Next Selection Previous Selection


1. General

Definition of Population.— It will be well, before referring to the statistical record of the census either of the country as a whole or of any of the many areas into which it has been divided, to define precisely what is meant by population. For the country as a whole the criterion is definite and of general application. The population consists of all persons who were enumerated as being alive and present within England and Wales at midnight, 19th-20th June, 1921, persons on board vessels which were in port on census night or which reached port on the following day being regarded as within the scope of the definition. It comprises visitors as well as residents, persons of British and alien nationalities, civilians and non-civilians, and in all tables the totals for England and Wales refer, unless specifically defined, to the aggregate population so described. Similarly, for individual portions of the whole country, the 1921 census population is once again, following the precedent of all earlier censuses of England and Wales, the de facto population, that is the sum of the individuals enumerated in an area, and with an important exception referred to presently the tables conform to this basis of allocation throughout. Adaptation to the basis had to be made in respect of persons who were travelling during the night, were engaged on night work or were otherwise inaccessible, and these, if they had not already been enumerated, were counted as belonging to the population of the place at which they arrived on the following day. In this class were also included persons on board vessels which reached port on the day following the census.

The adoption once again of the de facto basis of classification was due primarily to its simplicity. The criterion of location at a given point of time involves a minimum of difficulty in determination individuals are included in their appropriate schedules by reference only to their actual presence, and the collection of the schedules immediately after the appointed day automatically ensures a rapid and accurate computation of the populations of all areas from the smallest to the largest units. The distribution is theoretically an accidental one, and the justification for its use—apart from administrative convenience—has lain in the fact that by a judicious selection of the census day it has usually been possible to secure a distribution substantially equivalent in the majority of areas to a classification by place of normal residence or to the de jure basis of distribution.

With the de facto basis, it must be the aim of the census authorities to choose a date at which the movement of the population, not only to and from the country as a whole, but between the several areas within the country, may be expected to be at a minimum. In the selection of a suitable date it is possible as a rule to avoid large scale movements affecting wide areas but no one date applying to the country as a whole can be expected to meet with an absence of movement in every locality. Small tidal migrations are almost continuously occurring, of which examples may be seen in the seasonal movements in some industries, particularly in connection with agricultural operations such as harvesting, hop picking, pea and fruit picking, etc. Again, the universities and large public schools, which often bulk largely in the life of the districts in which they are situated, regularly expand and contract their populations with their alternating periods of term and vacation. Such movements are usually limited in range, but their number, variety and frequency render it improbable that the result of a general enumeration at any given point of time could ever be exactly representative, in the case of every area, of the normal resident population; the approximation to the intended representation has however been regarded hitherto as sufficiently close to justify foregoing the costly and laborious processes involved in the identification of each person's residence and the consequent re-arrangement which adherence to a strict residence basis would involve.

The first census was taken in March, 1801, and though for the next four, viz. 1811-41, dates near the end of May were selected, the subsequent censuses, 1851-1911, have invariably been taken at the end of March or the beginning of April.

Postponement of the census from 24th April to 19th June.— It has already been stated that arrangements had been brought to a state of completion with a view to the 1921 census being taken on the 24th April when circumstances intervened which necessitated its postponement to the 19th June. While the latter date succeeded in avoiding the recognised industrial holiday season, there is no doubt that the periodical summer movement had by that time begun and that the divergences between the "enumerated" and the "resident" populations are in many areas greater than would have been the case if the census had been taken in April. An estimate of the approximate amount of the divergence in individual boroughs, urban districts, and rural districts has been made and is described in Appendix A. It is sufficient to state here that the disturbance chiefly affected holiday and health resorts, the enumerated populations being generally in excess, sometimes considerably so, of the resident populations.

It has been urged that for some of the statutory and other purposes upon which the censal population has a bearing, a census taken in March or April does not adequately reflect the mean population of a town which is inhabited for several months in each year by an influx of visitors out of all proportion to its winter strength. Thus, an inflation which may render the returns more representative of the average than of the more permanently resident population, while undesirable for some reasons (e.g. mortality comparisons, since deaths are allocated to area of residence), is not necessarily to be regretted in connection with other purposes. Conversely, if it is customary for the residents of an inland or industrial district to spend a portion of the year, however small, away from home, it may be equally reasonable that the fact should find statistical expression in the census figure. The matter depends, of course, upon the nature of the purpose to which the census figure is applied.

Exceptions to place of enumeration as basis of areal classification.— With certain of the enquiries falling within the scope of a census, the subject itself has an areal significance, and in such cases the statistical classification must of necessity conform to the subject areas as well as to the area of enumeration. Examples are the birthplace enquiry, which has been a feature of past censuses as well as of the present, and also the enquiry regarding place of work introduced for the first time on this occasion. In each case, in so far as one, if not the main, object has been to establish a relation between the place of enumeration (i.e. residence) and the place of birth or place of work, as the case may be, the use of the subject area as a function of the classification does not necessarily constitute a departure from the normal basis of tabulation by place of enumeration. The availability of an alternative basis, however, renders possible its adoption in connection with any other portion of the census material, and advantage has been taken of the opportunity on this occasion to present the industrial classification of the people (see page 142) by local divisions representing, not the areas of enumeration, but those of workplace, the man-power of the several industries being thus shown in relation to the local seats of the industries themselves. As the scope of the census widens, further occasions may arise necessitating the adaptation of the areal criterion to meet the needs of a particular subject of enquiry the present may properly be regarded as the first in the history of the censuses of England and Wales on which it has been possible.

Population of England and Wales.— The total population enumerated in England and Wales on the night of the 19th June, 1921, numbered 37,886,699 persons, of which 18,075,239 were males and 19,811,460 females.

In the following table these figures are shown in continuation of the series of corresponding figures of past censuses and they are shown pictorially in the diagrams on page 13.

The present total is the largest ever recorded in this country and exceeds the number returned at the enumeration of the 2nd April, 1911, by 1,816,207, a growth in 10 years and 2 months of 5.04 per cent., corresponding to a decennial increase of 4.93 per cent. The increase is numerically only about one-half of the increase in the preceding intercensal period and is less than any similar figure since 1811, while proportionately it is far lower than any hitherto recorded.


