CHAPTER X

CHANGES IN THE RELATIVE IMPORTANCE OF DIFFERENT

INDUSTRIES.

§1. Light may be thrown from another angle upon the problems which we have been considering in the last five chapters by a study of the relative importance of different industries or, more broadly, of different employments, in any community, and of their changes from time to time.

Following the principle adopted in the British Census of Production of 1907, the relative importance of different industries, or employments, may be measured, for our present purpose, by the relative money value of their annual net outputs. This principle is capable of extension, beyond the range of industries covered by the Census of Production, to the net output of personal services, as distinct from material commodities. In the case of the medical, legal, and other private professions, which offer services for sale, the valuation of net output presents no great difficulty beyond the collection of the relevant statistics. In the case of public servants, employees of the State or local authorities, who render services which are paid for out of public revenue and not specifically, it is a matter of greater difficulty to decide what should be included in the value of net output in addition to the salaries of these employees. But difficulties of this type can be solved by arbitrary decisions.

The absolute share of any factor of production in the total income of the community is the sum of the absolute shares of that factor in all the employments which contribute to the total income of the community. Similarly the relative share of any factor is the weighted average of

Q

the relative shares of that factor in all these employments. The weighting of the average is according to the relative importance of the different employments. As between different employments in any community, the relative shares of factors show great variation. Thus in this country before the war the annual net output of certain industries was divided in the following percentages.1

1 Bowley, Division of the Product of Industry, PP. 42-5, following in

the main the Census of Production.

17 7

67

56

78

74

62

75

II

19

81

27

66.5

18

30

32

39

70.5

48

40

Two of the most important causes of these variations in percentages are, according to Professor Bowley, the amount of capital employed per worker, and the sex of the workers employed. Thus the amount of capital per worker is exceptionally large on the railways and in the tobacco trade,' while the proportion of male to female workers is high in coal mining, the metal trades, building and gas works, and low in the textiles and in tobacco. It will also be noticed that, as a general rule, the higher the percentage of wages to net output, the higher are the

1 On the other hand Professor H. L. Moore (Laws of Wages, pp. 55-70) argues that the relative share of labour increases with the amount of capital employed per worker, or in other words that the relative share of labour is greatest, and of capital smallest, in the most highly capitalised industries. Professor Moore seems to think that this result is involved in the specific productivity theory of wages, of which he is an adherent. His argument consists in applying the theory of correlation to certain statistics of the French coal-mining industry between the years 1847 and 1899. These statistics are (1) the ratio each year of daily wages to the value of a day's product per worker, and (2) the amount of machine power employed per 100 workers. He obtains from these figures a positive correlation of .599. This is not a high degree of correlation and the argument is unconvincing in other respects. For it only takes account of a single industry in a single country, and it allows nothing for changes in the efficiency of labour or of improvements in machinery over a period of 52 years. It illustrates the theoretical difficulty of measuring both quantities of labour power and quantities of capital.

Marshall (Industry and Trade, pp. 831-2) quotes the following statistics from American censuses regarding the production of factories in the United States :

The proportion of wages to net output thus fell from 41.6% in 1900 to 41.5 in 1905, and to 40.2 in 1910. Between 1900 and 1910 the number of wage-earners increased from 4,712,000 to 6,615,000, i.e., by 40%, the money value of the capital employed from 8,975 to 18,428 million dollars, i.e., by 105 %, and the horse power used from 10 to nearly 19 million units, i.e. by nearly 90%. This is the type of statistical information, which, if it was forthcoming on a large scale over considerable periods of time, would be of great interest and value.

average earnings per worker. But to this rule there are one or two striking exceptions.

As regards some of the industries and employments not included in the above table, it is certain that the percentages of wages and salaries are exceptionally high in the professions and in most branches of the public service, are exceptionally low in agriculture and are practically non-existent as regards income from foreign investments.

As regards the broad division between workers and property owners in any particular industry, the percentage of wages plus salaries gives a minimum for the relative share of the workers, but something must be added to this, in so far as profits represent work done by employers, as distinct from income derived from their own property.

As regards the relative share of land, as distinct from other property, no information is forthcoming from the above statistics. In agriculture and other extractive industries, however, it is probable that economic rent generally forms a larger proportion of cost of production than in manufacture. It is also obvious and important that in the shipping industry the proportion is exceedingly small, the only payments of economic rent which are involved here being in respect of the sites of offices, warehouses, etc., ashore. For, though human ingenuity, prompted by the desire for gain and aided by indulgent laws, is ever seeking new extensions of private property, the sea has hitherto resisted attempts at appropriation, except to a limited extent in the close vicinity of land. Physical difficulties have thus hitherto prevented the rise of a sea-owning class. Nor does it seem likely that, as

' Compare Marshall, Principles, p. 473 n. The statement in the text has no relevance to the rather tedious and ill-worded controversy, as to whether rent enters into" cost of production, in the sense of helping to determine it.

"

the industry of transport by air develops, a class of private air-owners will be permitted to take toll of it. The legal rights of the landowner, indeed, according to the old textbooks, extend "usque ad coelum et ad inferos "-up to heaven and down to hell-but, whatever may happen to the others, it seems clear that the landowners' celestial rights, which would in any case be mostly unenforceable, will not be very substantial in the future.

§2. The question suggests itself, whether distribution might not be advantageously modified, in the direction of reducing the inequality of incomes, by the deliberate encouragement of certain industries and employments at the expense of others. Thus the relative share of land might be reduced by the discouragement, by taxation or otherwise, of those employments in which rent forms a high proportion of cost of production, and the encouragement, by bounty or otherwise, of those in which rent forms a low proportion. This policy would be illustrated by a bounty to shipping out of the proceeds of a tax on agriculture, the converse of what is, in effect, proposed by most Protectionists. Similarly, in order to encourage employments, in which wages form a high proportion of cost of production, and to discourage those, in which wages form a low proportion, a bounty might be given to the production of coal or of buildings out of the proceeds of a tax on brewing or the manufacture of tobacco.

The effects of such policies upon the inequality of incomes depend upon the relation between categories and persons, which requires further discussion, and also upon the reactions on the total volume of production, on relative prices, on interdependent industries and on

1 Deliberate manipulation of the relative importance of different industries has been proposed, from a different point of view, by Protectionist writers. See e.g., Ashley, Tariff Problem, pp. 105-113, 133-4, etc.