You may think that Donald Trump isn’t doing well with young people, or with black people, and especially not doing well with young black people. But the polls say you’re wrong. Well, one poll. Well, one guy in one poll.
There is a 19-year-old black man in Illinois who has no idea of the role he is playing in this election.
He is sure he is going to vote for Donald J. Trump. …
He’s a panelist on the U.S.C. Dornsife/Los Angeles Times Daybreak poll, which has emerged as the biggest polling outlier of the presidential campaign.
The USC Dornslife/LA Times poll isn’t really a “poll” in the classic sense. It’s more of a glorified focus group; a “panel” composed of handpicked people across the nation. They don’t call up new people for each “poll,” they just … check in with their panel.
Then they apply a weighting factor to the responses that’s extraordinary.
Our Trump-supporting friend in Illinois is a surprisingly big part of the reason. In some polls, he’s weighted as much as 30 times more than the average respondent, and as much as 300 times more than the least-weighted respondent.
That’s far outside the weighting that goes on in most real polls. So much so that the presence or absence of this single voter can flip the results of the poll. And, because the USC Dornslife/LA Times numbers are treated like a normal poll in polling averages, their results can swamp values collected in a more reliable manner.
This is definitely one place where the polls do need to be unskewed.
Election Day is fast approaching, and we need all hands on deck. With the PCCC and Daily Kos, no matter where you live, you can call key voters in districts where progressive Democrats are in tight races. Click here to get started.
The USC Dornslife/LA Times group has trended to be the only pro-Trump vote among national polls, reporting Trump up for a solid month until Hillary took the lead in just the day. It’s been solidly unresponsive to news, often contrary to the movement of actual polls, and the best way to describe its movements is unmitigated noise.
Why did Hillary finally move into the lead? It’s back to our friend in Illinois.
He is also the reason Mrs. Clinton took the lead in the U.S.C./LAT poll for the first time in a month on Wednesday. The poll includes only the last seven days of respondents, and he hasn’t taken the poll since Oct. 4. Mrs. Clinton surged once he was out of the sample for the first time in several weeks.
That the presence or absence of a single person in a group of almost 3,000 could make this kind of difference, shows the ridiculous extremes by which the USC/LAT results are being weighted.
Including this information in any national polling average isn’t adding insight. It’s suppressing it. The USC/LAT results are so off from the rest of the results, so heavily weighted by subgroups, that their inclusion in polling average decreases the accuracy of any result.
The national polls really do need to be unskewed—by dropping the USC Dornslife/LA Times not-a-poll from their picture.
If the U.S.C./LAT poll were a normal poll, the 19-year-old from Illinois might have been in the poll only once. Most of the time, the heavily weighted young black voters would lean toward Mrs. Clinton — ensuring that the poll both had the appropriate number of black voters, and a relatively representative result.
But the U.S.C./LAT poll had terrible luck: The single most overweighted person in the survey was unrepresentative of his demographic group. The people running the poll basically got stuck at the extreme of the added variance.
Bad panel design, bad execution when it comes to weight, and bad luck when it comes to outliers. Bad poll.