Intercensal Movement.— The several components of the net intercensal increase of 1,816,207, expressed in thousands of population, are approximately as follows:—

  Births registered in England and Wales + 8,281
Decrease —    
  Deaths registered in England and Wales—    
  Civilian - 5,200
  Non-civilian - 68
  Deaths of non-civilians belonging to England and Wales which occurred abroad at the various theatres of war (estimated) - 577
  Excess of outward over inward migration { civilian - 590
non-civilian1 - 30
Net intercensal increase + 1,816

Of these, only the figures relating to births and deaths registered within the country, of which detailed records are obtained under a careful system of registration, can be accepted without reserve. The deaths of non-civilians abroad, including many cases of "presumed" death, have been stated for the United Kingdom as a whole by the War Department responsible for the records, no doubt with a certain degree of approximation, but we have no direct evidence showing how many of the total could be regarded as occurring among the English and Welsh populations separately, and it has been assumed, in the absence of better evidence, that the United Kingdom deaths can be apportioned according to the numbers of recruits to the forces contributed by the several countries of the United Kingdom. The net migration shown is a balancing item necessary to complete the table and will accordingly be subject to the same error, with an opposite sign, as is contained in the estimate of war deaths abroad.


Population of England and Wales 1801 to 1921


Intercensal increases in population in England and Wales 1801 to 1921


Population growth rates in England and Wales 1801 to 1921

In Table II the several movements are shown in comparison with those of earlier intercensal periods, and in Table III they are further analysed by individual years of occurrence locating more precisely the important changes both in amount and direction, which have taken place since 1911.



So large a portion of the decennium has been subject to the influences of the war that it will be impossible to observe from the successive annual records any of the more permanent tendencies which might be expected from such a series of figures. Only the first three years, 1911-13, can be regarded as completely normal in character. With the outbreak of hostilities in 1914, the usual migration currents were at once curtailed, and a little later the hitherto regular series of births and deaths became subject to greater and greater disturbances, which had not come to an end at the determination of the decennium.

Births which in the period 1901-10 had averaged about 930,000 per annum, with a slightly decreasing tendency, fell below the 900,000 point for each of the years 1911-14, and thereafter diminished at a much greater pace to 1918, when a minimum of between 662,000 and 663,000 was registered Though the direction of the movement changed upon the cessation of the principal hostilities in 1918 and large increases were recorded after the middle of 1919, the actual amount of the increase within the intercensal period has not been nearly sufficient to compensate for the deficiencies of the war years.

Deaths, on the other hand, if account be taken of those of non-civilians belonging to England and Wales which occurred overseas, showed no such decline between 1914 and 1918; the peace-time average of 507,000 registered in 1911-13 rose to between 600,000 and 700,000 in each of the years 1915-17, and probably reached 750,000 in 1918 owing to the further imposition of the severe influenza mortality experienced towards the end of that year. After 1918 the death-rate, like the birth-rate, at once took a favourable turn, the numbers ultimately showing a notable reduction even in comparison with those of the years preceding the war.

To the combined effect of diminished births and increased deaths resulting over the ten years in a natural increase of 2,436,000 (bringing all war deaths into account), as compared with 4,044,000 in the previous intercensal period, must be ascribed principally the exceptionally low increase in population now recorded.

Migration, of course, contributes to the total movement, but, notwithstanding the restrictions placed on normal passenger traffic, the net population loss from this cause over the whole decennium did not exceed about 620,000, a figure which is only 119,000 in excess of the corresponding figure for the preceding intercensal period. It must be borne in mind that the migration figures of Tables II and III represent the balance between two much larger movements in opposite directions, and that a small change in either of the direct currents will involve a relatively much greater change in the net difference between the two for this reason, and also from the fact that the movements are likely to be more readily responsive to changing social and economic conditions, the records of successive periods will tend to vary to a much greater degree than those of births and deaths. This is illustrated in the last columns of Table II; the 1911-21 figure for females is rather higher than any corresponding figure shown in the table, but for the two sexes taken together they are not out of character with their antecedents and are in no way indicative of. the abnormal movements of which they form the net result.

For a knowledge of the migration of a period other than that beginning and ending with a census, recourse must be had to the returns collected by the Board of Trade. These have been extended and improved since 1911, but they do not afford the means of accurately determining the number of migrants from and to England and Wales as a separate unit. The migration figures shown in the final columns of Table III are therefore approximate they are based upon the Board of Trade Returns for the individual years, which have been modified to secure a general correspondence in the aggregate with the net intercensal figure deduced above. There appears to have been a considerable outward flow in the years 1911-13 which was reversed in 1914 and after a period of quietude again restored in 1919. The suddenness both of the reversal in 1914 and its later restoration is no doubt largely accounted for by the influx and repatriation of war refugees.

Revision of Intercensal Estimates of Population.— A brief reference may be made here to the estimates of population in respect of years of the past intercensal period given by the Registrar General in his successive Annual Reports for those years. The estimates were prepared on such data, much of it incomplete, or imperfect, as were at hand at the time the estimates were made, and the whole intercensal series falls naturally to be revised in the light of the results of the recent census figures. A description of the methods used in the estimation of each year's population is given in the report for the particular year, and no more need be said here than to point out that after 1913 any assumptions based upon peace-time continuity in the several factors contributing to population changes became untenable, and that with the inevitable reticence which had to be observed in respect of the large and unusual movements taking place continually between this country and other parts of the world, it became extraordinarily difficult to identify the numbers within the country at any given point of time. The close correspondence of the final estimate with the 1921 enumeration—the Preliminary Report of the Census stated that the estimate was in excess by about 1 per 1,000, but later examination suggests that the difference was rather greater—might have been regarded as generally confirming the whole series of intercensal estimates further evidence which has now become available indicates however that such inference, acceptable over a series of peace years, could not be applied to the past decennium, and the following statement, compiled with the aid of more complete data regarding intercensal movements as well as of the 1921 census itself, may probably be accepted as giving a truer picture of the yearly changes in population.


