2. Trade and wages: a theoretical digression
In a simple 2 country, 2 commodity, 2 factor (skilled and unskilled labor) framework the
links between trade and wage gaps are clear. Assume that Mexico is abundant in unskilled labor, and
that the US is abundant in skilled labor. Let NAFTA reduce the barriers to trade between the two
economies: Mexico will export and produce more unskilled-labor intensive products, and the demand
for unskilled labor will increase; while it will import more skill-intensive goods, which will lead to a
reduced demand for skilled workers, and a fall in the skilled wage. Wage inequality in Mexico will
thus fall; unskilled Mexican workers should favor free trade, while skilled workers should oppose it.
Of course precisely the opposite will occur in the US: skilled wages will rise and unskilled wages
will fall; thus in the US it is the unskilled who should be protectionist, and the skilled who should
favor free trade.
There have been several recent trade-theoretic papers exploring the ways in which the links
between trade and income distribution become more complicated once we move away from simple
2x2x2 models. These contributions have been largely motivated by the fact that, while simple
Heckscher-Ohlin logic might seem to suggest that globalization should lower skill differentials in the
South, in fact differentials have widened in several developing countries (DCs) during the past 20
years. One possibility is that FDI, associated with trade liberalization, might lead to new skill-
intensive activities being introduced into DCs (Feenstra and Hanson 1996). Under such
circumstances, the relative demand for skilled labor could rise in the South. Alternatively, if skilled
labor and capital are complementary to some natural resource (e.g. minerals), then liberalization in a
resource-rich DC might increase skill premia and inequality overall (Kanbur 1999). Clearly, allowing
for more than 2 factors of production, or for links between trade and factor flows, or between trade
and technology transfers, leads to theoretical ambiguity regarding the relationship between trade and
wages. Nonetheless, it is interesting to ask whether the predictions of simple 2x2 models help to shed
light on policy preferences generated in the admittedly complicated real world.
Even within a simple two-factor framework, however, there is another set of complications