Mr. RANKIN. If this provision had been in the law in 1920, would it have enabled you to solve those difficulties that you had with reference to taking the census in the rural districts?

Mr. GOSNELL. Absolutely.

Mr. JACOBSTEIN. Why do you say that? What was the difficulty in 1920 that will be overcome by this bill?

Mr. GOSNELL. For instance, take the southwest part of Arizona as an illustration. We were only allowed to pay the enumerators 4 cents a name and 30 cents a farm, or employ them on a per diem basis. Now, the per diem employment, as I have explained, has not proved at all satisfactory. The supervisor, in many instances, is 150 to 200 miles from him and cannot, of course, supervise his work. Whereas, if he is enumerating on a piece price basis, he has to turn in the work to get the money.

Mr. JACOBSTEIN. You mean the difficulty in 1920 was that the per diem worker did not do justice in the enumeration of the population of the rural district?'

Mr. GOSNELL. He did not.

Mr. JACOBSTEIN. You believe that?

Mr. GOSNELL. Yes.

Mr. RANKIN. I will say in this connection that we had a representative from your State, Mr. Moorman, who had charge of this work in Kentucky; and he said he found it utterly impossible to get men to do this work on the compensation provided by the Government.

Mr. GOSNELL. I remember that case very well, Mr. Rankin. Mr. JACOBSTEIN. May I ask another question at this point? Did the rate of compensation permit you to get the quality of enumerators that you wanted?

Mr. GOSNELL. No, it did not.

Mr. JACOBSTEIN. So that the rate was too low; aside from the fact that the system was not right, the rate was too low. So there were two improper conditions?

Mr. MOORMAN. You anticipate getting the proper class of service for the next census, do you?

Mr. GOSNELL. Yes.

Mr. JACOBSTEIN. That is, through higher pay?

Mr. GOSNELL. Through higher pay and more equitable rates of compensation.

Mr. JACOBSTEIN. Do you find that the higher pay itself is an assurance that you will get a better quality of enumerators? Mr. GOSNELL. I think it will attract better people and more qualified people.

Mr. JACOBSTEIN. You think that is true?

Mr. GOSNELL. Yes; I think that is true. Another difficulty that we had was that the enumerator in a district where the work was easy would get the same rate of compensation as the enumerator in a district where it was difficult to enumerate.

Mr. JACOBSTEIN. Suppose you raised the pay and more people wanted the positions; and it became a political proposition and you had no civil service features to protect you: How could you assure us that you would get a higher quality of enumerators?

Mr. GOSNELL. We would have to depend on the judgment and honesty of the supervisor who appoints these enumerators.

Mr. JACOBSTEIN. You think it is safe to go ahead on that basis, even though the civil service does not apply?

Mr. GoSNELL. Yes; I do. And there is another reason for carrying this provision in the bill: At the 1920 census of agriculture, it was necessary to employ enumerators in Oklahoma on a per diem basis. The census of agriculture in that State in 1920 cost $120,000 as the enumerators' compensation. In 1925, when we had a law which permitted us to pay any rate of compensation acceptable to the director, we took the same census of agriculture for $60,000. There was a clear saving of more than 50 per cent in the compensation of the enumerators on a job that was not taken in conjunction with a census of population.

Mr. JOHNSON. Covering the same data?

Mr. GOSNELL. Covering practically the same data, yes. The schedule was, I believe, somewhat different.

Mr. RANKIN. Let us not overlook this element, however, Mr. Jacobstein, and that is that the census of 1920 was taken at the peak of high prices, when a man could go out and get $6 or $8 a day, whereas in 1925 the same man could not get $3 a day for his work. So that was one thing that made it difficult, in my section of the country, at least, to get enumerators to do this work.

Mr. GOSNELL. Yes, that is very true.

Mr. RANKIN. The pay simply did not measure up to the pay that they could receive in some other business.

Mr. GOSNELL. Yes.

Mr. JACOBSTEIN. How does this rate of pay that you have in mind compare with the rate of pay in taking a State census?

Mr. GOSNELL. Why, the State that I know something about the rate of compensation in is New York. I believe that our rate of pay in the last census, 1920, was considerably lower than that paid in the New York State census.

Doctor HILL. That is a fact.

