RELATION BETWEEN RENTS AND INCOMES 87 



and fcHxl decreases and the per cent for clothing increases. The 

 decrease in the per cent for food as incomes increase is slight and the 

 increase in the per cent for clothing is especially marked in the higher 

 incomes within the range covered. Thus, it appears that among 

 families of moderate incomes as incomes rise the increase is spent 

 by preference for clothing rather than for food or rent. The relative 

 decrease in expenditure for rent as incomes increase is significant in 

 rental analysis. This means that while a 10 per cent difference in 

 rents among the lower rents in a city indicates an average difference 

 in income of about 10 per cent, a similar difTerence among the higher 

 rents indicates a difference in income of much more than 10 per cent. 



Rent Levels in Various Cities. As nearly as may be determined 

 from the Bureau of Labor Statistics data, there is no regular tendency 

 for Eastern, Western or Southern cities to differ from the average 

 of all cities, either in the amount of wage-earners' incomes or in 

 amounts spent for food or clothing. In Southern cities somewhat 

 less is paid for rent than in other cities. This refers only to w^hite 

 families. Negro families have smaller average incomes than white 

 families and at any given income they spend less for rent, more for 

 food and, to a less degree, more for clothing than white families. 

 The size of a city, so far as may be told from these data, does not 

 determine either total incomes or expense for food, rent or clothing. 



In different cities the difference in rent levels, that is, variation in 

 rentals paid for substantially similar dwellings, is considerably greater 

 than the difference in levels of prices for food or clothing. The varia- 

 tion in price levels is about twice as great for rents as for food or 

 clothing, reckoned as percentages of the amounts spent for each class. 

 The expenditure for food by the lower middle class families included 

 in this investigation is more nearly the same in different cities than 

 is the expenditure for rent or for clothing. Food expense is the only 

 one of these classes in which all cities are as closely grouped as in 

 total expenses, considering deviations from the averages on a per- 

 centage basis. The amounts spent for rent show relatively wide 

 variation between cities. It appears that if a workman moves from 

 one city to another to secure increased wages a large proportion of 

 the increase in income goes for increased rent. This is to be expected 

 since land rents and, to a less extent, construction costs are peculiar 

 to each individual city, much more than food or clothing costs. 



A comparison of rent data from the 1918-1919 investigation of the 

 Bureau of Labor Statistics and data from Commercial Surveys leads 

 to the conclusion that differences in average rents in various cities 

 are due at least as much to differences in the level of prices for rents 



