Beveridge: Benefit Rates and the Problem Of Rent

193.     The rates of benefit or pension provided by social insurance should be such as to secure for all normal cases an income adequate for subsistence, on the assumptions :—

(a)   that assistance will be available to meet abnormal subsistence needs;

(b)   that voluntary insurance and saving to provide for standards of life above subsistence minimum will be encouraged and made easy.

It is assumed, further, that there will be allowances for all dependent children, while the parent on whom they are dependent is on benefit or pension. The rates of allowance required for subsistence needs of children are considered in this section. The general question of children’s allowances, both when the parent is earning and when he is not earning, is discussed in paras. 410-425.

194. Determination of the minimum income required for subsistence, though simplified by the foregoing assumptions, remains a difficult problem on two grounds:—

(a) No reasoned forecast can now be made of the cost of living at the time when the insurance scheme may be expected to come into force. All that can be done is to make estimates of subsistence income at pre-war prices (for convenience the year 1938 is taken) and to say how these estimates would be changed for any assumed increase in prices above the pre-war level. The cost of living is not the same for all families, or in all parts of the country. The main difference is in respect of rents, which differ markedly
between London, the rest of England and Scotland, and between industrial and agricultural households.

  1. The problem presented by inequality of rents is discussed in paras. 197-206, as a preliminary to determining rates of benefit. Thereafter sub­sistence needs are considered separately for persons of working age, for retired persons, for young persons and boys and girls of working age, and for children below working age.

  2. It will be seen, from the discussion which follows, that any estimate of subsistence income for the population as a whole during unemployment and disability, even at known pre-war prices, is to some extent a matter of judgment.   Nor can any single estimate, such as is necessary for the deter­mination of a rate of insurance benefit, fit exactly the differing conditions of differing households;   while the main differences in the cost of living arise through variation of rents, there are differences also in the cost of fuel and other articles.    The calculations made have been prepared in consultation with a Sub-Committee including Professor A. L.Bowley, Mr.Seebohm Rowntree, Mr. R. F. George and Dr. H. E. Magee, and in respect of items other than rent, have been approved by the Sub-Committee as affording a reasonable basis for fixing rates of unemployment and disability benefits which at 1938  prices would provide a subsistence minimum in normal cases. In regard to rent, the Sub-Committee express the view that no single figure can be justified on scientific grounds as fitting the needs.

The problem of Rent

197.  Rent has three characteristics differentiating it from other forms of expenditure:—

(i) Rent varies markedly from one part of the country to another,

(ii) Rent varies markedly as between different families of the same size in the same part of the country,

(iii) Expenditure on rent cannot be reduced during a temporary interruption of earning as that on clothing, fuel or light can.

  1. The first of these differences between rent and other necessaries is illustrated by Table IV giving, for industrial households in each division of Britain and for agricultural households, the weekly expenditures in 1937-38 on rent and on food, clothing, fuel and light respectively, as deduced from the Ministry of Labour Family Budgets. (The main results were published in the Labour Gazettes for December  1940 and January 1941.)  In order to allow for the differing numbers of persons per household in different divisions of the country, the expenditures are shown in columns 4 and 5 for ” standard households ” of the same size in all cases; adjustment to standard households reduces the expendi­ture shown where the number of persons per household is above the average for the country as a whole and increases it where the number is below average. In columns 6 and 7 these standardised expenditures are shown for industrial households in each division as a percentage of the mean for all divisions.

  2. It will be seen that, for food, clothing, fuel and light, the percentages for the separate divisions all lie in a narrow range, between 94-2 and 104-9,with agricultural households spending 76-6 per cent, of the industrial mean. For rent, the range is much greater, from 70-4 per cent, in Scotland to 148-1 per cent, in London in industrial households, with agricultural house­ holds spending only 43-5 per cent, of the industrial mean. The average of the actual rents runs from 16-0 shillings a week in London to 7-6 shillings in Scotland for standard industrial households, and is 4-7 shillings in agricultural households.  The expenditures of agricultural households relate to a time before the recent raising of agricultural wages.

