Bob Rae was premier of Ontario from 1990 to 1995. His book, What's Happened to Politics?, will be published by Simon & Schuster in August.
Hardly a day goes by without a new poll about something. All political junkies, this writer included, follow them – and even rely on them a bit. So, every once in a while, it's important to give our heads a shake.
Pollsters got it completely wrong in the British Columbia election, predicting an NDP win and not the Liberal victory that ensued. They missed Kathleen Wynne's victory in Ontario. They captured the NDP wave in Alberta, but completely botched the vote in Britain.
Popular U.S. statistician Nate Silver and U.K. professor Ben Lauderdale ran a daily blog on the British election on the website 538.com. They never came close to predicting David Cameron's victory, or the dismal showing of Labour or the Liberal Democrats. In a postmortem, Prof. Lauderdale had this to say: "So our conclusion is that if we had put a reasonable amount of probability on the right national vote shares, we would have also had a reasonable probability of the right seat totals. … As it is, we have a lot more work to do refining the national vote-share analysis to protect us against the possibility of the kind of error we saw in 2015." So, it's back to the drawing board.
After the famous Literary Digest debacle of 1936 in the U.S., when the well-known publication relied on bad samples and bad data to predict Franklin Roosevelt's defeat, a new generation of social scientists insisted on statistical rigour, reliable sampling and better questions, and felt confident they had it right. But the paradigm has shifted again, and a new humility is required, as well as a public less mesmerized by never-ending polls telling us what we think.
The premise is that people have formed opinions about the questions that are being put to them. But people today may not answer the phone, or may not answer the questions. If they do, many take longer to "come to judgment," and are less likely to reach their conclusion based on how they voted before.
The renowned British pollster David Butler wrote a book in the late 1960s predicting a "natural Labour majority," because Labour voters were reproducing more quickly, and family trumped other predictors. Conservative Edward Heath won in 1970 and Margaret Thatcher shattered the Labour consensus in 1979. The pundits who point to deep-seated trends often get it wrong because the electorate has a way of upsetting the apple cart.
Events in Alberta show how quickly things can change, and how a sudden shift in the perception of leadership can combine with a volatile public mood to light a spark. Rachel Notley entered the campaign without, one suspects, much thought about becoming premier. But an early election call, a falling economy and an effective, positive campaign completely changed the perception of the public about what was possible. It showed that dramatic change can come when least expected. Voting patterns seen as impossible by professors, pundits and pollsters suddenly change once a tipping point is reached. It should teach humility to each of us, and also provide us with hope. Change can happen.
Or not. Anyone predicting the result of the next federal election is just making a lucky guess. The truth is we don't know anything except anyone can win. Partisans will want to rule out the possibility of anyone else going first past the post, but wishing doesn't make it so. Our politics are unpredictable at the moment, and we would all be wise to enjoy where we are, and relish the thought that this is one election where a campaign will matter.