Harper's Weekly 08/19/1899
The Plans for the
ALL through the summer preparations have been
going busily on in Washington for a great
publishing enterprise, which will be launched
promptly on the first day of the coming June.
The results of the undertaking will begin to
appear in finished form two years from that date,
and will continue to be brought out at intervals for three or
four years thereafter. The publisher is the government;
the publication will be designated as the Twelfth Census
of the United States.
The Twelfth census will differ in several particulars
from any of the preceding ones. It will be conducted
on a larger scale, as there are of course more people to
be enumerated. it will embrace a greater area; for
the first time the inhabitants of Alaska, Hawaii, and
Puerto Rico are to be included in the count. Moreover,
the coming census will be the first in which all the work
of recording and computing statistics is to be done by
mechanical means. Electric tabulating-machines were
introduced for this purpose toward the close of the elev-
enth census, but in the coming enumeration they will be
relied upon entirely.
The thorough organization necessary in order success-
fully to carry through such an undertaking as this may
be appreciated when one reflects upon the labor involved
in counting seventy-five millions of anything—a task that
would require one man's undivided energies for twelve
hours a day during more than a year and a half. In the
case of the census the labor is multiplied by the considera-
tion that the seventy-five million units are human beings,
concerning each of whom a dozen facts must be record-
ed, and that they are scattered over some four million
square miles of the earth's surface.
The task of taking the census will require altogether
the services of more than forty thousand persons. They
will be separated into two main divisions—the field forces,
and the headquarters staff in Washington.
The former will include by far the greater number—
nearly forty thousand, all told. These will be the enu-
merators, who will gather the required information from
all parts of the country, and the superintendents in charge
of this branch of the work. The data thus collected will
be compiled and prepared for publication by a staff of
three thousand clerks in the central office.
Roughly speaking, there will be one enumerator for
each township throughout the country, or, in the cities,
one for each ward. The enumerators will be local resi-
dents appointed by the Director of the Census, on the
recommendation of some influential person, usually the
Congressman from the district. The superintendents will
have charge of divisions generally the same in limits as
the Congressional districts. In the case of the larger
cities, however, there will be but one superintendent to
each city, although his territory may include several Con-
gressional districts. In Massachusetts, where an efficient
census bureau exists under the direction of the State
authorities, there will be a single superintendent.
The enumerators are expected to start on their rounds
on June 1, 1900. They will be supplied beforehand with
portfolios containing blank schedules on which to enter
the name of each person in their districts, together with
the information provided for by law. Most of them can
complete their tasks within a few days, and will receive
from $50 to $150 for their services, according to the
amount of work involved. As soon as the schedules are
completed and revised, under the direction of the district
superintendents, they will be forwarded to Washington.
Here is where the work of putting the census data into
intelligible and valuable form will be done, and here is
where the tabulating machinery will come into play.
These machines, by-the-way, are the invention of a former
census employé, Mr. Herman Hollerith. They were de-
signed with a special view to use in the census, although
they have proved valuable for other statistical work.
By this system the statistics concerning each person
will appear on a separate punched card. About seventy-
five millions of these cards will be required, therefore, to
contain all the data collected for the census.
The cards are numbered to correspond with the num-
bers opposite the names in the schedules. They contain
two hundred and eighty-eight symbols, each of which is
an abbreviation representing some fact within the range
of the census enumeration. They are punched by means
THE CARD FOR INDIVIDUAL STATISTICS.
of a machine something like a type-writer in appearance,
which has on its key-board a reproduction of all the sym-
bols on the cards.
In recording the statistics a clerk reads from the sched-
ules the information entered opposite a certain name to
an operator seated at the key-board of the punching-ma-
chine. With a little practice this punching-machine can
be operated as fast as an ordinary type-writer. Experi-
ence has shown that the average number of records that
FREDERICK H. WINES,
one clerk can transfer from the schedules to the cards is
seven hundred per day. It is the intention of the Census
Bureau to put one thousand clerks at work with these
machines as soon as the returns are in, so that this branch
of the work should be completed in about a hundred
From the punching-machine the record cards go to the
electric tabulating-machine, which is even more ingen-
ious. In form it is something like an upright piano. In
the face of the upper part of the box are set a number of
indicator dials, each one devoted to some one set of facts
comprehended in the census. Inside the machine is a
complicated system of electric wiring connecting these
indicators with the operating apparatus.
It is the mission of this machine to total the various
facts recorded on the punched cards. To do this the
punched cards are slipped into the machine beneath a set
of electric needles, mounted on spiral springs. The op-
erator presses these needles down upon the card. Wher-
ever there are punch-holes the needles pass through and
did into a cup of mercury placed beneath. An electric
circuit is thus completed, which moves up the indicators
on the connected dials one point and records the particu-
lar fact indicated by each punch-hole. The totals are al-
ways in view on the indicators, and are copied off on
slips at the end of each run.
The old plan of computing these statistics was by a la-
borious system of hand-tallying. It may be readily be-
lieved that, with the increasing amount of labor involved
in each succeeding census, this would become an inter-
minable task. The improvement effected by the substi-
tution of an automatic process in this work can be judged
by the fact that one machine does the labor of twenty
clerks under the old system. Each machine is capable of
disposing of five thousand cards per day. About one
hundred and fifty of them will be required to keep up
with the clerks at work with the punching-machines.
Not only is the system of mechanical computation
more rapid than that of hand-tallying, but it is more ac-
curate. If one of the details—say that of sex—is not
punched. the electric plunger will not register, and the
automatic bell at the side of the machine which announces
the completion of the record will not ring. It is, then, a
comparatively easy matter to go back and supply the
missing information. Under the hand-tallying system it
was almost impossible to discover a mistake.
The statistics computed by the machines will be copied
on record slips and turned over to another force of one
thousand clerks, whose business will be to make up tables
and prepare copy for the printers.
In the eleventh census seven years elapsed before the
final volume of the principal report—that on popula-
tion—was off the presses. In the census of 1880 the last
volume was published in 1889. By the act of Congress
providing for the coming enumeration it was stipulated
that the four principal reports—on population, mortality,
agriculture, and manufactures—must be ready for pub-
lication on July 1, 1902. Director Merriam and his as-
sistants expect to accomplish this feat without difficulty,
and they will be greatly aided in doing so by the use of
the automatic tabulating-machines.
The Director of the twelfth census is Mr. William R.
Merriam, ex-Governor of Minnesota. The actual work of
preparing the statistical information of the census for pub-
lication will be in charge of Assistant Director Frederick
H. Wines. Mr. Wines has had long experience in this
sort of work. He was in charge of one department of the
eleventh census, and was employed also in the census of
1880. As assistant to Mr. Wines there are five chief sta-
tisticians, all experts in their lines, to each of whom will
be assigned one department.