International Changes in Population.— It will be of interest to compare the rate of growth in this and in other countries, and this is done in the following table for the most important countries for which statistics are available. Where, in the case of countries which participated in the late war, boundaries have been changed during the decennium, the populations, both of the pre-war and post-war territories, have been given as far as possible in order to preserve comparability in the percentage increases or decreases. Most of the changes shown for the past intercensal period must be regarded as abnormal in character according to the degree to which the populations were affected by the war. All the belligerents will have suffered direct loss by actual casualties, and where, at the termination of the war, boundaries were altered there will often have been considerable adjustments in the resettlement of the populations of the transferred areas the more indirect effects of the war, such as the reduction in the birth-rate, have been more generally distributed and have extended in varying degree to neutral countries as well as to those directly involved.

One of the prominent features of the table is the general decline in the rate of growth only five countries, viz. Australia, Denmark, Norway, the Netherlands and Italy, show a higher percentage increase in population between 1911 and 1921 than that of the preceding decennium. In Denmark and Norway with comparatively small populations the acceleration in the rate of increase has been considerable, whereas in Sweden the 1901-11 increase has not been quite maintained. Two of the accelerated rates of growth are noteworthy as being in respect of countries which participated directly in the war, viz. Australia and Italy, and it will also be observed that the increases in other British Dominions and the United States of America, though rather lower than they were ten years ago, are much above the European average of this or past decades. Of the European belligerents, the German increase has diminished to a fraction of the high rates of previous decades, while in France the population which had been almost stationary for many years showed the only important decrease in the table. Belgium, the only other country showing an actual decrease in population, is noteworthy in that the decline, negligible in itself, follows a period of steady and not inconsiderable growth.


Density of Population.— The area of England and Wales, including land and inland water, but excluding tidal water and foreshore, according to the latest survey returns received amounted in June, 1921, to 37,340,338 statute acres or 58,344 square miles. The density, assuming an even distribution, of the 1921 population therefore represents .99 acres per person or conversely 1.01 and 649 persons per acre and per square mile respectively.

The area of the country on which the density depends varies so little from time to time, that rate of change in the density corresponds almost exactly with the movement of the population, and its graphic representation will be a facsimile of the population chart on page 13. The actual densities at each of the past censuses of England and Wales are given in the following table.


The principal interest of the average density lies in its use as an index of the national population pressure in relation to the similar figures both of colonial and of foreign countries, and for this purpose the corresponding densities of the current populations of other countries have been shown in the final column of Table V. In this comparison the position of England and Wales is noteworthy as being the highest in the list; the inclusion of the more sparsely populated areas of Scotland and Ireland results in a United Kingdom density which is about 60 per cent. of the England and Wales figure and very closely corresponds with that of Japan. It is only exceeded in the table by the densities of Belgium and the Netherlands. Other countries showing a density exceeding 100 persons per square kilometre are Germany and Italy. France, with its stationary population, has only 71 persons per square kilometre, while in the newer countries of the Dominions and the United States the densities are almost negligible in comparison.

Divisions of the Country.— The bulk of the population being working adults or persons economically dependent upon them, the distribution of individuals throughout the several areas of the country will be governed largely by the relative wage-producing capacities of the several areas, and the variations in distribution from time to time may be expected to reflect the change in industrial activities, using the term industry in its widest sense, more than any other factor. For this reason the full significance of the changes between one census and another will only be seen by an analysis which subdivides the country into relatively small units. Broad geographical divisions or the conventional county areas, separate population statistics for which have to be prepared for local government and other purposes, will rarely in themselves provide a guide to the characteristic variations owing to their usually heterogenous composition from an industrial point of view; the best illustration of this is found in the wide differences of their respective urban and rural components. On the other hand the simple economic basis of distribution may be modified by other factors, such as the uneven incidence of housing accommodation for the workers concerned, by the increasing development of cheap and rapid transport which enables a worker to live at a considerable distance from his work, and in individual areas by special local features as, for example, the natural attractiveness of seaside and health resorts for the infirm and retired sections of the community, or the existence of schools or institutions whose inmates may form a significant proportion of the local population. And when, as was the case in the latter portion of the past decennium, the whole industrial organisation of the country was diverted from its normal peace-time channels and temporarily recast for the service of a national war, the consequential population adjustment will be more than ordinarily complex and may be expected to remain abnormal until the effect of the dislocation has wholly disappeared.

Geographical Divisions.— The following table shows approximately how the variations in the population, which, for the whole of England and Wales amounted to an increase of 5.0 per cent. in the period 1911-21, as compared with one of 10.9 per cent. in the previous decennium, were distributed in six geographical divisions of the country.


The divisions are placed in their order of increase in 1911-21 and it will be observed not only that the order of the 1901-11 increases has been roughly maintained, but that there is a similarity in the changes in the individual rates of increase in that each of the current rates, like that of the whole country, is between 40 and 55 per cent. of the corresponding rate for the earlier period. Between themselves the smallest groups show the greatest divergence, Wales standing highest with an average increase of 9.7 per cent., while the Eastern and Southern areas only advanced by 3.0 and 3.9 per cent. respectively.

2. Administrative Areas.

County Areas.— Various forms of comparative county statistics are given throughout the General Tables Volume. Tables 3-5 show the growth of the population in each county since 1801, both in absolute numbers and also in relation to the population of the whole of England and Wales, the earlier records referring to the ancient county and those since 1891 corresponding to the areas within the more recently constituted administrative boundaries. In Table 6 the administrative county (with its associated county boroughs) is treated in greater detail but over a shorter range of years.

In the subjoined tables geographical situation has been ignored and the 62 administrative counties (including associated county boroughs) are classified (a ) according to the sizes of their population in 1921, and (b) according to their rate of increase in 1911-21.



A general feature brought out by the first arrangement is the unsatisfactory nature of the county unit as a basis of comparison owing to the extreme differences in the sizes of their populations. Lancashire heads the list with a greater lead over London than it had ten years ago, and these two counties together account for approximately one-quarter of the whole population. The West Riding of Yorkshire forms an outstanding third in the series, after which the gradation is more regular down to Rutland, the smallest, with a population less than one two-hundred and fiftieth part of that of Lancashire 31 of the least populous of the 63 county divisions shown in the table contain in the aggregate but 10 per cent. of the total population, and in 45 the total population is but slightly in excess of that of Lancashire and London combined.

The principal feature to be observed in connection with the changes of the past decennium is that, notwithstanding the variety and magnitude of the movements which must have occurred during the war, the resultant changes throughout the country, as indicated by a comparison of the county distribution of 1921 with that of 1911, are far more restricted than those of the preceding decennium.

