Random Justice

Justice has to be seen to be done. There is a concept of fairness that says that people who commit the same crime should get the same punishment. There is a simple logic to this, but I think that it leads to an inefficient penal system.

One problem is that if all punishments are to match perfectly, this requires the legislature to define the nature of every crime very precisely, and then to set down a fixed punishment for these. In reality, though, all crimes are individual cases, and this method leads judges and magistrates to be forced to give harsher or more lenient sentences than they or the public feel to be appropriate.

Another snag is that our prisons are becoming rather full, and making prison sentences compulsory for such crimes as burglary means that our prisons will remain crowded for a long time, and making prison sentences for the same crimes impossible will mean that hardly anyone does time for such nasty behaviour, which will annoy the public quite a bit.

A third disadvantage of the approach to justice outlined above is that potential criminals are very likely to plan their activities according to their knowledge of the potential consequences of being caught. If a criminal has a gun and is surrounded, and knows for certain that his punishment will be very severe, he will try to shoot his way out, and further endanger police and civilians. If a man contemplating burglary knows that there is no risk of his going to prison, or even of being punished at all if this is his first offence, then he is all the more likely to go ahead and commit that crime. If he isn’t caught, then he is far more likely to try it again. When eventually he is caught, he will admit solely to the one burglary – the one he got caught for – and after getting away with a caution or something similar, he will be an experienced burglar, and far more likely to return to crime as a result.

What I propose is that the approach to justice should be changed. Instead of fairness being represented by equal punishments for equal crimes, instead there should be equal risk for the same crime. Let me give you an example from wartime of equality of risk.

Atkins, Quinn, Watts, Peterson and Smith were cut off from the main body. The enemy counter-attack had been swift and in great force. They took cover as best they could in the farm buildings, but it became clear that this position was untenable. The enemy artillery shells were raining down and smashing their cover to pieces. Already their retreat was cut off by enemy machine guns. Advance was hopeless with just five of them against such numbers. On one flank the enemy had penetrated the lines in force. This just left the other flank, but that way was across an open field. The men conferred and agreed that this was their best move. At one signal, all five broke cover, and clutching their rifles, they ran as fast as their legs could carry them over the low straw stubble. Almost immediately, a mortar bomb landed next to Watts and he crumpled in a heap. He was at the back, though, and the others ran on unheeding. The enemy had got one machine gun forward and this raked the field with lead. Just a dozen paces short of safety, Atkins went down with a bullet in the belly. Quinn and Smith dragged his body for a few yards but Atkins yelled in agony and insisted that they leave him. A burst of fire narrowly missed them and they leapt the last few yards, landing in a drainage ditch from which they crawled to safety.

Atkins died four hours later of his wounds. Watts was taken prisoner and survived another year in a POW camp, before succumbing to TB just before the war ended. Quinn ended the war as a major, and went on to make his fortune as the owner of a chain of laundrettes. Peterson emigrated to new Zealand and ran a sheep farm. Smith retired from the motor industry to spend more time with his eight grandchildren. He organised a reunion of the chums, and at this showed them the letters he had received from Watts before he died, in which Watts looked forward to seeing them all again.

The fates of the five men were very different, I think you’ll agree. Do you think that Atkins spent his last four hours cursing his comrades? Why do you think that Watts was not bitter at the unfairness of his fate? The fact is that all five men understood that the situation was fair, in that all five shared the same risk as they tried to cross that open field. The mortar bomb landed and claimed Watts, but it could just as easily claimed someone else. No one know before setting off that two men would be hit, and that these would be two specific men. The men who got away with their lives were doubtless very grateful for their reprieve, and probably determined to live their lives more carefully as a result.

I think that a certain amount of randomness in justice would be a good thing. If criminals knew that they were risking harsh punishments, then they would think twice before committing crimes. They could not plan ahead, in the knowledge of what light punishments awaited them. Similarly, if criminals in a dire fix knew that there was a decent chance of getting a light punishment, then they would not so readily resort to violence, and would perhaps be more ready to deal with the police.

Game theory research has shown that it pays to forgive people about one third of the time. I need to explain that. If you are bringing up a child, and that child does some naughty things every now and then, then it is an unwise policy to be too forgiving, because this teaches the child that it can get away with anything, and he will continue to behave badly. Then again, it is unwise to punish him swiftly every time, because this means that there will be no chance for you to build up any level of mutual trust, and the child will never admit to any wrong doing, because he will know that he will be punished, and never forgiven. If you forgive the child more often than not, then it will frequently choose to be naughty, knowing that it will probably get away with it, because even if caught, he will usually not be made to pay for his crime. The ideal amount of forgiveness turns out to be almost exactly one third of the time. Game theory calculates this mathematically, using a method too complicated to describe here. Not one to ignore the findings of science, I reckon that criminals should be treated lightly about a third of the time. It strikes me as right that they should be treated with moderate severity about a third of the time, and harshly about a third of the time.

How should this be applied? The criminals would be caught, charged and tried in the usual manner, but then when it came to sentencing, three levels of punishment would be specified. These would be fixed by statute, which would please the body of opinion that holds that penalties for given crimes should be fixed. This way, a criminal who had committed a serious crime would always be risking a serious punishment. The opinion of the judiciary would then come into play, which would please those who would wish every criminal case to be considered individually. A die would then be rolled in front of the accused, the judges, and the legal representatives involved, and the result modified by the opinion of the judge. There are a few simple ways the die could be tested to demonstrate its lack of bias.

For example: Bert Crackitt has been found guilty of burglary. The statute books define the three levels of punishment as: prison sentence; community service; small fine and suspended sentence. The judge considers that Bert is a particularly unpleasant criminal who has offended before, and seems utterly lacking in remorse. Accordingly, he judges that +1 will be applied to the die result. The die modifier applied by the judge can never be so much that it makes any of the three possible punishments impossible. The number of sides on the die might not be six. In this example, a twelve-sided die will be used. Results: 1-4 light sentence; 5-8 medium; 9-12 severe. The maximum die modifier would therefore be +/- 3 (because for example +4 would make a light sentence impossible). The die is cast and comes up a 3. +1 is added, for a final result of 4. Master Crackitt sighs with relief as he has got off lightly. He knows, though, that he risked a nasty prison sentence, and so will not want to repeat his offence. The prisons remain that little bit less full, and a less good training ground for future criminals.

The die roll method might also be used to determine the length of prison sentence, amount of fine etc.

This system is actually fairer than the existing one, in which for some crimes the opinion of the judge counts for so little, and in which some criminals can be certain of walking away free and smug. A criminal getting off lightly in the current system may leave the court feeling that since he got away with it once then he can do so again. Equality of risk is something people instinctively understand, and criminals walking free after some random justice will be doing so knowing how close they came to a fate much worse, and hoping never to have to sweat like that again; while those who take the harshest punishments will continue to serve as a warning to others, so the benefits of both extremes are preserved without the drawbacks.


Click here to go back to the home page