Equitable data collection for Gypsy, Roma and Traveller communities in health services

The terms ‘race’ and ‘ethnicity’ are often used interchangeably, although they are arguably different concepts. There are no universal standards for defining the cultural aspects of ethnicity, and they tend to change over time and in cultural contexts. Differences in terminology, methods of data collection, individual perceptions of group identity, and changing demographics present challenges in determining racial and ethnic categories that are specific and acceptable to all individuals. Even self-reported identities can change over time depending on one’s affiliation with their community, perceptions of acceptance, advantages or disadvantages, and safety concerns. In essence, one’s ethnicity is not an ingrained or permanent trait, instead it is influenced by social and environmental factors over time.

There are thought to be four main reasons for collecting ethnicity data: counting to control; not counting to support social integration; counting or not counting for national hybridity; and counting to justify positive action. There are contradicting theories about whether collecting ethnicity data allows for better awareness, understanding and representation of certain groups or if it highlights differences and divisions that could exacerbate discrimination. Due to historical persecution, many Gypsy, Roma, and Traveller groups may be more likely to see collection of ethnicity data as 'counting to control' and would prefer a 'not counting to support social integration' approach.

Ethnicity data may be recorded through observational, community-based and self-identification approaches, which may conflict with one another. Although self-identification can present challenges, such as people not declaring their ethnicity honestly due to fear of discrimination, individual interpretation, and changes over time, it is still thought to be the most respectful of individuals' rights and is the recommended practice for healthcare settings.

Ethnicity categories vary across and within England, Wales, Scotland and Northern Ireland's health datasets, with varied inclusion (or lack thereof) of Gypsy, Roma, and Traveller groups. This raises questions as to the logic of including or excluding certain groups, presenting them in various combinations, and how this limits the ability to understand population-level health in these communities.

Ethnic identity is often connected to other identities such as education and socio-economic status (see Kisfaulidy, 2018) and, thus, it should be collected along with other key demographic variables to be meaningful. Contextual data has the potential to enhance our understanding of population dynamics by providing insight into the factors that shape disparities, trends, and outcomes. For instance, identifying the socio-economic status and education background of Gypsies, Roma, and Travellers may aid with addressing exiting structural inequalities.

Ethnic identity is not stable or monolithic. It needs to be recognised that individuals’ identity may shift between or span across categories over time (e.g., identity pluralisation). In other words, people’s identities can change due to life experiences, exposure to new cultures, personal growth, and shifts in societal attitudes. Individuals may identify differently in their youth compared to later in life, based on their evolving understanding of culture, heritage, or personal experiences. For instance, the importance of cultural belonging may increase for Gypsies, Roma, and Travellers as they age and improve their knowledge of and appreciation for their own ethnic background.

As populations change, cultures develop, and new social dynamics emerge, the way ethnicity is understood and categorised should reflect these transformations. Adaptive ethnic categories allow for more accurate, inclusive, and context-sensitive data collection and better understanding of human diversity. Given that ethnic identity is shifting, and changeable, ethnic categories should reflect social changes and be revisited (and revised if deemed necessary) in regular intervals. For instance, over time identities transform by both gaining new and losing old elements, which might explain why Gypsy Travellers in Scotland incorporated Scottishness into their identity and prefer to be called Scottish Gypsy Travellers.

- Pertaining to the focus of this review, the following ethnic categories should be separated when health-related ethnic data are collected: Gypsy, Roma, and Traveller1 (Please see the Gypsy, Roma & Traveller communities: equitable data collection report where we present six empirical evidence-based categories to be used across health services in the UK to replace the current classifications)

In the UK health services, a minimum of 19 ethnic categories should be deployed across the sector, in line with the latest recommendation on ethnic categories in the health service by Quattroni et al. (2024). Whilst this source was not part of the review, given the dated ethnic categories used across the UK health services, the recommendation is pertinent to note.

Whilst ethnicity data can be and has been collected for multiple reasons and Morning (2008) and Sookrajowa (2021) have detailed a four-fold typology for ethnicity data collection, it is recommended that ethnicity data should only be collected to promote positive social action and that the rationale for data collection to be clearly and effectively communicated to the population.

Standardisation of terminology of ethnic categories is essential. By having consistent and generally understood terms, we can ensure that the information gathered is accurate, comparable, and useful in addressing social disparities for promoting clarity, accuracy, and inclusivity in medical literature (Lewis et al., 2023). For instance, standardised terminology would allow for consistency in how ethnic categories are defined and used across different surveys, studies, and datasets. This ensures that data can be accurately compared across time, locations, and research projects.

Self-reporting for ethnic data collection should be used as a form of data collection whenever possible. Allowing individuals to define their own ethnic identity provides more reliable and meaningful data, while also ensuring that individuals are empowered to express themselves as they see fit.

Out of 103 responses (8 non-responses), a total of 2 (2%) people indicated that their ethnicity was often recorded when using health services, while 25 (24%) noted that their ethnicity was sometimes recorded. As 76 participants (74%) indicated that their ethnicity was never or hardly ever recorded, this information may be helpful in partially explaining the current lack of health data on Gypsy, Roma, and Traveller populations.

We asked participants if their ethnicity was listed on forms used by health services. Out of the 64 participants who responded to this question 29 (45%) said yes, 3 (5%) responded sometimes and 33 (50%) reported that their ethnicity was not listed, were never asked, or were not sure. A high number of participants (46) did not respond to this question, which may be because most participants (74%) had never been asked about their ethnicity when using health services.

We asked participants if they thought that their ethnicity should be recorded by health services. Out of 104 responses, 76 (73 %) participants noted that recording their ethnicity would be important or very important, while 8 (7.5%) deemed it somewhat important, 17 (16%) deemed it not important or not very important and 3 (3%) did not know the answer to the question. 7 people did not respond to the question.

Out of 106 responses:
o 34 (32%) of the participants identified as Roma (out of which 4 participants added Czech or Slovakian to their ethnic category)
o 22 (20%) identified as Irish Traveller or Irish or Northern Irish
o 16 (15%) identified as Welsh Gypsy or Romany Gypsy
o 14 (13%) identified as English or English Gypsy or Gypsy
o 8 (7.5%) identified as Scottish Gypsy Traveller (out of which one identified as Nacken)
o 7 (7%) identified as English Traveller
o 3 (3%) used the combined Gypsy or Irish Traveller category which is still used in the healthcare system
o 2 (2%) identified as Showman
o 5 people did not respond to the question.

It can be seen from the data that there are more ethnic categories than currently listed across UK health services and that in many instances ethnic and national identities overlap.

In terms of declaring their ethnic identity within health services, most participants reported never being asked about their ethnicity and, when asked, their ethnicity was often not listed. This omission was perceived by community members as meaning: ‘You are not known here’ and ‘You are not welcome / do not belong here’. Members in every focus group expressed some degree of reluctance to declare their ethnicity, but it was less prevalent in the three Roma groups.

Other main concerns identified by participants were:
1) the limited availability of appropriate interpreters (Roma only)
2) their own articulations of health needs were not listened to
3) the lack of understanding held by health care staff about their cultural needs
4) the mode of communication between community members and health service providers was often ineffective
5) a preference for working with trusted, local community-run charities to receive essential health information
6) discriminatory experiences of the past still impact how community members interact with people outside of their trust circle
7) general lack of understanding of the need to collect health data.