DW - Banks
[Democracy Watch Logo][Op-ed]

Polls only helpful if information accurate and available to everyone

(This editorial, authored by Duff Conacher, Coordinator of Democracy Watch, appeared in The Ottawa Citizen, Sunday, May 28, 1995 on page A9)

In July 1988, then Liberal leader John Turner's decision to force an election by allowing the Liberal-dominated Senate block the Canada-U.S. free trade deal was not based upon the principle of "letting the people decide", as he claimed, but instead upon results of secret polling which showed Canadians favoured an election before ratification of the deal. This is one of many examples from the 50 years since so-called "scientific" polling came to Canada in which the public reasons for a politician's decision seem laudable, but the private reasons show that the decision was simply practical.

Do public opinion polls help or hinder political leadership? Their main claim is that polls can help leaders know what is wanted by "the people", as selected by the pollster using demographic information, and why they want it (all results qualified by a comfortable margin of error, of course). However, the helpfulness of polls depends entirely on the results being accurate. It is now well known that polls can be completely inaccurate; they can provide a distorted snapshot of public opinion to politicians; and they can be used by politicians to distort public opinion.

A few small errors in the polling process can significantly change the results, but few polls are analyzed by the media or the public in enough detail to determine potential errors. Ask yourself: how much do I know about the latest poll besides the results? Do you know all the details about the what, where, when and how of the poll? Add to these questions unpredictable factors such as people lying to pollsters and it becomes clear that calling polls "scientific" is as much an exaggeration as many poll results, and media reports of poll results, are themselves.

Worse than the designed-in and unintentional flaws in the polling process is the manipulative use of polls. Currently, governments do not release poll results regularly, often using the completely ridiculous excuse that the polls represent advice to cabinet and are therefore protected under so-called "access" to information legislation. Polls paid for by political parties are not even subject to access laws. This means that any party can ask in a poll the same question worded ten different ways, and then release the one result that is most favourable while withholding the other nine results. This partial or strategic release of polls is fully intended to distort, not inform, public debate, usually in an attempt to increase public support for the polling party's agenda.

On May 11, 1994 the federal Liberals announced vague and supposedly "new" guidelines that stated they would routinely release polls as part of fulfilling their campaign promise of "open government" and "governing with integrity." The change in guidelines has not changed government practice, however, and one year later we are still waiting for many polls to be released.

While we wait, the federal government spends tens of millions of taxpayers' dollars annually polling on such issues as changes to social welfare, the federal budget, and gun control. Probably the worst effect of polls is that they have provided a convenient excuse for Canada's political elite to avoid significant democratic reforms that would give Canadians a direct role in government decision-making. So we have the government, the opposition parties, interest groups, and the media all claiming to know, through polling, what the public wants, while individual Canadians are shut out of the process and left only with the choice of which poll to believe.

Instead of gathering the often uninformed, or misinformed, opinions of Canadians about issues, wouldn't it be better to inform Canadians and involve them in meaningful consultations on these issues? Wouldn't we have better political leadership (and a better democracy) if, during its decision-making on an issue, the government distributed information containing facts and figures and possible options for action, and then facilitated informed discussions between Canadians and an opportunity to provide feedback to the government?

This fundamental change in government decision-making process, along with the following simple steps, would put polls in their rightful place as a tool, but not a very meaningful tool, in Canadian politics: