1. Methodology background
|National Statistic|| |
2. Executive summary
The Living Costs and Food Survey (LCF) is an annual survey, designed primarily to measure household expenditure on goods and services. It also gathers information about the income of household members. Respondents, including children, keep a detailed diary of expenditure for 2 weeks. Respondents also record the weights and volumes of food and drink items bought.
A household expenditure survey has been conducted each year in the UK since 1957. From 1957 to March 2001, the Family Expenditure Survey (FES) and National Food Survey (NFS) provided information on household expenditure patterns and food consumption. Both surveys were well-established important sources of information for government and the wider community, charting changes and patterns in the UK’s spending and food consumption since the 1950s. In April 2001, these surveys were combined to form the Expenditure and Food Survey (EFS). The EFS was renamed as the Living Costs and Food (LCF) survey in January 2008 when it became part of the Integrated Household Survey (IHS). The change saw the insertion of a core set of questions, common to all of the separate surveys that together comprise the IHS.
The long history of the survey has inevitably meant that discontinuities have been introduced. More information on this can be found in the section on Comparability and coherence.
The LCF sample for Great Britain is a multi-stage stratified random sample. Addresses are drawn from the Postcode Address File (PAF). In Northern Ireland, the companion survey to the Great Britain LCF is conducted by the Central Survey Unit of the Northern Ireland Statistics and Research Agency (NISRA). A systematic random sample of private addresses is drawn from the Land and Property Services Agency’s database.
Results from the LCF are published in two main publications: Family spending and Family Food. Family spending, produced by Office for National Statistics (ONS), provides a comprehensive overview of household expenditure and income and also includes background information on the survey’s history and methodology. Family Food, produced by the Department for Environment, Food and Rural Affairs (DEFRA), provides detailed statistical information on the purchase and consumption of food and drink.
The UK Data Archive at Essex University provides free access to the various LCF datasets, and can be contacted via the UK Data Archive website. The archive also provides detailed user guides on the survey.
This report contains the following sections:
- Output quality
- About the output
- How the output is created
- Validation and quality assurance
- Concepts and definitions
- Other information, relating to quality trade-offs and user needs
- Sources for further information or advice
3. Output quality
This report provides a range of information that describes the quality of the output and details any points that should be noted when using the output.
We have developed Guidelines for Measuring Statistical Quality; these are based upon the five European Statistical System (ESS) quality dimensions. This report addresses the quality dimensions and important quality characteristics, which are:
- timeliness and punctuality
- coherence and comparability
- output quality trade-offs
- assessment of user needs and perceptions
- accessibility and clarity
More information is provided about these quality dimensions in the following sections.Back to table of contents
4. About the output
(The degree to which the statistical outputs meet users’ needs.)
The LCF meets the needs of a range of users. These are summarised in this section.
One of the main purposes of the LCF is to define the basket of goods for the Consumer Prices Index (CPI) and the Retail Prices Index (RPI). The CPI is the main UK domestic measure of consumer price inflation for macroeconomic purposes. It forms the basis for the government’s target for inflation that the Bank of England’s Monetary Policy Committee (MPC) is required to achieve. The CPI is also used to uprate benefits, tax credits and public service pensions. The RPI is the most long-standing general purpose measure of inflation in the UK.
Information from the LCF is also a major source for estimates of household expenditure in the UK National Accounts.
LCF information on expenditure and income is used by HM Treasury and HM Revenue and Customs to study how taxes and benefits affect household incomes and to analyse the effects of policy in these areas.
Regional LCF information is one of the sources of regional estimates of consumer spending and other regional statistics.
Department for Environment, Food and Rural Affairs (DEFRA), which sponsors the collection of specialist food data, uses LCF data in its Family Food publication and other reports. Many other government departments use LCF data as a basis for policy-making, for example, in the areas of housing and transport.
Users outside government include independent research institutes, academic researchers and business and market researchers.
The main results from the survey are published by ONS in the annual Family spending statistical bulletin.
The LCF is the primary source of official information on household expenditure on goods and services and is also an important and detailed source of income data. The survey design and outputs are tailored to user needs (see the section on How the output is created). This is reflected in feedback obtained from main users, as detailed under Assessment of user needs and perceptions. However, some limitations are noted.
