Page:American Journal of Sociology Volume 3.djvu/374

 360 THE AMERICAN JOURNAL OF SOCIOLOGY

can be used in this examination and statements that cannot be controverted." Colonel Wright then proceeds to quote statistics the incomparability of which has already been so conclusively demonstrated in this JOURNAL that he has attempted no reply. Let us call Colonel Wright himself to testify as to the alleged facts on which he bases his conclusion of a large increase in the earnings of wage workers and a decrease in the number of children employed in manufacturing industry. In an official communication to the Chairman of the Committee on Census United States Senate, dated Feb. 15, 1895, Colonel Wright, as superintendent of the Eleventh Census, said: "The tendency of the questions used in 1880 was to obtain a number of employes in excess of the average number, while it is believed the questions used in 1890 obtained the average number. The questions used in 1890 also tended to increase the amount of wages as compared with 1880. The enumeration of establishments in certain lines of industry was more thorough at the Eleventh than at the Tenth Census. For these reasons the average annual earnings per employe, as obtained from the totals for the two censuses, are not compar- able. Mr. Waite states: 'This great increase is due chiefly to the fact that the census of manufactures for 1880 was worked up upon an entirely different basis from that of 1 890. In the former census the officers and firm members were reckoned among the number of hands employed, but were not accredited any wages except in exceedingly few cases. In 1890 the hundreds of thousands of officers, firm members, and salesmen were each accred- ited with large salaries, aggregating upwards of $300,000,000. Some salaries were equal to that paid the President of the United States. On the other hand, in the census of 1880 the figures purporting to represent the "average number of hands" were for about half the establishments identical with the "greatest number of hands employed at any one time during the year." In the other half they represented for each establishment the average number employed during the few months when the establishment was running a full force. As a result they were almost always