Trump and Hillary are leading in the poll. Carson on the Republican side and Sanders on the Democratic side, are close seconds. Now, we are a year away from the election, and a lot will change. Even a week before the election, with all the massive data gathered, the experts may be wrong. Why?
- The proportion of people favoring one candidate or another changes. So, polls become less related to the overall population with time and events. One particular event is when a candidate drops out of the race.
- There is a significant undecided group or people only vaguely familiar with the candidates.
The polls generally are trying to find out who will win in the primaries not the election. When you hear that Trump is 5% above everyone else, it is with registered Republicans. Another problem is polls ask too many questions, and people will at some point, get tired of responding. Imagine pollsters reading off a list of 16 candidates and then asking whoever has agreed to a survey, which one is their favorite, or second favorite, or who they would not vote for in any case. They may say they like Jindal, because the sound of his name, or Trump because he is so well known. They are not seriously going to vote for these people.
So, how much more agony until we know which two candidates we have to choose from? It’s a good question. From the party’s perspective, it would be nice to channel the maximum amount of contributions to just one candidate. I’ve heard estimates of April to June 2016 for the final decision. It is well decided before the primary, so the nominating conventions are just big pageants, promoting a single candidate.
Now, the job of the pollsters should be in theory easier after the conventions, as they have only a 50% chance of picking the wrong candidate to win. Also, by this time, the general platforms of each party are well defined, so the undecided or poorly informed groups should be smaller.
All attention will be on the swing states. Candidates will work the hardest in the swing states where the polls say they are lagging. Thus, while the polls attempt to identify who is ahead in the half dozen key states, the candidates frustrate this effort by pulling out all stops to improve their numbers. Television plays a huge role. Huge numbers of people polled does not necessarily translate into better predictions, because it takes time to poll many people and in that period of time, people are changing their opinions.
So, if you want to shut all this stuff out for the next 8 months- I don’t blame you.