ECCENTRIC OFFICIAL STATISTICS 527
employment little if any more than half the year. Because of this fact their wages are higher than wages in factories where employment is constant. The effect of including these trades more largely in the census of 1890 than previously, and divid- ing by the average number of employes instead of the whole number, cannot but largely increase the apparent average annual earnings at this census. To illustrate : an establishment in one of the building trades may employ 100 men 100 days at $3 per day; 60 of these men another 100 days at $2.50 per day, and 10 of them another 100 days at $2 per day. This would make the total wages for the 300 days $47,000 and the average num- ber of employes 56%. Dividing the total wages by the average number employed we have apparent average annual earnings of $829. As, however, there are 100 employes, the correct average annual earnings 'would be $470, which would make a very appreciable difference to the wage-earner if not to the statis- tician. 1
Mr. Steuart falls into the same error in his tabulation and remarks accompanying the tables of the manufacturing reports of the eleventh census. Presenting the annexed table Mr. Steuart remarks:
"The average annual earnings of employe's as obtained from the total for 1880 was $346.91, while the average obtained for all employe's for 1890 was $484.49, and for the employe's exclusive of officers, firm members, and clerks, $444.83. Owing to the difference in the form and scope of the inquiry of 1 890, as compared with that of 1880, previously referred to, neither of these average annual earnings for 1890 should be accepted as the exact increase during the decade. While the number of employes reported for 1890 is, as nearly as possible, the true average number employed during the year, the result obtained by using this number as a divisor into the total wages must not be accepted as the actual annual wages per employe* for any particular class or occupation. The total number of employe's and the total wages include not only officers, firm members, and clerks, and their salaries, but the number and wages of females, apprentices, children, and adults, employe's working
'The writer has not undertaken in this illustration to show the fact as to the average employment in the building trades, but as one engaged in one of these trades he is confident from his own observation that the illustration is not far from the truth.
