RELATION OF LAND INCOME TO LAND VALUE. 23 
It is clear, then, that the ratios of rent to value are highly con- 
sistent from year to year, even as revealed by data which are not 
entirely adequate. One thing more, however, is to be noted. All 
the ratios drop from 1919 to 1920. The greatest decrease occurred 
in Iowa and the least in Wisconsin. If the decline had occurred 
only in Iowa it might have been suspected that it was due to a bias 
in the estimates of value for that year. The percentage increase in 
value from 1919 to 1920 was much greater than for any other pair 
of years in this 10-year period. Furthermore, the average land value 
for Iowa, as estimated by the Crop Reporting Board, was 823 per 
acre higher than that estimated by the census. That this decline in 
the ratio of rent to value in 1920 was due to an upward bias in the 
estimated values of 1920 is controverted by the fact that the same 
decline in ratios is found in both Wisconsin and Ohio, where the land 
values are based on actual sales of land and both the latter areas 
showed relatively great increases in land value from 1919 to 1920. 
The evidence seems to indicate clearly that the ratios of rent t< 
value in 1920 were below their normal trend. It is probable, then, 
that the ratios of rent to value in 1920 are not as typical of the norma' 
trend in these ratios as the ratios of 1917, 1918, or 1919, would have 
been. If normality, then, be denned as conformity to trend, the 
ratios of rent to value in 1920 were below normal. But it is necessary 
to go beyond this and find out why these ratios were below the normal 
trend in this year. It may have been due to a lag in cash rents in 
1920 behind the rents imputed to lands by buyers and sellers, or to 
the fact that in the excitement of the land boom buyers may have 
offered prices for land that bore little relation to its earning power, 
or to the fact that a greater amount of anticipated increases in income 
was projected into the future at this time. 
Since in most areas cash rent showed a greater increase from 1919 
to 1920 than in any other year it is believed that there was no lag 
in cash rents. The explanation, then, lies between the other two 
causes and is in all probability partly explained by both. Given 
the rate of capitalization, the value of land is determined by the 
income it yields and the most probable future changes in that income. 
The most probable changes in income are estimated by buyers and 
sellers of land, Normally these estimates are not wild guesses but 
are based definitely on the past experience of buyers and sellers of 
land with respect to changes in land income. Since the experience 
of the past is a changing one, each year will bring forth varying 
estimates of the future. For instance, in 1920 the average increases 
in incomes in the preceding years indicated a greater future increase 
in income than had been shown in the past for any other year pre- 
ceding 1920. (See Table 11). Therefore, in 1920 a greater future 
increase in income was naively anticipated, and this made the ratios 
of rent to value fall below their trend for this year. It is also likely 
that the farmers who bought and sold land, instead of being skeptical 
of the soundness of the high land incomes prevailing in 1920 and the 
preceding years were made unduly optimistic by these high incomes 
and the great increases during the war years, so that they were willing 
to pay higher prices for land than they would normally have felt 
justified in doing. This again would tend to lower the ratios of rent 
to value in 1920 below their trend. The latter cause is belies ed to 
be of less importance than the first one. The relatively great in- 
