The Economics of Welfare

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

Part IV, Chapter VIII


§ 1. WE now turn to what is in practice by far the most important field of possible disharmony. In a great number of ways, and for a great variety of reasons, poor people in civilised countries are given help, in the main through some State agency, at the expense of their better-to-do fellow-citizens. In Great Britain in 1925 the contribution from central government funds towards social services, chiefly for old-age pensions, education, unemployment insurance, health insurance and housing, amounted to 113 millions, and the contributions from the funds of local authorities, chiefly for education and Poor Law relief, to 79 millions.*58 Evidently a large part of these 192 millions represent what is, in effect, a transfer of income from relatively rich for the benefit of relatively poor persons. Prima facie the transference, which this help implies, must increase, and it can certainly be so arranged that it shall increase, the real income available for the poor. Hence the question whether any particular form of help to the poor involves disharmony is often equivalent to the question whether its indirect consequence is to increase or to diminish the national dividend. This question, in one or another of its various aspects, will be the theme of the next four chapters. But, before we embark upon it, a brief comment is needed upon two popular arguments, one of which asserts that no transference of resources to the poor is possible, because, in effect, all money taken from the rich is really taken from the poor, while the other asserts that it is not possible, because the beneficiaries will give back what they have received by agreeing to accept lower wages.


§ 2. The position taken up in the former of these arguments is that any levy of money, whether voluntary or compulsory, from the well-to-do for the benefit of some poor persons necessarily implies the infliction of a substantially equivalent burden upon other poor persons, through the reduction which the rich are compelled to make in their purchases of the services rendered by them. The foundation of this view may be set out as follows. It is obvious that a great part of the expenditure of the rich involves, directly or indirectly, the employment of labour; and it is equally obvious that, if the incomes of the rich are diminished by, say, £20,000,000 of taxation, their expenditure for consumption and for capital investment together must be contracted to a corresponding extent. Some persons, concentrating attention upon these facts, immediately conclude that the workpeople, whose services this expenditure would have called into being if the tax had not been there—and an exactly analogous argument applies to a voluntary contribution—must suffer a loss of income approximating to the twenty million pounds levied in taxation. To argue thus, however, is to ignore the fact that the twenty million pounds collected from the rich is, ex hypothesi, transferred to the poor, and that the expenditure of it by them is likely to be no less productive of employment than the expenditure of it by the rich would have been. No doubt, if we are contemplating the immediate effect of the addition of twenty millions to the taxation of the rich for the benefit of the poor, it is relevant to observe that the men who lose jobs on the one side will not be the same persons as those who find them on the other; and that, therefore, a certain number of men, who have been trained to special aptitudes, may find their immaterial capital of acquired skill rendered permanently worthless. This loss, however, is the result, not of taxation, but of change in taxation, and would emerge equally in consequence of a reduction by twenty millions in imposts levied on the rich for the benefit of the poor. Our problem is not concerned with incidents of this character. The comparison we have to make is between one permanent system, under which nothing is collected from the rich and handed over to the poor, and another permanent system, under which twenty millions is so collected and handed over. To this comparison the incident we have just been discussing is irrelevant. Speaking broadly and apart from special circumstances, we may say that it makes very little difference to the employment of, and wages paid for, labour, whether twenty millions is annually transferred or not transferred from any one class to any other class. The idea that reactions in this field will render attempts at transference of no effect is, therefore, illusory.


