have taken censuses recently. Have they made any changes in their technique in getting population statistics, etc., that it might be of advantage to adopt here? Mr. GOSNELL. I could not say. Doctor HILL. I am not familiar with recent developments in Europe; but my impression is that they are proceeding along, the same lines as heretofore. Mr. JACOBSTEIN. I thought we might learn something as to technique from the recent censuses they have taken over there. Doctor HILL. We will examine into that matter. Mr. JACOBSTEIN. I do not think we ought to use Turkish methods. In Turkey, I am told they ordered that a certain day be designated as census day; and everybody had to stay at home that day; the police were in charge of the matter and they saw to it that everybody stayed at home on that day, and the census was completed on that one day for the whole Turkish Empire. Mr. SELVIG. Mr. Chairman, I have one or two questions to ask regarding the proposed agricultural census. I have had occasion to study the census, especially with reference to the agricultural schedule, for the purpose of making a comparison during the census years with various estimates that come from other sources, especially the Department of Agriculture. And I find that every 5 or 10 years we have figures that do not agree between those two sources for that particular period. And I am wondering whether or not there will be anything in the new census which will probably result in greater accuracy, and also in providing more data with reference to agriculture. Now, there was a paper presented at one of the recent scientific meetings in Washington during the holidays, which I have found very interesting. It was prepared by Doctor Davis, of Leland Stanford University. I would like permission to insert in the record a brief abstract of that paper. I do not care to take the time of the committee this morning to read it. The CHAIRMAN. Certainly; if it is agreeable to the committee, we will insert it in the record. (The statement referred to was subsequently submitted by Mr. Selvig, and is as follows:) In considering the taking of the regular decennial census of 1930, I wish to call to the attention of the committee certain facts regarding the agricultural census in order to secure, if possible, greater accuracy in the forthcomming census and data of greter value to the country. There are some shortcomings in the basic agricultural data in the agricultural census of 1925 with respect to crop acreage and production to which I wish to call your attention. In comparing census figures for acreage and production of certain crops with revised estimates subsequently published by the Department of Agriculture, it is found that the department figures for acreage are 2 to 4 per cent higher than the census figures, for most crops, and much higher for several of the lesser crops (e. g. rye, rice, potatoes, sweet potatoes, tobacco). The production figures of the Department of Agriculture are much further above those of the census-in wheat 8 per cent, in most other crops considerably more. I will insert a table showing that these differences are very marked. The data of animal population in the 1925 census and the statistics showing the production of meat, lard, milk, butter, and other animal products differ materially from data issued by the Department of Agriculture. The output of the animal industries is a large and increasing portion of the total of the agricultural production and is, therefore, of the greatest importance in connection with agricultural statistics. Unfortunately there have already been three changes in the date when the census for animal products has been taken. For the years 1850 to 1900 the figures were as of June 1, for 1910 as of April 15, and for 1920 and 1925 as of January 1. The bill under consideration provides that the census should be taken as of April 1. These changes in date make it difficult to secure figures that are properly comparable. In this field of animal statistics there are in some instances radical differences between the estimates of the Bureau of Agricultural Economics of the Department of Agriculture and the Census figures. In the case of sheep and lambs the census figures show an increase of sheep and lambs of 556,000 or 11⁄2 per cent, while the Department of Agriculture figures show a decrease of 2,131,000, of 5 per cent. For hogs the census data shows a decline of 8,492,000, or about 14 per cent, while the Department of Agriculture shows a decline of only 4,245,000, or only 7 per cent. There should not be this difference in animal population figures for the same year. Both figures can not be correct. In most of these the Department of Agriculture figures are higher than those given in the census for the same year. I will insert a table showing these differences. Since 1907 the Department of Agriculture has issued annually estimates of the annual production of meat and other livestock products. The differences between the