Are Polls Meant To Inform Or Influence?

While this is a complete assumption, I would have to say that when polling started out on a serious level, it was probably intended to be a reflection of what people thought at any given time about any subject. As the practice has evolved, things have gotten more complicated and more outlets have ventured into the polling game, things have likely changed. The results of polling from the 2016 Presidential Election are a major indication of this fact.

In 2016, many pundits felt that the polls where horrifically inaccurate. Well, that depends on how you read them. The national polls were actually pretty good. They generally had the popular vote tally down to a correct result. However, it was the state by state polls that were indeed atrocious. The lesson we learned, although it remains to be seen if we will heed it, is that national polls, even when taken as averages of many, are relatively useless. The reason for this is simple. If one or two very large states, such as the state of California for example, have an overwhelmingly lopsided victory for one candidate over another, the results from that state alone can inaccurately skew the result of the popular vote. This was the case in 2016 when California voted for Hillary Clinton by a margin of more than 5 million votes. If you remove that, Donald Trump won the rest of the nation by near 1.5 million.

Another reason for the major flop by the polls in 2016 was flawed and outdated methodology. Like anything else, the process to get accurate results from polls has to change over time. However, many polls still use essentially the exact same methods as they did decades ago. Some completely discount Independent voters who are not registered with either major party. That’s a big problem because today Independents make up more than 40% of the electorate, more than either the Democrats or Republicans. Some oversample Democrats, usually by a margin of around 8 points or more. Historically, Democrats have outnumbered Republicans but not by such a large margin.

It seems as if there are two main sources for polling and in my opinion that makes all the difference.

First, you have the polls that are conducted my major media outlets like the newspapers The New York Times the Washington Post or television stations like NBC or CNN. Generally speaking, the majority of the mead has become overwhelmingly biased, one way or the other. That being the case, whether intentionally or unintentionally, they are more likely to use a method that will come to the conclusion they want to hear and thus use this to influence the public. Along with the media polling, you also have some done by major universities. These seem to follow a similar process and are more likely used to influence than inform.

Then, there is the other type of polling. This is from groups that exist primarily for the purpose of polling science and study. Some of these are likely to include organizations like Trafalgar and Rasmussen. Sometimes, their results seem to be an outlier to what is the average compiled by sites like Real Clear Politics and FiveThirtyEight. But that stands to reason because they often use a different process and do not align in the same methods as those taken by media outlets.

What are we to take away from all of this? It’s simple really, when it all comes down to it, either group of polls can turn out to be correct. It just depends on whether or not they succeed at doing the job they are intended to do. Do the polls attempting to sway you vote get the job done? Or do the polls attempting to inform you about the views of the public carry more weight?