Public sector - understanding equality data collection: main report

This research describes and explores the range of equality and socio-economic disadvantage data collected by public sector organisations. Findings offer insights into what works best in terms of collecting, utilising and safeguarding robust data, highlighting major barriers to its collection or use.

Annex 6: Collection of information on individual protected characteristics

An overview of the questions, definitions, response categories, and variation on these, in relation to each individual protected characteristic is presented below.


  • The collection of age information was seen as relatively uncontentious by interviewees, with some saying that people were very used to providing this information in different contexts.
  • Information was gathered in four different ways by asking for either (i) age (ii) age group (iii) date of birth or (iv) year of birth.
  • Most commonly individuals were asked to provide their date of birth, thus offering an accurate and dynamic indicator of age. The second most common way of gathering age information was by asking people to indicate an age range to which they belonged. There was variation in the age range options offered between and within organisations. However, it was often clear that the options offered were linked to the focus of the relevant service - for example, in the case of age-related eligibility for services.

Religion and belief

  • Most often questions asked about 'religion', with no reference to 'belief', but in a few cases, questions asked about 'religion or belief'. One organisation asked two separate questions, one on religion and one on belief.
  • Questions were often similar to that used in the 2011 census question - in terms of both question wording and response options - although slight variations were also common (e.g., 'Christian -Protestant' instead of 'Christian - Church of Scotland').
  • Questions typically asked 'What religion, religious denomination or body do you belong to?' Other forms of question included: 'What best describes your religion or belief?', 'What is your religion?', and 'Which group do you most identify with?
  • Response categories typically included: Buddhist, Christian - Church of Scotland, Christian - Roman Catholic, Christian - Other, Hindu, Jewish, Muslim, Sikh, Prefer not to say. Additional / alternative Christian denominations were offered in some cases. There was also occasional inclusion of 'Pagan', 'Spiritual', and 'Bahai'.
  • There were a number of different variations of 'no religion' offered by organisations, including 'none', 'no religion', 'non-religious (atheist, humanist etc)', and 'atheist, agnostic or no religious affiliation'.
  • The number of response categories varied - one organisation offered six response categories, while a small number of organisations offered very extensive lists of categories.


  • There was variation in the questions used and response options offered both between and within organisations. However, questions often appear to be modelled on the 2011 census question which uses response options structured around six main categories and sub-categories within those (although with wide range of variations also apparent in the detail of the questions).).
  • Where information on questions / field names was provided, these asked about 'ethnicity' rather than 'race' - as is the case with the 2011 census question.
  • There was variation in terms of phrasing and terminology, with examples including the following: 'What is your ethnic group?' (i.e. the question used in the census), How would you describe your ethnicity?', 'Which of these groups do you consider you belong to?', 'What best describes your ethnic group?', 'Which of the following best describes your ethnic group or background?'
  • The following (different) terminologies were all used: ethnicity, ethnic group / sub-group, ethnic background, ethnic origin, national identity.
  • There was a mix of approaches in term of using a single 'layer' of response options, or using main categories and sub-categories.
  • While there was a lot of commonality in the response options offered, there was also a great deal of (often slight) variation. The number of options ranged widely (e.g. six options used in one case; over 70 used in another case). There was also variation in the use of 'please specify' options, and the wording of 'prefer not to say' options.
  • There was some limited use of multi-part questions - e.g., asking about ethnic origin and national identity.


  • The information gathered on disability was wide ranging and often reflected the needs and practicalities of different services - i.e., the information was often used to ensure that the needs of individual service users could be properly met. As such the question and response options were also very varied with some organisations simply aiming to determine if an individual had a disability or long-term condition, and others gathering very detailed information on the nature of a person's physical and mental health and the impact on their daily functioning.
  • Two-part questions were common, with individuals first asked if they had a disability - with some either directly referencing or alluding to the definition in the Equality Act - and then asked for further details of any disability or long-term condition, either by using a list of tick-box option or inviting a free-text response.
  • Those using a single part question included those asking a simple yes / no / prefer not to say question as to whether somebody has a disability, and others offered a list of different disabilities / conditions for individuals to respond to.
  • There was also some limited use of separate observed and self-defined ethnicity.