figures covering this part of our agricultural production in the census and the estimates prepared by the Department of Agriculture are in many cases radical where it is possible to make comparisons. It is important to obtain more adequate basic data within the range of work to be undertaken by the census. My only purpose in calling this matter to the attention of the committee is to secure if possible a solution to this problem and to bring before the committee concrete suggestions looking to the improvement in procedure in respect to the 1930 census as it relates to the agricultural schedules. I wish to emphasize first the importance of securing future census data on agriculture that will be as accurate and as complete as possible. The taking of the census entails a considerable expenditure of public funds. Producers as well as consumers are vitally interested in having as accurate and complete agricultural data in the census as it is possible to secure. It is likely that additional funds will be needed but the importance of the work justifies this additional expenditure. Trained agricultural economists should be secured to assist in making up the agricultural schedules and to suggest studies that should be undertaken to insure more accurate and complete data in the census of 1930. It is suggested further that there should be close coooperation between the members of the staff of the Census Bureau with the Department of Agriculture and associated State agencies. It has been suggested also that studies should be made that will lead to carefully checked and thoroughly reliable revisions, State by State, for a considerable series of years in the past in the statistics involving both crop and animal products. These studies should from time to time be issued and would assist materially in improving subsequent census data. Annual livestock and animal products data taken by the Census Bureau would seem to be a necessary step in order to give the food producers in the United States reliable information for their benefit. I am submitting two tables which were compiled by Dr. Joseph S. Davis, food research institute, Stanford University, California, in a paper read before the American Statistical Association in Washington, D. C., December 27, 1927. Table I gives the acreage and production of certain crops in 1924 according to census reports and revised estimates of the Department of Agriculture. You will note that the differences in acreage are very great with respect to some of these crops, particularly rye, rice, and potatoes. In each instance the estimate by the Department of Agriculture is higher than the census data for the same year. The differences between these two sources of information in the production of important crops is radical, as will be seen by referring to Table I. Table II gives the livestock count January 1, 1920 and 1925, according to census reports and revised estimates of the Department of Agriculture. In looking over these tables, it is found again that the estimates by the Department of Agriculture are higher than figures given in the census. TABLE I.—Acreage and production of certain crops in 1924, according to census reports and revised estimates of the Department of Agriculture 1 - Dets from Census of Agriculture, 1925, and Crops and Markets, Monthly Supplement, December, 1926, Table II—Livestock count January 1, 1920 and 1925, according to census reports and revised estimates of the Department of Agriculture 1 1 Data from Census of Agriculture, 1925, Crops and Markets, and Agriculture Yearbooks. Mr. JACOBSTEIN. Does that relate to the farm indebtedness and farm mortgages? Mr. SELVIG. That is dealing with the agricultural statistics of the census. The CHAIRMAN. Perhaps one of the gentlemen from the Bureau of the Census might inform us in regard to this matter. Mr. HIRSCH. Mr. Austin, the chief statistician of the agricultural census, can tell you as to that. The CHAIRMAN. Can you give us any information, Mr. Austin? Mr. AUSTIN. The census of agriculture is an actual enumeration of farms. The country is divided into supervisors' districts, and heretofore those districts have conformed to congressional districts. Then each of those districts are divided into smaller districts, called "enumeration districts." Enumerators are appointed, local men living in the district and supposed to know something of the population and agriculture in that district. That local man makes his visits around the farm houses in his district. While there, he reports the farm information; he makes the farm inquiries. He gets that information himself. Now, that is the way the census of agriculture is made. The information furnished by the Bureau of Economics of the Department of Agriculture, to which reference has been made, is based upon estimates. They make no actual enumeration. They revise their estimates, and use as the base the decennial census figures, and now the five-year figures on agriculture as a new basis, on