How would you feel if the three times out of 20 you are wrong in your forecasts, and it all happened in the same election? That was the pollsters’ dilemma with last week’s election in the United States. They were just flat out wrong.
But you have got to have some sympathy for them. Historical data did not mean much. They only experienced one other election such as this for president and they got it wrong then. In the largest turnout in American history, the republican voter turnout set records. Luckily, so did the democratic voters.
But some states that were expected to be won by democrats were often won by the republicans and visa-versa. The Senate ended in the hands of the republicans and the democrats kept the house majority by a slim margin. It all took four days of concern and counting.
What surprised us Canadian political watchers that the pollsters had no way of verifying their polls. I think I was studying polling before I ran my first political campaign. I enjoyed politics more than working for the news media but I had always been fascinated by the studies that told you what publication readers read some or most of advertisements and editorial content.
And then I met a federal politician who used telephone calls to determine how voters intended to vote. This was back in the day when he had to use a card index for each voter. He used those cards to tell his workers who to get to the polls on election day. What we are so routinely doing today with computer programs, he was doing the hard way almost 60 years ago.
But I think his card system relied on better input. Today’s smart phones do not always get answered with blocked numbers. Those automated Robocalls calls might be cheap but they are less and less effective. Political parties might just have to revive the science of knocking on doors with what we call the ground game.
-30-
Copyright 2020 © Peter Lowry
Complaints, comments, criticisms and compliments can be sent to [email protected]