Expenditure on Necessaries 1937/8
Table 4

Source: Ministry of Labour Family Budget Enquiry

  • The figures in brackets give the numbers of households in each division for which budgets were obtained. The divisions are those of the Ministry of Labour and National Service.
  1. Table IV gives averages for large regions. It conceals the second characteristic of rent, that within any region there are great differences of expenditure by individual households. The cost of necessary food, clothing, fuel and light is much the same, or can be made much the same, for all households of the same size in the same region. This is not true of rent. The wide variation of the rents paid by individual households is sufficiently illus­trated by the following table from the Report of the Unemployment Assistance Board for 1938.
Distribution of Rents Paid by Applicants for Unemployment Assistrance 1938
Table 5
  1. The average weekly rent paid by all industrial households included in the Ministry of Labour Family Budget in 1937-38 was 10/9. If any single figure is to be taken as the pre-war requirement for rent in fixing rates of benefit, a figure of 10/- a week is the natural suggestion.  The figure should be below rather than above the average shown in the budgets, since these cover families living well above the subsistence level; the average rents of applicants for unemployment assistance deduced from the table given above for each of the three regions would be appreciably below those shown by the family budgets for those regions.  Mr. Rowntree by independent enquiry reached 9/6 a week as the best single figure for rent requirement in 1936. The figure of 10/- is for a household. For solitary individuals a figure of 6/6 is taken.

202. But neither 10/-, nor any other single figure, can fit the true require­ments even reasonably well. A glance at Table V shows that an allowance of 10/- a week for rent in 1938 would have been anything from 2/6 to 7/6 too much for more than two-thirds of the Scottish households and anything from 2/6 to 10/- and upwards too little for half the London households. In no part of the country would it have been within 2/6 of the actual rent for as many as half the households. It would have been at least twice as much as was needed for more than half the agricultural households.   With the present variety of rents, it is not possible to fix any uniform rate of insurance benefit as meeting subsistence requirements with any accuracy. Even when differences in the size of the family have been provided for by allowances for dependants, any uniform rate must be many shillings a week too high for many cases and many shillings too low for many other cases,

  1. These misfits are made more serious by the third characteristic of rent—that it is a form of expenditure in which adjustment or postponement, when earnings are interrupted, is difficult or impossible. A man who feels able to commit himself to a high rent because his earnings are high, remains liable for it when his earnings stop; unless he can meet the rent, he has not enough for subsistence. This consideration is relevant particularly to un­employment, sickness or accident, whose onset cannot be foreseen;  for retirement, which can be foreseen, there is greater possibility of adjusting rent to prospective resources.

  2. In the social surveys which have been made shortly before the present war of living conditions in various British towns, Mr. Rowntree and others, in judging of subsistence needs  and estimating what   proportions of the population are in poverty or in varying degrees above it, have invariably and unavoidably allowed for the actual rent paid by each family and have calculated the income required for the other necessaries—food, clothing, fuel and so on—according to the size and constitution of the family. The question arises whether the rate of insurance benefit ought to be determined on the same principle, securing to each insured person when his earnings are interrupted the actual rent paid by him and a fixed weekly sum on top of that for his other needs. The suggestion that benefit should be determined in this way was made by the Association of Municipal Corporations, speaking from then: practical experience in dealing with need, and by the Fabian Society. It was strongly urged by some members of the sub-committee who were asked to advise on the rates of benefit that would ensure adequate subsistence.

205.   There is no doubt that if insurance benefit could be made variable, so as to allow for actual rent and not for an assumed average rent, it would meet subsistence needs more closely. This is the main argument for the proposal put to the Committee by the Association of Municipal Corporations and the Fabian Society. The proposal has other advantages also.    First, if the actual rent is covered in every case, children’s allowances can safely be made exclusive of rent. There will be no risk of benefit being inadequate because a large family has made a large rent necessary. Second, the differentiation suggested elsewhere (paras. 111—113) between the benefit rates for single and for married women would take place automatically. The former will have rent to pay; the latter, apart from exceptional cases, will not be responsible for rent. Third, the future course of rents is, if possible, even more problematical than that of prices generally, and may be different in different parts of the country, according to the housing policy of the Local Authority. It will be anomalous that of two people in the same town getting the same benefit for subsistence, one should be comfortable on a low rent in a subsidised municipal house or with a rebate because he is unemployed, while the other is left in want because he is paying a rack rent.

  1. If benefit is to be related to needs, the case for adjusting it, if possible, to the actual rent paid is strong. This adjustment can be made and is made in paying assistance. Should it be made in determining benefit? Should men who pay the same contribution get different rates of benefit when they are unemployed or sick, according to the rent which they pay? The suggestion raises questions of principle and of practice.

