Introductory

Next Selection Previous Selection

CENSUS OF 1911.


TO THE RIGHT HONOURABLE LORD RHONDDA,
PRESIDENT OF THE LOCAL GOVERNMENT BOARD, ETC.


CENSUS OFFICE,
LONDON, S.W.
28th December, 1916.

MY LORD,

We have now the honour to submit our General Report on the twelfth Census of the population of England and Wales, which was taken on April 3rd, 1911. Various urgent demands upon the staff of the General Register Office in connection with the war have unfortunately delayed the preparation and publication of this Report, but owing to the method adopted on this occasion (described in Vol. I, Administrative Areas), by which full reports have been published for the whole country with each subject volume of Population and Social Statistics, the delay in the publication of the present volume, which is largely in the nature of a summary of the previous reports, has not had the result of withholding information as to the results of the Census.

A number of changes introduced in the present census have had the effect of considerably extending its scope. These increases in the material dealt with, as well as various questions connected with the collection and tabulation of the information, will be discussed before the results of the enumeration are reviewed.

We shall not on this occasion repeat the very full historical account of previous censuses which was a feature of the General Report on the Census of 1901, but confine ourselves on this part of the subject to such a summary of the main points as will explain our comments on the present position.



I.—INTRODUCTORY.

1. HISTORICAL SUMMARY.

A census of the population is taken in England and Wales decennially, the first enumeration having been made in 1801. There is no permanent Act; of Parliament authorising censuses to be taken at regular intervals, an arrangement common in other countries and one which obviates much administrative inconvenience, but each successive enumeration is the subject of a special Act, which prescribes the subjects of inquiry, fixes the date on which the enumeration is to be made, and designates the central authority for carrying out the requirements of the Act.

On the 20th November, 1800, a bill "for taking account of the population of Great Britain, and the increase or diminution thereof," was introduced into the House of Commons, passed through both Houses without opposition, and received the Royal Assent on December 31 of the same year. It may be noted that this Act did not embrace Ireland, and that none of the succeeding Census Acts have included the whole of the United Kingdom within the scope of a single measure. The Act of 1800 provided that the oversees of the Poor (or, in their default, some substantial householder) of every Parish, Township and Place should take an account of the number of persons who shall be actually found at the time of taking such accounts to be within the limits of such Parishes, Townships or Places respectively. This census, therefore, related to the population de facto and not to the population de jure (i.e. , the population usually resident), and the same course has been adopted at each subsequent census. The scope of the inquiry was defined in the three questions specified in a schedule to the Act, which the Overseers were required to answer, viz.:—

  1. How many Inhabited Houses are there in your Parish, Township or Place, by how many Families are they occupied; and how many Houses therein are Uninhabited?
  2. How many persons (including Children of whatever age) are there actually found within the limits of your Parish, Township or Place, at the time of taking this account, distinguishing males and females, and exclusive of men actually serving in His Majesty's Regular Forces or Militia, and exclusive of Seamen, either in His Majesty's service, or belonging to registered Vessels?
  3. What number of persons in your Parish, Township or Place are chiefly employed in Agriculture; how many in Trade, Manufactures or Handicraft; and how many are not comprised in any of the preceding classes?

Two further questions, to be answered by the Rectors, Vicars, Curates, or other Officiating Ministers of the several parishes, relative to the numbers of Baptisms and Burials in every tenth year from 1700 to 1800, and to the number of Marriages in each year from 1754 to 1800 were put with the view of determining the increase or decrease of t he population. The abstract of the results of this first census were presented to Parliament by the Secretary of State for the Home Department.

The three succeeding censuses did not differ in principle from the foregoing, though some minor alterations were made in the form of the questions. Thus in 1811, instead of asking how many persons were engaged in Agriculture, Trade, &c., this question related to families . Whether it was the intention at the first census to ascertain the numbers actively engaged in the principal kinds of occupation or the numbers dependent on those occupations is not clear, but the change of question indicates that in 1811 the latter object was held to be the more desirable. In 1821 an interesting addition was made to the questions, which otherwise remained practically identical with those asked on the preceding occasion. This addition consisted in instructing the Overseers, if of opinion that the Ages of the several individuals could be obtained in a satisfactory manner and without inconvenience to the parties, to state the number of those who are under five years of age, aged between 5 and 10, 10 and 15, 15 and 20, 20 and 30, and so on in ten-year periods of age. The return of ages was thus made on a voluntary basis, and as it was furnished for about 90 per cent, of the population, it may be regarded as furnishing a reliable guide to the age-constitution of the whole.

For some reason or other, however, the inquiry as to ages was not repeated in 1831, but at that census the question as to occupation was considerably amplified; in other respects this census did not differ materially, either as to the nature of the inquiries or as to the machinery and organisation for dealing with the returns, from the three preceding censuses.

Before the census of 1841 was taken two Acts of Parliament were passed, in 1834 and 1836 respectively, which have had a decisive effect on all subsequent censuses. The first of these two Acts—the Poor Law Act, 1834—created Poor Law Unions which, with the exception of a few large parishes not combining with others to form Unions, extend over the whole country. The other Act—the Registration Act, 1836—established the civil registration of births and deaths in England and Wales, and for the purposes of this Act, the newly created Unions were generally adopted as the areas for registration. A superintendent registrar was appointed for each registration district; these districts were divided into sub-districts each under a separate registrar, all these local officers being responsible to the Registrar-General as head of the central office dealing with registration matters.

When, therefore, the Census Act of 1840 was drafted the recently established system of registration districts and sub-districts was utilised as the basis of the census organisation, and the Registrar-General was appointed by the Act as one of the Commissioners for the purpose of taking the census. The Act further provided that each registrar's sub-district was to be divided into the necessary number of enumeration districts, for each of which a special enumerator was to be appointed; and the date of the enumeration was fixed for Thursday, 1st July, 1841. The selection of the middle of the year as the census day was a new departure, the first census having been taken on March 10th and the next three on May 27th, 28th and 30th respectively. There is obviously some statistical advantage in ascertaining the population either in the middle or at the end of a year, but when, for other reasons, the 1840 Act was amended in the following session, the date of enumeration was altered to June 7th. The precedent of taking the census on Monday was again followed (as it has been on every subsequent occasion), although the objection that many people are not at their usual place of residence on Sunday night was pointed out in the report on the 1831 census. The most important alteration, however, which was made by the amending Act was the provision of a separate schedule to be filled up by each householder with the particulars of himself and his family. Hitherto the returns had been merely numerical summaries prepared by the Overseer for the whole of his parish or district from information elicited by personal inquiry of the several householders. From 1841 onwards, the householder himself recorded the particulars for each individual person, and the enumerator has been responsible only for seeing that obvious errors were corrected in the schedules, and that the necessary particulars were not omitted. Thus both the method of making the returns and the organisation for collecting them were completely altered at the census of 1841; and at the same time the question as to ages was re-introduced and a new question, viz., "where born," was added to the schedule. With regard to ages, the replies were not, as in 1821, optional, but precise statements were not required, persons being allowed to give their ages as a multiple of five years; and with regard to birthplace were required to state merely whether born in the county in which they were enumerated, or the name of the country, if born out of England and Wales.

The 1851 Census Act did not add to the subjects of inquiry, but the forms and instructions issued under the provisions of the Act by the authority of the Registrar-General prescribed greater exactness in the replies. No abstracts of the numbers of baptisms, burials and marriages were, under this Act, required to be made from the Parish Registers, but in the belief that the collection of statistics with regard to places of religious worship was authorised by the Act, schedules were distributed for the purpose of obtaining information as to the amount of accommodation, the number of worshippers and various other particulars. The question was raised, however, whether such returns were compulsory, and it was finally decided to continue the inquiry on a voluntary basis, a ml the information so collected was embodied in an interesting report which was issued separately from the general report on the census. This is the only attempt which has been made in England and Wales to use the census machinery for the purpose of obtaining statistics as to religion.

