2 BULLETIN 1433 ? U. S. DEPARTMENT OF AGRICULTURE 
the relative importance of tenancy in the North Central States would 
be shown to be even greater. 
The tenant-landlord situation in this region is also distinctive, as 
compared, for instance, with the South, by reason of the extent to 
which tenancy is a stage in the so-called agricultural ladder, the extent 
to which it is a method of farming deliberately chosen for economic 
reasons, the large proportion of cases of relationship by blood or 
marriage between landlord and tenant, and other characteristics. 
The present study was undertaken with a view to throwing as 
much light as possible on such questions as the following: 
To what extent is land ownership concentrated and has the degree of concen- 
tration been increasing or decreasing? 
Where do the landlords live? Are they sufficiently near their farms to be able 
to give them their personal interest and attention, or are they distant absentees? 
What are the personal characteristics of these landlords? How old are they? 
How are they- occupied? What previous farming experience have they had? 
To what extent are they related by blood or marriage to their tenants? 
How did they acquire their farms? 
To what extent do they contribute personal supervision and advice? 
To what extent, in the opinion of the landlords, are the rented farms decreasing 
in fertility? 
METHODS AND AREAS SELECTED FOR STUDY 
A study of the ownership of rented farms was made in connection 
with the census of 1900. 3 Using methods similar to those employed 
by the Bureau of the Census in indexing the ownership of rented 
farm land in 1900, the Division of Land Economics and Land Utiliza- 
tion of the Bureau of Agricultural Economics has indexed the owner- 
ship of rented land located in 85 counties of the North Central States 
so far as such ownership could be ascertained from schedules for the 
rented farms enumerated in connection with the census of 1920. 
From each schedule for a rented farm, the total and the improved 
acreages, the value of the real estate, the State and county in which 
the farm was located, and the name and address of the owner were 
transferred to a slip of paper. These slips were filed by county and 
by town in which the owner resided, and the owners resident in each 
town were indexed alphabetically. In this way, and to the extent 
that the name and address of the owners were ascertainable, all of the 
farms which any one landlord owned in the counties studied were 
brought together. 4 
In connection with the study made of conditions in the 85 selected 
counties in the North Central States, a study was made at the same 
time along similar lines of conditions in 99 counties located in other 
representative parts of the United States (fig. 1). The results of the 
study of all 184 counties are published in Department Bulletin 1432, 
in which it was possible to go into more detail in comparing condi- 
3 The results are published in summary form in vol. 5 of the Twelfth (1900) Census. 
4 Tabulations presented in connection with the census of 1900 show that 5.9 per cent of the rented farms in 
the North Central States were not indexed as to ownership by that census. Some duplication arises where 
a tenant reports that he rents parcels from two or more landowners, for unless an index card was prepared 
in the joint name of the owners one would need to be prepared for each owner. The former procedure 
would fail to bring all the properties of any given owner together and the latter procedure would increase 
the apparent number of farms indexed. If such duplication were not allowed for in connection with the 
index to ownership of rented farms in the 85 counties studied as to ownership in 1920 the apparent percentage 
of farms not indexed would be 7.8, whereas, making allowance for such duplication, it appears that 11.6 
per cent of the rented farms were not indexed for ownership. The number of rented farms in the 85 counties 
of the North Central States studied, as reported by the census of 1920, was 92,728. When the ownership 
of 81,949 of these farms was indexed it was found necessary to recognize them as consisting of 85,472 parcels, 
of which 92 per cent were complete farms in themselves. 
The population schedules were referred to in doubtful cases of ownership in preparing the index for the 
census of 1900, but this practice was not resorted to when the 1920 index was prepared. 
