TESTIMONY OF ISADOR LUBIN, COMMISSIONER OF LABOR STATISTICS, DEPARTMENT OF LABOR, WASHINGTON, D. C.

(The witness was duly sworn and testified as follows:) Mr. FAHY. Please state for the record, Doctor, your full name. Dr. LUBIN. Isador Lubin.

Mr. FAHY. And your present position?

Dr. LUBIN. Commissioner of Labor Statistics, Department of Labor.

Mr. FAHY. Now, will you also state, Dr. Lubin, not in elaboration but to give the committee a fairly good idea about it, something of your academic training and experience up to the time you became the officer of the Government in the office you now occupy?

Dr. LUBIN. For 11 years prior to becoming Commissioner of Labor Statistics I was a member of the staff of the Brookings Institution. Prior to that I was a member of the faculty of the University of Michigan, in charge of labor courses.

Prior to that I was statistical expert on the War Industrial Board. Prior to that I was an economist in the Food Administration, and prior to that I was an instructor at the University of Missouri.

Mr. TOLAND. Mr. Chairman, I am willing to concede the qualifications of Dr. Lubin.

Mr. FAHY. I am not proving his qualifications.

Mr. TOLAND. As an expert in the field.

Mr. FAHY. And you have been Commissioner of Labor Statistics

since

Dr. LUBIN. June 1933.

Mr. FAHY. June 1933-the last 7 years. As a part of the duties of your present position, are there compiled by you or under your supervision statistics as to the number of strikes in the United States and their causes, duration, and so forth.

Dr. LUBIN. There are.

Mr. FAHY. And has that been done, to your knowledge, since 1933, each year, currently?

Dr. LUBIN. Yes; currently.

Mr. FAHY. And from your knowledge of the records of the office was it done prior to that time?

Dr. LUBIN. Yes; it was done, in fact, way back-it started in the late eighties, although there was a period of 5 years between the eighties and the present date in which certain types of figures were not collected.

Mr. FAHY. Now, Doctor, do you have certain charts bearing on the question of strikes and their trends and their causes which you think .would be useful to the committee in the study of the effect of the National Labor Relations Act on industrial peace?

Dr. LUBIN. I don't know that I can say that they will be useful to the committee, necessarily. We have these statistical tables kept current. They are charted for our own use, and I have brought them with me, and they tell the story of what has happened to strikes since the eighties.

Mr. FAHY. Would you like to explain them or discuss them?

Dr. LUBIN. If I might, I perhaps might describe how we collect the figures.

Mr. FAHY. Very well. You take charge of the testimony now and go right ahead. I am sure the committee would permit you to do so. Dr. LUBIN. We define a strike as a temporary stoppage of work by a group of employees in order to express a grievance or to enforce a demand. The Bureau counts, however, only such strikes in its records which involve six or more workers and last for 1 day or more. Frequently you may have a temporary stoppage of an hour or two, which we do not include in our official records as strikes. A strike must last at least 1 day before it becomes part of our official record. In addition to compiling statistics each month on the number of strikes, we also compile statistics on the number of workers who are involved in these strikes. For reasons of practicability, we confine the number of workers involved to include only those employees in the plant or establishment who are made idle because of the strike. We do not attempt to separate within the establishment those who initiate the strike and those who are thrown out of work because of a strike of a portion of the employees. In other words, if there is a strike and the plant shuts down, we include all of the workers in the plant as involved in this strike, and not merely the number of people who called the strike.

On the other hand, we do not include workers in other plants who might be affected by the strike. In other words, a plant may shut down because of a strike. Having shut down, it becomes impossible for some other plant to secure certain raw materials. The people in that second plant would not be included in our figures on the number of people directly affected by the strike. These lines of demarcation, of course, are somewhat arbitrary, but are necessary, due to the particular difficulties involved in using any other form of measurement. Since this method has been used throughout the Bureau during all the years in which strike data have been collected-and that goes back some 50 years-comparisons between years and periods are therefore valid. We use the same basis today with these computations as was used 50 years ago, so a comparison of 1939 with 1889 would be on exactly the same basis.