1901-11. 1911-21.
Mean Increase
England and Wales.
Mean Increase
England and Wales.
Per cent.
Per cent.
Increase or
Decrease of
Increase or
Decrease of
Increases exceeding 30 per cent. 3 692,611
" 25-30 per cent. 1 191,917
" 20-25     " 3 464,973
" 15-20     " 7 632,978 1 13,912
" 10-15     " 10 669,477 6 559,752
" 5-10     " 20 839,050 22 848,562
" 0-5     " 10 76,432 24 457,341
Decreases 0-5     " 8 21,502 8 53,470
" 5-10     " 1 3,287 2 9,890
" exceeding 10 per cent.

Although the number of counties showing an increase of population on this occasion is not very different from the number of 1911, the number of decreases in each period being only a fraction of the total, the range of the increases of 1911 was very much wider. Thus in 1911 24 counties showed increases exceeding 10 per cent., of which the highest, viz., those of Middlesex and Monmouthshire, were as much as 42 per cent. and 33 per cent. respectively. On the present occasion only seven of the increases exceed 10 per cent. of these Flintshire is the highest (15.0 per cent.), followed successively by Monmouthshire (13.9 per cent.), Glamorganshire (11.7 per cent.), Warwickshire (11.4 per cent.), Middlesex (11.2 per cent.), Sussex West (11.1 per cent.) and Surrey (10.0 per cent.).

In five English counties, viz., Cumberland, Isle of Wight, Kent, Sussex East and Westmorland, and in six Welsh counties, viz., Anglesey, Cardigan, Carnarvon, Flint, Merioneth and Radnor, the present increase is greater (or the decrease less) than it was ten years earlier. Many of these areas however contain a number of holiday resorts which were subject to temporary inflation by visitors in 1921 (see Appendix A) and to that extent their apparent acceleration in growth may obscure the real position.

Increases above the average are recorded for all the Home Counties, but, with an exception in the case of Kent, where the 1901-11 development was somewhat low and where the 1911-21 movement is overstated by the 1921 inflation, the present increases are much below what they were in either of the two preceding intercensal periods.

Metropolitan Counties. Increase per cent. in Population.
1891-1901. 1901-11. 1911-21.
Middlesex 45.9 42.1 11.2
Surrey 25.3 29.4 10.0
Kent 15.7 8.8 9.2
Hertfordshire 14.1 20.5 7.0
Essex 38.4 24.6 8.8

The County of London itself, dealt with subsequently in greater detail, again shows a small decrease. In nine county areas (other than London) a decline is registered amounting in no case to more than 6.8 per cent., with the exception of Rutlandshire, where the loss of 1,970 persons represents a decrease of 9.7 per cent.

Most of the counties in which mining is an important industry—Monmouth, Glamorgan, Carmarthen, Durham, Northumberland, Denbigh, Nottingham—show increases well above the average for the whole country, but in these, as in most others, the current increases are substantially lower than those of 1901-11. Of other important industrial counties, the increases in Warwick, Cheshire, and Stafford are above the general average in the textile areas of Lancashire and the West Riding of Yorkshire they are somewhat below, as they are also in Worcester and Leicester, while in Northamptonshire a decrease is recorded. The predominantly rural counties occupy, with few exceptions, the lowest position in Table IX; they account for most of the recorded decreases, and the increases, where they occur, are usually of a low order.

In Table 6 of the General Tables Volume the net increase or decrease of population in each administrative county and in each county borough is divided into two portions: (1) showing the natural increase (civilian), or excess of births over civilian deaths registered in the country, and (2) the balance, or what in ordinary circumstances would represent the gain or loss by migration. A note at the heading of that table states that war deaths which occurred outside the country are included as losses by migration since, theoretically speaking, the individuals ceased to be part of the de facto population on their leaving the country. Had it been possible to allocate the war deaths to their appropriate districts it would have been better to have separated them from the true migration component. But as already stated the number of war deaths, among the English and Welsh populations as a whole, representing about 3.6 per cent. of the male population, or 1.7 per cent. of the total population, is only an approximation, and in analysing the county increases or decreases we cannot do more, without an amount of research that the subject would not warrant, and which might very well be negative in its result, than assume that such deaths were distributed proportionately over the country.

Densities of Population in Counties.— The density of the population in each of the county areas is shown in the last column of Table IX. The difference suggested by these figures is very wide, ranging from 59.9 persons per acre in London to 0.1 in five of the smaller counties, but it must be remembered that densities in this composite form will be governed largely by the relative proportions of dense urban units, sparse rural areas, practically uninhabited moorland, and the inland water included within the several areas, and are of little interest as guides to population pressure. All important populations are massed together in well defined towns, and even here, though the populations are more evenly distributed, the presence of parks and open spaces and the frequent inclusion of a fringe of undeveloped land within the boundary will tend to under-rate the effective density and reduce its value as a relative measure of living conditions. Excepting in London, which is many times as dense as any other county, there is so much room for population expansion, or in other words the densities are so far from any maximum which could be regarded as approaching saturation point, that the county increase between 1911 and 1921 bears little or no relation to the density. If anything, the table suggests that generally speaking it is the denser counties which record the greater growth and vice versa, but the degree of correlation does not appear to be a high one.

Urban and Rural Divisions.— Of the 1,789 urban and rural districts into which the country is divided, the population of the 1,126 urban districts (including county and municipal boroughs and counting the Administrative County of London as one district) amounted at the date of the census to 30,035,417 persons, while the population of the rural districts (numbering 663) was 7,851,282. So that the proportion now living under urban and rural conditions may broadly be regarded as 79.3 and 20.7 per cent. respectively.

The following table giving similar statistics of earlier censuses shows the steadily increasing predominance of the urban as compared with the rural population. After 1851, when the proportions were about equal, the urban element gained a definite lead which was rapidly and consistently increased in each decennium up to the year 1901, when it reached 77 per cent. of the total population the much slower advance in the urban predominance since that date has been due more to the gradual encroachment of towns on to adjacent rural territory as part of their natural development than to the more definite tendency of migration from country to town which marked the latter half of the last century.


