Introduction

Romani minorities belong to Europe's most visible minorities (Szalai and Schiff, 2014). In legal discourses, the concept of ‘visible minorities’ was associated with non-white migrants in settler states. For example, the Canadian Employment Equity Act (1995) defined members of visible minorities as ‘persons, other than Aboriginal peoples, who are non-Caucasian in race or non-white in colour’. Yet Romani minorities have been visible not only as migrants, but also as traditional minorities in their countries of citizenship (Taylor et al., 2018). Recently there has been a growing consensus among scholars that it is predominantly practices and discourses of racialisation that make Roma visible as a minority throughout the European public space (Yuval-Davis et al., 2017; McGarry, 2017; Kóczé and Rövid, 2017; Yildiz and De Genova, 2018): the novel form of racialisation is connected to ascribing fixed cultural characteristics to Roma, which are seemingly incompatible with liberal democratic states. As these scholars have shown, whilst racialisation constructs fixed boundaries around ascribed cultural rather than biological differences (Gilroy, 1987; Balibar and Wallerstein, 1991), it also intertwines concepts like mobility and nomadism (Sigona, 2003; van Baar, 2015) and poverty and benefits abuse (Geddes and Hadj-Abdou, 2016) as interchangeable characteristics of Roma. The merging of such concepts as interchangeable is locally specific and can be different from context to context, yet it serves as a justification for the civic marginalisation of Romani minorities.

This chapter seeks to approach the civic marginalisation of Roma in Europe from a different perspective: it argues that Roma have been a visible minority but invisible as citizens. The EU and its Member States have manifestly constructed (supra)national citizenship on the foundation of fundamental rights and minority protection. Nevertheless, the chapter questions why such a construction still leaves many Roma, who are clearly in need of having their fundamental and minority rights protected (Pogány, 2004), on the fringes of citizenship: can this be attributed to the uniqueness of Roma as a minority?

On the basis of the chapter's analysis, I will show that whilst public discourse might represent a failure to protect Romani minorities as a unique (East) European challenge, the types of civic marginalisation processes that Roma face are not exceptional. Liberal democracies around the globe have employed similar processes with other minority populations and placed them on the fringes of citizenship too. The chapter will focus on three aspects of civil marginalisation. First, it examines the politics around minority naming and counting. Second, it explores the development of targeted ‘multicultural’ approaches in legislation and policy for minority recognition and how they have contributed to marginalisation rather than protection. Third, it analyses how these multicultural approaches coexist with the breaches of fundamental rights marginalised minorities have experienced as citizens.

A human rights litmus test for East European regimes?

After the fall of the Berlin Wall in 1989, human rights advocates and minority activists dubbed the position of Roma the litmus test for human rights and respect for diversity during the process of democratisation of postsocialist countries (Stewart, 2002). ¹ Today, even when human rights – including the rights of minorities, pluralism and tolerance – are enshrined in Article 2 of the Lisbon Treaty as part of the EU's core values and when most of the postsocialist states in Europe have become EU Member States, the position of Roma has become not a litmus test but rather one of the greatest stains on these core values in the EU as a whole. It is not only the ‘new’ postsocialist EU Member States that have failed to hold up core EU values for all, but also the founding Member States, such as France, which have not achieved these standards in their treatment of Romani minorities (Parker, 2012; Parker and Toke, 2013; Ram, 2014a). In today's Europe, it appears that the original 1993 Copenhagen criteria for accession have not achieved their goal. Although postsocialist states did implement requested legislation, such moves did not necessarily improve human rights in those states (Blitz, 2013). Moreover, evaluations show that subsequent actions, such as those under the European Social Fund 2007–13 (Kostka, 2018) and the EU's soft law approach such as the EU NRIS Framework, have not improved the position of Romani minorities significantly (Carrera et al., 2019) in either the founding or newer EU Member States.

EU policies have contributed to the visibility of Roma as a minority in the Member States and candidate countries (Vermeersch, 2012). However, such visibility has not always been positive: it has also made Roma more visible as a target for hostile politicians and media with antigypsyist stances, who have reinterpreted Roma as the minority privileged by the EU and international organisations (Stewart, 2012; Vermeersch, 2012; McGarry and Agarin, 2014; van Baar and Vermeersch, 2018). A number of politicians have echoed deeply rooted antigypsyist sentiments, present in the majority population, that equate Roma with beggars, thieves, nomads or misfits who are not compatible with the mainstream social order (Okely, 1994; Willems, 1997; Van de Port, 1998, Lucassen et al., 1998).

The increase in the visibility of Roma as a minority was not followed by a discussion as to why certain parts of citizenry do not enjoy equal rights. Debates on citizenship frequently highlighted the position of Romani minorities, but mainly in terms of citizenship exceptionalities far removed from the experience of ‘ordinary citizens’. The position of Roma was illuminated as a difficult case of minority integration rather than in terms of equality among fellow citizens.

This chapter examines how even the benevolent policies for integration of Roma can have adverse effects when policymakers do not critically reflect upon their own assumptions about Roma as well as assumptions about all-inclusive citizenship. It analyses commonsense assumptions on citizenship, minority protection and the way Roma are expected to integrate into society as minority citizens. Previous studies and more recent policy analyses claim that antigypsyism positions Roma at the margins of society and citizenship (Alliance Against Antigypsyism, 2017; Cortés Gómez and End, 2019). Here I also look at another angle: the perceived multiple normalities of what a citizen should be and how Roma are excluded from these normalities even when they are ostensibly included. I argue that European states have introduced multiple normalities through their own visions of multicultural (Kymlicka, 1995) or differentiated (Young, 1989) citizenship which include group-differentiated rights and some limited political representation of minorities. Yet all these states, separately, have deemed Roma incompatible with such visions.

