Food Health and Income: Appendix 6 Comparison of estimated food consumption in income groups with figures obtained from family budgets

If all the statistics were perfect, the average consumption of each food found by calculating an average of the consumption per head in the various income groups, weighted by the proportions of the population within those groups, should agree with the national average consumption figure obtained by dividing total supply by total population. Such precise agreement cannot, of course, be expected, partly because of the margin of error inherent in the estimates themselves, partly because the two sets of figures—the aggregate national supply, and the consumption in households as derived from family budgets—relate to somewhat different totals.

At all stages it has been necessary to make estimates from insufficient or barely sufficient data, and the figures used have throughout been approximations, sometimes reasonably close, sometimes subject to a fairly wide margin of error. The figures of total supplies, which are the most satisfactory of the data, are themselves merely estimates which, though believed to be reasonably accurate, cannot be accepted as precise beyond dispute. An indication of the error which may be involved is given by the impossibility of reconciling the official estimates of milk consumed in liquid form with corresponding figures derived from the published statements of the Milk Marketing Boards.

What applies to the estimates of total quantities applies with even more force to the two separate factors which should, theoretically, also give us figures of total supplies. As we have seen, the number of family budgets used was something less than 1,200. They included an undue proportion of families in the industrial north, of families with small incomes and relatively large numbers of dependants. Moreover they were not distributed seasonally throughout the year, but tended to be concentrated in the spring and early summer months. Family budgets for “black-coated” workers were few; those for the middle classes were poorly represented and for the rich completely lacking. There is good reason to believe that the average proportion of the income spent in providing food is fairly accurately shown by the family budget figures for the various income groups—at any rate by those in the lower income groups—but the actual quantities and values of the different items in the family dietary would no doubt be somewhat altered if a collection of family budgets thoroughly representative of the whole country were available.

The estimates of the proportions of the population falling within each income group on a per head basis are similarly the result of work done with inadequate material, and a margin of error of as much as 10 or 15 per cent, in any one group would not be surprising.

Apart from these possibilities of error, however, there are a number of points of difference between the estimated national supply and the quantities consumed in family households. In the first place, the national supply figures are “gross” ; they cover the total quantity of each food available at the first point of sale—the farm, factory or port. Between that point and the purchase by the housewife there is a considerable loss of weight. The total supply of fish and meat includes a large proportion which is not passed on to the consumer, but is wasted or is sold for industrial purposes. Other commodities also are subject to a similar, although smaller, degree of wastage, e.g., milk, eggs, fruit and vegetables. Secondly, not all the food consumed in the country is included in the “family food bill” ; some part of it is eaten in institutions, residential hotels, and other “non-family” establishments, while a considerable proportion is served in restaurants, eating-houses and canteens. It has been assumed that the average consumption per head in institutions and hotels is the same as the average for the whole country.

Consumption of food in restaurants, etc., is additional to the food provided by the housewife, and yet must be included in the average per head consumption of the families concerned. It is assumed that food bought and eaten away from the home constitutes a very small addition to the food consumption of the poorest groups, is un­important for most foods even in the fourth group, but increases rapidly in the fifth and sixth groups. Mr. Feavearyear, in his estimates of national expenditure, assumed that 10 per cent, of the nation’s food is sold through hotels, restaurants and eating-houses. It is probable that this is an over-estimate; but even if the proportion be taken at only 5 per cent., and this be divided equally between the fifth and sixth groups, it means an addition of 12 ½ per cent, and 25 per cent, respectively to the per head consumption in these groups. For some foods of course, the increase will be much larger than for others.

Since it is the upper groups (V and VI) which are mainly affected by the “meals out” problem, and since in any case the family budget data for these groups were scanty, it is the average consumption in the upper groups that has needed most adjustment to secure agreement between the national average and the weighted average of the groups. The proportion of the income spent on food in the lower groups being fairly well established, no material alteration in the average consumption of any one food can be made in these groups without a corresponding alteration in the opposite direction in some other food: otherwise the food expenditure of the groups would be altered.

