even less than in the record year 1902. The mean percentage of the unemployed in the year 1905 was 11.2 as compared with 16.9 in 1904, 17.5 in 1903 and 14.8 in 1902, or an average of 16.4 per cent for the three years. The accompanying chart exhibits the CENT JAN FEB MAR APR MAY JUNE JULY AUG SEP OCT NOV DEC 1016 23 1992/41 relative number of the unemployed at the end of each month in 1905, as compared with the mean for the three years 1902 4, and makes clear the wide difference between condi. tions of employment in 1905 and those of the preceding years. With only a few exceptions, every industry made a better showing in 1905 than in 1902, 1903 or 1904; in five groups of trades (metal trades, musical trades, tobacco trades, restaurants and retail trade, and public employment), the ratio of the unemployed was slightly higher than in 1902, when all trades were exceptionally busy (see table opposite). While the unwonted activity in the building trades and the generally peaceful relations between employers and employees therein, were influential factors in keeping down the percentage of idleness in 1905, they were not of themselves sufficient to account for the difference. In the combined trades, exclusive of the building industry, the ratio of the unemployed was smaller than usual in nearly every month; only in February, August and September did the 1905 percentage rise above that of 1902, the most favorable of the years immediately preceding. TABLE 2.-UNEMPLOYMENT IN THE COMBINED TRADES EXCLUSIVE OF THE BUILDING IN Number reporting. MONTH. January. April. May June. July. August... September.. October.. November.. December..... 1905. 1902. 1903. 1904. 1905. 1902. 1903. 1904. 1905. TABLE 3-PERCENTAGE OF UNEMPLOYED MEMBERS OF REPRESENTATIVE TRADE UNIONS: BY INDUSTRIES-Concluded. Feb. Mar. Apr. May 1902. 6731 6223 8848 343 1261 8523 6595 1227 131 1245 4572 7851 +756 1903. 331 22.4 1903. 13.5 4.2 †Included in Group IV. That one of the principal factors in the favorable labor market of 1905 was the lessened importance of labor disputes is demonstrated by the statistics concerning the causes of idleness, which may be summarized as follows: NUMBER OF WAGE EARNERS IDLE IN CONSEQUENCE OF STRIKE OR LOCKOUT. ..... .... YEAR. Jan. Feb. Mar. Apr. May. June. July. Aug. Sept. Oct. Nov. Dec. 1902. 131 6,690 246 497 1903. 1,086 2,083 5,274 9,668 15,983 9,991 7,755 3,803 1,821 1,479 1,645 1904. 2,541 1,559 6,820 2,963 3,712 1,647 4,982 4,824 4,829 3,194 2,660 2,760 1905...... 3,036 2,775 3,288 2,197 1,266 1,197 561 675 487 609 756 708 ..... = In 1902 the causes of idleness were tabulated for only four months. These are nevertheless sufficient to show that strikes and lockouts were on a much smaller scale than in 1903 and 1904. The earlier months of 1905 indicated a continuation of the labor troubles, but they gradually quieted down and in the latter half of the year compared not unfavorably with 1902. TABLE 4.-MEMBERS OF REPRESENTATIVE TRADE UNIONS IDLE ON ACCOUNT OF LABOR DISPUTES. Building Trans- Clothing, machine AT END OF January. February. 198. 202. Food All Wood and Other trades. trades. portation etc. trades. Printing working. liquors. trades. 3,036 2,200 2,775 1,936 Metals, 612 26 612 25 UNEMPLOYMENT IN PREVIOUS YEARS. The system of monthly reports on the state of employment was instituted in 1902 but data for comparison with earlier years can be drawn from the quarterly returns of all trade and labor unions, beginning in 1897. For three years, 1897-99, quarterly reports were collected by the Bureau four times a year, thus furnishing a complete record for each year. Since 1899 the Bureau has been unable to collect such reports for every quarter and has therefore confined its record to the first and third quarter of each year. From the monthly and quarterly reports the follow ing figures have been compiled concerning the proportion of the unemployed at the four dates in each year: TABLE 5.—PERCENTAGE OF TRADE UNIONISTS IDLE AT THE END of B.-Representative unions, containing about one-fourth of all unionists. While these figures give some indication of the trend of business conditions, they do not furnish an adequate record. The series is complete for only two months, March and September, and the mean percentages for those months would indicate that there was more idleness in 1905 than in 1902, and considerably more in 1904 than in 1903, whereas complete figures for 12 months in each have already proved the contrary. More satisfaetory statistics, however, may be drawn from the reports regarding the duration of employment, which are concisely summarized in the table opposite. In the first three months of 1905 it appears that 31,638, or 8.7 per cent of the 364,544 wage earners reporting, did not work at all, whereas in the third quarter the number idle fell to 7,491 or only 2.0 per cent. Of the number employed (column 5), nearly all reported the duration of their employment (column 6), which the first quarter averaged 65.67 days and in the third quarter 73.55 days. If, however, the aggregate number of days of employment (column 8) be distributed over all wage earners including those who were idle as well as those who had employment (column 9), the average duration of employment in the first quarter was only 59.9 days and 72 days in the third quarter-the total number of working days in a quarter being 77. Combining the averages of the last column for 1897 to 1899, it is found that the duration of employment in 1897 averaged 227 days, in 1898 232 days and in 1899 258 days,-out of 308 working days in the year. As figures exist only for two quarters of each year thereafter, estimates have been made for |