PEOOEEDINGS OJ" SECTION G. 515 



short interval of time it passes out of mind unless we are in inti- 

 mate contact with the case; but, as far as the fund is concerned, 

 the draift upon the fund is a continuing one, going on year by year 

 during the life-time of the pensioner. 



In passing, I may mention how fallacious have the estimates as 

 regards the number and cost of old-age pensions and of invalid 

 pensions proved to be. 



The whole experience but exemplifies the aphorism : " The sup- 

 ply is greater than the demand." 



The Salary Rate. — The pensions usually depend upon the salary 

 received by the employe. It is, therefore, important to form what 

 is known as a salary scale. Naturally, every service has its own 

 salary scale, but to measure it satisfactorily is by no means easy. 

 In any particular case, when examined in bulk, we find that the 

 average salary per age forms a steadily increasing scale up to about 

 age 35, after which there is no regularity whatever. An increase 

 of staff usually means an increase in the numbers of the lower-paid 

 ranks, which means a lowering of the general salary scale. 



The average salaries at the various ages are of only minor im- 

 portance in themselves — the point of importance is the rate of 

 salary increase. It does not matter whether the actual salaries 

 are iOO, 110, 120, 130, 140, 150, 160, &c., &c., or 50, 55, 60, 65, 

 70, 75, 80, 85, &c. ; what is of importance are the ratios, and these 

 are the values which offer the more difficulties in determining as 

 regards the future. Of recent years the salary-increase rates have 

 shown a marked increase at the lower ages going to extend to the 

 higher ages. Is this increase at the lower ages going to extend to 

 the higher ages in the future, or is €he present increase shown at 

 the lower ages going to gradually disappear in the future 1 If the 

 former alternative is the one that will actually come to pass, then 

 a valuation in the future based on rates constructed from the past 

 data will be required to be stiffened. 



W ithdraival . — Most of those who enter a service at the younger 

 ages do not remain in it till the pension age — ^the majority leaving 

 it by withdrawal (by voluntary resignation or dismissal, being paid 

 off). This feature of the employment must not be ignored. To say 

 definitely what the effect of this element is is impossible; usually 

 the withdrawing member receives back a portion of the contribu- 

 tion made to the fund, and the result of making an assumption as 

 to the future rate of withdrawal as compared with that of making 

 another apparently quite as reasonable, all other things being 

 equal, can only be determined by trial. 



How the withdrawal rate affects the mortality rate, or the inva- 

 lidity rate, or the salary rate, or how the salary rate affects the 

 withdrawal rate are all points of importance in examining into the 

 future experience of a fund, but at present they admit of no 

 answer. 



