Change in the Southern Electorate

This paper presents an analysis which measures the extent to which each of five sources of change in the party loyalties of the Southern electorate has been active during the periods 1952-1960 and 1960-1972. The measure of party loyalty is Converse's normal vote, with a set of updated parameters for the 1960-1972 period provided by Arthur H. Miller. The basic data sources are the six SRC presidential election studies ranging from 1952 to 1972, supplemented by data from the Matthews-Prothro and CSEP studies. In addition, data from the U.S. Census are used when possible in place of the estimates provided by the SRC samples. In both time periods, we find that change in the partisanship of individual Southerners is the dominant mechanism of change in the normal vote. Inand out-migration are of moderate importance, while the impacts of death and the entrance of youth into the electorate are negligible. Using Key's terminology, we find that white conversion has resembled the process of secular realignment over the past twenty years. Black conversion, on the other hand, is best interpreted as critical change, with a sharp reversal of partisanship between 1960 and 1964.