Ronald W. Jones and José Scheinkman show that under very general conditions the factor returns change with output prices as predicted by the theorem.
An additional robust corollary of the theorem is that a compensation to the scarce factor exists which will overcome this effect and make increased trade Pareto optimal.
Therefore, the early versions of the theorem could make no predictions about the effect on the unskilled labor force in a high-income country under trade liberalization.
Indeed, Feenstra called the Heckscher–Ohlin model "hopelessly inadequate as an explanation for historical and modern trade patterns".
[6] Papers that compare output prices with changes in relative wages find moderate-to-strong support for the Stolper–Samuelson theorem for Chile,[7] Mexico,[8] and Brazil.