Be Very Afraid: Trump’s Support Among Republicans May Run Deeper Than His Poll Numbers Show

Donald Trump has upset the status quo in the Republican party, and pushed the already crazy GOP even further into insanity. But now, there is even more reason for everybody to worry about how appealing The Donald appears to be to Republicans– the polls, which have shown him leading the Republican field for weeks, may actually be underestimating Trump’s popularity.

A Democratic party research analyst named Jon Robinson had taken a close look at polling data, and he has found an interesting phenomenon. Donald Trump’s support tends to run higher in “robopolls” than in polls where respondents talk to a live interviewer.

The phenomenon, known as “mode effect,” isn’t limited to polls about Trump. Pew Research, one of the major polling organizations, conducted an experiment that found that when asked the same set of questions by a live pollster versus an automated poll, there was a significant difference in the answers that respondents provided. When talking to a live person, people were less likely to be critical of political figures, and more likely to say that they approve of their lives in general. But when it comes to the current crop of presidential candidates, both Democrats and Republicans, the only one who seems to be producing a significant mode effect is Trump.

Robinson used HuffPost Pollster to compile the data he used to come to his conclusion. He divided polls into two categories: those that used a live interviewer, and those that had no interviewer involved, including both polls conducted via phone and those on the internet.

Without getting bogged down in the weeds of statistical analysis, the chart below shows what Robinson found.

via Huffington Post

via Huffington Post

You can see that after mid-June, when Trump’s popularity began to surge, his support in polls with no live interviewer began to top his numbers in polls conducted by a human.

Natalie Jackson, who works as a data scientist for The Huffington Post, adjusted Robinson’s chart to compensate for random events. That is what is indicated by the shaded bands on either side of the lines on the chart. The gap between the two bands for much of August shows that the difference in the poll results is statistically significant. Jackson also did the same for polls involving the other candidates, and found no significant mode effect for any of them.

It’s still a long way from November 2016. But if Robinson’s observation holds over the weeks and months to come, the Republican nominee could well be Donald Trump. Before you start celebrating that possibility, and the likelihood that it will guarantee a Hillary Clinton or Bernie Sanders presidency, remember that in 2000, Al Gore should have been a lock. Gore was a former U.S. senator, and the sitting vice president in an administration that had brought peace and prosperity to America. He lost to a bumbling recovered alcoholic who had largely failed at every job he had ever been given. We all know how that worked out.

Donald Trump. George W. Bush 2.0? Be afraid. Be very afraid.

Featured image via Gage Skidmore/Flickr