Wednesday, June 29, 2016

THE PRAIRIE EDITOR: Polls? For Now, Ignore Them

If you begin to read an article that’s based primarily on poll
results, my advice is to stop reading and go on to something

My reasoning is based on the now irrefutable evidence that
political public opinion polls in competitive or controversial
contests in virtually all major Western nations, particularly
in the U.S. and Great Britain have been chronically wrong for
some time.

Dramatic poll numbers in the U.S. presidential primaries, and
now in the match-up between Donald Trump and Hillary
Clinton flip-flop in a matter of days or a few weeks, and that’s
assuming the dubious premise that these results are in any
way accurate at the time they were taken. Even exit polls have
been wrong.

Public polling has been an honorable profession when the
pollsters have maintained high standards, as many have. But
the rise of the use of the cell phone, internet and social media
has introduced a new level, if you will, of Heisenberg’s
uncertainty principle (the measuring device alters the
measure), and the disinclination of the millions of angry
voters to disclose their true feelings to pollsters has reached
epidemic levels.

In normal times, when political rules and traditions dominate
the political environment, polling errors exist, but are relatively
infrequent. Usually, those errors are caused by the pollsters
themselves, i.e., by some bias, by flawed questions, by flawed
samples. In a time such as we are now in, specifically the 2016
presidential campaign season, the flaws in public opinion
polling might be well beyond the pollsters best efforts and

A portent of this phenomenon came in the 2014 U.S. national
elections, and the two most recent British parliamentary

There is now a full-blown mutiny taking place among the mass
of voters in virtually all Western nations, with elections in
in the U.S., Spain, France, Austria, Italy, Iceland, Netherlands,
Belgium and elsewhere seeing anti-establishment candidates
either upsetting major party candidates, or almost doing so.
These voters seem to be wary of candidly responding to
traditional poll takers or participating in most public polls.

This voter reluctance exists among voters of the left, right and
center. They are not only angry and frustrated, as most analysts
now concede, they seem determined to upset the political
shopping cart of candidates and policies put forward by the
major political parties.

With the discarded litter of bad polls for both the Democratic
and Republican presidential nomination contests still visible,
and the erratic behavior of recent polls of the November
contest between Hillary Clinton and Donald Trump recently
published, and so much turbulence ahead at both national
conventions, why should anyone pay any attention to polls
about this race?

The Colorado Republican Party has just nominated an
unknown black military veteran Darryl Glenn to be their
candidate against incumbent Democratic Senator Michael
Bennett. The initial media comments have been that
Democrats should breathe a sigh of relief for this race which
has previously considered to be potentially competitive. The
black veteran is also an outspoken conservative, and was not
favored by most in the state party establishment. Mr. Glenn
might indeed lose, but it is extremely early to make
pronouncements about this race, especially in the year of
Bernie Sanders, Donald Trump and Brexit. Polls will now no
doubt come out heavily favoring Mr. Bennett. They might,
however, bear no resemblance to the final poll in early

This will also be a political season in which the two largest
“third” party candidates for president might well receive very
large numbers of votes. The Green Party on the left, and the
Libertarian on the “sort of” right, could complicate both
polling and elections. Both these parties' tickets will be on
almost all state presidential ballots.

No doubt, by mid-October and later, the polls will become
more and more accurate, not only in the presidential race,
but also in many down-ballot races.

Until then, however, don’t count on opinion polls and analyses
based primarily on them to tell you much about what is going
to happen on election day, 2016.

Copyright (c) 2016 by Barry Casselman. All rights reserved.

No comments:

Post a Comment