THE JOURNAL

OF

POLITICAL ECONOMY

JUNE 1903

THE TWELFTH CENSUS OF MANUFACTURES. THE main purpose of this article is to consider the scope and methods of the census of manufactures and to raise certain questions relating to these subjects. Some of the questions are old and general, but they require to be asked again. Others are particular, and may or may not be new, but at a time when the census of manufactures is in a formative condition they merit attention. Changes and modifications of no small significance have taken place in 1890 and 1900 in the treatment of manufactures, and changes which may be called radical still remain to be made if the census is to realize its possibilities of usefulness. For the census of manufactures has been hitherto bound up too much in the swaddling clothes of tradition and legislative prescriptions. It needs freedom. Methods which were adequate in dealing with a simpler subject-matter, such as population, vital statistics, or even agriculture, break down in dealing with forces at once so plastic and organic as those of manufactures. This is especially the case when the function of the census of manufactures is interpreted as broadly as it is in the United States. Other censuses, especially the Belgian, are more perfect so far as they go, but in the comprehensiveness of its scope and in the vital character of its investigation the United States census Vol. XI. No. 3.

343

of manufactures is unique. And yet, though the scope of the census is already so wide, examination will, I think, show that it must be further widened. There are at present important questions affecting the progress and well-being of industry— questions imperatively requiring methodical and periodic investigation, which are untouched by the census. Place must be found for these subjects; and the question rises: How is it to be made? The size of the census is already a serious problem. If the present proportions are to be maintained, if there is not to be a very considerable pruning in certain directions, the natural. expansion proportionate to the industrial expansion of the country will involve a considerable increase in the census of manufactures. A more critical policy must therefore be pursued. What is required is that the question of census values be threshed out. We need a clear understanding of what the census of manufactures tries to do and cannot do, and does and ought not to do, and does not and ought to do. By such means whatever is unfittest may be judiciously removed, and place made for those inquiries which are considered to be worthy of and suited to a census investigation.

Before approaching the thorny problem of the scope and method, the simple question of the function of the census should be raised. What is the end or purpose of the census of manufactures? To be clear on this is of fundamental importance. It affects the whole consideration of scope and method. It is the basis of the judgment of values. Several of what we regard as the defects of the present census may be traced either to an uncertainty or to the neglect of the root-purpose of a census of manufactures, or else to a real difference of opinion as to what that purpose is. If the latter alternative is the case, then the question all the more deserves attention.

The purpose of the census of manufactures is practical. It is to present an accurate and organized body of relevant fact to the statesman and to the student of social and economic problems. The furnishing of comparisons or the compiling of detailed information which does not give promise of a practical issue should have no place in it. Such things are worse than

useless. They distract attention from that which is vital and they occupy space which other topics should have. The census is a strictly utilitarian investigation. The more it is so the better it is. To say this does not mean that it is to be one whit less impartial or less minute in its investigation, but it does mean that the clear understanding of the census function is the best insurance against the inclusion of irrelevant and inorganic matter. Further, to say that the census is for students and statesmen is not, as it may seem, exclusive and undemocratic. For democracy as it advances fits its tools to suit the hands of experts, not those of any and every one. As it is, there exists at grave danger of the census of manufactures failing to realize what Plato would have called the idea of the census and of becoming instead in parts a census, in parts a technical manual, in parts an industrial history, and in parts a field for statistical exercises of a kind which from their very nature are doomed to be barren of practical results.

The census of manufactures of 1900 marks important changes. and improvements. Each state and territory has been treated as a distinct entity, and the statistics for the state, its cities and counties, are presented together in one volume, a brief account of the origin and character of its leading manufactures prefacing the statistics of each state. A separation of manufactures from hand-trades has been made and a step advanced toward eliminating the hand-trades from the census of manufactures. The number of special reports on selected industries has been increased from twenty in 1890 to fifty-nine in 1900. An advance has been made in the classification and grouping of industries in the differentiating of classes, e. g., of salaried officials and wageearners; in the definition of the meaning of terms, and in the elimination of duplications. At the same time, the field of the expert enumerator has been widened and the industries of 1,341 cities and towns, as against 1,042 in 1890, has been reported on. Considerable improvements have also been made in the administrative organization of the census of manufactures. Contiguous localities have been grouped together under chief special agents a measure which has insured a more uniform and efficient can

vass. In several states a chief special agent to supervise the canvass of the state was appointed with satisfactory results. Further, a force of inspecting special agents was organized with marked success. These measures in the field, supplemented by developments in cataloguing at the census office, have rendered the work of the census of manufactures at once more uniform and more thorough. At the same time progress has been made easier by the increased willingness of manufacturers to give the desired information. Lastly, consideration has been given as to the best means of insuring co-operation between different investigating bodies, federal and state, so as to reduce, on the one hand, as far as possible the annoyance to the manufacturers of repeated inquiries, and, on the other hand, to prevent duplication of work and secure uniformity of method.

As regards the scope of the census of manufactures, no radical change has been made. The law required, so far as the schedule was concerned, an exact parallel of the census of 1890. And as the scope of the census has been "one of the most difficult and important problems which the office has been obliged to solve," and as it is around the question of the solution that such criticism as we have to present mainly revolves, it will be well to recall the subjects on which the census of manufactures presents returns. Roughly they are as follows: number of establishments; capital—in land, buildings, machinery, cash and sundries; number employed as salaried officials or wage-earners; number of men, of women, and of children under sixteen employed; total of salaries and of wages; miscellaneous expenses, viz., rent, taxes, interest, and contract work; cost of materials, purchased in raw state and in partially manufactured state; cost of fuel, rent of power and heat, cost of mill supplies, and of freight; number of establishments reporting power, total horsepower; value of products, including custom work and repairing. Returns are made also of the organization of establishments, whether individual, or firm and limited partnership, or incorporated company, or co-operative; of the average number of workers, men, women, and children, during each month of the year; of establishments, both in the hand-trades and manufactures,

classified by the number of persons employed; of the product of the census year and of the preceding year; of power used, whether owned or rented, whether engines, steam, gas, or gasoline, or water-wheels, or electro-motors. These subjects, or a number of them, are recorded in connection with an exhaustive list of specified industries, including hand-trades, and also as regards a group of fifteen leading industries. At the same time their geographical distribution as regards the states and territories is shown. Such is the subject-matter of the twelve tables. in Part I. In Part II the same subjects are treated of in connection with the different states, and the distribution of industries in the cities, towns, and counties is shown. In the case of the towns of over 20,000 the specified industries are given. Parts III and IV contain the special reports on selected industries, in all fifty-nine.

Such, roughly, is the scope of the census of manufactures. Let the census be its own first critic. We shall limit ourselves to noticing its comment in three cases. It should be said, however, that the census has passed many excellent criticisms on its own handiwork, and time and again one is led to think that it is its own best critic. Nor is this a new feature of the census work, but there is at least no diminution in the frankness with which difficulties and doubts are avowed.

The first case is the criticism of the practice of including the hand-trades in the census of manufactures. It is shown that the canvass is expensive, lacking in uniformity, arbitrary, and incomplete, and in view of "its demonstrated inaccuracy and the impossibility of making it otherwise," as well on account of its costliness, a strong recommendation is made that the investigation of the hand-trades should be abandoned in future censuses. This is a very important step. It is one which all who realize the gigantic task of the census of manufactures will approve. The hand-trades, as the census says, are different from manufactures. Much of their work is repairing work or job work. They are extremely difficult to canvass properly. Their number is very great, and they keep the most imperfect records of their income and expenditure. Under present conditions the investi