tim's experiment

The Court of Appeal is continuing its custom, I understand, of sentencing appellants whose cases have no merit in their Lordships’ view to a further three months in the clink. This is justified on the basis that it discourages hopeless appeals, and consquently leaves their Lordships’ valuable time for other matters. There are reasons for supposing that this hope is entirely baseless, which we will come to in a minute.

First, a couple of distressing features about this arrangement. One is that lawyers are exquisitely polite to each other. Consequently punishable levels of frivolity are never found in appeals presented by a member of the legal fraternity. Or to put it another way, if you can afford your own legal landshark you don’t have to worry about this at all. Defenders of the scheme might argue that no lawyer would hire himself out to present an appeal without merit, and I might then roll around the floor laughing. The second is that consequent on this first feature, the victims of the bonus three months are all lay people who are not really in any position to judge the merits of their own appeals. They are being punished for ignorance and poverty.

Now let us consider the argument that this arrangement is a useful deterrent, and our prisons are full of people who decided not to bother with an appeal because it might arouse their Lordships’ ire. Actually this is not very likely. Consider a small experiment, which you can conduct on your friends – it is harmless and the result is quite predictable. If you offer someone a choice between a definite $900,000 and a 90 per cent chance of winning $1 million, then the vast majority of them will go for the sure thing. This is an affront to conventional economic theory because both choices are actually worth the same. But now let us suppose we change the plus signs to minus. Ask your friend if he would prefer a certain loss of $900,000 or a 90 per cent chance of losing $1 million and a ten per cent chance of losing nothing. Under these circumstances people overwhelmingly go for the gamble. Economics again wilts. The discoverers of this interesting effect (which can be found lucidly summarised in Kahneman: Thinking, fast, and slow, at chapter 31) sum it up as follows: people are risk averse for gains and risk seeking for losses.  Or to put it another way, when both the outcomes on offer are unacceptable people will accept very poor odds and a large penalty to have a small chance of a satisfactory outcome. Once you have this idea in your head it explains a lot of things, from Napoleon’s choices in the closing stages of the Battle of Waterloo to the willingness of litigants with very poor cases to fight to the end.

Now let us put ourselves in the position of the man in the Stanley prison cell. The Court of First Instance has sentenced him to ten years in prison. The choices before him are to stay where he is and serve the ten years, or to appeal, and have a very small chance of being released and a very large chance of being sentenced to an extra three months. Which would you go for? Actually, leaving human irrationality in these matters aside, someone with a sentence that long would be justified in pursuing a very faint hope of acquittal. Ten years is 120 months. So the extra three months is a 2.5 per cent increase. Even if the guy is guilty he only has to believe the Court of Appeal makes mistakes in one of every 40 cases for the gamble to make sense.

Actually of course the Court of Appeal is over-ruled by the next court up in far more than one out of every 40 cases. On the other hand the fly in this ointment from the point of view of our Stanley inmate is that the chance of their Lordships making an error in favour of a convicted felon with no lawyer are virtually non-existent. But this is beside the point. Potential appellants are in no position to assess the statistics and are unlikely to be swayed by economic theory, traditional or otherwise. They will follow the instinctive path, which is to take the gamble. however hopeless.

I infer from this that the belief that people are deterred from hopeless appeals by the custom of penalising them is a mere superstition. The habit gratifies judges, no doubt, at some considerable expense to the taxpayer, who has to pay for the extra three months of room and board. But I do not expect this to change. Judges, like academics but perhaps with more excuse, are impervious to research about the effects of their activities.