Professor Sellin³⁵,⁴⁰ compared homicide rates for states with similar outlook in the U.S.A. He selected five sets of three states each and compared their crude homicide death rates. The comparisons covered a 43 year span for each set of states, extending from 1920 to 1963. In each set, at least one of the three states did not provide the death penalty for all or a part of the period while the others did provide it. Each of the three states in each set borders one or both of the other two. The figures showed clearly that homicide death rates in all the states have followed the same trends, whether or not the death penalty was provided. In all of the 15 states covered by the comparisons, homicide death rates reached peaks in the 1920's and early 1930's, then followed a general downward trend, leveled out in the 1940's and continued through 1960 at about that level. Comparison of trends and rates revealed no differences among adjacent states with and without the death penalty which can be ascribed to either its presence or absence. Professor Sellin found that:

The inevitable conclusion, therefore, is that the presence of the death penalty, either in law or practice, does not influence homicide death rates. As Prof. Sellin puts it:

"The important thing to be noticed is that, whether the death penalty is used or not, or whether executions are frequent or not, both death penalty States and abolition States show rates which suggest that these rates are conditioned by other factors than the death penalty".