Politics works better in some areas than others. Fairfax County has wonderful parks, libraries, and schools. French politics has brought about a good health care system. There are many other examples. The mechanisms are various and often involve accident. Sometimes governments luck into good institutional arrangements. Sometimes a far-sighted visionary is at work. Tiebout competition, or efficient Beckerian bargains across interest groups, may kick in. Sometimes the median voter rules, that median voter is a good judge of outcomes, and what is good for the median is good for the nation as a whole. Sometimes venal interest groups control politics, yet the desires of those groups happen to coincide with welfare maximization.
We can, in principle, rank policy areas along a spectrum. Assign a "10" to the most efficient policy-generating areas and assign a "0" to the least efficient. Don’t worry too much about what the scale means, this is Blog Land.
Where do you put trade restrictions along this scale?
I give them a 1.5, at best a 2. I think trade restrictions are hardly ever generated by processes which coincide with the general welfare. I view trade restrictions as almost always motivated by the classic, crude "diffusion of costs, concentration of benefits" logic.
If we redefine the problem more broadly, it could be said that trade policy as a whole gets an 8 or a 9. Most of the wealthier countries, agriculture aside, have fairly free trade, as they ought to. But what if we focus on evaluating only the quality of the trade restrictions? How good are the restrictions we get in tracking the ideal welfare-improving restrictions?
I say 1.6438. That is why I think Bhagwati is essentially correct to be fighting "the last war." What is your number?