Introduction
Next Selection Previous Selection
The thirteenth Census of the population of England and Wales was taken on 19th June, 1921. This volume gives an account of its working, the difficulties encountered and overcome, and comments on the recorded results. Very full accounts of the earlier censuses of this country are given in the General Reports on the censuses of 1901 and 1911, and it is therefore not considered necessary on this occasion to repeat this retrospection. (a) Legislation.
Legislation was, of course, necessary to confer the powers under which the census was taken. Unlike all previous Census Acts, each of which had been limited in. operation to a single census, the Census Act, 1920, is a perpetual Act having application not only to the recent census but also to all future censuses in Great Britain. Further, it enables a census to be taken at quinquennial intervals and provides for the taking of local censuses at any time at the request and charges of the local authority for the area concerned. It also contains provisions as to the supply of statistics in intercensal intervals with a view to their being brought into closer relation with the periodical census statistics in pursuance of a common statistical policy. As a natural consequence of the general form of this Act, it reserves the character of the enquiries to be included in the census on each successive occasion for determination, subject to certain conditions, by Order in Council, power being also conferred to make regulations covering the requirements of administrative machinery. In pursuance of these provisions, an Order in Council was made on 21st December, 1920, prescribing the date of the Census (24th April, 1921), the persons by whom and with respect to whom census returns were to be made, and the nature of the particulars to be furnished in those returns. Regulations were also made on the same date prescribing the procedure of enumeration and the forms of return. (b) Preparation of schedule and method of presenting results.
Considerable preparation had, however, been previously necessary. Active arrangements commenced to be made at the end of 1919, when steps were taken in particular to consider the nature of the enquiries to be adopted and to secure the important object of co-ordination between the census returns of the several parts of the United Kingdom. For this purpose the Ministers responsible for the census in the several parts of the United Kingdom appointed by arrangement a Census Joint Committee of three officers, and to this committee was entrusted the task of considering the nature of the census enquiries with a view to their rendering the fullest measure of service to the common requirements of all parts of the United Kingdom and to the attainment of a maximum degree of comparability in the resulting statistics. The numerous enquiries proposed or submitted for inclusion in the schedule were grouped according to subject matter, and for each group an expert Sub-Committee was founded on the administrative basis of the Joint Committee to examine and advise upon the questions involved as affecting the United Kingdom as a whole. As the schedule already appeared to have reached the extreme limits of its capacity expansion, the problem presented was to decide upon the selection of enquiries which promised results of greatest general utility for present and future needs. As a result of this process' of sifting, and after numerous consultations with authorities, scientific bodies, and industrial undertakings, the final contents of the schedule were settled for submission to Parliament in the draft Order in Council and Regulations. A copy of the schedule prescribed for general use in England is shewn to Appendix B. Reference may be made to those respects in which the schedule thus adopted differed substantially from the schedule of 1911. An enquiry was added as to the number and ages of children under 16 (including an enquiry as to orphans), in view of the increasing importance for many administrative and public purposes of statistics as to the extent of the burden of dependency upon different sections of the community. The original estimate formed in 1920 as to the prospective value of this enquiry has been fully justified by the service which it has rendered in the preparation of the financial framework of the Widows', Orphans' and Old Age Contributory Pensions Act, 1925, and in the material which it has afforded for the use of several important public investigations. An enquiry as to place of work, which was felt to be of great value for transport, housing and general industrial purposes, was also added. On the other hand, it was decided to omit the enquiry as to "infirmities" included in previous censuses, in view of the generally recognised fact that reliable information upon these subjects cannot be expected in returns made by or on behalf of the individuals afflicted. Further, it was concluded, after very careful examination, that the "fertility" enquiry of 1911 (viz., as to duration of existing marriages and the number of children born of such marriages) could be omitted in 1921, notwithstanding its importance, with less disadvantage than either of the new enquiries proposed, particularly in view of the long range covered by the 1911 enquiry and of the fact that the wealth of material which it provided had not been completely exhausted. This is the first time in the modern history of census-taking in this country that any enquiry once introduced into the schedule has been omitted therefrom on a subsequent occasion. The fact is indicative of