17,000 a year. The total number of laborers who were unemployed at any time during the census year 1900 (the latest for which data are available) was 40,1081 and the rate of unemployment in 1904-1906, as far as can be judged, was about the same as in 1900.2 It is safe to say that the labor agencies were able to place outside of New York City less than one half of all unemployed laborers. The majority stayed in New York waiting for work "to pick up."

We come next to the cyclical fluctuations of business which result in variations of the number employed from year to year. A temporary decline in business means unemployment for a number of wage-earners who were employed the year before. Such fluctuations of business occur in countries with a net emigration, like Great Britain, or in France whose population is practically stationary, as well as in the United States. The effect of these fluctuations of business activity "is the requirement in each trade of reserves of labor to meet the fluctuations of work incidental to years of prosperity."3 An illustration of the range of these cyclical fluctuations is given in Diagram VI.4

The most important cause of unemployment in point of numbers affected, however, is to be found in the fact that "the actual demand is that of each of many separate employers in many different places." The effect of this "dissipation of the demand for labor in each trade" upon

migrant Laborers in United States," Bulletin of the Bureau of Labor, No. 72, p. 417.

I

Occupations at the XII. Census, Table XLIII., p. 634.

According to the statistics of the New York Bureau of Labor, the ratio of unemployed in 1900 was 21.0 per cent; in 1904, 30.5 per cent; in 1905, 14.6 per cent and in 1906, 8.6 per cent, averaging for 19041906, 17.9 per cent. (Report of the New York Bureau of Labor Statistics, 1908, p. xviii., Table 10.) While these figures relate to members of labor unions only, they may serve as a standard of comparison of general business conditions in various years. 3 Beveridge, loc. cit., p. 13.

4 Based on Report of the New York Bureau of Labor Statistics, 1910, s Beveridge, loc. cit., • Ibid., p. 13 p. 10.

p. viii.

00

1902

1906

1903

1907

-1904

1908

-1905

1909

the mass of wage-earners can be measured by the difference between the aggregate number of hands required by all establishments at their busiest seasons and the aggregate minimum number employed when business is at its lowest ebb. To be sure, this difference does not represent the amount of unemployment at any given date, since the seasonal variations are not simultaneous in all industries and a considerable portion of the unskilled laborers readily shift from one industry to another. But every change of position involves some period of unemployment, as it requires some time to find a new job. This is clearly demonstrated by the census statistics of manufactures. The difference between the greatest and the least average number employed during any month in the year is far short of the difference between the aggregate greatest and the aggregate least number employed in all establishments during the same year, as shown in Table 16 next below:

TABLE 16.

RANGE OF FLUCTUATIONS OF EMPLOYMENT, 1899 AND 1904'

'Compiled from XII. Census Report on Manufactures, Part I, p. 59, Table 3; Census Report on Manufactures, 1905, Part I, p. lxxix., Table

In order to ascertain the total number of wage-earners who had steady employment during the year, the number employed by each individual manufacturer on the slackest day in his own business must be added to similar numbers for all others, though the days may not have coincided. It appears that this number was less than the lowest monthly average, which means that even in the worst month of the year more people are wanted on an average than can be given permanent employment. On the other hand, while during the best month of the year 1899 only fourteen additional wage-earners were needed over and above every one hundred who had permanent employment, actually four times as many persons were hired for temporary jobs during the same year. In 1904, the proportion was still more striking, the number of persons hired temporarily was nearly six times greater than the actual temporary force needed at the busiest time during the year. In either year there were over one half as many temporary jobs as permanent positions. Of course, these figures comprise many duplications; there may have been few wage-earners over the greatest number that could actually have been employed at the same time; in that case each of the temporary employees had about four different jobs during the year 1899 and six during the year 1904. The unemployment intervening between one temporary job and another clearly did not depend upon the number of applicants for jobs, but was determined by the vicissitudes of business.

The whole problem of unemployment is admirably elucidated in Mr. Beveridge's exhaustive treatise on the subject, from which the following is condensed:

A general and normal excess of the supply of labor over the demand appears to be explicable only by an excessively rapid increase of population. But such an explanation does not square with the facts showing irreducible minimum of unemployment precisely in those industries which have grown with exceptional rapidity in recent years.

The general formula for the supply of labor in an industry appears to be this: for work requiring, if concentrated at one spot, at most ninety

eight men, there will actually be eighty in regular employment; there will be a hundred in all, so that at all times two at least are out of work. The twenty, however, are as much part of the industrial system as are the eighty; the reserve is as indispensable as the regulars. The idleness, now of some, now of others, of the reserve is mainly responsible for the irreducible minimum of unemployment. The figures here given have only an illustrative value; the proportion of regular and reserve and irreducible minimum vary from trade to trade. The principle is of the greatest generality. The rule for each trade is to have more men than are called for together even at the busiest moment.

The normal state of every industry is to be overcrowded with labor, in the sense of having drawn into it more men than can ever find employment in it at any one time. This is the direct consequence of the work of each industry being distributed between many separate employees each subject to fluctuations of fortune. It depends upon the nature of the demand for labor, not upon the volume of the whole supply. To speak of the reserve of labor in a trade may become, in fact, only another way of speaking of the whole volume of unemployment in it. The change, however, is not one of words alone. It implies a revolution of mental attitude. It involves perception of unemployment, not as a thing standing by itself—an inexplicable excrescence on the industrial system-but as a thing directly related to that system and as necessary to it as are capital and labor themselves.

Unemployment is not to be identified as a problem of general over-population. Unemployment arises because, while the supply of labor grows steadily, the demand for labor, in growing, varies incessantly in volume, distribution, and character. This variation in several of its forms at least flows directly from the control of production by many competing employers. It is obvious that, so long as the industrial world is split up into separate groups of producers each group with a life of its own, and growing or decaying in ceaseless attrition upon its neighbors-there must be insecurity of employment. Unemployment, in other words, is to some extent at least part of the price of industrial competitionpart of the waste without which there could be no competition at all.1

It has been shown that unemployment does not depend upon the volume of the supply of labor, but is determined by the nature of the demand for labor, which produces a "relative surplus-population. "a Is it not possible, however

'Beveridge, loc. cit., pp. 70, 76, 99, 100, 103, 235. Karl Marx: Capital, Vol. I., ch. xxv., sec. 3.