Up to and including 1851, one Act of Parliament bad authorised each census for the whole of Great Britain, Ireland—in which the first census was taken in 1821—having always been dealt with separately. In 1860, however, separate Acts were passed for England and Wales and for Scotland, though in recent years the practice of providing for the whole of Great Britain in one Act has been reverted to. As, however, the census in each of the three Kingdoms is under the control of a separate authority it is impossible to secure that uniformity in, the tabulation of results which would be desirable, in spite of the careful consultation between them now usual. No change in procedure was made in the 1861 census, and only comparatively minor alterations and additions as regards the detail of the tabulation distinguished the successive enumerations up to 1901. The additions included (1) in 1861 an extension of the inquiry as to persons suffering from infirmities so as to show the number of lunatics and imbeciles in addition to the number of blind and deaf and dumb who had been returned for the first time in 1851; (2) in 1891, the addition, as the result of a motion by a private member in the House of Commons, of an inquiry as to the languages spoken by the inhabitants of Wales and Monmouthshire; (3) an inquiry as to the number of rooms occupied by families living in less than five rooms; and (4) an inquiry, in respect of persons engaged in trade or industry, whether they were employers, employees, or were working on their own account. The two latter changes introduced in 1891 were due to the recommendation of Departmental Committee appointed by the Treasury arid presided over by Mr. Leonard Courtney which reported in May, 1890—the only official committee which has ever examined into the objects and conduct of a census in this country.

No further additions to the subjects of inquiry were made at the census of 1901, and that census differed from the preceding mainly in the adoption for the first time of the method of publication of results by counties in advance of the figures for whole country a method to which it will probably be found desirable to revert at the next census.

2. CHANGES ADOPTED IN 1911.

We may now briefly describe the principal additions to the scope of the report upon The census of 1911 as compared with its immediate predecessor.

Volume i. Administrative Areas. —Under each rural district is displayed a list of its constituent civil parishes, with their areas and populations. This information was previously given only for urban districts, the rural civil parishes being tabulated solely as constituents of registration areas. The new tables have added greatly to the bulk of the volume, but it was felt that the relative importance of the administrative area has increased to such an extent as to make it essential that the information given with regard to it should in all respects exceed rather than fall short of that shown for any other type of area. This accordingly has been made a general rule throughout the report, many facts being tabulated only for administrative areas.1

For the numbers of "houses," previously tabulated for each area for every type dealt with, have been substituted the numbers of "families or separate occupiers," i.e., of census schedules collected. These figures are accurate whereas those referring to houses necessarily left much to be desired, since the definition of a "house" has always been found a matter of great difficulty. Extended and improved figures relating to houses have, however, been published in Volume VI for administrative areas only. As the significance of the figures relating to separate occupiers is prejudiced by the existence of institutions, &c., accommodating considerable populations returned on a single schedule, the numbers of these and the populations they contain have been tabulated for administrative as well as other types of area. By deducting them from the total of separate occupiers in any area the number of private families and the corresponding population may be approximately ascertained.

Volumes ii-iv. —These volumes, relating to registration, parliamentary, and ecclesiastical areas respectively, give very much the same information regarding them as has been contained in previous reports. The substitution of "separate occupiers" for "houses" applies to these volumes in common with Volume I, and for the same reasons. The maps showing the sub-division of each county into registration and administrative areas, which were published with the 1901 report, have not been repeated, as a much superior set of maps, published by the Ordnance Survey Department, giving the same information was already available for those who desired them, while the inclusion of the maps if not required entails an unnecessary addition to the cost of the reports. The accuracy of the facts tabulated for ecclesiastical areas has been immensely increased by means of the investigation of their boundaries and the resultant set of maps referred to on page 10.

Volume v. —This volume, consisting of a topographical index to the whole report, has been considerably extended both in regard to the number of localities included, and to the information afforded about each. In accordance with the policy followed generally of recognising the predominant importance of the administrative area, the situation of each locality is for the first time described in regard to these areas as well as, on the plan formerly followed, to registration areas.

Volume vi, Buildings, —The information given in this volume has not appeared in the report of any previous English census. The difficulty of defining what constitutes a "house" is here avoided by the enumeration, for each urban or rural district, of the number of various classes of buildings used as dwellings—ordinary dwelling houses, blocks of flats and the separate flats composing them, shops, institutions, &c., &c., with the corresponding populations.

The numbers both of the buildings and of' the separate occupiers housed in them are recorded in each case, so that it is possible for instance to ascertain for each town the numbers of blocks of flats, of separate flats contained in these blocks, and of sub-divisions of individual flats by lodger occupations. The different descriptions of uninhabited dwelling-houses and of those in course of erection are also shown, as well as of buildings such as Churches, Offices, &c., which are not used as dwellings. In this way a fairly complete census of the buildings in the country has for the first time been secured, and will be available for future comparisons, the nature of the various buildings enumerated being defined to such an extent as to admit of such comparisons being valid no matter to what extent the housing conditions may in the meantime have changed (as, for instance, by substitution of large blocks of flats for small dwelling houses), comparison merely of "houses " without allowing for such variations being evidently valueless. The particulars for compiling these tables have to be collected by the enumerators as a part of the routine common to this and previous censuses in order that no occupied building may be overlooked in their delivery and collection of schedules, so that the new departure made on this occasion applied to tabulation and publication only, and, therefore, involved comparatively little expense.

Volume vii. Age and Marital Condition. —The new feature of chief importance under this head is the statement of age in single years of life in combination with marital condition for the country at large, and for the various classes of administrative area contained in it. This advance has been rendered possible only by the adoption of mechanical tabulation. It has proved of the greatest value as a means of estimating the trustworthiness of the quinquennial tabulation of ages still for the most part retained for separate areas, and, in conjunction with a similar detailed tabulation of ages at death, of securing improvements in Me table construction. The detailed statement of males and females at the earlier ages living in each administrative and registration area has in this report been extended so as to include single years of age up to 21. Marital condition is also shown for the first time, both for administrative and registration areas (other than counties), and a record of the number of cases in which ages were not stated on the schedules, has been introduced into the volume.

Volume viii. Tenements. —The inquiry on the schedule as to the numbers of rooms in tenements, restricted in 1901 to tenements of less than five rooms, was in 1911 extended to tenements of all sizes, the tabulation of the results being expanded so as to deal separately with each size of tenement up to and including those of nine rooms, all of ten or over being grouped together. This change has not only increased the volume of information under this head, but has also increased its accuracy, for there can be no doubt that in 1901 a number of tenements containing four rooms or less were assumed to contain more because no statement of number of rooms had been made. In addition to the increased scope of the tables as regards size of tenement, they now distinguish fox the larger towns the number of children under ten years of age amongst the occupants, the resulting table being the most elaborate in the whole report, and one which it would have been quite impracticable to prepare without mechanical means of tabulation.

Special care was taken in preparing the tenement tables to limit their reference to private families only (see page 169), and in this volume will accordingly be found the most accurate returns we have been able to compile of the numbers of such families, these returns, however, being available for administrative areas only. Similar care in discriminating between private families and other aggregations of population as in hotels and boarding houses, residential business premises, and institutions, has been exercised at no previous census.

Volume ix. Birthplaces —Information under this head' has been increased by the separate tabulation of all persons born in County Boroughs and other towns of over 50,000 population in 1911. Formerly the county of birth (if in England or Wales) alone was distinguished, with the administrative county or county borough of enumeration, birt the counties of birth and of residence alike are in the present report the administrative counties, exclusive of the county boroughs, which are independently tabulated, but inclusive of the other large towns, which are also independently tabulated. This great increase of detail in the tabulation has the result of throwing light upon the movement of population between country and town, and is of value on that account. The further advance required to place England on a par with certain other countries in regard to this matter is probably not feasible without very radical changes in our methods of census taking and of vital statistics generally.