  2. On the question of principle, it can be argued that, as men for the same contribution get different rates of benefit, according to the number of dependants they must maintain, there is nothing unjust in differentiating benefit by reference to the rent which men have to pay. When a man contributes, he does not know just how much rent he will be paying when he next falls out of work or is sick. He ought to be insured against that liability for rent, whatever it may be;   social security, as the representatives of the Fabian Society put it, should guarantee the home, and an income for other necessaries.

  3. Against this, it can be argued that the analogy between providing for dependants and providing for rent is a false one, because the amount of rent which a man pays is not independent of his will or of his financial resources; a high rent is to some extent at least the sign of a larger income; it is paid not for subsistence but as part of a standard of living above the minimum.   Men with larger incomes do not as such have more dependants, so that giving allowances for dependants does not mean that high wage workmen get on the whole more benefit than low wage workmen when unemployed.  But to adjust benefit by rent would mean this: on the average the skilled men would be found paying higher rents than the unskilled; for equal contributions they would get, on an average, higher rates of benefit when unemployed or sick.   Granted that the man with greater liabilities for rent should be insured against them, it can be urged he ought to be so by voluntary insurance to supplement the compulsory subsistence benefit.

  4. This argument raises the fundamental question of fact as to whether rents above the average represent the wishes or the necessities of those who pay them. It is a question to which a clear-cut answer cannot be given:  the answer probably varies ‘from time to time and from one town to another, according to the housing situation. Such direct evidence as is available about the relation of rent and income shows that while rents, on an average, increase with income, the increase is not rapid, so that the proportion of income devoted to rent falls as income rises.   The following Table VI relates to a random sample of more than 6,000 applicants for War Service Grants in 1939-42. The facts as to voluntary insurance payments are referred to in para. 286 and in Appendix D on Industrial Assurance. Here the table is used only in relation to rents.


Rents and voluntary insurance payments of applicants for war service grants in relation to income, 1939-42

Average    Voluntary

.Income (*) Average Rent per household

Insurance Payments

shillings per week Shillings Per cent of Income


Per cent of Income.

Under 40 10.2 33.5



40-60 10.2 20.4



60-80 11.2 16.0



80-100 11.9 13.2



100-120 12.4 11.3



120-140 14.0 10.7



140 or over 15.3 8.4



All households 11.3 15.1



(*) Pre-service income.

There is no difference between the average rents paid by households with incomes below 40/- a week and between 40/- and 60/- ; thereafter the average rent rises slowly to 12-4s. in the group between 100/- and 120/–of income ; beyond that it rises more rapidly. The table suggests that rents behave in main as necessaries in households with incomes up to £6 a week. The percentage of total income spent on rent falls from a third in the poorest group to a twelfth in the group with incomes of more than 140/- a week. These results for the country as a whole in war-tune are similar to those for York in 1936. Mr. Rowntree’s investigation showed that the average rents paid by the income classes into which he divided the working-class population differed only by l-3s. a week as between the poorest and the most prosperous. This appears from Table VII below.

TABLE VII rents in relation to income in York, 1936

Net income after deducting rent and rates

Rent including rates Shillings

(incl. amount paid for rent) . Shillings Rent as per cent, of Income

Class A

Under 33/6




Class B

33/6 and under 43/6




Class C

43/6 and under 53/6




Class D

53/6 and under 63/6




Class E

63/6 and over




Source : Rowntree, Poverty and Progress, p. 262.

The net incomes shown are those left after allowing for rent so that a high rent in itself tends to put the household in a lower income class ; this accounts no doubt for the fact that the rents in Class A are higher than those in Class B. But subject to this caution, the table supports the view that the level of rents is largely a matter of necessity, not of choice. Undoubtedly, those whose work compels them to live in London cannot in general avoid paying rents materially higher than their fellows elsewhere; those who in any region pay rents above the regional average often do so because they have no alternative. On-the other hand, individuals can and do choose to some extent between expenditure on house-room and other forms of expenditure ; sometimes they pay high rents because they can afford it or because they value., good accommodation more than other things ; sometimes they pay low rents at the cost of heavier expenses in travelling to work. The amount of rent that a man pays represents often both a necessity and a comfort or a luxury, but is not divisible between these purposes. On the question of principle raised by the proposal to adjust security benefit to the actual rent paid by the individual, reasonable men may take different views.

