Next Selection Previous Selection


At the 1921 census, occupied persons in England and Wales were asked for the first time to give particulars of their place of work. Figures, partly provisional, relating to London and the Home Counties, were published as a Supplement to the London County Volume * That volume and the Workplace Volume for England and Wales, presenting in statistical form the results of this enquiry, thus signalise the first occasion on which any attempt has been made to classify the population of this country by reference to any basis of local distribution other than the place of residence or the place of birth. Indeed, so far as the Department is at present aware, this is almost the first occasion upon which statistics of workplaces have been obtained and presented by any country in the world.

The importance of this new factor in demography will be readily appreciated. In a less highly organised and industrialised community localities may tend to be more or less self-sufficient, each local resident population being served and supplied with the bulk of its needs by the same population in its working capacity, and thus experiencing little necessity for interaction with other localities apart from the occasions of periodical fairs or markets. Where such conditions prevail, workplace and residence will ordinarily be near enough to be both represented by a single classification on the basis of the latter, or on the basis of the normal census distribution which accepts place of enumeration as a criterion of residence.

But this stage has long been passed in the history of this country for the great majority of its workers. As those conditions were due, it may be said, to the dispersion of necessary services and production which, in the absence of transport facilities, had to be located in proximity to the population served, so the great development of transport and communications fostered a concentration which has changed the whole face of industry. The growth of large manufacturing, distributing and commercial centres has not only given rise to concentrations of workers beyond the residential capacity of their immediate neighbourhoods, but has exaggerated that deficiency by substituting factories, warehouses and offices for dwellings in the centres themselves; while the very development of transport which has made it necessary for the worker to live at a distance from his work has also made it possible for him to do so.

Thus it happens that at the present time in many parts of the country masses of population move in tides of daily ebb and flow. These movements obviously have a direct bearing upon many difficult problems of traffic, transport and housing, and it has seemed necessary that there should be some means of measuring their direction and volume. Statically considered, moreover, the new position is important. The resident population of any locality is no longer the sole matter of concern to that locality. During the day it may be peopled by a body of workers numerically far exceeding and even very differently composed from its so-called permanent population. Local public services must be provided for these invading armies, and for many practical and administrative purposes we have now to reckon with the fact that, for localities situated within a region of highly organised industry, separate account must be taken of both a night and a day population, the two often differing widely from each other in number and constitution.

The increasing divorce between residence and workplace has called for reconsideration of the areal basis of some census statistics. While it is relevant to present the working population in its occupational capacity as part of the resident population of which it constitutes the bread-winning element, an industrial tabulation by area of residence will clearly give a distorted picture of the industrial map. Hence in presenting the industrial classification1 of the people, its local distribution was based not upon the area of enumeration but upon that of workplace, the man-power of the several industries being thus shown in relation to the local seats of the industries themselves. Occupation (as distinct from industry) has a close connection with the health, mortality and other personal aspects of the individual, and its statistical presentation cannot lightly depart from the residence distribution upon which the great body of comparable statistics on those subjects is based. But in retaining the residence basis as an index to some important aspects of environmental influence, it must not be forgotten that a full half of the waking day—even, it may be said, a full half of the active life—of the worker is often spent in an entirely different environment, which cannot fail to leave its mark upon him.

Thus, even apart from those problems upon which the subject of workplace seemed more directly to bear, it gave promise of breaking new ground in so many fields of administration or study that a serious attempt to exploit it statistically appeared to be called for notwithstanding the difficulties inherent in the enquiry. An account of these difficulties, and the manner in which they were met is contained on pages 3 and 4 of this Report. The question on the census schedule asked for the address of each person's place of work, and intimated that persons with no regular place of work or whose work was carried on mainly at home should state the fact by means of the statement "no fixed place" or "at home" respectively The brief question on the schedule itself was supplemented by instructions to the enumerators and other local officers whose duty it was to provide any necessary assistance in the filling up of the schedules, indicating more precisely those cases to which the "no fixed place" classification was intended to apply. Thus a worker though usually working in different places, who operates from a regular centre such as his employer's premises and who visits these premises in the course of his daily work, was asked to give his employer's address. Similarly, a worker regularly employed in the vicinity of his employer's address, as in the case of a farm labourer, was required to give that address. On the other hand, in the case of commercial travellers, costermongers, a building contractor's employees, etc., the more appropriate reply would generally be that of "no fixed place" In regard to persons out of work, the census schedule directed that the place of work enquiry be left unanswered. This instruction, however, was based upon normal expectations of unemployment, and, if adhered to in the wholly exceptional industrial situation which prevailed at the time of the census, would have deprived the enquiry of a large proportion of its value. In order to avoid this, enumerators were instructed to obtain a statement in respect of all such persons of the address of their last place of work.

