RELATION BETWEEN RENTS AND INCOMES 83 



will be shown, are a measure of the incomes of their occupants, are 

 comparatively readily collected and classified. Rent data obtained 

 in the course of a commercial survey show the "character" of a city 

 and are used as a basis for estimates of the future residence telephone 

 market. 



In view of the importance of these rent data, it seems desirable 

 to study them in some detail to find out just what their limitations 

 are. 



There may be set down in advance certain things which it is desirable 

 to know, as, for instance, the relationship betw-een money incomes 

 and house rents and methods and limitations of comparison of different 

 cities on the basis of rental values. On the first point, as applied to 

 any particular city, a knowledge of the relation between incomes and 

 rents is desirable in a general way, although there is no necessity to 

 translate the telephone market expressed in terms of rent types into 

 a scale of incomes. On the second point, the comparison of different 

 cities, it should be ideally possible by a study of rent data to compare 

 the inherent economic markets for telephone service, and also to 

 measure differences in the strength of the telephone habit, but in 

 practice only rough approximations may be made. 



Certain limitations to work of this kind are fairly evident. The 

 most obvious difficulty is the fact that rent levels have changed 

 along with the general price level. Rent lev^els in various cities differ 

 according to the varying degrees of housing congestion and the 

 varying social standards of the population. Furthermore, the varia- 

 tion in rent levels extends to different sections of any one city. The 

 mere fact that a given family paid say $30 rent is not an indication 

 of that family's economic condition or its value as a telephone pros- 

 pect, unless there is also known the city and the part of the city in 

 which that family lived, and the time when the given rent was paid. 

 Therefore rent data from different cities and of various dates are 

 not directly comparable at their face value. To adjust the money 

 values of house rents for an accurate comparison of the telephone 

 markets in different cities w'ould require a knowledge of the relative 

 proportions of income spent for rent in the different cities, of the 

 relative levels of incomes and rents at the time of the surveys as 

 compared w^ith their levels in some base year, and perhaps of other 

 factors equally difficult to estimate. 



Various rent tabulations can not be compared one with another 

 without knowing something of the way in which rents are distributed 

 about their average. The nature of the distribution is determined 

 by the house count data, but from those data in their usual form it 