The survey does not provide a balance sheet of income and expenditure, either for individual households or for groups of households. The majority of expenditure information collected relates to the 2-week period immediately following the interview, whereas income components refer to a much longer period.
Experience of household surveys in the UK and in other countries indicates that reported expenditure on a few items (notably tobacco and alcohol) is below the levels that might be
expected by comparison with other sources of information. National Lottery spending has also been under-recorded in the LCF, particularly for scratch cards.
Timeliness and punctuality
(Timeliness refers to the lapse of time between publication and the period to which the data refer. Punctuality refers to the gap between planned and actual publication dates.)
ONS - Family spending
Family spending is timetabled to be published 11 months after the end of the reporting year. Once published, several deliveries of LCF data are released to other divisions of ONS and to other government departments.
Household Income and Expenditure (HIE) Branch
Processed data are supplied to HIE Branch of ONS both quarterly and annually. Quarterly files are timetabled to be delivered approximately 10 to 11 weeks after the end of the reporting period. The data contained in the quarterly files are only used for checking the income data and not for publication.
Annual files are timetabled to be delivered approximately 6 months after the end of the reporting period. The data are used for the Effects of Taxes and Benefits on Household Income publication.
Household and Non-Profit Institutions Serving Households (HNPISH) Branch
Processed data are timetabled to be supplied to the HNPISH Branch on a quarterly basis approximately 9 to 10 weeks after the end of the reporting period. The data are used for the Consumer trends publication.
RPI Production and Commodity Analysis Branch
Processed data are timetabled to be supplied to the RPI Production and Commodity Analysis Branch on an annual basis. These are timetabled to be delivered approximately 2 months after the end of the reporting period.
DEFRA receives raw data from the LCF on a monthly basis. DEFRA is timetabled to receive an initial supply of data for quality assurance purposes 10 weeks after the end of the reporting period and a re-supply of the data 6 weeks later, which includes any amendments made to the food coding arising from the quality assurance process.
Eurostat receives LCF data every 5 years for inclusion in the European Household Budget Survey.
Other annual outputs
ONS has a service level agreement with Department for Business, Energy and Industrial Strategy (BEIS). ONS send BEIS data on fuel expenditure following the publication of Family spending.
Deliveries to DEFRA, HIE Branch, HNPISH Branch and the RPI Production and Commodity Analysis Branch are usually made to timetable. Publication of the 2009 edition of Family spending was delayed because of procedural problems encountered in preparing the data. The delay was
conducted in compliance with the Code of Practice for Official Statistics, Protocol 2: Release Practices. The decision was appropriately authorised and publicised.
For more details on related releases, the GOV.UK release calendar provides 12 months’ advance notice of release dates. If there are any changes to the pre-announced release schedule, public attention will be drawn to the change and the reasons for the change will be explained fully at the same time, as set out in the Code of Practice for Official Statistics.Back to table of contents
5. How the output is created
(This section outlines the methods used to produce LCF outputs. More information is available in the publication Family spending and the user guides deposited with the UK Data Archive.)
The LCF is a voluntary sample survey of private households. The basic unit of the survey is the household. The LCF (in line with other government social surveys) uses the harmonised definition of a household: a group of people (not necessarily related) living at the same address who share cooking facilities and share a living room or sitting room or dining area.
Each individual aged 16 or over in the household visited is asked to keep diary records of daily expenditure for 2 weeks. Information about regular expenditure, such as rent and mortgage payments, is obtained from a household interview along with retrospective information on certain large, infrequent expenditures such as those on vehicles. Children aged between 7 and 15 keep a simplified diary.
Detailed questions are asked about the income of each adult member of the household. In addition, personal information such as age, sex and marital status is recorded for each household member.
The fieldwork, and the editing and coding of data are carried out on a monthly basis. Interim monthly datasets are compiled and quality assured before being delivered to Department for Environment, Food and Rural Affairs (DEFRA). Additional quality assurance checks are carried out at the completion of each quarter and quarterly datasets are produced. Once the fieldwork for the calendar year has been completed, further quality assurance of the data is carried out before producing the annual dataset.
