THERE APPEAR TO HAVE BEEN NO EFFORTS TO TEST THE DETERRENT EFFECT OF PUNISHMENT SCIENTIFICALLY UNTIL THE 1950'S. EARLY STUDIES OF THE RELATIVE DETERRENT EFFECTS OF THE DEATH PENALTY AND LIFE IMPRISONMENT WERE LIMITED BY INADEQUATE TECHNIQUES OF STATISTICAL ANALYSIS. MORE RECENT STUDIES IN DETERRENCE HAVE COME PRIMARILY FROM ECONOMISTS AND SOCIOLOGISTS. ECONOMISTS TEND TO CONDUCT THEIR STUDIES UNDER THE ASSUMPTION THAT PUNISHMENT DOES DETER CRIME; SOCIOLOGISTS PRESUME THAT IT DOES NOT. ALL OF THE STUDIES HAVE BEEN HAMPERED BY INADEQUATE DATA. STUDIES APPLYING ECONOMIC CONCEPTS TO THE PUNISHMENT-DETERRENCE QUESTION REPEATEDLY HAVE FOUND SUPPORT FOR THE DETERRENCE THEORY. SOCIOLOGISTS, USING DIFFERENT STATISTICAL TOOLS THAN ECONOMISTS AND GENERALLY INTENDING TO CONFIRM THE CONVENTIONAL BELIEF AMONG SOCIOLOGISTS THAT CRIME CANNOT BE DETERRED BY PUNISHMENT, ALSO HAVE PRODUCED EVIDENCE SUPPORTING THE DETERRENCE THEORY. ALTHOUGH RESEARCHERS GENERALLY HAVE FOUND THAT THE FREQUENCY WITH WHICH A PUNISHMENT IS APPLIED IS OF GREATER SIGNIFICANCE TO DETERRENCE THAN IS THE SEVERITY OF THE PUNISHMENT, THE FREQUENCY-SEVERITY QUESTION IS RELATIVELY UNIMPORTANT AND NOT PARTICULARLY AMENABLE TO STATISTICAL TESTING. IN CONSIDERING THE CONCEPT OF DETERRENCE, IT SHOULD BE REMEMBERED THAT THE FACT THAT A CRIME CAN BE DETERRED BY A PARTICULAR PUNISHMENT IS NOT A SUFFICIENT ARGUMENT FOR IMPOSING THAT PUNISHMENT. THE PERSISTENCE OF REHABILITATION AS THE DOMINANT RATIONALE FOR PUNISHMENT DESPITE THE ABSENCE OF ANY SCIENTIFIC SUPPORT REFLECTS THE COMMON FALLACY IN THE SOCIAL SCIENCES THAT 'ALL GOOD THINGS GO TOGETHER.' GIVEN THE CHOICE BETWEEN PREVENTING CRIME BY TRAINING THE CRIMINAL TO BE GOOD (REHABILITATION) OR DETERRING CRIME BY IMPOSING UNPLEASANTNESS ON CRIMINALS (PUNISHMENT), PEOPLE WOULD LIKE TO CHOOSE THE FORMER. IN FACT, THE PUNISHMENT APPROACH HAS BEEN SUCCESSFUL AND THE REHABILITATION APPROACH (AS YET) HAS NOT. THE REAL CHOICE IS BETWEEN DETERRENCE AND HIGHER CRIME RATES. (LKM)