Setting punishment to ''fit the crime'' has always been a primary goal of the US criminal justice system. Unfortunately, that objective is not always lived up to in practice in this age of plea bargaining and overcrowded prisons, both of which tend to work in favor of short sentences and early release dates. Thoughtful attention should therefore be given to the conclusion of a new six-year study by the Rand Corporation: namely, that sentencing practices based on a consideration of overall criminal behavior by a wrongdoer could help to reduce the crime rate and the prison population.
The Rand study recommends that ''chronic'' criminals - heavy repeaters, in other words - be given longer terms of confinement. Less active wrongdoers would receive shorter sentences under most circumstances. The result, according to the study, would be to cut the robbery rate, for example, by as much as 15 percent, and the number of robbers in jail by 5 percent.
Some civil libertarians argue that such an approach may be overly tough, ignore individual differences, and scuttle the concept of rehabilitation in favor of an emphasis on punishment alone. Yet that does not follow. A wrongdoer would still have the opportunity for rehabilitation.
At the same time there would be an element of clear justice in such an approach. Since a disproportionate number of crimes are committed by repeat offenders who are in effect professionalizing their wrongdoing - making a career of crime - it would seem logical that they should be given longer confinements. Rand found, for example, that some robbers committed less than five robberies a year, while others committed more than 80 a year.
Those persons who repeatedly prey upon society, or who do so despite numerous periods of incarceration, should be dealt with in a way that best protects the public as well as themselves.
The Rand study deserves careful consideration by jurists and legislative bodies as one way of helping to reduce the crime rate - and the criminal population.