THE CAUSES OF THE DECLINING BIRTH RATE 145 
States make a serious attempt to compel such registration by law. 
While physicians and midwives may comply with the regulation 
for reporting births, there are many children born without attend- 
ance, and which, therefore, are frequently not registered. More 
care has been taken recently in compiling data on births with the 
result that a larger number are reported. The rise in the birth 
rate of several of our states is not improbably due largely to this 
cause. Massachusetts has for many years compiled data on 
births and has passed laws compelling birth registration, but the 
U.S. Children’s Bureau has made a thorough study of a limited 
district in that state with the following results: ‘‘99 births were 
found to have been registered twice, 10 births were registered 
which actually occurred outside the limits of the municipality, 
10 births occurred in another year from that in which they were 
registered;’’ 123 births for one reason or another were not regis- 
tered. The errors, which were considerable, happened to offset 
each other fairly well since the record showed only 14 fewer births 
than actually occurred. 
The birth rate is undoubtedly affected by changes in the age of 
marriage and in the frequency of marriage, but it is evident that 
neither of these causes can account for more than a small part of 
the general decline in the birth rate during the past fifty years. 
Marriage statistics suffer greatly from inaccuracy of data on the 
age of marriage. As most people do not consider it a matter of 
much importance to report the true ages of the contracting par- 
ties, the age of the woman especially is frequently stated to be a 
few years younger than it really is.1_ Conclusions in regard to 
the effect of the marriage rate and age of marriage on the birth 
rate, so far as the United States is concerned, must be regarded as 
tentative. According to the U.S. Census for 1910, there has been 
for both sexes a gradual advance since 1890, in the percentage of 
married persons and in the percentage of married, widowed, and 
divorced persons combined. ‘In the age groups 15 to 19 years, 
1 For a discussion of what might be called the coefficient of mendacity for differ- 
ent ages of Australian brides see Knibbs, The Mathematical Theory of Population, 
Appendix A, of the Census of Australia for 1911, Vol. 1. 
