SUBMISSION TO WOMEN & EQUALITIES SELECT COMMITTEE INQUIRY INTO RESULTS OF THE RACE DISPARITY AUDIT

WRITTEN EVIDENCE SUBMITTED BY ROMA SUPPORT GROUP Monday 27 November 2017

Executive summary Roma Support Group (RSG) is the longest established voluntary organisation in the UK supporting and representing the interests of the Roma diaspora from Central & East Europe who have made their homes in the UK, with particular focus on London. The evidence submitted by RSG on the Race Disparity Audit: • • •

• •

•

•

Questions the value of the Audit in providing evidence about relations between public services and Roma communities. Finds that very few of the published datasets contain any information about Roma. Concludes that the few datasets that contain information are conceptually flawed as Roma are combined with ‘Gypsy’, which creates a problematic and unsound category – for both migrant Roma and English Gypsy communities. Finds that in some sections, the narrative departs from accepted statistical principles (e.g. varies the baseline for understanding trends). Finds that the emphasis on differences between ethnically described populations, has hidden differences and trends within particular ethnic populations (e.g. gender, age, locality). Outlines that a major contextual issue is that since 2014, the circumstances of many Roma families and communities has been determined as much by their national origin and migrant status, as by their ethnicity. This is underwritten by the changes in relation to access to social security provision, for families and individuals. Concludes that the UK government should adopt a national Roma integration strategy, to ensure that the needs and aspirations of Roma and Gypsy and Traveller communities are met.

How useful is the data? 1. For migrant Roma communities in the UK, this audit has (very) minimal significance. 2. There are 120 datasets on the Ethnicity: Facts and Figures website. 3. There are 98 datasets, excluding the Education, Skills, Training domain. None contain any information about Roma as users and/or beneficiaries of public services.

4. Of the 22 Education, Skills, Training datasets, 3 failed to include any reference to Roma. They utilise the “five standardised ethnic groups based on the 2011 census”. These three datasets relate to apprenticeships, further and higher education. • White; • Mixed/Multiple ethnic groups; • Asian/Asian British; • Black/African/Caribbean/Black British; • and Other ethnic group 5. Of the remaining 19 Education, Skills, Training datasets, all included Roma (as pupils or children or young people or students), but only in conjunction with Gypsy children/young people. And within each of these 19 datasets, there is also the category of ‘Irish Traveller’. 6. To reinforce this, the only references to ‘Roma’ in the Race Disparity Audit report are located in the section on Education. 7. In relation to Roma women, men and children living in the UK, these datasets do not challenge their invisibility in UK government policy.

Where and what are the most significant gaps? 8. The Department for Education now collects pupil data separately for Gypsy and for Roma children and young people. This categorisation was in use from January 2016, and therefore includes the 2015-16 and 2016-17 school censuses. The Department provides the following guidance on definition: 9. ‘Gypsy’ is an ethnic category, “This category enables the separate identification of Gypsy pupils. Gypsy refers to: all pupils who identify themselves as Gypsies. This includes all children of a Gypsy ethnic background, irrespective of whether they are nomadic, semi-nomadic or living in static accommodation”1. 10. ‘Roma’ as an ethnic category, “This category identifies the separate identification of Roma pupils. Roma refers to: all pupils who identify themselves as Roma or Romany, part of a diverse community of related groups whose ancestors are believed to originate from the Indian subcontinent but who have more recently migrated from Central and Eastern Europe. Many Roma speak a form of dialect of the Romani language as their first language and for many this is in addition to their national language (e.g. Czech or Romanian).”2 11. This data should now be published separately (for ‘Gypsy’ and ‘Roma’) for the whole range of datasets for children and young people of school age relating to attainment, attendance and inclusion.

1

School census 2016 to 2017 Guide, version 1.6, Department for Education, (April 2017)

2

Op cit

12. The absence of any information for Roma, Gypsy & Travellers in the datasets of Work, Pay, Benefits is a major omission. It is a widespread view within migrant Roma communities that both (a) rates of economic activity are high, and (b) rates of accessing/claiming state benefits are low – particularly since 2014 with the government policies to reduce access to social security provision for EU migrants. This view is fundamentally different to the government’s constant reference to ‘welfare tourism’ and other distortions. The absence of any data around economic activity rates, and claiming benefit is a serious gap. 13. There is no information on access to, provision of, and impact of Early Years provision. 14. There is no information on access to, progression within, and impact of Higher Education.

