BUREAU OF CROP ESTIMATES. ' 331 



1918 within 2.7 per cent of the final ginning figures. The wider 

 deviation in the estimate of the 1917 crop was due to the price 

 of cotton in the winter of 1917-18, which made it profitable to pick 

 closer than had ever been done before, the picking continuing as late 

 as February and March in some States. The underestimate of 1918 

 was due to the comparatively mild winter, which made it possible 

 for more bolls to mature than could be foreseen by the growers, and 

 prevailing prices made it profitable to pick all that opened. The 

 practice of harvesting " bolly " cotton i. c, unopened bolls has 

 gradually extended throughout the cotton belt because of the demand 

 for cotton and relatively high prices. 



In 1918 for the first tiuie the bureau had a fairly complete check 

 on its estimates of wheat production, afforded by the records kept 

 by the United States Grain Corporation. At the close of the fiscal 

 vear the Grain Corporation had reported approximately 730,000,000 

 bushels of wheat received from farms. The 1918 crop as estimated 

 by the bureau was 917,000,000 bushels, with 8,000,000 on hand July 

 1, 1918, from the previous crop, making a total of 925,000,000 bushels 

 available. Of this amount approximately 105,000,000 bushels were 

 used for seed, 25,000,000 bushels of damaged or inferior grain is nor- 

 mally fed to poultry and other live stock, and after deducting the 

 730,000,000 reported by the Grain Corporation as marketed from 

 farms, there is left 65,000,000 bushels to be accounted for. It should 

 be remeinbered that the Grain Corporation gets no reports of wheat 

 ground in mills of small capacity, which includes many small country 

 grist mills in every State in which wheat is gro\\Ti and consumed 

 locally, the annual capacity of which is far in excess of 65,000,000 

 bushels. Apparently the bureau underestimated the 1918 wheat 

 crop slightly, perhaps as much as 2 per cent. In all probability the 

 estimates of the United States wheat crop were about as accurate 

 as an agricultural census by ordinary methods. 



DEMANDS FOR IMPROVED CROP AND LIVE STOCK REPORTING 



SERVICE. 



The war emergency not only demonstrated the efficiency and the 

 value of the crop-reporting service but also showed the need for im- 

 provement to meet continuing demands upon the service, which can 

 only be partially met with the present force and funds. These de- 

 mands ma}^ be classified as follows: 



(1) The demand for estimates and forecasts of the consumption, 

 market, export, and import requirements and available surplus of 

 agricultural products, not only in this country but in foreign coun- 

 tries as well. This information is needed as "a basis for intelligent 

 planting and marketing programs, the importance of which was fully 

 demonstrated during the war and will continue to be of equal impor- 

 tance in the years to come, as population increases without a corre- 

 sponding increase in arable land in this country, and as competition 

 develops with foreign countries. 



(2) The demand for information in advance of planting time or 

 while planting is in progress as to acreage to be planted to crops, in 

 order that any prospective surplus or deficiency may be equalized by 

 modifying later plantings in accordance with prospective supply and 

 demand. Estimates of farmers' intention to plant were made in 



