When you bet on football, by and large you know whether you’ve won or lost. There are rules. There’s a referee. You can argue about whether a goal should have been given but you can’t argue about whether it was given. There is little room for interpretation most of the time.
Betting on politics is not always so straightforward. The refereeing is more complicated. Even when dealing with something as granular as vote-counting, the participants can and do sometimes argue about what is a vote and whether it can be counted. And often political betting is not about something as granular as that.
British bettors have tended to forget this too often. It’s very rare for election results to be challenged here, and when they are it doesn’t usually work out well for the challenger. In 1997 the Conservatives lost in Winchester by two votes. They successfully challenged the declaration in court on technicalities about the voiding of some votes, so the vote was rerun. They then lost by 21,000. That has to date been enough to push litigation by and large out of British electoral politics, at least as far as the counting of votes is concerned.
The same is not true in the US, as is now abundantly obvious this year. It should have been obvious beforehand though. The 2000 presidential election was decided ultimately by a Supreme Court decision. It is relatively common for close elections to be followed by litigation about some aspect of the process, for a while at least. The 2004 election for governor in Washington state was contested by the Republican candidate for months after the winning Democrat had been sworn in.
And this year, Donald Trump had made his strategy known for months in advance, growling incessantly and baselessly about fraud in postal voting. As early as July, the idea that he might refuse to accept defeat was being seriously mooted. So what happened after the vote on 3 November was entirely foreseeable.
Much has already been said about Betfair’s handling of their “next President” market. I’m not in a position to complain too much about it because I happily continued betting against Trump long after the date. I will, however, note that all of Betfair’s problems came from their interpretation of their own rules.
Peter Smith has already looked at this in detail. Betfair seemed to shift position as to what their own terms meant – no doubt under intense pressure from intense Trumpsters – from day to day. By the end, it was not at all clear what their own criteria were. While I would still have happily bet and in practice thought I might well be betting against the proposition that Donald Trump would be inaugurated next month at the odds on offer, that was not what the wording of the bet said.
For a market where way over £1 billion was traded, this is just not good enough. This is always the premier political betting market (the political equivalent of Wimbledon and the US Open combined) and the drafting of the terms of the bet needed to be bombproof. I’m not some mad red-pilled redneck who unthinkingly believes that there was MASSIVE FRAUD and that the kraken is about to be unleashed, but those who are might reasonably asked what changed on Monday night, given that there remain routes by which the electoral college decision could be set aside, in theory at least. Long before Monday, Betfair had implicitly moved away from the idea that the market was to be settled on projected electoral college votes. Betfair will be hoping that the aggrieved Trump bettors will subside once Joe Biden is inaugurated. Next time round, they will no doubt be sharpening up the wording of the bet considerably.
Things get much harder when there is no external arbiter. One market that has been getting a lot of attention recently is Smarkets’ market on whether there will be a deal between Britain and the EU this year. I have avoided this market simply because I find the definitions too tricky.
Smarkets’ rules for this market are as follows:
“If the UK and EU sign a trade deal between the 16th January 2020 and the 31st December 2020 this market will be settled as yes. If the trade deal is agreed in this period but comes into force at a later date, this market will be settled for yes. This market covers any trade deal, either sector by sector or a complete deal. If no trade deal is signed in this period between the EU and UK this market will be settled for no.”
This looks clear enough on its face. But again, once you get into really granular stuff, black and white merges into grey.
Two quite likely scenarios really concern me. First, the current overarching trade deal being negotiated might not get there. In order to avoid chaos, the UK and the EU may well in those circumstances negotiate mini-deals covering shipping, air travel and such like. On a literal reading of these rules, this would settle for Yes, which is not what most people would expect.
Secondly, and this looks really quite likely right now, the UK and the EU may shake hands on terms but not be in a position to put ink to paper even provisionally in time for 1 January. Logistically it may not be possible, particularly on the EU side, to get the necessary authority. There could be horrible arguments for this market about what the words “is agreed” mean. Does it mean “agreed between negotiators”? Or does there need to be some agreement of principals? Your hard cash may well depend on how this is resolved.
Helpfully, Smarkets have set out their view on both of these points:
“Mini-deals are covered in the rules and were a conceivable outcome. I could see them as a valid alternative to an overarching deal and a stop-gap between the confusion of a no-deal outcome as you note.
The term ‘agreed’ is purposely ambiguous. We’ve created enough scope that if representatives for both parties agree in any way in 2020 we’d settle the market for yes as the actual ratification steps could take place in a longer time. It is common in negotiations (these included) for the delegates to assume the powers of their bodies for the purpose of the legislation. We didn’t want to tie up what could be easily discernible as a deal in complex procedural points. An agreement in 2020, even on the basis of a handshake would work for us as long as it (a) happened in 2020 and (b) was reported in some reputable high quality source.”
This is clear, though not necessarily what every gambler might have expected when betting on whether there would be no deal. In particular, those on the no deal side might not have anticipated that no deal required no safety net at all.
I have no easy answers either for bettors or for market-makers. As a lawyer, I’m keenly aware that apparently-clear words can come under stress in unexpected circumstances. Those placing money on such markets need to think all round the risks. It would be a crying shame to lose money to a risk you hadn’t thought of.