Buildings, Dwellings, Rooms and Families

Next Selection Previous Selection


Scope of the Inquiry.—The detailed examination of the buildings in each area, which is necessary at a census to ensure the complete enumeration of all population, provides an opportunity for the collection of valuable data in regard to the housing accommodation of the people and of the types of buildings in which they are distributed. In view of the exceptional interest and importance attaching to the subject at the present time, this side of the census enquiry was carefully organised in advance, with the view of securing that the records should be as complete and consistent as possible, having regard to the variety of buildings met with in practice, and the conditions affecting their occupation.

Both in the manner of obtaining the original data and in the subsequent tabulation of the results, the treatment of the subject follows generally the principles adopted in 1911 and described in considerable detail in the report for the census of that year. Such modifications as have been introduced, are described hereafter and owe their presence partly to the experience gained at preceding enquiries which suggested certain directions in which the methods of collection and presentation of the statistics could usefully be varied or amplified, and partly to the change in the angle of interest from which the subject is viewed to-day as a result of the peculiar problems of the post-war housing position.

Each enumerator employed at the census of 1921 was instructed to explore carefully the whole of his district, and to enter in his enumeration book the class of each building as he visited it on his route, e.g., whether private house, block of flats or model dwellings, workhouse, hospital, school, hotel, church, chapel, warehouse, block of offices, block of shops, theatre, washhouses, town hall, stables, etc., etc. He then ascertained and noted in respect of the building whether it was actually used or intended, partly or wholly, for habitation. It may be observed here that whereas, in 1911, certain types of building, like churches, theatres, government and municipal buildings were classified in the tables as buildings not used as dwellings, and the numbers of buildings assigned to these classes were slightly deficient in that the exceptional cases with resident caretakers had to be included amongst miscellaneous buildings used as dwellings, the form of the 1921 tabulation contemplated and provided for the enumeration of population in all types of building, and the separate identification of the number of each type, distinguishing in each case those which were actually used or, so far as could be inferred from the nature of the premises, were intended for habitation.

For buildings not used or intended for habitation no further information was required but in all other cases the next step was to ascertain whether the building was designed or adapted for the occupation of one or of a number of families. To avoid the ambiguity which may hitherto have attached to. terms like "house" "tenement" etc., a new term, "structurally separate dwelling," has been introduced as the housing unit serving as the basis of the returns. It was denned in the enumerators' instructions as "any room or set of rooms having separate access either to the street or to a common landing or staircase." Each flat in a block of flats is a structurally separate set of premises. A private house which has not been structurally subdivided is similarly to be reckoned as one set of premises. But where a private house has been subdivided into maisonettes or portions each having a front door opening on to the street or a common landing or staircase to which visitors have access, then each such portion must be regarded as a structurally separate set of premises. The majority of buildings are of course private houses, each comprising but a single structurally separate dwelling in itself, but the rule is not invariable and both for private houses and for the more composite buildings which serve as habitations for various types of families or individuals it was found desirable to arrange for the examination of each building with a view to ascertaining the number of structurally separate dwellings it contained and to make the record of such number an integral part of the enumeration procedure.

Having ascertained and recorded the number of structurally separate dwellings within a building, the next step was to find out what "separate occupations" were comprised in each dwelling, that is to say how many families there were in each dwelling, for whom separate census schedules had to be provided, how many rooms* the dwelling contained, and how these were apportioned amongst the said families or occupiers within the dwelling. After the census had been taken the number of persons comprised in each family was noted in each case, and the general result of this enquiry is that particulars are available as to the number and type of buildings the number of structurally separate dwellings, the number of rooms in the dwellings, the number of private families occupying them, the number of rooms occupied and the number of persons in the families, together with much valuable information derivable from the combination in various ways of these particulars.

In addition to the term "structurally separate dwellings" denned above, the meaning attached to the census use of the terms "family" and "rooms" may also be noted.

Private Family.— For census purposes generally a family is the person or group of persons included in a separate census schedule and comprises, in addition to the ordinary domestic household, a heterogeneous variety of aggregations of individuals enumerated in premises like hotels, boarding houses, hospitals, workhouses, schools, prisons, barracks, ships, etc., etc. These groups may be broadly divided into two categories, private families, with which the housing enquiry is mainly concerned, and the rest which are included under the general term "non-private" families. The term "Private Family" applies to the ordinary domestic household inclusive of indoor servants, a resident caretaker (with or without family) of a house to let or other premises, an outdoor servant (with or without family) occupying separately a building or rooms detached from the house to which it belongs or having no internal communication therewith. A lodger occupying part of a house or flat and not boarding with the family is treated as a separate private family, and where two or more lodgers share a part of a house in similar circumstances, they are treated as one family. A lodger boarding with a family is not treated as a separate family, but is included with the family with which he boards. The residents of business establishments and boarding houses are only included in the private family class when the number of trade assistants or boarders is not greater than the number of members in the employer's or householder's family (including domestic servants).

Private families so deemed comprise about 95 per cent. of the total population.

Rooms1 .— For the purposes of the census, the rooms enumerated are the usual living rooms, including bedrooms and kitchens, but excluding sculleries, landings, lobbies, closets, bathrooms, or any warehouse, office or shop rooms. The term scullery occasioned some difficulty in certain parts of the country and enumerators were instructed to exclude such a room from the count of living rooms, when it adjoined a living room but was used only for purpose of washing, cleaning or cooking, and was not used as a place in which meals are eaten.

Buildings and Dwellings.— The classification of buildings, summarised for the whole country, is shown in Table XV.

In comparing the present distribution with that yielded by the Census of 1911, two small variations introduced in 1921 have to be noted.

  1. Structurally divided private houses are separately shown for the first time, with a view to affording an indication of the extent to which houses originally built for the occupation of single families have been structurally divided to provide separate and independent occupation for two or more families. In some cases where structural modifications had been introduced for the convenience of separate occupiers, the enumerator experienced a little difficulty in deciding whether the alterations were sufficiently complete to bring the several portions of the building within the category of structurally separate dwellings in accordance with the definition, but normally the interpretation of the definition was clear, and doubtful cases were usually cleared up in the course of subsequent correspondence. There is evidence in the returns that it was not always possible to distinguish between buildings originally erected to comprise flats or maisonnettes, shown as group III of the buildings in the following table and those subsequently converted for the purpose, and therefore falling within group II.
  2. Inns and Public Houses in which the business of the establishment is primarily that of retailing liquors are classed with shops, instead of with hotels, the numbers of the latter being thus limited to those of a residential character.


