quotients derived from these data due consideration should be given to the effect of possible errors in the divisors.
The quotients computed show that the share that railway labor took out of the national income rose steadily from 1916 to 1920. In 1921 there was a small recession; and in 1922 a greater one. The last is ascribable to improved efficiency, the railways having managed to get along with fewer men, while the average compensation per employee remained substantially unchanged.
I do not think that the above table which begins with 1916 for the reason previously stated, tells the whole story. Making comparison between the three calendar years 1912-14, in which the national income averaged about 33.5 billion dollars, and the three fiscal years ending with June 30, 1914, during which the total compensation to railway employees averaged $1,295,224,000 per annum, it appears that the railway employees received 3.87 per cent of the national income. If we consider this a fair exhibition of prewar, normal conditions, railway labor was not treated as well in 1916-17 as might have been expected. In all subsequent years, however, the share of railway labor on the whole has been supernormal.
I say “on the whole,” for if we analyse railway labor into train service and all other kinds of service, especially maintenance, we find different results. Broadly speaking, although the men who function in train service are members of the brotherhoods, which is one of the strongest groups of labor unions, it is found that their compensation lagged behind the general average. This is shown in the accompanying table.
