cost per job protected

In Eco340 today I mentioned that estimates of the cost to consumers (lost consumer surplus) per 'domestic job protected' from tariffs and other forms of trade protection in the United States ran quite high, but I didn't have any actual numbers at hand to show you. So I checked a frequently cited policy study on this matter by Kimberly Ann Elliot and Gary Hufbauer from the Institute for International Economics in Washington (full disclosure: I worked there before graduate school). From the book's back cover:

the annual consumer costs per American job "saved" by "special" protection range from $100,000 to over $1 million and average $170,000. Consumers thus pay over six times the average annual compensation of manufacturing workers to preserve each job. In terms of net national welfare, the cost per protected job is about $54,000. This figure far exceeds the cost per worker of the most generous adjustment program entailing income support, retraining, and relocation..

Their conclusion seems to be that we could have yet more generous adjustment programs and still save a lot of money.

To an economist it seems like there is obvious space for a 'Pareto improvement' (like I proposed, why not give each 'protected worker' a $50,000 a year vacation retirement package, which would appear to still save consumers an average of $120,000 per job saved per year). The complication with pulling this off is political. The workers who stand to lose their jobs would never believe a commitment to pay them off that handsomely, in large part because if such costs became so explicit, taxpayers would probably balk at delivering on them.