The structurally separate dwellings within the several classes of buildings (other than those in course of erection) are shown in columns 5.8 of the table and number 8,029,863 in all for the whole of England and Wales. The comparable figure derived from the slightly different form of tabulation in 1911 was 7,753,468, so that dwellings may be said to have increased during the decennium by 276,395, or 3.6 per cent., concurrently with an increase in population of 5.0 per cent. Undivided private houses, the class of building to which nearly 90 per cent. of existing dwellings are assigned, have increased their numbers by at least 4.2 per cent., while dwellings scheduled in class III under flats, tenements, etc., are 3.6 per cent. greater than in 1911. The 1911 counterpart of the 44,194 dwellings in the category of structurally divided private houses was not separately identified on that occasion and these must have been included with either private houses or flats, though exactly how they were distributed between the two classes is not known. The fact that they were not so identified and the numbers merged with other groups prevents an exact comparison in respect of either of the three individual classes, and it may be assumed that both the above-mentioned increases of 4.2 and 3.6 per cent. understate the full growth, perhaps to a greater extent in the case of flats, where the total numbers are small in comparison with the numbers of private houses. Putting classes I, II, and III together, that is in the aggregate of buildings designed solely for private residences, the dwellings have increased by 4.8 per cent. during the decennium.

Of the total dwellings in all classes of buildings, 7,759,821, or 96.6 per cent., were in the occupation of private families, 51,209, or 0.6 per cent., in the occupation of non-private families and 218,833, or 2.7 per cent. were classified as vacant.

Vacant or unoccupied dwellings numbered 434,048 in 1911 so that the present numbers represent a decrease of 215,215 or nearly one half of the earlier figure. In view of the extreme shortage of houses in 1921 it is perhaps surprising that the decrease shown is not greater, and that existing unoccupied premises are apparently so numerous. It must be borne in mind, however, that in both the 1911 and 1921 classification, unoccupied dwellings include both tenantless dwellings and also premises merely vacant on census night through the temporary absence of the occupier and that both the returns, therefore, overstate the numbers of dwellings vacant and available for occupation. Moreover, in view of the much larger number of visitors enumerated in seaside and health resorts in 1921 (see reference to inflation in Appendix A), it is probable that the houses vacant through the temporary absence of their occupiers were relatively more numerous in 1921 and that considerably more than one half of the tenantless dwellings of 1911 must have been absorbed by 1921.

Proportion of Families to Structurally Separate Dwellings.— From the following table, which compares for England and Wales the numbers of families with the dwellings they occupy, it will be seen that while the occupied dwellings have increased by 6.7 per cent., the growth in the number of private families has been at about one and one half times this rate, or 10.0 per cent., so that the average number of families per occupied dwelling has increased from 1.09 in 1911 to 1.12 in 1921, though, the families being smaller in average size, the population per occupied dwelling has decreased from 4.93 to 4.85 persons per dwelling.


If it could be assumed that each private family normally requires a separate dwelling, and that the demand for dwellings is accordingly proportionate to the number of families, the increase of the "families per dwelling" ratio from 1.09 in 1911 to 1.12 in 1921 might be interpreted as representing an increase in the deficiency of dwellings of rather more than 200,000 as compared with the position in 1911. A number so obtained, however, will throw little light upon current housing conditions without an examination into the composition of the families and the accommodation in the dwellings they occupy.

The classification of structurally separate dwellings (other than those occupied by non-private families) by the number of living rooms they contain, and by the numbers of families occupying them has been introduced for the first time on this occasion, and an analysis has been given in respect of each urban and rural area in Table 10 of the county census volumes, and for England and Wales as a whole in Table 20 of the General Tables volume.


From the proportion figures of Table XVII it will be seen that self-contained dwellings of three rooms or less account for 16.9 per cent. of the whole, 48.8 per cent., or very nearly one half, contain four or five living rooms only, 28.6 per cent. six, seven or eight rooms, while in 5.8 per cent. there are nine or more rooms. Plural occupation, that is occupation of a single dwelling by more than one family, is present in each of the types distinguished, and, as is natural, its frequency increases with the increase in size of the dwelling, though, owing to the relative scarceness of dwellings of the largest type, the numbers are highest in dwellings of 6.8 rooms, where of 2,278,444 total dwelling units, 334,583 or 15 per cent. were in the occupation of two families each, and 91,537 or 4 per cent. each contained three or more families. Even in the smallest dwellings, that is, those of three rooms or less, 17,718 out of 1,349,390 were occupied by two families and 885 by three or more families each.

Alternatively, it may be stated that of the 8,739,197 private families enumerated in the country at large:

7,006,707 or 80.2 per cent. were living in the single occupation of separate dwellings.

1,195,614 or 13.7 per cent. were living two families to a dwelling.

536,876 or 6.1 per cent. were housed in dwellings containing three or more families each.

Conspicuous as is the difference between the number of families and the separate dwellings available for them and the evident need, therefore, for a considerable increase in the latter, it must be remembered that, in the sense that the term "family" is used for census purposes, the ideal of one dwelling for each family is an impracticable one since a single lodger boarding separately from the occupier is regarded—as on previous occasions—as a separate family. This and similar cases involving a degree of inter-dependence between one "family" and another, render plural occupation inevitable and the total families must always be in excess of the occupied dwellings.

Throughout the country the 4 and 5-roomed group of dwellings usually shows a marked predominance over the other groups distinguished in the classification, but there is considerable variation in the incidence of the smaller and larger premises, as may be seen from the following table, which contrasts counties with the highest and lowest average number of rooms per dwelling.


Broadly speaking, the largest dwellings are to be found in the rural and residential counties and the smallest in the industrial areas, particularly those associated with the mining industry. Of the latter, Northumberland and Durham are prominent in respect of their very high proportions of dwellings of the smallest type; in Northumberland more than half of the total are of three rooms or less and in Durham the proportion of 45.8 per cent. in this group is greater than that of either of the other sizes distinguished. And not only are the small dwellings predominant in these two counties, but they are occupied by a larger number of families on the average than obtains elsewhere, though in the relatively fewer number of large houses the occupation is under average. Counties with the highest average number of rooms per dwelling naturally show an excess in the proportions of the larger sizes, but they are not so regularly constituted in respect of the smaller types London, for example, standing fifth on the list, contains an excessive proportion of 1.3 roomed dwellings, due no doubt to the flats and tenement buildings, which are characteristic of the central portion of this region, while at the same time there is a notable deficiency in the proportion of 4.5 roomed premises. In Surrey and Sussex West, with the sixth and seventh largest overall averages, the 4.5 roomed type is more numerous than all other types combined, but on the other hand, the proportion of 1.3 roomed dwellings is very low indeed.

