PROCEEDINGS OF SECTION G. 509 



value of that pension in question. For this to be so this cash value 

 must be built up during the term of active service of the employe 

 by contributions prior to the pension being entered upon. 



For the purposes of the fund it is immaterial who makes this 

 contribution, whether it is made wholly by the employer, as is 

 common among American railway companies, or made wholly by 

 the employe, as in some trading corporations, or made partly by 

 the employer and partly by the employe, which is the most usual 

 method. 



The contribution is then the soul of the fund, and it must be 

 adequate for the purpose. ^ 



In the early days of such funds scientific treatment apparently 

 was non-existent. Very generous schemes of benefits were arranged 

 between employer and employed. The benefits, being for the most 

 part not enjoyable for many years in the future, were underes- 

 timated, and utterly inadequate coi^tributions were levied. The 

 subsequent experience of many of these funds has been anything 

 but satisfactory, and a similar reconstruction awaits others in the 

 near future. 



As mentioned above, scientific treatment at the inauguration 

 of these early funds was non-existent — indeed, you may search 

 actuarial literature in vain until within the last twenty years for 

 guidance on this point. 



Not only was the treatment unsatisfactory, but adequate data 

 were absent, so that an approximation to a rate of contribution was 

 struck. 



From some cases that have lately come under my notice, I should 

 imagine that in these cases in which it was attempted to measure 

 the incidence of the benefits, and thereby calculate the necessary 

 premium, the rock upon which the estimates foundered was as- 

 sessmentism — that very widespread, alluring, and, in practice, 

 most deceptive and treacherous principle. 



In early days it was out of the question to quote anything but 

 an approximation or tentative rate of contribution, principally 

 owing to lack of data. At the present moment data are more 

 plentiful and a much closer approximation is possible, but the 

 ideal rate and ideal fund have not yet been reached. 



In any ordinary life assurance contract the principal element to 

 be taken into account is the rate of mortality, and that is well 

 understood for practical purposes. No harm is done by charging 

 a rate of contribution in excess of the true risk. 



To arrive at this ideal rate it is necessary to estimate — 



(1) The future rate of interest. 



(2) The future rate of mortality to be experienced among 



employes in active service. 