a stage at which the limits of expansion have been approximately reached, and a new problem presented to the census authorities. Special attention was given to the enquiries as regards occupation and industry with a view to securing an improvement in the statistics derivable therefrom. It appeared clear from experience of the 1911 results that a fuller and more scientific classification, both of occupations and of industries than had hitherto been available was essential to the proper statistical treatment of this subject, and to the utility of the census results for comparison with the Census of Production. The Census Joint Committee accordingly arranged for the subject to be considered by a special Sub-Committee, assisted by representatives of the Board of Trade, Home Office, and Ministry of Labour. This body, working with the co-operation of the Departments mentioned, drew up occupational and industrial classifications which have been adopted for the purpose of the census returns and it is understood that those departments have expressed their intention of conforming to these classifications for the purpose of their own departmental statistics. Thus the actual achievement in this respect has gone beyond the immediate object of facilitating the census work and has secured a valuable advance in the co-ordination of official statistics by providing standard classifications to which the departments mainly interested have agreed to adhere. The second stage in the work of the Census Joint Committee was the decision of the form of the tables to be published by way of presentation of the results of the census, so far as this involved questions of comparability and co-ordination between the several parts of the United Kingdom. Practically every government department was consulted with a view to the census statistics being rendered as useful as possible; and suggestions were invited with the same object from the organisations representing local authorities. On the basis of the conclusions thus arrived at, the detailed programme of tabulation operations at the Census Office was then prepared. The Imperial Statistical Conference in London in the early part of 1920 afforded exceptional opportunities, of which full advantage was taken, of consultation with the statistical officers of the Overseas Dominions on the subject of census-taking within the Empire. As a result of the discussions which took place, agreement was reached with regard to the major points upon which uniformity of action within the Empire is desirable, and plans were concerted to secure that common Imperial requirements should be, as far as possible, observed in the results of the separate censuses of the several Dominions, Colonies, and Protectorates which were due to be taken in the year 1921. In view of these efforts to promote improved co-ordination within the United Kingdom and within the Empire, it was a source of particular regret that conditions in Ireland were ultimately found to render it impracticable to proceed in that country with the census as originally planned. (c) Preparatory work at Census Office.
The administrative procedure of enumeration did not differ substantially from that previously adopted in recent years. The country was sub-divided by the 1913 local Registrars of Births and Deaths into 38,563 enumeration districts and for each district (with certain exceptions) an Enumerator was appointed on the recommendation of the Registrar to undertake the distribution and collection of schedules. Officers of H.M. Customs and Excise enumerated persons on board mercantile shipping, fishing vessels, etc., and the Corporation of Trinity House made the requisite arrangements in the case of lighthouses and lightships. Returns of homeless persons were obtained by the police. Registrars were, however, instructed directly from headquarters, and not, as on previous occasions, through Superintendent Registrars. This change of plan was made largely in the interests of economy. It did not appear that the inclusion of Superintendent Registrars in the scheme of organization could be justified by the essential requirements of the executive operations, although, as anticipated, the change imposed an increase of active responsibility upon the headquarters staff. The regulations provided, however, for the appointment in each area of a Census Advisory Officer; and Superintendent Registrars, with few exceptions, accepted these appointments and acted in an honorary capacity. The Department thus had the inestimable advantage of the local support of Superintendent Registrars as a measure of precaution against any local emergency; and while no emergency arose which necessitated any serious demands being made upon them, their advice and support in many instances was of great assistance. Reference may also be made to one or two novel features in the procedure. For the first time definite provision was instituted to enable separate confidential returns to be made by those persons who would otherwise have suffered hardship by disclosing particulars to some other person charged with the duty of making the return. Consequential arrangements had, however, to be made to enable the separate return to be subsequently associated with the household return in which it should normally have been included, in order that statistics based upon the household unit might not be vitiated by the concession. Substantial administrative machinery had also to be devised to give statistical effect to the new "place of work" enquiry. While the scheme of enumeration itself ensures that the places of enumeration of