What Went Wrong With the Polling?

   By Florida Bill  

                                 After the unanticipated election of Donald Trump, contradicting 99 per cent of the polls, the pundits exhibited amazement.   "Just what went wrong and how did it happen" asked super-liberal Anderson Cooper from his loft at the CNN station.
                                  It was  going to be the greatest blowout drubbing of a Republican since the defeat handed Barry Goldwater in 1964 by President Johnson.  Make book on that, the analysts advised.
                                Actually Cooper and his wise German-born sage and colleague, Wolf Blitzer, were simply drowning their sorrows that Trump, whom they disdained, had handily defeated their heroine, Democrat Hillary Clinton, and the polls which they commissioned, along with others, were way, way off. 
                                  The long and short of it is that the polls ran wild and were not a bell ringer of the preferences of a nation, but rather served only as the source of inane and endless chit chat by the talking electronic heads.  Polls and surveys make news in the political season and the pundits rely upon them for endless babble.  Some shotgun polls come out so fast that you have  to wonder if there ever was a poll which involved the telephoning of 500 persons.  Did someone simply find a way to write that there was a poll, deliver a press release to the TV stations and submit a bill for work done?  
                                   In any case, if all of the polls and surveys which supposedly had been taken really were legitimate, then 99 per cent of them got it wrong. It is likely that pollsters like to produce the results which would most please the companies and organizations which commission them to take the survey.  Presto, if you want to make Clinton the preferred candidate, just ask the residents of Washington D.C. or the African Americans in Harlem.  It all depends upon whom you ask and how and where you ask it, sometimes even when. 
                               For me, and for others, including the intuitive former Chicago Tribune Labor editor Jim Strong, we see the polls as just so much horse feathers. Everyone is eating up those polls, and the pundits have plenty to talk about.  This is what the political season is all about, it seems: extrapolating the intentions of the nation, based on a handful of supposedly representative telephone and exit surveys.  "Rubbish," quipped Strong.  
                                  I have for quite a spell been skeptical of the polls and the so-called error factor of three per cent. This three per cent nonsense is just so much invented scientific rhetoric so as to add legitimacy to the "art" of surveying and polls based upon telephone inquiries to some 500 to 1,000 persons.  Does anyone really buy the theory that asking a few hundred persons on the telephone of their preferences will reveal the inner heartbeat of a nation of 310 million Americans, residing in 19,000 cities, towns and villages,  spread over 50 states and extending over 3 million square miles of land and water?  If they do, I have a bridge for the highest bidder. 
                                   There may be some real legitimacy in surveys designed to tell us the desires of a particular area with 10,000 persons based upon telephoning 500 persons.  But saying that a review of answers from that number of persons will expose the true desires of the USA is beyond reality. 
                                     In one polling book I found in the library, the aim for pollsters is to get answers by interviewing a selected sample of Americans.  These targets must look and act like the larger population they come from in every important way.  The sample must have almost exactly the same proportions of men and women, blacks, whites and Hispanics; Democrats and Republicans and old and young people as in the entire population. This small sampling is supposedly accurate to within plus or minus three percentage points.  I wonder how often pollsters adhered to those guidelines. And even if they slavishly stuck to their "snapshot," what if large numbers of one representative group or another simply read their caller IDs and decided not to answer the phone? Wouldn't that throw everything off?
                                    In this political season which got moving back in the early months of 2015, there have been polls and surveys in the thousands and thousands. They are commissioned by everyone and anyone.  Prominent pollsters firing out results almost daily are NBC, CBS, ABC, CNN, and FOX  from the TV tube; and singular or combo polls results from the New York Times, Washington Post, Chicago Tribune, Los Angeles Times and from a host of smaller papers and organizations.  And then there are the trusted polling granddads like Gallup, Rasmussen, Bloomberg, Emerson, Landmark, Reuters, and Quinnipiac. 
                                     Now, in this race between Hillary Clinton and Donald Trump, the big shots of polling like the New York Times were reporting that Clinton was an 85 per cent certainty to "trounce Trump."  Other pollsters simply took their lead from the  polling guns and followed up on their findings based on who knows what.  And always we heard about the 3 per cent margin of error, all of that playing to the scholars to get them to believe in the science of the inquiries. 
                                      I have an idea that the pollsters have discovered a way of making a whole lot of money by doing not much at all.  Consider rounding up a some ten telephone talkers and signing them on for minimum wage.  Find a big table and rig it with telephones and a script with questions and let them fill in the responses.  Give these telemarketers or what ever they are called numbers to call and pick the areas.  If you want some big Clinton Democratic numbers call Chicago which hasn't heard the word "Republican" since the 1930s.  Its Clinton cruising to victory.
                                          The pollsters then assemble the data and fashion it into a press release for distribution to the television talking news heads and the reporters on a news paper.  Presto, you have a poll and when you put it all together with the so called plus and minus 3 percentage points nonsense, the prediction is out there.  Clinton is way ahead.  But be mindful of the old admonition,  "garbage in--garbage out." 
                                           It seems that the election of 2016 may be the final proof that polling is neither an art nor a science.
                                            
                                                  XXX