I use a similar approach to Morag Goodwin and Roosmarijn Buijs (2013) as well as Tina Magazzini (2017), who have looked at how policies are made for Roma without taking Romani perspectives into account. Romani representatives have usually been included in the consultation processes but have not had a veto power over the position of the community they represent (McGarry and Agarin, 2014). To deconstruct the normalities of citizenship and what the position of Roma conveys about these normalities, I embrace ‘critical whiteness’ as it has been transposed into critical Romani studies by Violeta Vajda (2015, see also Silverman, 2018; Shmidt and Jaworsky, 2020). Vajda suggests that non-Romani scholars should reflect on their positionality when conducting their research on Roma. I also use Gloria Wekker's ideas on ‘white innocence’ (Wekker, 2016) and Gurminder Bhambra's line on ‘methodological whiteness’ (Bhambra, 2017a; 2017b). Wekker shows how the Dutch state and society employed racist discourses and practices even though the state was portrayed as benevolent towards immigrants and its former colonial subjects. Catherine Baker has suggested using this approach when discussing the concept of race in the former Yugoslav region (Baker, 2018). Bhambra (2017b) notes:

This chapter shows that the ‘dominance of “whiteness”’ can be found in discussions of legislation and policies that are either inclusive of all citizens or specifically benevolent towards Roma but, in fact, create invisible edges of citizenship that construct unbridgeable difference and exceptionalism.

The first dilemma that policymakers and scholars face when addressing the position of Romani minorities in any field is the dilemma of naming and of numbers. First, ‘Roma’ is an umbrella term for diverse populations across Europe: not all of them identify as Roma, and the term itself can unify in a common fight against discrimination, but also can be used to flatten differences among these populations and specific challenges they face in different contexts (Carrera et al., 2019: 29). Second, according to the EU data, Roma are the largest ethnic minority, numbering around 10 to 12 million people in Europe (European Commission, 2018c). Nevertheless, in each individual country in Europe Romani minorities represent a very small proportion of the whole population according to national population censuses. Most of the relevant academic literature and some policy papers acknowledge that many Romani individuals do not identify as Roma in population censuses (Messing, 2014). The underrepresentation of Roma in population censuses shows, on the one hand, that Roma want to avoid being constructed as a visible minority and hence an easy target for extremist groups and, on the other, that these scarcely reliable data are still the basis for certain legislation and policies or the lack thereof (Open Society Foundation, 2010; European Commission, 2018b). The first part of this chapter explores this situation in more detail.

In the second part of the chapter, I examine the laws and policies introduced to protect Roma by analysing primary sources, such as the minority and citizenship acts, country reports and opinions of the CoE's Advisory Committee of the Framework Convention for the Protection of National Minorities (FCNM). I focus particularly on the last cycles of monitoring in each country before 2019 (usually the fourth cycle). In the 1990s, the initial approach that international organisations, such as the CoE as well as the European Commission, recommended that states should take towards Roma was minority protection. The Roma Decade and the NRIS shifted the focus back onto socio-economic integration. At state level, some countries, such as Italy, took a more culturalist approach towards Roma, whilst others, such as Spain, leaned more towards the socio-economic (Magazzini, 2017). Yet both approaches cemented the view of Roma as an exceptional minority. Is the position of Roma unique? I argue that the focus on ‘Roma exceptionalism’, even when benevolent, has concealed the responsibility of the states and how they have contributed to the placing of Roma on the margins of society and the fringes of citizenship. Such a view is possible only when one ignores minorities around the globe who have faced similar treatments by the states of which they were citizens. Whilst Romani as well as African American civil society activists and scholars (see Jovanović and Daragiu, 2010; Brooks, 2018; Rucker Chang, 2018; Matache and West, 2018; Chang and Rucker-Chang, 2020) have been arguing for decades that there are similarities between different marginalised minorities around the globe, it is only in recent years that some policymakers have become more open to such comparisons. The question, though, remains: what can be set as the common ground for such comparison, given that both the contexts and the histories of marginalised minorities around the globe have their own distinct specificities? The chapter considers the position of Aboriginal people in Australia, New Zealand Māori and Native Americans as citizens to show how states use practices towards Indigenous people similar to those used towards Roma in Europe. The positions of Indigenous people and African Americans are indeed grounded in different contexts.

What is in a number, what is in a name?

The European Commission estimates that 10 to 12 million Roma live on the European continent and around 6 million of them are EU citizens (European Commission, 2018a). However, how are these estimates made? Who counts as Roma, and who counts Roma (Surdu, 2017)? What impacts do external categorisation and self-identification have on Romani minorities in different contexts?

These questions form the foundation of any research examining the position of Roma. As McGarry (2017: 16) comments, ‘[e]very academic article or book must explain very early on the appropriate nomenclature to refer to Roma. The matter has diverted attention away from the myriad complex puzzles relating to the marginalisation of Roma communities.’ Still, the debates on how Roma should be named and counted have real consequences for policymaking (Messing, 2014) and can either diminish or deepen the stigma around Romani identity.

The designation ‘gypsy’ was based on the misconception that Roma are a nomadic group coming from Egypt (Tremlett, 2013), whilst the etymology of the word cigan (also cigany, cikan and other similar variations in different contexts) is less clear (Liégeois, 2007: 17). On the basis of a more recent analysis, linguists have shown that the Romani language is strongly connected to Indian languages and have traced the origin of Roma to India (Matras, 2004). In April 1971, Romani activists from around Europe held a congress in London to challenge the name ‘gypsy’, since it had been given to them by outsiders and had derogatory connotations, choosing instead the name Roma, which in Romanes means a human being (Nirenberg, 2009). There have been disputes around the name Roma also within groups that have previously been known as Gypsies: some wanted to keep the old name, and others claimed that the name Roma privileges certain subgroups and neglects the heterogeneity of subgroups, for example, like the dispute between Roma, Ashkali and Egyptians in post-Yugoslav Kosovo (Lichnofsky, 2013). Yet there has been an agreement by many Romani activists (Costache, 2018) that a certain degree of ‘strategic essentialism’ (Spivak, 1988) is needed to counter discrimination and the denial of European belonging faced by all these groups despite the fact that they have been present in Europe since the Byzantine Empire at least (Crowe, 2007).