An experiment has been made in the direction of tabulating age in conjunction with birthplace by statement of the ages of all persons born in certain areas, including London, wherever they were enumerated, and of all persons enumerated m certain other areas, also including London, distinguishing their birthplaces Complete tabulation on this plan would be so bulky and laborious, that it will only be justifiable if, as a matter of experience, it is found possible to put the information now recorded in this way to profitable use. The ages and marital condition of those born in foreign countries have been tabulated at this as well as at previous censuses.

Minor changes include tabulation of the county of origin of those born in Scotland and Ireland (this information not having been asked for on the schedules previous censuses) and the distinction of residents from visitors amongst those born outside the United Kingdom.

Volume x. Occupations and Industries —The information tabulated under this heading involved a new question in the schedule. In addition to the statement of personal occupation, demanded at each census from 1841 onwards, inquiry was made as to the service or industry (as indicated by the nature of the employer's business) in connexion with which such occupation was carried on. The information so obtained was valuable both in itself and from the light thrown by it upon the statements of personal occupation, which have often proved to be explicit when read in connexion with the statement of industry though not so otherwise. The results of the new inquiry have been tabulated with distinction of sex and age for England and "Wales only, and, without the distinction of age, for small groups (20 in all) of adjacent counties, since it is held that this information is chiefly of value in its bearing upon the industrial position in the country as a whole. The tables relating to personal occupation on the other hand apply, in varying degrees of detail, to a large number of separate areas. They also distinguish for the first time widowed from married females, and (for the country as a whole only) state the numbers and proportions of married males of different ages engaged in the various occupations.

In response to a considerable demand for additional information as to the occupations of young persons and the changes of employment taking place amongst them, the ages of male and female workers in each occupation have been tabulated by single years from 10 to 21 inclusive. This somewhat laborious undertaking has provided much fuller census material than was before available for the study of the question of "blind-alley" occupations and of the problems of youthful labour in general.

The list of occupations has been considerably increased; and the precise constitution of each heading included in it has been published as an Appendix to Volume X. of the Report (Cd. 7660). This shows the occupational designations allocated in tabulation to each heading in the list, and is, therefore, essential to a complete apprehension of the meaning attaching to the headings in census practice. Corresponding volumes had been prepared for the purposes of several previous censuses, in order to secure uniformity in the practice of the tabulating staff, but the information has now for the first time been placed at the disposal of the public.

Fertility, —Finally, the whole of the report upon fertility covers quite new ground, so far as English censuses are concerned, and represents a considerable proportion of the work involved by the entire report. This subject is not dealt with in the present volume. The material for England and Wales is too bulky for inclusion in a general report, and therefore can be shown only in the special report dealing with it.

United Kingdom, —Comparative figures of the population of the United Kingdom and its several divisions have been shown, for each census from 1821 onwards, in Tables 93-95 of the Summary Volume; and the distribution in foreign countries of natives of the United Kingdom (or of British subjects) in 1911 is given in Table 96 of the same volume. Further statistics showing the ages, marital condition, occupations, and birthplaces of the population will be found in Appendix C, pages 263 to 285, but considerations of space do not, in the present circumstances, admit of detailed comment on these figures such as was given in the report on the census of 1901; moreover, the fact that this detailed comment was given so recently as 1901 makes it less useful to go over the same ground again on the present occasion.

Census of the British Empire. —In the final paragraph of the Preliminary Report on the Census of England and Wales 1911 (cd. 5705, 1911) we intimated our intention to present a Report on the census taken in that year throughout the British Empire somewhat on the lines of the only previous Report on the subject, that presented for the census of 1901 (cd. 2260, 1906). With this object we had, at a period anterior to that census, been in communication with the Colonial Office in order to secure that a census of some kind should be taken in all the Dominions, Colonies and Protectorates oversea, and to direct attention to the subjects which are of importance in an imperial comparison; and we had caused copies of the schedules and instructions prepared for the Census of England and Wales, as well as of the previous Empire Report to be communicated to to the various Governments concerned. The great mass of material with which we have had to deal for the Reports on the home census and the delay which necessarily occurs in the receipt of the reports from some outlying portions of the Empire made it impossible to take the work seriously in hand before the outbreak of the present war, and since that time the causes which have prevented the earlier publication of our General Report have retarded the preparation of a report on the census of the Empire and will continue to do so for an indefinite period. We have therefore decided to print as an appendix to the present Report a series of comparative Summary Tables (see Appendix: D pages 286 et seq. ) similar to those which formed a party the "Empire Report above alluded to, thus making accessible materials from which deductions may be drawn by inquirers at a much earlier date than would otherwise be possible. By this decision we do not wish to prejudice the question of the ultimate publication of such a Report as we foreshadowed in 1911; but a further consideration of the Report on the Census of the Empire published in 1906, has raised a doubt in our minds how far a Report on these lines which, after all, can only be a brief epitome of the returns serves any very useful purpose. It can never in any case supersede in interest the separate reports prepared by those familiar with the local circumstances and problems of the various portions of the Empire; and it is a question whether future progress should not be sought for in the development and improvement of these reports and an increase in their scope and interest, and whether simple tabular summaries of such results as are comparable may not be all that can advantageously be attempted by any central authority. Our procedure on this occasion has at all events the merits of bringing together in a single volume population statistics relating to every part of the British Empire.

3. COLLECTION OF CENSUS DATA.

The means employed to produce the mass of information embodied in the report, both that common to previous reports, and also more particularly the new features of the present report which have just been described, may be classified under three heads, the schedule of questions asked regarding each individual, the methods of distribution, collection and supervision of these schedules, and the tabulation of the information thus collected.

The Schedule. —In addition to very careful revision throughout, new questions were included with reference to the duration of marriages and numbers of children born and surviving (Nos. 6-9 on schedule), and as to the industry or service with which the work of occupied persons was connected (No. 11). Besides these entirely new inquiries, that regarding number of rooms was extended to refer to all sizes of tenement, and various additions were made to the inquiries relating to birthplace and nationality.

We had proposed, as against these considerable additions to the bulk of the schedule, that the inquiry as to the infirmities should be abandoned, on the ground that, as the Report on the census of 1881 puts it, the replies are not worth the cost and labour of collection and tabulation. Information on this subject obtained from householders through the mechanism of a population census can be of little value except possibly as a basis for further inquiry, and with the establishment of school medical inspection and national health insurance, new machinery has now been established by means of which it has become possible to collect very much more accurate information on this subject than can possibly be afforded by the unskilled and frequently prejudiced replies to the census inquiry.

The proposal that the inquiry should be abandoned did not, however, commend itself to the Local Government Board, and it was, therefore, continued. In order to make the inquiry as definite as possible, and to avoid the difficulty of calling upon the householder to apply a standard defining the degree of defect to be recorded on the schedule, the questions as to blindness and deafness were limited to cases of total disability.

This limitation has met with much disapproval on the part of those interested in the welfare of the blind and deaf, who naturally desired to obtain information of all cases of economically disabling, even though only partial, defect. This most desirable object is however, unfortunately impossible of attainment by the ordinary machinery of a population census, where responsibility for the accuracy of the information is thrown upon the householder.

Distribution and Collection of Schedules. —The organisation of this work was in essentials the same as at former censuses from 1841 onwards, the work being entrusted to the local registrars of births and deaths. Each of these officials (2,035 in all) was instructed to prepare a plan for the division of his district into areas—"enumerations districts"—of such a size and population that one enumerator could call at each house and collect every schedule in a single day, and so arranged as to avoid as far as possible inclusion in one enumeration district of portions of more than one area of any type for which tables are published. The total number of enumeration districts amounted to over 35,000.

The registrar having made what he considered a suitable division of his sub-district into enumeration districts, forwarded his "plan of division" to his superintendent, who, after examining and approving it, transmitted it to the central office. There it was again examined and compared with the official map of the sub-district (see below), and finally returned to the registrar corrected and amended where necessary.

