THE BLOG
09/25/2014 03:51 pm ET Updated Nov 25, 2014

Severe Flaws in Polls

Political polls have become daily news in recent years. It begins to appear that we have put the cart before the horse.

Leadership and followership are, of course, Siamese twins in many ways, but we have begun to confuse the two. Horses trying to push carts rarely work.

True leadership is based on clear and strong headed judgment of people qualified by experience to think through issues and negotiate balanced solutions.

The basic problem with polls is that public sentiment is quite like our body temperature; it goes up and down, somewhat randomly, and taking action too quickly in response to changes can lead to serious unintended consequences.

That is why experienced doctors only make judgments after observing various factors over an appropriate period of time.

Democratic societies presumably try to pick political leaders, or they should be, based on their perceived ability to know more than average voters and to use their best collective judgments to steer that society through a complex and rapidly changing world.

When the voter--think patient--says to a pollster what his/her view is on a question, that is only one point of information for an elected leader--think doctor -- to consider in prescribing a right long term treatment.

What is happening today is a very dangerous thing. The polls are spurring and spawning too many short term and ill-considered decisions.

Polls could -- and probably would if they had their way -- take the place of some parts of the legislative process. Therefore voters should be demanding new consideration of how polls are conducted and how and when their results are interpreted and used.

It is also increasingly clear that many polls are tainted by respondents lying to pollsters because they believe they cannot trust where their answers might end up, and lead to political retaliation. Who needs that problem?

That can lead to unintended consequences of self-fulfilling promises because misinformed poll results obviously do not tell the real story. That is very dangerous in misdirecting election results and leading to voter sentiment being distorted and thus driving wrong-headed policies opposite to the actual underlying public view.

Perhaps keeping the identity of poll responders BLIND to the pollster could be managed?

What could be done to change and dampen the other bad effects of poll distortions?

A first issue to be addressed is whether there might be some 1st Amendment problems. Pollsters might argue that they have the right to ask and speak. People polled might insist on their right to respond any way they like. And, of course, the press -- which pays for a lot of polls because the public likes them --- would argue that constraining their use of polls in any way would infringe freedom of the press.

All that may be correct, but one well-understood limitation of 1st Amendment rights is that people cannot cry "FIRE" in a crowded theater. Many published polls today get massive coverage [which surely is the equivalent of crying fire] which frequently over leverages and misrepresents the public's real views which often leads to distorted consequences.

One idea that seems reasonable, if there could be a consensus to address any of these problems, would be to prohibit purely political polls too close to election dates. We have done something like this already by prohibiting publishing partial results on election days until the actual polls have closed.

Another idea would be to require specific issue polls to put questions into different forms, from different angles which should also be published, to be sure the question is not leading, and over enough of a time span to offer more than a momentary peek at opinion in response to breaking news. This would be hard to manage, but doing it right could lead to a significant improvement in making public policy.

There is probably no need to worry too much about private polls as long as they stay private. Managed polls for any advertising purposes (e.g., the notorious "push polls" intended not to measure but to shape public opinion through questions such as "Would it change your vote if you knew candidate X is a registered sex offender?" when candidate X is no such thing) are very close to koshering fraud and, hard as they are to control, we should try to eliminate them.

A good debate about polls and how they effect and distort what we do in modern society would be a very healthy thing.

The medical analogy of picking doctors for their judgment and experience--as long as one has confidence in them--is an easy analogy to apply to the process of picking elected leaders and making public policy which is based on solid grounds, not thin ice of simply reflecting the ever-changing momentary whims of random groups of patients/voters.

If we cannot manage that challenge, we may face a future much too dominated by a lot of distorted polling instead of wise heads looking at societies' longer term needs.