  1. On the question of practice, it is clear that to make the benefit vary with the actual rent would involve additional administration. Every applicant for unemployment or disability benefit or retirement pension (if the plan applied there) would be required to state the rent for which he was responsible, and the benefit to be paid to each applicant would have to be calculated separately. This in itself is not a serious objection ; unemployment benefit payments have now to be adjusted individually, according to the number of  dependants. The extra work involved in this question on the application form and in checking the rent paid, where necessary, would be relatively small. A much larger practical difficulty would arise in determining the amount of the rent for which the applicant was responsible, in cases in which he was not the only earner in the house—where for instance he was one of an earning group, or was a householder with independent lodgers or a lodger himself. It is clear that such questions would sometimes prove very difficult; they would arise often enough to add appreciably to the task of administering benefit. All that can be said is that the administrative difficulties of adjusting benefit for rent cannot be regarded as insoluble.  They should be faced, if the proposal is necessary and right in principle.

  2. As a modification of ,the proposal to adjust benefit to rent which would simplify administration while preserving its principle, it has been suggested that a standard allowance for rent, say 8/- a week, should be included in the benefit, and that  anyone paying more rent than this could apply to have his benefit increased accordingly, up to a maximum rent of, say, 16/-.  This plan would reduce the number of cases in which the actual rent had to be taken into account and would thus reduce administrative labour. But it would not touch the main practical difficulty of determining for what rent any particular applicant in a composite household was responsible, and, unless the standard allowance was made low, it would increase appreciably the cost of benefit as compared with the proposal to allow for actual rent in every case.   There would be no saving on the persons with low rent, (including all the agricultural labourers) to set against the additional payments to persons with high rent; if, to avoid this result, the standard allowance was put low, the saving on administration would be small. It can be urged, on the other hand, that unless some allowance is made, either in this, way or in another, for the exceptionally high level of rents in London, substantial supplementation of benefit by assistance may prove unavoidable in London. Even if differ­entiation for rents in the manner now suggested increased expenditure on benefit, this might be less than the saving which it would bring about on assistance.  If the principle of adjusting benefit to rent without any difference of contributions is admitted, the suggestion made in this paragraph is probably the most practical way of giving effect to that principle.

  3. As a second modification, it has been suggested that allowance for actual rent should be confined to the working age benefits, for unemployment and disability, and that retirement pensions should be uniform. There is much to be said for this modification on grounds of principle and of practice, While men cannot foresee unemployment and disability, they can foresee the coming of old age and its date; they have time to adjust expenditure to the means which they have been able to provide, through compulsory insurance and through voluntary additions to it. Pensions, moreover, will as a rule be drawn for much longer periods than unemployment or disability benefits. It will appear indefensible that those who just before retiring have been able to secure good accommodation at a relatively high rent should thereby retain this advantage for the rest of their lives, in kind if not in cash, as compared with those who have been less fortunate or less foreseeing. On the other hand, if those who are already drawing pension on the basis of one rent are free to move to more expensive accommodation and have their pension increased accordingly, pensions will come to look like a subsidy to landlords. . On all grounds of principle and of practice, the basic provision made by compulsory insurance for retirement pensions should be uniform for all citizens in all regions.

  4. With these two modifications, the proposal to vary the benefit in individual cases so as to allow for the actual rent paid cannot be dismissed either as impracticable or as clearly wrong in principle. Nevertheless the balance of argument in the end appears to be against the proposal, even with these modifications. The principle that a flat rate of insurance contribution should lead to a flat rate of benefit has a strong popular appeal and is much easier to defend than any departure from it. Some at least of those who advocated variation of benefit according to individual rents objected to any application of “insurance principle” to social security;   they favoured the financing of social security not by insurance contributions, but by graduated income-tax. If insurance contributions are retained, as is proposed in this Report, it seems best also to keep a rate of benefit without adjustment for individual rents. This decision is supported by the hope that the launching of the Plan for Social Security will coincide with a determined and successful effort to deal with urban congestion and shortage of housing. If and so far as this hope is realised, inequalities of rent bearing no relation to the accom­modation obtained will disappear ; a high rent will then represent a free choice by the householder and it will become indefensible to favour that form of expenditure over other forms of expenditure in fixing scales of benefit.