The enquiry being a new one, the treatment and arrangement of the data must be regarded in a measure as experimental. The form of the tables has largely been governed by the multifarious character of the uses to which the figures may be applicable. It would have been an easy matter to have arranged the statistics so as to have rendered them more particularly significant for one or two specific purposes. But at the present stage it appeared preferable to give the full figures in themselves without any special adaptation, in order that they might be more fully available for varied—perhaps at present unforeseen—uses. Attention must, however, be drawn to certain limitations affecting the material which should be borne in mind in any particular application of the statistics derived therefrom.

A feature of which account must always be taken, and which is common to many census or other statistics of a national character, is represented in the tables by columns j and k of Table I of the Workplace Volume, recording the numbers enumerated in each area from whom a statement of workplace was not forthcoming. Those numbers comprise persons out of work at the date of the census from whom the enumerators failed to obtain particulars of the last place of work in accordance with the special instructions referred to above. But they also include persons in respect of whom the addresses given were insufficient for purposes of classification, with some consequent understatement in the other columns. For the whole country the "not stated" cases amount to 2.8 per cent. of the total occupied population.

Another defect, which is partial in its incidence, but which, unfortunately, will have the effect of considerably modifying the value of the figures in a limited number of localities, arose through the postponement of the census until the 19th June While this date succeeded in avoiding the recognised programme of industrial holiday, some holiday movement was, largely owing to the abnormally fine weather, then in progress. This has been duly reflected in the enumerated population figures in the inclusion in the case of holiday and other resorts of varying and sometimes large proportions of visitors. In such areas the census or de facto population will be in excess of the resident or de jure population. Moreover such of the visitors as are included in the occupied population usually be recorded amongst those working outside the holiday area, and the normal tidal movement between workplace and residence may not be truly represented by the relation between recorded workplaces and place of enumeration, the divergence varying with the amount of the local inflation. For a fuller statement regarding census inflation reference may be made to Appendix A in connection with which attention should be directed particularly to the table on pages 199 et seq., showing the Boroughs, Urban Districts, and Rural Districts in which the inflation or proportion of the enumerated population, in excess of the estimated resident population, was more than 3 per cent. In a majority of the 1,817 areas dealt with in the tables the difference between census and resident populations is unimportant, there being only 209 cases in excess of 3 per cent., in 92 of which the inflation exceeds 10 per cent. of the census population.

The statistics are presented in detail in three tables in the Workplace Volume. Table I classifies the occupied population of each sex enumerated in every area according to the four descriptions of workplace, viz:—

  1. workplace in the area,
  2. no fixed workplace,
  3. workplace not stated,
  4. workplace outside the area,

and in the adjoining columns the complementary movement to (4) is shown, viz.:—

  1. the numbers working within the area but enumerated elsewhere.

From (4) and (5) (or, in the table, columns 1 and m, n and o), an indication of the daily displacement of the working population may be obtained from which, by making arbitrary assumptions in regard to those with "no fixed workplace" or with "workplace not stated," a relative measure of the respective "day" populations may be arrived at. For convenience in the use of the tables, two final columns, p and q, have been inserted, showing the excess of the outward over the inward movement, both as an actual number and as a proportion of the total enumerated population.

In Table II, the movements shown in aggregate form in columns 1, m, n, and o of Table I are analysed in detail but without distinction of sexes. The general direction of the movement from and to each area is indicated by the county totals appearing within each section, and these are supplemented by the actual places of origin or destination together with the numbers involved in all movements consisting of 25 or more persons. The volume is completed by the addition of a third table, summarising for each county and every urban area with an enumerated population in excess of 20,000 persons, the enumerated population, the day population (persons with "no fixed workplace" or "workplace not stated" being treated as working in the area of enumeration), the gross movement in and out of the area, and the net movement (an inward balance Being indicated by a negative sign, as in Table I).

In this first and, therefore, experimental treatment of the relation of workplace and place of enumeration, it is necessary to observe that the movements identified in the tabulated statistics can only be a fraction of the total number which take place daily throughout the country. The areal unit from and to which movements are recorded is the complete sanitary district—borough, urban district or rural district, as the case may be—and only when the place of work is in a different district from that in which the person was enumerated, that is, when the movement from and to place of work involves the crossing of a boundary, is the fact recorded in the tables. Movements, however numerous and whatever their importance from the point of view of the distance travelled, which begin and end in the same area could only be identified by the adoption of a unit smaller than that of the complete sanitary district, and though it was originally intended that, for this purpose, urban areas should be sub-divided by their constituent wards and rural districts by civil parishes, the proposal had to be abandoned on account of the very large addition to the expense and labour that would have been involved by so great an extension of the tabulation scheme. The result is that there is a wide diversity in the sizes of the basic areal units. These range from the large county boroughs numbering their populations in hundreds of thousands—Birmingham is the largest single unit with a population exceeding 900,000—down to comparatively unimportant urban and rural districts, of which the population may number little more than a thousand individuals and in the extreme cases is even less than a thousand.

The magnitude of the movement shown in respect of an area and the proportion this bears to the total movement, identified and unidentified, which may be said to take place daily from, to, or within the area will, to a not unimportant extent, be dependent on the size of the area in question. In a large town much of the total movement will be within the existing boundaries and will remain unrecorded in the tables, while in the small area a larger proportion of the whole will involve the crossing of the boundary and will be recognised accordingly if within the scope of the census enquiry.