the population on the census night are automatically allotted to the several local sub-divisions of the country (boroughs, urban or rural districts, wards, and civil parishes, etc.) in which they are situated, no statistical expression could, of course, be given to the place of work addresses until a similar allocation had been made and the local sub-division accurately identified in the case of each address. It was not to be expected that the officers in charge of the enumeration, though familiar with the topography of their own districts, would be possessed of the precise and expert knowledge of localities and boundaries in other and, often, distant parts of the country to permit of their correctly assigning all place-of-work addresses entered in the schedules passing through their hands. Nor, indeed, without enormous expense (if at all) would it have been possible to have collected and trained a staff at headquarters which would have been competent to identify in all cases the wards or civil parishes within the boundaries of which every place of work throughout the country was situated. The following procedure was accordingly decided upon. The local registrar receiving the returns from enumerators was required to identify and code the local area of each place-of-work address within the district for which he was responsible; but with regard to other place-of-work addresses, it was arranged that a simple Postcard form should be written bearing the address of the place of work and the reference number of the census schedule upon which it was entered and an arrangement was made with the General Post Office whereby these postcards, when posted, were delivered to the registrar for the locality in which the place-of-work address was situated. This Registrar having the required knowledge of the boundaries and local sub-divisions in his own district was, of course, competent to make an accurate assignment of the address of each postcard received by him, and it was arranged that the registrars receiving such postcards should code them and transmit them to headquarters. On receipt at headquarters the postcards thus coded were sorted back according to the districts whence they originated with a view to their ultimate association with the schedules to which they respectively related. This procedure appeared to offer the only means whereby full advantage could be derived from the invaluable material afforded by the place-of-work enquiry. It involved, of course, large-scale operations in connection with the many millions of place-of-work addresses, and detailed preparations had to be made to provide for the various contingencies which were bound to arise. In the subsequent course of the tabulation of this material it was found necessary on financial grounds to restrict the range of areas for which work-place figures were given to those of the order of urban and rural districts and upwards. Had this originally been in contemplation it might have been possible to have adopted other means of procedure but in the circumstances no other course was available by which the full information originally deemed necessary could have been obtained. All the foregoing arrangements had been brought to a state of completion with a view to the census being taken on the 24th April. Conditions occasioned by the coal dispute and the expectation of a strike of railwaymen and transport workers gave rise, however/to serious doubts as to whether the enumeration could be successfully carried out in all parts of Great Britain during the period originally fixed; and in view of the heavy loss which would have resulted had the enumeration proved abortive, a postponement was decided upon. A draft Order in Council to give effect to that decision was accordingly submitted to Parliament on the 14th April, and on the 25th April a further draft Order was submitted substituting the 19th June for the date previously prescribed. The new date decided upon was the earliest date of which the requirements of Parliamentary procedure admitted it was, on the other hand, the latest date before the commencement of the important series of public and industrial holidays which continues well into the autumn. Any census taken during those holidays would have substantially misrepresented the distribution of the population in the areas affected, It was, in any case, inevitable that some change in the distribution of population would take place between the original date and the new date, and the statistical consequences of this change are commented on later in this report. Regulations were made on the 27th April to define the position of Enumerators who had undertaken to serve on the basis of the earlier date, and to afford them an opportunity of obtaining release from their obligations. Arrangements were made to fill the places of Enumerators who took advantage of this option, and in many other respects to bring the administrative machine to a state of readiness for the new date in the altered circumstances resulting from the postponement. The enumeration accordingly took place on the 19th June, and the collection of the returns was satisfactorily concluded in all parts of the country. A census was taken in the Isle of Man and Channel Islands on the same day. The provision of the requisite schedules was undertaken by this department, the actual enumeration being carried out by the respective Island Governments. A census was simultaneously taken of naval, military, and air force establishments abroad. It was necessary, in order to ensure that the subsequent coding and tabulating processes should be carried out smoothly and expeditiously to decide the method and form of publication well before the date of enumeration. With regard to the form of the reports, some choice of alternative lay between the method of publishing a separate volume for each field of subject matter for the country as a whole, and the converse method of "county volume" publication, viz., of presenting in respect of each county separately the relative statistics on a number of different subjects. In some cases the nature of the subject matter or the conditions governing its presentation precluded any option, but, broadly speaking, it was necessary to decide between publication in the first instance by subject matter volumes, the contents of which could, if necessary, be subsequently broken up and regrouped in county volumes and publication in the first instance by county volumes, the contents of which might similarly be regrouped and republished according to subject. The course adopted in respect of the 1921 Census Reports Was to proceed by the publication in the first instance of county volumes. The requirements of the tabulation procedure in any event entailed the building up of figures on any subject for the whole country out of the figures for each locality taken in turn and it seemed desirable that the figures for each county should be made available for local purposes as soon as possible. A full list of all 1921 census publications is printed on the cover of this report, while Appendix C contains a subject index to the various volumes. Presentation by area-the method of local distribution of the figures relating to any given subject—constitutes one of the most important elements in the tabulations, because common to them all. The general principle necessarily adopted is to increase the elaboration of the subject matter in inverse ratio to the degree of local sub-division. Thus, for the smallest sets of areas—which are necessarily the most numerous—the simpler sub-divisions of the subject matter only can be shown while for the larger sets of areas a far greater wealth of detail can be given in the form of the combination of subject matter and elaboration of their classifications. Naturally, account cannot be taken in the census statistics of any classes of areas save those which cover the whole country and which have a basis of statutory authority for the purposes of various forms of public administration. A statement of the several sets of local areas for which figures are separately shown will be found in Appendix C above-mentioned. Others not represented in the census tables, such as Lieutenancy Sub-divisions, Coroners' Court Districts, Highway Districts, Polling Districts, Relief Districts, Archdeaconries, etc,, either are aggregates of the basic areas for which figures are given, or serve purposes having too little connection with population figures to have given rise to a sufficient claim to be separately represented. Others, again, such as the areas defined by local Acts for the supply of gas, water and electricity and for drainage purposes, possess a special and limited interest, and can best be dealt with, should occasion arise, by means of private arrangements with the interested parties in accordance with the provision made by the Census Act to meet such cases. (a) Duties of local officers and issue of Preliminary Report.
During the weeks succeeding the date of the census, the local officers subjected the schedules and enumeration books to a careful scrutiny, which frequently involved reference back to the person making the return. In addition to these duties, "place-of-work" postcards were received by them from other districts, and were coded according to the area of the place of work by stamping on each the appropriate code number furnished by the Census Office. In the larger business centres, e.g., City of London, this particular duty was extremely onerous, although arrangements were made to lighten the burden by transferring some of the actual coding to headquarters. In view of this and of the novelty of the scheme, a tribute is due to the way in which the local officers responded and performed their allotted functions. After the examination of the enumeration books and schedules was completed, they were forwarded to headquarters. The first step in post-censal activities at the Census Office was the preparation of a Preliminary Report. The tables comprised in this volume consisted mainly of population figures (by sexes) of (a)
administrative areas, i.e., each county, borough, urban and rural district, and (b)
parliamentary constituencies. The rapid compilation and publication of these tables had been rendered possible by the summarising of returns furnished by the local registration officers, and although the figures obtained were provisional in character, no material discrepancy manifested itself in the subsequent statistical operations at the head office. In this connection it is interesting to observe that the finally ascertained population total for England and Wales of 37,886,699 differed only by 1,457 or .0038 per cent. from the preliminary count. This difference compares with .0132 per cent. at the census of 1911 and .0054 per cent. in 1901. The Preliminary Report was published on 23rd August, 1921, just over two months from census day. The special procedure adopted for the purpose of the Preliminary Report could not, however, be applied to the main contents of the census returns, each entry in which had to be separately examined and classified before the information could be expressed in statistical form. (b) Tabulation Procedure.
