PredictIt Still Favors Clinton Over Trump 2:1

11:00 PM CST 7/30/16

With only 100 days to go in the election, Polls show that Donald Trump has pulled virtually even with Hillary Clinton and may even be slightly ahead. The Reuters tracking poll shows Trump gaining 17 points in only two weeks. If one were to only look at these polls and the stories behind them, one might conclude that the momentum was with Trump.

Yet, for at least the last three months, through many presumably important pro-Trump events - Trump clinching the nomination, opposition to him at the convention collapsing, a fairly well-received Trump speech, an in many ways disastrous Democratic convention, the burgeoning Wikileaks and Clinton email scandals and of course the growing migrant/terrorist crisis in Europe - the predominant prediction market, PredictIt, has seen the election odds go virtually unchanged - 2:1 for Clinton over Trump.

As I write, Trump is at 34 cents and Clinton is at 67 cents.

The prediction markets - which make use of the "wisdom of crowds" - have generally been much more accurate at predicting election results than polls. Obviously, they would seem to know something that the rest of do not, either that the above events are relatively meaningless - terrorist incidents do not help Trump - or that their occurrence was foreseen and thus already baked into the odds.

I admit to finding this somewhat weird. I'm not claiming that because I think Trump will win - although if I had to bet, I would bet on him - but because it seems that nothing that happens in the world and nothing the candidates say or do seems to be really having much of an effect on things.

It might be that the markets are saying that demographics are the all-important thing, and the demographics in 2016 favor the Democrats just as they did in the last two elections. Indeed, I was one of those people who thought Romney would beat Obama in 2012 because everything seemed so different in 2008. In fact, Romney did almost exactly the same in 2012 as McCain did in 2008.

So maybe that's it - X amount of people will vote Democrat, almost no matter what, and Y amount of people will vote Republican almost no matter what, with very few people in the middle. And unfortunately (if you're a Republican), X is 5 percentage points or so greater than Y.

I actually think this is at least half-right. The country is such right now such that an "establishment" Republican will almost certainly be a loser. Romney lost in 2012. McCain lost in 2008. Bush, an incumbent President, came close to losing in 2004 and did lose the popular vote in 2000. Dole and Bush Sr. both lost. You have to go back to Reagan to find a big Republican win (with Bush Sr. winning the first time partly based on his memory). And some would say that Reagan was not an establishment Republican...

That's one of the reasons why, after deciding that Trump was not the quasi-demagogic huckster that I initially thought him to be, rooted for Trump to win the nomination. Whatever you think of Trump, he's a manifestly different candidate than all the Republican nominees of recent history. Many of his positions are different. In some ways he's realigned the party. You may like that or not (and I'm mixed) but you can't deny that he has done so.

So, my position is, while Trump may go down in flames, a Rubio or even a Cruz certainly would have done so.

But PredictIt doesn't seem to agree. Whether that's because the markets believe that the Trump difference is illusory, or whether it's because they believe that the man's seeming obnoxiousness and weirdness cancels out that difference, I have no idea.

I think PredictIt is wrong. But I fully admit I haven't plonked down $500 to put my money where my mouth is.

See, that's the flaw of PredictIt - it doesn't take into account the views of cowards.