faith in democracy has this very solid foundation: that no leader's
wisdom can be applied unless the democracy comes to approve of it. To
govern a democracy you have to educate it: that contact with great masses
of men reciprocates by educating the leader. "The consent of the
governed" is more than a safeguard against ignorant tyrants: it is an
insurance against benevolent despots as well. In a rough way and with
many exceptions, democracy compels law to approximate human need. It is a
little difficult to see this when you live right in the midst of one. But
in perspective there can be little question that of all governments
democracy is the most relevant. Only humane laws can be successfully
enforced; and they are the only ones really worth enforcing. Voting is a
formal method of registering consent.

But all statistical devices are open to abuse and require constant
correction. Bribery, false counting, disfranchisement are the cruder
deceptions; they correspond to those enrolment statistics of a large
university which are artificially fed by counting the same student
several times if his courses happen to span two or three of the
departments. Just as deceptive as plain fraud is the deceptive ballot. We
all know how when the political tricksters were compelled to frame a
direct primary law in New York they fixed the ballot so that it botched
the election. Corporations have been known to do just that to their
reports. Did not E. H. Harriman say of a well-known statistician that he
could make an annual report tell any story you pleased? Still subtler is
the seven-foot ballot of stupid, good intentions--the hyperdemocratic
ballot in which you are asked to vote for the State Printer, and succeed
only in voting under the party emblem.

Statistics then is no automatic device for measuring facts. You and I are
forever at the mercy of the census-taker and the census-maker. That
impertinent fellow who goes from house to house is one of the real
masters of the statistical situation. The other is the man who organizes
the results. For all the conclusions in the end rest upon their accuracy,
honesty, energy and insight. Of course, in an obvious census like that of
the number of people personal bias counts for so little that it is lost
in the grand total. But the moment you begin inquiries into subjects
which people prefer to conceal, the weakness of statistics becomes
obvious. All figures which touch upon sexual subjects are not