The latter circumstance may be very common. Two economies with similar relative prices will gain little from trade (in the limit, a gain of zero), yet as your diagram shows, movement toward equilibrium may nonetheless produce huge dislocations even near the zero-gain limit, albeit with lesser driving forces, hence slower change.

Thank you for deepening my understanding of this issue.

Regarding current dislocations, the Economist recently observed that the entry of China into the world economy is creating large dislocations (with large driving forces) precisely because its entry is changing the relative prices of labor, capital, and energy. Because China has very different relative costs, its entry pushes toward large changes at equilibrium, large gains from trade, and large, hard-driven dislocations.

Considering dislocations more broadly, I see a substantial nonlinearity between the aggregate magnitude of the dislocation and the magnitude of its effects. At least two mechanisms seem clear: First, individuals encounter new modes of suffering as the inputs to suffering increase -- eg, passing the threshold of losing a house. Second, social effects are nonlinear in the number of individuals affected -- eg, if one person in a hundred encounters a disaster, each has 99 others to help, but if all encounter a disaster, each has no one to help, hence the effect is more than 100 times greater.

I expect dislocations resulting from technological change to increase sharply over the next few decades (I base this judgment on my own assessment of specific emerging developments). These technology-driven dislocations will likely be difficult to avoid, for various reasons. From the nonlinearity of effects, one may argue that dislocations from trade will therefore be more costly than they seem.

For the record, I say this as a habitual, reflexive free-trader, and I think that greater support for freer trade is probably, at the margin, a good thing, in part because so much of the opposition to trade is based on false beliefs.

Words and ideas I offer here may be used freely and without attribution.