Might I say in connection with that A. A. A. data, it is my understanding that in most cases the reports of individual farmers were checked with threshers' records and other sources of information, and also I believe that the information was published, so that the neighbors might see what farmers near them had reported.

The records of the Crop Reporting Board go back, of course, much farther than 1930, and in the case of several States, notably Kansas and North Dakota, there are records further back, for a considerable number of years.

One of the biggest difficulties has been that the Crop Reporting Board in its work has used yields per harvested acre. Naturally, they need that for calculating production. In the case of crop insurance it is necessary to have yields per seeded acre, and I think this point needs emphasis, because I have the impression from testimony I have heard that the estimates as to yields that most persons have in mind are estimates per harvested acre rather than per seeded acre, and per seeded acre yields are substantially lower.

I have a few tables here that might be useful. I have a table here which shows the average yield per harvested acre for various periods for the States in the Great Plains region, and for one or two other States, and in that table I have shown the average for the 25-year period 1911 to 1935, the 15-year period 1921 to 1935, the 10-year period 1926 to 1935, the 8-year period 1928 to 1935, the 5-year period 1929 to 1932, and the 6-year period, the base period that we have used, 1930 to 1935. These, however, are yields per harvested acre, not per seeded acre, and are computed from published figures of the Department.

Senator MCGILL. What States do you include in those figures?
Mr. Rowe. I have Kansas here, Senator.

Senator MCGILL. But what others? Kansas is always included. Mr. Rowe. North Dakota, South Dakota, and Nebraska, Montana, Texas, Oklahoma, Washington, Oregon and Ohio. And I would be glad to put this in the record.

Senator POPE. You might identify that table by calling it table 1, illustrating your testimony.

(The table referred to follows:)

TABLE 1.-Average yields of wheat per harvested acre for various periods in specified

States 1

1 Based on annual yield data of the Crop Reporting Board, Bureau of Agricultural Economics, U. S. Department of Agriculture.

Senator MCGILL. Now, there are other States that produce quite a quantity of wheat besides those. Why are they left out?

Mr. Rowe. These were taken just as illustrative, as the principal States. We had the same data for the other States.

Senator MCGILL. You have it for all the States?

Mr. RowE. The yield from harvested acres we have for all the States. This is the data prepared by the Crop Reporting Board. But, as I say, this is per harvested acre, and is of limited use in the crop-insurance work.

Senator POPE. We will make this table part of the record.

Mr. RowE. Now, the thought we have in mind with regard to adjustments is to take the data prepared by the Crop Reporting Board for counties over a longer period, and with that data to adjust the individual farm data for the 6-year period. For instance, under the plan as worked up with the 6-year data for individual farms in one of the counties in North Dakota, the average insurance coverage would have been 3%1⁄2 bushels. If the 10-year period, 1926 to 1935, were used and the original data corrected for that 10-year period, the coverage would have been 5.06 bushels.

Senator POPE. What was that first figure for the 6-year period? Mr. RowE. 3.52 bushels. And for the 10-year period 5.06 bushels. That would be a substantial increase. I am inclined to think that such adjustments would be desirable, because I believe, particularly in North Dakota and other States of the principal spring-wheat area, that the 6-year period is a very bad period and not a fair period on which to base the insurance program.

Senator POPE. You do not have figures for a 15-year period?

Mr. RowE. Yes; I have the figures for a 15-year period as well. In that county the figure for the 10-year period would be 5.88 bushels. Senator MCGILL. What do you think would be the period that should be used?

Mr. RowE. I would offhand be inclined to think that the 10-year period would be the desirable period, at least would be a satisfactory period.

Senator POPE. It is interesting to note that your 10-year average and your 15-year average is almost the same.

Mr. RowE. Yes; that is very true. At least it is true in that

county.

Senator POPE. How does it run in the other counties that you have checked?

