Another point, however, is worth noting. Our real wealth would be somewhat increased in the case supposed; but if we were to turn to the money measure of wealth, the opposite result would be far more likely, For the price of boots would most likely fall to nothing, and the total value of boots, in the commercial sense, would accordingly be nothing also. This shows that money values may be a most imperfect measure of aggregate wealth; for what money values represent is the product of the quantity of the commodity and its _marginal_ utility, while aggregate wealth is _total_ utility, which is a very different thing. This, it may be observed, makes all attempts to compare the wealth of different countries or different times, and no less to construct Index Numbers of Prices, imperfect of necessity, and arbitrary in their foundations. §4. _Criteria of Policy_. The point has now been reached at which we must take into account the very important fact which was mentioned at the close of Chapter III. The maximum utility which the laws of supply and demand tend to bring about is a maximum _total_ utility indeed, but one still measured in terms of money. An unequal distribution of wealth destroys any necessary correspondence between that and the maximum _real_ utility. This consideration, however, does not affect the general validity of the conclusion that the laws of supply and demand represent what is socially desirable now or under any system. For what is at fault here is the distribution of wealth; and it is that which should be changed, in so far as it is possible to do so. Now it is important to realize that whenever it is possible to supply a commodity to poor people below cost price, it is possible to alter the distribution of wealth, for that in effect is what is done. Purchasing power, which may be taken from richer people by taxation, or which may be obtained from "collective" profits on other trading, is in effect transferred to the poor people in question, though the transference is coupled with the condition that the purchasing power must be expended in a particular way. It is _in general_ desirable that the transference should be made without this condition being attached. To this general statement, exceptions indeed exist so numerous and important as possibly to justify a great extension of social expenditure of this type. Education should certainly be provided free of charge, there are strong arguments for subsidizing housing; the provision of milk to expectant mothers, the feeding of school children, such instances can be multiplied into a very extensive list. But it is important to observe that in each case the justification of the policy rests in the presumption that the service supplied is one which it is particularly important that the beneficiaries should have, _as compared with_ the other things upon which they might have preferred to expend the equivalent purchasing power, had it been transferred to them without conditions.