Can We Really Trust The Numbers from Pollsters?


Public pollsters, especially those measuring presidential candidate preferences, approach their work with high integrity and professionalism. Polling expertise at Gallup, Quinnipiac, Pew Research, et al, is excellent. Reading their published methodologies, I can’t find anything to criticize.

And yet…

Most of the results are meaningless, especially when we are more than a year away from the 2016 presidential elections.

But let’s pretend for the sake of argument, that the polling results released to the media are 100 percent accurate. Even so, they are open to interpretation and commonly contribute to fictional narratives.

The current Republican candidate results, averaging all major polling by Real Clear Politics as of October 4, provides a textbook example of confusing polling data.


(Note: Former Virginia Governor Jim Gilmore isn’t listed. Most results omit him since he polls less than one percent.)

The media and pundits declare Trump as the clear leader. That’s technically true, but 76.8 percent selected a candidate other than Trump. As time rolls on 13 candidates will gradually drop out. We need to ask ourselves who respondents will support after each unsuccessful candidate drops out. That may not seem to be a big deal when Pataki, Graham, Jindal or Santorum bows out.

But which candidate would benefit if Bush drops out? Would Bush supporters gravitate to Trump? That’s possible though unlikely. If Cruz drops out, would his supporters move to Trump? My educated guess would be yes. But if Fiorina drops out would her support move to Trump? That would be highly unlikely. Few Rubio supporters would move to Trump. Some Kasich supporters might go to Trump. But very few Carson supporters would move to Trump unless he was the last man standing.

Even if media and pundits reported evenhandedly, polling data might still be corrupted.

Here’s why:

High-level polling organizations do excellent work in processing the data that has been collected. Unfortunately, they primarily use telephone polling, especially robocalling. Looking at the ways people use phones today, polling by phone makes many results questionable.

Most phone users, cell phone or land line, can view a calling number and optionally decide not to answer. Many people refuse to answer any unrecognized number thereby avoiding telemarketers, donation requests, etc. Some pollsters may leave voice mail requesting a return call, but very few people will honor those messages.

Who are the people who actually answer calls from pollsters? How many calls must a pollster make before reaching someone who is willing to talk? Do they have more time on their hands than those who don’t answer?

And for that matter, who are the people who don’t respond? Are they too busy? Are they less engaged? Would the data be different if they participated?

Technology has also created a dichotomy of a different kind. Only 60 percent of the population now uses land lines. Many landline users also have a cell phone, though there’s no data that shows whether the cell phone is their primary connection or only for occasional use. This is relevant because virtually 100 percent of people younger than 40 use cell phones as their primary connection. And the vast majority of older people primarily use landlines. So if successful connections are skewed toward land line or cell phone users, they may be skewed by age. Polling companies may try to compensate for that question by asking interviewees for their age, but results would be questionable. Age is the one question that draws the most lies.

Perhaps the worst factor, however, is that many respondents know very little about the issues, and may not even know the candidates by name. Ironically many people will answer questions about candidates and issues they know nothing about and pretend rather than be embarrassed.

At best, “it’s complicated.” Media and Pundits want to get ahead of the story. They did that in 2012. There was no Trump-like candidate then, so they projected a different winning candidate every few weeks. At various times, they predicted people such as Tim Pawlenty, Herman Caine, Newt Gingrich, Michelle Bachmann, and Rick Santorum.

So what do the polls tell us? They may offer a current day snapshot of attitudes of people who know little and care less. They do make cable news more interesting. But do they predict the future? I’ve personally polled 19 people, and 15 say “No”. Let’s go with that for now.

My book “The Victory that Wasn’t” offers a fictional alternate history with a different kind of Military, and better outcomes for all Americans. It’s available on Amazon at






Author: Steve Vachss

Steve Vachss has enjoyed a career that permitted him to perform diverse roles. He has been a reporter, a broadcaster, an editor, a tech executive, a tech marketing consultant, and entrepreneur-founder of a company providing online business services. He’s also a US Army veteran. Through all of these experiences, his first love has always been writing. Prior to creating “The Victory that Wasn't,” he wrote literally hundreds of online articles, web pages, and “how-to” books, as well as guest editorials for print media. Born in Stamford, CT, he now lives in Dublin, CA, a San Francisco Bay Area suburb.

One thought on “Can We Really Trust The Numbers from Pollsters?”

  1. 4 years ago, Karl Rove and his excellent polling savvy had Romney inaugurated and into his first hundred days as president right up until obama crushed him. So polls are somewhat suspect just like the pollsters. I subscribe to Rasmussen daily poll reports and they are routinely at great variance from most of the other polls, primarily because they poll a larger number of people. Still there is only one poll that counts and we all know which one that is.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s