United Kingdom 1986 - documentation

Table of Contents

• A. GENERAL INFORMATION
• B. POPULATION AND SAMPLE SIZE, SAMPLING METHODS
• C. MEASURES OF DATA QUALITY
• D. DATA COLLECTION AND ACQUISITION
• E. WEIGHTING PROCEDURES
• F. DETERMINATION OF SURVEY UNIT MEMBERSHIP
• G. CHILDREN AND SPOUSES
• H. AVAILABILITY OF BASIC SOCIAL AND DEMOGRAPHIC INFORMATION
• I. AVAILABILITY OF LABOUR MARKET INFORMATION
• J. AVAILABILITY OF GEOGRAPHIC INFORMATION
• K. SOURCES AND AMOUNTS OF CASH INCOME
• L. TAXES
• M. BIBLIOGRAPHY OF MAIN PUBLICATIONS

A. GENERAL INFORMATION back     (see note 1)

Official name of the survey:
Family Expenditure Survey 1986

The United Kingdom Department of Employment has overall responsibility for the Family Expenditure Survey (FES), and for processing and publishing the results (see note 2). The Social Survey Division is responsible for all stages of the management of the FES up to and including coding. This includes the sample selection procedures, design of the questionnaires and other documents used in the day-to-day work of the FES, field-work, the editing and coding of completed schedules, including any recalls on respondents for further information (see note 3) Detailed plans for each year's survey are considered by an inter-departmental committee under the chairmanship of the United Kingdom Central Statistical Office. Close liason is maintained between the Department of Employment, the Central Statistical Office and the Social Survey in the day-to-day running of the Survey (see note 4 ).

Additional information to that provided here can be obtained from:

 The FES data for 1979, for the years 1961-1963, and from 1968 through to the present are stored at the Economic and Social Research Council (ESRC) Data Archives. Copies of the original codebooks, questionnaires and other documentation are also available from the ESRC Data Archives.

The principle users of the data collected in this survey are UK central government departments. It is used by the House of Commons and other government departments as a basis for economic policy decisions. It is also used by local authorities, employers, trade unions, and research workers in universities and independent research institutes (see note 5 ).

Historically, the main purpose of the survey has been to provide information on spending patterns for the UK Retail Prices Index (also referred to as the "Cost of Living Index"). The Family Expenditure Survey originates from a recommendation of the Cost of Living Advisory Committee (now the Retail Prices Index Advisory Committee) in an interim report published in 1951 (Interim Report of the Cost of Living Advisory Committee) that an enquiry should take place into the patterns of expenditure of private households as a source for the weighting pattern of the Index of Retail Prices, to be followed by smaller-scale enquiries at frequent intervals. A large-scale Household Expenditure Enquiry was undertaken in in 1953-546 and was followed by the Family Expenditure Survey, which has been in continuous operation since 1957
(see note 7).

Users of the 1986 FES data should be aware that some major differences exist between the 1986 and earlier Family Expenditure Surveys, including the 1969 and 1979 surveys in the LIS database. Some definitions have been changed, as has the recording of certain income sources, (see note 8) some of which affect those variables available in the LIS versions of the FES. In addition, beginning with the 1986 Family Expenditure Survey, major changes have been made to the FES sample design for the first time since 1967 (see note 9). Consequently, care should be exercised when making comparisons between the 1986 and earlier Family Expenditure Surveys.

Detailed information on changes in the sample design can be found in Section B of this chapter. For information on changes in variable definitions the reader should consult Section M.

Concerning access to the FES data, it should be noted that it is:

"provided to enquirers on request, subject to stringent undertakings on the use of the data. All applications to the Archive are examined and approved by the Department of Employment before data are made available."(see note 10)

The Family Expenditure Survey microdata have been made available to the Luxembourg Income Study Project under the condition that each user or potential user of the FES data sign a Confidentiality Undertaking Form11 before he or she is given access to the data. This applies whether the applicant is accessing the data himself, or whether LIS carries it out on his or her behalf. Users or potential users must agree to:

1. not supply copies of any tape or database of individual FES data to anyone else without the prior permission of the Central Statistical Office; 
2. acknowledge the source of the data in any work based on analysis of data supplied from the survey, and to state clearly that those who carried out the original collection and analysis of the data bear no responsibility for further analysis and interpretation of the data;
3. not include details relating to particular households or persons in any publication or circulated document;
4. send to the UK Central Statistical Office, if requested to do so, a copy of any report, whether for publication or for wider circulation to a conference or seminar, which is based on further analyses of the data.