Our third measure of strike activity is what we term man-days idle because of strikes. Please note that this is not man-days lost because of strikes, and there is very definite distinction between man-days idle because of strikes and man-days lost; because certain workers are idle during a strike it does not necessarily follow that the day's work or wages have been lost. Anticipating a strike, there may have been overtime in building up for stock before the strike began, which means. more employment before the strike occurred than otherwise would have taken place. There may have been overtime after the strike closed to fill delayed orders, or there may exist such a chronic oversupply that a prolonged strike does not affect materially the year's output.

In other cases certain employers and workers may suffer a loss due to cessation of work during the strike but other employers and workers would have more work to do. In such cases there is no net loss because of the strike to the economy as a whole.

A good example is that of the automobile industry in 1937. While this industry suffered a great deal of lost time during the spring of

1937 because of strikes, the employers were able to fill all of their orders for the season and began to lay off workers by October.

Another very important example of an instance where a strike may mean lost time but not lost wages in the coal industry. If you go back over the history of the coal industry, you will find that in virtually every instance where there has been a prolonged strike in a given year, despite that fact the number of days employment afforded to the workers in that industry have usually been larger than in normal years when there have been no strikes. In other words, the time lost during strike has, in that industry in particular, been more than made up later in the year, as a result of increase in employment.

In the automobile industry in 1937 the total number of man-hours worked did not show the effect of the strike. In other words, after the strike was over there was more work available making up those unfilled orders that could not be filled during the strike.

In discussing the figures of the Bureau of Labor Statistics, we have divided our strike materials into two general classifications-total strikes as compared with strikes for union recognition and in protest against discrimination. In the latter group, mainly the group that cover strikes for union recognition and in protest against discrimination, other issues were very frequently present, such as wages and hours. However, the group includes only those strikes in which the question of recognition and discrimination were the major issues. Strikes for closed shop have not been included in our statistics under union recognition and discrimination, since the National Labor Relations Board has considered the issue of a closed shop outside the purview of its law.

We secure our information from notices or leaves that are obtained by the Bureau from 700 labor and trade papers, from daily newspapers in all cities of over 50,000 population, and in some industrial communities of less than 50,000. On the basis of these notices that are brought to our attention-and may I say, Mr. Chairman, that every time there is a strike in a local area it is a news item as far as the newspapers are concerned, and we feel that by following the trade press and the labor press and the general newspapers of the country we have a pretty accurate coverage of all of the strikes that take place. It is rarely that a strike occurs without some notice getting into either a trade paper or a labor newspaper or a general newspaper.

On the basis of these notices we send schedules to the unions and to the employer, both, as well as to neutral parties who might know Something about the strike. In a few cases where no replies are received to our inquiries, agents of the Bureau call on the parties to get the required information.

Now, I have brought with me certain charts which the Bureau maintains and other charts which give, in a pictorial form, the statistics of the Bureau of Labor Statistics on strikes. The first chart that I should like to present is this chart called, "Number of Strikes and Workers Involved from 1880 to 1939." This chart covers not only the number of strikes that actually occurred, which is this solid black line, but also the number of workers who were affected by these strikes, which is this broken line, and also the number of man-days idle as a result of the strikes, which is this dotted line here.

You will note that the data on man-days idle started in the twenties. They were not collected prior to that date. The result is that it is

impossible to estimate the number of man-days idle back prior to the middle of the 1920's.

There are certain significant factors that are brought out by this chart. In terms of the numbers of strikes that have occurred since 1881, two important things should be brought out, namely, in 1939, the end of this black line, there were 2.500 strikes, approximately, in the United States. I say "approximately" because the final figures for December have not yet been received and tabulated. In 1938 that number was 2,772; in other words, approximately 275 more than in 1939.

Mr. HALLECK. Where is that point indicated on the chart?

Dr. LUBIN. It is the midpoint between these two points. Incidentally, I might say for the purpose of the record that I have based these all on an index, the average of 1927-29, being 100. The index here is a little over 600, which means there were six times as many strikes in that year, 1937, as there were in 1927-29. In terms of actual numbers that is 4,740 strikes, which was the peak year.

However, it is well to compare that peak figure with the situation back here in 1917, when there were 4,450 strikes. In other words, approximately 22 years ago the actual number of strikes in the United States was 4,450, or approximately 200 less than at the peak year, 1937. Again, it should be pointed out that back here in the early part of the century, in 1901, 1902, 1903, the number of strikes averaged more than 3,000 per year, as compared, as I said a minute ago, to approximately 4,700 in 1937.