  5. But while the difficulties of principle and of practice, in the way of adjusting benefits to individual rents are probably sufficient to justify rejection of this proposal, the difficulty of meeting the differing requirements of house­ holds by a uniform rate of benefit in all parts of the country remains. The difficulty presents itself in regard to benefits and contributions alike. A rate of benefit adequate for needs, where rents are high, will be more than adequate for regions of low rent, and will involve contributions that may appear excessive in relation to the wages earned.   The preceding discussion of rents raises the question whether a rate of benefit and of contribution uniform in every part of Britain and in every occupation is right. On an average, the rent of an industrial household in London in 1937-38 was about 6/- a week above the average of industrial households in the rest of Britain, and the latter in turn was about 6/- a week above the rent paid by agricultural households. It would be inequitable to pay different rates of benefit in differing regions while charging a uniform contribution; some people outside London pay higher rents than some people in London. But there would be nothing inequitable in making regional or occupational differences in the rates of benefit and of contribution alike. To increase unemployment and disability benefits in London by, say, 3/6 a week for a single person, with 2/6 a week more for wife or dependant, making 6/- for man and wife, would mean adding about 6d. to the joint contribution of employer and employee in the case of a man and 4d. in the case of a woman.   To provide, either for agricultural households as such or for rural regions, benefits lower by the same amounts than the general standard for industrial households outside London, would make possible  a corresponding reduction of joint contributions any regional or occupational differentiation of benefits and contributions would detract from the simplicity of the social insurance scheme and would raise some difficulties in administration. But so long as differentiation applies to contributions as well as to benefit it is equitable, and so long as it is confined to working age benefits, the administrative difficulty, which is chiefly that of determining the rate to be paid to a man who has moved from one region or occupation to another region or occupation with different rates, cannot be regarded as insuperable or, indeed, as extremely serious. For pensions, it is right and necessary to maintain a uniform basic rate on retirement at the minimum pension age with a uniform contribution, and this is proposed in the following section. For unemployment and disability differentiation of insurance contribution and benefit may well appear preferable to a situation in which benefit in London had to be supplemented in a large proportion of cases in order to cover rent, and in which agricultural labourers and their employers found themselves paying unwillingly the high contributions which are needed for benefit to suit urban conditions.

215. The practical conclusions which emerge from this discussion are as follows :—

(1)  For the purpose of determining standard rates of unemployment and disability benefit household rent at pre-war levels is taken as 10/- a week for a household and 6/6 a week for a solitary individual.

(2)  The practicability and desirability of differentiation of both benefits and contributions, regionally or occupationally, in the way suggested in the last paragraph should be examined further in consultation with the persons and bodies affected.    If some such body as the Social Insurance Statutory Committee proposed in Change 22 of Part II is established, this question might conveniently be referred statutorily to that Committee, as the question of applying unemployment insurance to agriculture was referred to the Unemployment Insurance Statutory Committee under the Unemployment Act of 1934.

(3) The proposal to adjust benefit according to the rent actually paid by individuals should, provisionally, be rejected.

In accord with the first of these conclusions and with the estimates of other needs made below, a scale of uniform benefits and of the contributions needed to provide them has been prepared. If the further examination suggested in the second conclusion leads to regional or occupational differentiation of working age benefits and of the corresponding contributions, this differentiation can be introduced without affecting in any important way either the structure of the social insurance scheme or its financial implications. If, contrary to the provisional view taken here, it should be decided that adjustment for individual rents is desirable and practicable, this change also could be made without affecting the main structure.

216. These practical conclusions are suggested to make the best of a difficult situation, not as a solution of the problem of finding a subsistence basis for rates of benefit. The extreme variation of rents, between regions and in the same region, for similar accommodation is evidence of failure to distribute industry and population and of failure to provide housing according to needs.  No scale of social insurance benefits free from objection can be framed while the failure continues.  In this, as in other respects, the framing of a completely satisfactory plan of social security depends on a solution of other social problems.

Persons of working age

217.  In considering the minimum income needed by persons of working age for subsistence during interruption of earnings, it is sufficient to take into account food, clothing, fuel, light and household sundries, and rent though some margin must be allowed for inefficiency in spending. Persons not working will not have to spend money in travelling to work, and it is assumed that there will be provision for excusing insurance contributions, both compulsory and voluntary, during interruption of earning. In so far as these contributions are not excused, they must be covered by the margin. So also must be covered by the margin any diversion of expenditure from things which are necessary to things which, though not necessary, may appear prefer­able to the individual.