Approximately 6,000 responding households in the UK per year.
Postcode Address File (PAF).
The LCF sample for Great Britain is a multi-stage stratified random sample. Addresses on the PAF with “small user” postcodes are used as the sample frame. Postal sectors are used as the Primary Sampling Units (PSUs), with 18 addresses selected from each PSU to form the monthly interviewer quota. A total of 638 PSUs are selected annually after being arranged in strata defined by government regions and two 2001 Census variables: socio-economic group of the head of household and ownership of cars.
In Northern Ireland, the companion survey to the Great Britain LCF is conducted by the Central Survey Unit of the Northern Ireland Statistics and Research Agency (NISRA). A systematic random sample of private addresses is drawn from the Land and Property Services Agency’s database.
The fieldwork is conducted by ONS in Great Britain and by NISRA for the Department of Finance and Personnel in Northern Ireland using largely identical questionnaires. Differences between the two questionnaires reflect the country-specific harmonised standards for ethnicity, nationality and national identity, and the different systems of local taxation used in Great Britain and Northern Ireland.
Households at the selected addresses are visited and asked to co-operate in the survey. In order to maximise response, interviewers make at least four separate calls and sometimes many more, at different times of the day to households that are difficult to contact.
Interviews are conducted by Computer Assisted Personal Interviewing (CAPI) using laptop computers. Respondents complete a face-to-face interview and each individual aged 16 or over in the visited household is asked to keep a diary of daily expenditure for 2 weeks.
A detailed description of the weighting process and its effects can be found in the Living Costs and Food Survey Technical Report.
The weighting procedure comprises two stages. First, a non-response adjustment is made using weighting classes derived from an analysis of respondent and non-respondent LCF households using addresses linking LCF response status to the 2001 Census data; for LCF addresses selected around the time of the census. Second, the non-response weights are calibrated to population totals for a set of calibration groups constructed from the numbers of males and females in different age groups and, separately, for regions.
From 2007, the results from the 2001 Census-linked study of non-respondents have been used to carry out the non-response weighting. The LCF took part in the Census Non-Response Linked Study from the 2011 Census, the results of which may be used to update non-response weighting in due course.
Calibration is used to adjust the non-response weights in such a way that weighted totals in the calibration groups match their counterpart population totals, based on a population definition of people living in private households. From 2006 onwards, the population totals have been projections based on estimates rolled forward from the 2001 Census using births and deaths registrations data and migration estimates.
Definitions of response outcome categories
A full response denotes a household in which:
- all adults aged 16 and over co-operated with the interview
- no income questions in the questionnaire were refused
- all adults kept a 2-week record of their expenditure
- the information given was complete and usable
There are three types of partial responses on the LCF:
- one or more adults, that are not the Main Diary Keeper (MDK) in the household refuse to keep the diary, but all adults complete the full income section of the interview
- one or more adults refuse the full income section but all adults in the household keep the diary and main income information is collected for all adults
- one or more adults (that are not the MDK) refuse to keep the diary and one or more adults refuse the full income section but main income information is collected for all adults
All partial responses must contain a diary from the MDK. If the MDK refuses to complete the diary, the household is classified as a refusal.
An outright refusal is a household that refuses to respond to the survey and the interviewer feels that there is no chance of an interview at the given time. In addition to outright refusals, there are also refusals when some of the information has not been collected. Refusals on the LCF are defined as:
- all adults complete the full income section, but the MDK of the household refuses to keep the diary
- one adult refuses to give the main income information in the questionnaire
A non-contact arises when an address is occupied but where it has not been possible to contact any member of the household in the field period.
A re-issue is an address where the interviewer has been unable to make contact or the household has refused to take part in the survey, but an interviewer has judged that the household is suitable for re-issue in a later field period month. These cases are re-distributed to field managers for assignment to interviewers at a later date. Re-issues are allocated on an ad- hoc basis to interviewers who are able to accept them alongside their main quota of work. Interviewers are expected to complete the re-issues within the field period of their main quota. Re-issued addresses are processed independently of the routinely collected interviews and diaries and then added to the annual dataset.