Mr. Rowe. Well, in North Dakota, in one of the other counties the 10-year period 5.22; 15-year period, 5.50. In still another county the 10-year period is 5.34 and the 15-year period 6.45.

Now, coming back to the first county I referred to, the 6-year figure is 3.67; then as I gave you, the 10-year figure is 5.22, 15-year, 5.50.

For the second county the 6-year figure is 2.38 bushels; 10-year figure, 5.34. That is a substantial increase, and the 15-year figure is slightly more, 6.45.

Senator POPE. Those figures would indicate that the 10-year period would be fairer than the 6-year period, because in some of those figures the 10-year average would be twice what the 6-year average would be.

Mr. RowE. Yes.

Senator MCGILL. Was that 6-year average affected by the drought which they have been having in that territory?

Mr. RowE. Very severely.

Senator MCGILL. So that would not be a fair criterion to go by in that section?

Mr. Rowe. I do not believe so. I might add, however, that if you increase the coverage above the 6-year period it is necessary also, in order to collect sufficient premiums to cover this higher yield, to increase or adjust the premium rate as well.

The comparable figures for the premium rate are as follows: In the first county that I gave you the premium under the 6-year plan was 1.25, or a bushel and a quarter; under the 10-year plan it would be 1.58; under the 15-year plan it would be 1.63.

For the second county, under the 6-year plan as we have figured it out it would be 1.09; under the 10-year plan 1.82, and under the 15year plan, 1.65.

Senator POPE. Well, you mean that the premiums would be lower, based on the 6-year period than they would be on the 10- or 15-year period?

Mr. RowE. That is right.

Senator MCGILL. Your average production is less?

Mr. RowE. The average production is less. However, in making this adjustment we have found some counties where if you lengthen the base period you increase the coverage, but you decrease the premium cost.

Senator MCGILL. That 6-year period, due to abnormal conditions which have prevailed in the last 2 years, is not a basis on which you can arrive at the average production per acre in that territory? Mr. Rowe. It is not a representative period.

Senator MCGILL. Not representative at all. And that is true in Nebraska, and true in Kansas, is it not?

Mr. RowE. Yes.

Senator MCGILL. It is true in virtually all of the Middle Western States?

Mr. RowE. Yes. In some Kansas counties however, to lengthen the period would reduce somewhat the coverage.

Senator MCGILL. Yes, that is true because of the extremely heavy crops produced in 1931.

to?

Mr. RowE. That is right.

Senator POPE. In what States were these counties that you referred

Mr. RowE. The counties that I gave you were in North Dakota. Do you care for the name of the county?

Senator POPE. No. Did you make any calculation as to the counties in Ohio or other States where the drought did not prevail?

Mr. Rowe. I did not have the 10-year and the 15-year figures in Ohio, as we had in North Dakota, because in North Dakota these 10-year and 15-year yields per seeded acre were obtained from assessors' data. In Ohio, however, we do have an 8-year period which we have obtained by splicing together two series, our 6-year and a 5-year series that was prepared in connection with the A. A. A. work from 1928 to 1932. Splicing those two series together we have an 8-year period.

Take in Ohio, for instance, if we took the 6-year period, the original 6-year period, the coverage for one of the counties there would be 16.85 bushels. If we took the 8-year period, 1928 to 1935, that coverage would be reduced somewhat to 14.66.

Now also for the 6-year period, the losses per acre in that county were 0.49, or about half a bushel. If you took the 8-year period to which I have reference, the average losses per acre were over 2

bushels. So it would appear that if you extended the base period you would lower somewhat the coverage and increase the premium charge that would be necessary, and I believe that for a system to operate satisfactorily you would want to stretch out this 6-year period by the use of other data.

Senator POPE. And you said a minute ago, I believe, that you thought a 10-year period might be fair all the way through?