Thus the population of the rural districts as constituted in 1921 is actually less by 56,274, or 0.7 per cent., than that of the rural districts as constituted in 1911, but compared with the 1911 population within the boundaries of the rural districts as existing to-day it is greater by 4.3 per cent., showing that the growth within the rural districts is not abnormally low, but that it is counteracted by reductions in area occasioned by the gradual extension of urban at the expense of the rural districts.

With the continued extensions of urban area there was a concurrent increase in the separate urban administrative units up to 1911, when they numbered 1,137, but in the past decennium the new creations have been less than the number amalgamated or absorbed, and though the total area has been again enlarged, the number of units has diminished to 1,126.

These are classified in the following table according to the numbers of their population in 1921.

One-quarter of the total population of the country is massed in the larger aggregates exceeding 250,000 persons each, the exact proportion, viz., 25.5 per cent., being practically unaltered from what it was in 1911, viz., 25.4 per cent. Below 250,000 the tendency towards increased urbanisation is more evident; in towns of between 100,000 and 250,000 persons the proportion of the total population has increased from 12.6 per cent. in 1911 to 13.6 in 1921, between 50,000 and 100,000 the increase is from 9.9 to 10.2, and between 20,000 and 50,000 from 12.8 to 13.1 per cent. in the ten years. In the smaller towns, many of which are entirely non-industrial in character and primarily exist as the necessary market centres for the service of adjacent rural areas, the proportionate population like that of the rural districts themselves shows a decrease.


These changes in proportions fail, however, to present a true picture of the relative rates of growth of towns of different sizes owing to the fact that there is a continual process of transfer going on between them a number of the towns near the upper limit of one category in 1911 will have passed the limit in 1921, and so have been classified in a different group at the second census. It is necessary, therefore, to supplement the above figures by others comparing the populations of the towns as they are to-day with the populations of the same areas in 1911, and this is done in the next table, additional columns being added giving similar information in respect of the two preceding intercensal periods.


The increases in the whole country during the three decennial for which figures are given vary from 12.2 per cent. in 1891-1900, to 10.9 per cent. in 1901-11, and 5.0 per cent. in 1911-21, and without exception the groups follow the order of national change showing broadly that the actual movement within an area is dependent primarily on general factors which affect all areas alike. The table does, however, appear to disclose a relation between the several groups, for in each decennium the rate of growth in the smallest towns is above the average, and this rate increases with an increase in size up to about the 50,000-100,000 category and thereafter consistently diminishes, the growth in the largest units being rather below normal. This seems to suggest that a figure in the neighbourhood of between 50,000 and 100,000 roughly marks a limit of effective aggregation beyond which the advantage of further accretion begin to be counterbalanced by increasing disadvantages. Further, in the later decades, the rate of increase in the smallest towns appears to have been relatively higher and that of the larger units lower than before. This tendency is probably associated with the changes in the organisation of industry which are gradually being brought about by factors tending to the dispersion rather than the concentration of population, such as the recent development in transport, the increasing use of electrical power, which can be transmitted over long distances with comparative economy, the necessity of providing workers with more adequate houses and healthier environment; so that the most effective concentration of individuals in urban units may in the future be a diminishing one and the decline in the rate of growth set in earlier than it has in the past.

London and Great Towns.— Among the 1,126 urban areas are 101 (including the Administrative County of London as one district) each of which had in 1921 a population exceeding 50,000. Altogether they account for 18,694,648 persons, nearly one-half of the total population of the country thus being found in large and relatively dense aggregates.

In the following table these are arranged in the order of their 1921 populations and comparative figures for earlier years are added relating to the towns as constituted in 1921.


Again the most noticeable feature shown by this table is the restriction of movement which has taken place in the past ten years. Classifying the towns by their rates of increase,

  1901-11 1911-21
Increase 50 per cent or more 6 towns 2 towns
Increase 30 per cent.—50 per cent 12 1
Increase 10 per cent.—30 per cent 38 17
Increase 0 per cent—10 per cent 42 69
Decrease 3 12

it will be seen that in only three towns does the 1921 population exceed that of 1911 by 30 per cent. or more, as compared with 18 towns in the preceding decade. Even this comparison over-favours the later period, because, of the three towns in question, two, viz. Southend on Sea and Blackpool, are seaside resorts in which the 1921 population was unduly swollen by temporary visitors (see reference to inflation in Appendix A).

The 20 large towns in which the recorded increase was 10 per cent. or more are specified below; and in the adjoining column are given those, numbering 12 in all, in which an actual decrease was shown.

Growth of Population 1861-1921 in each County Borough of which the Population exceeded 250,000 in 1921.

Note.— Population added by extension of boundary is Indicated by the dotted portions of the columns. These portions in each case represent the population of the added area at the Census last preceding the extension of boundary.
The black portion shown above the 1921 column represents not the population of The town in question but the population of all contiguous urban areas.

Population growth in County Boroughs 1861 to 1921
Increases greater than 10 per cent. Decreases.
  Per cent.   Per cent.
Blackpool 64.0 Blackburn 4.8
Southend on Sea 50.0 Bury 4.5
Hendon 44.3 Burnley 3.4
Coventry 20.5 Halifax 2.4
Eastbourne 18.1 Oldham 1.7
Barrow in Furness 16.4 Bolton 1.2
Bournemouth. 15.9 Merthyr Tydfil 1.1
Wallasey 15.7 Bradford 0.9
Darlington 14.9 London 0.8
Luton 14.2 Bath 0.7
Croydon 12.5 Rochdale 0.7
Wimbledon 11.7 Norwich 0.7
Doncaster 11.6
Birkenhead 11.3
Southampton 11.0    
Ealing 10.7    
Chesterfield 10.7    
Newport (Mon.) 10.4    
Grimsby 10.3    
Southport 10.0    

Excluding the seaside towns where inflation was prevalent in 1921, some of the largest increases might be specially associated with war production were it not for the fact that in practically all—Barrow in Furness is an exception—the rate of development was very much larger in the preceding decade.

It is interesting to note that Southend on Sea, the second in the above table, showed the highest increase in the corresponding group of towns for the period 1901-11, while Ealing, Coventry and Wallasey were third, fifth and sixth respectively in that period (see the 1911 Census Report).

A number of the large towns are included in "Greater London," and many of these show increases well above the average.

