36 BULLETIN 1224, U. S. DEPARTMENT OF AGRICULTURE. 
Table 12 shows the relation in Iowa between the calculated i's and 
the average increase in rents of the preceding years for each of the 
f)ast 11 years, exclusive of 1911 for which no data were available on 
and values. 
It will be noted that the anticipated increases in income as calcu- 
lated by the formula gradually increased from 1910 to 1920, and a 
comparison of the calculated -i's with the average increases in rent of 
the preceding years explains why a constantly greater future increase 
was anticipated and capitalized into the value of each succeeding year. 
The calculated i's and the average increases in rent of the preceding 
years move up together. The anticipated increases in income in each 
of these years were based upon the experience of the past; that is, 
these anticipated increases changed each year because of a changing 
past experience, just as the anticipated increases were found to be 
different from area to area because of varying past experiences in 
those areas with regard to increases in income. 
Since the calculated i's in Table 12 are based upon gross rent and 
on land values which have an upward bias, they can not be compared 
with the average increases in rents for different past periods to show 
how long a past period was most influential in determining estimates 
of the future. The calculated i's correspond to the average rent 
increases of the preceding 12 to 15 years. This can not be taken to 
mean, however, that the experience of the past 12 to 15 years has 
been most influential in determining the forecast of the future for any 
given year. The average rents are gross rents, and when used in the 
formula to calculate i give a value for it that is too low. Table 11 
shows that buyers and sellers of land, in 1920, made up their forecasts 
of the future on the basis of changes in land income during the past 
7 to 10 years. Table 12 shows that i calculated on the gross-rent 
basis in 1920 corresponded to the average increase in rent of the past 
14 to 15 years, or a longer period than on a net-rent basis. Now, 
since i calculated on the gross-net basis corresponds in each year from 
1913 to 1920 to the average increases in rent of the preceding 12 to 15 
years, it seems highly probable that buyers and sellers of land were 
influenced by about the same past period in all other years in making 
their forecasts of the future as they were in 1920; that is, on the net- 
rent basis about 7 to 10 years. To calculate this accurately, changes 
in taxes relative to rents would have to be known. 
Buyers and sellers of land have based the prices at which they 
were willing to buy or sell on past experience, but naively so. In 
1920 they projected the average increase in income of the preced- 
ing 7 to 10 years into the future just as they had done in 1910, 
regardless of the fact that their 1920 forecasts were based upon 
the abnormal experience of the war period. The forecasts of this 
generation of buyers in all the years preceding 1920 had apparently 
been justified, for the buyers in any year had found that land incomes 
actually increased faster in the following years than had been 
anticipated at the time of their purchases, so that buyers in 1920, 
and in the other years immediately preceding it, were not led to 
question the normality of the conditions that produced the high 
incomes of these years or the relatively great increases of the pre- 
ceding years. Therefore, in 1920 they did what they had always 
done, and made up their forecasts of the future on a past experience 
which was uncritically accepted as a safe guide. 
