THE COURSE OF PAUPERISM BEFORE THE WAR

Before discussing the statistics of pauperism during the war, it is important to examine the pauper returns for the period preceding the war in order to determine whether a new tendency set in as a result of war conditions or whether a tendency already present was continued, retarded, or accelerated. Table I, which gives the ratio of paupers to population for five-year periods from 1885 to 1914, will illustrate the general tendency of the pre-war period.

This table shows that the ratio of indoor pauperism to population had remained almost stationary for the quarter of a century preceding the war and that only a very slight decrease in the outdoor-relief ratios had occurred until the period of the old-age pension acts, 1908-11. The marked decline in outdoor pauperism in 1910 and in the years following was due to the old-age pension legislation. The removal of the pauper disqualification and the extension of old-age pensions, after December 31, 1910, to old persons who had been in receipt of poor relief, explains the decline in the number of outdoor paupers from 513,242 in 1910 to 392,596 in 1911, a decrease of more than 120,000 in a single year. This de

1907* and the recently published results for 1924.5 Consequently, this information has been surprisingly deficient.

6

Since the publication of the 1924 census reports, there has been a new opportunity for studying the question, and one index of 1924 production has been made by the Balfour Committee on Industry and Trade. As will be shown, however, this estimate leaves much to be desired. Its result is far out of line with previous estimates, and its method is both highly questionable standing by itself and is unsuitable for securing the details by industry which are essential to a full knowledge of the situation.

8

It seems appropriate, therefore, to construct the best index possible for general purposes from this new information, to compare this with previous estimates, and to examine the chief findings laid open by such an index. Such is the purpose of this article. First, however, it is necessary to examine what is meant by an index of production and to set up the formula that will be used for measurement.

II. FORMULA

In many cases of indexing, it may be assumed that any approximation to the concept to be measured will give results quantitatively near the truth. Unfortunately, this is not true for the problem in hand. Information is available for but two years, and

* Great Britain, Parliament, "Final Report of the First Census of Production of the United Kingdom (1907)," Accounts and Papers (1911–13), Vol. CIX, Cmd. 6320.

'Published originally in preliminary reports, Nos. 1 to 30, issued as supplements to the Board of Trade Journal between February 24, 1927, and January 5, 1928, entitled, "Third Census of Production, 1924." We shall not be concerned with the similar reports on Northern Ireland which appeared subsequently. All these preliminary reports are obtainable as separate reprints from H. M. Stationery Office.

6 Great Britain, Committee on Industry, Trade, Further Factors in Industrial and Commercial Efficiency (1928), pp. 52–67.

'Although corroborated by this study (cf. Table II, p. 13). Cf. p. 10, infra. 'Our problem would not be quite so difficult if we had data on the intervening periods between the first and last years. As the number of divisions of our total period increased, those disturbing changes which are a function of time would be negligible, since the lapse of time during each period of comparison would approach zero. We might in such a case link these intermediate comparisons so as to cumulate the changes in production which we wished to measure, while not doing so for the disturbing factors.

these two single points of comparison are separated by seventeen years of rapid change which resulted in very divergent movements of the separate commodities to be indexed. Selection of the particular formula thus becomes a crucial matter.

It is commonly recognized that a valid index of production must be independent of changes in the general price level. Comparisons of aggregate sale value10 produced by industry have thus been generally rejected as measures of production because they include this disturbing price element. Realizing this, many have concluded that an index of production should measure "physical volume," and thus avoid the pitfalls of value relationships.

This counting of physical units is, of course, the method that would be used if only one type of goods were to be measured. So long as every unit to be counted has the same significance for any practical purpose, the physical measure is as good as any. But whenever a true index problem is present, various kinds of goods have to be aggregated together. It then becomes vital to select the particular aspect of the goods, common to all types, which one wishes to measure. An index is, indeed, a quantity statement about this common quality. The selection of the specific quality or common denominator of the group one wishes to measure, is, therefore, the controlling issue in determining the best method of weighting.1

11

The selection of this common quality among the various components of an index should be controlled by the purpose of the index. It would be possible, of course, to aggregate the number of pounds of all goods produced, and thus secure one sort of in

10 That is, a summation of the price of each good times the number of units of it produced. "Total expenditure," "total exchange value" (cf. Taussig, Principles, 2d ed., I, 128), and occasionally "gross (national) income" are terms used to mean the same thing (cf. note 17, infra). The British census calls this simply, "sale value," and we have adopted that term.