The presence, among the 12 cases of decrease, of six Lancashire towns is remarkable, particularly having regard to the fact that the population of Lancashire as a whole increased by 3.6 per cent., a rate not greatly below the average of England and Wales. It may be that some of their residents were enumerated as visitors in Blackpool and other neighbouring holiday centres, but this is hardly likely to account for the full loss, and the inclusion in the list of two Yorkshire towns associated with the textile industries is not without significance.

As a measure of caution it should be remarked that, while the successive census returns provide general evidence of the growth or otherwise of various types of areas, the actual figures of individual districts, particularly of the large towns, can only be viewed in relation to the boundaries which delimit them. Development will often occur in suburbs, which may be outside the administrative boundary and would not, in that event, be associated in a census classification with the main area from which it derived its existence. Boundaries are extended from time to time to take in the new accretions, but the changes can only take place at infrequent intervals and would normally lag behind the growth of the population until the latter had reached a comparatively advanced stage. In the denser regions, such as the area immediately outside the boundary of the Administrative County of London, and certain portions of Lancashire, Yorkshire and the Midlands, will be found examples of urban districts either adjoining one another or surrounding and contiguous to an important central area. Where this occurs the movements will be inter-related according to the amount of interaction between the respective populations, and an examination must proceed on a wider basis than that of the administrative unit for a complete account of the changes taking place.

The development of the ten largest towns (other than London) over the past 60 years is illustrated in the adjoining diagram, the open columns represent the successive census populations, the light dotted portions show the populations added by extension of boundary as at the date of the census last preceding the extension, and the black portion shown above the 1921 column in each case represents, not the population of the town in question, but the population of all contiguous urban areas.

London Administrative County and the "Outer Ring."— In the Administrative County of London and the immediately surrounding areas conveniently designated the "Outer Ring" (which together comprise what is usually known as Greater London, and coincide with the total area covered by the City of London and Metropolitan Police Districts) 7,480,201 persons were enumerated, representing about one-quarter of the population of all urban areas and about one-fifth of the total population of England and Wales.


The intercensal increase in the combined area is seen from the above table to have been at the rate of 3.2 per cent., which is less than one-third of the corresponding figure for the preceding decennium and is only about 60 per cent. of the rate of increase in the whole country.

In the administrative county alone a decrease was recorded, representing a numerical loss of 37,162 persons, or between two and three times the corresponding loss which occurred for the first time in 1901-11.

It is possible that, but for the temporary absence of persons who went to swell the numbers in the seaside resorts at the date of the census, a small increase might have been registered in the county. In distributing the excess population of those areas for the purpose of providing estimates of resident populations for use with vital statistics the Registrar-General has credited London with a population in 1921 very slightly in excess of that of 1911. But whatever may be the exact position the decennial change is inconsiderable and signifies little else than that, in the process of the decentralisation of the resident population which accompanies the development of an important commercial or industrial centre and which has been gradually taking place during several decades in the London area, the county boundary for the past 20 years has roughly marked the line within which the normal increase and the outward movement, due to decentralisation, approximately neutralize one another.

This movement is well illustrated in the following statement in which the constituent areas of Greater London are aggregated into a series of zones.

Increase of Population
(- = Decrease.)
1901-11. 1911-21.
Amount. Per Cent. Amount. Per Cent.
City and Inner Metropolitan Boroughs of Holborn, Finsbury, Shoreditch, Bethnal Green, Stepney, Bermondsey, Southwark and Westminster 1,049,473 1,100,274 -8.0 -104,912 19.1
Remainder of London A.C. 3,435,050 85,692 2.6 67,750 2.0
Outside the A.C. but wholly or partly within a radius of 10 miles from Charing Cross 2,662,774 621,601 34.1 220,688 9.0
Outside the 10 miles circle but within Greater London (approximately 15 miles radius) 332,904 62,937 28.0 45,317 15.8

The latest decrease in the population of the innermost Boroughs is marked. It has probably been aggravated somewhat on this occasion by the repatriation of aliens from the East End during the war years. Even so, there is no evidence of any special inward movement with a view to the occupation of the dwelling space vacated, and the decline has proceeded simultaneously with an increase of more than 220,000, or 9.0 per cent., in the zone between the county boundary and the 10.mile circle, and an increase exceeding 45,000 (15.8 per cent.) outside the 10.mile circle. The increase in the whole of the "Outer Ring" is markedly less at 9.7 per cent. than it was 10 years ago (33.5 per cent.), and it will be observed that, whereas on the earlier occasion the percentage increase was greater on the inner side of the 10.mile circle, in the past 10 years the position of the maximum development as measured by the rate of increase has shifted a further stage outwards.

Details of the constituent areas of the several zones are set out in Table 12 of the General Tables Volume. From this table it will be seen that the individual movements, though very uneven, are with few exceptions smaller both in amount and range than those of the period 1901-11.

Civil Parishes and Wards of Urban Areas.— The modern civil parish, which has been defined in several Acts of Parliament as, in the words of the Interpretation Act, 1889, "a place for which a separate poor rate is, or can be, made, or for which a separate overseer is, or can be, appointed," is the unit not only of the poor law system, but also of the more recently reorganised administrative system. It should in consequence have a place in the tables relating to each, but this would involve duplication, and, except therefore where the parish itself is a complete poor law area, the fullest statistics of individual civil parishes have been shown once only in the tables dealing with administrative areas in the County series of Census publications.

Civil parishes differ considerably both in area and population. The numbers and population classification within each administrative county will be found in Table 13 of the General Tables Volume, and in the following table the 14,483 civil parishes in England and Wales as a whole are summarised according to the size of their respective populations.

Number of Civil Parishes   Population in 1921.
  22   No Population
  2,104   Under 100
  4,413   100-299
  2,368   300-499
  2,275   500-999
  2,305   1,000-4,999
  432   5,000-9,999
  263   10,000-19,999
  301   20,000 and over
Total 14,483    

The civil parish, which generally provides an adequate unit as a subdivision of the more extensive rural districts, is not always suitable in this respect in urban areas, some of which, containing relatively dense populations in small areas, comprise but single parishes in themselves. For boroughs and urban districts, therefore, the wards, into which many of them have been divided for local purposes, have been adopted as alternative units of area, and population statistics have been published in the individual County Census Volumes in addition to the figures for civil parishes. The information has been supplemented on this occasion by a statement of the approximate acreage of each ward and the consequent density of population in terms of persons per acre.