Following my previous work (Sardelić, 2015), I use the term ‘Romani minorities’ when emphasising the multitude of heterogenous and hybrid identities of minorities around Europe, including those who do not necessarily identify as Roma. For example, here I include minorities who identify themselves as Ashkali and Egyptians from South-East Europe, Manush from France, Gitanos from Spain and Sinti in Germany and elsewhere, among others. Whilst acknowledging the hybridity and heterogeneity of Romani identities (Tremlett, 2014) in different contexts, I also recognise that individuals who might ascribe different identities to themselves have been similarly subjected to antigypsyism. I take the view that antigypsyism is connected to civic marginalisation: this goes beyond the ethnic or socio-economic marginalisation usually highlighted by scholars (Ladányi and Szelényi, 2006) and policymakers. When using the term ‘Romani minorities’, I underline the varieties, but also ambiguities, of self-identification and external categorisation. By using the term ‘Roma’, I am highlighting two issues, the first of which is political mobilisation based on a specific identity (Vermeersch, 2006). Anna Mirga-Kruszelnicka (2018), following Stuart Hall's (1996) notion of new ethnicities, has referred to Roma so as to avoid essentialist understandings and also to confirm that what it means to be Roma is continuously being remade by individuals who identify as such. Secondly, this word is used by both international institutions and state authorities that have accepted it as an official umbrella term for Romani minorities in their documents.

The European Commission has identified Roma as the largest ethnic minority on the European continent on the basis of the estimate that there are up to 12 million Roma living in Europe as a whole (European Commission, 2018a). However, in the official state censuses, the number of individuals identifying as Roma rarely exceeds 5 per cent of the population. In Romania, for example, according to the 2011 census, 621,573 residents declared themselves to be Romani, corresponding to 3.3 per cent of the population. ² Yet according to the CoE estimates, there are 1.85 million Roma living in Romania, just over 8 per cent of the population. Similarly, in Slovakia, the 2011 population census reports that 105,738 individuals self-identified as Roma (European Roma Rights Centre (ERRC), 2012: 7), but EU data stated that there were approximately 500,000 Roma in Slovakia (9 per cent of the total population). ³ Similar high discrepancies occur in Bulgaria, where 335,343 residents (4.4 per cent) identified as Roma but the CoE estimated a total of 750,000 Roma (10 per cent of the total population), and in Hungary, where in 2011 315,000 identified as Roma (Messing, 2014: 814) but estimates go as high as 700,000 (or 7 per cent of the total population).

Both underestimation and overestimation of the number of Roma have very real effects when translated into policy, especially given the drive for policymaking to be based on hard, quantifiable data. Already in 2010, the Open Society Foundation (2010) published a mid-term evaluation report on the Roma Decade which highlighted that ambivalent data on how many Roma live in different countries had caused further difficulties with the gathering of data on the position of Roma in different policy fields (such as the Decade's key areas of housing, education, healthcare access and employment). Similarly, the report on the evaluation of the EU NRIS up to 2020 stated ‘that data collection, monitoring and reporting systems should be strengthened. Lack of reliable data disaggregated by ethnicity and lack of transparency and accountability mechanisms are key challenges that make the process of monitoring difficult and unreliable. This makes it difficult to measure progress’ (European Commission, 2018b: 11).

Yet the drive for quantifiable data can produce material (Messing, 2019) that can then be used in less than positive ways. An illustrative example is a research project undertaken by the University of Salford in the UK which, on the basis of surveys sent to local authorities, published estimated figures stating that there were 197,705 migrant Roma in the UK (see Brown et al., 2014). Previous estimates had suggested that there were not more than 6,000 Roma from Romania in the UK (Cahn and Guild, 2010). Messing (2019: 24) took issue with the project and its published findings:

Leggio (2019: 82) noted that ‘during the interview that accompanied the report's release, the lead author, Philip Brown, argued that local authorities were struggling to cope with such numbers due to the cuts implemented by Cameron's government. He added that knowing the actual number of Roma migrants was needed for local authorities and third sector agencies to better target EU funds and compensate for the lack of governmental support.’ Despite the project leader's call for stronger data, the published figures were used for fearmongering in public debates in 2013 before the UK labour market limitations were lifted for Romanian and Bulgarian citizens and, later, in the Brexit debates on the future of the UK's EU membership. ⁴ The published figures were used to generate hostile media headlines such as ‘UK Roma Population One of Biggest in Europe’ (Jenkins, 2013), ‘Roma Surge Threatens to Add to Estimated 200,000 Population Already in UK’ (Dawar, 2013) and ‘Roma Army: 200,000 are Already Here with MORE on the Way’ (Perkins, 2013).

Whilst the British right-wing parties, like the United Kingdom Independence Party (UKIP) and the Conservatives, used these figures to create and support a moral panic about the free movement of Romanian and Bulgarian citizens, the left-wing parties issued similar warnings about the difficulties that local authorities would have in dealing with the huge ‘inflow’ of Romani migrants (Leggio, 2019: 82). Ultimately, the main issue was not whether or not the data were reliable, but the fact that more data did not bring better and more equal inclusion for Romani minorities in the UK. Instead of focusing on integration, political parties from across the spectrum described Roma variously as increasing in numbers in the UK and as ‘problematic migrants’ for the state (Leggio, 2019: 69–88). Romani migrants themselves and their position were not seen as the significant issue, but rather taken as merely a symptom of a more contentious debate (Sardelić 2019b, 2019c). The very publication of these figures, whether or not they quantifiably overestimated the number of Roma in the UK, raised concerns, and they were used as a proxy for a debate about the nature of EU citizenship and the rights of EU citizens vis-à-vis UK nationals. Arguably, and no doubt regrettably, these then served as precursor debates for the UK referendum on EU membership (Leggio, 2019).

