The concept of Pareto optimality leads us to situations where it is impossible to make someone better off without making others worse off. But it does not offer any guidance to the choice involving making some better off and some worse off. In Chapter 3, we discussed welfare criteria that provide sufficient conditions for a social improvement. But what we obtained there is still incomplete in two aspects. First, in cases where the compensation test is met but the distributional one is not, or vice versa, we have no answer. Second, the criterion (Little’s) we find acceptable is based on some judgements about distributional desirability. How do we obtain such judgements? In Chapter 4, the discussion of the magnitude of welfare changes reduces (not completely) the inadequency of Chapter 3 in the first aspect. But the incompleteness with respect to the second aspect has not been overcome. Thus we still need something more. If we have a specific social welfare function (SWF), the vacuum can then be filled. But how do we get a specific SWF? If there is a super-Stalin, he may say that, ‘The SWF should just be my preference function. Whatever I prefer or whatever I think is good for society should prevail.’ But the solution by resorting to a super-dictator is not a palatable one to most people. The problem of social choice is to see whether we can derive our social preference based on the preferences of individuals, satisfying certain reasonable conditions. This seems simple enough and nothing more than the basic requirement of (minimal) democracy. But a formidable difficulty is encountered in this problem of social choice – the impossibility theorem of Arrow.
Unable to display preview. Download preview PDF.