ABSTRACT
In the previous chapter I discussed the aggregation problem at some length. The problem as discussed in the first two sections has been called one of ‘consistent aggregation’ by Green (1964), among others. He says that aggregation is consistent ‘when the use of information more detailed than that contained in the aggregates would make no difference to the results of the analysis of the problem at hand’ (Green 1964: 3). This terminology is not used above, because I think it is confusing. Aggregate variables contain less information than individual variables do. It is obvious that in general the use of more detailed information yields more determinate predictions than the use of less information. In other words, it is not surprising that aggregate variables contain enough information to yield determinate predictions only when some restrictive conditions hold true. This suggests another way of looking at the aggregation problem, a way that goes back to Nataf (1948).