Size of Family in relation to Number of Rooms Occupied.— The essential aspect of housing and overcrowding questions is dealt with in detail in Table 11 of the County Volumes, where the private families of each area are classified according to size, i.e. the number of persons in the family, and also by the unit of occupation, the number of rooms occupied by each individual family. Similar analyses for England and Wales as a whole and for the aggregates of county boroughs, urban districts and rural districts are given in Table 22 of the General Tables Volume.

It is to be noted that the unit of occupation in these tables is the number of rooms occupied by a family whether this number forms the whole or part only of a structurally separate dwelling 1,732,490, or nearly 20 per cent. of the total of 8,739,197 private families, were living two or more in a dwelling, and in respect of each of these the number of rooms scheduled as the unit of occupation will be less than the total rooms in the dwelling.

The analysis for the country as a whole is summarised in the tables which follow—



  1921 1911
Average size of private family (persons) 4.14 4.36
Average number of rooms occupied per family—    
(a) In all units of occupation 4.55 *
(b) In units of occupation of 1.9 rooms only 4.35 4.52
Average numbers of rooms occupied per person—    
(a) In all units of occupation 1.10 *
(b) In units of occupation of 1.9 rooms only 1.06 1.05

* Information not available.

From Table XIX it will be seen that the commonest unit of occupation is that consisting of four rooms which forms 24.4 per cent. of the whole 20.8 per cent. consist of 5 rooms, 15.5 per cent. of 3 rooms and 12.9 of 6 rooms, so that nearly 74 per cent. of the units of occupation consist of from 3 to 6 rooms, the balance being distributed amongst smaller and larger units, rather more than half going into the former category.

As regards the size of families Table XX shows that 20.8 per cent. of the total private families consist of 3 persons, 18.6 per cent. of 4 persons, 17.7 per cent. of 2 persons and 13.9 per cent. of 5 persons, 71 per cent. of the total thus consisting of from 2 to 5 persons each. Persons scheduled as living alone form 6.0 per cent. of the separate families, or 1.5 per cent. of the total private family population.

It will be seen from Table XX that as compared with 1911, the increase of 796,060 in the number of private families is compounded of two portions, a decrease in the numbers at each size from 6 upwards aggregating to 116,805 in all, coupled with the much larger increase of 912,865 in respect of families of 5 persons or less.

The average size of family has thereby been reduced from 4.36 persons in 1911 to 4.14 persons in 1921, a drop of 5 per cent. The general decline in the size of families between 1911 and 1921 has been consistently observed in almost every section of the country it has been foreseen and commented upon in the successive Annual Reports of the Registrar General and may be regarded as a natural consequence of the increase in the marriage rate in association with a heavily reduced birth rate and the increased (allowing for war deaths) death rate.

The reduction in the size of families, while it may not abate the demands of individual families for separate dwellings, obviously justifies a review of the general position by reference to the reduced requirements of families in point of accommodation. Accordingly, in any comparison of present housing conditions with those of 1911, it is very necessary, in view of the fact that approximately 20 per cent. of the families occupy a portion only of a structurally separate dwelling each, that changes in the size of the individual occupations should be brought into account. Complete comparisons in respect of this latter fact are not available, owing to the absence of information, for 1911, of the number of rooms in units of occupation of 10 rooms or more and such comparisons as are. possible and as are made hereafter, must be understood to be limited to families living in units of occupation of nine rooms or less. These 1911 families, for whom particulars are forthcoming, number 96 per cent. of the total number of families then enumerated, so that their degree of incompleteness is hardly likely to be enough seriously to prejudice the comparisons.

The tables show that whereas, as already stated, the average size of family has dropped by about 5 per cent., the average unit of occupation has decreased from 4.52 rooms per family in 1911 to 4.35 in 1921, a decline of less than 4 per cent., with the consequence that the average number of occupied rooms per person in the country as a whole has improved from 1.05 in 1911 to 1.06 in 1921.

Disregarding, therefore, differences in types of accommodation and size of rooms (on which the census returns are, of course, silent) the population as a whole is, on a general average of rooms per person, rather less densely housed than it was in 1911. The significance of the comparison, however, lies not so much in the amount of the improvement, which is very small, but in the fact that in the face of the widespread housing shortage experienced throughout the country at the present time, present conditions should, on any basis of comparison, appear more favourable than in 1911, when the pressure was certainly less acute than it is to-day.

From the following table, which shows for units of occupation of less than 10 rooms, the numbers of persons living under various conditions of room accommodation and the proportion of those numbers to the total population in private families, it will be observed that the slight improvement in the general average density referred to in the previous paragraph has been occasioned mainly by a redistribution of the numbers recorded at the more favourable end of the density scale.


It will be seen that 32.5 per cent. of the population are returned as living at densities of less than 1 person per room, that for 54.5 per cent. of the total the densities are from one to two persons per room, and that for 9*6 per cent. they are more than two persons to a room. The ratio of more than two persons per room was selected in the census reports for 1911 and earlier years as an approximate comparative index figure for the purpose of measuring the prevalence and distribution of overcrowding conditions. Since, however, that ratio has been commented upon, as though it had been propounded as an absolute standard or a definition of overcrowding, it may be well to observe that its use in this sense in the census statistics implies no judgment whatever as to what in fact constitutes overcrowding. That persons living at this density are more crowded than others and that they can accordingly be regarded as more badly housed in general than those at more favourable densities of occupation, is however, incontrovertible, and it is significant to note that both the numbers enumerated in the class and the proportion they bear to the total show an increase over the corresponding figures of 1911, and further that within the group itself this unfavourable increase in numbers and proportions is most strongly marked at the worst densities shown in the table.

Some further examination of the "rooms per person" ratio is, however, necessary before the general average can be utilised as a basis for the consideration of housing conditions in any given area. The actual unit of distribution is neither the person on the one side, nor the room on the other, since in practice it is always a group of persons, classified in the census as a "private family," who share in common a group of rooms here denominated the "unit of occupation." Hence the ratio of rooms to persons, while forming a simple basis of comparison, is incomplete as a measure of conditions relating not to the individual, bat to the family as a composite whole and it is to the movements in the size and accommodation of families that attention must be directed for a more definite knowledge of the changes in the standards of housing.

