A new poll after the nomination of the old hag as the Evil Party candidate suggests she got a mild boost from the show. Of course, Trump got a boost after his convention, but then the polling companies changed their methods in order put Clinton back in the lead. There are other polls showing Trump with a big lead and probably polls showing a dead heat. With the election more than three months out and most Americans enjoying the summer, the wild swings in the polls seems logical. It’s why partisans are prone to dismiss any poll that does not make them feel good.
Polling science is said to be much better now than in the past. After each election we’re told the pollsters got it close to right. Once in a while they miss, like we saw with Brexit or the last Parliament election in Britain. Obviously, the polling was wildly off with Trump early on and he did over-perform against the polls throughout the primary. That suggests the polling companies have not yet figured out how to identity the voter pool. Or, maybe the critics are right and the pollsters are lying to help the establishment.
It’s easy to be skeptical of polling. The sample sizes are so small, it is hard to see how they can be representative of the voter pool. What is never disclosed is the number of people who refuse to participate. It’s reasonable to assume that the hard thumping fanatics want to be polled, while normal people have better things to do with their time. A generation ago normal people may have been inclined to participate, feeling it was their duty as a citizen, but those days are long gone. The normies are woke.
Then there is that other reason to be skeptical. Everywhere you look the media is conspiring to deceive the public. A Muslim shoots up a gay club and we get stories about how he was a homosexuals struggling with his sexuality. All of those stories were lies. We get a dump of DNC e-mails showing a clear conspiracy between the media and the party, but the story they tell us is about Boris Badenov secretly conspiring with Trump. If you’re willing to lie like that, rigging polls is no great shakes.
That said, quantitative types will argue that some polls are pretty good. They get within a point or two of the results. Nate Silver’s new model was laughably wrong in the primaries, but his old model was pretty close to right in most of the primaries. He may have been off a few points, but he was picking the correct winner in every case. Investors Business Daily has been within a point the last few elections. They missed on the 2012 winner, but that was a close election and they were better than the rest.
The counter to this is that the range of possible results in any election is pretty small. Since the end WW2, the average difference in the popular vote is a little under nine percent. The big outlier was Reagan beating Mondale 58% to 40% in 1984. Most elections are within a 5% range so that means about five possible outcomes. In most of these elections, it was long clear who would win. Of those 16 elections, only six had any mystery to them and that’s counting 1968 and 2000.
|Barack Obama, Dem. defeats Mitt Romney, Rep.||3.86%||2012|
|Barack Obama, Dem. defeats John McCain, Rep.||7.27%||2008|
|George W. Bush, Rep. defeats John Kerry, Dem.||2.46%||2004|
|George W. Bush, Rep. defeats Al Gore, Dem.||-0.51%||2000|
|Bill Clinton, Dem. defeats Bob Dole, Rep.||8.51%||1996|
|Bill Clinton, Dem. defeats George H. W. Bush, Rep.||5.56%||1992|
|George H. W. Bush, Rep. defeats Michael Dukakis, Dem.||7.72%||1988|
|Ronald Reagan, Rep. defeats Walter Mondale, Dem.||18.21%||1984|
|Ronald Reagan, Rep. defeats Jimmy Carter, Dem.||9.74%||1980|
|Jimmy Carter, Dem. defeats Gerald Ford, Rep.||2.06%||1976|
|Richard Nixon, Rep. defeats George McGovern, Dem.||23.15%||1972|
|Richard Nixon, Rep. defeats Hubert Humphrey, Dem.||0.70%||1968|
|Lyndon Johnson, Dem. defeats Barry Goldwater, Rep.||22.58%||1964|
|John Kennedy, Dem. defeats Richard Nixon, Rep.||0.17%||1960|
|Dwight Eisenhower, Rep. defeats Adlai Stevenson, Dem.||15.40%||1956|
|Dwight Eisenhower, Rep. defeats Adlai Stevenson, Dem.||10.85%||1952|
The point here is that claiming you nailed 12 of the last 16 elections means nothing. Where pollsters are measured is when the final result is a mystery or debatable. Silver getting the 2012 election right made him a star because everyone else got it wrong. His star has now faded because he blew the primaries so badly. It suggests he was just lucky for a while or maybe his great insight was just a moment in time. The mood of the country has changed and the polling methods have changed, so his algorithm is now worse than guessing.
There’s also a new element here that we have not seen in our lifetime. The people in charge universally hate Trump. The media of both parties, the leadership of both parties, all sides of the chattering skull class, all of the beautiful people, everyone. They all hate Trump and the people backing Trump. This is a revolt of the elites and it is reasonable to assume that the pollsters feel pressure to put their thumb on the scales. If you’re going to do that, this is when you do it because everyone is doing it.
Even if the pollsters are playing it straight, they are facing an impossible task. What will this electorate be like compared to previous elections? We know lots of new voters are turning up. That was the story of the primary. We know lots of people are changing teams. Nationals Review, The Federalist and Red State are now wearing their woman cards, backing a candidate they excoriated just a year ago. At the same time, old Lefty warhorses like Susan Sarandon are flirting with Trump.
At least for now, no poll, even those that make you feel good, should be trusted. We are in uncharted territory in many ways. The pollsters, even those playing it straight, are just as lost as everyone else. More important, the people we tend to rely on for information are feverishly working against our interests to a level we have never seen. If they are willing to claim Trump is working for the KGB, they will say anything and do anything. All bets are off now, so trust no one.