The registrar was also required to engage suitable enumerators to perform the actual work of delivering and collecting the schedules, to instruct them in their duties, to see that the schedules were delivered to householders or lodgers within one week before the census day, and that they were all collected on the following day after being filled in by the householders or lodgers, to scrutinise the schedules In order to ensure that the enumerators had performed their duty in securing their proper completion before collection, and to deliver the books of schedules to his immediate superior, the superintendent registrar, by whom they were further scrutinised before being forwarded to the Census Office.

In one respect this routine differed materially from that followed at previous censuses. At these the enumerator was required to copy the replies to the questions on the schedule into an "enumeration book," in which form the information was very much more compactly arranged, blank spaces being eliminated. It was from these enumeration books that the census tables were prepared. This procedure had the advantage of presenting the information in a much more legible, compact and convenient form for tabulation, but it involved much, laborious work in copying which, especially in view of the increased schedule of the present census, would have had to be taken into account in fixing the enumerators' remuneration. It also led, despite the checking provided for in the regulations, to many copyists' errors, some of a purely accidental nature, and others due to mistaken emendations from, misinterpretation of the replies. It was, therefore, decided to omit this copying process in 1911, and to tabulate from the schedules themselves as received from the public. The inconvenience of this course, though undeniable, has been found by no means prohibitive, and we have no doubt that it has tended both to lessen the cost and to increase the accuracy of the present census.

In order to ensure that schedules were left at all inhabited buildings it was the duty of the enumerator, as at previous censuses, to carry a memorandum book" while delivering the schedules, in which he entered, street by street, the address and description of every building, inhabited or not, with the names of the heads of families. After collecting and numbering the schedules, and entering the total number of males and females on each schedule, all these particulars were copied, from the memorandum book and the schedules concurrently, into a summary book. These "enumerator's summaries," prepared primarily in order to enable the registrar to ensure by checking them that no housed population was omitted in the enumeration, played a very important part, as will be described later, in the tabulation.

Use of Maps in checking the returns. —The principal new departure, other than the elimination of the "enumeration book," in the method of collecting the data of the census of 1911, was the use made of specially constructed maps in order to ensure accuracy in the assignment of houses and population to the various types of area dealt with as well as suitable and convenient subdivision of registrars' areas ("registration sub-districts") into enumeration districts. The want of such maps haying been very much felt at previous censuses, permission was obtained to commence their construction nearly two years in advance of census day. A set of six-inch (in towns, 25-inch) ordnance survey sheets was prepared by a staff of clerks employed to trace as accurately as possible the boundaries of the ecclesiastical areas, making use of all obtainable sources of information, such as Orders in Council, and local maps and records borrowed for the purpose. This process proved especially difficult, as there was no set of maps in existence on which the boundaries were defined, and the local information was, in many cases, vague and conflicting. With the help of the Ecclesiastical Commission and of the Diocesan Registrars, these difficulties were largely overcome, and a record was mapped of the ecclesiastical parishes of England and Wales, which is as accurate as the information available on the subject permits. The set of maps thus prepared was forwarded to the Ordnance Survey Department, and that Department compiled a second set of maps primarily for the use of the local registrars. For this set a separate map was prepared for each registration sub-district, the boundaries of the various kinds of areas—civil parishes, urban districts, municipal boroughs, wards of urban districts or municipal boroughs, and parliamentary divisions—being inserted in distinctive colours from the Ordnance Survey records, while the ecclesiastical boundaries were transferred from the Census Office maps. These sub-district maps, on which the boundaries of the enumeration districts were afterwards indicated by the registrars, were of the greatest service in the conduct of the census. They have on the one hand enabled registrars to understand the constitution of their ages thoroughly, to divide them intelligently into enumeration districts, and to describe precisely and accurately all the boundaries dealt with; and on the other hand, they have for the first time supplied the Census Office with an adequate means of supervising this important part of the registrar.s duties, and securing its amendment when required.

Co-operation of the public. —While increased accuracy was thus secured in the tabulation of areas by the preparation and use of maps, an attempt was also made to improve the returns for individuals by increased effort to secure the interest and intelligent co-operation of the public.

With the assistance of the Board of Education, it was arranged, as at the previous census, for lessons on the subject of the census to be given in the public elementary schools, and an enlarged copy of the schedule was supplied for demonstration purposes. We feel confident that this measure has proved most useful in obtaining, in many cases, the proper filling-up of the schedules, frequently by children who had been allowed to fill up specimen schedules in the schools. The Parliamentary Committee of the Trades Union Congress also kindly helped by distributing explanatory memoranda to the various Trades Unions, and in this way undoubtedly facilitated the collection of accurate and complete information, especially with regard to the occupational inquiry. Shortly before census day, moreover, a wider appeal was made by means of a circular letter, describing the objects of the census, which, through the courtesy of the Press, was inserted in most of the newspapers.

Revision and Summarising of Schedules. —The ground having been thus prepared, the enumerators, who, subject to the approval of the Registrar-General, were appointed by the registrars and took instructions directly from them, were required to deliver the schedules during the week preceding census day, and to collect them all on Monday, 3rd April, 1911. While collecting, the enumerator had to see that the schedule was properly filled up, and to ask any questions necessary to enable him to correct it. A special form of schedule was used for institutions, which, if large, were enumerated by the chief resident officers; for persons on board vessels, who were enumerated by officers of His Majesty's Customs; and for vagrants, who were enumerated by the police.

Though the enumerators were required to collect all the schedules in one day, they were allowed six days in which to complete the revision and summarising. When the summary was completed it was taken with the schedules to the registrar, who was instructed to examine the schedules thoroughly and to check the enumerators' summaries. The registrars were allowed three weeks for this process and were required to forward all the schedules direct to the central office by the end of April, the summaries being forwarded through the Superintendent Registrars, who were instructed to satisfy themselves that no part of the district had been omitted from the enumeration. The population summaries furnished by the enumerators were aggregated by the registrars into the several local sub-divisions within the registration sub-district, and from these registrars' summaries a preliminary abstract of the population, giving the figures for each county, urban district, rural district, parliamentary area and registration district and sub-district, together with a preliminary report on the general results, was compiled and published within ten weeks of census day. Before the final tabulation was begun the schedules were revised and corrected as far as possible at the central office.

Defects of present method of collection and suggestions for improvement. —The system just described has on the whole proved to be efficient and satisfactory for the purpose for which it was created in 1841, i.e., for securing an enumeration of the people with a few very simple particulars regarding each individual. Even when considered only as a means to such a limited end as this, there has been in the late and probably all previous censuses a certain amount of wastage and inefficiency. Under the circumstances, involving as they do the improvisation once in ten years of a huge temporary staff, to be employed only for one week or less, and then dispersed again, tins is no doubt, in greater or less degree inevitable. We hope and believe that the care taken in the organisation of the late census in the light of the experience gained in 1901, had the effect of increasing considerably the efficiency of the machinery, and this improvement is, no doubt, susceptible of a still further development.

