the table reports the coefficient, with the !-statistic in parentheses below.
Equation 1 provides a preliminary examination of the impact of skill. Skill345 is a binary
variable which takes the value 1 if the respondent’s skill level is either 3, 4 or 5, and zero ifhis or her
skill level is 1 or2.18 In what follows, we will loosely speak of the variable as indicating whether a
respondent is high-skilled or not. The equation relates protectionism not only to this variable, but to
Skill345 interacted with the country’s GDP per capita (measured in thousands ofUS dollars).19 The
results indicate that the high-skilled are more predisposed towards free trade than the low-skilled, and
the interaction term suggests that this effect is greater in rich countries than in poor countries,just as
Heckscher-Ohlin theory predicts. Equations 2 through 9 indicate that the latter finding is quite robust
to the inclusion of other variables (although the former is not): since it is the interaction term which is
crucial in testing Heckscher-Ohlin theory, it seems that, so far, preferences are entirely consistent
with that theory.
Equation 2 in Table 4 establishes that patriotism and chauvinism are both positively related to
protectionism, as expected, with chauvinism having a larger impact. A glance across the table
confirms that these findings are also extremely robust to the choice of specification: the hypothesis
that nationalism is an important determinant of attitudes towards trade policy is, on the basis of these
results, confirmed. However, nationalist attitudes are not all-important: taking them into account still
leaves scope for skill differentials to have an effect in ways that, as noted above, are consistent with
economic theory.
National mobility has no consistent effect on attitudes, which is perhaps a surprise, but those
who consider themselves to be internationally mobile, as well as those who have lived abroad in the
past, are more positively disposed towards trade, supporting Rodrik’s (1997) arguments. Older people
tend to be more protectionist, as are rural dwellers, women, Roman Catholics, and those in marital
18 We will explore other skill-related specifications later.
19 These are the World Bank’s data for 1995 PPP-adjusted GDP per capita, in 1995 international
dollars.
15