Coal, Gas and Electroicity

218. Food.  For food it is possible to make estimates on the basis of dietaries.  At 1938 prices, it would have been possible to provide an adequate dietary, either on the scale laid down in the 1936 and 1938 Reports of the Technical Commission on Nutrition by the League of Nations or on the scale laid down in 1933 by the Committee on Nutrition of the British Medical Association, for a man and woman together at a cost of about 13/- a week. Reference to these dietaries does not imply a view that they are themselves incapable of  improvement  or would be accepted today as final by all authorities.  The science of nutrition, like other sciences, progresses and shows how health may be improved by different ways of feeding. But a better dietary is not necessarily a more expensive dietary. Although the constituents of the League of Nations and the British Medical Association diets differ, there is no marked difference in their total cost.  The 13/- may reasonably be divided as 7/- for a single man and 6/- for a single woman. These figures exclude special needs, such as those of invalids and of expectant and nursing mothers. For the latter additional nourishment is necessary and the question may arise of providing it as part of a health service.

219. Clothing.  In respect of clothing it is not possible to determine the necessary expenditure with anything like the same accuracy as with food. The expenditures of industrial households in 1937-38 may be estimated from Ministry of Labour Family Budgets as 2/4 ¼ a week for a man and 2/6 a week for a woman or for a man and woman together 4/10 ½ ;  for agricultural households the figures are 1/8 for a man and 1/7 ½ for a woman, or 3/3 ½ together.    These expenditures are above the subsistence requirement, since they relate to households which are living above the minimum.    Moreover, clothing is an item of expenditure which can for a time be postponed.    In none of the Social Surveys undertaken in various towns before the war was the weekly cost of clothing for men and women together put as high as 3/-, the actual figures assumed ranged from 1/1 ½ to 2/11, and most of the estimates were below 2/-.   It is reasonable to put the allowance for clothing in unemploy­ment or disability benefit as 1/6 each for a man and a woman, or 3/- together. This is very little below the actual expenditure of agricultural households in 1937-38.

220. Fuel, Light and Household sundries. The average weekly expenditure of industrial households of two persons in 1937-38 on coal, gas and electricity as given in Table VIII varied from 4/1 in the Northern Division to 5/4 in the North-East, averaging 4/10 for the whole of Britain. For households of one person the divisional averages varied from 6d. in Wales (very few households) to 3/1 in the North-West, the average for Britain as a whole being about 2/8. Subsistence expenditure can clearly be put below these figures, which relate to households living on an average well above the minimum. In a sample tabulation of households in the lower quartiles for fuel consumption, taken as representative of the poorer households, the average expenditure (as given in the last column of Table VIII) was 3/7 for a two person family, as compared with 4/10 in all industrial households. There is also some possibility of reducing or postponing expenditure on fuel and light, though not as much as in the case of clothing.  On the other hand, some provision must be made for household necessaries other than fuel and light. It is suggested that 4/- for a man and woman together and 2/6 for a man or a woman separately to cover these other necessaries, as well as fuel and light should be adequate.

221. Margin. The  foregoing calculations, particularly that for food, assume complete efficiency in expenditure, i.e. that the unemployed or disabled person buys exactly the right food and cooks and uses it without waste. This assumption is clearly not likely to be realised.  Some margin must be allowed for inefficiency in purchasing, and also for the certainty that people in receipt of the minimum income required for subsistence will in fact spend some of it on things not absolutely necessary. It is suggested that a margin of 2/- a week for a man and a woman together and 1/6 a week for a man or a woman separately should be allowed.

  1. From these considerations there emerges as a basis for fixing the rate of subsistence benefit at 1938 prices the following table :—

TABLE IX Requirements for adults of working age at 1938 prices

Table 9
Table 9

223.   Strictly the figures for clothing and one or two minor items relate only  to short periods of unemployment and   disability, during which expenditure on renewals can be postponed; more will be needed in prolonged interruption of earnings. On the other hand, there should be room for re­adjustment in such matters as rent or of retrenchment in the margin.  On the whole, it seems fair to balance these considerations against one another and make no change in the rate of benefit as between short and long interruption of earnings during working years.