The reduction in the average size of family within the several units of occupation or, to speak more accurately, the increase in the numbers of the smaller and the decrease in the numbers of the larger sized families has already been remarked. For the purpose of obtaining comparative figures relating to their accommodation it has been necessary to develop the 1911 statistics rather more fully than had hitherto been done in order to show the relation between different sizes of families and the number of rooms in their occupation. The resulting figures are shown in Table XXII in which comparable figures for 1911 and 1921 are given showing the average number of rooms per family and rooms per person in families of each size.


It will observed that while the number of rooms per family increases consistently with the increase in the size of the family, the increase in the former is at a much slower rate, with the consequence that the individual's share, or the average number of rooms per person usually declines as the family becomes larger. This is consistently true for the whole range of families living in units of accommodation of 9 rooms or less comprising in all 96 per cent. of the total private families, but when the occupants of larger dwellings are brought into account column 5 of the table appears to show that for families of 12 or more persons an increase in the size of family is more than compensated for by the increase in the rooms occupied. The circumstances of large families living in large houses, some of which are probably akin to those excluded from this analysis under the category of non-private families, are, however, somewhat exceptional, and for the bulk of the families the general rule of a decrease in the average rooms per person with the increase in size of family may be said to hold good. A rule of this sort might reasonably have been inferred from general observation, but what might not so easily have been foreseen is the degree of contrast between the housing of large and small families. The maximum individual density shown for the smallest families is in each case about 2f times as generous as the average for the whole and this diminishes by progressive stages to a minimum of less than half of the general average in the case of families of the largest size. The existence of a similar grading in 1911 when the number of empty and available dwellings was far in excess of present numbers indicates that it is a circumstance governed by economic or other conditions which are independent of the presence or absence of houses.

Examination of the comparable figures for 1911 and 1921 shows that during the past decade an improvement in the rooms per person ratio is recorded only in respect of single person families where the density was already at its most favourable point, while a deterioration has taken place for all other sizes of family, including the larger families whose density was already approaching the region of overcrowding.

This deterioration is not, of course, inconsistent with the improvement in the general average density from 1.05 to 1.06 rooms per person. That improvement follows from the fact that the increase in population since 1911 has been accompanied by a relatively greater increase in the total number of rooms occupied. But owing to the increase in the number of families, the average number of rooms per average family has been reduced from 4.52 in 1911 to 4.35 in 1921; although, the average family being smaller in 1921 than in 1911, the density per person in the average family, so far from being worsened has, in fact, improved, as indicated above. But it is in the actual, and not the average, allocation of rooms per family that the important movements above mentioned have taken place. The re-grouping of the population into smaller family units has been accompanied by a re-allocation of the available rooms. As families are now smaller, the weight of their distribution has in consequence moved on the density curve in the direction of the more favourable housing standards. In other words, a larger proportion of the population have become qualified for the housing advantages enjoyed by small families, and have thus absorbed out of the available pool of rooms a larger share than previously fell to the lot of families of the same size, the share remaining to the larger families being correspondingly diminished. But the slight improvement in the total supply of rooms in relation to the total population has not been sufficient to support an increased number of smaller families at the full standards of density which prevailed for such families in 1911 the families of three, for example, which replaced the families of four, housed in 1911 at a density of 1.17 have had to be content with a density of 1.43 instead of reaching the density of 1.49 enjoyed by families of three in 1911. Thus, although as compared with 1911, more of the population is now grouped in the smaller families housed at a density, e.g. of over one room per person, and less in the larger families housed below that standard, the redistribution has been accompanied by a lowering of the 1911 density standards for families of every size save those consisting of single individuals.

The diversity of the room density amongst families of various sizes given in Table XXII clearly shows that an average density expressed by the ratio of total rooms to total population is not of itself an informative guide to the standard of housing in a given community. A density of 1 might be superlatively generous in a neighbourhood where large families was the rule, while the same figure in an area of small families would indicate a relatively inferior housing position. For the purpose of providing index figures by means of which the housing conditions of different areas may be compared with one another and which will at the same time take into account the variations in the sizes of the families found in those areas, a method of standardisation has been adopted throughout the census reports based upon the density ratios for England and Wales, 1911 (shown in column 4 of Table XXII) by multiplying these ratios into the numbers of families of each size in an area, a hypothetical number of rooms is obtained, which, by comparison with the actual rooms enumerated, indicates how far the accommodation in the area exceeds or falls short of the general standard of housing in England and Wales in 1911. The use of such a standard thus enables comparison of the conditions in a series of areas to be made not only with those of the area which provides the standard bat between the individual areas themselves or between the conditions in the same area at different points of time. The choice of the standard is not a matter of unimportance as with the standard of overcrowding referred to on a previous page, it implies no judgment whatever on the part of the census authorities as to the sufficiency or insufficiency of any given number of rooms for families of a particular size it merely represents the conditions actually obtaining in a wide area—England and Wales as a whole—at a comparatively recent date, viz., 1911, The 1911 figures were adopted in preference to those of 1921 for the reason that when the earlier county volumes in which these comparisons were instituted were published, the full 1921 material had not been tabulated and recourse was therefore had to the most recent data then available in respect of the country at large; the latter may from one point of view be preferred since it will be free from any abnormality which might be introduced by the varying degrees of house shortage, experienced throughout the country in 1921.

To have maintained the 1911 standard densities in respect of the families of various sizes enumerated in England and Wales in 1921, approximately 38,149,000 rooms would have been required, whereas the actual returns account for a total of 37,007,000 only in occupation, so that by a comparison of this kind, the position may be said to have worsened in the 10 years by approximately 3 per cent., or alternatively, that the addition of extra accommodation comprising about 1,142,000 living rooms would be required to bring the present standard up to that of 1911.

Comparison of Structurally Separate Dwellings and Units of Occupation.— In the following statement the distribution of structurally separate dwellings (i.e. in regard to the number of rooms they contain) is contrasted with the similar distribution of units of accommodation—whether structurally separate or not—in the actual occupation of individual families containing two persons or more each.

Number of Rooms.* Units of Occupation
Inhabited by
Individual Families.?
Structurally Separate
No. Per cent. No. Per cent.
1-3 2,245,348 27.3 1,305,111 16.8
4-5 3,830,364 46.6 3,803,804 49.0
6-8 1,786,773 21.8 2,210,835 28.5
9 or more 349,856 4.3 440,071 5.7
Total 8,212,341 100.0 7,759,821 100.0
* Exclusive of bathroom, scullery, etc. ? Exclusive of units in the occupation of single person families.

