"Instead of attenuating this paper’s theses, heterogeneity amplifies its importance. Contemplate a scenario where Schumpeter’s fruitful capitalist destruction harms a really sizeable fraction of the future U.S. population and, say, improves welfare of another group and does that so much as to justify a calculation that the winners could be made to transfer some of their gains and thereby leave no substantial U.S. group net losers from free trade. Should noneconomists accept this as cogent rebuttal if there is no evidence that compensating fiscal transfers have been made or will be made? Marie Antoinette said, “Let them eat cake.” But history records no transfer of sugar and flour to her peasant subjects. Even the sage Dr. Greenspan sometimes sounds Antoinette-ish. The economists’ literature of the 1930s—Hicks, Lerner, Kaldor, Scitovsky and others, to say nothing of earlier writings by J.S. Mill, Edgeworth, Pareto and Viner—perpetrates something of a shell game in ethical debates about the conflict between efficiency and greater inequality. Policy aside and ethical judgments aside, mainstream trade economists have insufficiently noticed the drastic change in mean U.S. incomes and in inequalities among different U.S. classes. As in any other society, perhaps a third of Americans are not highly educated and not energetic enough to qualify for skilled professional jobs. If mass immigration into the United States of similar workers to them had been permitted to actually take place, mainstream economists could not avoid predicting a substantial drop in wages of this native group while the new immigrants were earning a substantial rise over what their old-country real wages had been."