22 THE FALL OF THE BIRTH-RATE 



with their own. Our fertility rates used for stand- 

 ardisation do not represent a law of nature for women, 

 but only effective fertility in certain circumstances, 

 and there is an element of arbitrariness as indeed 

 there is in all methods of standardisation. 



As regards duration of marriage, in addition to age, 

 it is difficult to see any possible method of correction, 

 and, it may be added, the data as to the effect do not 

 seem to be consistent. 



There is another possible source of discrepancy, in 

 the earlier part of the period, for which Dr Dudfield 

 also attempted some allowance, namely omissions to 

 register. Such omissions will have been greatest at 

 the beginning of the period considered, and will have 

 tended to make the apparent fall in the decade 

 1871-81 too small. A correction is, however, very 

 hypothetical. 



The next question that presents itself is this : has 

 the fall in fertility affected all classes equally, or has 

 there been any differential incidence on the various 

 social classes? The answer is, almost certainly, that 

 the fall has affected successive social strata from the 

 top downwards in a rapidly decreasing degree. 



There can be no doubt that at the present day the 

 upper classes and unskilled labour stand at opposite 

 poles as regards fertility, though the evidence is 

 scrappy and indirect owing to the imperfections of 

 our registration system. I give in Table VI some 

 evidence from the Report of the Registrar General 

 for 1912. 



The births in 1911 were classified according to the 

 occupation of the father, which has to be stated at 

 the time of registration. These occupations were 

 grouped into five main classes, i and 2 representing 

 broadly the upper classes and middle classes and 3 to 

 5 the manual working classes, graduated from skilled 



