OFFICIAL STATISTICS 627
with their salaries has been excluded and the average given as $444.83. Then the only possible factors that entered into this average wage of $445 for 1890 that would possibly tend to abnormally increase it, as compared with the average of $347 for 1880, and which could not be eliminated except by the application of an arbitrary estimate, are the more thorough enumeration and the inclusion of additional industries previously referred to and explained ; also a question which required the classes of employe's and their wages to be reported separately ; and the greater care, if any, taken in the editing or preparation of the schedules for tabulation.
If we exclude from the totals for 1890 not only the officers, firm members, and clerks, but all the employes and the wages paid in the industries that were possibly omitted or not thoroughly canvassed in 1880, we still have an average annual wage of $429.47 for 1890. This average is evidently still too high for those who have criticised it, but I do not believe the change in the form of the question has had any material effect on it, for the obvious reason that the total amount paid in wages is the item of all others that the manufacturers were able to report with exactness : it was the item most readily ascertained at both censuses. This being the case, the same total would invariably be given, no matter whether a lump sum was required, as in 1880, or an itemized statement, as in 1890. Then, having eliminated the officers, firm members, and clerks as a class that was pos- sibly more fully reported in 1890, we have left as an abnormal factor, so far as the questions are concerned, that part which related only to the number of employe's. In 1880 the greatest, the least, and the average number employed during the year, or the time in operation, were required to be reported. In 1890 the average number only was required to be reported by the classes previously enumerated.
There is a possibility, which is explained fully in the final reports, that the question of 1890 resulted in securing a smaller number of employe's than did that of 1880, but if the difference was great enough to make any perceptible difference in the