Some of the issues that arise from the datasets 15. Many Gypsy & Traveller NGOs welcomed the 2011 census as it enabled people to define as Gypsy or Irish Traveller. The ethnic categorisation based on the 2011 census underpins much of the Education, Skills, Training datasets. As we have said, this also includes Roma, but within a Gypsy/Roma category. However, there are a number of datasets which are based on post-2011 service user profiles which unnecessarily use the 2001 census categories – and so exclude Gypsy/Roma and Irish Traveller ethnic categories. There is therefore no justification for this data to exclude Gypsy and Traveller categories. 16. Relevant datasets include: • Crime, justice and the law; policing – stop and search (Home Office, 2006/07 – 2015/16) • Health; access to treatment - detentions under the Mental Health Act (NHS Digital, 2016/17) • Health; preventing illness – overweight children (Public Health England, 2014/15 – 2015/16) 17. Whilst it is accepted that the data for stop and search predates the 2011 census, the justification for not using the 2011 census categories is disingenuous; This is because these ethnic categories are easier to compare with the categories used when recording stop and search. 18. There is a completely unprincipled approach to using the datasets to suggest trends that are not justified. And an apparent avoidance of describing some of the negative conclusions from the data. For example, in destinations for pupils after KS4, the narrative following a time series chart says: in 2014/15, a higher percentage of pupils from every ethnic group went into education, employment or training, compared to 2010/11 – the total percentage of pupils going into education, employment or training went up by 5 percentage points over this period (from 89% to 94%) – our emphasis Gypsy/Roma and Traveller of Irish Heritage pupils saw the largest percentage point increases in pupils going into education, employment or training – there was a 17 percentage point increase for Traveller of Irish Heritage pupils between 2011/12 and

2014/15 (from 50% to 67%), and an 8 percentage point increase for Gypsy/Roma pupils in the same period (from 60% to 68%) – our emphasis. 19. The time series was from 2010/11 – 2014/15. Each bullet point in the narrative section reports the changes from 2010/11 – 2014/15; except (as above) in the section on Gypsy/Roma and Irish Traveller – where the base comparator year is 2011/12. If the base comparator year is 2010/11, the percentage increase for Irish Traveller pupils going into education, employment, training would have been 8% (59% to 67%); and for Gypsy/Roma pupils an insignificant 1% (67% to 68%). For Gypsy/Roma pupils, this is actually the equal smallest percentage increase for all ethnic categories – not the largest! 20. The narrative does report the low level of Gypsy/Roma pupils going into education, employment or training. However, it does not report that the same table shows that 32% of Gypsy/Roma pupils either went into ‘no sustained education or employment’ or their destination was ‘not known’.

Explanations for disparities and the historical context 21. The absence of any clear definitions of ethnicity is sharply revealed in the Race Disparity Audit. This is highly relevant in the case of (migrant) Roma. 22. The ONS reported on the decisive change in 2011 that, for the first time in the UK, allowed people to ascribe themselves as Gypsy or Irish Traveller; the ONS said: “The ethnic group question on the 2011 Census was based on the respondents own perceived ethnic group and cultural background. The 2011 Census figures for Gypsy or Irish Traveller are solely based on respondents who chose to identify with that ethnic group either through the tick box or writing Gypsy or Travellers in the ‘any other White’ background. The tick box referred to ‘Gypsy and Irish Traveller’ and not other groups such as ‘Roma’ as they have different geographical and territorial associations.”3 (our emphasis). 23. The only datasets that describe the experience of Roma are those in the Education domain, based upon the school census. The datasets based on the 2011 census (e.g. Housing domain, and some in Health domain) exclude Roma. 24. Our experience is that when Roma are required by an agency or service provider to describe themselves, they will do so by referring to their national origins. However, where Roma have a stronger relationship with the agency or service (e.g. with a primary school), they will be much more likely to ascribe as Roma. 25. In comparison to Gypsy & Irish Travellers, access to and the use of public services by Roma is increasingly determined by their nationality, and their status as EU migrants4. Apart from the years between 2011 (the end of the Workers Registration Scheme which had governed 3

2011 Census analysis: What does the 2011 Census tell us about the characteristics of Gypsy or Irish travellers in England and Wales, ONS, January 2014 4

Further curbs to migrants access to benefits announced, DWP press release, 8 April 2014 https://www.gov.uk/government/news/further-curbs-to-migrant-access-to-benefits-announced (accessed 26 Nov 2017)

employment status since EU A8 accession states in 2004) and 2014 (with the range of changes to access to benefits for EU migrants), this has underpinned Roma peoples relationship to work and social security. For Roma from Romania and Bulgaria (so-called A2 countries), their national and migrant status has always determined their relationship to the labour market and benefit entitlement.