These figures have little direct relation to the present housing shortage indeed, had it been possible to construct a similar table for 1911, it must have pointed to substantially similar conclusions. But if it be true that, as suggested in a previous paragraph, the size of the unit of occupation is primarily determined by economic conditions, then the distribution of the different sizes of units shown above (columns 2 and 3) may be taken as a proportionate measure of the effective requirements of the population for varying quantities of house-room, and, as such, may be compared with the distribution of the available structurally separate dwellings with corresponding accommodation (columns 4 and 5).

It will be seen that in the case of accommodation of five rooms or less the units of occupation exceed the number of structurally separate dwellings, the excess being specially marked as regards the units of 1.3 rooms, occupied, as the table shows, by nearly two and one quarter millions, or 27.3 per cent. of the total families in the country. Conversely, in the case of accommodation of six rooms and over, the structurally separate dwellings are in excess of the corresponding numbers of units of occupation, with the result that a large proportion of such dwellings are occupied by two or more families.

How far it is desirable, or indeed possible, that a structurally separate dwelling should be available for every family is not a consideration with which this commentary is concerned, or which these figures are intended to elucidate. But as all building operations for housing purposes obviously contemplate the erection of separate dwellings for occupation by single families, such figures (or better still, the corresponding figures for smaller areas which can be compiled from the available data provided in the county reports), may afford some useful guidance by illustrating the relative sufficiency or deficiency of separate dwellings of different sizes, and thus suggesting the type and size of dwelling which, so far as these indications go, could most suitably supplement the existing supply.

Families and Housing in Urban and Rural Districts.— The next series of tables compares for London and the aggregates of county boroughs, smaller towns and rural districts, the incidence of families of various size, the rooms they occupy and the relation between them in the manner so far discussed for the country as a whole.


In the order of their respective frequencies of families of different sizes, the county boroughs, smaller towns and rural districts are similar to one another and correspond to the country at large. Families of three persons are the most numerous, followed by those consisting of 4, 2 and 5 persons in each case, the principal difference between them being the rather greater proportion of both very small and very large families in the rural districts than in either of the urban categories. London, on the other hand, is distinctly different in this respect from other areas, for the predominant family is that of 2 persons and the presence of large numbers of lodgers results in a proportion of single person families more than twice that of other towns and nearly as greatly in excess of the corresponding rural proportion. In each case it will be observed that the average family is smaller now than it was in 1911, the amount of the reduction being relatively smallest in the rural areas and greatest in London, The London average it may be noted was markedly lower than that of either of the other areas distinguished above in 1911 and the disparity has accordingly been widened during the succeeding years.


The contrast between London and other portions of the country in respect of the numbers of persons in the family is repeated in the above table in the distribution of the units of occupation according to the number of rooms occupied by the families.

As with the analogous distribution of structurally separate dwellings referred to on an earlier page, the smaller units predominate to a far greater extent in the metropolis than in either of the other sections. The most frequent unit of occupation in this area is that of three rooms, which account for nearly one quarter of the whole, and next to this comes the almost as numerous two-roomed unit. Single room occupation is recorded in respect of 13.2 per cent. of the total families as compared with 3.4 per cent., 2.0 per cent. and 0.7 per cent. in the county boroughs, other urban areas and in the rural districts respectively. Outside London, the four-roomed unit of occupation occurs with the greatest frequency and is followed by that of five rooms in each case, the two together accounting for between forty-six and fifty per cent. of the total units of occupation in each of the other aggregates. Generally speaking, in these areas the frequencies of the smaller units decrease, and those of the larger units increase with decreased urbanisation, so that whereas the proportion of families living in 1.3 rooms is 30.5 per cent. in the county boroughs, 24.0 per cent. in the smaller towns and 20.2 per cent. in the rural districts, the order is reversed in respect of units of occupation of six rooms or more, the proportion of 22.9 per cent. in the aggregate of county boroughs rising to 27.7 per cent. in the smaller towns and to 30.3 per cent. in the rural districts. For all families taken together the average unit of occupation is highest in the rural areas at 5.09 rooms and lowest in London (3.62 rooms) and it will be observed that the average unit of accommodation has, like the average number of persons in the family, decreased in each case since 1911.

But whereas in the county boroughs and also in the smaller towns the amount of the decreases in each of these functions have been practically similar to one another, so that the individual space allotment or average density in terms of rooms per person has remained unaltered, in both London and the rural districts, the decrease in the size of the family has been greater than the corresponding decrease in the rooms occupied, with the result that the overall densities, widely different as they are in the two areas, both show an improvement over the corresponding figures of 1911.

This is shown in the following statement, which also brings out the marked and consistent superiority of the housing position in the rural areas for all classes of families distinguished and the almost invariable deterioration in density as the degree of urbanisation increases from the rural aggregate at the top to London at the other extreme.

Persons in Family. Rooms per person.
London. County
1 2.92 1.82 2.91 3.32 3.70
2 2.01 1.46 1.95 2.14 2.28
3 1.47 1.16 1.43 1.53 1.61
4 1.18 0.98 1.14 1.22 1.28
5 0.98 0.86 0.94 1.00 1.07
6-7 0.79 0.72 0.75 0.81 0.88
8-9 0.63 0.61 0.58 0.63 0.72
10 or more 0.58 0.60 0.50 0.54 0.74
All private families, 1921 1.10 0.96 1.04 1.13 1.22
Private families occupying 1-9 rooms 1921 1.06 0.91 1.02 1.09 1.15
1911 1.05 0.88 1.02 1.09 1.13

The overall average density in the aggregate of rural districts amounts to 1.22 rooms per person, a figure about 11 per cent. more generous than the mean for the whole country. The smaller towns are represented by an average density figure of 1.13 which is also above normal, while below the mean are the county borough, 1.04 rooms per person, and London, the worst in this series, at a figure of 0.96.

The relative position of housing in rural and urban areas and the rather more favourable nature of the intercensal changes recorded for London and the rural areas as compared with the other urban areas is also brought out in the next table, which shows the numbers and proportions of individuals living at densities falling within the conventional overcrowding standard of more than two persons per room, or, in the reciprocal form shown in the table, of less than 0.5 room per person.


