how criminals communicate

A quick lesson in political plane-spotting: we observe, about 2.25pm today, a small business jet type, with minimal wing sweep and a tail about half-way up the fin, in an approach profile heading northwest over North London. Conclusions? It's an RAF Hawker 125 heading into Northolt, and Gordon Brown is probably back in London.

This article, meanwhile, is one of the most factual I've read so far.

Something I think is worth pointing out: Labour plus the Liberals, plus the sister parties who take the whip automatically, only need three seats to reach the 322 mark (don't forget the Sinn Feiners). Plaid Cymru would do. And it's not a question of forming a three- or four-party coalition. You can have a coalition with the Liberals and a toleration agreement with Plaid (or the SNP, or whoever). Arguably there are constitutional issues with Scottish, Welsh, or NI parties having ministerial posts with UK-wide responsibility - I'm on record as saying that no-one has ever been killed as a result of the West Lothian question, but it's a point.

Also, Labour has leverage on the Scots and Welsh parties; Labour did well in Scotland, and could only do better campaigning against an SNP that put Tories in national office. The Tories did unexpectedly well in Wales, and a similar effect might be expected for Plaid Cymru.

Another point is that the bargaining payoffs are quite interesting (I finally get to use my International Relations MSc!) - the Tories must get Liberal support to get Labour out, so they have an incentive to bid high. Labour can stay in office to the wire, and then dare the Liberals to vote in a Tory government - because of the King-Byng Thing and the Senex letter, there is no requirement for a second election in the event that the government is voted out on the Queen's Speech, so this would make Nick Clegg into a suicide bomber. Therefore, they have an interest in starting the bidding low (although not so low as to risk insulting the Liberals).

On the other hand, the Tories probably think they are still winning, so they have an incentive to cheat, making a high offer to the Liberals without any intention of carrying it out, rather than calling an election at the first opportunity. As significant numbers of Tories are potential rebels against electoral reform, the Tories' bid incentives are high offer but low credibility. Labour's are lower offer but higher credibility.

In classical IR theory, we'd be looking at this point for a costly signal, as described so well in Diego Gambetta's classic book Codes of the Underworld: How Criminals Communicate. The reason why the Tories and Labour can try to obfuscate their credibility is that talk is cheap. For a signal to be credible, it has to cost the signaller something. This could be either general or specific; whatever the cost is, the fact it exists lends greater credibility to the signal, but signals can also be "cost-discriminating", when it costs someone who is telling the truth less than it would a liar.

Labour has apparently already hoisted a costly signal - offering the Liberals a referendum on strong proportional representation and several cabinet seats. The cost here is that Labour might lose out from PR, and that offering cabinet seats to Liberals means sacking existing cabinet ministers.

But the Tories' offer is startlingly puny. Offering a Speaker's conference real-soon-now pretty much defines the concept of a cost-free and therefore worthless signal. Perhaps they are trying to signal that they don't think they need the Liberals, so as to bargain us down? If so, they're very close to the point of making an insultingly pathetic offer. Of course, there's no reason to assume the Tories are competent, or that they have an accurate assessment of their own capabilities - the Dunning-Kruger effect will be playing a major role here, especially as no Tories have ever operated in coalition since the time of Winston Churchill. The "Tory coup" strategy, which is the Tories' bargaining threat, seems to be going the way of the Schlieffen plan - once it starts losing time, it's doomed.

And von Schlieffen famously wanted to keep the right wing strong; Tory unity is far from given.

What's the Liberal position in signalling terms? Obviously, the more Labour thinks it can count on Liberal votes, the less it's going to offer - if they are certain we won't vote with the Tories, their optimal strategy is to sign up Plaid or the SNP, form a minority government, and take it to the Queen's Speech. The Tory position is similar, but marginally less so - they don't have the option of simply digging in on the high ground.

So, we need to signal Labour that they have to make us a real offer. We also have to make an opposite signal to the Tories that a centre-left coalition is a serious prospect. And the signalling has to be costly to be credible, although obviously the least costly signal is to be preferred. The simplest way of doing the first of these is to let the Tories keep talking. It keeps the press hanging on, but it doesn't involve any actual policy commitment. And, if they act rationally, the longer they wait, the higher they'll go. It pisses off most of the party, which would appear to be the cost of the signal.

So I suppose I agree with Nick. That leaves a question; what should the corresponding signal to Labour be?