Media polling a toxic addiction

Another day passed and another federal poll has been published along with endless speculation by the “experts” about what it may mean for the timing of the next election. It has become the same routine day after day, week after week. All major media outlets do it to varying degrees and frankly it is pushing me to want to find the tallest building in the city and take a leap.

It has gotten so bad that pollsters are beginning to bicker amongst themselves over the implications the media’s poll addiction is having on democratic discourse in this country.

Pollsters like Allan Gregg of Harris-Decima and André Turcotte have warned that the basis of making conclusions about public opinion based on polling is faulty. The principle of random selection is violated as phone polls skew disproportionately to the elderly, poor and rural Canadians on top of plummeting response rates.

I don’t share the pollsters’ concerns on accuracy, at least for simple survey designs based on vote choice. Pollsters have been accurate in predicting elections by making corrections for the response bias they know exists.
They miss the main point: media obsession with polls and election speculation is crowding out debate on actual relevant issues and forcing parties to chase the polls.

Why does the media insist on continual focus on the “horse race?” I think this is because the mainstream media is horrendously biased. And by biased I don’t mean ideologically, but simply biased towards laziness and journalistic shortcuts.
There is an increasingly intense 24-hour news cycle. Traditional news outlets are under more competitive pressure than ever before, particularly from New Media sources.

This encourages them to churn out content quickly with no focus on quality or importance of the product. Gone are the days of quality journalism and in-depth reporting. It’s a lot easier to spit out verbatim political party talking points or craft a template story around the most recent poll.

This becomes even more evident given how the press sensationalizes each poll that comes out regardless of whether the results are actually newsworthy. If one political party gains three points from the previous week’s poll it’s considered “momentum.”

Yet in actual fact it simply lies within the margin of error and means absolutely nothing. Most polls taken in the past two years have shown a consistently deadlocked political landscape that fluctuates around a mean. Yet that doesn’t stop the media obsession with weekly polling.

The media’s practice of putting polls on pedestals has also altered how poll results are treated. The reality is that most Canadians do not have well- formed views on political issues for perfectly rational reasons.

Their responses on surveys therefore can be wildly skewed by things as subtle as the question ordering of a survey to word choice of the question. Studies using panel data have shown that individual respondents tend to answer a question one way one week and one way the next. This fact is masked to a degree by stability in aggregate results. It gives an illusion of a crystallization of public opinion.

The media, however, treats each poll as a snapshot of a concrete public opinion on any given issue. This hems political parties into backing the position supposedly taken by the polled public, which to the media is the “right” position to take. Parties begin to chase the polls when crafting policy. We see this time and again with the current federal government.

We also see the consistent mainstream media behaviour to be critical of any party that is offside with so-called public opinion. The media needs to stop treating polls as the sole measure in judging a policy.

The polling industry is being cheapened by this behaviour. They are no longer engaged in research or thorough analysis of public opinion on the issues. The industry needs to be upfront with the public and force the media to acknowledge the limitations of their methodology as a condition of its use. And they need to stop simply feeding an addiction with no worry of the consequences.

Pollsters sell surveys to the media like drug dealers sell crack to addicts. They need to stop.

And for the media, Turcotte suggests they “should really consider what is the basis for [their] addiction and maybe enter a ten-step program.”
I couldn’t agree more.