10 November 2016

What Did The Polls Get Wrong?

Why was the 2016 election win of Donald Trump so surprising?

Mostly because some key polls were considerably off the mark.

The states with the narrowest pre-election polling, followed by the actual results, follow, with errors of more than 4 percentage points highlighted.

                                  Predicted                   Actual            Underestimate of Trump
Washington                 D+13.2                   D+17.6                               (4.4)
Connecticut                 D+12.9                   D+12.2                              (0.7)
Illinois                         D+12.8                   D+16.0                              (3.2)
Delaware                     D+12.7                   D+11.5                               1.2
New Jersey                  D+11.7                   D+12.8                              (1.1)
Oregon                         D+9.2                     D+11.0                             (1.8)
Maine - statewide        D+7.8                     D+2.7                                 5.1
New Mexico                D+5.9                     D+8.3                                (2.4)
Minnesota                    D+5.7                     D+1.4                                 4.3
Virginia                        D+5.5                     D+4.9                                 0.6
Wisconsin                    D+5.3                      R+1                                    6.3
Michigan                     D+4.3                      R+0.3 (not final)                 4.6
Colorado                      D+4.0                      D +2.1                                1.9
Pennsylvania                D+3.7                     R+1.2                                  4.9
New Hampshire           D+3.6                      D +0.2                                3.4
Nevada                         D+1.2                      D+2.4                                (1.2)
Maine 2nd District       D+0.8                      NA                                     >0.8
Florida                         D+0.6                      R+1.3                                 1.9              
North Carolina             D+0.6                     R+3.8                                  4.4

Ohio                             R+1.8                     R+8.6                                  6.8
Nebraska 2nd District  R+2.3                      NA                                       NA
Arizona                        R+2.3                     R+4.3                                  2.0
Iowa                             R+2.7                     R+9.6                                  6.9
Georgia                        R+4.2                     R+5.7                                  1.5
South Carolina             R+7.2                     R+14.1                                6.9
Alaska                          R+7.8                     R+15.2                                7.4
Texas                            R+8.8                     R+9.2                                  0.4
Missouri                       R+10.0                   R+19.1                                9.1
Utah                             R+11.2                    R+19.0                               7.8
Indiana                         R+11.5                    R+19.3                               7.8
Tennessee                    R+12.4                    R+26.2                              13.8
Kansas                         R+13.0                    R+21                                   8.0
Mississippi                  R+13.1                    R+18.5                                5.4

It was not particularly notable that Trump won North Carolina, Florida and Maine's 2nd District, all of which had razor thin margins in favor of Clinton. But, Trump's performance was widely underestimated by large margins in a great many states and was overestimated in just a few states, most of which were in the western United States.

FiveThirtyEight looked at what the polls got wrong. The trends that it identified in the errors were as follows:
While the errors were nationwide, they were spread unevenly. The more whites without college degrees were in a state, the more Trump outperformed his FiveThirtyEight polls-only adjusted polling average,1suggesting the polls underestimated his support with that group. And the bigger the lead we forecast for Trump, the more he outperformed his polls.2In the average state won by Trump, the polls missed by an average of 7.4 percentage points (in either direction); in Clinton states, they missed by an average of 3.7 points. It’s typical for polls to miss in states that aren’t close, though. The most important concentration of polling errors was regional: Polls understated Trump’s margin by 4 points or more in a group of Midwestern states that he was expected to mostly lose but mostly won: Iowa, Ohio, Pennsylvania, Michigan, Wisconsin and Minnesota.
Some of the main explanations for this include the following:
James Lee of Susquehanna Polling & Research Inc. said his firm combined live-interview and automated-dialer calls, and Trump did better when voters were sharing their voting intention with a recorded voice rather than a live one. 
Women who voted for Trump might have been especially reluctant to tell pollsters, said David Paleologos of Suffolk University. The USC Dornsife/Los Angeles Times poll corroborated that: “Women who said they backed Trump were particularly less likely to say they would be comfortable talking to a pollster about their vote.” 
Gourevitch offered a theory for why polls underestimated Trump support: “that some percentage of the Trump vote is distrustful of institutions and distrustful of poll calls.” 
Pollsters also cited lower-than-expected turnout, particularly in the Midwest. “Democrats had a turnout problem,” Gourevitch said, and therefore so did pollsters. “The turnout models appear to have been badly off in many states,” said Matt Towery of Opinion Savvy
It also looks as if Trump pulled late support from many Republican voters who had been undecided or were supporting a third-party candidate. Libertarian candidate Gary Johnson’s recent decline coincided with Trump’s gains in the polls.

No comments: