Accredited official statistics

Family Food 2020/21: About Family Food

Updated 25 April 2023

1. About Family Food

Over the 80 years of the survey, we estimate around half a million households have participated in Family Food and its predecessors. Our thanks go to all those respondents, without whose cooperation this invaluable historic data resource would not be possible, and especially to those who freely donated their time in 2019/20.

2. Survey organisation

Family Food 2020/21 is a report on the 2020/21 Family Food Module of the Living Costs and Food Survey (LCFS). This report provides statistics on food purchases by type of food. Datasets and methodology notes are provided on the website with some statistics back to the 1940s. The survey covers about 5,000 households across the United Kingdom each year. Food purchases are reported at a detailed level and demographic patterns and trends are identified.

A total of 22,516 addresses were selected in 2020/21 for the LCFS in Great Britain, of which 6,842 households co-operated fully in the survey. The overall response rate for the 2019/20 LCFS from eligible households was 40 per cent in Great Britain. In Northern Ireland 371 households co-operated fully, a response rate from eligible households of 44 per cent.

The technical report for the Living Costs and Food Survey gives more details of the methodology.

Defra is the main user of the statistics in its coordinating role on food policy across Government. The statistics feature in high level indicators on healthy diet and food security. In Scotland the statistics are used to monitor the health of the Scottish diet. The data is placed on the National Data Archive and is accessed by academics and used in research.

Family Spending is a separate report on the Living Costs and Food Survey published by the Office for National Statistics. It covers all forms of household expenditure but without as much detail on food and without quantities of food purchases.

3. Comparisons between ONS and Defra reports

Family Food uses LCFS data on food purchases is used to estimate food consumption. It should be noted that in Family Food, food consumption is at person level.

Family Spending reports expenditure at household level, meaning that the figures cannot be directly compared to those presented in Family Food. The different approaches reflect the different analytical purposes of the two publications.

Family Spending includes data on the share of household expenditure that is spent on household food. Engel’s law is an observation in economics stating that as income rises the proportion of income spent on food falls, even if actual expenditure on food rises. Although these estimates are of proportion of expenditure not income, they are consistent with that observation.

4. National Statistics

Family Food conforms fully to National Statistics standards. The term ‘National Statistics’ is an accreditation quality mark that stands for a range of qualities such as relevance, integrity, quality, accessibility, value for money and freedom from political influence. More information is available from the UK Statistics Authority.

4.1 Revisions

In the production of the 2017/18 estimates, ONS discovered an issue with the weights used to generate the 2016/17 purchases estimates. This affected, to a small degree, most estimates of average purchased quantities. We have published revised estimates for 2016/17 as part of this release. The corrected estimates vary by around 1% from the previous ones.

5. Survey development

5.1 Weights and measures

When completing the food diary portion of the LCFS, a sub-sample of respondents are asked instead to fill out an expanded diary, which as well as expenditure on food, seeks to also record the weights and details of purchased foods. ONS apply a custom calibration to certain survey variables, meaning that for each specified constraint the weighted sub-sample is adjusted so that it sums to the totals from the full sample. The calibration builds on the base LCFS calibration on age/sex and region, and adds Alcohol (P519t), Food (P518t) and Non-alcoholic drink (FS12).

Table: Comparison of standard errors for different calibration sets

COICOP
  Total alcoholic beverages, tobacco Total clothing and footwear Total housing, water, electricity Total furnishings, household equipment Total health expenditure Total transport costs Total communication Total recreation Total education Total restaurants and hotels Total misc. goods and services
  (P602) (P603) (P604) (P605) (P606) (P607) (P608) (P609) (P610) (P611) (P612)
2016 Q2                      
Base calibration 0.59 1.65 4.02 4.4 0.79 3.73 0.41 4.55 1.48 2.28 1.58
Defra 0.58 1.6 4.02 4.2 0.75 3.7 0.41 4.33 1.46 2.2 1.58
2016 Q3                      
Base calibration 0.69 1.21 3.39 3.14 0.71 3.41 0.44 4.75 0.89 2.87 1.72
Defra 0.68 1.2 3.38 3.15 0.71 3.39 0.44 4.73 0.91 2.88 1.7
2016 Q4                      
Base calibration 0.83 1.77 3.15 3.18 0.7 3.32 0.59 4.16 2.85 2.08 1.8
Defra 0.82 1.76 3.15 3.17 0.7 3.31 0.59 4.14 2.82 2.06 1.79

5.2 Accuracy of reporting and coding

Survey participants record their food and drink purchases in a two-week diary. They are able to attach till receipts or to write in diary entries to cover amount spent and quantity purchased for each individual item. In some cases, there is insufficient detail recorded on the diary to identify the correct food code, or quantities are not properly recorded. Whilst every effort is made by the survey team to correct these during household visits it is sometimes necessary to tolerate this in order to maintain goodwill and high response rates.

