|Publication date:||17 April 2019|
|Who this is for:||Users and producers of statistics|
|Type:||Harmonisation guidance and principles|
The following guidance sets out how to collect and report statistics about demographic information to ensure statistics about this topic are as comparable as possible across the Government Statistical Service (GSS).
This section provides guidance on the survey questions to use when collecting information about demographic information.
Name or other unique identifier within the household?
[enter name/ identifier]
What is your date of birth?
if month not given, enter ‘6’ for month
if day not given, enter ’15’ for day
What was your age last birthday?
If respondents refuse to give their age, then give your best estimate.
To ask respondents if they are aged 16 and over
Are you …
- single, that is, never married and never registered in a civil partnership
- separated, but still legally married
- in a registered civil partnership
- separated, but still legally in a civil partnership
- formerly in a civil partnership which is now legally dissolved
- surviving partner from a civil partnership
- A person whose spouse/civil partner has been working away from home for over six months, for example on a contract overseas or in the armed forces, should still be coded as married and living with husband/wife or civil partner if the separation is not permanent.
To ask respondents if they are aged 16 and over, there is more than one person in the household and respondent is neither married or in a civil partnership.
May I just check, are you living with someone in this household as a couple?
Are you living with someone in this household as a couple?
- Yes – opposite sex couple
- Yes – same sex couple
- Only respondents who are living with their partner in this household should be coded as living together as a couple.
- You may code No without asking the question ONLY if all members of the household are too closely related for any to be living together in a de facto partnership relationship.
Code relationship of each household member to the others:
- Civil Partner **
- Cohabiting partner *
- Son/daughter (incl. adopted)
- Foster child
- Foster parent
- Brother/sister (incl. adopted)
- Foster brother/sister
- Other relative
- Other non-relative
- You may want to introduce this section. A possible introduction is: “There are a lot of changes taking place in the make-up of households/families and this section is to help find out what these changes are. I’d like you to tell me the relationship of each member of the household to every other member.”
- The list is detailed, but interviewers should not probe for relationships that are not volunteered or queried by respondents. The full relationships grid, showing relationship of each household member to all the others, will enable the computation of units within the household, such as family units and benefit units. Coding of such units directly by interviewers is error prone and it is difficult to correct errors later in the office
* Includes same-sex couples who are not married or in a registered civil partnership. Treat relatives of cohabiting couples in the same way as relatives of spouses.
** Treat relatives of civil partners in the same way as relatives of spouses.
- The section must be asked for all households consisting of more than one person. Please ask in every case. You should not make assumptions about any relationship.
- Treat relatives of cohabiting members of the household (both opposite and same sex) as though the cohabiting couple were married, that is, the mother of a partner is coded as mother-in-law. Other relatives include cousins, nieces, nephews, aunts and uncles.
- You should probe on this question, but be sensitive. It may be that someone described as a ‘son’ or ‘brother’ earlier is actually a stepson or half-brother. Where possible, we want to know the true relationship. If you have doubts about any relationship, record as much information as possible to allow changes to coding later if appropriate.
- Half brothers/sisters should be coded with step-brothers/sisters.
Checks of the data
Check that the ages are consistent:
- hard check: Child should be younger than their parent/guardian
- hard check: A foster-parent should be older than their foster-child
- hard check: A foster-child should be younger than their foster-parent
- hard check: A grandparent should be older than their grandchild
- hard check: A parent or grandparent should be older than 15. check that age has been recorded correctly
- soft check: A step/ in law-parent is usually older than their step/ in law-child
- soft check: A step-child/ child-in law is usually younger than their step/ in law-parent
- soft check: You have recorded this persons relationship as ‘cohabitee’ but both partners living arrangements are not stated as ‘cohabiting’
- hard check: You have recorded this persons relationship as ‘spouse’ but both partners partnership is not stated as ‘married’
- soft check: You have recorded this persons relationship as ‘civil partner’ but both partners partnership is not stated as ‘civil partnership’
- soft check: [person] is not married or cohabiting so he/she cannot have a partner in the household
- hard check: You have coded this person as married but age is under 16
- soft check: You have coded this person as civil partnered but age is under 16
- soft check: You have coded this person as cohabiting but age is under 16
- soft check: HRP may only have one (main) spouse or partner
The household reference person is:
- the member of the household in whose name the accommodation is owned or rented, or is otherwise responsible for the accommodation. In households with a sole householder that person is the household reference person
- In households with joint householders the person with the highest income is taken as the household reference person.
