MICHIGAN ACADEMY OF SCIENCE. 33 



coiiii)Oiiiul interest mounts up so rapidly. Assuming 5% as the 

 net return on laud, the value of a site must double in fifteen years 

 to make it ])rofital)le to hold it idle. 



Althou*?h the effect of almost any kind of local tax is to reduce 

 the value of land, and that of all well-planned governmental ex- 

 penditures is to increase land values, it is not by any means a mat- 

 ter of indifference on whom and on what basis the taxes are im- 

 posed. Total value, land value, gross rental and net rental, though 

 equally available for the tax-collector, do not stand in the same 

 proportion to each other in a comparison of different properties. 

 I mention this point only to avoid misunderstanding. Even if 

 the owners as a class must in any case ultimately bear the whole 

 incidence of taxation, the share of each owner would vary accord- 

 ing to the base chosen, and the duration and incidental results of 

 the process of shifting might be widely different; and I do not 

 maintain that the loholc incidence of local taxes is on the land- 

 owners. 



It is a matter of common knowledge that not all pieces of land 

 are fit for building sites, and that the difference in value between 

 those that are and those that are not is not determined by the ex- 

 penditure on the land, in laying out streets, putting down sewers, 

 and so on, plus the original value of the land, if any, for agricul- 

 tural jmrposes, but depends on the demand for building accommo- 

 dation in that particular locality, and the cost of constructing the 

 kind of accommodation demanded. It is not so commonly under- 

 stood that there are among the higher grades of land relations 

 similar to those between suburban agricultural lands and building 

 sites. If for example a certain site is well suited for a bank or a 

 department store, it will not for long be devoted to any other 

 form of development, for such sites are rare, and the return to 

 the land will be much greater, over and above the costs of build- 

 and maintenance, than if a warehouse or a tenement were put on 

 the site. Occasionally the difference in site rent to be obtained 

 from two competing methods of exploitation is slight, but in most 

 cases the difference is very great. It may not always be certain 

 that the development which promises the most site rent is bound to 

 succeed, but granting its success it will usually yield much more 

 than any other development could. 



Marginal land for any kind of development is land on which 

 capital invested in that certain way would yield, besides a reason- 

 able return on the capital, no more than would be earned by the 

 land if developed in some other way. A tax on buildings on such 

 5 