One further preliminary point should be made clear. Since the income groups are on a per head basis, any one group will contain a heterogeneous collection of occupations, wages, earners and non-earners. Even in the wealthier groups there will be a small proportion of working-class families—skilled workers with only one dependant, or families with several earners each in receipt of good wages. Of no group can it be said that its needs or tastes are noticeably different from those of the groups immediately above and below it. This would be equally true if the income groups were to be made much narrower—with one or two shilling ranges instead of the wider ranges selected. Hence it follows that any curve showing variation in average consumption per head at different income levels should be a smooth curve.

The procedure adopted was as follows:—The family budget figures were first entered on a diagram and a smooth curve drawn as closely as possible to the points plotted. The curve was continued in groups V and VI, its course being determined by the trend of consumption as shown by the middle-class budgets. The group averages were then read off from the diagrams, and the weighted average of all groups compared with the national averages. Reasonable approximation was regarded as satisfactory, but if a serious disparity appeared, which could not be explained by a necessary difference between the national and the weighted budget averages, the disparity was removed either by slight amendments throughout the groups, or by further adjustment in the two upper groups; for the lower groups could not be substantially altered without affecting the proportion of income spent on food. The result is a compromise, but one which is believed to be not far from the truth. Such errors as are contained in the picture are likely to be mainly in groups V and VI, and would merely involve transfers between these two groups.

In the following tables are shown the actual budget data and the figures finally adopted.

Appendix 6 Table 1
Group I Group II Group III Group IV Group V Group VI Weighted average of Groups
Proportion of the population 10% 20% 20% 20% 20% 10%
Number of budgets 411 152 233 156 136 64
Beef and veal ozs. 9.5 11.5 11.7 11.3 10.2 9.5 10.8
Mutton and lamb ozs. 2.1 3.1 4.3 6.2 6.8 9.7 5.3
Bacon and ham ozs. 2.6 4.1 4.6 5.7 5.7 6.6 4.9
Sausage, corned beef and pork ozs. 2.8 2.9 4.2 5.4 3.6 3.5 3.8
Total meat (above) ozs. 17.0 21.6 24.8 28.6 26.3 29.3 24.8
Bread and flour ozs. 64.5 62.0 63.3 64.7 54.6 47.7 60.1
Milk fresh pints 1.1 2.1 2.6 2.9 4.5 5.4 3.1
Milk condensed liqid milk equivalent pints 0.6 0.4 0.4 0.3 0.2 0.1 0.3
Eggs no. 1.9 2.8 3.7 4.8 4.7 5.2 3.9
Butter ozs. 2.7 5.7 7.4 8.8 8.9 9.7 7.4
Cheese ozs. 1.5 2.1 2.8 3.2 2.9 2.5 2.6
Margarine ozs. 4.9 2.9 2.2 1.9 2.5 1.4 2.5
Tea ozs. 2.2 2.5 2.5 2.8 2.5 1.4 2.5
Potatoes (not including chips) ozs. 51.2 50.8 55.5 57.4 42.8 39.4 50.4
Lard, suet and dripping ozs. 2.5 3.4 4.5 4.7 3.5lard only 3.2lard only 3.8
Fish (excluding fried or tinned) ozs. 2.4 2.6 3.9 5.4 5.9 8.1 4.6
Sugar purchased as such ozs. 13.5 15.9 18.1 20.1 19.0 18.1 17.8
Jams jellies and syrups 4.3 5.5 5.7 5.8 6.5 5.6 5.7

Group V has been calculated as a straight average of one working class group with income of 30s. to 40s. per head and two middle class groups with family incomes of £200-£00 and £300-£400 (30s. and 40s. per head per week respectively.

Group VI has been calculated as a straight average of four middle class groups with family incomes of £400-£500, £500-£600, £600-£700, £700-£800 per annum.