We have, however, impressed by the fact that the present system is capable of meeting adequately only the simpler requirements of census taking. We feel that without a change of system which would leave less to the unchecked discretion of the householder further elaboration of the schedule beyond the point reached in 1911 would probably be found ineffective. It is true that the regulations of this, as of previous censuses, have made enumerators responsible for seeing at the time of collection that all schedules were filled up satisfactorily and registrars for ensuring that enumerators performed this duty. In practice, however, this provision is very much a dead letter. With 300-400 schedules to collect in a single day it would doubtless be impossible for the most zealous and well-instructed enumerator to go fully through every schedule collected and obtain and insert the missing particulars where the information supplied by the householder was defective. Even if this were possible the effort involved would postulate a staff of enumerators more zealous and well instructed than we can ever hope to secure. Nor do we consider that the Householder is to blame for the failures met with. On the whole, lie struggles, heroically with an unaccustomed task, and produces quite as well filled a schedule as it would be reasonable to expect of him. All forms, however, are more or less difficult for those unaccustomed to them to fill in perfectly. The census schedule is an elaborate and in the nature of things a difficult form to fill in, and the average householder is a person without much clerical or literary training, and quite unaccustomed to the formidable form with which he is confronted. Little wonder, then, that his efforts, creditable to him as we feel them on the whole to be, frequently fall short of the desired success,

As an instance of the difficulties met with we may quote a question which has appeared in some shape or other on the schedule of every English census and which it might seem at first sight a very simple matter to answer, namely, that relating to occupation. The present classification of "occupations" is only truly occupational in parts, being, as regards manufacture, for the most part frankly industrial. Thus the driver of a motor car is classified as a chauffeur—his occupation; but the workman engaged upon its manufacture is classed either as a motor car chassis maker or as a motor car body maker, headings which indicate merely the industry in connexion with which he plies his trade, whether it be that of tinsmith, moulder, fitter, sand-blastman, painter, or what not. The schedule aims at eliciting both types of information and asks for a statement of the precise branch of manufacture and the particular kind of work done. On this occasion, with a view to future revision, a very thorough test was made of the extent to which the departure of the list of "occupations" from a truly occupational basis is dependent upon the nature of the information on the schedules, and therefore, under the present system of census taking, inevitable. For a large number of the headings in the list, some of the workers under which were believed to be engaged in processes involving special risk to health, cards were prepared upon which the description of the occupation was exactly copied from the schedule, it being hoped by this means to ascertain the numbers subjected to the unhealthy conditions. The attempt was a complete failure, for it was found as the result of the test that such workers commonly describe their calling merely in terms of the article produced, that under which it is classified in our present tables of occupations. Thus there are many separate processes in needle making, some dangerous to health and others not, but it was found that the great majority of the Redditch workers had described themselves simply as "needlemakers." In the case of the Sheffield toolmakers, where it was particularly desired to obtain the distinction of the dangerous processes, the results were better insomuch as a majority described the nature of their work as well as the product, but in this, as in all other cases, the proportion who did not afford the necessary information was so large as to make a truly occupational tabulation impossible.

The questions as to occupation are probably, with the exception of that relating to infirmities, the worst answered, because the most difficult to answer, upon the schedule; but all the questions are liable in greater or less degree to be incorrectly or carelessly replied to. Incorrect answers did not, however, in the case of most questions affect the total number of persons shown on the schedule, and did not therefore lead to mis-statement of the actual population. As an exception to this rule, the questions with regard to the number of children (Nos. 7-9) appear to have been misunderstood by some people with the result that the names and other particulars of children who were not present on census night, including those who were returned as dead, were entered on the schedule, and the population was overstated on this account. Special instructions were issued to the registrars to be on their guard against this particular form of error, and most of the incorrect returns were rectified before transmission to the central office; the further revision at the central office did, however, reveal a large number of cases in which the erroneous entries had not been detected, and further reductions in the population as returned had, consequently, to be made. This appears to us to be a necessary consequence of the system of reliance upon the practically unchecked returns of householders which characterises English censuses; and if marked advance is in the future to be made as we think it should be made upon the results attained in 1911, we believe that this object will be secured neither by further puzzling the unaided householder by adding to the complexity of the schedule nor by the improvements still possible in tabulation which has already been brought to a high pitch of perfection, but by some radical improvement in the method

In this connexion it is interesting to note the two fundamental changes of method introduced with the census of 1841 Prior to that time the data had been both recorded and tabulated by local officials but from 1841 onwards these two duties have been transferred the first to the householder and the second to the census staff in London The second transference we regard as an unmixed benefit, and so far as it has remained incomplete in respect of the compilation of population returns from "enumeration summaries" we are of opinion that it should for the future be completed (see page 16). We believe however, that the transfer to the householder of the duty of record can be regarded as advantageous, if at all, only provided that the scope of the census inquiry is to be severely restricted. In the face of a demand for increase and improvement of census tabulation, such as precedes every census taken, the question should in our opinion be considered whether this duty of record should not be retransferred from the householder to the enumerator, or at least much more effective assistance and supervision provided for the former than he has hitherto received.

In the United States of America, which produces a more elaborate census probably than any other country, the duty of record has for many years been thrown upon the enumerator, but with the disadvantage, under the system there employed, that the compilation of the schedules for each enumeration district occupies a considerable time, two weeks to one month we understand, during which appreciable movement of population must take place, so that it would seem possible for some individuals to be enumerated in two or more localities and for others to escape enumeration altogether. For this reason we are especially attracted to the system employed in India, which we understand to be in broad outline as follows:—A definite census day is fixed, as in this country, to the state of affairs on which all the returns are to refer. During a short period prior to this date the enumerators visit every house in their districts, as do ours, but instead of merely leaving schedules to be filled in by householders (of whom a large proportion are illiterate) the enumerator by questioning the inmates of the house obtains the necessary information with regard to each of them, which he then and there fills in on the schedule for every person habitually living in the house. When the census day comes round he revisits the house, inquires who have slept there the previous night, strikes out the entries of any absentees and has to record on the busy day itself the facts only with regard to newcomers, who must form but a small proportion of the total population.

We believe that some such system as this would represent a great improvement upon our present method. Instead of trying to train the great mass of the public, as at present, through lessons to the children in the schools, trade union appeals, newspaper paragraphs, &c., it would only be necessary for each registrar to train his small staff of selected enumerators. If the work of registrars can be linked up, as we hope may ultimately be possible, with that of local administrative authorities, local patriotism might find a better opportunity than at present to provide efficient enumerators. Not only would the records be made under this system by persons familiar with the schedule and its requirements, but they would in every case be made by responsible paid officials, who could be called to account for avoidable omissions and imperfections, while the public would be spared the troublesome task of trying to puzzle out the census requirements. The chief objection to the system would probably be increase of cost owing to the increase of the enumerator.s duties. He would not only have to write down all the information (as at all censuses prior to 1911), but also to elicit it by obtaining an interview with each householder or some responsible representative of the householder. This would no doubt increase his work considerably, but all except the revision of schedules on census day could be done at any convenient time without interruption of his ordinary work. In any case it is not our purpose here to discuss the finance of such a change, but merely to point out the method which, we think should be adopted if the development of the English census is to continue

An incidental advantage of the Indian system is that it might be made to provide automatically the material for a tabulation not only of the population actually present on census date which alone is taken account of in our present census, but also of the number of persons habitually resident in each locality, as enumerated for many other states. To do this it would only be necessary for the enumerator at his preliminary visit to enter particulars for every habitual resident and for no others, the necessary additions and eliminations being made on census day in ink of distinctive colour.

Further points for consideration before it becomes necessary to make preparation for the next census will be those arising out of the present war. So great a disturbance has been caused in the distribution of the population as to invalidate all calculation of rates for the purpose of vital statistics. The results of the National Registration of 1915 are being utilized for the year 1915, but unless this is maintained, or some substitute in the way of a count of the population, or a census of a simple kind, put in hand at the conclusion of the war, it will probably be necessary to abandon any attempt to publish rates. It may be hoped that before the year 1921 there will have been some return to normal peace conditions, that industry will have settled down into some defined lines, and that emigration, if emigration, should be stimulated by the war, will have done its work. One result may certainly be anticipated, namely, reluctance on the part of the Government to spend liberally on a census however important the statistics obtainable from it (and they will certainly be more important and interesting than ever before) may be. It is to be hoped, therefore, that the experience gained from the National Registration of the notable assistance which can be rendered by voluntary workers in such an enterprise may bear fruit, especially as such assistance would only be required for the work of enumeration and not for tabulation.