To deal with quantity omissions on the diary the validation team collect proxy quantities by searching on-line supermarket websites and matching the item description and expenditure. If there is insufficient information to allocate a food item to a specific code, default codes may sometimes be used. Default codes are based upon the most commonly occurring product within a category; e.g. a diary entry of ‘sausages’ gives insufficient information to distinguish between pork/beef/other meat, so in this case it would be allocated to the ‘pork’ food code by way of default as the most commonly bought variety.

5.3 Checks on portion sizes to improve the quality of eating out estimates

Quantities are not recorded against eating out foods on the Family Food diaries because purchases are often in the form of meals and quantities are unknown. In the eating out section of the Family Food diary the survey participant records an itemised list of meal components. Defra uses a set of standard portion sizes for eating out food codes. These were reviewed in 2013, and no significant changes were made.

6. Family Food production team

Isabella Worth, Leigh Riley, Chris Silwood, Jonathan Smith, David Lee and Matthew Keating.

7. Feedback

We would welcome feedback and suggestions from users of Family Food and its datasets. Contact the team at familyfood@defra.gov.uk.

8. Data downloads

Data in spreadsheet format are available to download from the Defra website.

The Family Food data are spreadsheets containing survey estimates for years 2001/02 onwards. The UK household consumption and the UK household expenditure spreadsheets show results for 1974 onwards. Historical estimates going back to 1940 in some cases are available from the National Archives.

Information is available at United Kingdom level for both household and eating out on:

  • Purchases
  • Expenditure

There is a further breakdown by:

UK regions

  • Scotland, Wales, Northern Ireland, English NUTS 1 Region
  • Rural and Urban: England, Wales and Scotland

and other breakdowns across the UK sample:

  • Gross income quintile
  • Equivalised income decile
  • Household composition
  • Age group of household reference person
  • Age at which household reference person ceased full-time education
  • Ethnic origin of household reference person
  • Socio-economic classification of household reference person
  • Economic activity of household reference person

8.1 UK Data Service

Survey data for the Expenditure and Food Survey (2000/01 to 2007) and subsequently the Living Costs and Food Survey (since 2008) is available to download, as well as the National Food Survey data from 1974 to 2000, from the UK Data Service.

9. Glossary

9.1 Consumer Price Index (CPI)

The Consumer Price Index is a measure of consumer price inflation produced to international standards and in line with European regulations. The CPI is the inflation measure used in the government’s target for inflation.

The CPI is produced at the same level of detail as the CPIH in the accompanying dataset and accompanying data time series.

9.2 Equivalised income

The income a household needs to attain a given standard of living will depend on its size and composition. Equivalisation means adjusting a household’s income for size and composition so that the incomes of all households are on a comparable basis. To calculate equivalised income using the ‘Modified OECD’ equivalence scale, each household member is given an equivalence value. This scale, first proposed by Haagenars et al. (1994), assigns a value of 1 to the household head, of 0.5 to each additional adult member and of 0.3 to each child. Additional household members are assigned smaller values to reflect the economies of scale achieved when people live together. Economies of scale arise when households share resources such as water and electricity, which reduces the living costs per person.

9.3 Household Reference Person (HRP)

The HRP is the person who: owns the household accommodation, or is legally responsible for the rent of the accommodation, or has the household accommodation by virtue of their employment or personal relationship to the owner who is not a member of the household. If more than one person meets these criteria the HRP will be the one with the higher income. If the incomes are the same then the eldest is chosen.

9.4 Main effect regression

A statistical technique that does not allow the effect of an explanatory variable (e.g. age) to change when another explanatory variable (e.g. region) changes.

9.5 Multiple regression modelling

A statistical technique that predicts values of one variable (e.g. intake of fat) on the basis of two or more other variables (e.g. age, region and income).

Return to Family Food 2020/21 home page