16
THE STATISTICAL VERIFICATION OF
IV. The United States and Britain—a present contrast.
If we refer to the United States in particular, we
immediately see a great improvement in recent years.
First, the body of statistical material has grown ; secondly,
technical methods of statistical analysis are much better ;
thirdly, the funds and agencies for the laboratory work are
much increased, with numerous endowed organizations
for research. In this country it would, I think, be true to
say that aggregated and national statistics have consider-
ably improved, although we are still without any know-
ledge of total production, and have not much knowledge
of changes in distribution of income and the net product,
while our banking statistics are very difficult to interpret
and not so complete as they used to be. But in the field
of sampled or ad hoc inquiry we are very weak indeed,
depending on the efforts of individual statisticians and
research students, with very little team work. Our tech-
nical equipment is just as good in quality as that in the
United States, but there are fewer craftsmen. Thisis partly
because in the third respect we are completely behind the
United States, and we have practicallyno endowed statis-
tical inquiry. The London and Cambridge Economic Ser-
vice has to rely largely on annual subscriptions from
business firms. The importance of this difference cannot
be exaggerated, for as Wesley Mitchell rightly says, the
quantitative method, unlike the qualitative which needs
only a thinker and his books and pencil, has tofaceaheavy
burden of routine labour, computers, and field workers.
The list of institutions in the United States which are
responsible for co-ordinated team work and diagnosis is
NOW very imposing.
V. Examples. (a) Time series with published statistics.
It has been shown that there are two distinct aspects
of statistical inquiry. In the first, the statistics are
gathered and massed for their own sake, uniformities
and variations are noticed, and hypotheses constructed
