18
Manufacturing export orientation produced a similar pattern in line with
theoretical expectations, although its effect was somewhat less dramatic. Across the full
range of observed values, manufacturing export orientation would decrease the
probability that a respondent would take an isolationist position from 0.33 to 0.11.
Although the military contribution to personal income and the agricultural sector
variables had statistically significant total effects using the standard recommended by
MacKinnon, et al. (2002), they were very small. The manufacturing sector had the most
substantively important overall effects on support for isolationism.
Table 3 presents two tests of the robustness of these findings. The analysis
presented thus far has focused exclusively on the impact of the international economy.
As the opening section noted, however, factors such as race, gender, rural residency, age,
and party identification have all been shown to influence foreign policy opinion. Strictly
speaking, these considerations have no necessary relationship to the economic
considerations discussed here. However, they are strongly related to the variables used to
indicate self-interest: income, education, and holding a professional job. These
relationships raise the possibility that the apparent relationship between the self-interest
variables and foreign policy attitudes is primarily the result of omitting these other
variables. The first column in Table 3 presents a model that includes several additional
demographic variables suggested by previous research.
[Table 3 about here.]
4
While each of these variables is interesting in its own right, their role in this context is simply to test the
reliability of the findings concerning economic interests. Considerations of space preclude a complete
discussion of their effects here. Suffice it to say that readers of the works on public opinion and foreign
policy cited in the opening section will find no surprises here.
Convention