The problem of the tabulation of the raw material obtained on census schedules and of the presentation of the results has always been one that demanded careful consideration. Prior to the census of 1911, such mechanical devices as existed as an. aid to statistical tabulation were crude and afforded no material assistance to the census authorities. By that time, however, mechanical tabulation had made great progress, and that system of tabulation was accordingly adopted for the preparation of the 1911 Census Reports. An account of the methods employed will be found in Appendix B to the General Report—Census 1911 (pages 259—262). By 1921, these systems had still further developed and were again employed. The information thus mechanically recorded related to sex, age, marital condition, orphanhood, birthplace, nationality, education, occupation, industry, place of work, numbers and ages of dependent children, number of rooms occupied by the family, and—in the case of Wales and Monmouthshire—language spoken. The following is a copy of the card used for tabulation. The relationship between the columns (or "fields") of the, card, and those of the schedule is easily traceable. In order to transfer the information contained in the schedule to the card it was necessary to "code" the verbal statement, i.e., to classify it under its appropriate heading and assign to it the reference number signifying that heading in the pre-arranged code. The chief coding processes were those for birthplace, industry, and occupation, and necessitated the training of staffs specialising in each of these branches. The books of schedules were examined (revised where necessary) and. coded for each item of information in accordance with a pre-arranged topographical order to accord with the scheme of publication (see
page 4). The enumeration books, from which the housing and ecclesiastical statistics were to be derived, were also examined, and the totals checked for the purpose of verifying at a later stage the machine count of punched cards. When a sufficient reserve of coded material had accumulated, the mechanical processes started with the punching of a card in respect of each individual person, the code numbers being thus recorded upon the cards by means of the punching machines. As a rule, each column of the card, when punched, contained only one hole. It was found, however, that, if the restrictions of punching one hole in a column were observed throughout the card, it would be impossible to place the whole of the information contained in the schedule on a single card. Multiple punching was thus resorted to in some columns, especially those used only for the identification of the card in relation to the entry on the schedules, and in columns 37 to 40, headed "Dependency Ages." In this field, column 37 contained a hole for each child under 10 years of age in a family and column 38 a hole for each child of the age 10 to 15 years, while columns 39 and 40 were used to punch information respecting twins and triplets. This device permitted the punching of all the information required on one card and resulted in the saving of many million cards, in addition to the saving of time and labour that would have been involved in punching the second card. For the tabulation two classes of machines were used one, a sorting machine, which operated on a single column of the card only, throwing the cards into the several categories 0, 1, 2, etc., of the columns under operation the other a counting machine, which was capable of counting the number of cards punched, 0, 1, 2, etc., in any number of columns up to three, and was further capable of simultaneously sorting those cards to a new order. Thus, when the preliminary sorting had been performed for the counting of the several items of information under one heading, it was possible in the course of that count to perform the sorting preliminary to the subsequent counting under another heading. The total number of cards punched was 37,886,699 (the total population enumerated) and the total number of passages of the cards through the machines required in order to obtain the detailed information tabulated averaged 24.77, of which 18.77 were sorting and 6.00 counting processes. The combined sorting and counting runs, however, averaged 3.95 per person, and the total number of passages was thus reduced to 20.82 per person—of which 14.82 were sorting and 6.00 counting processes. At the census of 1911 the average number of runs was 13.00 per person for sorting only and 4.44 for counting only. The speed of counting varied considerably on the different processes and fell very low where the analysis of the cards was made in great detail. On such processes the time occupied in copying results and in feeding and discharging the machine greatly exceeded the time it was actually running; and the need of an automatic device for printing the results without loss of time and risk of error in transcription was felt acutely. The results thus ascertained were entered up on working sheets properly planned with a view to the final composition of the published tables. (c) Publication.