- If both householders have exactly the same income, the older is taken as the household reference person.
Note that this definition does not require a question about people’s actual incomes; only a question about who has the highest income.
In whose name is the accommodation owned or rented?
You have told me that [names] jointly own or rent the accommodation. Which of them has the highest income (from earnings, benefits, pensions and any other sources)?
The household reference person has been computed as:
Presentation of outputs
This section provides guidance for outputting the survey questions to use when collecting information about demographic information.
Single year of age is preferred, but if data quality does not allow this, other groupings provided may be used.
Harmonised Principle Age Band Group 1
Harmonised Principles A – D are directly comparable and may be collapsed into each other. Principles A – D can be used as standalone age groups. Principle E may be appended to Principle C and D.
|All ages||All ages||All ages||All ages|
Harmonised Principle Age Band Group 2 (only to be used for demographic outputs)
Harmonised Principle E and F are designed to be in addition to Harmonised Principle Age Bands Group 1 (A -D). Most high profile demographic outputs, including the Census, produced by the Office for National Statistics (ONS) are at single year of age but when lower levels of geography are used they can be aggregated up to quinary age bands. Principles F may be used as a standalone age group or principle E may be appended to Principle F. Principle E may be used as a standalone age grouping when presenting data on older or high age estimates or may be appended to Age Band Group 1 (C and D).
*Use additional age band when appending to Age Band Group 1.
For analyses related specifically to employment, a secondary classification may be appropriate, given below. Further detail (5 or 10 year bands) may be appropriate in some cases. Age groupings to choose will depend on:
- the sample size of the source used (narrower age bands will be more suitable for outputs from surveys with a larger sample size or large administrative sources)
- the subject of interest in the statistics. For example, employment analyses should start from age 16 in most cases.
Harmonised output categories for employment and pension related analyses:
|65 and over||XXX|
Changes to the state pension age came into effect from April 2010 and will continue progressively until 2028.
Users often require statistics for the state pension age at the reference date of any output. Producers of outputs are asked to consider their users and their possible need for outputs on state pension age basis. Given this, it is suggested that producers provide this in addition to the age groups suggested above, where it is both practical to do so, and does not present any disclosure control risks.
Age in years
We recommend that the date of birth input data is translated into age at time of survey for most published outputs. For specialist analyses other categories could also be presented. These might include analysis based on ‘academic age’ (i.e. age on the preceding August 31st), which is useful for studies concerned with education and issues relating to youth. Another example is ’rounded age’, in which a child’s age is rounded to the nearest integer (e.g. all children aged 14½ to 15½ are counted as being aged 15). Rounded age is used to calculate standard growth curves for children. Some health-related analyses may need to use this to allow comparison with published data on children’s heights and weights by age.
Survey managers may wish to consider including derived variables on these definitions on their databases.
Aggregation of age groups for publication
The level of aggregation used in published analyses will depend on the purpose of the analysis and the quality of the data, including sample size in the different age groups. For general purposes, including reference tables, the categories above are recommended, though not necessarily with full breakdown.
We propose two sets of harmonised output categories: ‘marital or civil partnership status’ and ‘living arrangements’ which will be appropriate for different purposes. The former is based solely on the legal status of the relationship and the latter combines information on the legal status of the relationship and whether the respondent is living as part of a couple or not. It is hoped that both will be made available in publications and datasets where possible, but this will not be appropriate for all tabulations. Where only one set of categories is used, and in discussion, it should be made clear in a footnote or through a clear context whether ‘marital or civil partnership status’ or ‘living arrangements’ are being referred to, as some of the category names are the same.
