RELATION OF LAXD IX CO ME TO LAND VALUE. 25 
used for direct comparisons with Figure 3. That is, a comparison of 
the ratios in Figure 3 with those in Figure 5 for the same areas can 
not be made the basis for a statement that the ratios of rent to value 
were higher in 1921 and 1922 than in 1920. Figure 5 certainly shows 
higher ratios throughout the country than Figure 3 does; but if the 
Crop Reporting Board had made estimates of cash rent and land value 
in 1920 would the ratios based upon these estimates have been lower 
than the ratios based upon the estimates of 1921 and 1922 ) It is 
likely that they would have been slightly lower, but it is improbable 
that the difference would have been as great as the difference in the 
ratios based on census data and those based upon the Crop Reporting 
Board estimates for 1921 and 1922. In other words, the cen-us data 
and the Crop Reporting Board data were obtained in different ways 
and are not directly comparable. 
Although Figure 5 can not be used to show the change in the ration 
of rent to values since 1920. it does bring out the fact that the varia- 
tions in the ratios as between different sections of the country were 
not peculiar to 1920. All the major differences brought out in Figure 
3 are also brought out by Figure 5. In both maps the South and the 
irrigated regions of the West stand out as areas where the gross rate 
of return on land value is highest. Both maps (figs. 3 and 5) bring 
out the Corn Belt as the area having the lowest per cent of return. 
Their outstanding similarities are obvious at a glance, but the two 
maps are alike even in smaller details. The eastern part of the Corn 
Belt (Ohio and Indiana,) shows higher ratios than the western part in 
both maps. The high and low areas in Minnesota, Iowa, and Illinois 
are about the same in each map. The Red River Valley in Minnesota 
and Xorth Dakota shows higher ratios than the region just below it = 
Both maps show the Xew England States to be relatively high and 
variable. Where the two maps do not show the same relative differ- 
ences in the ratios of rent to value from area to area, the lack of simi- 
larity can usually be explained by the fact that the two maps were not 
constructed exactly alike^ In Figure 3, based upon census data, the 
county is the unit, so that variations by counties are shown. In 
Figure 5, on the other hand, the crop-reporting district, which is 
usually made up of several counties, is the unit for which average 
rents and average values are reported. 
This difference in the bases of the two maps results in some con- 
siderable apparent changes, from 1920 to 1921 and 1922. in the ratios 
of rent to value in some of the irrigated regions of the West relative 
to other parts of the country. Imperial County, Calif., and Maricopa 
County, Ariz., for instance, show a gross rate of return in 1920 that 
is relatively high, but in 1921 and 1922 both these counties appar- 
ently show a rate of return that is relatively low. The two maps, 
however, are not comparable for these areas, because of their different 
bases of construction. The census data are based upon rented 
farms, the great majority of which were in the irrigated regions of 
these two counties. The data of the Crop Reporting Board are 
simple averages of estimates for a region including several counties. 
Maricopa County was therefore only one of several counties included 
in the average for this district, and the same is true o\ Imperial 
County. The regional variations in the ratios of rent to value found 
in 1920 were not due to erratic fluctuations in either rents or values 