Proposals for a Quinquennial Census. —While the present system of census enumeration is, in our opinion, scarcely adequate to all the requirements of an elaborate population census, such as it is the policy of this country to carry out once in ten years, it is quite conceivable that it would be found needlessly expensive for a simple count of the population in the middle of the inter-censal decennium, to which, desirable in itself for many reasons, the sole objection appears to be the expense involved. In this connexion the proposals of a year or two ago with regard to the automatic registration of voters are of great interest. If such a method of parliamentary registration should ever become law it might be considered whether the returns involved should not be extended to embrace all householders, who might in certain years at least be required to state the age and sex of each member of the household. Whatever method of securing information as to voters was adopted this addition could add but little to its difficulty or expense, and would provide the means of tabulating the number, sex and age of population, not only in Parliamentary areas, but in administrative or any other areas with which it was desired to deal. The maps and other information as to the relations to each other of various types of area, which have to be constantly kept up to date even at present, would provide the means of sorting the forms received for any parliamentary area according to the administrative and other areas to which they referred.

In the case dealt with it is unnecessary to consider means of distribution and collection of forms, for the methods adopted would be governed by the requirements of their primary purpose, the suggestion now made being merely that a slight extension of whatever scheme was adopted would provide most economically for a simple form of census at frequent intervals as a by-product. Failing such a convenient opportunity of cheap enumeration we believe that one or two other plans for carrying out the work without incurring the expense of a specially created staff may be worthy of consideration, though we recognise that the circumstances of the present year have made it impossible to arrange for a census in 1916.

The first possibility that suggests itself is that of delivery and collection of schedules through the post. We do not claim to have worked out a complete scheme on these lines, and objections readily suggest themselves. We believe, however, that most of these could be overcome if the goodwill and co-operation of the authorities concerned could be secured.2 As an alternative scheme, local administrative bodies might be encouraged to carry out enumerations in their own areas on a fixed date, the necessary forms being provided and the tabulation carried out by the Census Office. The local authorities are in a favourable position for organising and carrying out such work, and in fact are in the habit of doing a portion of what is required yearly in the revision of their rate-books. The staff in their own employ would make an admirable nucleus of that required, and indeed if the police be included might in many cases suffice for the whole work. Such a system, which would provide a census for those areas desiring to have it and willing to contribute to its cost, but not for others, would have the disadvantage of providing no enumeration for the whole country. In so far, however, as the more important areas whose populations are the most difficult to estimate, joined in the scheme the estimate made for the unenumerated portion of the country would be improved, and it may be hoped that as enumeration at local expense is already not unknown in this country progressive areas would in many instances join in a scheme organised and supported by the Government.

On some such lines as these it seems to us there may be a possibility in future of devising plans for a simple intermediate census by which the sole objection to the proposal its cost, may be overcome. No attempt has been made as yet to fill in the details of any of the shadowy outlines sketched, but we hope that their possibilities and those of any others which may suggest themselves will be seriously considered before it is decided ten years hence that the old and frequently repeated demand for a quinquennial census must once more be refused. If, however, it were a question of deciding between the improvement of the decennial census and the provision of a simple quinquennial enumeration we should have little hesitation in preferring the former, believing that it is more important to improve the quality than the quantity of our enumerations.

4. TABULATION.

The means employed for tabulating the information on the schedules differed very materially from those in use on former occasions, and represent, it is believed, a great improvement upon them. A description of the apparatus employed and the method of its use will be found on pages 259 to 362. The fact, already referred to, that the work was carried put from the schedules themselves instead of, as hitherto, from copies of the schedules in enumeration books caused on the whole less inconvenience than might have been anticipated and had the advantage of eliminating one very considerable source of error. The schedules were received tied up in cases in book form, one case or book for each enumeration district, and were kept in this form throughout, a fireproof building with metal racking to accommodate the 35,000 books being provided.

The first step after the books had been arranged was to classify, and record in numerical form, where it did not already possess it, the information contained in them by the assignment of a code number to each item of information to be tabulated. These code numbers represent the smallest and most detailed subdivisions used in the various tabulations. Some of the items to be tabulated did not require any special coding process either because, as in the case of age, number of rooms in tenement, duration of marriage, and children born to a marriage, they were already in numerical form, or because, as in the case of sex and conjugal condition, the alternative possibilities were so few that no special coding process was required.

The two chief coding processes were those for birthplace and for occupation, including industry. The latter, by far the more difficult and laborious of the two, formed a single process because the code was so arranged that the number for the occupation represented tlie industry as well in all cases except when the industry differed from that most commonly associated with the occupation in question, in which case a separate industry number was added. The relative importance of this part of the work may be judged from the fact that of a total sum of nearly 9,000 expended upon clerical work on the preparation of the occupation and industry tables, no less than 4,000 was for coding alone. The staff engaged upon it was distinct from that coding birthplaces, and both staffs specialised upon their particular work.

In addition to the subject codes, area codes were arranged by which an identifying number was provided to represent each area dealt with; of whatever type.

When all the code numbers required had been inserted in a book of schedules and the necessary checking of this work had been completed, the book passed into the hands of another staff by which the coded information was transferred to cards as described on page 259 by means of punching machines which punched out holes in the cards in the appropriate numbered positions. These cards, one for each individual, were then checked to ensure that they had been punched correctly and afterwards tabulated by the special machinery provided.' Of the two types of machine used for this purpose one, the sorting machine, assembled the cards which had been similarly punched for any of the subjects dealt with, say, all of a given age or occupation in any area (the distinction of sex and area was maintained throughout, male and female cards for a given area being kept by themselves in appropriately labelled boxes) and the other machine then counted these cards, providing at the same time sub-totals of cards similarly punched in regard to some other matter for which they had not been sorted. Thus in preparing the tables of detailed age and conjugal condition in combination (tables 1-7, Volume VII) the cards for each district might have been sorted so as to assemble those for each individual year of age and then the cards for each age counted, the machine showing sub-totals for the single, married and widowed. As a matter of fact, however, it was found more convenient to sort first for conjugal condition and then count batches of cards of the condition and of the same decade of age, the machine showing sub-totals for each individual year in the decade. The tables, though published in this detail only for large sections of the country, were obtained by aggregating district tables similarly prepared, and thus the more detailed tables are frequently available in manuscript form for many more areas than those for which it has been possible to publish them. It was necessary in the case in point to prepare such tables in full detail for every administrative area because tables showing age in less detail had to be prepared for individual areas and the same count had to serve both purposes,

It was found impossible to provide, on a single card of the size used, for all the subjects To be tabulated, and the facts as to the fertility of marriages were therefore punched on another sot of cards, one for each married couple. (See page 261.)

In addition to tabulation from the schedules as described, certain other tables, those, namely, showing numbers of males and females in areas of various types, Volumes 1 to IV, mid the foibles relating to buildings, Volume VI, were prepared from the enumerators. summaries described on page 10. These summaries give the population in each numeration district (or portion of it undivided by any area boundary line) made up by the enumerator and cheeked by the registrar, and from them the preliminary report on the census is prepared. It was originally our intention not to publish any portion of the final report until all these population totals had been thoroughly checked by count of, the individual cards punched from the schedules. The preparation of the card, however, with the necessary preliminary processes of coding and scrutiny involved, occupied more time than had been allowed for, and it was deemed inadvisable to delay publication until the cheek in question had been carried out for all areas. Sufficient cards (about a quarter of the whole) had been prepared and counted before Volume I was published to show that the errors brought to light in the enumerators' totals, as amended after scrutiny in the Census Office, were quite inconsiderable. When this fact had been thoroughly established it was decided for the remainder of the population to accept the summary book totals without further check, and from these, cards were prepared for each enumeration district or portion thereof not situated in more than one area of any given type showing the inhabited houses and population and the area of each type dealt with in which the district was situated (see page 260), When these cards, about 02,000 in all, had been prepared, it was a simple matter to assemble those relating to a parish, an urban or rural district or a diocese, &c., on the sorting machine and count the number of dwellings and of males and females on the tabulating machine provided for the purpose (page 261) which adds not units only but numbers of any size as punched on cards relating to aggregations of individuals.

