542 



INSURANCE. 



it proportionally less; and provided the number of 

 those among whom it was distributed were very 

 considerable, it would hardly occasion any sensible 

 inconvenience to anyone in particular. Hence the 

 advantage of combining to lessen the injury arising 

 from the accidental destruction of property : and 

 it is the diffusion of the risk of loss over a 

 wide surface, and its valuation, that forms the em- 

 ployment of those engaged in insurance. Though 

 it be impossible to trace the circumstances which 

 occasion those events that are, on that account, 

 termed accidental, they are, notwithstanding, found 

 to obey certain laws. The number of births, mar- 

 riages, and deaths , the proportions of male to fe- 

 male, and of legitimate to illegitimate births ; the 

 ships cast away ; the houses burned ; and a vast 

 variety of other apparently accidental events ; are 

 yet, when our experience embraces a sufficiently 

 wide field, found to be nearly equal in equal periods 

 of time: and it is easy, from observations made 

 upon them, to estimate the sum which an indivi- 

 dual should pay, either to guarantee his property 

 from risk, or to secure a certain sum for his heirs 

 at his death. It must, however, be carefully ob- 

 served, that no confidence can be placed in such 

 estimates, unless they are deduced from a very 

 wide induction. Suppose, for example, it happens, 

 that during the present year one house is accident- 

 ally burned, in a town containing 1,000 houses; 

 this would afford very little ground for presuming 

 that the average probability of fire in that town 

 was as 1 to 1,000, For it might be found that not 

 a single house had been burned during the previous 

 ten years, or that ten were burned during each of 

 these years. But supposing it were ascertained, 

 that, at an average of ten years, one house had 

 been annually burned, the presumption that 1 to 

 1,000 was the real ratio of the probability of fire 

 would be very much strengthened; and if it were 

 found to obtain for twenty or thirty years together, 

 it might be held, for all practical purposes at least, 

 as indicating the precise degree of probability. 

 Besides its being necessary, in order to obtain the 

 true measure of the probability of any event, that 

 the series of events, of which it is one, should be 

 observed for a rather lengthened period, it is 

 necessary also that the events should be numerous, 

 or of pretty frequent occurrence. Suppose it were 

 found, by observing the births and deaths of 

 1,000,000 individuals taken indiscriminately from 

 among the whole population, that the mean dura- 

 tion of human life was forty years ; we should 

 have but very slender grounds for concluding that 

 this ratio would hold in the case of the next ten, 

 twenty, or fifty individuals that are born. Such a 

 number is so small as hardly to admit of the ope- 

 ration of what is called the law of average. When 

 a large number of lives is taken, those that exceed 

 the medium term are balanced by those that fall 

 short of it ; but when the number is small, there 

 is comparatively little room for the principle of 

 compensation, and the result cannot, therefore, be 

 depended upon. It is found, by the experience of 

 all countries in which censuses of the population 

 have been taken with considerable accuracy, that 

 the number of male children born is to that of fe- 

 male children in the proportion nearly of twenty- 

 two to twenty-one. But unless the observations be 

 made on a very large scale, this result will not be 

 obtained. If we look at particular families, they 

 sometimes consist wholly of boys, and sometimes 

 wholly of girls ; and it is not possible that the boys 



can be to the girls of a single family in the ratio 

 of twenty-two to twenty-one. But when, instead 

 of confining our observations to particular families 

 or even parishes, we extend them so as to embrace 

 a population of 500,000, these discrepancies disap- 

 pear, and we find that there is invariably a small 

 excess in the number of males born over the fe- 

 males. The false inferences that have been drawn 

 from the doctrine of chances, have uniformly al- 

 most, proceeded from generalising too rapidly, or 

 from deducing a rate of probability from such a 

 number of instances as do not give a fair average. 

 But when the instances on which we found our 

 conclusions are sufficiently numerous, it is seen that 

 the most anomalous events, such as suicides, deaths 

 by accidents, the number of letters put into the 

 post-office without any address, &c., form pretty 

 regular series, and consequently admit of being 

 estimated a priori. The business of insurance is 

 founded upon the principles thus briefly stated. 

 Suppose it has been remarked that of forty ships, of 

 the ordinary degree of sea- worthiness, employed in a 

 given trade, one is annually cast away, the probabil- 

 ity of loss will plainly be equal to one fortieth. And 

 if an individual wish to insure a ship, or the cargo 

 on board a ship, engaged in this trade, he ought to 

 pay a premium equal to the one-fortieth part of the 

 sum he insures, exclusive of such an additional sum 

 as may be required to indemnify the insurer for his 

 trouble, and to leave him a fair profit. If the pre- 

 mium exceed this sum, the insurer is overpaid; and 

 if it fall below it, he is underpaid. Insurances are 

 effected sometimes by societies, and sometimes by 

 individuals, the risk being in either case diffused 

 amongst a number of persons. Companies formed 

 for carrying on the business have generally a large 

 subscribed capital, or such a number of proprietors 

 as enables them to raise, without difficulty, what- 

 ever sums may at any time be required to make 

 good losses. Societies of this sort do not limit 

 their risks to small sums; that is, they do not often 

 refuse to insure a large sum upon a ship, a house, 

 a life, &c. The magnitude of their capitals affords 

 them the means of easily defraying a heavy loss ; 

 and their premiums being proportioned to their 

 risks, their profit is, at an average, independent of 

 such contingencies. Individuals, it is plain, could 

 not act in this way, unless they were possessed of 

 very large capitals; and besides, the taking of large 

 risks would render the business so hazardous, that 

 few would be disposed to engage in it. Instead, 

 therefore, of insuring a large sum, as 20,000 upon 

 a single ship, a private underwriter or insurer may 

 not, probably, in ordinary cases, take a greater risk 

 than 200 or 500 ; so that, though his engage- 

 ments may, when added together, amount to 

 20,000, they will be diffused over from forty to 

 a hundred ships ; and supposing one or two ships 

 to be lost, the loss would not impair his capital, 

 and would only lessen his profits. Hence it is, 

 that xvhile one transaction only may be required in 

 getting a ship insured by a company, ten or twenty 

 separate transactions may be required in getting 

 the same thing done at Llyod's, or by private in- 

 dividuals. When conducted in this cautious man- 

 ner, the business of insurance is as safe a line of 

 speculation as any in which individuals can engage. 

 To establish a policy of insurance on a fair foun- 

 dation, or in such a way that the premiums paid by 

 the insured shall exactly balance the risks incurred 

 by the insurers, and the various necessary expenses 

 to which they are put, including, of course their 