Ideally, all adult members of the household should be present during the interview, so that the income section can be asked personally. However, where a member of the household is not present during the interview, another member of the household (such as spouse) may be able to provide documentary information about the absent person. The individual interview is then identified as a proxy interview.
From 2001to 2002 the EFS or LCF has included households that contained a proxy interview. Around a fifth of LCF households contain at least one proxy interview. Additional information on proxy interviews is available in the Living Costs and Food Survey Technical Report.
Statistical disclosure control methodology is also applied to LCF data. This ensures that information attributable to an individual is not disclosed in any publication. The Code of Practice for Official Statistics and specifically Principle 5: Confidentiality, sets out practices for how we protect data from being disclosed. The Principle includes a guarantee to survey respondents to “ensure that official statistics do not reveal the identity of an individual or organisation, or any private information relating to them”.Back to table of contents
6. Validation and quality assurance
The degree of closeness between an estimate and the true value.
The LCF is conducted with people who volunteer their time to answer questions about themselves and keep a diary for 2 weeks. The voluntary nature of the survey means that people who do not wish to take part in the survey can refuse to do so. Reasons for not participating in the survey include people who “can’t be bothered” and those with a “bad experience of a previous survey”. The sample is designed to ensure that the results of the survey represent the population of the UK. The risk of the survey not being representative may increase with every refusal or non-contact with a sampled household (survey non-response). One measure of the quality of survey results is therefore the response rate.
Response rates are available in the Living Costs and Food Survey Technical Report. In 2015 to 2016, the response rate was 46% in Great Britain. A long-term decline in response has been observed for the LCF, in common with other social surveys. Non-response weighting is applied to help mitigate non-response bias.
Prior to publication LCF data are subjected to a rigorous process of quality assurance. An initial series of automatic checks are applied to raw household and income data as they are collected from respondents and entered onto the CAPI version of the questionnaire. These data are further checked by a team of editors within ONS who also impute for missing values. Missing data are imputed using a combination of the following methods:
- by reference to tables based on external (non-LCF) data published elsewhere
- by reference to tables based on LCF data from previous years showing average amounts according to household income
- by using information collected elsewhere in the questionnaire or by referring back to the interviewers
Respondents’ diaries are checked after a few days to ensure the process of recording daily expenditure has been understood and before they are returned to ONS to ensure recorded information is complete. In manually entering diary data to ONS systems, further checks are made to these data.
Prior to the initial supply of data to Department for Environment, Food and Rural Affairs (DEFRA), further checks are made to raw data to ensure consistency between household, person and diary files and to check the coding of specific products known to have been coded incorrectly in the past.
DEFRA complete further checks of the data, feeding back their observations to ONS, before final versions of these raw datasets are supplied to DEFRA.
Further quality assurance is also undertaken once the data have been aggregated into quarterly and annual datasets. Household and Non-Profit Institutions Serving Households (HNPISH) Branch, of National Accounts Group, identify unusual movements in the aggregate data and HIE carry out detailed checks on the income and benefits data. Suspicious data are further investigated by the LCF team and may be corrected if found to be in error.
The LCF team also carries out a series of checks on the time series data to identify odd movements and to identify extreme values or outliers in household and person level datasets. In general, outliers are not treated on the LCF, with the exception of extreme values of income. However, only a few cases are treated each year by using the “surprise stratum” method. This method treats each outlier as if it were the only extreme unit in the stratum population and involves giving the outlying case a weight of one. This technique effectively reduces the impact of the outlying value without deleting it altogether.
Sources of error
Survey results are subject to various sources of error. The total error in a survey estimate is the difference between the estimate derived from the data collected and the true value for the population. It is helpful to distinguish between systematic and random error.
Systematic error, or bias, covers those sources of error that will not be expected to average to zero over repeats of the survey. Bias may occur, for example, if a certain section of the population is excluded from the sampling frame, because non-respondents to the survey have different characteristics to respondents, or if interviewers systematically influence responses in one way or another.
Substantial efforts are made to avoid systematic errors and these include:
- processes to ensure that households are selected in accordance with the sample design
- extensive measures to minimise non-response
- training to ensure that interviewers ask questions in such a way as to avoid biasing response
Additionally, the data are weighted to compensate for non-response; each respondent is given a weight so that they represent the non-respondents that are similar to them in terms of the survey characteristics. The sample distribution is weighted so that it matches the population distribution in terms of region, age group and sex.
