"Charm City" -- or if you prefer, "the city that reads" -- is being transformed into "the city that bleeds." The number of homicides in Baltimore last year hit 335, topping the former high (330) set back in 1972.

The response to that tragedy was as disturbing as the murders themselves. Our law-enforcement officials explained that the murder and mayhem have something to do with drugs and guns. Mayor Schmoke blamed the Reagan and Bush administrations.

No serious student of crime claims that the solution to the crime puzzle is simple; there are many interrelated and complicated causes. However, one cause is clear-cut: the economics of crime.

Gary Becker of the University of Chicago was awarded the Nobel prize last year for his work on that topic, among others. Professor Becker's work shows that crimes are not irrational acts, but are voluntarily committed by people who compare expected benefits with expected costs. Hence, if crime rates are surging one reason is that some criminals calculate that crime pays.

The cost of committing a crime can be measured by determining the "expected punishment" associated with various criminal acts. Expected punishment is calculated by first multiplying four probabilities times each other: that of being arrested for a crime after it is committed; that of being prosecuted if arrested; that of being convicted if prosecuted, and that of receiving punishment if convicted. The product of that arithmetic is the probability of being punished. To complete the calculation, we must next multiply the probability of being punished times the penalty for an offense.

The cost of committing a crime can be increased or decreased by adjusting either the probability of being punished or the penalty. Another Nobelist, George Stigler, used to illustrate that point to students at the University of Chicago with the following conjecture:

Suppose during one day of the year, chosen at random, one parking violator was caught and then publicly immersed in boiling oil. Professor Stigler suggested that despite the low probability of being punished, parking violations would be dramatically reduced.

Let's look at some numbers. About 7 percent of the burglaries in the United States result in an arrest; of those arrested, 87 percent are prosecuted; of those prosecuted, 79 percent are convicted, and of those convicted, 25 percent are sent to prison. If we multiply those four probabilities together, we find that a burglar has only a 1.2 percent probability of being punished with a prison term. Since the average prison time for burglars is 13 months, the average risk is only 4.8 days in prison (1.2 percent times 13 months). So, burglary pays if the prospective thief values stolen goods more than 4.8 days behind bars.

The National Center for Policy Analysis in Dallas keeps tabs on the expected punishment for various crimes. The following cost numbers show the expected prison time for each crime category: murder 1.8 years, rape 60 days, robbery 23 days, arson 6.7 days, aggravated assault 6.4 days, burglary 4.8 days, larceny-theft 3.8 days and motor-vehicle theft 1.5 days.

The expected punishments are not only low, but falling like a stone.If we lump all serious crimes together, the average time expected in prison has fallen from 24 days in 1950 to 9 days today. No wonder crime rates have surged over time.

Baltimore's lads and lasses may have trouble with their school math, but they are street-smart, and can calculate the likelihood of punishment for crime.

If we are serious about reducing crime, we must demand more effective law enforcement and a tougher criminal-justice system -- including, if not Professor Stigler's boiling oil, stiffer penalties to make sure that crime doesn't pay.

Steve J. Hanke is a professor of applied economics at The Johns Hopkins University.

@