Two Harris-Trump Debate Related Posts

These two articles both arose from that damnable US Presidential debate, which, against my better judgement, I did watch for the first half-hour.  I have nothing useful to add to the gazillion words that have been written about the debate itself, other than to say that I switched back to watching a baseball game after 30 minutes because I was getting depressed.

Other than arising from that one event, these posts have nothing in common.

I. The BBC and the Election

The BBC website for the US and Canada is one I check regularly, despite my belief that the BBC, once among the most reliable and fact-oriented of all news-reporting institutions, is now no better than any other, and worse than some.

On Sept 12, I found the following story posted on the US and Canada BBC News website.

What the world thought of US debate

The first showdown between Kamala Harris and Donald Trump was closely watched not only in the US but around the world.

The debate in Philadelphia featured some tense exchanges on foreign policy between the two presidential candidates.

From Beijing to Budapest, here’s how the debate went down, according to BBC foreign correspondents.

…..

What then follows are brief segments from BBC reporters stationed around the world, purporting to reveal reactions to the debate from people in the countries/regions where they are stationed. In fact, there is almost nothing, beyond one cryptic quote from Ukrainian president Zelensky and two sentences from two media outlets in Hungary, that tells you what anyone around the world, be they leader, media organization or citizen, thought of the debate. It is all what the BBC correspondents think such people thought of the debate, which makes it not reporting but yet another opportunity for said correspondents to write what they wish. Not a fact to be seen anywhere. Gathering facts, going out and interviewing foreign individuals or leaders, that takes effort, and resources, and that clearly didn’t happen. So, the piece offers the reader no reporting or news-gathering, just more reporters saying things.

More sadly, one part of the piece is much worse than that, as one correspondent takes a dip into conspiracy theory. Here is one entire section of this report, nominally about the Middle East.

White House race keenly watched in Middle East

By Paul Adams, international correspondent, Jerusalem

The two candidates did not stray much from their previously stated positions last night, even if Trump did add, with characteristic hyperbole, that Israel wouldn’t exist in two years if his opponent becomes president.

Here in the Middle East, the race for the White House is being keenly watched.

With the war in Gaza raging and a ceasefire deal still elusive, some of Benjamin Netanyahu’s critics suspect that Israel’s prime minister is deliberately stalling until after the election, in the hope that Trump will be more sympathetic to Israel than Harris.

There’s a whiff of history perhaps being about to repeat itself.

In 1980, Ronald Reagan’s campaign team was suspected of urging Iran not to release American hostages held in Tehran until after he had beaten President Jimmy Carter, saying Reagan would give Iran a better deal.

Could something similar be afoot now? Certainly Netanyahu’s opponents believe he is now the chief obstacle to a ceasefire deal.

That’s Mr. Adams entire contribution, and all I can say is – Oh, my god. As I wrote, he talked to no one, quotes no one who live in the Middle East, but then – A whiff of history? The BBC’s intrepid correspondent in Jerusalem tells readers nothing about what anyone in the Middle East actually said about the debate, and then does not mention that this suspicion (he does at least use that word) about the Reagan campaign’s interactions with the Iranian government in 1980 was investigated by two separate US Congressional committees, who found no credible evidence of its truth. Indeed, the Chairman of the House task force, a bloody Democrat, went to the trouble of publishing an editorial in the NYTimes in 1993, shortly after his investigative committee published its findings, saying there was no credible evidence to support the claims of Reagan campaign efforts to delay the hostage release, and most of the people who testified that it had occurred could not be believed. No mention of any of that in the BBC story. You can actually read the report of the Congressional Task Force here – all 998 pages of it. Yea, I know, not likely – I’m betting Mr. Adams hasn’t read it, either, but there’s no need to if your only intent is to find another bad thing to attach to Netanyahu, if only by implication.

As I said, how the mighty have fallen. A BBC correspondent reviving what can only be seen as a (rather old) conspiracy theory, in the guise of reporting on an ongoing US presidential campaign.

II. The Debate’s Impact on Two Prediction Markets

I got this idea from a blog I follow by Rajiv Sethi, called Imperfect Information. He is very interested in prediction markets, and this post is about how two such markets reacted after the debate. If you want to read his (more techy and detailed) take on this, it’s freely available here.

