ON THE NATURE OF INSURANCE. 2l 



could be made as stable as the duration of human life, 

 I could then see no objection to an immediate and con- 

 siderable reduction of the premiums charged, to an 

 amount at least equal to that proposed by Mr. Finlai- 

 son. But here lies the difficulty; that these tables, at 

 3-J- per cent., already involve a rate of interest which 

 the office cannot much exceed, if at all ; so that the secu- 

 rity which the precautions, nominally made against mor- 

 tality, really afforded against fluctuations of interest, is 

 partially or wholly destroyed, while no safeguard is in- 

 troduced to supply its place. 



An office raises its premiums either because its pre- 

 vious notions of existing mortality were wrong, or be- 

 cause it finds that it had calculated upon too high a 

 rate of interest. A mistake on either of these points 

 might be compensated by a contrary mistake as to the 

 other. Now, though the offices which existed during 

 the war have demonstrated that the mortality and rate 

 of interest together yielded a large profit, it by no 

 means follows that one of those causes of profit may be 

 fully corrected, while the other has been correcting 

 itself. To make both perfectly accurate, would bring 

 the office to the very line which divides security from 

 insecurity ; a position which it would not be safe to en- 

 deavour to maintain. We are already in a very dif- 

 ferent position as to the rate of interest, which has been 

 gradually falling since the war. The opinion as to the 

 extent to which tables of mortality may be safely cor- 

 rected, is formed upon arguments which dwell on the 

 favourable rate of mortality, without sufficiently consi- 

 dering the counterpoise (for, as far as it goes, it is a 

 counterpoise) existing in the alteration of the value of 

 money. 



Assuming the necessity of calculating upon a rate of 

 interest something less than that which can actually be 

 attained, I should think that no office would be justi- 

 fied in supposing more than 3 per cent., with tables 

 which are sufficiently high to come any ways near to the 

 actual experience of mortality. With regard to one 

 s 3 



