'^HO 



oo{ 



ANNUAL EEPOrtTS OF DEPAHT.MEXT OF AGEICULTUEE. 



results will be available before the final estimates of acreage and pro- 

 tliiction are issued for publication. When so checked for complete- 

 ness and correctness, assessors' returns of agricultural data, which 

 have no connection whatever with taxation, form an annual census of 

 the greatest value. During the winter and spring months of 1918-19 

 a number of State legislatures amended old laws or passed new laws 

 providing for the collection of agricultural data b}^ assessors. The 

 States which now have such laws are shown in the shaded portions 

 of figure 2. It is ho}>ed that all States will join in this movement to 

 ascertain accurately the agricultural resources of each county and 

 S(ate annually. 



ACCURACY OF THE CROP ESTIMATES. 



Tlie value of the crop and live-stock estimates depends upon their 

 accuracy and timeliness, and the amount of detail. The degree of 

 accuracy of the estimates can be determined only approximately for 

 most crops, by checking against such figures as are available for 

 crop movement, receipts at primary markets, and exports. In the 

 case of the cotton crop, however, an absolute check is afforded by 

 the annual census of the number of bales ginned. Tlie following 

 table shows the annual estimates by this bureau in December in 

 comparison with the annual report of bales ginned by the Bureau 

 of the Census the following March : 



Annual cotton estimates of the Bureau of Crop Estimates, compared loith annual 



Census reports of cotton ginned. 



It Avill be noted that the bureau estimated the cotton crop of 1915 

 within three-tenths of 1 per cent, the crop of ,1916 within five-tenths 

 of 1 per cent, the crop of 1917 within 3.1 per cent, and the crop of 