Across the political spectrum, UK politicians portrayed Roma as a ‘challenge’ exported from an enlarged EU. The assumption was that the new postsocialist EU Member States were not dealing properly with the integration of Roma and, together with the EU Free Movement Directive (see Chapter 2), this had become a problem for the Western EU Member States. UK political representatives, as well as the media, did not reflect upon their own tropes of stereotypes that they held both about Roma as well as Eastern Europeans. In these debates, Roma faced dual stigmatisation: on the one hand, by the UK public as unwanted immigrants generally, and on the other, by other Eastern European immigrants who considered Roma to shed a negative light on them as immigrants. In the UK, they were not positioned as fellow EU citizens, but as a question mark on the very concept of EU citizenship itself. At the same time, other Eastern European migrants distanced themselves and dismissed Romani minorities as their national fellow citizens. Equal EU citizenship was thus shown to be unavailable, and instead there almost existed different taxonomies of EU citizenship, as Rigo (2005) noted in his discussion of postcolonial subjects in the colonial empires. The constructed position of Romani migrants (Magazzini and Piemontese, 2019) in this debate was illustrative of the existence of citizenship taxonomies.

Whilst there is overwhelming agreement among researchers that the most reliable data on the number of Roma should be based on self-identification rather than external categorisation, the fact that self-identification is fluid and contextual (Messing, 2019) can also be used by the state authorities to diminish the rights available to Romani minorities. For example, Croatia amended the 2002 Constitutional Law on the Rights of National Minorities in 2010 and 2011. In 2010, Croatia had amended its Constitution to recognise twenty-two national minorities, and among them Roma were named as a national minority for the very first time. The constitutional recognition of Roma has had to be contextualised within a broader minority politics since Croatian independence. For the Serbian minority, who had been named as a constitutive nation in the Socialist Republic of Croatia before the disintegration of Yugoslavia, recognition as a national minority seemed like a downgraded status and was deemed to be ‘constitutional nationalism’ (Hayden, 1992). At the same time, Roma had never been recognised as a minority in any of the Yugoslav constitutions (Sardelić, 2013b), but there was a consensus that they belonged to the ethnic groups within the constitutional hierarchisation of constitutive nations–nationalities–ethnic groups. ‘Ethnic group’ was a designation for territorially dispersed minorities without territorial claims, and this translated into fewer minority rights than those held by nationalities (Sardelić, 2015).

Because of EU conditionality (Sardelić, 2011), Roma were nominally recognised as equal to other minorities in the Croatian Constitution for the first time. In principle, the Constitutional Law on the Rights of National Minorities reflected this recognition and should have given access to a range of minority rights to enhance the presence of minorities in the public space as Croatian citizens. For example, Article 12 gave all the national minorities the right to use their minority language and script in the local administrative units where the number of minority members exceeded at least one third of the total population. According to the Croatian population census, Roma in Croatia did not reach this threshold in any of the local administrative units. Whilst there was a possibility of having this right on paper, in reality it was not reachable. In the FCNM country report (FCNM, 2014b: 66), the Croatian government stated that even in cases where the minorities do not constitute one third of the local population, there are still informal agreements that local administration personnel are sometimes fluent in the minority language. However, the FCNM Advisory Committee's opinion (2016b: 23) in the fourth cycle stated that this did not extend to Roma since the language is not spoken by the majority or spoken as often as other minority languages. Although Roma in Croatia were constitutionally ‘upgraded’ from an ethnic group to a national minority, this did not necessarily translate into a broader scope of rights.

Yet the question is, why do more individuals not self-identify as Roma? One answer lies in the diversity among Romani minorities. Another answer lies in the widespread fear that identifying as Roma in an official capacity provides very few protections and rights for citizens on the ground despite their having targeted minority protection. There have been numerous cases in Croatia highlighting, for example, ethnic profiling and harassment by the police (ERRC, 2001) as well as the segregation of Romani children in schools (see Chapter 3).

In many European countries, the trade-off between having rights as a minority and having rights as citizens is a very real one for Roma. Self-identification of Roma, as in the case of other minority identities, is not fixed and is context-dependent. In addition, self-identification is not only a part of subjective processes, but is created in dialogue (Messing, 2014). For example, in an interview undertaken for this research, one highly educated woman, who identified as Roma both privately and publicly, noted that when she was looking for employment her employment advisor ‘benevolently’ recommended she should hide her knowledge of the Romani language (among the listed languages she speaks) in her CV and in any job interviews: ‘I think it is still true that no matter if you have education they would rather take someone else, not Roma. The only exception is if the job requires knowledge of the Romani language or is in any way connected to Roma. I still remember when I was at the Employment Office, the employment advisor there was shocked that I wanted to write in my employment profile that I can speak Romani. She told me between the lines that she was afraid this would be a reason why I would not get a job’ (quoted in Sardelić 2019a: 235, for more context, see Chapter 5). As this interview excerpt indicates, being visible as a minority often positions Roma at the fringes of citizenship, in this case with no access to employment.

Minority protection and socio-economic integration: and citizens’ equality?

The dilemmas around the numbers and naming of Roma also reflect regional geopolitical developments and policy orientations that international organisations and states have taken towards Romani minorities. Whilst the name Roma emerged from the Romani movement in 1970s, international organisations, especially the European Commission, only gradually started adopting this name rather than ‘gypsies’ in the 1990s. Before the fall of the Berlin Wall, European Commission documents referred to Roma as a social group which was disadvantaged because of its itinerant lifestyle (Simhandl, 2009). It was only with the prospect of EU enlargement that the discourse changed and EU institutions started using an ethnic denomination, gradually referring to Roma by this name as well as describing them as an ethnic minority especially from Eastern Europe. As argued by Simhandl (2009), the use of Roma as an ethnic denomination served to monitor the progress that former socialist countries were making in following the 1993 Copenhagen criteria, especially with respect to human rights, rule of law and minority protection. Whilst the accession progress reports were initially produced because of fears over minorities with potentially destabilising territorial claims, as in the case of the former Yugoslavia (Vermeersch, 2006: 3), they soon started underlining the position of Roma and their minority rights as the fear of their westward migration grew (Guglielmo and Waters, 2005).

