I call the solution to wrong political polls the National Unity Poll but that is a facetious name. The correct name would be National Base Survey. Let me explain.
Way, way back in history, 1965, I was a hot figure in marketing research and very active in the survey research field. Being Righteous Michael, I was offended by all the straw polls taken in Iowa barbershops that were treated as legitimate polls and considered by the press to be the equivalent of more reliable random sample polls.
So I solved my problem by writing up a professional ethics statement for survey researchers and went to San Diego and got my statement passed by the Pacific Chapter of the American Association of Public Opinion Researchers PAAPOR, of which I was an active and influential member. Within a year and a half the national AAPOR adopted my statement of professional ethics. My ethics required professionals to publish their questionnaires, sample size and survey methodology.
At that point, I started a series of classes for journalists to help them understand survey research sampling and help them tell the difference between junk and decent polling.
It all worked as I had planned. I was young and able to move mountains easily.
Today, political polls are just as wrong as straw polls were forty years ago, for different reasons, but the solution to the problem is different. I am older, old people can rarely move mountains and I have no influence in the survey research field.
The solution we need is a national survey, face to face, with a very high completed-interview rate, maybe 90% completed interviews. Such a survey would be done every five years. It would ask a range of questions that could serve as a base for corporate marketing research as well as public opinion polling and would connect to Census data as well.
The National Base survey would be used to make more reliable weighting corrections in the cheap ongoing surveys that we see every day in political polls. Correct weighting would solve the current wrong polls problem.
Such a survey would be very expensive, over $5 million each time it is done.
I could get AAPOR to pay for such a survey if I were young, interested and active. I’m not.
So in lieu, a few members of Congress should start a bill along the same lines with the Bureau of Labor Statistics doing the job. To keep the government out of corporate surveying, AAPOR would quickly do the job themselves.
I call such a poll a National Unity Poll out of irony. The irony is that such a national base poll could not be done today by AAPOR because they are a chaotic bunch of stray cats that cannot agree on anything.