The process of data collection (i.e. field work) for the first wave of LIS data was begun in January 1969 and completed (see note 12) in December 1969, and for the second wave of LIS data was begun in January 1979 and completed in December 1979, and for the third wave of LIS data was begun in January 1986 and completed in December 1986.

A bibliography of the most important publications based on the Family Expenditure Survey data is provided in Section N.

Also included in the bibliography in Section N is a list of available user documentation for the Family Expenditure Survey data.

Notes: 

1. Much of the information provided in this section on the Family Expenditure Survey was taken from documents published by Her Majesty's Stationary Office. In the interests of style and readability, direct quotations from these documents have in most cases not been placed within quotation marks. To do so would have required the use of cumbersome introductions to each quoted passage. Given that the technical documentation of the LIS database is to be provided mainly to users of the LIS database, and is not intended for general circulation, nor will be available for purchase, it seemed to us more appropriate to present the material in this manner. In any case, references to the documents from which passages have been excerpted have in all cases been provided.
2. Department of Employment. 1988. Family Expenditue Survey 1986 (Revised). London: HMSO. p. vii
3. Family Expenditure Survey Handbook, 1980, p.1
4. Family Expenditure Survey 1979. p.1
5. Family Expenditure Survey 1979. "Introduction" p. iii.
6. Report of an Enquiry into Household Expenditure in 1953- 54, 1957, HMSO
7. Family Expenditure Survey 1979. "Appendix 1" p. 143.
8. Department of Employment. 1988. Family Expenditue Survey 1986 (Revised). London: HMSO. p. v.
9. Department of Employment. 1988. Family Expenditue Survey 1986 (Revised). London: HMSO. p. vi.
10. Family Expenditure Survey 1979. "Appendix 1," p. 144.
11. Confidentiality Undertaking forms are available from LIS.

B POPULATION AND SAMPLE SIZE, SAMPLING METHODS back

1. The Sample Design and Sampling Frames  

For the Family Expenditure Survey two separate sample designs are employed: one in Great Britain (England, Scotland and Wales), and one in Northern Ireland.

Although the sample design for the Northern Ireland survey has not been changed, in 1986 the Great Britain FES sample design was changed for the first time since 1967. Instead of using the Parliamentary Electoral Registers as the final stage sampling frame (see note 13), the 1986 FES uses the "small users file" of the Postcode Address File (PAF) (see note 14)

For Great Britain a four stage, stratified sample design is employed. In Northern Ireland a three stage sample design is employed (see note 15).

No attempt was made to obtain interviews from households where it was discovered after the first interviewer call that the household contained:

 i. Members of the Diplomatic Service of any country (except U.K.)
ii. Members of the USA Forces
iii. Roman Catholic Priests if they are living in accomodation  provided by the parish churc (see note 16).

 In addition, because the FES is concerned with collecting information about domestic, not business expenditure, if a commercial establishment17 fell into the sample interviews were not obtained, except in those cases when at the selected address completely separate accomodation where the household is responsible for all its domestic expenses exists.

Despite its name, The Family Expenditure Survey is actually a survey of households, not families. Consequently, the survey unit is comprised of households. These are defined as:

 i. one person living alone; or
ii. a group of people living at the same address having meals prepared together and with common housekeeping.

Only private households are included. Resident domestic servants are included. The members of a household need not necessarily be related by blood or marriage (see note 18).

Unoccupied units were not excluded from the sample frame as there was no indication in advance that any unit might be unoccuopied (see note 19)

Households were not excluded if some or all members were not British subjects.

The total number of respondents who were actually surveyed (see note 20) was 7,045.

No case weights have been assigned in the Family Expenditure Survey. Theoretically, the sample design is intended to be self- weighting, but in practice this is not completely realized due to deficiencies in the data available for units listed in the sampling frames (see note 21).

Notes:

12 This time period does not include data processing or other survey operations.
13 See Section B of the Family Expenditure Survey Chapters for 1969 and 1979.
14 Department of Employment. 1988. Family Expenditue Survey 1986 (Revised). London: HMSO. p. vi.
15 Family Expenditure Survey 1979, 1980, pp. 146-147
16. Office of Population Censuses and Surveys. Social SurveyDivision. 1979. Family Expenditure Survey. Instructions to Interviewers. p.8
17. Such as public houses, hotels of all types, guest houses/commercial boarding houses, private households containing 4 or more boarders at the first interviewer call, and institutions (e.g. hostels, schools, prisons, hospitals, religious establishments). Family Expenditure Survey.Instructions to Interviewers. 1979. p.8
18. Family Expenditure Survey 1979. 1980. HMSO. p.153)
19. In cases where the sampling unit was comprised of housing units.
20. The number for which at least the minimum amount of data were obtained.
21. Family Expenditure Survey Handbook, 1980, p.6

C. MEASURES OF DATA QUALITY back

1. Response Rates

The overall response rate to the survey was approximately 69% (see note 22). Of the approximately 31% of households which were not surveyed, most of these represented households which did not wish to take part in the survey. In only about 1.5% of the sample was the interviewer unable to contact anyone living at the selected address. There is evidence that the characteristics of these households differ from those who cooperate, and it is therefore possible that their patterns of expenditure and income differ.

