Isn't there something profoundly unhelpful about a concept of efficiency which would define some allocation of resources as "unimprovably efficient", while there is a feasible re-allocation of resources which would make, say, 95 percent of people 10 percent better off, and 5 percent of people who are, let's suppose, the currently best off 5 percent, 1 percent worse off? Or is this just too obvious to even bother asking about? Peter