Mr. GOSNELL. In New York they gave them a per diem allowance of $2 or $3 a day, and they so much additional for each name that they enumerated.

The CHAIRMAN. Let me ask you this question: When this bill becomes a law, the Census Bureau will go before the Bureau of the Budget, and later before the Committee on Appropriations, in order to get the appropriation for the taking of this census. Have you made an estimate as to how much officials you will ask an allocation for, or a lump sum appropriation for, in order to pay their salaries? What I am trying to get at is, how many people will be employed under this bill, in addition to those that are now employed? Mr. GOSNELL. May I answer that, Doctor Hill?

Doctor HILL. Yes.

Mr. GOSNELL. There will be approximately 375 supervisors in the field. There will be from 9,000 to 10,000 clerks, special agents and interpretors combined, and about 90,000 or 95,000 enumerators. The CHAIRMAN. And on that estimate you will go before the Budget Bureau and the Committee on Appropriations asking for the amount necessary to pay them?

Mr. GOSNELL. Yes, and specifying

The CHAIRMAN (interposing). Specifying that as the basis?

Mr. GOSNELL. And specifying the minimum and maximum rates of compensation.

The CHAIRMAN. The Committee on Appropriations would ask you that anyhow.

Mr. JACOBSTEIN. Would they not ask for a lump sum to cover that? The CHAIRMAN. I presume it would come in as a lump sum, but information would have to be in the hands of General Lord and the Committee on Appropriations of the House. Of course we have nothing to do with that.

Mr. RANKIN. I was going to say that the limitation on expenditures is going to be looked after by the Committee on Appropriations. The CHAIRMAN. Yes, by the Budget Bureau and the Committee on Appropriations.

Mr. RANKIN. Yes.

Mr. JACOBSTEIN. What is the estimated population for 1930— about 120,000,000?

Doctor HILL. 120,000,000, or a little over that.

Mr. JACOBSTEIN. That would be 120,000,000 at 4 cents per capita? Mr. GOSNELL. It will average higher than that; it will average 8 cents.

Mr. JACOBSTEIN. That is about $10,000,000 for the population census?

Mr. GOSNELL. Approximately that-well, it would not be quite that much; it would be about $7,000,000.

The CHAIRMAN. You will have the figures to submit to the Budget Bureau and the Committee on Appropriations, on which they can make their allocation?

Doctor HILL. Yes.

Mr. JACOBSTEIN. When will those enumerators be selected? Doctor HILL. About 30 days before they make the enumeration. Mr. JACOBSTEIN. Under this bill when will that be?

Mr. GOSNELL. This bill provides for the taking of the population census as of the first day of April.

Mr. JACOBSTEIN. Then these enumerators will be appointed in February or March, 1930?

Mr. GoSNELL. Yes, they will be selected at that time; but they will not receive any compensation until they actually begin the enumeration.

May I make one more statement about rates of compensation? We had in the last census act, in 1920, a provision which permitted us to pay $2 per day, and so much per capita and so much per farm for the enumeration. We found in practically all cases, or in many cases, that those enumerators would draw a compensation of $2 per diem and probably not enumerate any person or any farm on many of those days. And that is the thing we are trying to get away from by taking the compensation on a piece price basis.

Mr. THURSTON. Have you any idea about what is the average time necessary on the part of the enumerator to perform the work in his district?

Mr. GOSNELL. Do you mean the time that will elapse?

Mr. THURSTON. The total time?

Mr. GOSNELL. In the urban districts, from 15 to 30 days; in the rural districts, 30 to 40 days. In some cases, where bad weather,

heavy snows, etc., may prevent the enumerator from working at all for three or four weeks, it may take three months.

Mr. JACOBSTEIN.. Of course, in this bill you propose to adopt an earlier enumeration?

Mr. GOSNELL. This bill provides for the enumeration of population at one time and the enumeration of agriculture at another.

Mr. JACOBSTEIN. When is the enumeration of population to begin? Mr. GOSNELL. The census of population is to be taken as of the 1st day of April, and the farm enumeration as of the 1st of the preceding November.

Mr. JACOBSTEIN. When you say "farm enumeration", do you mean population on farms, or farm data?

Mr. GOSNELL. Farm data.

Mr. JACOBSTEIN. But the population of the entire country will be taken on the same date?

Mr. GOSNELL. On the farms?