For those who don’t know, a prediction market refers to a platform on which you can buy a ‘share’ in the outcome of some future event. In this case, you can either buy a share that Kamala Harris will be elected president, or buy a share that Donald Trump will be elected president. This is really placing a bet, because what buying a share in, say, Kamala Harris gets you is a payment of a dollar if Harris wins, and nothing if she loses. The terms of that bet then depend on what you paid for that share. In the graph below, which I reproduced from Sethi’s blog post, the prices at which one could buy a Harris share on two different prediction markets is displayed, during the hours just before and just after the debate.

Note that on PredictIt a Harris share was priced at 50cents four hours

before the debate. That means that betting on Harris by buying a share was an even bet; you pay 50cents and get a dollar back if she wins and lose your 50cents if she loses. Someone who thinks she has better than an even chance at winning might therefore buy a share, and someone who thinks she is more likely to lose would not. On the Polymarket site, a Harris share cost only 47cents four hours before, so that would be a good bet perhaps even for someone who thought she had only a 49% chance of winning. [It’s actually a bit more complicated than this, as Sethi explains in his long post, because Harris and Trump contracts are traded separately. What I write above is close enough for a basic understanding of these markets.]

As you can easily see, the price of a Harris contract jumped on both markets right after the debate, even though they continued to be different.

Sethi’s post delves into how these two sites can offer different prices on a Harris contract at the same moment in time, and the possibilities of making money by betting on both markets when this is true. I am writing about it here just to note that both markets saw the price of a Harris share (contract) jump immediately after the debate, reflecting the fact that people trading on those markets in general thought her chances of winning went up, and this was reflected immediately in the price of a Harris share. It would also be reflected in a drop in the price of Trump shares, which are not depicted here.

This contrasts with the fact that so far, none of the opinion polls that have updated since the debate, so far as I know, have seen any change in the ‘intend to vote for’ numbers for the two candidates. This is a big difference in how prediction markets and polls, both of which can be said to ‘predict’ the electoral outcome, operate. Markets are always open, so right after the debate, or even while it was ongoing, people can go and buy and sell shares if they think what is happening has changed the probability of a Harris win. Polls have to wait until a new survey is done and compiled before they can reflect the effect of an event like the debate.

However, along with the fact that markets react more quickly to events than do polls, there is the fact that they can offer different predictions. As I say, no poll of which I am aware thinks the debate changed either candidate’s support, whereas the two markets above reflect considerable movement. I admit I don’t know what has happened since the time frame depicted on the graph, prices can change constantly on these markets.

Now, whether polls do a better job than markets in predicting outcomes is – so far as I can see – undecidable. With an event like this election, one candidate wins, and all either a poll or market tells us any time prior to that is a probability. If the market says the probability of a Harris win is 0.55 at some point in time, and Harris eventually loses, there is no verifiable sense in which the market was ‘wrong’. There was, after all, a probability of 0.45 that Harris would lose at that same point in time. Unless a poll or market states a probability is either 0 or 1, it cannot be wrong. One can of course just declare that if the prediction of something happening is greater than 0.5 and that something does not happen, then the prediction is ‘wrong’, but I think that is based on a misunderstanding of the concept of probability. Similarly, if a poll says Harris has a 50% chance of winning and at the same time a market says it is 55%, and she wins, does that mean the market prediction was ‘better’? One can declare that to be so, but again, it’s not clear to me what that really means, since both allowed that her losing was a legitimate possibility.

None the less, one can easily find people who will argue for the superiority of one over the other. Folks do like to argue.

Interestingly, it is very difficult to operate a prediction market devoted to the US Presidential election in the US itself. The SEC and other US regulators have come down hard on the ones that did exist (one of the first was built by economist George Neumann at the University of Iowa) as being unregulated trading platforms, which forced them to shut down or leave the US. PredictIt is based at The Victoria University of Wellington in Australia, and Polymarket operates on crypto currency only, and although it now seems to be operating legally in the US, it was fined $1.2million in 2022 by The Commodity Futures Trading Commission. Wikipedia (not my favourite source for credible info) claims over $700million has been wagered on the presidential election on Polymarket.