35
The very richest year for this is 1992; what it suggests is that doing the full
postwar analysis with just working-age welfare should not be a problem. (Table 2) From
one side, the relationship between presidential vote and working-age welfare as a specific
referent is nearly identical to the relationship between presidential vote and structural
welfare as a general notion. From the other side, the relationship between presidential
vote and behavioral welfare is not even statistically significant. Moreover, none of the
other coefficients—for foreign, race, or social policy—are substantially affected by
letting ‘working-age welfare’ stand in for both ‘structural welfare’ and ‘behavioral
welfare’. Apparently, ‘working-age welfare’ accurately represents structural welfare,
while behavioral welfare adds nothing substantial to its impact on the actual vote. And
the same analysis done for 1996 and 2000 tells exactly the same story.
v
The story of international relations as a voting influence across the postwar years
stands in marked contrast to that of social welfare. Where welfare policy offers a
consistent pattern with only the occasional idiosyncrasy, foreign policy went through four
distinct periods:
•
In the late New Deal years of 1948 through 1960, the dominant aspect of
foreign policy as it affected presidential voting was foreign engagement, the
cluster of issues involving internationalism or isolationism as American
approaches to the wider world.
•
During the Vietnam War years of 1964 through 1972, foreign engagement
was displaced by a rising concern with national security, for which attitudes
toward war in Vietnam were merely the strongest embodiment.
Convention