What Explains The Wide Range Of Poll Results Between IBD/TIPP And Others?
In this, one of the strangest presidential campaigns in modern U.S. history, the polls in part reflect both the confusion and the division in America. Who's right?
The ABC News poll is not a radical outlier. The RealClear Politics average for the most-recent major polls gives Hillary Clinton on average a 48.0% to 41.9% lead over Donald Trump, a comfortable 6.1-point margin.
News stories suggest that, based on these polls and perhaps her own internal polling, Clinton is coasting and focusing instead on helping Democrats win seats in Congress.
For daily updates until the election and full details, including demographic breakdowns of results, follow the IBD/TIPP Presidential Tracking Poll.
For instance, the Washington Post, New York Times and online political site Politico report that the Clinton campaign was keen on claiming a large enough margin for a mandate, and is already looking beyond the election.
But that 12-point margin may be a chimera, says IBD/TIPP pollster Raghavan Mayur.
"Think back to 2008 and 2012, when Obama won," said Mayur. "He had a 7.2 percentage point margin in 2008 and a 3.6 percentage point margin in 2012. And those were elections where people were greatly excited about him. And still, that's not 12 percentage points, even combined."
"There is nothing like that excitement at the margin for Hillary Clinton," he added. "So a 12-point victory just doesn't look plausible."
And remember, in polls enthusiasm — excitement — matters. In the current IBD/TIPP Poll, Trump holds a 67% to 61% edge in "intensity of support." Whose followers will show up on voting day?
When you pull apart the polls, you can see where the differences lie — and the similarities.
Mayur notes for instance that, in recent days, the Los Angeles Times Poll and the IBD/TIPP Poll have fairly similar data when it comes to women voters.
The IBD/TIPP Poll has a 46% to 38% lead for Hillary Clinton over Donald Trump, while the L.A. Times has a 49% to 39% lead for Hillary Clinton over Trump. Basically very similar.
The ABC News Poll, on the other hand, reports a 55% to 35% lead for Clinton among women in its tracking survey — a 20 percentage point difference.
They can't all be right. But most of the polls can be wrong.
As evidence for that, we only have to go back to June of this year, and Britain's vote on Brexit, the June 23 referendum to have the U.K. either leave or remain in the European Union.
Starting in January, the majority of polls foresaw defeat for those who wanted to leave. Over nearly six months, out of 115 polls tallied by the Financial Times' Poll of Polls, just 54 showed that the "leaves" would beat the "remains." That's less than half.
And of the final 10 polls taken just before the vote, just three showed the "leaves" winning. The average for the Poll of Polls on the eve of the election was 48% "remain", 46% "leave." So everyone expected "leave" to go down in flames.
The final poll given in the FT's list, just before the election, was one by a group called Populus. It showed the "leave" losing by a whopping 10 points. It's fair to say that, among Britain's knowledgeable elites, confidence was high that Britain would "remain."
But they were all wrong.
The actual vote on June 23: 52% "leave," 48% "remain."
This should perhaps be a cautionary tale for those who call names and cast aspersions on the honesty of polls they disagree with. No pollster is ever 100% correct. But polls do ultimately have to be reconciled with reality.
Tracking polls are exactly what they suggest: A reading of likely voters based on near-term data. It says how the election would probably turn out, based on that day's data gathering. It is not a prediction.
IBD/TIPP is fairly transparent. The typical poll is intended to have between 750 and 900 respondents, a random sample of registered voters. Those are then further winnowed by identifying likely voters, as opposed to just registered voters, through both targeted questions and demographics of the respondents.
The Oct. 24 poll is pretty typical: It yielded 815 likely voters with a margin of error of plus-or-minus 3.6 percentage points at a 95% confidence level. That means, based on the sample, there's a 95% certainty that the "true" support levels for the candidates are within 3.6 percentage points of the reported results.
TechnoMetrica, IBD's polling company, conducts the survey by telephone. It uses both landline and cellphones, with about 35% coming from landlines and 65% coming from cellphones. All of the interviews are done live — no "robocalls" or other dodgy techniques that might bias the outcome.
The numbers are not reported raw. They are adjusted to match the presumed registration percentages of the political parties. That way no party is systematically underrepresented. The same is done for race, gender, region, and party affiliation.
This ensures a more accurate end result than simply relying on raw poll responses. On party affiliation, the presumed mix is as follows: Democrats 37% of likely voters; Republicans a bit over 29%; and independents at 34%.
In the end, that latter category may be key. IBD/TIPP in its latest poll has Trump ahead among independents and "other" by 41% to 32%. That's much wider than most other polls, and one possible explanation for why the poll differs from others.
Polls are, by their very nature, approximations. They use a wide variety of means to guess what literally tens of millions will do, based on just a small sample. Sometimes that yields very big differences, as it has this time.
What about political bias, as some darkly allege? Well, no pollster wants to be wrong. If any poll is really out of whack, it's likely because they missed something in their polling — not political bias.
No comments:
Post a Comment