subdivided as follows: In elementary school; in high school; in vocational school, day-trade school, part-time school, evening school; in higher education.

BESS GOODYKOONTZ,
Assistant Commissioner.

The CHAIRMAN. I also submit a communication from Mr. D. A. Skinner, secretary Chamber of Commerce of the United States, in relation to Honolulu and Hawaii.

(The statement referred to is as follows:)

CHAMBER OF COMMERCE OF THE UNITED STATES OF AMERICA,
Washington, April 28, 1934.

Hon. RALPH F. LOZIER,
Chairman Committee on Census,

House of Representatives, Washington, D.C.

DEAR SIR: The Chamber of Commerce of the United States has had some correspondence with the Chamber of Commerce of Honolulu with respect to the possibility of the extension of the surveys of the various governmental statistical agencies to include the Territory of Hawaii.

This request is being referred to you, as chairman of the Census Committee, in view of the proposal, now before your committee, to take a census of population as of November 12, 1934.

The Honolulu chamber, in its letter, says, "It is our desire that when surveys are made having to do with the business of the Nation that Hawaii be made part of such surveys, because in this way Hawaii will be better understood by the people of the United States and also because Hawaii is an integral part of the United States and pays into the United States Treasury more money in revenue than is paid by 12 of our States of the United States." (In 1932 Hawaii paid into the Treasury more than 5 million dollars in Internal Revenue and Customs receipts.) "On several previous occasions, when requests of this nature have been presented, we have been advised that our request arrived too late or limited funds did not permit the inclusion of Hawaii."

The Honolulu chamber also requests that the Territory of Hawaii be included in all Census and other statistical reports of the United States Government in its alphabetical place after "Georgia", and not grouped with the possessions of the United States or reported in a separate volume.

In this connection we might call your attention to the inclusion in the census of 1910 of the then Territory of Arizona in its alphabetical place among the States following "Alabama".

This treatment would seem to be analagous to the present request of the Honolulu chamber.

Very truly yours,

D. A. SKINNER, Secretary. The CHAIRMAN. Dr. Isador Lubin, of the Department of Labor, will please take the stand. Doctor, you may make such statement apropos the pending legislation as you deem proper.

STATEMENT OF DR. ISADOR LUBIN, COMMISSIONER OF LABOR STATISTICS, DEPARTMENT OF LABOR

Dr. LUBIN. Mr. Chairman, the question of unemployment in the United States has been attracting the attention of the entire administration. The difficulties we are facing are, in a measure, due to the fact that we have not any idea of how many unemployed people there are in the United States. As you gentlemen know, there are three different estimates currently published by so-called "reputable organizations." One, the National Industrial Conference Board, in a recent statement, estimated something in excess of 8 million; another, the Alexander Hamilton Institute, has estimated something in excess of 13 million, and the American Federation of Labor has estimated the unemployed as somewhere around 11 million. There

is a fourth estimate published by the Cleveland Trust Co. These estimates vary by as much as 40 percent. They are simply guesses, and one guess is just as good as the other.

All of them have certain weaknesses, and all of these estimators admit openly and frankly that their figures are in part based on guesses which cannot be definitely substantiated. We must know definitely just how many people in the United States are unemployed so that we can plan for our future policies. Concretely, the only information we have in the United States today that we are willing to say has any approach to accuracy are the figures of the actual employment, and not of the unemployment.

We know how many people are being employed in the various industries of the country. I am not saying it in a boasting spirit, but I believe that we know within 10 percent of accuracy the actual number of people employed in the 90 manufacturing industries that we cover. We know within that range, how many people are employed, month by month, but we do not know how many are unemployed. And there is no way of finding out the number of unemployed unless we take an actual census of unemployment. We must know what the potential labor supply is in the various industries in order to make our plans for the future.

There is no use in appropriating money for public works unless you know where the unemployed people are and where the employment should be provided. When I say there is no use appropriating money for public works under those conditions, I may be exaggerating in a sense, but the effectiveness of the public-works program can be increased 100 percent if we only know how many more workers could be absorbed through the public-works program and how many would be given employment in as a result of such appropriations in the heavy manufacturing industries, or in the manufacture of the so-called "durable goods." We know how many people are employed in those industries, and we know how many people were employed in those industries in 1929, and who are no longer being employed in those industries, but we do not know whether those people are now employed in other industries. The purpose of an unemployment census, as we see it, is to make it possible for us to know just what that situation is, and to know where we may absorb more of the unemployed.