In spite of some improvement since 1911 there are still more than 680,000 persons in the Administrative County of London, or over 16 per cent. of the total county population, housed at a density of less than 0.5 room each as compared with 9.6 per cent. for the country as a whole. The total metropolitan population amounts to 12 per cent. of that of England and Wales, while its share of the overcrowded population (less than 0.5 room per person) is nearly 20 per cent. of the whole and of the even more densely housed element, represented by a density of less than 0.3 room each, its share is increased to 25 per cent. Persons living at a density of less than 0.5 room per person have increased from 9.4 per cent. in 1911 to 10.8 per cent. in 1921 in county boroughs, and in the smaller towns the increase has been from 7.4 to 8.1 per cent. in the same period. The position in rural districts, far better than either of the urban areas in 1911, shows a slight improvement, the proportions for 1911 and 1921 being 6.6 and 6.5 per cent. respectively.

Returning once again to the total private families in each area, that is to say, disregarding the special circumstances of the more densely housed as a separate class, comparison of their relative housing positions in reference to the common standard, referred to on page 44, by which allowance is made for the characteristic differences in the distribution of private families by size, is provided in the following statement.

  1921 1911
Excess or
Deficiency (-).
Excess or
Deficiency (-).
Amount. Per
Amount. Per
England and Wales 38,149,087 37,006,712 -1,142,375 -3.0 34,506,500 34,506,500
London 4,741,072 3,696,142 -1,044,930 -22.0 4,336,626 3,495,952 -840,674 -19.4
Aggregate of C.B.s 12,759,668 12,124,939 -634,729 -5.0 10,512,619 10,379,363 -133,256 1.3
Aggregate of U.D.s (less London and C.B.s) 12,935,250 13,064,772 129,522 1.0 12,157,539 12,718,796 561,257 4.6
Aggregate of R.D.s 7,713,097 8,120,859 407,762 5.3 7,500,981 7,912,389 411,408 5.5

The actual number of rooms counted in each area, both in 1911 and 1.921, is contrasted with a "standard" number, which represents the number of rooms which would have been enumerated if the populations had been housed at the densities registered in respect of families of various sizes in the whole of England and Wales in 1911. The standard densities in question are those shown in column 4 or 7 of Table XXII and the "standard" rooms in each case is the aggregation of the products of the standard densities and the families or population in the families, as the case may be, of the several sizes shown in the table.

The relation of house room to population requirement in the four areas given by this comparison is different from that suggested by the average overall density figures of Table XXV only in point of detail. Conditions are at their worst in London where the number of rooms is more than a million, or about 22 per cent., less than the number required by the common standard. Between London and the county boroughs, which come next in order, the gap is nearly three times as great as between either of the succeeding pairs. County boroughs as a class show a deficiency of 5.0 per cent., while in the smaller towns and the rural districts the position is more favourable than the standard to the extent indicated by an excess in the number of rooms of approximately 1.0 per cent. and 5.3 per cent. respectively.

The relative positions were very similar in 1911, but it may be noted that the slight worsening of housing conditions in 1921, marked by the deficiency at the latter date of 3 per cent. for the whole country, has occurred wholly in the three urban divisions, for in the rural areas the position has remained practically stationary at a level of rather more than 5 per cent. in excess of the standard.

Families and Housing in Counties and Large Towns.— The importance of environmental influences in the life of an individual or community of individuals, and the large share contributed to those influences by the conditions under which people are grouped together for the many purposes of their common life, give to the housing enquiry a peculiarly local character and interest. This has been recognised in the census scheme of tabulation by the publication in the county series of volumes of statistics for every borough, urban district and rural district in practically as full detail as that given for the country as a whole. From these and the corresponding tables in the 1911 census report a full examination and comparison of the conditions in individual areas may be made. While considerations of space preclude discussion in a general report of the details of the 1,817 areas treated in this way, a general survey of the different types of area existing throughout the country and of the variations in the several factors bearing upon the housing of different sections of the population may be obtained from the following tables, which illustrate in a brief and comparative form the more prominent differences to be found in units of administrative counties (inclusive of their associated county boroughs) and in individual large towns.

Some idea of the relative types of residential property in the several areas may be obtained from columns 6, c and d of the tables in which private dwellings are classified according to the nature of the buildings in which they occur, and from column e, which shows their average size in terms of rooms. In terms of the three types distinguished by columns b, c and d, the average distribution of dwellings throughout the country as a whole is given by the proportions of 89 per cent. of structurally undivided private houses, 4 per cent. of flats/tenements, etc., and 7 per cent. attached to shops, business premises, etc.; in the great majority of areas, however, the structurally undivided houses occur with greater frequency than is suggested by the average proportions, while dwellings scheduled as in fiats, tenements, etc., are as a rule only a fraction of that indicated by the general average proportion. A prominent feature in the distribution of the types of property is the prevalence of flats, maisonettes, tenements, etc., in the well defined areas of London and the two northern counties of Northumberland and Durham, and their relative scarcity or almost complete absence everywhere else. In London the development of this class of residential accommodation is of course mainly due to the increasing pressure of commerce and industry upon an already completely developed area, for which reason its incidence, as represented by 21 per cent. of the total dwellings in the country as a whole, is naturally greatest in the central and inner boroughs, the proportions in the case of Westminster, St. Marylebone, Finsbury and Southwark reaching 40 per cent. or more. In Northumberland generally and in the adjacent portion of the county of Durham, where conditions approximate more to those on the other side of the Scottish border, tenement property is even more prevalent than in London. In Northumberland the proportion for the whole county is 37 per cent. and reaches a maximum of 68 per cent. in the City of Newcastle, while in Durham the even higher proportions of 79 and 75 per cent. are recorded in South Shields County Borough and Gateshead County Borough, although, owing to the limitation of this feature to the northern half of the county, the average for Durham as a whole is only 19 per cent. In Middlesex and Surrey, influenced no doubt by the proximity of the Metropolis, the proportions are 8 and 4 per cent. respectively, and in East Sussex the proportion is also 4 per cent., but in no other county does it reach as much as 2 per cent., the great bulk of the residential property in all these areas consisting of structurally undivided private houses.

Dwellings attached to shops and other business premises average 7 per cent. of all dwellings in the country at large and vary in the county areas from 3 per cent. in Northumberland to 11 per cent. in London Administrative County. In the exceptional circumstances of the City of London itself the proportion is as high as 67 per cent. the next hi order, but far below, being the Metropolitan Borough of St. Marylebone, 18 per cent. In 12 other Metropolitan Boroughs and in the County Boroughs of Brighton, Hastings and Plymouth, the proportions are not less than 10 per cent.

