Play the lottery, win a million pounds or a legal ruling

Random Justice by Neil Duxbury is a study of the role lotteries might usefully play in judicial decision making. Imagine you are the judge deciding a child custody case. The mother has a demanding professional career, a splendid house in a good school catchment area but sometimes gets on badly with her teenage daughter; the father is employed part- time,lives modestly but has time to look after his daughter whom he relates to well. The girl is ambivalent about which parent she prefers to live with. The "best interests of the child" principle cannot help you because the benefits each parent offers are incommensurable.

In such situations of indeterminacy, suggests Duxbury, reason has reached its limits. Fine tuning (investigation of each parent's pluses and minuses) would be protracted and agonising and still might not provide a cast-iron reason for choosing one parent over the other. If you therefore determine the case by tossing a coin, would you be satisfied with the decision? How would the parents and the child feel? Such questions are the focus of Random Justice .

Similar problems arise in the distribution of scarce resources such as kidney transplants and public housing. Invidious and almost impossible judgements have to be made about people's relative needs or deserts. There is a flourishing and prolific literature debating the desirability and morality of taking such decisions by lottery,as Duxbury's extensively referenced book demonstrates.

But the main purpose of this scholarly work is to ask whether lotteries have a role to play in judicial decision making in countries like Britain and the United States, where "faith in reason" is a long- established central tenet of jurisprudence. The merit of the book is not merely "to turn a ludicrous idea into a dubious one" (as Duxbury suggests sceptics might conclude) but to erect another landmark on the road towards acknowledging that some predicaments are beyond human reason and that it is rational to recognise this and rational, rather than irrational, to resort to sortition to resolve them.

Duxbury's book, like a butterfly, ranges widely, often recrossing or revisiting trails that he left earlier. One reason is, as he points out, that whenever he makes a case for lottery decisions he immediately sees objections to it, but the counter-case also has to be qualified and then the original case reviewed. For example, resolving the custody case by lot to avoid procrastination and fine tuning may cause the losing parent to feel that his (or her) case has been trivialised and his (or her) merits undervalued; but dragging the case through the courts, with the inevitable exposure of the parents' private lives and proclivities, and then reaching a decision based on (often spurious) reasons may leave the loser feeling demeaned and mortified, and with a residue of anger. But, then again, "randomly to grant custody ... simply renders explicit something which is likely to occur implicitly within the adjudicative process anyway".

Chance is endemic in our legal system, yet faith in reason persists. Some legal theorists consider the reasoning process itself more important than the quality of decision reached and, according to Duxbury, people want "not right answers but attributable answers". The introduction of a blind, peremptory decision-making device like the lottery, which has no human agency and cannot give reasons, is disturbing and disempowering. Given the many objections to legal lotteries that he recounts, why does Duxbury conclude, albeit tentatively, in favour of a limited use of sortition? Partly because he wishes to debunk the "pseudo-rationalism" of many legal academics by confronting them with this "counter-intuitive idea" and partly because he believes the lot really could function as a heuristic device.

Much of the book is a discussion of the lottery principle in general, including the use of sortition in ancient democracies and recent proposals to reintroduce lotteries into modern representative democracy. The book's critical apparatus seems disproportionately large if its sole purpose is to convince an audience of legal academics of Duxbury's relatively modest concluding proposition that some adjudicators might operate "in the shadow" of a lottery, which would be invoked to decide the case if they could not adjudicate within a given time. Perhaps lawyers take more convincing than other people. But because of its breadth and compendious references, the book is an excellent resource for the growing body of academics interested in chance and lotteries. Among legal academics, Duxbury's explicit challenge to the supremacy - or hubris - of reason should spark some fireworks.

Barbara Goodwin is professor of politics, University of East Anglia.