I have an interesting illustration of the sort of problem that we run across in the Department of Labor, which would be directly affected by such a census as that proposed by the Bureau of the Census. We have a ship-building code which originally provided 30 hours per week employment on ships that were being built with P.W.A. money; it provides for 32 hours of employment on ships being built in private shipyards under Navy contracts, and 36 hours for ships being built under private contract.

The shipbuilders came in and said that they could not operate their plants so as to meet the naval program on that basis, and that they must have a 40-hour week. This was done, they claimed, because during the months of April, May, June, and July, there would be a shortage of shipbuilding labor. Fortunately, we had the census figures for 1930 and for 1920. We knew how many people had been employed in those various occupations, and we came to the conclusion that there must be in this country thousands upon thousands of

skilled workers who were formerly employed as ship-building workers. But where those ship-building workers are, we could not say. That is something we do not know, and we will not be able to know without a census of occupations.

The United States Employment Service has ahead of it the job of trying to bring the worker into contact with the job. I have an another illustration of that: There is a big bridge being built across San Francisco Bay, and the claim is made that there is a shortage of certain types of labor there. The question presented is whether we should permit an extension of the hours of employment on that bridge, the hours of employment being specifically provided for in the law. The local contractor there says he cannot get enough of certain types of labor to do the work without an extension of hours of employment.

The United States Employment Service knows there is unemployed labor of the type needed; if not there, certainly in other parts of the country. We know that all over the United States there are workers of that sort. In order to meet those problems, we must know who the unemployed millions are, where they are, and what they can do, in order that we may shift them from one industry to another on the basis of their experience and training. A census of occupations will give us a picture of what those people can do.

We have still another problem, and that relates to the youngsters who have come out of high school during the past 4 years. We also have the problem of the young people who have come out of college during that period. Many of those people have been asking for jobs, and we should know what they have been doing for the last 4 or 5 years, what they are good for industrially, and what industries have picked up sufficiently to absorb them. It is to questions of that sort that we expect to get answers from this census of unemployment, and the answers to those questions are vital, if we are ever to have a program in this country which will not only meet their needs, but also keep up the morale of those youngsters and direct them into activities where they can make a living.

There is no use in training telegraphers if we find that nobody wants telegraphers, or that the telegraphers are being supplanted by the automatic machines that the telegraph companies are now using. There is no use in training carpenters if we already have too many unemployed carpenters in the country. We know that we have a lot of unemployed carpenters, but where they are and what they are doing, we do not know. With a census of employment, unemployment, and so forth, we will know how many unemployed there are, where they are, and what those unemployed people can do. We must have that information if we are to be able to find jobs for them. Furthermore, with that sort of data, it will be possible for us to estimate with some measure of accuracy the number of unemployed from month to month. If we can get a base line from which to measure, if we take November 12, 1934, for example, and know that on that date there were so many unemployed people, then the Bureau of Labor Statistics-on the basis of the employment and pay-roll figures that it collects each month from the most important manufacturers of the country, and from other service industries throughout the country-would have a fair measure of just how many people have been taken out of the pool of the unemployed from month to month, and how many have been added to it.

We now cover in our employment and pay-roll inquiries information which affects over 50 percent of the workers employed in the manufacturing industries. This is a very fair sample of what is taking place We are using this information in estimating how many new people have been taken on, and how many have been fired, and we are showing how the pay rolls have gone up or down.

We are extending our coverage to other industries, such as the service industries, including the distribution and transportation services. This we are doing through the aid and cooperation of the Civil Works Administration. We have put white-collar workers to work in covering the various establishments, and we have expanded the sample so that we can get a fair picture of what is happening in the nonmanufacturing industries. By the middle of July, we hope to have as accurate figures as can be gotten as to what is happening to the unemployed. If you will give us a base from which to measure, we will be able to tell you accurately, after the figures are available from the census, just how many people are unemployed, just as we can now tell you how many people are actually employed.

