370 THE AMERICAN JOURNAL OF SOCIOLOGY
ages would run about like this: For those 10 years old, 3 per cent.; n years, 5 per cent ; 12 years, 8 per cent : 13 years, 1 1.5 per cent ; and for 14 years, 15.5 per cent. Such an estimate of the percentage at work at each age from 10 to 14 seems to be about what is necessary to bring the general aver- age of 8.57 per cent, at work when all from 10 to 14 years are lumped together. From this it is manifest that, of those 1 5 years old, about 20 per cent., or 257,773, must have been workers. Adding this to 603,013, the num- ber of workers from 10 to 14 years, the result is 860,786. This number, or 10.34 per cent, of the whole number of children 10 to 15 years of age in 1890, represents very closely the number at work.
As it appears, there is a difference in the classification of the Eleventh from preceding censuses of one year. The purpose of this change in classification we can only conjecture. The result, as will appear, is to mislead the public.
Colonel Wright has the duty of explaining to the public the reason for changing the age classification in the tables of occu- pation, rendering them worthless for the only purpose for which such statistics are compiled, that of comparison.
Why are the figures of child occupation fifteen years and under suppressed ? If the proper tabulation had not been made, it was in Colonel Wright's power to have it made. Why does he not give the public the facts instead of his guesses ? Colonel Wright estimates the difference in the number of workers resulting from the change of classification to be 257,773, and says : " It is plain that the individual percentage for each of these years would run about like this." Surely this is not at all plain, and seems a strange manner of reasoning for one who so uniformly adopts the " true scientific method." The data furnished by the Massachusetts factory inspectors' reports furnish the basis from which to intelligently estimate the relative proportion of child workers at the different ages.
There is, however, an important element in this problem which Colonel Wright entirely ignores : that is, the fact that the classification is made from the returns made by the enumerators of population, and that the schedule of inquiry at the census of 1880, and 1870 as well, called for the age of last birthday, while the schedule of the last census called for age at nearest birthday.
