CONCEPTION AND EVOLUTION OF MODERN CENSUS The census, as now conceived, has two main objects. It provides an instantaneous picture of the community—a cross-section of the body-politic exhibiting its constitution at the point of time when it is made. It thus affords knowledge of numbers and conditions which is valid for the particular point of time and, with approximations, for adjacent periods. But all things are subject to change ; and knowledge of the force and direction of changes in the community are even more important than the facts at any given date. A single sounding will disclose what depth of water is below the keel; but not without repeated sound ings is it known whether the water is shoaling, and how rapidly. Hence in the modern census each enumeration is conceived not only as a source of static knowledge but as an item in a con secutive series. From the succession of these alone can be as certained and measured those great drifts and currents in the national life which, whether deemed good or evil, whether resistible or irresistible, must at least be known and reckoned with.
The first among the questions upon which certainty was deemed requisite was the number of the population. Such a numbering entails a house-to-house visita tion coupled with enquiries of the inmates. This process (primi tive, but not to be bettered) affords opportunities, on the one hand, and imports limitations on the other, which have sub stantially determined the scope and evolution of census-taking. In the first place, the itinerary may be planned so as to group the dwellings visited according to any designed arrangement of boundaries : thus information obtained respecting the population of each areal group may be separately recorded and becomes available, not only as a component part of the national total, but also by itself, or as a component part of larger areal units, for the purpose of exhibiting the local distribution of that total. Further, the need to ensure that no dwelling is omitted leads to a record being made of all premises on the itinerary. It is but a step to include particulars of each dwelling, e.g., whether in habited or vacant, and the number of its rooms, thus affording, in combination with other material, statistics as to housing con ditions, overcrowding, etc. The convenience of obtaining partic ulars from one spokesman on behalf of each family or household leads to a system of household returns which themselves afford a count of such families. But the germ of the fertile elaboration of census enquiries is to be found in the fact that the information is obtained by direct enquiry of the persons concerned, thus permitting of the inclusion of every subject on which the indi vidual can reasonably be expected to state his own position and that of his dependants. Age may be asked and given : the relation ship and marital status of persons comprised in the household group can be described. Birthplace and nationality may be stated. A further group of personal particulars has furnished a great body of statistical material relative to occupation and industry. Every occupied person can state his job in terms which permit of precise occupational classification. And in addition to stating that he is, for example, a crane-driver, he can add that he is employed by a firm of brewers, thus enabling himself to be classed to the industry which he serves. Other enquiries are regularly included where the conditions render them suitable, such as religion, language spoken and literacy. Comparatively compli cated enquiries may be included such as the "fertility of mar riage" enquiry in English, French, Netherlands, Spanish and Norwegian censuses (date of marriage : number of deceased and number and ages of living children of the marriage).
The common element in all these enquiries is that they are personal. It has been suggested that the whole field of statistics can be viewed as concerned with either personnel or with material: this is not an entirely water tight division, but it accurately represents an important difference between the main points of view from which statistical informa tion is sought and studied. The census as an instrument for the collection of information is predominantly concerned with aspects of personnel, and covers the major portion of that province. It is unable to elucidate any subject except as an attribute of the individual man or woman. The inclusion of agricultural statistics in some countries is an apparent exception : but such enquiries may be in a sense an extension of the record of dwellings. The homestead is visited; it is noted that so many acres of arable, etc., are annexed and that there is stock of a certain kind and number. This is consistent with the census procedure, which must locate as it counts : it cannot deal with wealth which is everywhere and nowhere. Subjects such as exports and imports are clearly out of the picture. But there are some limitations in the field of personal attributes. The enquiry must be capable of being simply put and answered. Intrinsically there is nothing to prevent the use of the census to ascertain the distribution of the national wealth or income ; but the preparation of a statement in the form necessary to furnish the answer upon a uniform basis would be beyond the power of most individuals. If all particulars needed for the calculation were asked they would need a census schedule to themselves; and if the final figure alone were to be given there would be no security that it was properly arrived at. An enquiry of all wage-earners as to the rate of pay received (as in the Spanish census) would not be subject to the same objection.
But in many countries another general limitation would apply. Census returns cannot be verified, in view of the magnitude of the operations ; and it is useless to make any enquiry respecting which individuals may have any conscious motive, real or fancied, for suppressing or distorting the facts. Hence information which in the popular apprehension might be used by the authorities for enforcing individual liabili ties, such as taxation, cannot profitably be asked; and truthful information upon subjects, such as illegitimacy, which are felt to be delicate is often unobtainable owing to a fear of it becoming public. These considerations have tended to emphasize the character of the modern census as a source of abstract statistical information only. Whatever expectations may have been formed in the early days of the census revival as to its incidental utility for executive purposes, census administrators have had to choose between making it an effective instrument for the collection of statistics only and making the worst of both worlds. Another minor limitation applies to the doubtful ability of the individual to give reliable answers to enquiries as to infirmities such as blindness, deafness, etc. Even if no bias due to prejudice arises, the terms employed are unavoidably indefinite; and the con ditions themselves cannot be properly assessed save by trained medical judgment. It is becoming recognized that the census can afford little assistance of value in this sphere.
Within limitations the census has the peculiar and enormous advantage that all the varied attributes which it records are related to the individual to whom they jointly belong. Hence every type and class of information obtained can be pre sented in combination with every other type and class. If it be assumed, for the sake of illustration, that the collection of occu pational data is omitted from the census and separately organized by means of returns from employers, information might still be forthcoming of the number of metal spinners, while the census would continue to state the number of men aged 49. But between the two sets of figures there would be a gulf ; it would be im possible to obtain the number of metal spinners aged 49. Given, however, the inclusion of both enquiries in the census any com bination of these and other results is possible; and a census en quiry, as distinct from one separately promoted, not only affords the desired information but adds a fuller and richer significance to all those which it supplements.
And as the scope of the census has expanded, so the sphere of its utility has been enlarged. It still discharges its original function as an intelligence service of the government by which it is promoted in both legislation and administration. Estimates cannot be framed, for example, of the money provision requisite in future for the payment of old age pensions, or of the school provision needed for the rising generation, without the help of census material. Social insurance schemes equally rely upon such material for their actuarial foundation. Electoral redistribution must be largely based upon population distribution ; in South Africa constituencies are re arranged on the results of each census. Census statistics are the common tools and material of the business of government in ways too numerous to detail; but they are equally indispensable to the direction of State policy. In matters of defence the most pacific nation cannot afford not to know where it stands and whither it is tending. Questions of unemployment cannot begin to be considered without a knowledge of the industrial disposition of the people. Policy respecting migration still depends upon a recognition of the forces of population pressure which have so often changed the face of history. Knowledge of all these facts, even where not directly contributing to the solution of the prob lems of statecraft or state-policy, nevertheless forms a back ground against which they must be viewed. And for the public service in its widest significance the census provides material for research and study, helping to define and clarify the issues on the great questions to which there is as yet no agreed answer, and which, so far from having been admitted within the sphere of government responsibility, are but vaguely stirring in the conscience of the community. Not among the least of the serv ices of the census is that which it renders to statistical method whereby, for example, mortality data may be converted into mortality rates, thus eliminating variables and reducing to a com parable basis a wealth of material which for many purposes would otherwise be useless.
