Second. In view of all the circumstances involved, these three censuses should be taken as of some day in November 1934, and preferably as of November 12, in accordance with the proposal of the Bureau of the Census.

Third. That while the relief of unemployment is an important reason for taking these censuses at this time, the primary consideration to be observed in the selection and control of personnel must be a high order of competence for census work, to the end that there shall be in all districts an accurate and adequate enumeration completed with reasonable promptness.

un

I think this proposal for a census serves 1 major purpose and 2 minor purposes. I should like to speak briefly on those three points. The major purpose is to furnish information on employment, employment, and occupations. That this census will give that sort of information is clearly specified, it seems to me, in the bill that is before you.

Someone at the hearing, I think on Monday morning, suggested that the census of unemployment taken in 1930 was almost worthless. I think it is only fair to say that there were important defects in that census, but it certainly was not worthless. It was a very valuable piece of information.

It was a first attempt, and it is impossible to foresee all the difficulties that will be encountered in any first attempt.

We have now had a good deal of experience that we can fall back upon. This experience includes the 1930 census, and also the trial census of unemployment that has been taken in the three cities of Bridgeport, Conn., Springfield, Ohio, and Lancaster, Pa. I think Mr. Lubin explained that trial census more fully the other day.

The Census Bureau itself is in a vastly better position to take an accurate census at the present time than it was in 1930. I think it is important to recognize that one of the principal functions of a census is to give comprehensive information that enables one to use current information on employment more intelligently.

The Bureau of Labor Statistics in the near future, partly through a C.W.A. project, and partly through a cooperative arrangement with the National Recovery Administration, is going to have vastly better current, monthly information on the amount of employment in the various industries than it has ever had before. The coverage will be much more complete in that more industries will be canvassed than previously. If the information derived from such reports is to be used satisfactorily, it is very important to have a bench mark, something to relate to in the way of a census. It seems to me that this is one reason why it is particularly important to have a census at this time.

The census in connection with the questions of unemployment and employment seems to me to have a double aspect. It is useful, on the one hand, in giving adequate and complete information as a basis for the relief of people who are unemployed. It is useful, on the other hand, in connection with a constructive policy for the adoption of measures to put these unemployed people back into jobs.

It seems to me, therefor, that for purposes of dealing with the unemployment situation, it is very, very important to have a census of employment, unemployment and occupations this fall. So much for the major purposes that this census will serve.

I want to say just a word about two incidental purposes. In the first place, such a census will give an enumeration of the entire population. A census of employment and unemployment cannot be taken

adequately without enumerating the entire population. We know pretty well today what the total population of the United States is, but we do not know where that population is. The situation, as some of the other witnesses have already pointed out, is so bad that it is impossible to make accurate estimates today of the population of most of the important cities in the country.

Population information is important for the purpose of understanding the employment and unemployment situation, and it is useful also for a great many other purposes. I do not believe there can be any agency working in the statistical field that does not have need for population data.

The second of the incidental purposes that I should like to mention concerns the census of agriculture. The census of agriculture has already been provided for in the Fifteenth Decennial Census Act. But I think it is the opinion of all who have studied the problems in connection with the taking of a census of agriculture, that such a census can be taken much more adequately and accurately if a population census is taken at the same time.

The CHAIRMAN. The fact that the 1930 census is now inadequate in many respects for use at the present time does not in any manner impeach the accuracy of the 1930 census, or reflect upon the Bureau officials who took it.

Mr. COPELAND. No; I meant only to say that conditions have changed; and conditions have changed at least as much in the past 2 or 3 years as they would ordinarily change in a period of 10 years. The CHAIRMAN. And the rapidly changing economic conditions have impaired the present usefulness of that census?

Mr. COPELAND. Yes.

The CHAIRMAN. But it was a good, adequate census when it was taken?

Mr. COPELAND. Yes; it was probably the best census we have ever had.

Mr. CRUMP. Who initiated this legislation? Who is asking for it? The CHAIRMAN. I might say to the gentleman from Tennessee (Mr. Crump), practically all the departments asked for it. This is an administration proposal.