Mr. JACOBSTEIN. No; just the population statistics?
Mr. GOSNELL. Yes.

The CHAIRMAN. The bill provides in section 6:

That the census of the population required by section 1 of this act shall be taken as of the first day of April and the census of agriculture shall be taken as of the first day of the preceding November, and it shall be the duty of each enumerator to commence the enumeration of his district on the day following, unless the Director of the Census in his discretion shall change the date. [Etc.] Mr. RANKIN. Why do you fix the date as of November 1 for the farm enumeration?

Mr. GOSNELL. That is a question that Doctor Hill can answer. Mr. RANKIN. All right. want to ask you some questions. What

is your position?

Mr. GOSNELL. I am in charge of the field division.

Mr. RANKIN. I do not know whether you are the one for me to ask this question of or not; but what does your agricultural enumeration contain?

Mr. GOSNELL. That is a question for Mr. Austin to answer. He is the chief statistician.

Mr. RANKIN. All right. I will reserve that question, then.

Mr. GOSNELL. May I explain this? About six years ago the director consolidated the administration of the entire field force into one division, instead of having it spread all over the office, the idea being that if we sent a man to one town to obtain information of one type he should get all the information necessary for the census at that place at that time, whereas formerly we sent different agents to the different cities to collect special data.

Mr. JACOBSTEIN. May I ask a question right there? Does the Census Bureau have any method of checking up the quality of the work done at the time it is done?

Mr. GOSNELL. Yes, it is checked by the supervisor. That is the purpose of the employment of supervisors' clerks, to check that work. Mr. JACOBSTEIN. What does he do?

Mr. GOSNELL. He checks it to see that, so far as he can tell, the answers are approximately correct. He checks it, I believe, with the population at the last census of that particular territory, the number of farms, etc.

Mr. JACOBSTEIN. When would he do that?

Mr. GOSNELL. Immediately after the completion of the enumeration.

Mr. JACOBSTEIN. So that if he did discover any errors there would be very little opportunity for correction?

Mr. GOSNELL. Yes, there is plenty of opportunity for correction, because the supervisor discovers that before he sends the report to Washington.

Mr. JACOBSTEIN. Do you mean he orders a new census of that particular district to be taken?

Mr. GOSNELL. Yes. Many times we have enumerated districts twice.

Mr. JACOBSTEIN. How many second enumerations were made in the last census?

Mr. GOSNELL. I could not say exactly, but I should say somewhat less than 500.

Mr. JACOBSTEIN. Out of a total of how many?

Mr. GOSNELL. Eight thousand or nine thousand.

Mr. THURSTON. Is it not a common practice on the part of the various chambers of commerce to complain that the census has not enumerated all the people in their town or village, and that there should be a recount?

Mr. GOSNELL. That is true in many cases. It is not as true as as it was 15 or 20 years ago. We find to-day that the chambers of commerce are more interested in getting an accurate census of the population than they are in "boosting" the population of their particular cities. We certainly find that to be the case in centers where chambers of commerce play an important part.

Mr. THURSTON. There is a decided improvement in that respect? Mr. GOSNELL. There is a decided improvement in that respect. Mr. MOORMAN. These enumerators are not paid until the supervisor sees that the blanks are properly filled out, etc.?

Mr. GOSNELL. That is true. Here is another feature of the system: The enumerators are furnished a sheet showing the population of that particular district at the last census, and also the number of farms in that district at the previous census.

Mr. THURSTON. You also check by reports from the Department of Agriculture, do you not?

Mr. GOSNELL. I am not sure as to that.

Doctor HILL. There was a question raised about this, and I think it is appropriate to say at this time that the Director of the Census (and I think it is a very good idea), instead of waiting until all the schedules have come to Washington and we have been over them before announcing the population, intends to have the supervisor himself announce the population in his district when the canvass is completed, so that if there are any "kicks" or complaints, they will come at a time when the organization is still there.

Mr. JACOBSTEIN. That is what I had in mind. There would be no point in telling us that there are errors in these statistics after it was all over and the organization had been disbanded. We want to know about it in time to be able to correct it, while the organization is still there.

Doctor HILL. That is the idea.

Mr. JACOBSTEIN. May I ask a question of the field supervisor? I have not gone into this matter carefully; but Germany and England