NCJ Number
56286
Journal
Social Science Quarterly Volume: 58 Issue: 4 Dated: (MARCH 1978) Pages: 616-631
Date Published
1978
Length
16 pages
Annotation
A MATHEMATICAL MODEL IS DEVELOPED TO TEST THE DETERRENT EFFECT OF BOTH IMPRISONMENT AND THE DEATH PENALTY. IT IS USED WITH CENSUS AND HOMICIDE DATA FROM 1950 AND 1960. REASONS WHY THE RESULTS ARE INCONCLUSIVE ARE SUGGESTED.
Abstract
THE MODEL INCORPORATES 15 VARIABLES: HOMICIDE RATES, PROBABILITY OF INCARCERATION, LENGTH OF INCARCERATION, PROBABILITY OF THE SANCTION OF EXECUTION, MEDIAN FAMILY INCOME, PERCENTAGE OF FAMILIES WITH LESS THAN HALF OF MEDIAN INCOME, PERCENTAGE OF NONWHITES IN POPULATION, PER CAPITA POLICE RESOURCES, UNEMPLOYMENT, RATIO OF MALES AGED 14-24 YEARS TO ALL MALES, URBANIZATION, MALE TO FEMALE RATIO, STATE POPULATION, A DUMMY VARIABLE DISTINGUISHING SOUTHERN FROM NORTHERN STATES, AND MEDIAN YEARS OF SCHOOLING. FULL DATA WAS AVAILABLE FOR 43 STATES FOR 1950 AND 47 STATES FOR 1960. A CORRECTION IS MADE FOR THIS VARIANCE. NO SIGNIFICANT RELATIONSHIP COULD BE FOUND BETWEEN PROBABILITY OR SEVERITY OF INCARCERATION OR PROBABILITY OF EXECUTION AND HOMICIDE RATES. IT IS SUGGESTED THAT THE MODEL MAY NOT CONTAIN THE CORRECT SET OF VARIABLES, THAT THE LOG-LINEAR FORM MAY BE INCORRECT, OR THAT THE PERCEPTION OF SANCTIONS IS NOT SIGNIFICANT ENOUGH TO DETER HOMICIDES. IT IS ALSO SUGGESTED THAT THE VARIOUS TYPES OF HOMICIDES NEED TO BE BROKEN DOWN AND EXAMINED. REFERENCES ARE INCLUDED. (GLR)