Census Non-Response Linked Study (CNRLS)
Along with several other social surveys, the LCF is taking part in the CNRLS using data from the 2011 Census. The CNRLS matches census returns to recent sampled addresses from select ONS social surveys. The characteristics of survey responders and non-responders will be compared and evaluated using census data, providing an estimate of non-response bias. Where appropriate, this may be used to redesign the weighting methods that are used to correct for non-response bias in social surveys such including LCF.
Random error is the difference from the estimates derived using the sample data from the true values for the population that occur through chance occurrences. Random error may result from sources such as variation in a respondent’s interpretation of the survey questions, or interviewer inconsistencies in asking questions. Efforts are made to minimise these effects through pilot work to ensure questions are understood by respondents and through interviewer training emphasising the need for consistency in dealing with selected sample members.
An important component of random error is sampling error, which arises because the estimate is based on a sample rather than a full census of the population. The results obtained for any single sample may, by chance, vary from the true values for the population but the variation would be expected to zero over many hypothetical repeats of the survey under the same conditions. Precision is usually estimated through the calculation of standard errors. Standard errors are estimated for some of the main variables on LCF. These are published in the Living Costs and Food Survey Technical Report.
The LCF uses a multi-stage stratified sample design. Consequently, it is inappropriate to estimate standard errors on the basis of a simple random design, as this will not reflect the true
sampling variation because of the complex sample design. The two-stage sample of addresses can lead to a substantial increase in standard error if the households or individuals within Primary Sampling Units (PSUs) are relatively homogeneous but the PSUs differ from one another. Stratification tends to reduce sampling error and is of most advantage where the stratification factor is strongly related to characteristics of interest in the survey.
The design factor, or deft, is the ratio of the standard error of an estimate calculated taking into account the complex design relative to the standard error that would have resulted had the survey design been a simple random sample of the same size. Design factors are estimated for some of the main variables on LCF. These are published in the Living Costs and Food Survey Technical Report.
The size of the deft varies between survey variables according to the degree to which a characteristic is clustered within PSUs, or is distributed between strata, and the impact of the weighting. For a single variable, the size of the deft also varies according to the size of the subgroup on which the estimate is based and on the distribution of the subgroup between PSUs and strata.
Defts below 1.0 show that the standard errors associated with the complex design are lower than those associated with the simple random design due probably to the benefits of stratification. Defts greater than 1.0 show the survey has produced less precise estimates than would be obtained from a comparable simple random sample because of the effects of clustering and weighting.
Comparability and coherence
(Comparability is the degree to which data can be compared over time and domain, such as geographic level. Coherence is the degree to which data that are derived from different sources or methods, but refer to the same topic, are similar.)
As already noted, a household expenditure survey has been conducted each year in the UK since 1957. A methodological discontinuity was introduced when the Family Expenditure Survey (FES) and National Food Survey (NFS) were combined to form the Expenditure and Food Survey (EFS) in April 2001. At the same time, the survey adopted the Classification of Individual Consumption by Purpose (COICOP), outlined in the section on Concepts and definitions. This meant a significant change to the categorisation of expenditure.
Until 2005 to 2006, annual data were collected and published on a financial year basis; from 2006 data were collected on a calendar year basis.
In 2008, the EFS was renamed as the Living Costs and Food Survey and became part of the Integrated Household Survey. This involved the insertion of a core set of questions, common to all of the separate modules that together comprise the IHS. However, collection of data on income and expenditure underwent little discontinuity over this transition.
Other more minor changes to definitions used in the survey have been introduced. Changes made since 1991 are documented in the Living Costs and Food Survey Technical Report.
The main comparator for LCF estimates of expenditure data are the figures on final household consumption expenditure (HHFCE) published in Consumer trends and used in UK National Accounts. LCF data feed into some of the estimates published in Consumer trends, but other sources are also used. While differences occur in the estimates published, the differences are credible. Research is ongoing into the different estimates produced and their causes.Back to table of contents
7. Concepts and definitions
(Concepts and definitions describe the legislation governing the output, and a description of the classifications used in the output.)