Western EU Member States were not put under the same scrutiny in the 1990s (Donert, 2017). At the end of the 1990s, some of the EU Member States started reintroducing visa restrictions for the postsocialist EU candidate countries because of Romani migrants, which meant compromising the right to seek asylum (Clark and Campbell, 2000). For example, in 1999, the Finnish Ministry of Labour reported that 580 Roma from Slovakia sought asylum in Finland that year. Around 380 of those asylum seekers arrived in the space of a few days, so Finland decided to create a new visa regime for all Slovak citizens. Similarly, the UK briefly reinstated visas for Slovak citizens and threatened to introduce similar restrictions for Czech citizens if they received more asylum applications from those countries. The number of asylum applications in 1999 did not go above a few hundred (ERRC, 1999). Even before the countries with the largest Romani populations joined the EU, there was a strong perception that Roma should remain contained within their own countries and not be ‘exported’ to Western European states where it was presumed human rights standards were already respected. Indeed, the protection of minorities under the FCNM was part of the EU conditionality for the EU candidate countries (Blitz, 2013), but some founding EU Member States (like France) have neither signed nor ratified this Convention.

There was no shared agreement by the countries that joined the EU before 2004 on whether or not Roma should be legally recognised as a national minority. Spain, Portugal, Denmark, France and Greece did not include a definition of national minority in their legislation at all. Italy recognised linguistic minorities in its Constitution but did not include Roma among these minorities. In the fourth FCNM country report (2014c: 10), the Italian government commented: ‘No specific piece of legislation of our legal system recognises and protects the Roma communities living in Italy as linguistic minorities as they lack a stable connection with the territory, as it is well known.’ Finland included Roma in its Constitution as a group, next to Sami, who were recognised as an Indigenous group (Section 17 of the Constitution of Finland), whilst Sweden included Roma as a national minority in the National Minorities and Minority Languages Act 2009 with other groups (Article 2). Austria recognised Roma through its 1976 Volksgruppengesetz (although not directly mentioned in the Act itself) (FCNM, 2002a); this translates literally as the Ethnic Groups Act, but the FCNM Advisory Committee (2016a: 6) referred to it as the 1976 National Minorities Act. Germany recognised Roma and Sinti as a national minority with a Declaration after signing the FCNM (FCNM, 2002b: 6). In the UK, Roma were recognised as a racial group (to be protected by the Race Relations Act 1976) by the court decision in the case of Commission for Racial Equality v. Dutton (1989, see FCNM, 1999: 14) and later by the Equality Act 2010 (FCNM, 2014d: 7). ⁵ Whilst Irish legislation has previously included special provisions for the Irish Traveller community, it was not until 2017 that the Irish government decided to recognise them as an ethnic group rather than a group with a specific lifestyle (FCNM, 2017: 21). However, Ireland does not recognise Travellers as a national minority.

The FCNM Advisory Committee commented that the legal recognition of minorities as national was not a necessary predisposition for recognising their rights (FCNM, 2014a: 6). For example, the Advisory Committee's Opinion on Spain noted that whilst Spain does not have legal recognition of minorities, it does have comprehensive policies of inclusion for both traditionally settled and migrant Roma (see Magazzini, 2017; Kostka, 2018). Others comment that for migrant Roma the main approach is still forced voluntary return rather than integration (Vrăbiescu and Kalir, 2018). The FCNM Advisory Committee also noted that even though Roma have in many cases been recognised as a minority in national legislation, they were still the most discriminated citizens in their own countries and, faced with school segregation, police harassment and hate crime, experienced hindered access to healthcare and coerced sterilisation, among other issues. All this was caused either directly by the states or by the lack of state protection. According to the opinions of the FCNM Advisory Committee in the period before the 2004 EU enlargement, many postsocialist candidate countries introduced a form of legal recognition of minorities including Roma either in their constitutions or in other legislation or documents. In 1993, Hungary introduced Act LXXVII on the Rights of National and Ethnic Minorities, which was later repealed and replaced by the 2011 Act CLXXIX on the Rights of Nationalities. In the 1993 Act, Roma were named Gypsies, whilst in the 2011 Act they were referred to in the first instance as Roma/Gypsies and thereafter as Roma. Both laws were based on the self-government of minorities and their cultural autonomy. However, although legislation has in theory introduced Roma national self-government (McGarry, 2010), it has been criticised by the Romani non-governmental organisations (NGOs), which have claimed that this structure, rather than representing the interests of Romani minorities, complies with the practices of the leading party in Hungary:

Whilst in Hungary Romani minorities were recognised as nationalities from 2011 (in the previous Act they were included in the list of ethnic and national minorities), Slovakia and the Czech Republic included Roma as national minorities in their legislation. In the Czech Republic, Roma were recognised as one of the national minorities in the 2001 Act on Rights of Members of National Minorities. In Slovakia, the term ‘national minorities’ was mentioned in Articles 33 and 34 of the Constitution. The article requires a specific constitutional Act to further elaborate the rights of minorities. Yet such an Act was, according to some scholars, never introduced (Constantin, 2010). There was a similar situation in Slovenia. Article 65 of the Constitution of Slovenia mentions the ‘Romani community’, yet the specific rights given to this group were not defined in the Constitution but required a new legal act. In the subsequent legislation adopted in Slovenia, Roma were recognised as an ethnic community, whilst other minorities, such as Hungarians and Italians in Slovenia, were recognised as national communities. Whilst Italian and Hungarian minorities had their rights recognised by the Slovenian Constitution (Article 64), that was not the case for Roma minorities. In contrast, Poland made a similar distinction between ethnic and national minorities, but constitutionally recognised their rights as equal (Article 35). In the 2005 Act on National and Ethnic Minorities and on Regional Languages, Roma were recognised as an ethnic group, as in Slovenia, but without the hierarchy of rights that Slovenia established for national and ethnic minorities.