11

This approach contrasts with the attempt to discover an ideal index for all purposes, selected because it possesses a set of formal statistical properties; e.g., Irving Fisher, The Making of Index Numbers (Boston, 1922). The identity of our final formula with his is the more interesting because of this difference in initial viewpoints.

dex of the "physical volume" of production.12 But the result would have very little significance for most economic and social problems.

What aspect of the goods produced is of the most general social interest? Speaking in common-sense rather than in technical terms, the significant aspect of production as a whole is its want-satisfying power. No one can pretend to measure this accurately, but we believe that the best practical measure is the relative exchange value of the different goods as expressed in the market in terms of price. Relative changes in the marginal wantsatisfying power of goods, at least, are likely to be reflected best in market prices even when these measure incompletely the absolute amounts of satisfaction.13 Hence the common objective characteristic of goods we shall have to index in measuring general production is their value aspect. Relative values of the different goods, therefore, are just as important for our purpose as the physical volume of each type of goods itself.1

As an approximate measure of this common-sense concept, then, we shall define a change in national production as any change in the (market) value of work done to goods (in value

12 This is the sort of formula that a "physical" index would seem to imply. But the indices of "physical volume" just referred to in the foregoing are not physical in this sense at all (cf. p. 6, infra).

13 (1) Because of individual differences in wants and sensitiveness to the means of satisfying them, and (2) because the exchange mechanism is influenced by the inequality of incomes which makes the dollars of the wealthy stand for less want-satisfaction than those of the poor.

"Total production obviously increases caeteris paribus with any increase in the physical volume of any type of goods. But an increase in production may be caused just as surely, without any change in the total physical volume at all—by increases in those goods which man values highly, offset by decreases in the physical output of other goods. For example, a 50 per cent decrease in the number of wagons turned out along with a 50 per cent increase in motor-car production would result in the same physical total of vehicles. But there would be a net increase in production—i.e., want-satisfying power—as between the two years because of the higher value per unit placed on the motor cars.

This is not to deny that other definitions of production may be more useful for specific purposes. Thus in the work which one of the authors has done in trying to arrive at a mathematical theory of physical production, it has been thought advisable to exclude from the index of physical product the influence of the changing relative value of the respective commodities.

distress was due to the fact that the organization for the payment of allowances was not everywhere complete when mobilization was ordered, and also to the fact that in some cases there had actually been no time for the men to make allotments. There was, of course, great dissatisfaction over the situation, and efforts were made immediately to provide suitable assistance for such cases. A further difficulty was the low scale of the allowances at the time of mobilization. By October 1, 1914, a new and increased scale of allowances was brought into operation, and further increases were made later.

In order that no man who had been called to the service of his country might be dishonored by having the members of his family driven to the pauper rolls for relief as a result of his war service, a grant was promptly made the National Relief Fund not only to provide for such cases in the future but also to repay to the Poor Law Guardians the sums they had given as relief, in order that the names of such persons might be struck off the pauper lists.

Not only wives and children but thousands of mothers, sisters, and other relatives who had been dependent upon the men who joined the forces were left in a state of destitution immediately after mobilization. Arrangements were shortly made, however, to grant allowances to these relatives in all cases in which there was satisfactory evidence of the fact of dependence. Allowances were also to be made to unmarried mothers and their children in all cases "where there was evidence that a real home had been maintained."

The failure of the government to provide for the dependents of men who had joined the forces was said to be due, first, to the failure of the Army Pay Department to meet promptly all the claims upon it; and, second, to the government's delay in accepting liability for claims which were subsequently admitted to be a proper charge against the public funds.

That the repayment to the poor-law officials of the sums given as emergency relief to the families of men in the forces was a popular step, and that public indignation would have been aroused if this had not been done, is indicated by a statement in