That matter was discussed and presented, Mr. Crump, in the first hearing, but unfortunately you had business in other committees on that day and were not able to be present.

Have you completed your statement, Mr. Copeland?

Mr. COPELAND. Yes, sir.

Mr. KINZER. How long have you been connected with this board? Mr. COPELAND. I have been connected with the Central Statistical Board since it was started last summer.

Mr. KINZER. In the summer of 1932?

Mr. COPELAND. In the summer of 1933.

Mr. KINZER. What were you doing before that?

Mr. COFELAND. I have been doing a number of things. I have been connected with Cornell University and the University of Michigan. In 1927 I was for a year with the Brookings Institution. Subsequently, I was connected for half a year with the staff of the National Bureau of Economic Research, when it was making a study for the President's committee on recent economic changes. I was with the Federal Reserve Board's division of research for about a year.

Mr. KINZER. You have said here today that this census will fit in with the agricultural census that has been provided for. Just how will this census augment that?

Mr. COPELAND. If the two censuses are taken concurrently, I think there are two aspects of the matter that will help the census of agriculture.

Mr. KINZER. What assurance do you have, or what reason have you for thinking that they would be taken concurrently?

Mr. COPELAND. That is what the bill provides for.

Mr. KINZER. You mean the bill introduced by Mr. Lozier?

Mr. COPELAND. Yes.

Mr. AUSTIN. They would be taken by the same supervisors, and be taken more economically and with greater efficiency.

Mr. COPELAND. You will find that in section 3 of the bill introduced by Mr. Lozier, H.R. 9391.

Mr. KINZER. Just what does the agricultural census embrace, or what will it embrace?

Mr. COPELAND. You say the census of agriculture?

Mr. KINZER. Yes.

Mr. COPELAND. That is already provided for in the act providing for the 15th decennial census.

Mr. KINZER. What I want to know is, how this will augment that? Mr. COPELAND. I do not know that I would say it augments it so much as it enables it to be taken more accurately.

Mr. KINZER. I wish you would elaborate on that.

Mr. COPELAND. In the first place, it is a rather difficult question to say just what is a farm. You are apt to skip a good many farms if you try to count only those things which you know are farms. It is very much the same situation that exists in connection with taking a census of employment. You cannot count the unemployed without counting other people, too. In the second place, it is difficult. to pay enumerators enough, without having a very costly service, if they go to scattered farms only for the purpose of getting farm information.

If the two censuses are taken concurrently, it will be possible to pay the enumerators a little more money and make it more worth their while to go to every place.

Mr. KINZER. Has it been your experience as a member of this statistical board, that the 1930 decennial census is no longer of any use?

Mr. COPELAND. I do not quite see what you have in mind. The 1790 census is still of use to anybody who is interested in looking back and trying to ascertain the perspective or background as to the use of these censuses. But the 1930 census is now out of date, so far as representing the present situation is concerned.

Mr. KINZER. How long would a census be in date, if taken; for instance this new census?

Mr. COPELAND. Ordinarily, I think it is satifsactory to have a census once every 10 years. But there has been such a tremendous upheaval in the movements of population during the past 4 or 5 years that it seems to me fully as important to have a census in 1934 as it would ordinarily be to have it after a 10-year period has elapsed. Let me refer to what I tried to indicate in my opening statement. One of the purposes of a census is to make possible an interpretation

of current information, monthly data. When you get a long way away from your base in interpreting current data, your interpretations become less dependable. For this reason I think it is important to have a new census which will provide a new bench mark to which to relate current information.

The CHAIRMAN. In what manner, and to what extent, in your opinion, would this census aid in solving the all-important problem of unemployment?

Mr. COPELAND. I tried to indicate that it seems to me there are two directions in which this census will be very helpful, in which, in fact, the information seems to me to be almost essential.

The first is in connection with the relief program. There is an unemployment problem, and something must be done about it. In administering the relief program it is necessary to know where the people are; what their condition is in terms of occupation or lack of occupation; and what occupational experience they have.