In 2007, Slovenia became the only country in Europe to introduce a special Roma Community Act after a 2006 pogrom of a Romani village committed by the local majority population (Vidmar Horvat et al., 2008). However, the Act only confirmed the hierarchy of rights between constitutive nation, national communities and Roma as an ethnic community, with far less scope for minority rights (Sardelić, 2013a). For example, whilst bilingual education was possible for national communities, Roma do not enjoy a similar right in Slovenia. National communities in Slovenia have a representative in the parliament, whilst Roma only have councillors at the municipal level. Furthermore, the Roma Community Act only recognised some Romani minorities living in Slovenia: it did not give rights to Romani minorities who were not arbitrarily considered to be autochthon in Slovenia. The reason for this was to deprive internal migrants and migrants from other parts of the former Yugoslavia to Slovenia of rights (Janko Spreizer, 2004; Sardelić, 2012a).

Whilst Slovenia clearly introduced a hierarchy of rights for different types of minorities, it allowed so-called dual voting for recognised minorities. Members of Italian and Hungarian national minorities can vote for their representatives in parliament, but they can also vote for general representatives. For Roma, this form of dual voting is possible at the municipal level: they can vote for a minority representative and for a general representative. The Venice Commission (2008) published a report on dual voting by national minorities which found that only Slovenia allows dual voting for minorities whereas other EU Member States do not, as they see it as a deviation from the ‘one person – one vote’ rule. For example, in Croatia, Roma have a representative in parliament (the same MP also represents other national minorities), but they cannot vote for a general representative. In 2010 the Croatian Constitutional Court decided, ‘in these cases it does not seem plausible that recognition of the dual voting rights would advance the relationship between the “privileged minority” and other citizens. Actually, such a privilege in a legal sense could lead to conflict’ (Constitutional Court of Croatia, 2011, my emphasis and translation ⁶ ). Whilst the decision of the Croatian Constitutional Court was mainly focused on dual voting by the Serbian minority, Romani minorities were affected as well. In this case, minorities could vote either as a minority members or as general citizens but not as both.

After the most substantial EU enlargement in 2004, the policy towards Romani inclusion shifted the discourse once again: whilst Roma remained named as an ethnic minority, the social and economic disadvantages they faced were again among the prominent themes before minority protection was enforced in legislation. Some scholars argued that Roma needed more social and economic inclusion than group-targeted rights (Pogány, 2006). Bulgaria and Romania did not introduce specific minority legislation that would include Roma prior to their accession to the EU. However, they were party to initiatives for the 2005–15 Decade of Roma Inclusion. The Roma Decade introduced new international actors, such as the Open Society Foundation, the World Bank, the United Nations Development Programme, the CoE, and the Organization for Security and Co-operation in Europe (OSCE) in collaboration of some EU and non-EU states as the main drivers of Romani inclusion (Brüggemann and Friedman, 2017). The Roma Decade specifically targeted Roma but had a similar logic to the ten-year Millennium Developmental Goals (MDG) plan: ‘As a result, in Europe, concerns about Roma exclusion have been, at the heart of the concerns about inequality in MDG implementation’ (UN, 2013: 1). Besides the aforementioned international organisations, the Roma Decade was a partnership with countries such as Albania, BIH, Bulgaria, Croatia, the Czech Republic, Hungary, North Macedonia, Montenegro, Romania, Serbia, Slovakia and Spain. Whilst these countries were full members which needed to introduce and implement action plans that would improve access to education, housing, healthcare and employment for their Romani minorities, Slovenia and the US took part as observers. One of the agreed successes of the Roma Decade was that there was a clear continuation of the EU NRIS Framework (Brüggemann and Friedman, 2017). However, it did not achieve the main goal that was set: to close the inequality gap between the Romani and non-Romani population. Željko Jovanović, director of the Roma Initiative Office at the Open Society Institute, stated that the reason was institutional antigypsyism: ‘Anti-Gypsyism, as a form of exclusion, is not haphazard. It is embedded in our domestic institutions and structures. It runs through public offices, schools, hospitals, the labour market, the welfare system, police, and elections. A Roma child denied schooling with everyone else is not the result of one rogue, racist teacher – a whole system, built and entrenched over time, has led to this’ (Jovanović, 2015).

The EU institutions, particularly the European Commission and the European Parliament, started developing parallel mechanisms for Romani integration, such as the Roma Inclusion Platform and the EU NRIS Framework (Brüggemann and Friedman, 2017). Like the Roma Decade, the EU institutions focused on four areas where there was a perceived gap between the majority population and Romani minorities: (1) education; (2) employment; (3) access to healthcare; and (4) housing. The EU NRIS Framework functioned as a form of ‘soft law’ for the EU Member States and also candidate countries. The countries were required to design their own national strategies and then report to the European Commission, which would conduct an evaluation of the progress they made.

The EU NRIS Framework was adopted in 2011 as a policy response to the evictions and expulsions of migrant Roma from postsocialist countries (Romania and Bulgaria in particular; see Carrera et al., 2019). Just before the adoption of the Framework, the EU action plan in the field of justice and home affairs – the so-called Stockholm Programme 2010–2014 – categorised Roma as a vulnerable group (Article 2.3.3): ‘The EU and the Member States must make a concerted effort to fully integrate vulnerable groups, in particular the Roma community, into society by promoting their inclusion in the education system and labour market and by taking action to prevent violence against them’ (European Council, 2009). However, soon after its establishment, the vision of integration proposed by the NRIS Framework came under criticism from a number of scholars and policymakers. For example, Goodwin and Buijs (2013) analysed the Framework, concluding that it took a set of majority values and expected assimilation rather than integration of Roma. Iusmen (2018) argued that instead of taking minority protection, anti-discrimination and the human rights of Roma as valuable in themselves, the Framework repeated the previous assumptions of organisations such as the World Bank that Romani integration is framed in terms of the economic prosperity of the EU: here, Roma were not seen as EU (and Member States) citizens, but primarily an impediment to the EU's economic goals.

Another criticism of the Framework was that, as with its predecessor, the participating states did not address antigypsyism, instead putting the burden of the integration of Romani minorities onto the rather power-limited responsible authorities (especially those at state level) and majority population (Carrera et al., 2019). Different civil society actors, such as the European Romani Grassroot Organisations Network, emphasised that any further policies on ‘Romani integration’ should not only include measurable results, but also redress the antigypsyism of state actors when such policies are implemented (ERGO, 2020). The two major achievements in relations to the EU NRIS Framework did not come from the states, but from the Romani and the pro-Roma civil society: An Alliance Against Antigypsyism was formed out of a plethora of such organisations, and this supported a new definition of antigypsyism (Alliance Against Antigypsyism, 2017).

