TAUBE: The growing irrelevance of polls

OPINION:

Anyone who reads a daily newspaper such as The Washington Times will regularly see references to public opinion polls. The polling data gathered from trends and insights has historically provided helpful guidance for consumers, academics and businesses.

Unfortunately, this doesn’t seem to be the case quite as much with politics. Today’s polls are becoming increasingly irrelevant factors, and mean little in the political process.

Here are two pertinent Canadian examples. First, most polls during Alberta’s 2012 provincial election consistently showed the mildly center-right Tories were going to lose power to a more center-right outfit, the Wildrose Party. Second, virtually every polling firm indicated the left-wing New Democratic Party would defeat the governing center-right Liberals earlier this month in British Columbia.

In both cases, the party that was supposed to finish a distant second surprised pollsters, pundits and politicos by winning the election. Moreover, the Alberta Tories and British Columbia Liberals formed majority governments and threw my country’s polling firms for a loop.

To be fair, this is part of a growing trend we’ve seen in other countries, including the United States. The 1948 presidential election between Harry S. Truman and Thomas E. Dewey is perhaps the most famous example of mistaking voter preferences. There have been plenty of modern examples, too.

During the 1980 presidential election, The New York Times-CBS polls insisted Jimmy Carter and Ronald Reagan were locked in a tight battle. The Washington Post made the same argument during the 1984 presidential election between Reagan and Walter F. Mondale. Neither election was even close.

In more recent years, the polls have been completely skewed. Take the 2012 presidential election between President Obama and Mitt Romney. Although Mr. Obama won the Electoral College by a significant margin (332-206), the popular vote was tight (51.1-47.2 percent). Many polls were all over the map in this regard. Gallup’s national tracking poll had Mr. Romney leading by 6 points (51 percent to 45 percent), Bloomberg had Mr. Obama by the same margin (49 percent to 43 percent), while Fox News, CNN-Opinion Research and Politico-GWU-Battleground all called it a dead heat.

Clearly, there is an escalating problem with public opinion firms and the information being collected. They must shoulder the blame with other individuals and groups, however.

Pundits and journalists have certainly contributed to this problem. Many of us consistently use polls as primary sources rather than secondary sources — and that’s just wrong.

Polls are little more than tiny snapshots of the electorate at specific period of times. The same goes for daily tracking polls, which provide some interesting insights, but should also be taken with a grain of salt. These small windows into voter preferences can change in a heartbeat, depending on daily and weekly events in the ever-turbulent and never-ending atmosphere of election cycles. Hence, pundits and journalists would be wise to only use polls as a means of complementing their work rather than proving their point.

Respondents are also to blame for this sea change in interpreting polling data. Many people dislike being called on their home phones and cellphones at all hours of the day (and night) to answer questions about their preferred candidate or party. There’s the issue of privacy in modern society, which turns off many potential respondents. Moreover, individuals are simply getting overwhelmed because there are so many polling companies calling about so many different things, from elections to our favorite brand of detergent.

Besides, let’s face it: Some people lie to polling firms. There’s the classic example of some New York-based Jewish voters who claimed they supported President Carter in 1980 because they were too embarrassed to admit they were going to vote for Reagan. Putting that aside, honesty is no longer the best policy when it comes to answering questions. Increasing voter apathy and disgust with the political process isn’t helping, and there are some people who are just plain vindictive by nature.

Hence, we’re putting way too much stock in polling to predict current and future trends. This needs to change, and fast.

The electorate should, therefore, minimize their examination of polling data to establish pseudo-opinions, and start educating themselves on the issues and candidates to make informed decisions. Meanwhile, the polling firms need to relax their relentless pursuit of numbers, simplify the contact process to regain the trust of respondents, and focus on late election polls to get more accurate readings.

A pipe dream, you say? Maybe we should take a poll about polls — and see what people think.

Michael Taube is a former speechwriter for Canadian Prime Minister Stephen Harper and a columnist with The Washington Times.