The LCF utilises a suite of standard concepts and definitions. Some of the main terms are defined in this section, while a list of definitions and changes to these definitions can be found in the Living Costs and Food Survey Technical Report. These classifications reflect the needs of the users and also the survey’s inclusion in the National Statistics Harmonisation Programme.
Classification of Individual Consumption by Purpose (COICOP)
The COICOP coding frame for expenditure items was introduced in 2001 to 2002. COICOP has been adapted to the needs of the Household Budget Survey (HBS) across the EU and is therefore compatible with similar national accounts and consumer prices indices. This allows the production of indicators which are comparable Europe-wide, such as the Harmonised Indices of Consumer Prices (CPI in the UK).
National Statistics Socio-economic Classification (NS-SEC)
In 2001 to 2002, the NS-SEC was adopted for all official surveys, replacing the social class based on occupation and socio-economic group. The long-term unemployed, which fall into a separate category, are defined as those unemployed and seeking work for 12 months or more.
Household Reference Person (HRP)
The HRP is defined as the person who:
- owns the household accommodation, or
- is legally responsible for the rent of the accommodation, or
- has the household accommodation as an emolument or perquisite, or
- has the household accommodation by virtue of some relationship to the owner who is not a member of the household
If there are joint householders the HRP is the one with the highest income; if the income is the same, then the eldest householder is taken. The concept of the HRP replaced the previous concept of the head of household in 2001 to 2002.
A household is defined as one person living alone or a group of people (not necessarily related) living at the same address who share cooking facilities and share a living room or sitting room or dining area.Back to table of contents
8. Other information
Output quality trade-offs
(Trade-offs are the extent to which different dimensions of quality are balanced against each other.)
Publication of Family spending and the final LCF annual datasets, is completed as soon as possible, compatible with completing rigorous quality assurance conducted after all data have been collected and processed (see the section on Timeliness and punctuality).
Internal customers are provided with provisional datasets to provide data to feed into main outputs. Datasets are delivered quarterly as an input to the household expenditure component of UK National Accounts and quarterly GDP estimates; a final annual dataset is later delivered. Annual mid-year data are supplied to the RPI approximately 2 months after the reporting period, also representing provisional data for part of the relevant period covered by the data.
Assessment of user needs and perceptions
(The processes for finding out about users and uses, and their views on the statistical products.)
The LCF convenes a steering group of important customers in ONS and other government departments. Meetings are usually held biennially. The steering group is also consulted between meetings on proposals and developments, for example, on variable usage and questionnaire changes. A recent consultation of the steering group revealed positive views about the survey’s products, though some concerns were expressed about the observed decline in response rates.
The LCF held a user engagement event in 2012 aimed at the academic and wider user community. This enabled users to communicate their uses of the data, and provided an opportunity for feedback and discussion. Views from users have also been obtained via a facility on the ONS website.
Views obtained from the wider user community also feed into formulation of outputs. For example, feedback from users of Family spending indicated that comprehensive coverage of housing costs would be useful to the user community. This led to a chapter on housing being included in the publication each year since 2004.Back to table of contents
9. Sources for further information or advice
Accessibility and clarity
(Accessibility is the ease with which users are able to access the data, also reflecting the format in which the data are available and the availability of supporting information. Clarity refers to the quality and sufficiency of the release details, illustrations and accompanying advice.)
Our recommended format for accessible content is a combination of HTML webpages for narrative, charts and graphs, with data being provided in usable formats such as CSV and Excel. Our website also offers users the option to download the narrative in PDF format. In some instances other software may be used, or may be available on request. Available formats for content published on our website but not produced by us, or referenced on our website but stored elsewhere, may vary. For further information please refer to the contact details at the beginning of this report.
For information regarding conditions of access to data, please refer to the following links:
- Terms and conditions (for data on the website)
- Copyright and reuse of published data
- Pre-release access (ended from 1 July 2017)
- Virtual Microdata Laboratory (VML)
Further methodology information is available on the Living Costs and Food Survey methodology pages, including access to datasets and the technical report.
A National Statistics Quality Review of the Living Costs and Food Survey was published in May 2016 and an ONS response in December 2016.Back to table of contents