The other underlying issue with the EU NRIS Framework is its perpetuating narrative of underdevelopment. In the EU accession processes, the question of underdevelopment was addressed as the democratisation of postsocialist candidate countries. When the countries joined the EU, the narratives shifted to representing Roma as a vulnerable community that needed development so that it could catch up:

Both the approaches of minority protection and socio-economic integration were addressed in a project-oriented manner (spanning in decade cycles), as if the inequalities of Roma as citizens could be eradicated by working on their position rather than addressing and deconstructing the structures that led to such inequalities. In addition, as international organisations have become the main financial supporters of advocacy for Romani rights, they have expected Romani NGOs to follow their developmental vision as to what Romani integration should look like (Plaut, 2016, see also Chapter 5).

The developmental discourse in EU approaches towards Roma in Eastern Europe was dubbed ‘Europeanised hypocrisy’ (Ram, 2014b) and portrayed as neocolonialism (Trehan and Kóczé, 2009) because it had double standards on Romani inclusion in Western European countries that also failed to protect Roma. The developmental approaches towards Roma in Eastern Europe also produced a backlash from Eastern Europe itself: the local political elites started portraying Roma as a minority privileged by the EU who, consequently, were then seen as neglecting the predicament of the majority population in this region (Vermeersch, 2012). As Vermeersch (2012) showed, Roma were described as an EU minority, but at the same time alienated from their national fellow citizens. Whilst the (trans)national legislation and policy made them visible as a minority, Roma remained invisible as citizens at the fringes of their own national citizenship.

Roma as an exceptional minority? Names and numbers

Building on Alana Lentin's (2004) theory (see also Lentin, 2020), van Baar (2019) shows that one of the main challenges of the Roma inclusion programs – both of the Roma Decade and of the NRIS – was the assumed superficiality of antigypsyism (anti-Roma racism): in their policy documents, states as well as international organisations highlighted the exceptionality of antigypsyism, which goes against principles on which these states are founded. They did not recognise antigypsyism as something deeply embedded within society. Exceptionality of antigypsyism can be understood also in another way. Antigypsyism has been ascribed to encountering Roma as an exceptional minority, and has not been primarily understood as a violation of rights that all citizens are supposed to have regardless of their background. In the early 1990s, Václav Havel claimed that antigypsyism is a ‘gypsy problem’ that remains salient even when functioning democracies have been established (Kamm, 1993) and therefore is a litmus test for a civil society in postsocialist countries. Yet the Romani civil society and the Romani movement (Vermeersch, 2006; McGarry, 2010) had been strongly present in the postsocialist EU Member States even before the fall of the Berlin Wall (Donert, 2017). Nevertheless, the structural inequalities that Roma face across the EU remain intact.

Structural inequalities are, I claim, a product of the current liberal democratic states and their citizenship regimes rather than the alleged underdevelopment of Roma or the lack of critical civil society. This can be shown by comparing the position of Roma with that of other non-immigrant minorities in liberal democratic states around the globe. Whilst Romani minorities are arguably in a different position from Indigenous people and African Americans in postcolonial contexts, the treatment of these minorities that states and international organisations have applied has been strikingly similar. The first similarity can be seen in the controversies around naming and numbers of marginalised citizens. For example, in the cases of both Native Americans and African Americans, states have used derogatory naming in legal documents that have made both groups visible as minorities but marginalised as citizens. Albeit based on misconceptions of the first colonisers, the term ‘Indian’ remains present in Canadian legislation, such as the Indian Act, despite the fact that the activists’ and representatives’ preferred term is First Nations (Ramos, 2006). One of the Smithsonian Museums in the US is officially named the National Museum of the American Indian. Like Roma in Europe, Native Americans (including those categorised as Indian Americans) in the US do not represent more than 5 per cent of the total US population according to the 2010 US population census (Leavitt et al., 2015).

As Leavitt and her colleagues (2015) point out, Native Americans remain invisible as citizens in the US context. They are, however, represented in a stereotypical manner in the mainstream public domain. This is best illustrated by controversies of naming mascots and sports teams, such as the Washington Redskins. Native American civil society activists have been advocating a change to this name since 1940 (Phillips, 2017) as it reinforces a derogatory term in the public domain. The owner of the Washington Redskins asserted that he would not change the name because of the majority's support of this name: ‘In justifying his continued support for the team's name, Snyder asserted that the overwhelming public support for the continued use of the term “Redskins” was reason enough to retain the name’ (Nteta et al., 2018: 474). Whilst Native American civil society, with the support of professional organisations such as the American Psychological Association, has been successful in limiting the use of mascots and derogatory names in the educational sector through legislative battles, it has had less success when it comes to registered trademarks: according to the First Amendment on free speech, the Washington Redskins could retain their name; in mid-2020 Snyder decided to change the name to Washington Football Team (Washington Football Team, 2020).

Whilst civil society activists have been advocating against using the term Indians for Native Americans, according to the 2010 US census there is a larger group in the US that identifies as American Indian rather than Native American (Leavitt et al., 2015). Similarly, in some countries in Europe, such as Spain, the traditional minority refers to itself as Gitano (Gypsy) rather than Roma. However, whilst there is a proven link between derogatory names and the marginalised position of certain minorities (Stegman and Phillips, 2014), the politically correct name does not necessarily diminish exclusionary stereotypes and practices towards these minorities. For example, both Native Americans and Roma have been portrayed as savages on the one hand and as being noble and free through positive stereotypes on the other. Certain cultural practices have reinforced the former image in the media. For example, scalping has been described as one of the cultural practices of Native Americans, despite originating from the settlers (Ganje, 2011), just as nomadism was seen as a cultural practice of Roma (Sardelić, 2019b). As Leavitt and her colleagues point out (2015), Native Americans do remain invisible as citizens, but they are also depicted as ‘frozen in time’ and as a community not on the same developmental level as mainstream society.