In respect of room accommodation, the counties with the highest and lowest average number of rooms per dwelling have already been referred to hi conjunction with Table XVIII on page 39. In the list of large towns, the highest positions are claimed generally by some of the metropolitan boroughs, residential areas just outside the London boundary and fashionable watering places, e.g. Hampstead 7.96 rooms per dwelling, Paddington 7.53, Kensington 7.09, Hornsey 6.97, Bournemouth 6.95, Stoke Newington 6.73, Willesden 6.59, St. Pancras 6.57, and Islington 6.54. At the other end of the scale are found towns of Northumberland, Durham, and the West Riding of Yorkshire, viz. South Shields 3.34 rooms per dwelling, Dewsbury 3.42, Gateshead 3.44, Tynemouth 3.62, Newcastle on Tyne 3.65, Halifax 3.72, Huddersfield 3.75, and Bradford 3.96; amongst these, also, occurs the City of London, with its large proportion of caretaker premises, standing sixth from the end, with an average of 3.70 rooms per dwelling.

The average sizes of private families in 1921 and 1911 are shown in columns g and h. In the county areas the 1921 figure varies between 3.76 and 4.62 persons, and in the large towns between the rather wider limits of 3.31 and 5.09 persons. The largest families are associated with the mining counties, Glamorgan 4.62, Stafford 4.62, Durham 4.61, Monmouth 4.56, Northumberland 4.45, Flint 4.42, Cumberland 4.40, Brecknock 4.40, and the smallest with agricultural areas generally. Apart from a broad division of this kind, the families of Wales appear to rule rather higher than those of comparable areas in England, and the position of London—lowest but two in the county list—is somewhat exceptional, though not unexpected, owing to the high frequency of families of the smallest types, in that the average family here is smaller than in many agricultural counties. Outside the County of London and its residential suburban area, the association of larger and smaller families with particular types of towns may be observed in the following lists of extreme cases recorded amongst the large towns of Table XXVII.

Largest average families Smallest average families
St. Helens C.B 5.09 persons Plymouth C.B 3.55 persons.
Warrington C.B 4.83 persons Bath C.B 3.59 persons
Stoke on Trent C.B 4.82 persons Halifax C.B 3.60 persons
Rhondda U.D 4.81 persons Hastings C.B 3.68 persons
Bootle C.B 4.80 persons Exeter C.B 3.71 persons
West Bromwich C.B 4.73 persons Cambridge M.B 3.74 persons
Aberdare U.D 4.71 persons Gillingham M.B 3.78 persons
Smethwick C.B 4.67 persons Rochdale C.B 3.79 persons
Walsall C.B 4.61 persons Brighton C.B 3.79 persons
Merthyr Tydfil C.B 4.59 persons Bradford C.B 3.81 persons

It will be observed that the families of 1921 are smaller than they were 10 years ago in every one of the towns given in Table XXVII for which comparative figures are available, and that the reduction is almost as consistent in the county aggregates, the only two exceptions being Cardiganshire, which was below the average in 1911 and now shows a small increase, and Flintshire, where the position has remained stationary.



For a measure of the relative housing positions in different parts of the country, several factors, expressing, in one form or another, the relation of the population to be housed with the accommodation available and each bearing upon a slightly different aspect of the subject, are given in summary form in the tables. They have been dealt with in previous pages in respect of the country as a whole and it now remains to be seen how far the conditions of individual counties or of important towns, as illustrated by these factors, differ among themselves and how widely they range from the mean values of the whole aggregated together. The factors in question are four in number, viz. (1) the average number of families per dwelling, (2) the average number of rooms per person, (3) the percentage surplus or deficiency of rooms in relation to a common standard number based upon density ratios of 1911 as described on page 44, (4) the proportion of population housed at a pressure of more than two per room, the ratio of two persons per room or its reciprocal -room per person being the dividing line adopted in the census report as a practicable, if rough, guide to the extent of overcrowding.

The first of these, the average number of families per dwelling, is given in column f of Tables XXVI and XXVII and may be read as a measure of the amount by which conditions in a given area fall short of the theoretical ideal represented by a structurally separate dwelling for each family. For the country as a whole there are on the average 1.13 families to each dwelling and the corresponding county figures (exclusive of the County of London) range from 1.01, the nearest approach to this ideal position being recorded in counties like Huntingdon, Rutland, Westmorland, Anglesey, etc., to 1.26 in Middlesex, 1.21 in Devonshire, 1.19 in Essex and Sussex (East) and 1.18 in Glamorganshire. The exceptional conditions of the metropolis, with its large proportion of single lodger families and its relatively large houses, place it in a position apart from all other counties, for here the family pressure is represented by the high ratio of 1.60 families per dwelling, only 38 per cent. of the families in the whole county having a separate dwelling apiece, 32 per cent. being housed two families to a dwelling and the balance of 30 per cent. in dwellings each containing three or more families.

In many of the counties where the number of families per dwelling ratio is high, it will be found that the houses are above the average in point of size inevitably so, perhaps, since the increase in the number of private families has been accompanied almost everywhere by a diminution in the size of the family and the corresponding change in housing requirement, viz. for a larger number of dwellings of relatively smaller sizes can be met, however inadequately, by expansion within existing dwellings to a greater extent where the houses are large than where they are small. Particularly is this the case where in conjunction with houses above the average in size, the families of the area are abnormally small. London is probably the best example of this combination, and the large disparity between the number of families and the number of dwellings serves to mark the extent of the change in the character of the area since the date of the erection of the existing property. The Metropolitan Boroughs of St. Pancras and Islington for instance, one-time fashionable residential neighbourhoods, with large houses built for the occupation of single well-to-do families, have now become closely associated with industries both within and without their borders and the houses, or so many of them as still retain their character of private dwellings, now serve the needs of a much larger number of families in a relatively poorer section of the population, the ratios in these particular Boroughs exceeding two families per dwelling.

For a more direct measure of comparative housing, that is to say of the pressure of population upon the available living accommodation, an index must be sought combining the effect of the variations in size of dwellings in association with the corresponding variations in size of families and number of families per dwelling. The simplest criterion satisfying these requirements is the room density or average number of rooms per person. It has already been stated that for England and Wales as a whole this factor shows a slight improvement over the corresponding figure for 1911, and now stands at 1.10 rooms per person. The corresponding ratio in county divisions will be found from column k of Table XXVI to vary between 0.77 and 1.45; the Counties of Durham and Northumberland are incomparably the worst in this series with densities of 0.77 and 0.80, and they are followed at some distance by London 0.96, Staffordshire 0.99, Monmouth 1.00, The West Riding of Yorkshire 1.01, Glamorganshire 1.02 and Lancashire 1.04; the most favourable densities are found generally in the non-industrial counties, e.g. Isle of Wight 1.45, Cardigan 1.42, Rutland 1.41, Merioneth 1.41, Radnor 1.40.