The tabulation of buildings was also carried out from the enumerators' summaries, in which, as already explained, they had entered an account of every building in their districts. In this case the summary book was the original and only source of information, so no question as to the propriety of its use could arise.

As to the compilation of population tables from the summary books, while the completion of the card count showed that the assumption was correct that no errors in the summary totals of any appreciable magnitude would be brought to light, we feel that the necessity for the course taken is to be regretted and that it should, if possible, be avoided in future. The only certain method ("speeding up" in the production of the cards being of doubtful benefit and uncertain of attaining the desired result) seems to be a reversion to the form of publication, followed in 1901, when all the tables for each county were issued in a separate part as the counties were completed. This plan, which was followed in publishing the report on the Scotch census of 1911, admits of all the checks being utilised which any one form of tabulation can provide in regard to another before any figures are published at all. The county part is a less suitable form of publication for many purposes, though obviously more so for local needs. As the latter, however, have been, catered for on the present occasion by subsequent re-grouping of the tables, originally published in subject volumes, into county volumes, the only question at issue is which class of reader shall have his requirements met first, and in regard to this we think that the very great advantages of the locality form of publication from the producers' point of view must in future lead to a decision in its favour.

In regard to this matter as to many others the new methods of tabulation employed rendered the experience of previous censuses of less value than usual. To some extent the experience now gained and which will be available in the future has had to be bought, though, we have much reason for satisfaction with the successful operation of plans produced in the light of a merely theoretical knowledge of the mechanism to be used. In order that full advantage may be taken in future, even by an altered personnel, of the experiences now gained, reports have been prepared by all those who were in charge of important processes in which are recorded their impressions of the manner in which the work has been carried out, with their suggestions for future improvements. The collation and study of these notes should form an important process in preparing for the census of 1921.

5. STAFF EMPLOYED.

The small permanent nucleus of the census staff is engaged during the intercensal period in recording the numerous changes of boundaries that are made from time to time in the several kinds of administrative, registration and ecclesiastical areas, and in ascertaining the populations affected by such changes. Early in 1909 this small staff of three clerks was augmented by the appointment of six more, who were engaged mainly on the preparation of the set of maps referred to on page 10, and partly on experimental tabulation. By the end of the year the number of temporary clerks was increased to 18, and in the following April, when the work of examining the registrars' old "plans of division" (i.e. , the written description of the several enumeration districts within the sub-district) was begun with a view to making recommendations for their improvement, the number of clerks reached 30, and was not again increased until after census day. Meanwhile provision for superintendence of the temporary staff and for general administrative work had been made by detaching from the permanent staff of the General Register Office six clerks who had had previous census experience.

During the three months following census day the number of male clerks was increased to a total of 66, who were engaged mainly at first on the preparation of the Preliminary Report, and afterwards on the revision of the schedules and enumerators summary books. Towards the end of this period a staff of 30 female clerks (under the superintendence of five permanent women clerks kindly lent for the purpose by the Postmaster-General) was engaged primarily in order to check the work of punching the cards which was to be done by girls. These checkers at first took over some of the revision of the schedules, thus gaining the necessary familiarity with the form and releasing some of the male staff for the "coding" processes described on page 15.

By the end of September, 1911, the male staff had been increased to 84 clerks, and 16 machine-boys, the latter being employed (under the supervision of the clerks selected to record the results) in attending to the machines then being installed. By the same date the number of women checkers had been increased to 67 and the girl punch-operators to the number of 79 had been engaged.

Before the end of 1911 the staff was at its maximum, the principal additions to the foregoing numbers being among the machine boys and the punch operators whose numbers were raised to 32 and 140 respectively. Thus during the period of greatest activity, which extended up to the middle of 1912, there were altogether about 350 persons employed, comprising 90 male clerks and their superintendents, 81 female checkers and their superintendents, 32 machine boys, 140 punch operators, and a small staff of porters and messengers.

From June, 1912, the numbers were gradually diminished as the various processes approached conclusion, and by the end of June, 1913, were reduced to about one-sixth of the maximum number. The whole of the female staff had been disbanded by this date, many of them having been absorbed in the offices of the National Health Insurance Commissioners where the experience in the handling of cards gained at the Census Office Was found to be most useful.

At the end of June, 1913, the staff, then consisting of 36 second division clerks and 19 machine boys, was employed mainly in preparing for the press the volumes of statistics of occupations and industries, birthplaces and infirmities, and in the machine processes in the tabulation of fertility statistics. Successive reductions brought the staff in the middle of 1914 down to eight second division and three assistant clerks, but of these two were accepted for military service, and one was lent to another public department before the end of the year. All census work, with the exception of the final stages in the tabulation of the fertility statistics, which had to be completed on account of the contract for the hire of machines, was practically suspended shortly after the outbreak of war. Those members of the staff not actually engaged on the fertility tables have since been more or less fully occupied in matters arising out of the international situation, e.g. , the preparation of statistics for the Committee for the Prevention and Relief of Distress, the Central Register of Belgian Refugees, and the National Register.

It may be observed that of the large temporary staff only the second division clerks, the assistant clerks, and the superintending women clerks were permanent members of the Civil Service. The women checkers, who were required to be between the ages of 18 and 21, were obtained principally from among candidates for the women clerks' examination, while the punch operators, who were under 16, were recruited mainly from the elementary schools in the neighbourhood of the office, and the machine boys through the schools and the labour exchanges. On the whole the juvenile labour obtained in this manner proved to be quite efficient, and we feel satisfied that, in view of the mechanical nature of the tasks to be performed, it was the most advantageous course that could be followed.

6. COST OF THE CENSUS.

Cost of Collection of Data, —The remuneration of the whole body of local official superintendent registrars, registrars and enumerators—devolved upon the Central Office, the system in this country differing therein from that in vogue in some other countries where the cost of enumeration is defrayed by the local administrative bodies. The scale of fees payable to local officials did not differ materially from those of other recent censuses. In view of the reduction of their work, owing to the decision to tabulate direct from the schedules instead of from copies made by the enumerators (see page 10), the question of reducing their fees was discussed, but having regard to the dissatisfaction expressed at the census of 1901, it was felt that no substantial reduction could be made without prejudicing the quality of the work. Enumerators were paid a fixed fee, irrespective of the area and population of their districts, and an additional fee rising proportionally in all districts in which the population exceeded 400; the majority of the enumerators—about three-quarters of the whole—received payments on this account. Allowances were also made in cases where the distance necessarily traversed in collecting the schedules exceeded 6 miles. Altogether the 35,000 enumerators received just over 80,000, of which 4,000 was paid in respect of mileage allowances, as against 77,000 (of which 5,000 was due to mileage) in 1901, when the number of enumerators was approximately 38,000. The diminution in the number of districts which was found to be possible in spite of the increase of population was due to the more compact arrangement which was facilitated by the use of the maps referred to (page 10); these maps also enabled a complete check to be made on the claims for mileage allowances, with the result that a fairer and more economical distribution of the allowances was secured. Registrars were paid, as before, a fixed fee with a proportional addition for population in excess of 1,200, and an innovation made in 1901, viz., a bonus for efficient work, was repeated. The fixed fee was raised, and the total amount paid, exclusive of bonuses, increased from 23,500 in 1901 to 27,300 in 1911; on the other hand, the bonus payments which had been 7,600 in 1901, decreased to 5,700 in 1911 as a result of the more stringent checks which were applied to the work of the registrars; thus these officers, whose numbers were approximately equal at the two censuses, received nearly 2,000 more on the latter occasion. No change whatever was made in the remuneration of Superintendent Registrars, and the slight increase under this heading, from 6,500 to 6,800, was due to the increase of population during the intercensal period. In all, fees to these three classes of local officers totalled to 114,500 in 1901 and to 120,000 in 1911. The only other considerable items connected with the collection of the census data were the printing of the schedules and forms, and their carriage from and to the Census Office, which amounted to 2,054 and 1,242 respectively.

