How the election polls led to the wrong conclusion

The CBS News/New York Times poll is considered one of the best in the industry; yet it and many others missed the mark. CBS News elections director Anthony Salvanto tells Scott Pelley how so many experts got the election wrong.

[Ronrook]

It was simple. Trump rallies had 10, 000 plus people showing up and Hillary's rallies had empty seats in the front row.

[johngordon]

So when are these "experts" going to be fired?

[edricsayrs]

Here's one issue: in post-election polling, 93% of people polled said they voted, but actual voter turnout was around 60%. People lie. Let me repeat that...people lie. Here's another issue: polls capture the opinions of those that respond to polls. Conservative and Republican voters are, generally, less inclined to share their opinions on matters; accordingly, all polls skew liberal/Democrat.

[jamesmercer]

They were:
A. Incompetent in collecting the data.
B. Biased in interpreting the data.
C. Dishonest in reporting the data.
The correct answer is D. All of the above.

[jamesscott]

This guy says it was because of a slight differential between intent and action ~ That does Nothing to explain the huge gap between prediction and actual outcome. Even when trying to explain their own malfeasance they are incapable of any type of responsibility or self accounting.

[johnculley]

The LA Times/USC pollsters used a different methodology than other major polls that for weeks before the election indicated a Trump victory.

[keldonmcfarland]

I'm so glad that the media treated the election as they did. They contributed greatly to Trump's victory.

[nakrat]

A lot of the problem is the whole "demographic" theory of voting, slicing and dicing people according to identity, and then picking and manipulating the data to fit the theory. If you listen to the election night broadcasts, they mention race and gender about every 30 seconds. The polling, media, and political industries have been trained in this for decades, so they'd have to throw out their whole theory of society and the electorate to really change.

[ruleten]

"How can polling fix this next time?" Fix nothing you people!

[cameron]

Some of these polls had Hillary by 10 and 12. This guy is saying late polling would’ve changed that big of a margin. Polls are meant to condition you. Don’t fall for it.

[JimInTally]

These talking heads continually miss the point: The reason the polls were wrong is actually two factors: 1. Physical attacks on Trump supporters at rallies; 2. The only way to have an accurate barometer is to have a stratified random sample in each state. They just polled for the national popular opinion. Yes, they claim they also sampled each state, but it could not have been a truly random stratified sample, because such a poll requires around 1, 000 in the sample size to be reliably accurate. Usually, the popular vote somewhat mirrors the final electoral vote, but not always, the 2000 and 2016 are two where it did not.

[neguys]

No admission of cbs bias turning off people like me

[aulorenzo]

The screws up & now try’s to justify his polling. The majority of the networks are hard core Democrat & that’s the problem

[strawberryplains]

You people were only polling yourselves.

[immunecookie]

Watching this a day before the Election in 2024, the polls are saying that Harris is leading Trump.

[rentslave]

The Democrats learned a lesson:Once you go black, you can never go back.All they have for their pandering is Sotomayor and Kagan.When they die, so will the donkeys.