Purpose: The aim of the paper is to show the relationship between food as a shared good (or public within the household) in the economic sense, and food as a shared meal in the sociological sense. Design/methodology/approach (mandatory): Quantitative data derived from a household budget survey in Cyprus are used to set up questions to which answers are suggested using the qualitative approach of in-depth interviews. Findings: The main finding is that the relatively high expenditure by elderly couples on food for home consumption may be explained by frequent inter-household, intra-extended family meals in Cyprus. Research limitations/implications (if applicable): The paper provides evidence that household expenditure on food may not be directly indicative of household consumption of food. Researchers interested in household consumption of food should therefore be aware of the differences between household and extended family and, where extended family continues to be significant, they should be wary of using data from household budget surveys to analyse food consumption. One limitation is that the results are derived from in-depth interviews with a purposive sample of nine households. It may be appropriate to replicate the study, either in Cyprus or in similar societies where extended family remains significant, at a larger scale. Practical implications (if applicable): The evidence that household expenditure may not be indicative of household consumption suggests that questions on consumption should be included in household budget surveys. Originality/value: This paper draws together, for the first time, economic ideas on expenditure on food derived from the quantitative research of Ernst Engel on one hand and implications of the theories of Georg Simmel on the sociology of the meal on the other. The paper shows that some issues arising from quantitative analysis of household budget surveys cannot be explained using data from that source; this is particularly so where consumption of food is inter-household.