Total Cost. —The cost of collection has always been the most considerable part of the total cost, thus, for 1901, the whole amount expended under the Census Vote for England and Wales was approximately 149,000, of which, as already shown, 114,500 was expended in payments to local officers, while for 1911 the amounts were 162,000 and 120,000 respectively. The following comparative statement shows the cost of successive censuses from 1851 onwards, exclusive of the expenditure on printing, stationery, maps, &c., which is not available for the earlier enumerations.

Date. Population enumerated. Expenditure
under Census Vote.
Cost per 1,000
of population.
    £ £ s. d.
1851 17,927,609 93,132 5 3 11
1861 20,066,224 95,719 4 15 5
1871 22,712,266 119,977 5 5 8
1881 25,974,439 122,876 4 14 7
1891 29,002,525 120,599 4 3 2
1901 32,527,843 148,921 4 11 7
1911 36,070,492 161,900 4 9 9

In comparing the cost for 1911 with that for previous years, it is necessary to bear in mind the additional stationery office expenses for hire of tabulating machines and purchase of cards incurred for the first time at the present census. By the inclusion of these and all other charges the whole cost of this census is raised to 196,000, against 165,000 for 1901, or to 5 8s . 8d per 1,000 of population, as compared with 5 1s. 5d., this increase in cost being we believe, fully justified by the more comprehensive information which the new method of tabulation has enabled us to furnish. From these amounts should be deducted 899 and 834 respectively, representing the cost of supplying additional statistics to local authorities in terms of section 9 of the Census Act, these sums having been repaid to the Treasury.

For the past three censuses the figures may be analysed so as to show the amount expended in each financial year. Dealt with in this way, they emphasise the increasing importance that has been attached to the preparatory stages of the work. Thus, prior to census day, 1891, the sum of 543 was expended in salaries at the Central Office; before census day, 1901, the amount was 1,357; while for the census of 1911, the preliminary expenditure on salaries was no less than 3,985.

Comparisons with foreign countries may be of some interest, but it was not always clear from the replies to the inquiries which were addressed to foreign census bureaux on the subject whether the figures given are inclusive of printing and stationery charges, and in some countries only the central office expenses are charged to the national funds and the total cost could not be ascertained. The following figures of the total cost, however, appear to furnish a valid comparison:—

Country. Census year. Population. Total cost
of Census.
Cost per 1,000
population.
      £ £ s. d.
England and Wales 1911 36,070,492 196,001 5 8 8
Do. do. 1901 32,527,843 164,916 5 1 5
Austria 1900 25,921,671 148,809 5 14 10
Belgium 1900 6,693,548 29,633 4 8 6
Italy 1901 32,475,253 112,919 8 9 7
Netherlands 1900 5,104,137 36,334 7 2 4
Norway 1900 2,240,032 18,460 8 4 10

The total cost of the last census in England and Wales may be summarised under the following headings:—

  £ £
1. Fees and expenses of Local Officers   120,094
  (a) Superintendent Registrars 6,832  
  (b) Registrars 33,036  
  (c) Enumerators 80,226  
 
2. Sales, &c., at Central Office   39,987
  (a) Male clerical staff 26,990  
  (b) Females clerical staff 6,236  
  (c) Machine boys and punch operators 5,146  
  (d) Office keeper's staff (porters, charwomen) 1,615  
 
3. Incidental expenses   27,120
  (a) Carriage parcels 1,242  
  (b) Other office expenses 295  
  (c) Supply of stationary 8,128  
  (d) Preparation and supply of maps 3,650  
  (e) Hire of machines and purchase of cards and punches 13,805  
 
4. Cost of publication of results   8,800
 
  Total   196,001

7. DIAGRAMS.

A special feature of this Report is the use of diagrams, now become essential in the presentation of statistical information. It is of interest to note that the Census Report of 1851 contained a few beautifully executed maps and diagrams, but that since that date nothing of the kind has been published in connexion with a Census. We stated in the Report to Volume I of the present census (Administrative Areas issued in 1912) that we intended to issue separately a volume containing maps and diagrams illustrative of the figures in some of the Reports, but we have now decided to embody these in the present volume. Some of the diagrams have already appeared in Volume VII of this series (Ages and Condition as to Marriage), but with this exception the diagrams are now published for the first time.

8. ASSISTANCE RENDERED BY GOVERNMENT DEPARTMENTS AND OTHERS.

Before bringing these introductory observations to a conclusion we desire to place on record the assistance which we received from various quarters during the course of the Census of 1911. We shall first mention the Government departments, other than the Local Government Board which had the general control and supervision of all the arrangements, and the points on which we consulted them. With the Board of Trade our consultations were specially close owing to the importance of endeavouring to make the statistics relating to occupations conform as far as possible to practical requirements. We have also to thank the General Post Office for assistance in connexion with the loan of Lady Superintendents for the female staff.

The Admiralty assisted by making arrangements for the enumeration of officers and men of the Royal Navy and Royal Marines; the Board of Agriculture by the circulation of an instructional memorandum to agricultural workers with reference to occupations; the Board of Customs and Excise by undertaking, through their officers, the enumeration of persons on board mercantile and fishing vessels in port at the time of the census; the Board of Education by providing instruction to school children in the objects of the census, and in the method of filling up the schedule; the Home Office, by obtaining the co-operation of the various police forces in the enumeration of the vagrant population; the Ordnance Survey Department by the revision of the acreage of the several administrative, parochial and parliamentary areas, and by the preparation of the invaluable series of maps described on page; the Seamen's Registry by advice as to the allocation of the shipping population; and the War Office by making arrangements for the enumeration of the Army at home and abroad.

We were further indebted to the Parliamentary Committee of the Trade Unions Congress, who were good enough to issue a valuable instructional memorandum with reference to occupations, to the Trinity House for the enumeration of lighthouse keepers, to the clergy and diocesan registrars for assistance in defining parish boundaries; to the Urban and Rural Councils for the revision of proofs of the houses tables; to representative bodies of Superintendent Registrars and Registrars of Births and Deaths for attendance on several occasions at Somerset House for the discussion of preliminary arrangements; to the London County Council on whose behalf Mr. E. J. Harper kindly advised us on the tabulation of tenements and buildings and Mr. Cyril Jackson on the tabulation of occupations with special reference to age detail for "blind alley" occupations; and finally to the Royal Statistical Society whose Census Committee made several important recommendations with which we were for the most part in hearty sympathy.

Impressed with the necessity of consulting outside experience in the conduct of a census, we familiarized ourselves as far as possible with foreign and colonial census reports, and personally visited both Paris and Washington with the object of studying census methods of tabulation on the spot, with a special view to the introduction of mechanical tabulation. We invited the co-operation and counsel of several eminent statisticians and economists, and desire to offer our thanks for their assistance to Mr. A. L. Bowley, Mr. Udny Yule, Mr. Edwin Cannan, Dr. Heron, Sir Timothy Coglilan, the Right Honourable the late Charles Booth, and also to Mr. George King, whose actuarial work in connection with age statistics we have already acknowledged in the return dealing with that subject. We must not omit to record our obligations to the Press, which never failed to respond to our appeals for assistance in arousing interest in the Census, and impressing on the public the importance of conforming to the instructions issued.


1 In this connexion it is appropriate to note that, since the year 1911, the administrative area has been substituted for the registration area as the local unit of tabulation for births and deaths in the Registrar-General's Reports. (See Introductory Remarks 74th Annual Report, cd. 6578, 1913.)

2 It is interesting to note that a proposal for postal enumeration in the United States has been put forward by the Director of the last census in that country (see Publications of American Statistical Association, Vol. xiii, "The Census Methods of the Future").

Next Selection Previous Selection