262 



REPORT — 1887. 



From this condition, owing to the unequal consumption ^ of difFerent individuals 

 it follows that the precision of our calculation cannot be great. That is to say, we 

 cannot he certain that between considerable limits some other ratio than the one 

 which we have chosen would not be as good as the one which we have chosen. 



It may be worth adding that even if we could suppose that all commodities 

 were consumed in the same proportions by all individuals, yet the mere difference 

 in the size of fortunes and of debts would introduce an inaccuracy. To show this 

 let us first suppose that all fortunes would be equal but for the payment of debts ; 



Fig. 1. 



and let us represent the average amounts of commodities consumed by the height 

 of the columns in the annexed diagram, the divisions of the horizontal line being 

 equal. Now suppose a person, from being a consumer of the average amount of 

 each article, becomes debtor to the extent of a certain sum, expressed in Units - of 

 tabular standard. Theoretically he would retrench something of his expenditure 

 on each article, contracting as it were the margin of final utility. He might thus 

 fall back upon the curve HjHj' instead of the original boundary. And if his debt 

 increased he might have to fall back upon an interior frontier, the next isohedone, 

 as we might call this family of curves. Conversely, in the case of a creditor. 

 Now, in order that our standard should be applicable to debts of various sizes it 

 is virtually assumed that the ratio HHj : H'll/ is the same as HjH., : H/Hj', 

 and so on for other columns and curves. But this assumption is without evidence, 

 or rather contrary to evidence. Or, if it be held sufiicient that the standard should 

 represent the utility corresponding to the avorage debt, still even for this purpose 

 our method of determining the proportions (by the totals consumed) is arbitrary — 

 ii fortiori when we admit all kinds of inequalities of fortune and other irregulari- 

 ties. Thus it may plausibly be contended in virtue of the analogies of Fechuer's 

 law that, ■\!vhere the total wealth of a people has increased, an equal quantity of 

 utihty is represented by a larger quantity of wealth.'' In tliis case Method A B c D 

 (explained below) might be the legitimate deduction from the principle on which 

 we here suppose method A B C D to depend. 



It is important to realise how loose is the character of the calculation even 



' Cf. Professor Marshall, Industrial Conference. 



^ The term Unit is here employed in the sense proposed by Professor Marshall, 

 Contemporary Rcriew, March 1887. 



^ The standard defined in this section, the Consumption Standard as it may be 

 called, appears to be particularly appropriate to the case in which National Wealth is 

 regarded as a constant quantity. Otherwise there is apt to arise a divergence between 

 two attributes which we have hitherto assumed to be conjoined, namely, the condi- 

 tion that the Unit should be constantly equivalent to the same quantity of valuables, 

 and that it should afford, on an average at least, the same quantity of value-in-use, 

 the same ' Final Utility.' For, according to the Lam of Diminishing Utility (ex- 

 pounded by Laplace, Jevons, and others), the same increment of means tends to 



