NCJ Number
50390
Date Published
1978
Length
47 pages
Annotation
A MULTIPLE REGRESSION ANALYSIS IS USED TO EXAMINE THE RELATIONSHIP BETWEEN CHANGES IN STATES' EXECUTION RATES AND HOMICIDE RATES FROM 1950 TO 1960. THE EFFECT OF IMPRISONMENT AND SOCIODEMOGRAPHIC FACTORS ARE EXAMINED.
Abstract
THE THEORETICAL BACKGROUND OF THIS STUDY IS GIVEN IN DETAIL, ALONG WITH THE METHODOLOGY USED. TO EXAMINE THE DETERRENCE HYPOTHESIS OF AN INVERSE RELATIONSHIP BETWEEN CHANGES IN EXECUTION RATES AND HOMICIDE RATES, PARTIAL REGRESSION COEFFICIENTS ARE UTILIZED. CORRELATIONS ARE THEN COMPUTED FOR LAGGED EXECUTION RATES (EXECUTIONS NOT TAKING PLACE IN THE SAME YEAR AS THE HOMICIDE). A FINAL STUDY COMPUTES THE RELATIONSHIPS BETWEEN HOMICIDE RATE, LENGTH OF IMPRISONMENT, AND CERTAINTY OF IMPRISONMENT; THEN DOES SOCIOECONOMIC CORRELATION. IT WAS FOUND THAT BOTH CHANGES IN THE LENGTH OF PRISON TERMS FOR HOMICIDE AND CHANGES IN CERTAINTY OF PRISON TERM FOR HOMICIDE WERE POSITIVELY ASSOCIATED WITH CHANGES IN HOMICIDE RATES. FINDINGS FOR THE EXECUTION VARIABLE ARE FAR LESS CLEAR. CHANGES IN NONLAGGED EXECUTION RATES WERE POSITIVELY ASSOCIATED WITH CHANGES IN HOMICIDE RATES BUT LAGGED EXECUTIONS HAD NO POSITIVE EFFECTS. SOCIOECONOMIC FACTORS SHOWED NO SIGNIFICANT INFLUENCE. PROBLEMS ENCOUNTERED IN AN ANALYSIS OF THIS TYPE ARE DISCUSSED AT LENGTH. CHARTS AND TABLES PRESENT STUDY DATA. FOOTNOTES AND REFERENCES ARE APPENDED. (GLR)