In other words, about 38 years ago, in 1903, there were 3,648 strikes in the United States; in 1902 there were 3,240 strikes in the United States; in 1901 there were 3,000 strikes in the United States.

In other words, I think that one of the significant things that will be brought out by this picture is that your peak of strikes usually occurs at points of marked rises in industrial activity. It happened in the early part of the century, it happened in the 5 years starting in 1914, it happened in the period of revival starting in 1934.

Now, as I say, in terms of actual numbers the situation during the last 2 or 3 years has been quite similar in many respects to the situation between 1914 and 1919, and in some extent similar to the situation that prevailed in the United States during the first 5 years of this century.

Now, in terms of the number of workers who are involved, you get another interesting, significant picture, namely, in 1939 we estimate 1,200,000 people were involved in strikes. However, in 1938 that number was 688,000, which is approximately 600.000 less. The answer to the reason for that situation is, for the most part, one fact, the big strike that occurred in the coal industry last spring, when virtually the entire bituminous industry, which employs something like 400,000 people, were affected.

Now, you will note, for example, that in 1917 the number of people who were affected by strikes or involved in strikes, which is at about that point, was 1,227,000, which was more than were involved last year. In 1937 the number was 1,860.000 as compared to 1,200,000 last year, and 688,000 in 1938.

Now, that number, for 1937, namely 1,860,000 people involved, is about 350,000 less than in 1934, when the number of strikes was rela

tively small, but the difference in the number of people affected was around 300,000. Or, if one goes back to 1919, you will note that in that year the number of people affected was 4,160,000, which is more than twice as many people who are affected by strikes in 1937.

Again, the reason for that is particular instances in particular industries. During the year 1919-20 you had a very important strike in the steel industry and a very important strike in the coal-mining industry. Again, in 1922, you had an important strike in the coal industry. In 1935 you had an important strike in the textile industry, the whole industry being shut down. In 1937 you had an important strike in the automobile industry and later on you had a strike in Little Steel and other industries. But in terms of the number of people who were out on strike, the peak, which came in 1919, was twice as great as at any other period for which we have any records.

Now, in terms of the man-days idle, you will note again that in 1927 the number idle was rather marked. As a matter of fact the loss was 26,000,000 days in 1927. On the other hand, in 1937 the loss was 28,000,000 days, a difference approximately of 1,000,000 days between the peak of 1927 and the peak of 1937.

I would like to move on now to the second chart.
Mr. FAHY. May I offer this chart in evidence?

The CHAIRMAN. I was just wondering, Dr. Lubin-this chart and your very elucidating explanation of it-what you think it proves. Dr. LUBIN. Mr. Chairman, I am not here to prove anything. After all, the Bureau of Labor Statistics is a non-partisan, scientific bureau that collects these figures, and I have been asked to tell the story of what has happened to the number of strikes in the United States over the period of years.

The CHAIRMAN. You don't attach anything that occurs on here to the Labor Board or any other board or anything else?

Dr. LUBIN. Of course, you, sir, have the privilege of deducing anything you wish from those charts. I shall show others in which I break these strikes down into strikes for wages and strikes for union recognition and strikes for closed shop, and I shall attempt to tie them with various types of legislation, but I am not going to say that any one particular thing was responsible for the way that curve

appears.

The CHAIRMAN. That is all I want to know.

Mr. HALLECK. Dr. Lubin, you did suggest, I think, in your statement, though, the fact that economic conditions apparently in the past have had much to do with the prevalence of strikes.

Dr. LUBIN. Economic revival and changes in the cost of living have been the two predominant factors in determining the number of strikes. In other words, in those periods where business was picking up, particularly after marked declines, the number of strikes tends to increase very markedly, and during periods of rising cost of living-if you remember in 1919 we had a threatened railroad strike, in 1920 we had a coal strike, in 1921 we had a shopmen's strike. The cost of living was a very important factor. If you remember, President Wilson appointed a commission to see what they could do about this cost of living. In periods of that sort workers demand increased wages to maintain their standard of living.