nevertheless , average net family income was appreciably higher in families with several children than in those with only one , many of which were incomplete families of younger parents with lower earnings , and of course with lower tax reliefs and no family allowances . the rise in net family incomes between 1954 and 1959 was greatest for childless couples , especially older couples ( probably because of the increase in retirement pensions ) , and somewhat greater in small than in large families . expenditure and consumption . 69 . table 24 gives indices of domestic food expenditure per head and quantities purchased by older and younger couples and families with different numbers of children , with 1954 as the base year . the quantity index was calculated by dividing the expenditure index by a price index of the Fisher ideal type , constructed for each group separately . the quantity index is thus confined to food purchases and takes no account of changes in free supplies . compared with 1958 , the expenditure index for 1959 showed increases of 4 to 6 per cent for couples without children and those with one child , and much smaller changes for couples with several children . the quantity index , which has risen only slowly since 1956 , was almost unchanged in 1959 for couples with two or more children , but rose by 3 per cent in the older two-adult households and by 1-2 per cent for younger childless couples and couples with one child . 70 . table 27 gives the total domestic food expenditure and value of consumption per person per week in 1959 in households of different composition . percentage standard errors of these estimates are given in table 10 of appendix A . all types of household spent more than in the previous year except families with three children , whose expenditure had risen sharply in 1958 . the increases ranged from 2 s 7 d per person per week in the residual group of households with adolescents but no children and 1 s 8 d in older two-adult households to 4 d in the families with two and with four or more children . the value of free food was greatest ( 1 s 2 d to 1 s 4 d per person per week ) in the five types of household containing no children , and varied between 9 d and 11 d in households with children , except in the largest families for which the average was only 5 d , as in 1958 . in families with three children , the slight fall in expenditure was made good by an increase in the value of free food . the value of consumption per person per week in 1959 ranged from 41 s 7 d for younger childless couples to 19 s 5 d in families with four or more children ; in 1958 the range was from 40 s 3 d to 19 s 1 d . 71 . table 27 includes an index comparing the price of energy for the various types of household with that for all households in the sample . as in 1958 , younger couples paid some 12 1/2 per cent more per calorie than the national average , and families with four or more children 19 per cent less . the only substantial change was in families with three children , for whom the index declined from 91 to 88 . table 27 also shows the corresponding values of an index which compares the prices paid by different types of household for the commodities constituting the average household diet in 1959 . for all foods the range was from 3.6 per cent above the national average in younger two-adult households to 4.8 per cent below in families with four or more children , compared with +3.6 to -5.4 per cent in 1958 and +2.8 to -2.6 per cent in 1957 . as with the price of energy index , the only noteworthy change was for households with three children , in which the index fell by 1.8 to 97.5 per cent of the average for all households in the sample . the price ranges for milk , cheese , sugar , bread and flour were very narrow . for most other foods younger childless couples paid the highest average prices and large families the lowest , the price gradients being steepest for carcase meat ( +7 to -7 per cent ) , other fish ( +8 to -11 ) , other vegetables ( +9 to -9 ) and beverages other than tea ( +12 to -16 ) . 72 . details of expenditure and consumption per head are given in tables 28 and 29 . most groups obtained slightly less liquid milk than in 1958 , the greatest decrease ( from 5.24 to 5.08 pints per head per week ) occurring in younger two-adult households . table 25 summarizes the changes in consumption of liquid milk ( including welfare and school milk ) between 1954 and 1959 by this group and by classified households containing children or adolescents . consumption by younger childless couples declined throughout this period , but that of the smaller families was maintained except for the slight fall in 1959 . in the largest families , particularly those containing four or more children , there was a tendency for consumption to increase between 1955 and 1957 , and thereafter to decline . graduated scales of family allowances were introduced in October 1956 , and the welfare milk subsidy was reduced in April 1957 . despite appreciably higher average prices for natural cheese in 1959 than in the previous year , consumption fell only slightly in most groups ; the decrease was greatest ( from 3.68 oz to 3.20 oz per head per week ) for younger childless couples , who transferred much of their demand to cheaper varieties . 73 . all groups , except families containing three or more children , increased their expenditure on meat , but total consumption was much the same as in 1958 , although there was some replacement of beef ( which continued to be in short supply ) by mutton and lamb . all groups spent more on fish , and most increased their consumption , particularly of canned fish . eggs were cheaper than in the previous year and consumption increased in nearly all groups despite fewer free supplies . 74 . all types of household substituted margarine for butter in 1959 because of higher butter prices , but all except the largest families continued to buy more butter than margarine . total consumption of butter and margarine declined only in households containing children . the displacement of margarine by butter in 1958 , when butter was exceptionally cheap , appears to have had some lasting effect ; the average price of butter in 1959 was higher than in 1957 , yet butter purchases in 1959 , although smaller than in 1958 , were greater than in 1957 except in families with four or more children or with adolescents but no children . 