THE NATIONAL CENSUS BUREAU AND OUR CITIES

BY ERNST C. MEYER

U. S. Bureau of the Census

The purpose of this paper is two-fold: first, to achieve a better understanding of the methods, problems, and results of the collection of statistics of cities by the National Census Bureau; and second, to arouse if possible a keener interest among the members of this Association in the accomplishemnt of the ultimate purpose for which this work is being carried on—an interest which it is hoped may lead to active assistance on the part of individual members along the lines indicated below.

Twelve years have now elapsed since the collection of statistics of cities was first regularly undertaken by the national government. From 1898 to 1902 this work was conducted by the Bureau (at that time Department) of Labor. In 1902 this inquiry was assumed by the Census Bureau, which, as is well-known, was made a permanent body at that time.

Originally only a rather modest inquiry into city finance was made. More recently however the work has been peculiarly favored because of the well-known deep interest in municipal affairs which Dr. Durand brought to his chair as Director of the Census, and also because of a similar interest on the part of the Assistant Director, Dr. W. F. Willoughby. And so it was but natural that under the able immediate direction of Dr. LeGrand Powers this work should blossom forth into an elaborate investigation of both financial and physical statistics, the latter including employees, equipment, and other physical property. During the current year a comprehensive study is being made of physical statistics covering the following departments of city service: sewers, refuse disposal, and highways, the latter including street cleaning and sprinkling.

It may be of some convenience to know at the outset that this paper is divided into three parts covering: methods, problems, and results.

METHODS

As is well known the Census Bureau has at times been the subject of criticism both because of the form and character of data published, and because of the delay in its publication. Intelligent criticism means progress. Some of you have in this wise contributed to progress. At the same time the writer believes that a brief exposition of the methods pursued in this work may be beneficial.

The subject of methods may be considered under three heads: first, the preparation of the schedule or form of inquiries; second, the collection of the statistics by special field agents; third, the publication of the statistics.

Preparation of schedule: The art of asking for everything that is needed; and of not asking for anything that is not needed, is a difficult one. The first rough drafts of the inquiries are prepared in the office. Copies are then sent to experts in the various fields of finance and physical statistics. In some cases mooted problems are settled by correspondence; in others conferences are held in which details are thrashed through in almost painful elaboration.

Aside from the determination of the subject-matter of the inquiries and their exact wording, much time and painstaking effort is devoted to the preparation of instructions to the special field agents who are to collect the information. Practically every inquiry is carefully interpreted and directions are issued in great detail as to how unusual situation in the various cities visited are to be handled. They also contain much purely instructional matter of a general character.

Some idea of the elaborate form which these instructions assume may be gathered from the fact that the instructions on municipal finance alone represent a 135-page volume of fine print.

Collection of statistics: At the present time the Census Bureau collects practically all of its information on cities not by correspondence, but, as already indicated, through the medium of special field agents. With the improvement of accounting systems and an increased interest in this work on the part of cities, the more economical, where practicable, system of correspondence can be gradually developed.

Every new agent goes through a practical course of instruction in the central office. Economy and expedition govern in the assignment of cities to agents. During the past year 35 field agents collected statistics in the 158 cities of over 30,000 population. Most

of these men have had several years of service, and some have had no less than 10 years of experience, having grown up, so to say, with this work. The demand of politics, it must be confessed, sometimes does inject new blood which neither invigorates nor increases efficiency while on the other hand the consciousness of salubrious political support, at times, proves a powerful antitoxin to undue assiduity even in case of some veterans. That however is merely stating a trite and outworn fact of the limitations of governmental machinery. Arrangements have been made under which a majority of the field agents spend several months of each year doing work in the central office at Washington, thus maintaining their familiarity with both ends of the work.

Publication of statistics: This includes the editing of schedules of agents; the construction of tables for the presentation of data; the tabulation of data; the interpretation of data, or preparation of the text; the printing of the completed report.

A large amount of the work of the central office has to do with the editing, or checking-up, of the schedules which field agents send in. This is done by making a close comparative study of the figures reported with those contained in the published reports, if any, of the city; and with the report of the agent for the preceding year. Wherever an explanation of discrepancies has not, as directed by instructions, been made by the agent, the specific facts are brought to his attention once more. Even correspondence with city officials may be conducted to clear up inconsistencies.

