that these will become less and less a striking machine, and that more and more the policy of arbitration will be substituted for that of force. The right of organization on either side must be recognized, and the two organizations must learn to cease trying to hurt each other, and try to be mutually helpful. It is not an easy task. task. Many difficulties are present in the lack of intelligence and of sympathy, in the mutual distrust, and in the want of responsibility. Probably the knowledge of how they can help each other will have to be obtained through experience in hurting each other. But eventually we shall without doubt secure more stable industrial conditions through these very struggles, and with a growing recognition of the necessity of substituting arbitration for strikes and lockouts, attain to "a more perfect state." ERNEST L. Bogart.

Oberlin College.

THE MEASUREMENT OF UNEMPLOYMENT: A

STATISTICAL STUDY.

II.

A few words will be required concerning the investigation of unemployment made in connection with the Rhode Island State Census of 1895. No detailed inquiry was undertaken, the only information published being that contained in the Tenth Annual Report of the State Bureau of Labor, showing, by towns and groups of occupations, the number and per cent. of industrial workers according to the four classes; employed at their principal occupation; employed at some other occupation; unemployed; and conditions of employment unknown. This information is given for each month of the census year. There is a recapitulation by trades, but none by occupations, a remarkable omission in view of the nature of the subject considered. The general results of this investigation are shown in the following table, which gives the per cent. of all persons engaged in industrial pursuits who were found in each class.

It should be noted that the method of investigation pursued by the Rhode Island bureau is entirely different from that adopted by Massachusetts. The report of the latter showed the total number of persons unemployed at any time during the year, while the table we have just given shows the record of the number of persons unemployed from month to month. In some respects this form of presentation is more valuable than that followed by Massachusetts. We cannot determine the total amount of unemployment that existed during a year, but we can see what was the condition at any one time, as well as the changes from month to month, a record which for most purposes is the more valuable of the two.

Thus examining the table, it will be seen that 5.8 per cent. of all workers were, on the average, returned as unemployed at a given time, and that the largest proportion unemployed in any one month was 7.3 per cent., in February, 1895. The influence of the seasons can also be seen, the per cent. unemployed being distinctly higher during the winter than during the summer'

months.

Altogether this investigation, directed as it was to a single distinct point, must be considered as of no little value. It is only the more to be regretted that the material was not analyzed more in detail by occupations, and a fuller account given of the circumstances under which the information was secured. The material in the detailed tables permits this to be done, but the labor involved is beyond the power of other than a regular statistical bureau.

But one foreign government has attempted a census of unemployment along the lines of the efforts that we have been considering. This attempt, however, may justly be designated a notable one, and its results constitute the most important information that we have in this direction.

Advantage was taken by the German government of the two censuses taken by it in 1895, the one relating to industry on June 14, and the other to population on December 2, to secure as complete information as was practicable concerning the amount of unemployment existing on those dates. The fact that these

censuses show both summer and winter conditions adds to their value.

In taking these two censuses the following three questions were asked of all persons belonging to the class of persons returned as working for a livelihood as employees: (1) Are you at the present time employed, or at work; (2) If not, how many days have you been without work; and (3) Is the fact that you are without work due to temporary physical disability? The inquiry covered all males and females of the employee class. It did not, however, include housewives without any other special calling, persons in the civil or military service who were pensioned, or the widows of such persons, or the recipients of accident and invalidity pensions who were totally and permanently disabled.

The information obtained is analyzed in great detail according to sex, occupation, industry, duration of unemployment, etc. It will be possible for us, however, to reproduce only the more general results.

Excluding the persons engaged in public services, the church and the professions, the returns show that of the 15,497,632 persons, male and female, engaged as employees in gainful pursuits on June 14, 292,678, or 1.89 per cent., were at the moment unemployed. Of this, 0.77 per cent. was due to sickness or other temporary physical ailment, and 1.12 per cent. to other causes. The figures for December 2 were, total 15,641,100; unemployed 762,668, or 4.88 per cent., of which 1.38 per cent. was due to sickness and 3.50 per cent. to other causes.

Following the plan that we have adopted, the degree of reliance that can be placed upon this investigation must first be determined before its detailed results are more closely considered. In this work, we are much assisted by the frank explanations given in the official report of the difficulties encountered and the extent to which they were met, and by a critical study by Dr. Georg Schanz.1

It should be stated, first of all, that these two inquiries differ radically from the American efforts in that a distinct effort was

1Die neuen statistischen Erhebungen über Arbeitslosigkeit in Deutschland. Archiv für soziale Gesetzgebung und Statistik, Band x, Heft III, 1897.

made to meet the peculiar difficulties attending an investigation of unemployment, and to present the results in such a form as to bring out the important facts concerning which it is desirable to have information. This is seen in the limitation of the investigation to persons of the employee class, to the exclusion of persons permanently invalidated or in receipt of pensions, etc., and in the great care exercised in seeing that the questions were correctly understood and answered.

In spite of this care, it is recognized that errors occurred. A certain number of schedules were returned incompletely filled in, and the answers had to be arbitrarily supplied. This was done by presuming, where no answer was given regarding unemployment, that the person was employed, and where the cause of unemployment was not given, that the cause was other than sickness. Again, it is certain that a greater or less misconception existed in regard to what constituted unemployment. Thus persons returned themselves as out of work while they were performing their military service; others, while voluntarily unemployed on account of their marriage or some other event. There were even extreme cases, such as where a person who had learned the trade of plasterer, but had not worked at it for several years, returned himself as unemployed during that time, although he had worked steadily at another trade. Finally there were included a certain number of professional vagabonds, beggars, persons permanently retired from work, etc., whom it was desired to omit.

In consequence of these presumptions the report says that the figures given should be taken as showing certainly the maximum amount of unemployment existing at the dates of the census. This position of the bureau, it seems to us, is the correct one, though the margin of error, in view of the results of the supplementary investigations undertaken by a number of German cities, which will be subsequently alluded to, must be considered as greater than the language of the report would seem to imply. Nevertheless, the investigation, on the whole, must be pronounced to have been successfully carried out, and the results to furnish a very fair indication of the normal amount of unemployment that exists in the German Empire. The errors that exist are