There was some disagreement with my assertion that Democrats were — effectively — not as protectionist as many have argued. Here are some more thoughts on the matter, as the Administration prepares the case for countervailing duties on Chinese imports .
In their April 11th edition, the London-based consultancy IDEAglobal observed that of the 38 House and Senate bills targetting China proposed since 2005, 21 enjoyed bipartisan co-sponsorship. Of the single-party sponsored bills, 9 were Democratic, and 8 Republican. The common denominator in the bills’ sponsors was geographical in nature — that is a relationship to industrial states such Pennsylvania, Michigan, Ohio, NY, North Carolina, Virginia, Illinois and Indiana.
Additional support for the conjecture that a specific-factors — rather than factor-proportions — interpretation of protectionist impulses comes from this paper by Michael E. S. Hoffman presented at the 2006 SGE conference, in which he examines some polling data on views regarding international trade policy:
An interesting picture emerges from the regression results when they are
considered together. Educational attainment is most clearly associated with proglobalization
attitudes. Income is not. The latter result is at odds with Scheve and
Slaughter (2001), and Mayda and Rodrik (2002). If education affects trade policy
attitudes irrelevant of income, then the effect of education may be based more on
information (see Coughlin  for a discussion) and less on factor endowments.
Blinder and Krueger (2004) conclude that greater education leads people to be better
informed about economic policy issues, which could certainly explain the effect of
college education in this case.
The partisan effect is perhaps the most interesting. Party affiliation is not
significantly related to opinions on globalization or free trade in general. However,
Republicans are much more likely to support the steel tariffs that President George W.
Bush approved in 2002. This became a signature issue for Bush (The Economist,
February 14, 2002). Democrats are significantly more likely to be pro-NAFTA, which
was a signature issue for Bush’s predecessor, President Bill Clinton. Rather than being a
pro-capital party, as predicted, rank-and-file Republicans instead take positions espoused
by party leaders. On the other side, Democratic voters are not synonymous with labor in
the Heckscher-Ohlin sense, and instead support NAFTA despite its vilification by
organized labor. Blinder and Krueger (2004) also find that partisan ideology is a major
determinant of opinions on policy issues. This suggests that opinions are highly
malleable when it comes to trade policy. Attempts at trade liberalization (or trade
intervention!) may be made feasible through vocal party advocacy.
There are some industry effects as well. People working in manufacturing are
anti-NAFTA and anti-trade in general, and individuals living in rural areas are less likely
to support the CAFTA, perhaps due to a presumed agricultural comparative advantage of
Central American nations. These sector effects, coupled with the fact that income is not a
significant determinant of trade policy attitudes, seem to support a specific-factors view
of the economy. This is consistent with Magee (1980), but inconsistent with Scheve and
Slaughter (2001), and O’Rourke and Sinnott (2002), among others.
You might not have thought it possible, but I’m going to squeeze in housing into this discussion:
Scheve and Slaughter (2001) put particular emphasis on the role of assets in their
analysis of trade policy preferences. They do not test the role of home ownership in
isolation (and interestingly, in the results reported earlier here, the coefficient on home
ownership is marginally significant in only one of the regressions), but by interacting it
with the degree of import competition. If some of consumers’ wealth is stored in their
homes, they should be concerned about how trade policy affects the value of real estate in
their area. Trade liberalization (barriers) should increase (decrease) the value of housing
in export-oriented communities and vice versa for import-competing communities.
To test this econometrically, I follow Scheve and Slaughter and interact the home
ownership dummy variable with the measure of export exposure. It is a significant
determinant of trade policy attitudes for only one of the dependent variables: free trade
in general. …
This leads me to my next thought: That it’s not really enough to “just say no” to protectionism. In order to foster a free trade environment, consumers, workers and firms have to believe that they will gain from freer trade — and that if they lose, then the government will insure that they will have some of the net benefits acruing to society at large transferred to them as partial compensation. Until there are programs that do this (such as wage insurance — see here), we will see continued erosion of free trade (discussion of current programs is in this CRS report on TAA and workers as well as on firms).