This post won’t surface until September, but it is being written in January after whiling away a few hours on a UseNet newsgroup where people obviously don’t take their medication on a regular basis. This time the discussion, such that it was, centered around “exit polls.” I’ve noticed various shows on NPR and other media outlets have been rehashing the whole exit poll election prediction thing again this past week so, you ought to be ready to hear about it again in September.
People quote exit polls as if they are Holy Scripture and not to be questioned. In truth, exit polls are just a touch less than criminal fraud but that touch is so small you need an electron microscope to measure it. Most fraud outlets will refer to an exit poll as “quantifiable research” to make it sound all scientific like so you buy the bull they are selling. It is actually less valid than an opinion someone pulls out of their back side simply because you _know_ how much the opinion is worth but these hucksters are trying to get you to pay a much higher price.
Calling an exit poll quantifiable research is like calling a rusted out car which takes a quart of oil each time you hit the ignition “a good deal.” It was for the person who sold it to you, but not for you. The same is true with an exit poll. Want to show a Republican victory should have happened take the poll in a location which has voted staunchly Republican for the past 30 years or so. The same tactic works for a Democratic victory, just take the poll in a staunchly Democratic district. Both are fraudulent yet will be quoted endlessly like scripture at a Bible thumpers gathering.
What is an exit poll in reality? Someone standing around outside of a polling location asking people how they voted. It is not random as most people believe. To start with only people who are both ardent supporters AND have the time to chat will volunteer to respond. It also relies on people being willing to tell the truth. If you are in a heavy Democratic polling location are you going to have the moxie to say you voted for Trump? Same is true going the other way.
Despite all of the press hoopla about exit polls, there is nothing scientific about them. We have known this since before the Truman election.
Results of an exit poll are always little more than trying to read dried chicken bones shaken from a bag onto the ground.
Exit polls aren’t the least bit scientific and they never match actuals. Here is a good rant.
Of course this entire discussion has avoided any talk of early voting and those who voted by mail. There is no method under the sun for exit polls to include people who voted by mail because they would be out of town, be it on business or out of country on military or other service. If you look at this link and pull down the spreadsheet there were over 5 million counted mail in ballots for California alone. Given a registered voter count of just under 18 million that means somewhere between 1/4 to 1/3 of all registered voters for that state voted by mail and were excluded from the exit polls. I didn’t take the time to dig for actual voter turnout numbers in California but we all know it wasn’t 100%. According to this it was around 14 million who actually voted which, if true, means more than 1/3 of actual voters were excluded from the exit polls.
Voting by mail is becoming more and more popular around the country making exit polls more and more useless. I don’t know if I trust this site since it didn’t split out vote-by-mail, but it did split out “early voting” without really saying where the vote-by-mail numbers were included. If it has any integrity then roughly 1/3 of all voters were in the “early voting” category and excluded from the exit polls.
Exit polls are a flawed sampling of people who waited until the very last day to cast their ballot skipping all voters who made up their mind well before and voted early in one manner or another. In order for an exit poll to be “accurate” it has to have a margin of error exceeding 33% not the typical 5-7% most put in the fine print.
So yes, people quoting exit polls to support one position or another are simply spouting what they wish to believe. There is nothing scientific about it and it isn’t Holy Scripture.