The preparations of tables for the presentation of the statistical data involves a procedure resembling that of the preparation of the schedules of inquiries. Tentative drafts of tables are made and submitted for criticism to experts in the various fields. Conferences and correspondence follow. As the work progresses from year to year and the tables assume their most practical, and therefore most permanent form this phase of the work will involve less and less time.

As soon as a sufficient number of schedules have been edited and corrected and the tabular forms have been determined the work of the tabulation of data is hurried along by specially trained clerks guided by expert tabulators.

The completed tables are next made the basis for an interpretative text discussion. Because of the unavoidable inaccuracy of some of the data collected, and the incompleteness of the records in others this task becomes a difficult one. Generalities are never of much

use, and definite conclusions are frequently rash and fearsome scientific ventures. Hence in the past the Census Bureau has probably been wise in refraining in large measure from both.

The text writing done, the scene shifts to the Government printing office which, though operating with well-known marvelous time-saving devices and machinery, is occasionally a serious accomplice in retarding the appearance of the bulletins of statistics of cities.

METHODS

Expedition in publication of the bulletins: This is a problem which has been vexing all who have been concerned with this work since its inception. The statistics of cities for 1908 are just now appearing; and even this is, the writer believes, a new record in speed. A thorough appreciation of the reasons for this delay may deaden somewhat the force of shafts directed at this apparently very vulnerable spot.

These reasons are to be found largely in the conditions which prevail in our cities; partly in the circumstances and unavoidable limitations of the work in the central office; and to a certain extent also, as already suggested, in the delay in printing.

It is difficult to portray adequately the obstacles which confront agents in the cities. Many cities publish no printed reports at all; others publish reports which are quite useless; some, and particularly New England cities, publish very valuable reports. In some cases accounts have been out of whack and balances have not balanced for years. Once helped back to an equilibrium by the field agent such cities have generally succeeded in keeping their balance. Even in large cities it not infrequently happens that a city has no record at all on important matters, particularly in the way of physical statistics.

The following extract from a letter written by the city electrician of a city of over 250,000 population in response to a call for information on street lighting will serve to set forth in concrete fashion what the Census field agents at times must face; I quote with due reverence to the grammar of the writer:

The series street lighting is a mixture there don't anybody seem to have any real definite idea of any information at all. . . The writer knows that there are more or less magnetite lamps and enclosed series arcs, but I do not think that there are any open arcs, etc.

To this must be added the constant confusion due to the adoption of new charters, or amendments to charters, the enactment of new state legislation, the upheaval in the city administration due to political upsets, the constant change in personel. Were it not for these far-reaching factors the annual visit of the field agent would serve to establish a personal relationship between the city and the central office; it would increase the interest of the city officials in this work, increase their ability in supplying the facts sought, and greatly accelerate the movement towards improved accounting.

The amount of time which a field agent must spend in a city, is probably far greater than the uninitiated would surmise. During the last year the following number of working days were spent in certain of our largest cities in the collection of physical statistics on sewers, refuse disposal, and highways, the latter including street sprinkling and cleaning: Philadelphia, 24; Boston, 35; St. Louis, 40; Chicago, 32, New York, 120. In the collection of financial statistics the working days put in were: Philadelphia, 105; Boston, 88; Chicago, 171; New York, 178.

The circumstances and unavoidable limitations of work in the central office also contribute to the delays. The careful editing involves an expenditure of a great deal of time. The employment of a larger editorial force would necessitate the laying-off of part of such a force during a large part of the year. The maintenance of a permanent force of experts would under such circumstances become an impossibility. Field agents, like other common mortals, do not all work with equal speed or equal ability. Their corrections of inconsistencies discovered by editors must be made in the field after they have left the particular city involved, and while carrying on their work in some other city. This occasions more delay.

The employment of a large force of field agents who might complete this part of the work in a short time meets with the same objection made to the temporary enlargement of the central office force. A corps of expert agents could not be maintained on that basis. Moreover, since the cities prefer to have but one or two agents visit them at any one time, the largest cities would set the limit within which the work could be completed at six months and more.

Gradually, however, solutions for these difficulties are found, and with them the speed of publication will be constantly accelerated. It is well to bear in mind that the investigation of city statistics has been constantly and rapidly growing, overwhelming those in charge with