2016 is just around the corner, which means we’re on the verge of another presidential election year.
While most are focused on the candidates and the issues, though, a major issue next year could be the accuracy of the myriad of polls the presidential contest (not to mention congressional and state races) will produce.
The greatest fear is that as land-line telephones have disappeared, the response rate continues to drop. Many believe this phenomenon is behind the inaccurate polls from recent nationwide elections in Great Britain and Israel.
According to Carl Bialik of FiveThirtyEight, pollsters are increasingly fearful that this development is harming their reputations.
“Some of the recent failures may be a harbinger that the response rate threat is finally materializing,” Chris Borick of Muhlenberg College told FiveThirtyEight. “However, I think the relatively older electorate and marginally higher response rates in that group may continue to buffer more dramatic declines in accuracy.”
Whether polls will be able to keep up with the changing technology will be perhaps the most underrated yet important storyline of Election 2016.
11 Responses
IT Professional: election fraud and ballot stuffing g happened when we only used paper ballots. People counting the paper ballots have to be honest for democracy to work in elections using paper ballots. Even hanging chads were a problem a mere 15 years ago, because of the way the “butterfly” ballots were presented, making the paper system used unreliable.
Polling for November 2016 from the usual suspects here in Harrisburg will read: Republican nominee is up 10 points, Toomey already declared winner.
It’s what will appear regardless of the reality.
Maybe it’s just Pennsylvania this site and Grassroots rely on keystone polling and Susquehanna, both of whom have been soooooooo wrong. Romney was winning PA and some Turkey farmer was supposed to beat Casey, Obama declared PA winner at 8:15 and Casey was declared winner at like 8:03.
If accurate the quotation from Borick is nonsensical: “I think the relatively older electorate and marginally higher response rates in that group may continue to buffer more dramatic declines in accuracy.” You don’t offset declining representativeness in the poll sample by adding respondents from a demographic group that’s *known* to be unrepresentative of the population. Older voters are more conservative and less educated than younger people, and their issue positions differ. Having MORE of those voters in our samples will only skew results even more. The landline-mobile problem isn’t going to disappear thru wishful thinking.
the response rate for polls goes down because people HATE to be bothered on their private phones by pollsters and robocallers. i won’t answer such calls and would not be surprised if most of America so agreed. so the few without caller ID who do answer the pollsters and robots get to swing the polls their way. so be it. no great loss..
IT Professional-
“If scientific surveys say one thing and vote totals say another, which are wrong?
If the exit polls don’t match the vote totals, then there might be fraud.
If scientific surveys say one thing and vote totals say another, which are wrong? One question nobody seems to ask is if the massive differences between the surveys and reported vote totals are evidence of election fraud.
Consider:
* From long professional experience, I can tell you computers are almost impossible to secure effectively; at least it’s very expensive. Any computer security professional will tell you the same. Even the most sentitive computer systems are hacked; think of all the news reports: People hacking into banks and businesses, stealing top-secret military technology, stealing the identities of CIA spies, stealing health records, etc. Even the leading computer security companies, such as RSA, Netopia, and others, have suffered serious hacks.
* The election commissions operate on constrained budgets with antiquated equipment. There is no way they have the resources to secure their systems. Also, there are many studies demontrating how easy it is to hack voting machines.
* Would someone do it? Ballot stuffing is a practice as old as elections; computers just make it easier: Instead of producing large numbers of paper ballots, you only have to edit some numbers in a computer.
Imagine how much the outcome of elections is worth to certain businesses, political groups, or foreign governments. The value is trillions of dollars, or the very survival of their countries. I’m not saying every businessperson or politician would do it; most wouldn’t. But would the worst of them do it? Absolutely. Other countries would have no qualms (think of Putin, China, Iran, etc., and the U.S. has long influenced elections in other countries). All we need is one person or group to do it. Certainly someone has the will and many have the means.
The answer is simple: Paper ballots. Eliminate the computers in elections altogether. The world conducted elections for centuries without computers; it’s easily done. Would it take a little longer to count votes? Who cares, compared with the very real risk to the foundation of our democracy.
I haven’t had a landline for years and I still get called up for polls every now and again.
You forgot going to prison. Fed charges may also be coming.
** EDIT **
Didn’t you have a dozen polls conclude that Kane had to resign?
Apparently so.
Didn’t you have a dozen polls conclude that Kand had to resign?
And – look at her now. Shaking things up. Exposing corrupt, women-hating racist perverts (perhaps her sister included). Her arch-rival DEMOTED in Philly.
What a difference a few months make.