W.F.F. Kelmsley, in a study of the 1971 FES sample which compared this sample with the 1971 census, (see note 23) found substantial response variations. Households without children, and those where the head was or had been self-employed, produced lower response rates. Particularly marked was a fall in response with age.

Both heads of households and housewives showed a loss of about 2% in response for each increase of 5 years of age. There is also some evidence from an earlier study that response rates appear to be lower among households living in dwellings with higher rateable values (see note 24). Additional information on the effect of differential response estimates of income redistribution can be found in an article by R.P. Harris (see note 25). There is no reason to expect that the 1979 FES sample is markedly different from the 1971 sample in terms of non-response rates by these groups.

Estimates of sampling variability (standard errors) have been computed for this survey. Available estimates are reported in the Family Expenditure Survey 1979, pp.148-152.

2. Reporting and Under-Reporting and Income Data Quality

An evaluation of the quality of the income data has been made by comparing the survey data with census records. More detailed information can be found in Kemsley, W.F.F. 1975. "Family Expenditure Survey. A Study of Differential Response Based on a Comparison of the 1971 Sample with the Census." Statistical News 31 (November)

Internal edits and consistency checks of the data are regularly performed by the FES staff.

Notes:

22. Department of Employment. 1988. Family Expenditue Survey 1986 (Revised). London: HMSO. p. vi
23. Kemsley, W.F.F., "A Study of Differential Response Based on a Comparison of the 1971 Sample with the Census," in Statistical News. No. 31, November 1975.
24. Family Expenditure Survey 1979, 1980, pp. 2-3
25. Harris, R.P. "Differential Response in the Family Expenditure Survey: the Effect on Estimates of the Redistribution of Income." Statistical News, No, 39, February 1977

3. General Assessment of Income Data Quality

Users of this data should be aware of the following problems regarding the quality of the income data collected in this survey:

1. Self-employment income. As mentioned above, the 1971 Census/FES comparison showed higher non-response rates for for the self-employed than for other employment statuses (see note 26).

 2. Investment Income. It appears that income from dividends and interest are understated. However, most other budget surveys appear to also experience the same shortfall (see note 27).

Notes:

26. Family Expenditure Survey Handbook, 1980, p.54
27 Family Expenditure Survey Handbook, 1980, p. 54

 D. DATA COLLECTION AND ACQUISITION back  

1. Method of Data Collection

The data were collected via a personal visit by an interviewer. The interviewer was required to make at least five calls on the household. The following describes the standard interviewing procedure. At the initial call the interviewer ascertains how many households exist at the selected address, how many eligible people 16 years and over are at each household, and if cooperation can be obtained. An appointment is then made for an interview with the entire household. During this interview the respondents are told about the uses of the survey, and the interviewer administers the household schedule to the household head. In addition, an income schedule is administered to each income recipient. A diary record book is left with each "spender" (individuals over 16 years) to fill out over the next 14 days, and a credit card sheet is left with each holder of a credit card. On or before the fifth day following this interview the interviewer makes a first check call to inquire about the progress of record keeping. On or soon after the eighth day the interviewer returns again to check the progress of record keeping, and to collect the credit card sheets. A second set of credit card sheets are then left with the respondents. A final call is then made as soon as possible after the fourteenth day to collect the second set of credit card sheets and the diary record books. If all the schedules have been properly filled out each respondent is paid 2, and arrangements are made during this final call to mail a postal order to each respondent. Participation in the survey was voluntary, and the privacy of the respondents was assured by the interviewers.

2. Survey Instruments

Four survey insruments are used in the FES: 

1. Household schedule,
2. Income schedule,
3. Diary record book,
4. Credit card sheet.

The household and income schedules were administered by trained interviewers, and the diary record book and the credit card sheets were completed by the respondents with help from the interviewers if needed.

3. Survey Respondent Rules

Each household member over 16 years of age was required to complete the diary record book and other relevant schedules. If this was not done the infromation for that household was not included in the FES. Each respondent was required to answer the questions for himself or herself.

Respondents were asked to consult or use pay records, tax returns, or other documents to help provide the most accurate income information possible; however, it is not known what percentage of the respondents used records to help them answer the income or other questions.

