Food Health and Income: Distribution of National Income and Food Expenditure at different Income Levels

The figures given in the preceding sections have dealt with average food consumption per head of the population, but clearly this is a case where a mere statistical average has little meaning. From the standpoints both of public health and of agricultural marketing, it is more important to know how many families fall below the average and what is the variation in consumption at different income levels.

The family budgets and dietary surveys were first arranged in groups according to the income per head of the family, the total family income from all sources being divided by the number of persons, irrespective of age and sex, supported by that income. The average food expenditure in each income group and the average amount and value of each food purchased per head per week were then ascertained. The results are given in full in Appendix VI, Tables I-III.

The next step in the enquiry was to estimate the proportions of the entire population falling into the several income groups.

There is no completely satisfactory means of classifying the population by income groups. In default of accurate information, which could only be obtained by a census of family incomes, it has been necessary to make indirect estimates based on a large number of calculations. The data and methods used in arriving at these estimates are described briefly in Appendix V.

Income per head depends on the size of the family and the combined incomes of all members of the family. It has been inferred from a random sample of the 1931 Census that roughly half the families in the country contain more than one earner. The earnings of other members of the family and supplementary incomes, such as unemployment benefit, pensions, public assistance and investments, have to be taken into account and the total related to the number of persons to be maintained.

Table 4 below shows the final approximate estimates of the distribution of the National Income, based on the computations described in Appendix V. The population has been classified into six groups, consisting of 10 per cent, at the top and the bottom, and four intermediate groups of 20 per cent.

Table 5
Group Income per head Estimated average expenditure on food Estimated population of group.
Numbers Percentage
I Up to 10s
II 10s. to 15s.


III 15s. to 20s.
IV 20s. to 30s.
V 30s. to 45s.
VI over 45s.
Average 30s

Here also the figure given for income per head is the income of the family divided by the number of persons supported. Thus, a man and wife with £2 10s. a week with no children or dependants would fall in group IV; with one child into group III; with two or three children into group II, and with four or more children into group I. The poorest 10 per cent, of the population consist in the main of families in which there is a disproportionate number of children or other dependants per earner. It is estimated that half the persons in group I are children under 14 and that between 20 and 25 per cent, of the children in the country are in the lowest income group.

Included in the table is a column showing average food expenditure per head per week. It is obvious, of course, that as income falls the percentage spent on food, which is a prime necessity of life, will increase. The average expenditure on food represents a proportion rising from below 20 per cent, in group VI to nearly 50 per cent, in groups I, II and III. Among the poorest some observers have found expenditure on food exceeding 70 per cent, of total income (10, 18), but food prices were then higher than they have been in the past few years. A study of family budgets in a number of countries shows that food accounts on the average for between 45 and 60 per cent, of the total expenditure of working-class families (2).

It is of interest to compare the results obtained here with estimates of income distribution in working-class families given in recent surveys of the economic condition of the people in particular areas. Thus the percentage of the population living below the poverty line computed from data given in the surveys of London (42), Merseyside (20) and Southampton (15), is estimated to be approximately as follows:—

  • London, 1929-30    8 %
  • Liverpool, 1929 13%
  • Merseyside, 1929-30 14%
  • Southampton, 1931    16%

Data regarding the dispersion of working-class family incomes in these surveys indicate that in Southampton the upper limit of income of the poorer half of the population was 15s. per head per week, in Merseyside 18s. 6d., and in London 21s. In the present enquiry, the corresponding figure based on income data for the whole population is 20s. The comparison suggests, therefore, that as regards the lower groups the classification of the population given in the above table is unlikely to involve serious error and can be accepted as a reasonable working hypothesis.