is that
the L10,000 is split between Jones and Robinson. Jones maybe has a life
interest in it, and Robinson a reversionary interest. You value Jones's
wealth by his prospect of life on a life table, and Robinson has the
balance. But the life table does not indicate the actual likelihood of
Jones's life being fifteen years. It only represents the actuarial
average expectation of all the lives. This may be useful enough for
insurance dependent on the total experience, but it may be a shocking
injustice to the individual in taxation. Only some 10 per cent. of the
Joneses will live for the allotted time, and for the rest your valuation
and your tax will be dead wrong, either too much or too little. Jones
will be coming to you two years after he has paid, or rather his
executors will come to you and say: "We paid a tax based on Jones living
15 years, and he has died; this ought, therefore, to be shifted to
Robinson."

DIFFICULTIES OF VALUATION

People often say that a Capital Levy merely imagines everybody dying at
the same time. This parallel is wrong in degree when you are considering
the ease of paying duty or of changing the market values by a glut of
shares, and it is still more wrong when you are thinking of ease of
valuation. When a man is dead, he is dead, and in estimating the death
duty you have not to bother about how long he is going to live! But
every time you value a life interest and take a big slice of it for tax
you are probably doing a double injustice. The charge is incorrect for
two taxpayers. On a flat rate of tax this difficulty might be made less,
but the essence of any effective levy is a progressive scale. Moreover,
whether you are right or wrong about Robinson's tax, he has nothing in
hand with which to pay it. He has either to raise a mortgage on his
expectation (on which he pays _annual_ interest) or pay you by
instalments. So far as his burden is concerned, therefore, there is no
outright cut. You will be getting an annual figure over nearly the whole
class of life interests and reversions. It is difficult to see how one
can escape making adjustments year after year for some time in the light
of the ascertained facts, until the expiry of, say, nine or ten years
has reduced the disparities between the estimated valuations and the
facts of life to smaller proportions.

Next come those valuations which depend for their accuracy upon being
the true mid-point of probabilities. A given mine may l