410 NATURE OF CAPITAL AND INCOME
operations need be done by rule of thumb. But at present
all but the first and often even that are left to the merest im-
pression.
There was a time when business men did not use bond
tables, when they did not calculate cost sheets, and even when
life insurance was contracted for in scornful disregard of any
mortality tables. Just as these slipshod methods have been
displaced by the work of expert accountants and actuaries, so
should the mere guessing about future income conditions be
replaced by making use of the modern statistical applications
of probability.!
§4 (to Cm. XVI, § 20)
Method of computing a Pure Level Life Insurance Premium
The chief peculiarity of life insurance is that each year’s
premiums, in properly organized insurance companies, are
calculated not on the basis of the chance of death in that
year, as in the case of fire insurance, but are computed as “level
premiums,” which exceed this chance in the early years of
the policy, and fall short of it in later years. The wisdom of
such an arrangement is justified by the incentive which
is produced for all “risks” to remain insured, whereas when
the “natural” premium only is charged, increasing with
age, there is a tendency for the better risks to withdraw,
making thereby a “selection” adverse to the companies.
A consequence of a “level” premium being charged is that
the policy acquires an increasing mathematical value with
time, so that the policy holder, after a few years, sometimes
possesses a very valuable property, which he can, if he chooses,
sell or use as collateral security for loans, ete. The value of
such a policy, computed on the mathematical basis, is of
1 Cf. “The Put and Call,” by L. R. Higgins, London, 1902, pp. 65, 66.
For some suggestions along this line, see Edgeworth, © Mathematical
Theory of Banking,” Jour. Roy. Statis. Soc., March, 1888; for a state-
ment of the modern statistical method, see Karl Pearson, Grammar
of Science, and his journal ¢ Biometrica® ; for an application of this
method to financial and industrial problems, see J. P. Norton, Statis-
tical Studies in the New York Money Market, New Haven (Tuttle, More-
house & Taylor), 1903.
