View report page

7
The difficulty of framing reasonably accurate estimates
of population towards the close of an intercensal period has always
been found very great—so great, in fact, that each new census,
when taken, has revealed serious errors in the Registrar-General's
estimates of local populations.
The difficulty is greatly increased at the present time
by the disturbing effects of the war ; and until the results of the
1921 census are available, all that can be claimed for the estimates
is that they are the best which the means at his disposal has enabled
the Registrar-General to prepare.
At the same time, while the war has rendered the methods
relied upon for estimating local populations in normal times quite
useless for this purpose now, it has fortunately brought in its
train new sources of information in the shape of the National
Register and Rationing Statistics, which have greatly mitigated
the difficulty. The mid-1919 estimates (also mid-1920) of local
populations have been mainly based upon the returns of the distribution
of rationing cards in November, 1919.
It may be mentioned that the general tendency is for local
estimates to rule higher than those prepared in the General Register
Office. In fact, the impression is left that if the sum of local
estimates could be ascertained, it would be found much in excess
of any reasonable estimate of the population of the country as
a whole.
The local estimates are very largely based upon comparison
of the numbers of occupied houses in 1919 and 1911, and upon the
fact that the margin of unoccupied houses then existing has since
disappeared. This method of estimation generally assumes that
the number of persons per house has remained at least as large
as at the last census. But the acuteness of the housing problem,
and the frequency of cases in which two families now occupy
one house because of the failure to obtain a second, invites the
assumption that an increase has taken place in the average number
of persons per house. Were this the case, the method of estimation
would understate the increase of the population.
But in a comparison between conditions in 1911 and 1920,
other factors must be set off against the more recent effects of the
housing shortage, which are so much under public notice. Many
households in 1920 were without one or more members, killed
in the war; many were without young children, who would have
been added to their number under peace conditions. Lastly, the
fall in the birth-rate, even apart from the war, has to be taken into
consideration. Probably this was a main factor in causing the
decrease in average numbers recorded up to 1911; and the decrease
in the birth-rate has continued since. There is no doubt that the
size of the average family has been on the decline for many years;