which they figure out their estimates. Now, it is the difference between an actual enumeration and an estimate. Mr. SELVIG. I am wondering if it would not be possible, in order to get more accuracy and more adequate data, to have the agricultural census taken oftener than once every five years, in order that we might have something that we could depend upon. The CHAIRMAN. That is a question of policy. Mr. SELVIG. That is a suggestion that was made by Doctor Davis in the scientific meetings that I have referred to; and it met with a good deal of approval among the scientists who were present. The CHAIRMAN. How often did they suggest? Mr. SELVIG. On agriculture, it is every five years now. The CHAIRMAN. It is every five years now; but what did they suggest? You said they thought that was not frequent enough. Mr. SELVIG. It seems to me that more frequent information with regard to livestock and agriculture would be very desirable. Mr. THURSTON. Has there not been considerable complaint on the part of farm organizations, for instance, that the collection of this data and these statistics has worked to the disadvantage of the producer and farmer, and therefore there has been considerable hostility to them? The CHAIRMAN. Yes. Mr. Rankin will remember the complaints about the cotton reports. Mr. AUSTIN. It is largely a question of cost; because it takes a great deal of money to make an agricultural census. Mr. RANKIN. Mr. Austin, have you one of the agricultural schedules with you? Mr. AUSTIN. I think so. Mr. JACOBSTEIN. Right along that line, let me ask if we will be able to see all the schedules in advance? Doctor HILL. Do you mean the schedules for the next census? Mr. JACOBSTEIN. Yes. Doctor HILL. Surely. Mr. AUSTIN. The farm schedule in 1920 contains 415 inquiries; and it took about three years to determine it. The one for 1925 contains 178 inquiries; and it took us 18 months to decide upon the schedule. Mr. RANKIN. Mr. Austin, is cottonseed on that schedule? Mr. RANKIN. Why did you not? Mr. AUSTIN. Well, the difficulty is in making up the schedule. When we start out to make a census schedule of agriculture we consult the Department of Agriculture; farm organizations; the editors of the leading farm publications; the State Departments of Agriculture; the extension service; and the people who are directly interested in that one subject throughout the country. We usually start in with a schedule of from 1,200 to 1,500 inquiries; and then it is a case of elimination, keeping, as nearly as we can, the most important questions for the country at large, and not any particular section. Mr. RANKIN. Now, you say "the most important questions for the country at large.' You are aware, are you not, that one of the chief crops in the United States is cottonseed? Mr. AUSTIN. Yes, I know that. Mr. RANKIN. And that we produce on an average about 500,000,000 bushels of cottonseed a year. And that is just as important as any other crop, and with the possible exception of about three crops, is possibly worth more than any other one crop produced in the United States. And I am at a loss to understand why that was not included in the 1925 schedule. a Mr. AUSTIN. That is the reason. It is a question of getting the schedule down within limits where it can be handled by an enumerator. Mr. MOORMAN. Well, as a Representative from Kentucky, State in which they use thousands of carloads of cottonseed each year, I will ask, in view of the question asked by the gentleman from Mississippi (Mr. Rankin) and the fact that it has been brought to your attention that a crop as important as that is has been omitted from the census, whether it will be included in the next schedule? Mr. AUSTIN. Of 1930? Mr. MOORMAN. Yes. Mr. AUSTIN. I imagine it will be. But here is another thing about cottonseed: The Bureau of the Census makes a monthly report on cottonseed oil. But when we came to the 1925 census we had to have a small schedule, because we had a small appropriation. That census cost over $4,000,000. The 1920 census cost $7,000,000. Mr. RANKIN. Now, Mr. Austin, you say that you make a report how often on cottonseed? Mr. AUSTIN. We make a monthly report. Mr. RANKIN. On cottonseed? Mr. AUSTIN. On cottonseed oil mills. The CHAIRMAN. You send it out on those little cards that you have? Mr. AUSTIN. Yes. Mr. RANKIN. Do you make that report on cottonseed produced in the country? Mr. AUSTIN. We make that report on cottonseed received at the mill and consumed at the mill. Mr. RANKIN. In other words, the reports you make are more for the benefit of the mills and the mill trade than for the benefit of the farmer? Mr. AUSTIN. Well, it is part of the reports of the Census Office which include the amount of cotton ginned during the ginning season; we have a report from the mills in the same way. We have a report from the manufacturing plants of the country, and we have three reports on cotton. Mr. RANKIN. Do you make a report on the amount of cottonseed meal produced? |