INPUT AS RELATED TO OUTPUT 23 
Similar tables could be presented to show the variations in inputs 
per unit of output for other cost items — labor, interest, use of build- 
ings — as different feeding practices were followed. 
WHEAT PRODUCTION ANALYSIS 
Since most of the operations performed on the wheat farms were 
performed only one, two, or three times, the analysis of the data 
did not reveal the operation of the law of diminishing returns from 
successive units in the same way as did the data for operations on 
potatoes and feed fed to beef cattle. The analysis already made 
of reasons for variations in the input of labor per acre showed that 
increased labor was due principally to using more labor on each 
operation rather than to more operations. Table 22 shows no ap- 
parent increase in yield with additional hours per acre. 13 When the 
effect of number of cultural operations was eliminated by partial 
correlation methods, the net correlation was practically zero. Ac- 
cordingly, the tentative conclusion may be reached that the higher 
labor inputs per operation were due solely to variations in char- 
acter of the soil, in machinery and equipment used, and in efficiency 
and composition of the labor, and to errors of estimate by the 
farmer; and did not represent more thorough tillage; that is, 
better performance of each operation. 14 
Table 22. — Average yields of wlxeat by input of man labor per acre 
Hours per acre 
Number 
of farms 
Average 
yields 
3.0to3.4 - .- 
11 
13 
14 
Bushels. 
8.35 
3.5 to 3.9 
8.85 
4.0 to 4.4. „ 
8 35 
4.5to4.9-._ 
15 8. 11 
5.0 to 5.9 
5 8.02 
6.0 to 6.9 
4 
The relative effectiveness of labor in the different cultural oper- 
ations, however, can be determined. Table 23 shows the average 
input of labor per bushel of wheat for several different operations. 
Harrowing seems to be a much more efficient way of using labor 
than disking, as far as the average effect goes. Although there were 
wide variations in labor input per operation, there was not very 
much variation in the number of times the different operations were 
performed. For this reason these data could not show the effect 
of various individual changes in culture — the number of cases was 
too small for reliable results. In an area where there were wider 
variations in cultural practices, it would be possible to determine 
the gains from individual additional operations of each kind. Thus 
disking once might be a more efficient use of labor than harrowing 
the third or fourth time. Where records were available covering 
a sufficient variety of practices, the same marginal analysis as has 
been used for potatoes and beef could be applied ; that is, curves could 
be drawn showing the net effect on output of each specific increase 
in input. 
13 The gross correlation between yield and hours of labor is only r= + 0.038 ±0.121. 
u Or if higher labor input did represent more thorough tillage, this better tillage had no 
measurable effect on yields. 
