James E.Alt and Michael Gilligan 339Furthermore, even if factors were perfectly mobile, high costs of collective
action might mean that many of the factor owners would have little incentive to
take costly political action to affect trade policy, free-riding instead on the political
action of others. Then, as in the northwest quadrant of the Figure 2, it is likely
that there would be no coalitions over trade issues (except perhaps in some cases
where capital is the scarce factor). In a capital-rich country, for instance, labor in
one industry would let labor in other industries lobby for protection, and as a
result very little lobbying would be done. Depending on how far you push the
assumption of mobility, it could even be that exit—in the form of moving to
another employment, emigration or capital flight—would be a far more common
response than lobbying.
Let us now revert to the assumption of easy collective action, but assume factors
are specific. The southeast corner describes just such a political economy, where
individual industries seeking protection for their products are opposed by the
consumers of those products. Assuming no collective action problems in this
domestic political economy, however, any consumers might participate in trade
politics, however small their stake in the issue; they would not free ride, relying
on their fellow consumers to do the lobbying for them. An industry interested in
protection could only really win, then, if it banded together with other industries
interested in protection and lobbied for protection for all of them. The coalition
that would emerge in such a situation would pit import competers against non-
tradeable producers and exporters. The problem with this coalition is that all the
protected industries might be worse off from this “universalistic logroll” than if
they simply accepted free trade, since the costs to them of the protection to all the
other industries might very well be higher than their gains from protection of
their own industry. Therefore, such a coalition is inherently unstable. The existence
of collective action problems is, thus, essential to the Ricardo-Viner explanation
of trade policy coalitions generating protection for individual industries, as this
quadrant serves to show.
The southwest quadrant contains the ideal type of trade policy “coalition”
described by Pareto, Schattschneider, Olson, and the endogenous tariff literature.
In that ideal type, collective action problems exclude most of the public from
participating in trade politics. In fact, there really are no coalitions at all: there
FIGURE 2. Coalition Possibilities: The Effects of Factor Mobility and Collective
Action Costs