Can Public Polling be Trusted Anymore?

icons V1 03Public opinion polling took one directly on the chin on Election Day 2016 after polls had shown Hillary Clinton heading toward a big win. We all know how that worked out.

Following Donald Trump’s stunning victory, some are saying polling may have lost its ability to measure public opinion accurately, or that the industry needs to undergo a massive change to get back on track.

Despite all of the problems nationally,’s pollster Dan Jones & Associates did an excellent job predicting the results of the just completed Utah elections – a bright spot for political surveying.

Our Political Insiders say they’ve lost confidence in public opinion polling following the election errors. More than 2/3rds of the Republicans and Democrats on our panel say they don’t trust polling.

Our readers are much more forgiving, with 46% expressing some level of confidence in polling and 54% saying they are not confident.


Selected anonymous comments:

“Opinion polls are fairly accurate. Clinton did beat Trump in the popular vote, but the polls were off on the electoral count.”

“The polls were only off by 2%; which is pretty typical during any cycle. Clinton, in fact, did win the popular vote. Trump is president because of the odd rules of the electoral college.”

“Regarding the many pollsters that missed the mark, at best, they are incompetent to control for their biases. At worst, they are intended to sway public opinion to influence outcomes of elections.”

“I believe that a lot of Republicans intentionally lied about their plan to vote for Trump because they were too embarrassed by their candidate to admit it, but they always knew they’d vote for the party, no matter who the candidate was.”

“When President Hillary can explain to me why the Brexit vote failed, then and only then will I be able to trust public polling again.”

“Pollsters need to do a 2016 election study and deep dive in the same vein as the RNC in the aftermath of the 2012 election. Although, let’s hope they actually follow their plan if they do such a study.”

“I trust local poll takers like Dan Jones. The national folks, not so much.”

“Public opinion polling on basically everything is nothing more than a campaign tool or used to push an agenda. It’s a complete joke, has been for a very long time, and the 2016 election was great evidence toward this. They weren’t “errors,” they were an attempt to sway the vote. I believe people are waking up to the fact that the media has lost all credibility, especially with the WikiLeaks revelations, and that polling goes hand in hand with that.”

“If the accuracy of polls were really within a certain error range, you would have expected them to be randomly distributed both ways. Instead, they nearly all leaned for Hillary. Clearly, pollsters were putting their thumbs on the scale for Hillary.”

“Extreme bias during this election. On top of the obvious bias of the media, most polling firms intentionally weighted Democratic voters slightly heavier. The gig is up.”

“The assertions by those analyzing and reporting on poll data is an issue for me.”

“Polling is challenging under the most ideal circumstances; when bias by the organization is present when pollsters value pushing a narrative above accuracy, and when a culture of political correctness punishes respondent answers…mischief will ensue. Implementation of the LA Times model warrants strong consideration by polling organizations and clients who care about the veracity of the poll.”

“Public polling has become a political tool used to influence elections.”

“The 2012 and 2016 presidential election results indicate that there is something amiss in public opinion polling methods. It looks like the samples are not representative of the population.”

“Polling is as accurate as the person paying for it wants it to be. Nearly always, it’s designed to obtain the desired result and not the truth.”

“There are too many alternative forms of communication technology, and there was too much misinformation for the public to digest. It was not possible for the pollsters to survey a good cross-section of the voting public, and the voting public may not have made up its mind until late in the game.”

“I’m a scientist, with interest and expertise in statistics. Surveys are only as good as their sampling techniques: to have a meaningful poll, you have to have a completely representative sample. In the case of the last election, the polls had a couple of problems: 1) the groups sampled were not completely representative, as those supporting Clinton were easier to sample than others, 2) the voting behavior obtained in the poll was not how some of those polled actually voted. I’m suspicious of many polls, especially those that have non-neutral questions (“push polls”) or small sample sizes. Even biased organizations may take meaningful polls, but they are less likely than neutral organizations to do so. So, yes, I have confidence in SOME, but not all polls – they are all only as good as the work that goes into setting them up.”

“It’s not the polls that are the problem; it’s the way they are read. The results were all within the margin of error.”

“Locally, Dan Jones did a great job and was spot-on with most statewide and local races. Nationally, I have more concerns. The industry faces some significant challenges as people disconnect from the sources pollsters have traditionally used to collect their samples.”

“The Utah ones were right on. I think that this national election will help change some of the models.”

“Polls are all over the place based on their modeling samples. Not sure if rural areas were given as much attention as urban this time around. Also new demographic and cultural trends should be given more attention in the future.”

“Brexit and President Hillary. Back to the drawing boards for public polling.”

“The pollsters are doing the best they can. But if the presidential election polling told us anything, it is that people lie to pollsters.”

“Dan Jones & Associates perfectly predicted the results in Utah. Polling nationally needs to innovate and evolve.”

“More people are lying to pollsters or not answering at all.”