E. WEIGHTING PROCEDURES back

Survey weights have not been assigned in the FES. However, the sample design is theoretically intended to be self-weighting, but in practice this is not completely realized due to deficiencies in the data available for units listed in the sampling frames (see note 28).

Note:

28. Family Expenditure Survey Handbook, 1980, p.6

F. DETERMINATION OF SURVEY UNIT MEMBERSHIP back 

A person at a household who usually has one meal a day at that household is normally regarded as a household member.

Persons who spend only part of their time in the household are considered members provided they usually spend at least four nights a week in the household. However, when a married person is a member, the husband (or wife) is usually counted as a member provided that either he usually goes home at least one night a week or or he will be staying with the household for all or most of the record-keeping period. A child under 16 at boarding school is also regarded as a member provided he spends his holidays with the household. A person staying temporarily with the household, or who has been living with the household for only a short time, is considered a member provided he will be staying with the household for at least one month from the start of records (see note 29).

The survey unit head is the head of household. The head of household must be a member of that household. He or she is the person, or the husband of the person, who:

1. owns the household accomodation, or 
2. is legally responsible for the rent of the acomodation, or 
3. has the household accomodation as an emolument or perquisite, or
4. has the household accomodation by virtue of some relationship to the owner who is not a member of the household.

When two members of different sex have equal claim, the male is taken as head of household. When two members of the same sex have equal claim, the elder is taken as head of household (see note 30).

The following relationships to the household head can be identified: wife or husband; son or daughter (including stepson/daughter; son-in-law or daughter-in-law; father or mother; father-in-law or mother-in-law; brother or sister; grandson or grand-daughter; other relative (e.g. neice, nephew, brother-in-law).

Notes:

29. Family Expenditure Survey Handbook, 1980. p. 76
30. Family Expenditure Survey 1979. 1980. HMSO. p. 153.

 G. CHILDREN AND SPOUSES back  

In this survey adults are defined as persons who have reached the age of 18, or who are married. Children are defined as persons who are under 18 years of age and unmarried.

A couple who are cohabiting and consider themselves as husband and wife are considered "married."

H. AVAILABILITY OF BASIC SOCIAL AND DEMOGRAPHIC INFORMATION back

In Table 20.1 are summarized the basic social and demographic information which is available in the FES.

Table 20.1
Basic Social and Demographic Information

 I. AVAILABILITY OF LABOUR MARKET INFORMATION back

Labor force status information is available for the FES for household heads and spouses. For both household heads and spouses the following information is available in the LIS database:

1. whether working or not working (Variables LFSHD and LFSSP)
2. whether working full-time or part-time (Variables HRSHD and HRSSP)

In addition, for both household heads and spouses information is available for:

1. 1. occupational classification (Variables D14 and D15)
2. 2. Industry classification (Variables D16 and D17). 

No groups were excluded at the time of interview through filter questions relating to the employment status of the respondent. Available labor market information is summarized in Table 20.2.

Table 20.2
Labor Market Information 

J. AVAILABILITY OF GEOGRAPHIC INFORMATION back

Two types of geographic information are available for the FES data in the LIS database:

1. D7. Geographic location indicator A
2. D20. Population density and whether a metropolitan area, and if in London.

K. SOURCES AND AMOUNTS OF CASH INCOME back    

Sources and amounts of income are recorded for a two week period. The amounts recorded were the actual amounts received. Income sources and amounts are recorded for all persons receiving income.

L. TAXES back   

Income tax information is available in the LIS database for the FES. See variable V11 in Section M for more information on taxation.

 M. BIBLIOGRAPHY OF MAIN PUBLICATIONS BASED ON THE FES 1986 back  

 Department of Employment. 1988. Family Expenditue Survey 1986 (Revised). London: HMSO. p.

 ESRC Data Archive. 1986. Family Expenditure Survey, 1986. Annotated Questionnaires. SN:2556

 ESRC Data Archive. 1986. Family Expenditure Survey, 1986. Coding Frames - Comparison of 1985 & 1986. SN:2556.

 ESRC Data Archive. 1986. Family Expenditure Survey, 1986. Derived Database Schema Document. SN:2556.

 ESRC Data Archive. 1986. Family Expenditure Survey, 1986. Interviewers Instructions. SN:2556.

 ESRC Data Archive. 1986. Family Expenditure Survey, 1986. Raw Database Schema Document. SN:2556.

 ESRC Data Archive. 1986. Family Expenditure Survey, 1986. Technical Details. SN:2556.

Copyright (c) 1998 Luxembourg Income Study all rights reserved
Send mail to Caroline de Tombeur
Last update: 30/08/00