Lessons for planners from the 2016 campaign

Original Post Source

We have just come off of a presidential election year in which the political pollsters concur that they really missed the boat. Few (at best) picked Donald Trump to win the election. Countless articles have since been written trying to decipher how the profession could have been so wrong. Business forecasters and planners can learn from their mistakes. They understand that fore¬casts and plans are never perfect; thus they would suspect pollsters are being hard on themselves. However, I believe pollsters (as a whole) did not do a good job—and that there are lessons for forecasters in their results.

When an outcome is binary, such as when calling a coin toss, around half of a group would likely be wrong. So if the election was reasonably close, on average around half of the pollsters should have picked the winner. However, President Trump was a long shot—meaning less than 50% should have picked him. The New York Times and FiveThirtyEight, both respected pollsters, gave Trump a 15% and 29% chance of winning, respectively. I believe the latter was more accurate, but even if the former was, then around 15% of the pollsters should have predicted Trump the winner. The unbelievably minuscule number of pollsters that did predict him to win lends some credence to the pollsters being biased.

My assessment

During one year of my tenure as a business fore¬caster, I forecasted a substantial change in the annual revenue trend of my division. That was a good year for me in that I forecasted a revenue turning point—from low double-digit annual percentage growth to flat revenues. It was a bad year because the executives, managers and co-workers refused to believe the forecast because the division was coming off of multiple years of revenue growth.*

We have just come off of a presidential election year in which the political pollsters concur that they really missed the boat. Few (at best) picked Donald Trump to win the election. Countless articles have since been written trying to decipher how the profession could have been so wrong. Business forecasters and planners can learn from their mistakes. They understand that fore¬casts and plans are never perfect; thus they would suspect pollsters are being hard on themselves. However, I believe pollsters (as a whole) did not do a good job—and that there are lessons for forecasters in their results.

When an outcome is binary, such as when calling a coin toss, around half of a group would likely be wrong. So if the election was reasonably close, on average around half of the pollsters should have picked the winner. However, President Trump was a long shot—meaning less than 50% should have picked him. The New York Times and FiveThirtyEight, both respected pollsters, gave Trump a 15% and 29% chance of winning, respectively. I believe the latter was more accurate, but even if the former was, then around 15% of the pollsters should have predicted Trump the winner. The unbelievably minuscule number of pollsters that did predict him to win lends some credence to the pollsters being biased.

My assessment

During one year of my tenure as a business fore¬caster, I forecasted a substantial change in the annual revenue trend of my division. That was a good year for me in that I forecasted a revenue turning point—from low double-digit annual percentage growth to flat revenues. It was a bad year because the executives, managers and co-workers refused to believe the forecast because the division was coming off of multiple years of revenue growth.*

Leave a Reply

Your email address will not be published. Required fields are marked *