Progress was first made with the County Part series, the first volume of which, dealing with the County of London, appeared on 27th October, 1922. During the ensuing nine months ten further volumes were issued, these including the largest counties and together with London representing 53 per cent. of the total population. The remaining county volumes (39 in number) were published by the end of March, 1924. Attention was then concentrated on the subject volumes these were issued at intervals up to July 1925, when the statistical series was finally completed by the issue of the volume of General Tables. This volume presented for the country as a whole, and certain sub-divisions of it, statistics with reference (subject to some minor additions and omissions) to that body of subjects which had already been dealt with in the several county volumes. New matter hitherto unpublished was included with regard to populations of county court circuits and districts seamen and fishermen not enumerated with the general population on census night; and persons enumerated in vessels or establishments under naval, military or air force discipline outside Great Britain. Conversely, statistics of the populations of petty sessional divisions and of registration districts, previously included in the county volumes, were omitted as already adequately dealt with; while the publication of separate subject volumes dealing with occupations and dependency and orphanhood relieved that report of the necessity of including any national statistics under these heads. Mechanical tabulation is, of course, now well established and its advantages are fully recognised. It may be interesting to note, however, that while it has provided an immense increase in facilities for statistical tabulation (without which statistics of the present degree of elaboration would be wholly unattainable) it also imposes conditions not always equally advantageous. Under the old methods, the staff could be concentrated upon a given portion of the work and the completion of that portion expedited without reference to the rest. By machine methods, however, there is a far greater specialisation of processes, giving rise to the necessity for organising an unbroken sequence of operations over the whole field to be covered. This may be illustrated by reference to the 1901 Census—the last census carried through without the advantages of machine methods. The first County Part based upon that census, viz., that for London, was published within nine months of the census day, whereas the first County Part (also for London) issued on the present occasion did not appear till the lapse of sixteen months from the date of enumeration. The latter is of course, fuller than the former, but the essential difference consists in the fact that, roughly speaking, staff had been concentrated upon the first County Part in 1901 to the exclusion of other areas, and upon its completion were similarly concentrated upon the counties subsequently appearing, taken in turn. Under the 1921 machine system, however, a very high proportion of the work in all successive stages for the whole country had necessarily been completed before there was anything to show at the final stage of publication. For example, at the date of publication of the 1921 London County Part, there had also been already completed 90 per cent. of the birthplace coding, 80 per cent. of the industry coding, 76 per cent. of the occupation coding, and 55 per cent. of the card-punching for the whole country. It followed, of course, that subsequent volumes appeared in steady and rapid succession but the fact that so relatively long a period must elapse by machine methods as compared with the old methods before the first results can be shown must be noted as a set-off (even if a trifling one) to the enormous advantages accruing from the former. A review of the progress made in the publication of these Reports naturally invites a reference to the criticism which is very occasionally heard as to the length of time taken in "completing the census." It should be borne in mind that this question depends entirely upon the scale and degree of elaboration of the statistical results which the country is accustomed to expect. It is never possible to "complete" the census in the sense of exhausting the material which it affords indeed, the volumes of utility for one purpose or another which could be produced from that material would go far beyond the limits of any programme usually contemplated. The practical limits to the programme undertaken and completed are, of course, set by the funds allotted for the purpose, on the one hand, and on the other, the scale and standard—high and exacting in this country—of the results demanded for the bare purposes of central and local government to say nothing of the needs of science and research. In any case this country has little to fear from a comparison with the census achievements in other countries and on a proper examination of the facts, such a comparison would prove an exceedingly favourable one. Complete