ABSTRACT
Based on the premise that threat results in higher authoritarianism and that one aspect of authoritarianism is the tendency to identify with powerful figures, McCann and Stewin (1987) hypothesized that the power of presidents (Winter, 1976) from Coolidge to Reagan in 1980 would be positively related to objective and subjective threat indicators preceding elections, and that the percentage of the popular vote received would be negatively correlated with the absolute degree of discrepancy between prevailing threat and the power of the winner. Support was gained for both hypotheses. In this replication, a new social, economic, and political threat (SEPT) index and revised power motive scores reported by Winter (1987) were used. Both hypotheses were supported, using the SEPT index and the original power scores from 1920 to 1984. SEPT and revised power scores supported the second hypothesis. Revised and original power scores were especially divergent for Truman and Kennedy, and when data for 1948 and 1960 were excluded, support was also found for the first-hypothesis, using the revised power scores and the SEPT index.