The simple density, like the families per dwelling ratio referred to above is however only valuable as an indication of where the worst conditions are likely to be found. Throughout the census reports, which have been separately published in respect of each county, it has repeatedly been shown that the ratio of the number of rooms to the number of persons in an area is a totally inadequate guide to the actual conditions within the area. In every urban and rural district, individual family densities vary very widely from the average, and it has been shown that to a large extent the individual family density depends upon the size of the family, and that the smaller families require or are able to command a share of the available house room which, in terms of the rooms per person ratio, is four or five times as great as that obtained by the largest families. The existence of similar conditions in 1911, when houses were relatively plentiful, further suggests that they arise from causes which are independent of the supply of houses and would not therefore be materially varied merely by the provision of additional accommodation.

Part of the differences disclosed between the average room densities of separate areas must therefore be due to differences in the incidence of large and small families within them. In the counties referred to above as showing the least favourable densities, the families in all, except London and the West Riding of Yorkshire, are well above the average in size, and a further and perhaps better index to areal conditions will be one in which variations associated solely with differences in size of families have been eliminated. The method of attaining this comparison has already been described and the resulting factors are shown in columns r and s of Table XXVI as the difference between the actual rooms in an area and a hypothetical number, obtained by allotting to families of each size in the area a fixed number of rooms appropriate to that size according to the standard adopted. By this comparison Northumberland and Durham, with rooms 25.2 per cent. and 24.7 per cent. below the standard allotment, are still at the worst positions in the county list, but the comparison is slightly less unfavourable than that suggested by the room density figure, and it will be observed that the two counties have changed places, Northumberland now taking the lowest place. The position of London, on the other hand, is considerably worsened, and it now ranks as a close third to the two northern counties, with a deficiency of rooms computed at 22.0 per cent. of the standard number. Other counties for which a deficiency is recorded are, in order, West Riding of Yorkshire (9.0 per cent), Lancashire (4.2), Stafford (3.2), Monmouth (3.1), Essex (1.8), Denbigh (1.6), Middlesex (0.3) and Devonshire (0.1). It will be observed that in the two last-named counties the deficiency of rooms is co-existent with an average room density above the normal, while in Glamorgan, Cumberland and Warwick, in each of which the average room density is below normal, the actual numbers of rooms enumerated are in excess of the standard number.

The fourth test, that given by the proportion of persons living in overcrowded conditions, which for the sake of securing comparative figures have been taken as represented by densities of less than one-half a room per person, is also influenced by the fact that the lowest densities occur amongst the largest families and where, therefore, there is a high proportion of large families, there will usually be an abnormal degree of overcrowding. For the country as a whole 9.6 per cent. of the population is found within the overcrowding limits and in Northumberland and Durham, again the worst in the series, the proportion rises to 30.8 per cent. and 29.5 per cent. respectively, followed by 16.1 per cent. in London, West Riding (11.5 per cent.), Stafford (11.1 per cent.), Cumberland (10.6 per cent.), Denbigh (10.1 per cent.) and the North Riding of Yorkshire (9.7 per cent.)

Comparative ratios for boroughs and large towns are given in Table XXVII. The units are smaller and the range of variations consequently wider than those discussed in relation to county units, but the types of distribution are precisely the same, and it will be sufficient here to illustrate the worst conditions shown amongst these areas by a selection of the towns both where average room density is low, say not more than 0.95 rooms per person, and where the comparison of the enumerated rooms with the standard number discloses a deficiency of, say, 10 per cent. or more. The towns are classified in order of the average density.

Borough or Town. Average
families per
Average rooms
per person.
per cent.
of rooms.
Proportion per
cent. of population
living more than
2 per room.
Shoreditch Met. B. 1.85 0.65 41.1 32.0
Finsbury Met. B. 1.84 0.67 42.8 34.0
Bethnal Green Met. B. 1.51 0.68 36.3 27.8
Stepney Met. B. 1.52 0.69 34.0 29.0
South Shields C.B. 1.07 0.71 32.2 36.5
Gateshead C.B. 1.08 0.72 31.3 37.0
Southwark Met. B. 1.62 0.73 36.3 23.5
Bermondsey Met. B. 1.57 0.75 31.3 23.2
Poplar Met. B. 1.62 0.76 28.5 21.2
Sunderland C.B. 1.46 0.76 28.9 33.8
Newcastle-on-Tyne C.B. 1.08 0.78 27.8 33.6
Tynemouth C.B. 1.07 0.78 27.8 34.4
St. Helens C.B. 1.07 0.79 15.2 21.0
St. Pancras Met. B. 2.30 0.82 36.1 22.4
Islington Met.B. 2.12 0.85 31.2 19.4
West Ham C.B. 1.43 0.85 19.2 16.4
Wigan C.B. 1.17 0.87 18.5 15.3
Dewsbury C.B. 1.00 0.88 24.2 18.0
West Bromwich C.B. 1.04 0.88 10.3 17.1
Middlesbrough C.B. 1.09 0.89 13.3 16.0
Dudley C.B. 1.03 0.91 11.4 16.8
Plymouth C.B. 1.84 0.93 25.7 16.9
West Hartlepool C.B. 1.10 0.93 11.3 17.8
Carlisle C.B. 1.04 0.94 13.3 19.0
Darlington C.B. 1.06 0.94 11.2 17.2
Tottenharn U.D. 1.40 0.95 13.2 10.0
Hackney Met. B. 1.67 0.95 18.6 11.5

From such guides as these given in Table 30 of the General Tables Volume for every borough, urban district and rural district in the country, a fair comparative view of the housing conditions in each locality may be obtained. They provide a simple if approximate means of classification, and from them a selection may readily be made, if desired, of those areas for the more complete examination of which the greater detail given in the county volumes or a more exhaustive knowledge of local conditions is desirable.

1 In 1911 the number of rooms in the occupation of each family or other occupier was inserted on the census schedule by the head of the family or other person in occupation or in charge of the dwelling (house, tenement or apartment). In 1921 the duty was assigned to the enumerator it appeared to form a natural step in the enumeration procedure and was thought to be of advantage with a view to securing a greater degree of uniformity in the treatment of this record particularly in those cases where a single separate dwelling was shared by two or more families and the rooms—some of which might be used in common—required to be apportioned amongst the several occupants.

Next Selection Previous Selection