In the second place, if you are going to do anything in a constructive way to attempt to place these people; if, for example, you are going to pass on a code under the National Recovery Administration, and on the question of the terms of this code with regard to hours, you need to know, not merely how many people are employed at that time in the industry, but how many people there are available in the various occupations involved, and where those people are located in the country, in order to pass intelligently on that question.

If you are working on a particular Public Works program, you need to know what kind of people there are in the locality concerned and what the relative proportion of the different kinds of people seeking aid in that place is.

Mr. CRUMP. You have given a very intelligent answer to that question, I think.

The CHAIRMAN. Now, I would like to ask you a question with reference to the industrial activities of the Nation. You have explained the farm feature of the measure.

What have you to say with reference to the interest the industrial classes have in a census of this character.

Mr. COPELAND. It seems to me, Mr. Chairman, this information will be of use to industry in its employment policy. It will also be of use to industry to know more accurately what the condition of people in various sections of the country is in terms of purchasing power. It will help in the more intelligent use of current data, and will help to determine what the distribution of the purchasing power in the country is, thus making possible better estimates of the market situation.

Mr. KINZER. Do you draw any distinction between use and the necessity for this census? You have used the word "use" right along, and I take it that you jump over any distinction between use and necessity.

Mr. COPELAND. I thought I had attempted to draw a distinction between "use" and "necessity." I should say that the census would be "useful" to industry; it seems to me it is "necessary" for purposes of intelligently facing the problem of unemployment by the administration.

The CHAIRMAN. Between 1920 and 1929 there was at least a strong drift of population to the cities and industrial centers. Since 1929

and 1930 the current of that movement has been from the cities and industrial centers to agricultural communities; is that true?

Mr. COPELAND. As far as we know, but we do not know how much. For instance, between 1920 and 1930 the Bureau of Agricultural Economics in the Department of Agriculture made estimates of the movement of population between the country and the city, and they got a long way out of line. I think they were around 2 or 3 million people out of line in their estimates of the farm population between 1920 and 1930. I suppose they are as much off now in their estimate as to what the farm population is as they were in 1930.

Mr. LEMKE. I am pretty well acquainted with the central western agricultural areas, and my own observation is that there has not been a movement in population from the cities back to the farms. Can you tell us where such a drift is taking place? Have you any information about that? It may be that the cities have decreased in population, but the people who lived in the cities have not gone back to the farms. They may be out in the highways and byways.

Mr. COPELAND. I think we have a better opportunity to know that the cities have been decreased than to know where the people have gone.

Mr. LEMKE. I would like to know where the people went, if there is not anything but guesswork about it.

Mr. COPELAND. That seems to me to be substantially the situation we are in. Take, for instance, a large city like Detroit. We do not know what the population is, within 75,000, today. We have no way of knowing definitely, but we believe and we have every reason to think we are correct, that there has been a considerable decrease in the population of Detroit. Where they have gone, I do not know.

Mr. LEMKE. I have discovered this, to be frank with you. I think the people that have been lost are sleeping in the alleys and in the gutters, and that is why you have not got a record of them. They are not on the farms in that part of the country. There has been a decrease of the population there right along and still is.

Mr. COPELAND. There is good information to substantiate the view that they are not there. I do not mean that there are any people sleeping in alleys, but that there is a definite decrease in the population of a considerable number of cities.

Mr. LEMKE. From my information, there is a decrease both in the cities and in the country, and I am interested to know where the people are.

Mr. COPELAND. I do not see how you are going to find out without

a census.

The CHAIRMAN. May I say that the agricultural experiment station at Purdue University made an extensive study of the shift in the population in two Indiana counties in marginal agricultural areas of the shift in population since 1900. They found in 31⁄2 years ending in June 1933, that the number of people returning amounted to an increase in population of 36 percent and 16 percent respectively. The tax levied on new wealth brought in by newcomers amounted to only 16 percent and 13 percent respectively of the amount of relief necessary to support them.

Apropos of the very pertinent observation of Mr. Lemke, I think that the drift to the farms is more pronounced in the Mississippi Valley, where the agricultural areas are close to the great industrial