Appendix 6 Table 2
Group I Group II Group III Group IV Group V Group VI Weighted average of Groups National Average
Proportion of the population 10% 20% 20% 20% 20% 10%
Average food expenditure per week 4s. 6s. 8s. 10s. 12s. 14s. 9s.
Beef and veal ozs. 10.5 14.5 17.2 18.9 19.5 18.9 17.0 20.0
Mutton and lamb ozs. 3.1 5.6 7.2 9.4 11.6 13.9 8.4 9.0
Bacon and ham ozs. 4.3 6.3 6.8 7.3 7.8 9.4 7.0 7.8
Sausage, corned beef and pork ozs. 5.2 5.2 5.9 5.9 5.9 7.2 5.8 7.2
Total meat ozs. 23.1 31.6 37.1 41.5 44.8 49.4 38.2 44.3
Bread and flour inc biscuits and cakes in terms of flour ozs. 66.0 68.0 68.0 67.0 65.0 60.0 66.0 61.0
Milk fresh pints 1.1 2.1 2.6 3.1 4.2 5.5 3.1 2.8
Milk condensed liqid milk equivalent pints 0.7 0.6 0.55 0.5 0.4 0.3 0.5 0.5
Eggs no. 1.5 2.1 2.6 3.1 4.2 5.5 3.1 2.8
Butter ozs. 3.0 6.5 7.5 8.5 9.5 11.0 7.8 7.8
Cheese ozs. 1.8 2.5 3.1 3.6 3.6 2.6 3.0 3.2
Margarine ozs. 4.5 3.5 2.5 2.0 1.6 1.3 2.5 2.4
Tea ozs. 2.2 2.7 2.9 3.0 2.9 2.7 2.8 2.8
Potatoes (not including chips) ozs. 53.0 56.0 57.0 57.0 57.0 54.0 56.0 64.0
Lard, suet and dripping ozs. 2.7 3.6 4.2 4.4 4.3 3.5 3.9 2.7 (lard only)
Fish ozs. 2.7 5.5 8.2 10.4 12.2 13.5 8.9 13.2
Sugar purchased as such (includes industrial consumption 40%) ozs. 13.5 16.0 18.0 19.0 19.5 19.5 17.8 27.7
Jams jellies and syrups 4.3 5.3 5.2 5.4 5.8 5.5 5.2 including in fruit and sugar
Sugar consumed in other forms 6.5 7.5 8.5 9.5 10.5 11.5 9.0 included in sugar above
Fruit 14.0 21.7 25.8 27.9 30.5 39.3 26.5 35.1 includes fruit used industrially at 25%
Vegetables excluding potatoes 16.0 20.0 27.2 30.6 32.3 34.0 27.0 30.2 includes shop wastage at 10%

Figures for meat and fish include wastage in distribution. Figures for condensed milk include use in complex foodstuffs. Figure for potatoes includes allotment production. Group quantities for fruit and vegetables have been estimated from expenditure after allowing for quality variations but the figures are subject to a wide margin of error.

Most of the differences between the figures in Tables 1 and 2 are due to the smoothing of the curves described above. In a few cases, however, more important alterations have been made. These are described below.

(i) Meat.—The national supply figures for meat include the whole of the dressed carcase weight. While the butcher manages to pass on to the consumer considerable quantities of bone and surplus fat, there is a proportion varying between 5 per cent, for mutton and 15 or 20 per cent, for beef, which is not sold to the consumer. Moreover some edible fat (lard, dripping, suet) is included in the carcase weight, but may be bought separately by the consumer. But even allowing for these factors and for meals in restaurants, etc., the family budget averages were quite inadequate to account for the whole of the meat supply, and it was found necessary to raise consumption throughout the groups, the increases ranging from 6 ounces per head per week in group I to 20 ounces in group VI— the latter, of course, including the allowance for meals out.

(ii) Eggs.—The national supply of eggs was insufficient to provide the quantities shown in family budgets. As the latter were obtained mainly in the season when eggs are plentiful, a reduction throughout the groups was necessary.

(iii) Cheese, Sugar.—Here the converse occurred, and it was found necessary to raise all figures slightly.

(iv) Condensed Milk.—The budget figures were increased throughout by an allowance to represent the condensed milk used in confectionery.

(v) Fish.—Much the same considerations apply here as to meat.

