the population.) As can be seen from the figure, however, the Barro-Lee figures for schooling in
several transition countries are very high: for example, average schooling is higher in Slovakia,
Bulgaria, Latvia and Poland than in the Netherlands, Ireland, and Austria. We doubt whether these
figures provide a genuine reflection of the economically relevant human capital endowments of these
economies, and thus doubt the usefulness of Figure 2asa test of Heckscher-Ohlin theory.
Leaving aside the testing of trade theory for a moment, what other insights can be gleaned
from Table 6? Patriotism and chauvinism are significantly related to protectionist attitudes in
virtually all countries, with the size of the coefficients being fairly consistent throughout. Clearly, our
findings regarding nationalism and trade policy preferences reflect a quite general phenomenon (at
least in this sample of countries), ratherthan depending on strong correlations in a few countries.
There are several exceptions to the general rule that the old are more protectionist: they are
significantly more likely to favor free trade in East Germany, Poland and Estonia, all former
Communist countries; and coefficients are statistically insignificant in several countries. The finding
that women tend to be more protectionist than men is, however, a fairly general one.
Table 7 investigates the data further, by estimating models separately for various sub-samples
of the data. The first four columns take a further look at cross-country variations in the data by
splitting the sample into its western and former Communist components.26 The high-skill coefficient
in column 2 is positive (and statistically significant at the 10% level) in the east, and the interaction
term between skills and GDP is negative and significant, consistent with both Heckscher-Ohlin theory
and Figure 1. The results in column 2 suggest that in countries with per capita incomes below $4215
(slightly lower than the Russian income) high-skilled workers are more likely to be protectionist,
whereas they are more likely to support free trade in countries with higher incomes. The interaction
term is insignificant within the west, which is again consistent with Figure 1: there may be too little
variation in the western data to pick up any relationship here.
The impact of gender appears to be much stronger in the west, although it is still strong in the
26 The Phillippines is not included in either sample.
19