, had more offices and had a developed strategy for targeting young university towns.
Sound familiar? Fair enough. But when Trump compared his polling situation to Brexit in August, we published this analysis, noting that even the "Remain will win" polling averages were quite close, and that The Economist's average ended in an exact tie. Here are a few additional data points worth considering when evaluating the Trump-as-Brexit hypothesis:
Why do folks continue to argue that most polls in Brexit showed remain winning? During June, Leave led in 17 polls, Remain in 14, 3 ties.— (((Harry Enten))) (@ForecasterEnten) October 17, 2016
That scatterplot looks a lot different than this one. For Trump to capture Brexit-style magic to shock the world and defy the conventional wisdom, he'll need to pull a lot closer to Clinton in the polling averages. As of this writing, Trump trails by roughly seven points according to the aggregated data at RealClearPolitics, whose algorithm just absorbed this Trump gut punch from Monmouth. If Trump supporters are looking for a better story to tell, they might consider pointing to Kentucky's 2015 gubernatorial race. Public polling showed Republican Matt Bevin down by three-to-five points. I'm told that Democrats' internal polling showed a similar margin. Final result: A nine-point Bevin victory. The polls were off by double digits. Is it possible that the polling will be that far off in a presidential race, and in a string of battleground states? I suppose so, but it's unlikely. The polls correctly predicted the winner in 2004, 2008 and 2012. And don't forget how Trump performed against the polling throughout the GOP primary season:
For the record, the five states in which Trump outperformed his polling margin are New Hampshire (2.3%), Nevada (1%), Alabama (4.6%), Arkansas (6.3%), Georgia (0.6%). All were contests were March 1st or earlier. The 15 states in which Trump underperformed his polling margin are Iowa (-8%), South Carolina (-3%), Alaska (-6.9%), Minnesota (-10.2%), Oklahoma (-13%), Tennessee (-3.8%), (-8.1%), Texas (-8.1%), Vermont (-12.7%), Virginia (-11.7%), Kansas (30.9%), Kentucky (-9.3%), Louisiana (-12%), Idaho (-28.3%), Michigan (-0.7%), Mississippi (-13%).
In other words, applying the "shy voter" theory to Trump may be a stretch. Kentucky-style "the polls are just wrong" is the better argument, albeit a weak one. There is much more polling -- in terms of quantity, variety and quality -- of the presidential contest than there was in Kentucky, a deep red state that has traditionally elected Democratic governors.