RACE SUICIDE IN THE UNITED STATES. 



BY WALTER F. WILCOX. 



[Walter F. Wilcox, export special agent of th(! United States government appointed to 

 investigate the fluctuations in the birth rate, has conducted other impcjrtant iiujuiries 

 for the department of commerce and labor; upon his graduation from collegia he began 

 a series of studies of social problems S(>curing his facts by original research, and his 

 published reports of his investigations into municipal conditions, and the; life of the 

 working classes have attracted wide attention.] 



The increase of a population aside from immigration de- 

 pends not merely on the number or proportion of infants an- 

 nually contributed to recruit or swell the ranks of the popula- 

 tion; it depends also on the number successfully reared. The 

 enumeration of children under 5 years of age is admitted by 

 every one to be far more accurate and complete than the enu- 

 meration of children under 1 year of age. The proportion of 

 children is thus an approximately accurate and a significant 

 clue to the amount of new blood that is being brought into the 

 country by nature's processes of reproduction and growth. 

 Even if the enumeration of adults is substantially complete 

 and that of children far from complete, no valid ground has 

 been shown for believing that the per cent of omissions among 

 children differs widely from census to census. Each census is 

 organized more efTiciently than the last and gathers its infor- 

 mation from a better educated, less suspicious, and more friend- 

 ly population. Hence such omissions should and probably do 

 tend to become relatively less frequent. In that case the re- 

 ported number of children would increase from census to cen- 

 sus faster than the actual number, and the tendency of such 

 a gradually disappearing error would be to mask rather than 

 to exaggerate the real decline in the proportion of children. 



It is a debatable question whether the population with 

 which the number of children is compared should be the total 

 population, the adult population, the women of child bearing 

 age, or the married women of child bearing age. Each method 

 has its advantages. The proportion to the total population can 

 be computed for a longer period than any other, and hence is 

 better adapted for a preliminary survey of the general trend. 



207 



