The usefulness - or otherwise - of economics as a tool is well illustrated by its application to problems which the ordinary person would not think of as a matter for economists at all. Take organised crime. A recent book* applies economic analysis to this thorny subject and comes up with some surprising conclusions.
The motivation for the research - many of whose authors are, not surprisingly, Italian - is the observation that the activities of organisations such as the Sicilian Mafia are driven by the desire to make money.
There is no reason they should not be analysed like any other business activity.
As in any other business, criminal activity can take several forms. Broadly speaking, criminal ''firms'' operate in either very competitive markets, such as loan sharking and prostitution, or in oligopolistic ones, such as money laundering and narcotics smuggling, where a few big organisations dominate. The table lists the factors tending to favour the second type of organisation.
The broader the range of activities treated as illegal by the government, the greater will be the incentives for these larger-scale organisations to form. Heavy fiscal burdens on legal markets will also create incentives for an illegal market.
Typically, the big businesses of the underworld tend to be co-ordinated from a central location with ''vertical'' links between groups which can provide the variety of skills needed - from violence to accountancy.
Crime is not quite like any other business. Contracts are not enforceable in any conventional way, although violence offers an alternative, while property rights are ill-defined.
Even so, the economic method offers some useful insights into crime. For example, it helps to understand why so much of the male US population is in prison if crime is seen as paying more than other inner-city job opportunities.
One of the papers in the book considers the design of effective deterrence policies - or ''regulation'' of the organised crime industry, as it puts it. There are some pretty obvious difficulties in trying to deter something like the Mafia. There is a risk that inappropriate action simply makes them fight back, increasing the criminals' investment in violence and corruption.
However, suppose the government has the objective of minimising the profit of the Mafia. The economists analyse the problem using ''game theory'', the mathematical technique which models behaviour as if it were like a game where the players adopt strategies to try for the best outcome.
They conclude that the crime business will be least profitable when there is a kind of collusion between government and criminals. The government's best strategy is to allow criminal firms to get away with low profits but to use the full force of the law to put them out of business if they start up activities which are too profitable. In this case, the firms have a choice between steady, low-profit business or a high risk of no business. Those that are prepared not to earn ''too much'' are rewarded with survival.
The analysis is complicated if the criminal firm is assumed to be able to save up its past profits, because building up capital reduces the probability that law enforcement will put it out of business. However, that would also increase the government's pay-off for destroying a firm that had enjoyed big profits in the past.
The authors conclude: ''The state, in designing its deterrence policy, should take advantage of the nature of the organised crime sector.'' It will do better to take advantage of the economic incentives rather than simply throwing money at law enforcement regardless.
This is all very well, but the economists' approach raises several objections. Some are obvious: how much is ''too much''? How should the analysis take account of the fact that big criminal firms operate in many markets?
A second objection is that the assumption that closing the ''firm'' is the only effective form of regulation is extreme. For instance, the anti- corruption movement in Italy seems to have decreased the number of criminal events while increasing the risk of each crime; but this has apparently raised the profitability on each single criminal event by raising the going rate for bribes demanded by officials and politicians who are prepared to take the risk of engaging in criminal activities.
More seriously, is it right to assume that the government's aim is to make sure crime does not pay too much? Or do governments actually want to minimise the amount of criminal activity rather than its profitability? In the case of, say, uranium or Semtex smuggling they probably want to do the latter. In addition, as Carlo Scarpa, commenting on the analysis, points out, although the idea of an implicit deal with the Mafia has found acceptance with some politicians, it ''has led to a substantial loss of credibility of the state in several fields''. There is what an economist would describe as a ''reputational externality'' for the government in the recommended form of regulation.
He continues: ''The layman would probably be puzzled by the idea of regulating something that in theory should not exist, that is criminal activity. Economists (cynical as we are) are instead quite used to this idea."
However, governments have considered both implicit and explicit bargains with criminals. Several years ago Italy's finance minister (an economist) proposed guaranteeing public sector jobs to cigarette smugglers who agreed to give up their business and surrender their boats. The idea was squashed by opponents who thought it would actually create an incentive for people to become cigarette smugglers in the first place.
A second example of a near miss for economics in influencing public policy is the periodic debate about whether cannabis should be legalised. The economists' line is that banning cannabis altogether raises its price and makes dealing an excessively profitable business. Declaring it illegal creates a high barrier to entry into the cannabis supply business, even though physical barriers are low - after all, many otherwise law-abiding citizens seem able to grow it on their windowsills.
Legalisation would destroy the excess profits and undermine the criminal cannabis-manufacturing and dealing industry. The government could even raise some revenue by taxing the drug like tobacco or alcohol. That would give the tax inspector rather than the police inspector an interest in the nation's window boxes and back gardens.
The useful economic analysis runs, of course, into profoundly non-economic objections. Yet even if public opinion is in the end moved by other considerations, economics remains a helpful tool for thinking about the appropriate policies for tackling crime.
* The Economics of Organised Crime, edited by Gianluca Fiorentini and Sam Peltzman, Cambridge University Press and CEPR, pounds 19.95
Conditions favouring mafia-controlled cartels
Product differentiation .......................... Low
Barriers to entry ................................... Low
Technology .......................................... Low
Labour ................................................ Unskilled
Demand .............................................. Inelastic
Number of firms ................................... Large
Size of firms ........................................ Small
Unionisation ........................................ Present