Poor Law Unions and Registration Areas.— Statistics of the areas which have been established for the administration of the poor law and which, with a few exceptions referred to in footnotes, are identical with those of the Birth and Death Registration Service, are given in Tables 14 and 15 of the General Tables Volume, the latter showing the 637 individual districts in alphabetical order, and the former aggregating the separate units into 55 County groups. The "boundaries of these poor law union (or registration) counties, it may be noted, differ in nearly every case from those not only of the newer administrative counties but also of the ancient counties, and the populations are not therefore identical with those of the areas already considered. Their range is, however, much the same—from Lancashire and London at the head of the list with populations each in excess of 4,000,000 down to small and numerically insignificant aggregates of less than 100,000—and no statistical interest would be served by a repetition of features hardly distinguishable from those of the administrative areas.

In Tables 14 and 15 (General Tables Volume) the simple population figures of each area have been supplemented, in the interest of the poor law authorities, by a brief age classification dividing the total population of each sex into three groups, viz., under 16 years of age, 16-70, and over 70 years of age.

3. Parliamentary Areas.

Parliamentary Constituencies.— With the redistribution of seats which has taken place under the Representation of the People Act, 1918, the statistical interest in the ancient or geographical county upon which the old parliamentary divisions were based has, except from its historical aspect, almost disappeared. Under that act, the boundaries of parliamentary boroughs and counties were brought into line with those of local government areas as constituted on the 1st October, 1917. This step in the direction of diminishing the confusion and overlapping of boundaries is, from a statistical point of view, most welcome, but the unification is, unfortunately, not a permanent one, since, with the subsequent and continuing changes in the boundaries of administrative areas, divergencies between the latter and the fixed constituency areas are again in existence and on the increase. Between the passing of the act and the date of the census, changes in administrative areas have occurred affecting the following constituencies, the areas and populations of which are no longer identical with those of the administrative areas to which, in 1917, they were assimilated.


Cheshire County.
Altrincham, County Division.
(area only).

Gloucestershire County.
Bristol P.B. (S. &W. Divns.).
Thornbury, County Division.

Lancashire County.
Newton, County Division.
Stretford, County Division.
(area only.)

Lincolnshire (The Parts of Kesteven) County
Grantham, County Division.
Lincolnshire (The Parts of Lindsey) County.
Lincoln, P.B.

Somerset County.
Weston-super-Mare, County Division.

Southampton County.
Portsmouth, P.B.
Southampton, P.B.
Fareham, County Division.
Winchester, County Division.

Worcestershire County.
Bewdley, County Division.
Evesham, County Division.
Yorkshire (West Riding) County.
Leeds, P.B.
Rotherham, P.B.
Sheffield PB
Rothwell, County Division.

Carmarthenshire County.
Carmarthen, County Division.
Lianelly, County Division.

Glamorganshire County.
Swansea P.B.
Gower, County Division.

The population of each constituency is shown in Table 17 (General Tables Volume) and to facilitate comparison with the corresponding electorate, the total population figure has been amplified by a statement of the numbers of each sex above the franchise ages of 21 and 30 for males and females respectively. The actual numbers of electors, male and female, on the autumn register of 1921 have been shown in adjacent columns.

Excluding university representation, there are altogether 509 parliamentary constituencies in England and Wales, 498 of which return a single member to Parliament and the remaining 11 two members each, making 520 members in all. In terms of averages, each member thus represents 72,859 persons of all ages, 38,945 persons above the franchise age, or 34,151 persons actually on the electoral registers.

The following statement shows how far the actual representation per member varies from the general average and the narrowing of the variations as compared with those of 1911, shown side by side roughly indicates the scope of the change wrought by the recent redistribution.

Total Population per
1921 1911
Over 100,000 20 20 83 86
90,000-100,000 30 31 33 33
80,000 76 78 46 48
70,000 167 169 50 53
60,000 142 147 70 73
50,000 63 63 85 88
40,000 10 10 57 60
30,000 20 23
20,000 16 17
10,000 7 7
Under 10,000 (City of London) 1 2 1 2
  509 520 468 490

In 1911 there were no fewer than 83 constituencies in which the population was upwards of 100,000 per member—the Romford Division of Essex returned one member in respect of a population of 313,000, while in the Harrow Division of Middlesex, the Walthamstow Division of Essex, and in the Borough of Wandsworth the similar populations numbered approximately a quarter of a million each. In 1921 the constituencies with populations in excess of 100,000 per member have been reduced to 20 in all, as follows, and some of these, e.g., Southend-on-Sea, Isle of Thanet, Blackpool, owe their presence in the list to the fact that their 1921 populations are abnormal through the presence of summer visitors (see Appendix A).


Constituency. Population, 1921. Electors, 1921.
Durham County Gateshead, P.B. 125,142 55,352
  South Shields, P.B. 116,635 51,054
Essex County Southend on Sea, P.B. 106,010 41,355
Kent County Dartford, County Division 105,440 46,122
  Isle of Thanet, County Division 115,758 38,088
Lancashire County Blackpool, P.B. 125,516 44,927
  Burnley, P.B. 103,157 50,342
  St. Helens, P.B. 102,640 43,757
Lincolnshire (Parts of Lindsey) County. Grimsby, P.B. 110,510 51,970
London County Deptford, P.B. 112,534 51,775
  Greenwich, P.B. 100,450 45,485
  Islington (North Division), P.B. 100,303 46,504
  St. Marylebone, P.B. 104,173 46,339
  Shoreditch, P.B. 104,248 50,156
Northumberland County Wansbeck, County Division 106,270 44,851
Warwickshire County Coventry, P.B. 128,157 60,857
  Nuneaton, County Division 106,617 46,478
Yorkshire (West Riding) County Huddersfield, P.B. 110,102 57,076
Carmarthenshire County Lianelly, County Division. 101,111 47,898
Flintshire County 106,617 48,068

Similarly at the other extreme there are at present only 11 constituencies where the population per representative is below 50,000, as compared with 101 in 1911. It will be observed that the City of London, with a resident population of 13,709 only, has retained its right to send two representatives to Parliament.