Output categories: Marital or civil partnership status*
|In a legally registered partnership|
|Married and living with spouse or registered in a civil partnership and living with civil partner||xxx|
|All in a legally registered partnership||XXXX|
|Not in a legally registered partnership|
|Single (never married or formed a civil partnership)||xxx|
|Divorced or formally registered in a civil partnership which has now been dissolved||xxx|
|Widowed or surviving partner from a civil partnership||xxx|
|All Not in a legally registered partnership||XXXX|
Further detail (splitting ‘married and living with spouse’, and ‘registered in a civil partnership and living with civil partner’) may be appropriate in some cases.
*Note: Information for this analysis is derived from the question on marital or civil partnership status only. The term separated covers any person whose spouse or civil partner is living elsewhere because of estrangement (whether the separation is legal or not).
Output categories: Living arrangements**
|Persons living in a couple|
|Married/ Civil Partnered||xxx|
|All persons living in a couple||XXXX|
|Persons not living in a couple|
|Divorced or formally registered in a civil partnership which has now been dissolved||xxx|
|Widowed or surviving partner from a civil partnership||xxx|
|All persons not living in a couple||XXXX|
** Note: Information for this analysis is derived from the question on marital or civil partnership status and the question on living arrangements (i.e. whether the respondent is living as part of a couple or not). The living arrangements analysis differs from the marital or civil partnership status analysis in that cohabiting takes priority over other categories. For example, if a person is divorced and cohabiting they are classed as cohabiting. Further detail (splitting ‘married’, and ‘civil partnered’) may be appropriate in some cases.
# Note: This category includes same-sex couples.
The harmonised definition of a household for the purposes of analysis and publication is identical to that of the household response unit, which is used in defining the sample and data collection.
The definition of a household is:
One person living alone;
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.
This definition of the household response unit has been used for the 2011 Census and used in social surveys conducted by the Office for National Statistics since the start of 2011. It differs only slightly from the definition used for the 1981, 1991 and 2001 Censuses which was:
One person or a group who have the accommodation as their only or main residence and (for a group of people) either share at least one meal a day or share the living accommodation.
To measure the impact of the introduction of the changed household definition, the Office for National Statistics developed a set of questions to be asked to interviewers for each address they visited. The questions were initially asked on the Opinions (OPN) survey in May and June 2010. Early analysis of the data collected highlighted issues in the comprehension of the existing questions related to housekeeping arrangements. A revised set of questions was therefore developed to be asked in the OPN questionnaire in July, August and September 2010. Overall only two addresses out of 2,188 productive cases (less than 0.1% of interviewed households) would change classification under the new household definition. As the two households change classification in the opposite direction, there was no overall change in the total number of households under the two definitions.
There may be other groupings which are required for analysis, such as family units, benefit units and domestic consumption units, but these are wholly contained within this household response unit.
As in the case of harmonised questions, where a survey needs to add information it can do so, provided that it can also identify the harmonised information (i.e. in this case, the harmonised response unit). For example, the Labour Force Survey adds students who live in halls of residence in term-time and residents in National Health Service accommodation to the coverage allowed in the harmonised definition, but these are clearly identified and the harmonised definition can be derived (details are given in Appendix A).
If a respondent has more than one address, their assessment of which is the main address is taken except in the following circumstances:
- Adult children, that is, aged 16 and over who live away from home for purposes of either work or study and come home only for holidays should not be included at their parental address.
- Anyone who has been away from the address continuously for 6 months or more should be excluded even if the respondent continues to think of it as their main residence.
- Anyone who has been living continuously at an address for 6 months or more should be included at that address even if they have their main residence elsewhere.
- Anyone who is searching for a permanent address in this country should be included at their temporary address, unless they are making a holiday or business visit only and remain resident abroad.
- Addresses used only as second homes, that is holiday homes, should never be counted as the main residence.
This page will be reviewed annually.