Retired Persons

224.   The subsistence needs of retired persons are in some ways less and in other respects greater than the needs of persons of working age during interruption of earnings. The differences in respect of the main heads of expenditure are as follows:—

Food— The food requirements of old people are placed by all authorities at substantially less than those of persons of working age. In calories they need about 75 per cent, of what is needed by working age adults Some writers in the past have reduced the money requirement for the food of old persons in something like the same proportion. But there is no adequate reason for putting the needs of old persons for constructive and protective foods more than slightly below those of working age adults, while on the other hand, the food of old people will be more expensive because of their tailing mastication and digestion. It is suggested that the food requirements of retired persons in place of being put at 75 per cent, of those of a working age adult should be put at about 85 per cent., or for a man 6/- a week and for a woman 5/6 a week.

Clothing:- The requirements of old People for clothing are not more than two-thirds of those of adults of working age. On the other hand, in contrast to unemployment and disability benefit for assumed temporary interruption of earnings, the pension must allow for renewals of clothing. Mr. Rowntree put the long period requirement for clothing at 4/- for a man and woman of working age, as compared with 3/- during temporary interruption of earnings, making the long period amount for a single person 2/-. It is suggested that the allowance for a pension for a retired person should be put at two-thirds of this, that is to say, 1/4 per week.

Fuel, Lighting and Household Sundries.—The requirements of retired persons are higher than those of persons of working age. It is suggested that they should be put at 3/- for one person in place of 2/6, and 5/- for two persons.

Margin.—This should be left for retired persons as for others at 2/- for two and 1 /6 for one.

Rent.—Retired persons should be able to adjust their rents. It is suggested that 6/- for a single person and 8/6 for a couple should be sufficient.

  1. The requirements for retired persons at 1938 prices become accordingly:—


Requirements For Retired Persons At 1938 Prices

                                                            Man and Wife                Man     Woman

  • Food                                                     11/6                     6/-         5/6
  • Clothing                                                2/8                    1/4         1/4
  • Fuel, Light and Sundries                  5/-                       3/-         3/-
  • Margin                                                   2/-                      1/6          1/6
  • Rent                                                       8/6                      6/-         6/-
  • Total                                                     29/8                    17/10     17/4

This suggests that at pre-war prices the subsistence pension on retirement for a man and wife would be about 2/4 a week below the rate of benefit for unemployment or disability.

Young Persons And Boys And Girls Of Working Age

226.    The food allowance for boys and girls of working age and of young persons up to 21 should be slightly more than that for adults, say 7/6 a week for males and 6/6 for females. The clothing allowance should be the same as for adults, say 1/6 with the same margin of 1/6. These items accordingly add up to 10/6 for a male and 9/6 for a female.  Rent can be presumed to be covered for boys and girls by their parents, and the actual contribution which such earners will make to other common costs will vary with circumstances. Adding 1/6 for this, to the figures of 10/6 or 9/6 for food, clothing and margin, the benefit needed for boys might be put at 12/- with girls 1 /- less.  For young persons of 18 to 20, who will in some cases already have separate households, a larger addition, including something for rent is needed. The determination of benefit rates in these cases, must to some extent be arbitrary. As a practical conclusion it is suggested that the benefit rate at 1938 prices might be 12/- a week for boys and girls alike, for young men 16/- and for young women either 15/- or the same as the rate for young men, according to the decision taken (in para. 230) as to equalising or differentiating benefits for men and women.

Dependent Children

227.   For children, 5/- a week has commonly been assumed in the past as a suitable allowance. It seems doubtful, however, whether a figure as low as this can be justified if it has to cover subsistence needs. On the League of Nations dietary the weekly cost of food for a child at 1938 prices would be as follows :—

  • 0-5 years                4/6
  • 5-10                        6/-
  • 10-14                      7/-
  • 14-15                       7/6

The actual expenditure on clothing in 1937-38 for each child under 14 may be estimated from the Ministry of Labour Family Budgets at 1/4¼  a week in industrial households and 9 ½ d. a week in rural households. In the case of children the allowance must cover the cost of clothing over a long period; that is to say, it cannot be reduced on the assumption that such expenditure would be postponed as in the case of an adult during unemployment or dis­ability. The minimum weekly cost of clothing for a child can hardly be put at less than an average of l0d. a week. If it is assumed that clothing costs vary with age in much the same way as food costs, the weekly requirement for each age group may be put as follows :—

  • 0-5 years           7d.
  • 5-10                    10d.
  • 10-14                 l/-
  • 14-15                 1/3

For fuel and light Table VIII, based on the Ministry of Labour Family Budgets, shows average weekly cost rising fairly steadily from 4.86 shillings for two person households to 6.33 shillings for eight person households, i.e. by just over 1/6 for 6 persons. This suggests an allowance of 3d. per week per child to cover fuel, light and household sundries.

