The Economics of Welfare

Pigou, Arthur C.
Display paragraphs in this book containing:
First Pub. Date
London: Macmillan and Co.
Pub. Date
4th edition.
68 of 73

Part IV, Chapter XI


§ 1. SO far we have been considering transferences of a direct kind. There remain transferences through bounties or devices substantially equivalent to bounties. These take three principal forms: first, bounties, provided out of taxes, on the whole consumption of particular commodities which are predominantly purchased by poor persons; secondly, bounties, similarly provided, but confined to that part of the whole consumption which is actually enjoyed by defined categories of poor persons; thirdly, authoritative interference with prices, so contrived that the richer purchasers of particular commodities have to bear part of the cost of what is sold to poorer purchasers. The first of these methods is illustrated by the special subsidies which were paid on bread and potatoes during the Great War to enable prices to be kept down to what was considered a reasonable level. The second and third methods are only practicable in connection with commodities and services which are non-transferable in the sense explained in Part II. Chapter XVII. The second is illustrated by the Irish Labourers Acts, under which, not all house-building in the districts affected, but only house-building for labourers was subsidised, and by the more general provisions which were adopted to meet the post-war house shortage. The third method is illustrated by special arrangements often made in connection with the services supplied by monopolistic "public utilities." Whether these services are actually produced by private concerns or by the public authorities themselves, public authorities can, if they choose, compel sales to selected poor persons to be made at a loss, and can arrange for this loss to be made good through charges to other persons higher than would otherwise have been permitted. This plan is adopted under a number of Tramway Acts, where provision is made for a convenient service of workmen's cars at specially low fares. Thus "a recent report of the Highway Committee of the London County Council estimates that the loss involved by running the workmen's car service is £65,932 per annum."*1 The same policy is illustrated in another connection by the (pre-war) practice of the municipality of Wiesbaden, where gas supplied by means of prepayment meters—a more expensive method of supply—was charged for at the same rate as gas supplied by ordinary meters to all persons the annual rent of whose house was less than 400 marks.*2 It should be noted that this method is not necessarily confined to commodities and services produced under conditions of monopoly. Provided that the goods are, or can be made, non-transferable, it is open to public authorities to fix a charge at which anybody undertaking a named business or profession must sell whatever quantity of service is demanded by persons in a given category. The result will be to limit the number of persons entering that business or profession, till the expectation of earnings therefrom—derived jointly from sales to the poor and to other persons, for whose purchases the charges are fixed by the normal play of demand and supply—becomes about equal to that ruling in other businesses or professions of a similar difficulty and disagreeableness and involving an equally expensive training. This, of course, implies that the low charges made to the favoured category of persons are associated with charges to other categories higher than would have prevailed if the low charges had not been enforced.


§ 2. To all these methods it has been objected that they necessarily benefit unequally different poor persons whose circumstances are substantially similar. Professor Knoop writes, for example: "It is difficult to see why artisans, mechanics, and day labourers, who travel in the early morning, should receive privileges which men and women serving in shops, clerks, and others, who are no better off financially, do not enjoy."*3 It may be replied that, if a thing is good in itself, the partial realisation of it cannot rightly be condemned on the ground that complete realisation is impracticable. We are not, however, concerned with the validity either of this objection or of the different and more forcible objection, from the side of fairness, which can be urged specially against the third method, namely, that it throws the cost of helping the poor upon particular persons, instead of upon the taxpayers generally.*4 For the present purpose it is enough to know that all three of the methods distinguished above have, as a matter of fact, been adopted over a fairly wide field.


§ 3. The first of the three necessarily, and, if the categories are so chosen that people cannot practically be drawn by the bounty into a benefited category, the other two also involve "neutral transferences" in the sense explained in § 3 of the last chapter, and not differential transferences. Hence the expectation of them operates on the productive activity of the poor only through their effect on the marginal desiredness which money has to them. But they differ from the kind of neutral transferences so far examined in one respect. They will check to a small extent the contribution of work made by the poor, if they are granted upon things for which the demand of the poor has an elasticity less than unity; but they will increase this contribution to a small extent, if they are granted on things for which this demand has an elasticity greater than unity. For in the former event the marginal desiredness of money to the poor will be lowered, since more is left over for other things; and in the latter event it will be raised. As a matter of fact, bounties are most likely to be given on things of urgent need and, therefore, of inelastic demand. The check to output resulting from the consequent relaxation of effort on the part of potential recipients means some, though probably a very small, diminution of the national dividend.


§ 4. So far it would seem that there is little to choose between help to the poor by bounties and by direct neutral transferences. If the amount of the bounty-fed commodity which each recipient is to consume is fixed authoritatively, as under the British system of free and compulsory elementary education, this is in fact so. It is so, too, if the amount is not fixed authoritatively, but is, for other reasons, not liable to change in consequence of the bounty. Thus poor people are accustomed to buy some things through a common purchase fund, so organised that the payment a member has to make does not vary with the amount of his individual purchases. Sick clubs are arranged on this plan. There will be no inducement to a member of a sick club to increase the amount of the doctor's services that he calls for in a year merely because the fixed amount, that he has been accustomed to pay for membership of the club, is taken over and paid by the State. These conditions, however, are exceptional. In general, when a bounty, or the equivalent of a bounty, is given on any commodity, the purchasers, having regard to the bounty, will buy more of the commodity than they would have done had they received an equivalent subsidy in the form of a direct money grant. In this way resources are diverted out of the natural channels of production, and there is a presumption—which may, of course, as was explained in Part II. Chapter XI., be rebutted by special knowledge—that this diversion will inflict an extra injury on the national dividend, over and above that set out in the preceding paragraph. If the bounty is large enough, it may happen that the output of the bounty-fed commodity will be expanded so far that to the poor themselves the supply price, not merely in terms of money, but in terms of satisfaction, exceeds the demand price, or, in other words, that the economic satisfaction they get from the last increment consumed is less than the economic dissatisfaction involved in producing it. In general the expectation of a transference to the poor through bounties on particular commodities is likely to damage the national dividend rather more than the expectation of a direct neutral transference of equal magnitude. In spite of this, however, the bounty method may still sometimes be better than the other, not only because there may be special economic or non-economic reasons for encouraging the consumption of the particular thing on which the bounty is given, as compared with other things, but also because the element of "charity" is less obvious and, therefore, less damaging to the morale of the beneficiaries, when it is concealed in a bounty than when it is displayed in a direct dole.

Notes for this chapter

Knoop, Principles of Municipal Trading, p. 266.
Ibid. p. 213.
Knoop, Principles of Municipal Trading, p. 266.
Cf. ante, Part IV. Chapter V. § 7.

Part IV, Chapter XII

End of Notes

68 of 73

Return to top