Mr. Rowe. I am inclined to believe so. Work is now in progress to obtain county estimates of yields per seeded acre for the 10-year period 1926 through 1935. I believe that this program in the long run should aim to increase the length of the base period for individual farms as data for subsequent years are gathered, so that ultimately you might be able to operate with a history on the individual farm for a 10-year period, and at that time I do not believe you will need the adjustments that we are speaking about here.

Senator POPE. Yes, as you gain experience.

Mr. RowE. I should be glad to submit a few figures from this table, Senator, if you think it is necessary.

Senator POPE. I think it would be helpful.

TABLE 2.-Illustration of average insurance coverage and average annual crop losses, by counties, for base periods of varying length

Mr. RowE. I want to call attention to a point with regard to the subject of hail insurance that was mentioned earlier. Hail insurance is damage insurance rather than yield insurance, and the basis for adjustment of losses is quite different.

Senator MCGILL. They insure so many dollars per acre, for instance, on a wheat crop.

Mr. RowE. Yes.

Mr. MCGILL. And then if there is damage they estimate the percent of the damage, and the loss is paid according to the policy of so many dollars per acre.

Mr. RowE. Right.

Senator MCGILL. It is altogether a different proposition.

Mr. RowE. It is a different proposition entirely than the proposal for yield insurance.

Senator POPE. Do you have anything further you want to offer, Mr. Rowe?

Mr. RowE. I believe not.

Senator MCGILL. Do you have a table of figures made up there that would be beneficial to put into the record?

Mr. RowE. I have a large number of figures in this table. I should be glad to submit a condensed table with some of these figures for the record.

Senator McGILL. I think it would be well to do that.

Senator POPE. Yes, if you will do that we will put them into the final record.

At this point I will read into the record a letter which I received from Mr. M. R. Devaney, president of the Northwest Country Elevator Association. He had expected to be here to give some testimony but was unable to do so.

Senator McGILL. Where is that from?

Senator POPE. From Minneapolis. As I indicated, the man who writes the letter is president of the Northwest Country Elevator Association, and the letter is addressed to me, dated March 2, 1937:

In connection with Senate bill (S. 1397) to create a Federal Crop Insurance Corporation and for other purposes, this association wishes to place itself on record as desirous of cooperating in the plan to store the grain taken in as insurance premiums at country points in the States of Montana, North Dakota, South Dakota, and Minnesota.

The association endorses paragraph (d) section 7 of the bill, which reads as follows: "Wheat acquired under the provisions of this Act shall be kept in Federally bonded or State licensed warehouses, or in such other manner as the Board determines will adequately protect the interests of the corporation and the producers insured.

Our members own and operate country grain elevators in these States, approximately in the following number of stations: Montana, 150; North Dakota, 600; South Dakota, 150; Minnesota, 300.

Our association cannot at this time offer a reduced rate of storage on wheat because the laws of the States in which we are operating prescribe the rates which will be charged to all storers of grain. Should State laws be amended to permit the granting of reduced rates of storage to the Government, we would willingly. grant such reduced rates.

We can offer protection to the Government as grain warehousemen operating under license and bonds which protect the owners of stored grain. The members of our association are reputable firms of long experience and of financial stability and their country elevators are efficiently operated by experienced managers.

We invite your attention to the fact that there are approximately the following number of country grain elevators now in the four States above named:

These elevators have a storage capacity of approximately 125,000,000 bushels of grain. The terminals to which grain produced in these States is shipped are Minneapolis and Duluth, Minn. and the storage capacity at these terminals is approximately as follows:

Minneapolis..
Duluth...

92, 000, 000
50, 000, 000

We offer this information because we believe that the present facilities are adequate to provide storage space to meet all the Government needs in the States of the Northwest and it is our opinion that it should be unnecessary for the Government or any other parties to undertake the construction of new grain storage facilities within this territory.

We desire this letter to be made of record and will furnish you any additional information desired in connection with the establishment of country storage points for crop insurance grain, if you will advise us.

Very truly yours,

(Signed) M. R. DEVANEY,

President.