  • In all the cases where a specific question was provided to the research team, the term 'gender' rather than 'sex' was used. For example, questions asked: 'What is your gender?', 'What is your gender identity?', 'What best describes your gender?'
  • Interviewees were often aware of this as an issue, with one specifically referring to sex and gender being 'conflated' in their collection of equality data.
  • In more than a third of cases, respondents were (as with the 2011 census question) offered a binary response choice of Male / Female, with or without a 'prefer not to say' option. Again, there was widespread awareness that this was not appropriate, or did not reflect current thinking on this issue.
  • Where additional response options were offered, these commonly included: 'Other' (with or without the option to specify) and 'Non-binary'. Other less common response options included: 'man / woman', 'intersex', 'gender fluid', 'other gender identity', 'I identify with another term', 'In another way', 'I prefer to self-describe', 'unknown', 'indeterminate'.
  • In three case, response options specified the inclusion of trans individuals: 'Male (including trans man) / Female (including trans woman)', or 'Male (including transgender) / Female (including transgender)'; in another case there was an additional question asking if the respondent had ever identified as transgender.
  • There was a small number of examples of organisations using 'title' (Ms, Mr etc) as a proxy for gender.

Sexual orientation

  • Questions on sexual orientation were formulated in a variety of ways, but most commonly asked people how they would 'describe' their sexual orientation: for example: which of the following best describes your sexual orientation / how you think of yourself / how you think of your sexuality. Questions generally used the term 'sexual orientation', although there was occasional use of the word 'sexuality'.
  • The response options commonly included the categories of heterosexual / straight, gay man, gay woman (or a combined category of gay man / gay woman), bisexual, other. There was some variation in the wording used, most notably in whether the term 'lesbian' was included and whether 'straight' was included alongside 'heterosexual'. Occasionally, additional options (e.g., 'asexual', 'in another way', 'not sure') were offered.
  • There was one example of an organisation using a different style of descriptor for the response option offered as follows: 'Please describe your sexual orientation: Towards people of a different sex (straight) / Towards women of the same sex (lesbian) / Towards men of the same sex (gay) / Towards people of both sexes (bisexual) / Other (e.g., asexual).'
  • Some response lists included an option for individuals to self-define or specify further if they selected 'other'.

Pregnancy and maternity

  • Information on pregnancy and maternity was collected in a few cases only.
  • The information gathered in the data collections studied mainly focused on 'pregnancy', rather than 'maternity' (defined in the Equality Act as the 26 weeks following birth). Additionally, this information was sometimes collected on a household rather than individual basis ('Is anyone in the household pregnant at the moment?'), reflecting the information needs of the service being provided.
  • In a number of cases, interviewees said that while information on pregnancy / maternity was not systematically recorded in a format suitable for quantitative analysis, it would be included in case notes in a narrative form where this was relevant, or (in a few cases) they said that a separate record may be created for an unborn child.

Marriage and civil partnership

  • Information on marriage and civil partnership was only collected in a handful of cases.
  • Questions varied in terms of whether they asked about an individual's 'legal status' or asked how an individual 'thinks about themselves'. Some questions referred to marital status only while others referred to marital or civil partnership status. However, all offered civil partnership response options.
  • Response options offered ranged from three categories (married / civil partnership / other, plus 'prefer not to say' in some cases) to lengthier lists including single, co-habiting, separated, divorced, widowed and the civil partnership equivalents.

Gender reassignment

  • Only in a small number of cases did the data collections identified by the research gather information about gender reassignment.
  • Generally speaking, questions asked about 'identity', asking if individuals identified (or had ever identified) as trans / transgender / a transgender person. One question asked if the individual had undergone or was undergoing gender reassignment, and two questions asked if an individual's gender identity was the same as the gender they were assigned at birth, or if their current gender (sex) was the same as they were born with.
  • Although questions commonly used the term 'gender', both 'sex' and 'gender' were used in a couple of cases.
  • A two-part question was used in one case ('Is your current gender (sex) the same as you were born with?' 'Do you identify as trans?').

Socio-economic disadvantage

  • In almost all cases, organisations gathered postcode information. However, this was generally collected for administrative reasons and was not routinely used as an indicator of socio-economic disadvantage.
  • Where organisations gathered information on income - and this was only reported in 13 cases - this was always because there was a specific need for this information related to service eligibility.



Back to top