RELATION OF LAND INCOME TO LAND VALUE. 
among the colored tenants, and the explanation is not to be found 
in the personal interest of the landlord in the welfare of an old tenant. 
The higher rents paid by the shorter-occupancy groups are in the 
main apparent rather than real. Landlords make arrangements with 
old tenants whereby these tenants keep the buildings and fences in 
repair and sometimes pay the taxes, and for this they are given a 
reduction in rent. But the differences in the rents of the various 
occupancy classes are not ail apparent. In leasing a farm to a new 
tenant the landlord incurs a certain amount of risk. A new tenant 
may be a poor one. He may use a farm in such a way that it will 
depreciate rapidly under his operation, so that when he leaves it the 
owner will have to lease it at a lower rent than before or restore it to 
its former state of productiveness at considerable cost. The land- 
lord therefore charges the new tenant a higher rent to cover the ele- 
ment of risk. If this new tenant proves to be a satisfactory one, the 
landlord does not raise the rent the following year. and. if he con- 
tinues to prove himself a dependable tenant, the landlord will continue 
to shut more and more of the responsibility of taking care of the farm 
to him, and for this the landlord charges him a lower rent. 
If it were known how much of the landlord's expenses are allowed 
for in the lower rents paid by the old tenants, it would be better to 
use the average rents paid by them as a measure of farm rents. But 
there is no way of determining this. Varying proportions of the bur- 
dens of taking care of a farm are shifted to the tenant, depending 
upon the length of his tenure, his character, and ability. It seems 
better to base the average cash rents on the rents paid by the one- 
year tenants, even though a small proportion of these rents are pay- 
ments for the risk a landlord assumes in leasing to a new and untried 
tenant. Furthermore, by using the rents paid by this group, all 
the cases where long contracts exist, made several years previous to 
the time the census was taken, are excluded. The average rents 
paid by the one-year tenants are the most free from other influences. 
What they are and what they are not is better known, so that it is 
advisable to use them, making a full deduction for the landlord's 
expenses, than to use the rents paid by the old tenants on the ground 
that a part — without knowing what part — of the landlord's expenses 
have been allowed for. The cash rents paid by the one-year white 
tenants are. then, the proper data to use for a measure of farm land 
income or farm rent. 
EFFECT OF INADEQUATE EMPROVEMENT ON CASH RENTS. 
Even these data may not always accurately represent average 
farm rents. Farm rent, it will be remembered, is the annual income 
of a farm when it is properly improved: that is. when it is improved 
according to the average standards of the community in which it is 
located, with a type of improvement that will enable its operator 
to carry on the type of farming for which the farms in that com- 
munity are best adapted. For instance, a farm in a dairy community 
which is not equipped with buildings adequate for dairy purposes 
would be underimproved. 
Land will normally sell for its capitalized farm rent plus the capi- 
talization of the anticipated increases in farm rent. The cash rent 
on a farm will normally equal the farm rent if the farm is improved 
according to the average standards of the community. But if a 