228.   Even without any margin for inefficiency in purchasing, this calculation yields at 1938 prices the following  amounts as required for children’s allowances to cover subsistence needs without rent:—

  • 0-5 years            5/4
  • 5-10                     7/1
  • 10-14                   8/3
  • 14-15                   9/-

This makes the average subsistence allowance at 1938 prices for each child about 5/11 for food, l0d. for clothing and 3d. for fuel and light or 7/- a week altogether instead of the 5/- which it has been common to assume in discussing children’s allowances in the past.


229.The foregoing discussion suggests that at 1938 prices and on the assumption of a uniform allowance of 10/- a week for rent of a household the subsistence benefit should be 19/- a week for men of working age and 18/- a week for women of working age, and 32/- for a man and wife together. The corresponding subsistence pensions for retired persons should be 17/6 single and 29/6 for a couple, assuming 6/- for rent in the former case and 8/6 for rent in the latter case.   The subsistence needs of children should be put materially higher than has been customary hitherto in discussions of children’s allowances.

  1. On these estimates of subsistence needs, two practical decisions have to be taken.   The first is whether differentiation by sex should be maintained. The strongest reason for giving women, as such, lower rates of benefit than men would be a desire to avoid imposing on women rates of contribution excessive for their wages. Women’s contributions will in any case be substantially below those of men, which have to carry provision for housewives. To make any further substantial difference between the con­tributions of the sexes women’s benefit rates would need to differ from men’s rates by much more than the I/- a week during unemployment and disability than can be justified by examination of subsistence needs. On balance it seems not worthwhile to maintain any difference between the rates of single men and of single women. The second question is whether the pensions on retirement should, in accordance with subsistence needs, be put below the working age benefits. The final answer to this depends on the discussion of the problem of old age in the following section. For the reasons given in para. 251 it is suggested that the insurance scheme should provide retirement pensions at the level of working age benefits and above subsistence needs.

  2. The rates named above are based on 1938 prices. At the date of this Report (November, 1942) the cost of living is about 30 per cent. above the level of 1938. The level at which prices will tend to settle after the war cannot be foretold;  it is unlikely that they will return in any short period to the pre-war level; it is reasonable to hope that they may be kept near the present level. As a basis of discussion, in framing the Social Security Budget, it has been assumed that the cost of living after the war will settle at about 25 per cent, above 1938. This yields in round figures a provisional post-war rate of benefit for a man and wife together of 40/- a week, and for a man or woman alone of 24/- a week, neglecting any difference of sex and rounding the single person’s rate up to the nearest 1/-. These provisional rates allow for a rise of 25 per cent, or a little more in the cost of all necessaries, including rent and including also the margin suggested in para. 221.  If rents can be kept at their pre-war level, the provisional rates are sufficient for a rise of about 33 per cent, on the other necessaries, including the margin. If, in spite of a rise of materially more than 25 per cent, over all necessaries, including rent, the provisional rate of unemployment or disability benefit for a man and wife together is kept at 40/- a week, this means cutting into the margin; if the margin is left out altogether, 40/- a week covers a rise of 33 per cent, in the cost of living, including rent. If retirement pensions are put at the same provisional rate of 40/~ joint and 24 /- single there is a further margin, because the actual subsistence needs are lower. For young persons of both sexes 18-20 a provisional rate of 20/- a week is suggested and for boys and girls one of 15/- a week.

  3. For children’s allowances it seems reasonable to take a provisional post-war rate of 9/- a week, covering an increase of nearly 30 per cent, in cost of subsistence above the 7/- required at 1938 prices. This does not include anything for rent or allow any margin. On the other hand, substantial pro­vision in kind is already being made for children through school meals and supply of free or cheap milk. Allowing for this, in framing the Social Security Budget in Part IV the average cost of allowances in addition to existing provision has been put at 8/- per head per week.