§ 3. The latter of the two arguments distinguished above asserts that, if any group of poor persons are accorded any form of subsidy, they will, in consequence, be willing to work for less than the worth of their services to their employer, and so will, in effect, transfer back the subsidy they have received to members of the richer classes. This view rests partly upon a priori reasoning and partly upon what is called experience. It needs, therefore, a twofold discussion. The a priori reasoning starts from the fact that a Poor Law subsidy enables a person to accept lower wages than it would be possible for him to accept otherwise without starvation, or, at all events, serious discomfort; and it proceeds to assert that, if a person is enabled to work for less, he will be willing to work for less. Now, no doubt, in certain special circumstances, when a workman, in receipt of a subsidy insufficient to enable him to live up to his accustomed standard of life, is confronted by an employer occupying towards him the position of a monopolist, this inference may be valid. In general, however, where competition exists among employers, it is quite invalid. A person who, by saving in the past, has become possessed of a competence, is enabled to work for less than one who has not. A millionaire is enabled to work for less even than a relieved pauper. So far from this ability making it probable that he will strike a worse bargain in the higgling of the market, it is likely, in general, to have the opposite effect. It is not the fact that the wife of a man in good work is likely to accept abnormally low wages. On the contrary, the woman who, for this or any other reason, can afford to "stand out," is, in general, among those who resist such wages most strenuously.*59 Let us turn, then, to the reasoning from what is called experience. This starts from two admitted facts. The first fact is that old and infirm persons in receipt of a Poor Law subsidy very frequently earn from private employers considerably less than the ordinary wage per hour current for the class of work on which they are engaged. The second fact—given in evidence before the Poor Law Commission of 1832—is that the refusal of guardians to grant relief in aid of wages "soon had the effect of making the farmer pay his labourers fairly." From these facts the inference is drawn that, where a Poor Law subsidy exists, workpeople accept a wage lower than the worth of their work to their employers. This inference, however, is illegitimate. There is an alternative and more probable explanation. As regards old and infirm persons, may it not be that the low wage per hour is due to the circumstance that the work they can do in an hour is poor in quality or little in quantity? As regards the old Poor Law, may it not be that the unreformed system of relief, so long as it prevailed, caused people to work slackly and badly, that, when it was abolished, they worked harder, and that this was the cause of the alteration in their wages? The view that the true analysis of experience is to be found along these lines, and not in the suggestion that relieved persons work for less than they are worth to their employers, is made likely by general considerations. It has been further confirmed by recent investigations, which tend to show that, where two people differ solely in the fact that one does, and the other does not, receive a Poor Law subsidy, their wages are in fact the same. Thus investigators appointed by the Poor Law Commission of 1909, as a result of their inquiry into the effects of out-relief on wages, write: "We found no evidence that women wage-earners, to whose families out-relief is given, cut rates. Such wage-earners are invariably found working at the same rates of pay as the much larger number of women not in receipt of relief, who entirely swamp them.... We could find no evidence that the daughters of paupers accepted lower rates than others, or earned less than others, because of their indirect relation to pauperism."*60 This argument, therefore, like that set out in § 2, breaks down. The direct transference of resources from the relatively rich to the relatively poor, by way of philanthropic or State action, whatever its ultimate consequences may prove to be, is at least not impossible. Of course, this conclusion does not deny that additional work by assisted, or any other, workers slightly lowers the general rate of wages.*61 But that proposition is quite different from the proposition that assistance to persons who are working anyhow has this effect.


§ 4. In view of this result we may proceed undisturbed to our main problem—that of determining the effect of various sorts of transference upon the size of the national dividend. Some sorts, it would seem, are likely to increase that dividend and others to diminish it. We have, therefore, to investigate the conditions upon which the occurrence of the one or the other of these opposing consequences depends. These conditions can be examined most effectively by means of an analysis in which the distinction between the effect of the fact, and the effect of the expectation, of transference is made fundamental. Of course, when we have to do with a levy, which is made once and for all to meet some exceptional need, and the regular continuance of which is not anticipated, effects operating through expectation do not have to be considered. In ordinary times, however, the fact of a tax levy imposed one year carries with it the expectation that it will be continued in future years, so that both fact and expectation are relevant. I shall consider first the expectation of transferences from the rich; secondly, the expectation of transferences to the poor; and thirdly, the fact of transferences.

Notes for this chapter

Cf. Carr-Saunders and Jones, Social Structure in England and Wales, p. 158.
For an illustration of this among home-working tailoresses cf. Vesselitsky, The Home Worker, p. 17.
Report of the Royal Commission on the Poor Laws, Appendix, vol. xxxvi. pp. vi-vii.
Cf. ante, Chapter III. § 10. Dr. Hourwick, in his book Immigration and Labour, seems to miss this point; for, having shown that immigrants into the United States do not earn less wages for equally efficient work than native Americans, he treats this conclusion as implying that they do not affect the wages of native Americans.

Part IV, Chapter IX

End of Notes

65 of 73

Return to top