What the data tells us about the most pressing issues 26. The Race Disparity Audit says (#1.12) that It is...often the case that the Audit is observing differences between ethnic groups 27. This can lead to overlooking some issues which the datasets produce. See below a table constructed from the variety of datasets in Education – 5 to7 year olds: Proportion of Gypsy/Roma children meeting expected standards at key stage 1 (adapted from Education: 5 to 7 year old tables)

Phonics Writing Reading Mathematics Science

Children receiving free school meals All Boys Girls 37 31 44 22 15 28 28 20 35 30 26 34 43 38 47

Children not receiving free school meals All Boys Girls 37 33 40 20 15 25 25 22 28 29 29 29 38 37 39

28. The Audit states (at #4.7) that Children eligible for FSM have lower attainment than non-FSM pupils in all ethnic groups.... 29. Clearly, the data at key stage 1 shows that this is not the case for Gypsy/Roma children; and especially for Gypsy/Roma girls – where those receiving free school meals out-perform the more ‘advantaged’ (?) girls. In the case of Science, this is clear 8% point difference, and in reading, a 7% point difference. The explanations for this are unclear. Eligibility for FSM for migrant Roma children was certainly massively reduced after 2014 with the restrictions on benefit eligibility for EU migrants – and so having their eligibility for FSM reduced5. 30. Within the Education domain only, there is extensive information about what differences exist at a local authority level. However, this data has been arranged so that, for example, Gypsy/Roma and Irish Traveller pupils attainment etc are not visible. At a national level, the datasets reveal a massive over-representation of Gypsy/Roma and Irish Traveller pupils in school exclusions, and in school attendance. 5

Free School Meals and Roma children; a recent review of the evidence from twenty local authorities in England, Roma Support Group, June 2016 https://drive.google.com/file/d/0B2lw1_Krq5gnS3pqSGRnb0xRYjQ/view (accessed 25 Nov 2017)

31. The Roma Support Group recent report on school exclusions6 – using published Department for Education data – shows extensive variations between cities and authorities in the way that Gypsy/Roma pupils experience school exclusions. The datasets could in future develop a more comprehensive way of documenting the experience at a local level. While this is clearly the case in Education (with the school census), it could also be replicated in Health, Housing and Policing domains. 32. This is particularly relevant for Roma communities who have historically been located in particular towns and cities in the UK. 33. In terms of future follow up work, we would commend the Department for Education. We understand that the Department for Education is about to re-establish7 the GRT stakeholder group, with an initial task to explore the greater use of school exclusions for Gypsy/Roma and Irish Traveller pupils.

How should the government respond? 34. As a continuing and present member of the European Union, the UK government has signed up to the European Council declaration signed in 2011 to ensure “greater Roma integration”. Each member state was required to produce either a national Roma integration strategy, or (the approach of the UK government), “...integrated sets of policy measures within their broader social inclusion policies for improving the situation of Roma (#14) 8. 35. The Council further declared that such policies be monitored and evaluated: “Appropriately monitor and evaluate the effectiveness of their national strategies or integrated sets of policy measures within their broader social inclusion policies. This could be done by means of measures such as setting baselines or measurable targets or by collecting relevant qualitative or quantitative data on the social and economic effects of such strategies or measures, in line with applicable national and Union law, particularly regarding the protection of personal data” (#3.4)9 36. The latest European Commission assessment of the UK government progress in meeting the ambitions of Roma integration states:

6

Fulfilling their potential? Exclusion of Roma pupils in the English educational system, Roma Support Group, October 2017 http://romasupportgroup.org.uk/?p=4994 (accessed 26 Nov 2017) 7

The department’s GRT stakeholder group was ‘paused’ in March 2015

8

COUNCIL RECOMMENDATION of 9 December 2013 on effective Roma integration measures in the Member States; European Council http://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32013H1224(01)&from=en (accessed 25 Nov 2017) 9

op cit

The mainstream approaches have not demonstrated sufficient impact on improving the situation of Roma.10 37. The assessment says that “no measures were reported” on the monitoring of the UK government policies, and recommended that “evidence gathering should also be developed to enable the assessment of the impact of the measures”. 38. We believe that without a dedicated and specific National Roma Integration Strategy, no progress will be made to meet the ambitions that the UK government has signed up to. 39. An essential component will be to establish regular and comprehensive consultation structures with Roma and Gypsy and Traveller communities, with an explicit funding strategy to support GRT NGOs to be able to make such consultations effective. 40. Finally, we see that Roma and Gypsy and Traveller communities are probably the sections of the UK population most exposed to the impact of national austerity policies which have been pursued by governments since 2010, and which are set to continue for a further five years or more.

Andy Shallice & Sylvia Ingmire Roma Support Group

10

EC NRIS assessment 2016 http://ec.europa.eu/justice/discrimination/files/roma-report-2016_en.pdf