34 BULLETIN 1224, U. S, DEPARTMENT OF AGRICULTURE. 
and sellers of land project the past increases in rents into the future, 
and capitalize them into a present value. 
But it may be asked what part of the past period is thus projected 
into the future? Table 11 shows that the longer the past period 
which is included in making up the average, the lower the average 
increases. 
What period, then, has the most influence in determining the buy- 
ers' and sellers' estimates of the future ? Light is thrown on this ques- 
tion by comparing the calculated i's with the average increases of 
the different past periods. In the table, average increases in cash 
rents of the 5, 6, 7, 8, 9, and 10 preceding years are shown. The 
past period which shows an average increase in cash rents most 
closely approximating the calculated i for each area is marked with 
an asterisk (*). It is then seen at once that the calculated -i's, the i's 
which were necessary to justify the valuations of 1920, are approxi- 
mately equivalent to the average increases in cash rents of the pre- 
ceding 6 or 7 years in every area except Minnesota and Iowa. 
In Iowa the calculated i corresponds with the average increase in 
rents of the 9 preceding years, while in Minnesota the calculated i 
agrees with the average increases in rent for a much shorter past 
period than in the other areas. 
Can it be said from the evidence presented in this table that the 
increases of a certain definite past period are the basis for future 
forecasts of increases in land income ? As a general proposition, such 
a conclusion can not be drawn. There are too many other influences 
which may bring about expected increases in income. Such influ- 
ences are: (1) prospective shifts in the type of farming from a less 
intensive to a more intensive type; (2) prospective improvements 
in transportation, such as public road improvement, or the construc- 
tion of new or better railway facilities; (3) prospective improvement 
in the land itself through drainage or clearing. 
Such improvements increase the earning power of land, and if 
foreseen would give the basis for an expected increase in land earning 
power and a higher present valuation. In areas where such improve- 
ments are anticipated there would not be so close a correspondence 
between the calculated i's and the average increase in rents of the 
preceding years. 
In the new agricultural regions expected increases in income are 
in the main probably based on prospective improvements of the 
kind mentioned above. But in old agricultural areas which consti- 
tute the major farming regions of this country the expected changes 
in land income are based in the main on average changes in the past. 
The areas included in. Table 11 are for the most part old farming 
areas where no improvements of the type mentioned are expected 
for the areas as a whole. Small sections of these areas no doubt have 
their land values based upon such expected improvements, but 
not the areas as a whole, with the possible exception of the Minnesota 
area. 
In the northern counties of the Minnesota area a relatively large 
part of the land is reported by the census as needing clearing, drain- 
age, or both. It is probable that the value is based upon the expected 
earning power when these improvements are made. This would 
account for the fact that the calculated i in this area corresponds to 
the average increase in rent of a much shorter preceding period than 
