« SebelumnyaLanjutkan »
in statistical enterprises can readily furnish examples of this. Indeed, Mr. Allen's simple illustration of the value of statistical method affords an equally vivid illustration of its defect. Efficiency does promote goodness just about as the time-clock and cash-register promote punctuality and honesty. Any one who has ever managed an office knows that the perfunctory punctuality and honesty inculcated by the abovenamed devices are a poor excuse for the real thing; that the inert employee will remain inert under these checks and safeguards, and the actively ill-disposed will find ways to "beat" the machine. Yet the machine is, after all, better than nothing.
Thus with statistics. In a certain city department a commissioner was desirous of making a better record than his predecessor. In the statistical report of the department the number of orders issued was made a leading feature-on its face, a reasonable enough unit. The commissioner in question, by giving instructions to prepare the orders differently, made three out of the same material which had formerly made one, and showed a tremendously increased volume of work. Another plan resorted to was to issue orders as a result of perfunctory and worthless inspections. This also increased the volume of work done, and in this case there was no means of testing the quality. In the former case there was a means, as another tabulation of orders, on the basis of separate items attended to, showed the fallacy. But nine people out of ten, reading the report, would have failed to note the discrepancy.
This attempt to seize quality in the quantitative net is always imperfect, even under the best of circumstances. In the department referred to, units of work are necessarily employed to keep the inspectors up to a standard. Here, again, it is possible for an inspector to make good his required points by not doing his work thoroughly. There is a perpetual dispute going on between those who think the number should be so low as to permit of thorough work and those who think the number of units is of the most importance. Of course, a tendency to slight in quality of work is detected sooner or later by supervisors sent over the ground to test this. And this is not to say that cash-register methods are not necessary, but that they are cumbersome and mechanical at the best.
There are also the hindrances that arise when one set of people is trying to throw light upon and get light from another set of people. Dr. Allen has well summarized the main reasons why the production of intelligence has not been undertaken by the governing officials of large enterprises, public and private: first, fact is subordinated to expediency; second, in public life officials have been chosen for service to party rather than for fitness, or perhaps good men have been placed in office to carry out a program that a knowledge of actual conditions would have shown in
advance to be impracticable; third, officers are changed too often to discover needs and devise remedies or to develop a continuous policy.
And, lastly, it will always be difficult to get the public to use even the clearest and most carefully prepared information. For obvious psychological reasons the question, "Has any money been stolen?" will always be more vitally interesting than "Are we getting our money's worth?" After all "goodness" and "badness"-the absolute-is what takes the public eye, not the how much or how many. Nevertheless, the statistical method is the necessary machinery of the future. Like that other method of dealing with matters in gross, the factory system,—it will never supersede the method of dealing at first hand with the concrete things, and may often follow very clumsily after; but it is necessary to supply the clamoring need of the world, which can no longer be supplied by individual effort.
KATE HOLLADAY CLAGHORN.
NEW SERIES, No. 82
PROBLEMS OF SOCIAL STATISTICS AND SOCIAL
BY FREDERICK L. HOFFMAN.
Problems of social research require for their practical solution an adequate and conclusive basis of data free from even the suspicion of bias in their collection or serious error in their analysis. The ever-increasing complexity of social relations demands a clear presentation of social facts and forces, which, unfortunately, is only too often wanting as an underlying basis for plans and purposes of social reform. The collection of social statistics is almost invariably a most difficult and complex task, involving what may often amount to impertinence in a scientific inquiry into the actual facts of domestic life and the more or less successful individual adaptation to conditions as they are. This, for illustration, is best made evident in the numerous efforts to collect data as to household expenditures among wage-earners and others, but with patience and skill some, at least, of these investigations have produced conclusive and very valuable results. The value of investigation into social conditions is increased in proportion as the field is limited and as the investigator brings personal qualifications of an exceptional character to bear upon the collection of the data required. Those who are most familiar with the life and labor of the wage-earner, the poor and the pauper class, are, by their knowledge and experience, the best qualified to secure the original data upon many of the most important questions
*Read before the American Statistical Association, Yale Club, New York, April 24,
which demand solution. Unfortunately, it is difficult to secure qualified investigators, whose judgment has not become impaired by repeated impressions of social misery resulting from circumstances or conditions which may have no connection whatever with the problems under consideration. The object in view being strictly a scientific one, every effort should be made to eliminate sentimental bias or prejudice strongly inclined towards unwarranted conclusions or an unwarranted interpretation of the facts collected. Nowhere is the risk of amateur work greater than in the field of social statistics and social research, and, per contra, nowhere is the necessity of exceptional ability and discriminating judgment greater than in this. In economic statistics, such as prices and wages, cost of production and hours of labor, errors of judgment are less likely to occur, in that the degree of variety in the units to be considered is much less. Such data also are much less elusive in character, and not so complex in their relation to other and still more involved problems.
In its finality social research, as the term is generally understood, may be said to have for its object the solution of the problem of poverty, with all its resulting problems. Such social investigations, therefore, are largely concerned with an inquiry into the actual circumstances of life on the part of the poor and the relation of their condition to the wealth and circumstances of the materially more fortunate, or the well-to-do and the rich. The question which is being asked with everincreasing frequency is whether, under modern conditions, it is necessary that there should be as large a proportion of the poor and pauper class as are actually met with in civilized countries. Social inquiries are being directed to ascertain whether poverty, pauperism, ignorance, and crime are not more the result of an accidental miscarriage of human effort than of inherent limitations of human society as it is organized. Those who have felt most strongly upon the subject of social misery have elaborated in detail plans of radical social reform, but the many ideal communities which have been established have all been more or less complete failures. There are those who deny that
social progress is actually being made and who, in the words of Henry George, believe that "the poor are growing poorer, and the rich are growing richer." Theories are being spread broadcast over the earth as to the ever-increasing duties of the rich, the well-to-do, and even the prosperous towards those who are living under less fortunate material circumstances and conditions. In the abstract it is a question of social justice of one group of human beings towards the other, and it must be admitted that within the last generation, at least, a sense of social responsibility has been developed which was unknown in earlier and even comparatively recent times. The evidence is overwhelming that much of what goes under the term of social legislation has been productive of decidedly beneficial results, having improved the conditions of life generally and eliminated, among others, the needless evil of child labor and of degrading work on the part of women formerly employed in many industries unsuitable to the sex and certain to produce physical and moral deterioration. Much good has also been accomplished by social legislation relating to factory inspection, hours of labor, employers' liability, etc., all of which warrants the conclusion that even greater results may be attained by still more effective legislation or associated effort for the benefit of the mass of mankind not in a position to help itself.
As an aid toward the solution of these problems, social statistics are indispensable, and it may be said without fear of contradiction that much of the miscarriage of effort in social legislation has been the result of misleading statistics and even more of misleading analysis, little short of amateur guesswork. By slow degrees the inadequacy of the present basis of fact along certain lines of social legislation has been recognized, and efforts are being made in every direction to make such investigations more qualified, trustworthy, and practically useful. The time has passed when a plain statement of absolute fact relating to social conditions possible of amelioration or change could go unchallenged or leave a problem unsolved, merely as a matter of complacency, indifference, or criminal neglect. The present age demands the truth, and, when the truth has