75 . the smaller families and the residual groups of households containing children reduced their consumption of sugar and of preserves , but in all other types of household a decline in purchases of the one was accompanied by an increase in consumption of the other . 76 . most groups spent slightly less on potatoes than in the previous year , but consumption was maintained except in families with more than one child and in the unclassified households with children or adolescents . all groups except families with four or more children or with adolescents but no children consumed more fresh green vegetables , especially in the first half of the year , although most reduced their consumption of other vegetables . purchases of quick-frozen peas and beans generally increased , but extremely wide group differences persisted , the averages ranging from 0.1 oz per head per week in families with four or more children to 0.9 oz per head in younger two-adult households ; average consumption by older couples and other wholly-adult households was 0.5 oz per head per week . the two latter groups consumed much smaller quantities per head of canned vegetables and canned and bottled tomatoes than any other group , and much smaller quantities of canned and bottled fruit than younger childless couples ; in households containing children , consumption of canned fruit fell off sharply with increasing family size , but there was no regular gradation in purchases of canned vegetables and canned tomatoes . all types of household benefited from the improved supplies and lower prices of fresh fruit compared with 1958 , but the increase in consumption was least in families containing children . 77 . total bread consumption was virtually unchanged , although most types of household bought less white bread and more rolls and speciality breads than in 1958 . most groups increased their purchases of puddings , cakes and biscuits , but obtained less flour . 78 . regression estimates of the expenditure on different commodities attributable to the adult couple and each additional child in a selected group of households consisting of childless couples ( both under 55 ) and couples with different numbers of children were given for 1952-56 in table 39 of the annual report for 1956 . the younger childless couples are broadly comparable in age and family income with the family households , so that differences in food expenditure may be associated with the presence of children . the analysis has been repeated for 1957 , 1958 and 1959 , but the results will not be given in extenso . household food expenditure in 1959 averaged 80 s 9 d for younger couples and 92 s 10 d , 102 s 0 d , 111 s 9 d , and 126 s 3 d for two-adult households containing respectively one , two , three and four or more ( average 4.64 ) children under 15 . from a straight regression line fitted to these averages , the basic element in household food expenditure associated with the adult couple is estimated at 81 s 11 d and the average increment for each additional child as 9 s 11 d . table 26 gives similar regression estimates for previous years . the effects of price rises are roughly eliminated by expressing the average expenditure associated with a child as a percentage of that associated with an adult couple . the relative expenditure per child declined from 1952 to 1956 , but rose in 1957 when the subsidy on welfare and national dried milk was reduced ; since 1957 it has again declined . most of the average expenditure associated with a child was on cereal foods , potatoes and milk ; for fresh green vegetables , fruit , cheese , fish and carcase meat , the incremental expenditure was slight . energy value and nutrient content . 79 . table 30 shows the energy value and nutrient content of the diets of households of different composition . the averages showed little change compared with those for the previous year , except for generally increased intakes of vitamins C and D . since physiological requirements vary widely with age , sex and level of activity , comparisons between families of different composition are only apposite when considered in relation to needs . 80 . estimates of the adequacy of the diets , assessed by comparison with allowances based on the recommendations of the British medical association , are also shown in table 30 . in comparison with the previous year , changes were small except for higher estimates for vitamin C . in families with four or more children the levels of adequacy for all nutrients other than vitamin C decreased slightly . for this fairly small group , comparisons between different years can not be made so precisely as in groups with a defined number of children . in 1959 the households in this group contained slightly more children ( average 4.64 ) than in the previous year ( average 4.53 ) . their total food expenditure increased less than that in other groups , and they purchased more of certain foods such as fish , poultry , eggs , canned vegetables , fresh fruit , chocolate biscuits and breakfast cereals which , in general , are more expensive sources of nutrients than those foods of which they purchased less , namely dried milk , potatoes , carcase meat , sugar , bread , flour and oatmeal and oat products . 81 . in all these estimates of adequacy , the conventional allowance of 10 per cent has been made for wastage of edible food . the limitations of the use of arbitrary wastage factors , regardless of family size or circumstances , were pointed out , and the effect of the use of graduated wastage factors examined in the annual report for 1956 . as in previous years , the percentages in table 30 for all nutrients decreased with increasing family size . the lowest estimates were for protein and calcium in families with four or more children ( 82 and 81 per cent respectively ) . during the ten years from 1950 to 1959 there were downward trends in the percentages for protein and calcium for all types of family and for all households , the steepest ( from 94 to 82 per cent for protein and , from 92 to 81 per cent for calcium ) occurring for the families with four or more children ; another considerable fall was from 91 to 83 per cent for protein in families with adolescents and children . 