demographic census returns have been received in respect of seven countries and an analysis of these shows that the average length of time which has elapsed between the date of enumeration and the date of publication of the final volume is forty-eight months. For England and Wales the time was forty-nine months. If, however, account be taken of the relative scale of publication which affords the only true basis for comparison, the seven countries had an average monthly output of 41 pages compared with 168 for England and Wales—a rate at which the average programme of the seven could have been completed in a single year. Extending the comparison to ten other countries which have not as yet completed their series, but from which a reasonably large number of census volumes has been received, it is found that their average monthly output to the date of the last volume received was 29 pages only. While the basis of comparison is, of course, a rough and ready one, the broad contrast which it indicates is sufficiently illustrative and it may be added that, so far as can be judged on the information available, the rate of output in this country does not seem to have been equalled in any other; but individual comparisons are not easy to make reliably in view of the wide differences in conditions. A permanent census staff consisting of three officers of the General Register Office is engaged during intercensal periods in recording the numerous changes of boundaries which occur in Administrative, Registration and Ecclesiastical areas; in amending the Ordnance Survey Maps used by the local census officers and in ascertaining the populations affected by such changes. About twelve months before census day 1921, this intercensal staff was augmented by twenty female clerical officers (loaned from the General Post Office) who were engaged in the examination and correction of the Plans of Divisions (description of the boundaries and contents of the local units of enumeration) prepared by the local census officers. At intervals onwards, permanent officers of the General Register Office and Ministry of Health were seconded for census duty in connection with the preparation of the Occupation and Industry Dictionaries, Appointment of Enumerators and other preliminary work. After the date of enumeration, the staff, which then numbered about fifty began to expand rapidly as the various coding processes commenced, and attained Its maximum of 550 in August, 1922. The constitution of this staff at its peak was as follows— The male staff consisted almost entirely of ex-service men recruited either as the result of the Clerical Officers' examination (under the Lytton scheme) or through the Joint Substitution Board. The punchers (girls aged about 15 to 17) were obtained through Employment Exchanges. On the 1st April, 1923, the total staff numbered 413, while twelve months later it was only 24. The total expenditure under the census vote amounted to 351,334. This sum is exclusive of the cost of printing, stationery, maps, cards, and hire of machines. The corresponding expenditure in 1911 was 161,481. This large increase is due entirely to the greatly enhanced cost of living, which at the date of the census was almost at its highest. Of the above-mentioned sum of 351,334, 118,432 represented the expenditure on salaries, etc., at the head office, the balance 232,902, being the fees and expenses of local officers, viz., registrars 75,414 and enumerators 157,488. As was explained on page 3 the services of Superintendent Registrars were dispensed with on this occasion, except in an unpaid advisory capacity. The expenditure on this account in 1911 was 6,832. The fees of registrars and enumerators showed a total increase over 1911 of 119,680. Included in this sum, however, was approximately 9,300 paid to Registrars as a gratuity in respect of the extra work which devolved upon them through the postponement of the census. The following comparative statement shows the cost of successive censuses from 1851 onwards, exclusive of the expenditure on printing, stationery, maps, etc.
CENSUS OF 1921.
PART I.—INTRODUCTION.
(1) Historical Survey.
(2) Preparations for the Census.
(3) Scheme of publication.
(4) Progress after Census Day.
(5) Staff employed.
Grade.
Number.
Work on which employed.
Males.
Females.
Executive and Clerical Officers (loaned from General Register Office, Ministry of Health and General Post Office)
21
19
Supervision
Clerical Officers
117
—
}{
Coding Occupations and Industries, Revising Schedules, Abstracting Buildings, Clerical Work on Counting Machines.
Temporary Clerks
72
—
Temporary Clerks and Typists
—
53
Checking punching, coding birthplaces, examination of enumeration books.
Machine Minders
43
—
Sorting and Counting Machines.
Punchers and Comptometer Operators
—
202
Punching cards and casting enumeration books and result sheets.
Messengers, Porters and Cleaners
12
11
Total
265
285
(6) Cost of the Census.
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
6
1911
36,070,492
161,481
4
9
6
1921
37,886,699
351,334
9
5
6