Constituency. Population, 1921. Electors, 1921.
Cheshire County Eddisbury, County Division 48,034 22,272
Cumberland County Northern, County Division 46,736 21,545
  Penrith and Cockermouth, County Division. 44,794 21,460
Dorsetshire County Northern, County Division 48,666 24,372
  Western, County Division 48,044 23,602
Durham County Barnard Castle, County Division 48,012 20,455
London County City of London (including the Inner and Middle Temples), P.B. 13,709 44,083
  Holborn, P.B. 43,192 26,449
Suffolk, West. County Sudbury, County Division 49,457 26,034
Worcestershire County Worcester, P.B. 48,833 23,801
Merionethshire County   45,087 21,384

In the last two columns of Table 17 (General Tables Volume) comparison is made between the male and female electorate and the corresponding populations above the ages of 21 and 30 respectively, the age limits which govern the franchise qualification under the Act. For the whole country the electorate forms 95.3 per cent. of the male population and 79.9 per cent. of the females in the said age classes, leaving approximately 5 per cent. and 20 per cent. as the proportions disfranchised respectively. In the case of males, however, the electoral register, representing voting strength, rather overstates the number of individuals entitled to vote, owing to the possession by the same person of residence qualifications and business premises qualifications in different constituencies, and resulting therefore in a certain amount of plural representation. The business premises votes number about 2 per cent. of the whole, and if these be excluded the proportion of men over 21 years of age who are disfranchised may in consequence be regarded as nearer 7 per cent. than the 5 per cent. suggested above. From its nature, the business premises vote does not affect all constituencies alike it may be expected to occur with greater frequency in commercial and business centres, and examples of this will be found in the table, rather more frequently in borough constituencies, where the proportion of electorate to total male population over 21 is in excess of 100 per cent., the outstanding case being the City of London, where there are more than 5 male electors for each resident male over 21. In the case of women the conditions are somewhat different. The bulk of the women on the register are married women entitled to a vote in respect of their husband's occupation of premises, so that the proportion of electorate varies rather more in the case of women than it does in the case of men, because of the additional variable introduced in the proportion of women married in the several areas. In only three constituencies, the City of London and two other London divisions, is the female electorate in excess of the corresponding population.

4. Ecclesiastical Areas.

England, the Channel Islands and the Isle of Man are for ecclesiastical purposes divided into two Provinces—those of Canterbury and York—containing between them 38 dioceses, each of which is subdivided into a number of ecclesiastical parishes, and extra-parochial places. Current population statistics of each of these areas, together with comparative figures for 1911, have been published in a separate census volume devoted to ecclesiastical areas, and in a foreword to that volume will be found an outline of the foundation and development of their territorial organisation up to the present time.

A point of importance to be observed in connection with the 1921 enumeration is that since the date of the preceding census (1911) considerable change has taken place both in the machinery of government and in the territorial arrangements of the Church.

In the first place, by the Welsh Church Act, 1914, the Church in Wales and Monmouthshire was disestablished. Disestablishment was suspended during the period of the Great War in pursuance of the Suspensory Act, 1914, but finally took effect, under the provisions of an amending Act (Welsh Church (Temporalities) Act, 1919) on the 31st March, 1920. Meanwhile, under the provisions of Section 9 (1) of the first-mentioned Act, the inhabitants of each of the border parishes (i.e. those ecclesiastical parishes situated partly within and partly without Wales and Monmouthshire) had elected, by means of a plebiscite, whether they desired their parish to be considered for the purposes of the Act to be in Wales and Monmouthshire or not. The Welsh Commissioners (created by the Act) then, in accordance with the general wishes of the parishioners as thus ascertained, declared what parishes were to be treated as wholly outside or inside Wales and Monmouthshire for the purposes of the Act. It is interesting to observe that of the 21 ecclesiastical parishes held to be "border parishes," only one was declared to be in Wales and Monmouthshire. In accordance with section 9 (2) the Ecclesiastical Commissioners distributed among various English dioceses the 20 ecclesiastical parishes either in Welsh dioceses and situated entirely in England or falling to be treated as situated in England. Eight ecclesiastical parishes, being wholly in Wales and Monmouthshire, were transferred from English to Welsh dioceses.

It will thus be seen that the term "England" as used in the volume devoted to ecclesiastical areas includes a considerable area actually in Wales and Monmouthshire, and excludes a small area really in England.

Secondly, by the Union of Benefices Act, 1919, new machinery was provided to facilitate the unions of ecclesiastical benefices and of ecclesiastical parishes through the Ecclesiastical Commission.

Thirdly, and most important of all, powers have been conferred by the Church of England Assembly (Powers) Act, 1919, on the Church to legislate for itself subject only to the approval of each individual measure by both Houses of Parliament.

Attention may be drawn to the large numbers of extra-parochial places and of detached parts of ecclesiastical parishes that still exist. The detached parts belong in the main to ancient parishes, and have for civil purposes long been amalgamated with the adjacent parishes under the operation of various Acts of Parliament. Examples, however, are not wanting of new ecclesiastical parishes created with a part or parts detached.

5. Other Areas.

The remaining classes of areal division of the country which have received separate recognition in the various census reports are as follows:—

  1. Petty Sessional Divisions. —The 1921 population and the acreage of each division, together with its constitution in terms of civil parishes, are given in Table 9 of the County Volume Series of publications. This information has not been repeated or summarised in the General Tables Volume.
  2. County Court Circuits and Districts. —Population statistics of these areas are given in Tables 18 and 19 of the General Tables Volume.
  3. (c) Education Authorities Areas.— Statistics of the areas of separate Local Education Authorities have been published both in the County Tables and in the General Tables Volume. The areas are in all cases coincident with those of sanitary administrative areas, and so far as the simple population figures are concerned they are but a repetition of those to which reference has already been made.

A statement as to other areal divisions for which census statistics are not available has been given on page 5.

For some of them population figures would probably be of but little utility, even were it possible to obtain them exactly, and in those cases where such figures are required it is almost superfluous to point out that the areal subdivision in respect of published populations is a comparatively fine one and that, for most practical purposes, a sufficient approximation to the populations of any larger divisions can be obtained, when necessary, by ascertaining their constituent units (civil parishes or wards of urban areas) and amalgamating the respective populations as given in the various census volumes.

1 There were approximately 30,000 more non-civilians stationed abroad in 1921 than there were in 1911.

Next Selection Previous Selection