36 EIGHTEENTH REPORT. 



of 110 more assistance to laud owners in the struggle to avoid tax 

 burdens than the bottom of the ocean to a man struggling in the 

 middle of the Atlantic. 



Having discussed the case of a site adapted for different uses, 

 let us take up the case of several sites competing with one another 

 for the same use. all of them being above the margin for that 

 particular use. The simplest case is that of residence land. 



The effect of higher costs of buildings in checking demand we 

 need not consider at length. The demand for housing accommoda- 

 tion among most social classes seems to be fairly elastic, and to the 

 extent that it is, taxes obviously cannot be shifted on to tenants 

 in the shape of higher rents, for they will put up with less accom- 

 modation rather than jyay more. When this occurs the tax, whether 

 on building or land, will probably fall to some extent on all land- 

 lords, but all except the owners of the most expensive residence 

 sites will be able to replace a large part of the tenants they lose 

 by other tenants formerly dwelling in more expensive apartments, 

 and thus escape a large part of the burden. The most expensive 

 sites however have no such relief. 



But even Avhen the demand for housing accommodation is not 

 at all reduced by an increase in rentals or a tax, I shall attempt to 

 show that the traditional distribution of the incidence between 

 building and site is unwarranted. 



This reasoning covers the two classes of persons whose demand 

 for housing is according to Professor Seligman inelastic — those 

 who are too cramped already to be able to get along with less than 

 they have and those who are so wealthy they are unwilling to get 

 along with less. It can best be brought out by stud3dng a few 

 typical cases. Suppose a man with an income of |2,500 a year 

 choosing between a central and a suburban dwelling. There are 

 three possible hypotheses; (1) he has no preference and is guided 

 only by pecuniary motives; (2) he prefers to live in the city; and 

 (3) he prefers to live in the countr3^ In the first case he will esti- 

 mate the extra amount he is willing to paj^ for the central site by 

 considering the saving in car-fare and in time, which we will put 

 at -fo a month. We will assume that the type of house he desires 

 could be supplied at a rent of $25 if there were no taxes and no 

 ground rent. The city lot obviously has a site rent $5 greater than 

 the suburban lot, and it will not invalidate the argument if we 

 take the ground rents to be respectively |5 and nothing. 



Now suppose a tax on gross rental, collected from the tenant, 

 of 25%. Since we are assuming that the tenants demand is inelas- 



