A thought for Sunday: The Abyss always looks back, Presidential polling edition

A thought for Sunday: The Abyss always looks back, Presidential polling edition

A point I have made about economic forecasting a number of times is that one can be an excellent forecaster, so long as one is a bug on the wall. Once a significant number of people begin to follow *and act upon* the forecast, to that extent it must necessarily lose validity.

Take for example the yield curve, much in the news this year. So long as everyone ignores or excuses a yield curve inversion, it is an excellent indicator for the period of 12-24 months ahead. But if everyone *acted* on a yield curve inversion, by, e.g., canceling investments or increasing savings, it would turn into a botched “nowcast” instead. That which people might have started doing a year later, they would be doing now, when the conditions don’t yet necessitate it.

Simply put, people will act upon forecasts. The more previously reliable or certain the forecast, the more people will act on it — and thereby change the result.

This past week’s publication of former FBI Director James Comey’s book shows how the same principle applies to Presidential election polling.  Here’s the passage that has been getting a lot of scrutiny:

It is entirely possible that, because I was making decisions in an environment where Hillary Clinton was sure to be the next president, my concern about making her an illegitimate president by concealing the restarted investigation bore greater weight than it would have it the election appeared closer or if Donald Trump were ahead in all polls.

Leave aside for now that it was not for Comey to decide whether or not Clinton would be “an illegitimate president” — that’s what we have criminal juries and Impeachment for —  or that he simultaneously withheld from voters that Trump’s campaign was *also* under investigation at the time. The fact is that he was led by polling and poll aggregators who claimed that a Clinton victory was a near certainty to take an action that he probably would not otherwise have done.  And that action caused a near-immediate decline in Clinton’s poll numbers by about 4%, while early voting was actually going on in many states. All because Comey knew that Clinton’s election was “in the bag.”

In a similar vein, why was Barack Obama so passive in the face of the intelligence community telling him that Russia was trying to intervene in the election by, e.g., planting “fake news” stories? He was President. He did not need Mitch McConnell’s permission to address the nation in as non-partisan a fashion as possible. He didn’t act because he knew that Clinton’s election was “in the bag.” Isn’t that what Biden was sent to Europe to reassure all of our allies about?

There’s also been some detailed analysis indicating that there were enough Sanders to Jill Stein voters in Michigan, Wisconsin, and Pennsylvania to swing the outcomes in those States and thereby alter the election outcome. I think it’s a near certainty that these people felt comfortable casting such protest votes because they knew that Clinton’s election was “in the bag.”

To paraphrase the title of this post: when you look into the Future, the Future always looks back.