From a general inspection of the figures it will be seen that in relation to the place of enumeration the areas for which a net outward migration during he day is registered are more numerous than those for which the balance is inward The totals of the "ins" and "outs" for all areas taken together are necessarily the same because every "exit" from an area has its counterpart as an "entry" to another area; but since the needs of industry and commerce have so far been best met by a concentration of the day population while for residence and domestic purposes converse conditions are preferable, industrial areas are generally associated with a larger number of dormitory areas and such interaction as takes place is characterised by a daily pulsation between the common industrial centre and a larger number of surrounding residential areas.

The outstanding example of this daily pulsation is to be found in respect of the region of which London is the centre. Owing to its more or less detached position it is comparatively undisturbed by the influence of cross currents, caused by the presence of other industrial centres in the vicinity, such as exists in respect oi many of the areas in Lancashire, Yorkshire and the Midlands, and the area is in itself so extensive that the analysis by boroughs, urban districts and rural districts, many of them large in themselves, is fine enough to permit of the classification of the areas and of the movements affecting them into well defined types.

This may be seen from the following tables:—



The most divergent types occur at the centre and just within the boundary of the region. The extremest type of all is that of the City itself, which is almost completely non-residential and which expands and contracts daily between the insignificant night population of 13,709 and the more than thirty times as large day population of 436,721, a number which takes no account of the large miscellaneous movements represented by visitors for shopping and other purposes and by the traffic of all kinds passing continuously through the City in transit between areas on either side.

At the other extreme are the residential areas of the "Outer Ring" In respect of these, the outward movement during the day is, with few exceptions, substantial in itself and considerably in excess of such inward movement as is recorded so that the resultant day population is as a rule much lower than the corresponding night or resident population. The rule is not invariable, however, for in most districts there is some inward flow to be set against the daily exodus, and in those in which there is local industrial development the inflow may be large and even greater— Erith U.D., Acton U.D., and Richmond M.B., are examples—than the corresponding outflow.

The intermediate areas are more definitely both residential and industrial in character, the significance of the latter aspect increasing as the centre is approached. In nearly all these cases both the outward and inward movements are of importance, but whereas in six of the eight inner boroughs the immigrants during the day exceed the emigrants resulting in an increased day population—in the outer boroughs the position is reversed, only six increasing their population during the day as compared with fourteen boroughs in which the population is reduced.

The daily change in the zonal distribution of the population in the whole of Greater London is summarised in Table XC, from which it appears that the central areas receive a net addition of rather more than 800,000 persons during the day, approximately one-half going to the City and the remainder to the inner boroughs at the expense of the outer areas, the loss in this case being divided between the outer boroughs and the "Outer Ring" the larger number coming from the "Outer Ring."

From the resultant day population shown in the final column of Table XCI— and for this purpose it may be noted that persons with "no fixed workplace" or persons who failed to return particulars of their workplace have been treated throughout as working in the area in which they were enumerated—it will be observed that the most populous area during the day is the City of London with a total of 436,721. Of the 28 Metropolitan Boroughs, Westminster ranks first as regards its day population (385,984) in contrast with a position of 15th in respect of the enumerated figures, and is followed by Islington, Wandsworth, Stepney and Lambeth with day populations ranging between 273,000 and 260,000. Holborn, with the smallest night population, and Finsbury ranking 25th in this respect, advance during the day to positions of 22nd and 14th respectively in order of magnitude.

Elsewhere than in the Metropolitan region the unit of area adopted as the basis of tabulation is so large in relation to the population movement that much of the latter must remain unidentified. The more important towns are in themselves large enough to embrace not only a more definitely industrial or commercial central nucleus, but as a rule a large proportion of the related residential zone as well and though they may be subject to the same kind of daily pulsation of population as is manifested in London, on an appropriately smaller scale it is only when and so far as the residential area extends beyond the town boundary that: the movement can be recognised in the form of tables as published. Notwithstanding the limitation, the recognised movements are so numerous throughout the country at large as to occupy 144 closely printed pages for the identification of the more important of them (Table 2 of the "Workplaces" Volume), and since they are on the whole representative of the longer-distance journeys they should be of service in the solution of the more important problems relating to traffic, transport and housing.

Examples of towns subjest to various types of population movement are given In the following Table:—


The contrasts shown in these statements have been obtained by the use of percentages based, in all cases, upon the respective total census populations. It should be noted, however, that the movements themselves relate only to the "occupied" populations, and that their portrayal in the form adopted will accordingly be subject to an element of variation due to differences in the proportion of "occupied" to total. For the country as a whole the total "occupied" population is less than half the occupied and unoccupied combined, so that, apart from the effect of local variations, the substitution of the former as the basis of comparison would have the effect of approximately doubling the relative magnitude of the daily movements.

1 Census of England and Wales, 1921, Industry Tables .

Next Selection Previous Selection