Appendix 6 Table 3
Group I Group II Group III Group IV Group V Group VI Weighted average of Groups National Average
Proportion of the population 10% 20% 20% 20% 20% 10%
Average food expenditure per week 4s. 6s. 8s. 10s. 12s. 14s. 9s.
Beef and veal d. 4.7 7.1 9.8 12.0 14.5 15.4 10.7 10.4
Mutton and lamb d. 1.7 3.1 5.0 7.1 9.2 11.0 6.1 5.9
Bacon and ham d. 2.9 4.4 5.1 5.9 6.7 8.0 5.5 7.3
Other meat d. 4.1 5.1 6.0 6.8 8.0 10.0 6.6 5.5
Total meat d. 13.4 19.7 25.9 31.8 38.4 44.4 28.9 29.1
Bread and flour inc biscuits and cakes d. 9.0 11.0 12.4 13.8 15.3 17.5 13.2 8.1 (not inc cakes and biscuits)
Milk fresh d. 3.4 6.4 8.5 10.2 13.2 17.8 9.8 8.7
Milk condensed d. 1.4 1.2 1.1 1.0 0.8 0.6 1.0 1.0
Eggs d. 1.7 2.3 3.1 4.0 4.9 7.6 3.8 4.3
Butter d. 2.1 4.7 5.6 6.8 8.0 10.1 6.2 5.4
Cheese d. 1.0 1.4 1.7 2.0 2.4 2.1 1.8 2.3
Margarine d. 1.6 1.3 1.0 0.8 0.7 0.6 1.0 0.8
Tea d. 2.5 3.6 4.1 4.6 5.0 4.8 4.2 3.8
Potatoes d. 2.5 2.9 3.0 3.0 3.1 3.0 3.0 3.6 (inc allotment output)
Lard, suet and dripping d. 1.2 1.7 2.0 2.2 2.2 1.9 1.9 1.1 (lard only)
All Fish d. 1.0 2.4 4.1 5.8 7.6 9.3 5.0 5.1
Sugar purchased as such d. 1.9 2.4 2.7 2.8 3.1 3.2 2.7 4.8 (inc manufactured)
Jams jellies and syrups d. 1.3 1.5 1.7 1.9 2.2 2.4 1.8
Vegetables excluding potatoes d. 1.5 2.6 3.9 5.2 6.5 8.5 4.6 3.9
Fruit d. 2.4 4.6 6.6 9.5 13.0 20.0 9.0 11.7 (inc manufactured)
Miscellaneous d. 0.1 2.3 8.6 14.6 17.6 14.2 10.1 12.0
Total d. 48.0 72.0 96.0 120.0 144.0 168.0 108.0 105.7

Miscellaneous includes such items as coffee, cocoa, condiments, sauces, etc., and, for the groups, is the difference between the enumerated items and the total group food expenditure.

The budget figures show little consumption of fried fish, and it would seem probable that fried fish purchases may sometimes be regarded by the housewife as falling outside normal household expenditure.

All groups have been raised in the same proportion and sufficiently to give an average approximating to the national average, less an allowance for wastage.

Other differences are accounted for by the necessity of obtaining a smooth curve. The figures shown in the budget data for the poorest groups have seldom been reduced and have more often been increased. Apart from eggs, only margarine has been given a lower figure than that indicated by the budgets in group I, and this slight decrease has been balanced by an increase in butter. Only jams are reduced in group II; and only jams and lard in group III. On the other hand, in each of the groups there are several increases.

The adjustments, however, are not such as to alter materially the average composition of the diets of the different income groups, as calculated from the budgets and dietary surveys. In general the changes made to bring the results into harmony with the estimates of total food supplies have tended to raise the level of consumption throughout.

In Table 3 corresponding figures are given for expenditure per head per week in each income group. This provides a useful check on the figures of quantities in Table 2. The price per unit in each group can be roughly estimated. The figures in each group added together must then correspond with the total expenditure on food in the group and the weighted averages of expenditure on each food in all groups must add up to the average expenditure on all food for the whole population. The chief causes of discrepancy between the national and weighted averages, which is not serious, are referred to in the footnotes to Table 3.