Mr. KERR. For how long do you think a census of that kind would serve as a criterion to measure these matters by? How long would a census of this sort, with labor shifting from one place to another, be authentic and useful?

Dr. LUBIN. I am glad you asked that question, because of an inquiry that we have just been completing in the Bureau of Labor Statistics. We have made three censuses of unemployment in three cities in this country, namely, Bridgeport, Conn.; Lancaster, Pa.; and Springfield, Ohio. The purpose of those tests, or inquiries, was to find out whether there is any way of actually measuring or estimating, by simply taking a sample here and there, the total number of unemployed people in a community. In each of these cities we have covered the whole city. At the same time, we made all sorts of sample tests of unemployment, using for the purpose school blocks, gas-meter registrations, and so forth, taking the samples at random, and comparing the unemployed found among those samples with the total number of unemployed in the city. We hope that with the experience we have had in these three cities, it may be possible to work out some way of estimating unemployment, after this census has been taken, by making sample studies in 50 or 100 industrial centers, say, every other year. On the basis of that estimate, we can get a picture of what is happening in the country as a whole. By and large, I would say that with the coverage we have of the manufacturing and nonmanufacturing industries, with the data that we have with respect to the graduates from schools, and the data we have respecting immigration and emigration, the results of this census will be good for at least 5 years, and with such lesser investigations as we may care to make every other year, I should say that they might last for 10 years. Of course, if we have a repetition of what happened between 1900 and 1933, our estimates would not be worth much.

The CHAIRMAN. You may proceed with your statement, Doctor. Dr. LUBIN. Now, the Department of Labor, as I was saying, has data on unemployment in these three cities. Incidentally, I would say that the Bureau of the Census cooperated with us in that inquiry. That was because we felt that if the time came when they would have

to take another census of unemployment, they would like to profit by the experience we had in taking that census. Consequently, in cooperation with the Bureau of the Census, we drew up certain schedules. In the schedules used in these three cities, we used different types of questions to see which was the best way to get the information that we wanted. As I have said, the Census Bureau has been in constant touch with us in that work, and they will undoubtedly profit by the experience that we have had in this work.

That, Mr. Chairman, in short, is the interest of the Department of Labor. We need this data, first, so that the Secretary of Labor can recommend to the President and to the Congress such policy as is deemed essential, and such a policy cannot be well formulative unless we know what our problem is-not only the magnitude of it, but where the problem centers. In the second place, we need the information in order to develop a plan that will be of help in the activities of the United States Employment Service, and, thirdly, we need this information in order that we may know what to do with those youngsters who have not found positions after graduating from the high schools, and who have been doing nothing for the last 3 or 4 years.

There is one other matter that I would like, with your permission, to bring to the attention of the committee. That is the matter of two amendments to the bill, which I would like to suggest for your consideration. The first amendment concerns section 1, line 7. I would like to suggest, with your permission, that in line 7, after the words, "Director of the Census," a comma be inserted, and the following words added: "In consultation with the Secretary of Labor." I suggest that because the Department of Labor is vitally interested in the problem of unemployment.

This amendment does not mean that the Secretary of Labor will have, in any sense, any direction of it, but, for the reasons I have stated, we would like to have those words, "In consultation with the Secretary of Labor", added.

The second amendment concerns the material on employment and unemployment, which was collected by the United States Department of Labor through the unemployed white-collar workers. This material is of vital importance, not only in informing the public as to what the situation was in the three cities in which these investigations were made, but it is of vital importance in order that others working in the field of unemployment may know the methodology employed, and how they can use the data. In the case of the employment and pay-roll data, we have visited more than a hundred thousand big firms in this country, and secured their cooperation. We have covered all types of industry, for which no data have ever been available before. Our data show how many people are employed in wholesale and retail grocery stores, in hardware distribution, in laundries, dry-cleaning establishments, and many other industries on which we must depend, apparently, judging from present experience, in absorbing more and more workers of this country.

This information has been collected, and it is being tabulated, but we believe that these data should be made public for the use of those who want to work in those fields, and so that the industries that are concerned will know what has been happening to their employment and pay rolls. Consequently, I suggest for your consideration an amendment in section 5.