Human rights breaches and the biopolitics of citizenship

Besides the problems with names and numbers, the human rights breaches that different marginalised minorities face as citizens are similar in content. Whilst monitoring the minority rights situation and socio-economic integration of Roma, the FCNM Advisory Committee also continuously marked systematic breaches of rights that Roma experience as citizens. As the Advisory Committee has noted, Romani children have been systematically placed in either segregated classes or segregated schools in both Western and Eastern Europe. The US formerly treated its African American populations in a similar manner (this is discussed in Chapter 3). The next chapter shows how liberal democratic states have systematically limited the rights of minorities on their territories when they clashed with other state's interest. Chapter 4 discusses how Romani minorities and other similarly marginalised minorities around the globe experience impeded access to citizenship.

Women identified as belonging to marginalised minorities have often been faced with violence that needs to be studied with an intersectional approach (Yuval-Davis, 2011). One of the practices that Romani women have faced in order to prevent the ‘wrong’ biological nation being reproduced (Yuval-Davis, 1997; Kóczé, 2009) is coerced sterilisation. Throughout the twentieth century, several countries (such as Sweden, the Czech Republic, Hungary and Slovakia) employed coercive sterilisation for Romani women who were described as having too many children. Whilst this new century has banned coerced sterilisation as a state-sponsored practice, it still occurs because of deeply ingrained prejudice against Romani minorities. The ERRC has taken the Czech Republic, Slovakia and Hungary to court over this issue. Coerced sterilisations have happened without Romani women's informed consent, as they have usually been requested just before they undergo caesarean sections and are already under the influence of painkillers and sedatives. In Hungary, a woman who underwent sterilisation without informed consent was given compensation by the state. However, in I.G. and Others v. Slovakia (2012), the ECtHR did not recognise the practice of coerced sterilisation as a direct discrimination of Romani women: ‘The Court decided that in view of the documents available, it could not be established that the doctors involved acted in bad faith, that the applicants’ sterilization were a part of an organized policy, or that the hospital staff's conduct was intentionally racially motivated. At the same time, the Court insisted that shortcomings in legislation and practice relating to sterilisations were liable to particularly affect members of the Roma community, so their discrimination in these cases would be only indirect and unintentional’ (ERRC, 2016: 20).

The practice of coerced sterilisation existed in different Canadian provinces throughout the twentieth century (Dyck, 2013). In 2018, two Indigenous women from the Canadian province of Saskatchewan filed a class action lawsuit after they were sterilised without prior free and informed consent. However, in this latest class action lawsuit the similarities between the experience of the two Indigenous women is strikingly similar to those Roma women: they were both sterilised whilst undergoing caesarean sections. Similarly, like some European countries, some Canadian provinces formerly had eugenics laws that included coerced sterilisation. Whilst such laws were abolished, the practices of coerced sterilisation continued: ‘The practice has continued into the 21st century. Approximately 100 Indigenous women have alleged that they were pressured to consent to sterilisation between the 1970s and 2018, often whilst in the vulnerable state of pregnancy or childbirth’ (Stout, 2019). The authorities in both Czechoslovakia and Canada eugenically justified forced sterilisation as a means of control over the under-developed population and as a method of preventing poverty prevention among large families. As activists against the coerced sterilisation of Romani women have stated, this act was ‘racism's cruellest cut’ (Rorke and Szilvasi, 2017). In Foucauldian terms, coerced sterilisation can be seen as the biopolitical act (van Baar, 2016) of cleansing citizenry. In both cases, the responsible authorities legitimised such procedures as practices ‘for their own good’ (Stout, 2019).

The developmental logic of decades

The ‘Decade’ approach was not an original invention for international intervention for ‘improving’ the position of Roma. Other marginalised minorities around the globe were scrutinised by the ‘Decade’ approach as well. In December 1993 the UN General Assembly adopted Resolution 48/163, which established the first International Decade of World's Indigenous Peoples (1995–2004). A year later, the Office of the UN High Commissioner for Human Rights (OHCHR) was appointed as the coordinator of the Decade and an action plan for how to improve the position of the world's Indigenous peoples was introduced. The aim of the Decade was to enhance international cooperation to address the pressing issues faced by Indigenous people in the realm of human rights, territorial rights, development education and access to healthcare. In 2004, at the end of the first Decade, the coordinator submitted a report to the Economic and Social Council which emphasised that the Decade had brought about institutions, programmes and projects that highlighted the position of the world's Indigenous peoples, but the general conclusion was distressing:

Whilst also addressing specific challenges that Indigenous communities around the globe face, the conclusion drawn after the first Indigenous Decade has been strikingly similar to that after the Roma Decade. The General Assembly adopted another resolution in 2004 to start a second Decade of the World's Indigenous peoples (2005–14). Over this period, the most important international document relating to the position of Indigenous peoples came into existence: the UN Declaration on the Rights of Indigenous Peoples of 2007. The Declaration identified specific rights that should be guaranteed to Indigenous peoples around the globe by their states. Besides the general anti-discrimination provisions, it included Article 3 on self-determination of Indigenous people, Article 8 on mechanisms for prevention and redress, Article 19 on the necessity that the states acquire informed consent from Indigenous people and Article 26 on the right of Indigenous people to their traditional lands. This corpus of rights was specifically designed to address the position of Indigenous people around the world. However, Australia, the US, New Zealand and Canada voted against the adoption of the declaration and did so in a clear action to protect the current structures of their liberal states (Moreton-Robinson, 2011; Lightfoot, 2012; O’Sullivan, 2020). The four countries where the largest number of the world's Indigenous populations live presented multiple arguments: against self-determination because it could destabilise the territories of their countries; against returning traditional lands to Indigenous people because it would ignore current ownership; and finally, against prior informed consent in matters concerning Indigenous people, which was interpreted as causing inequality among citizens by giving Indigenous people a ‘veto right’ which other majority citizens do not have. In subsequent years, the four countries in question supported the declaration, albeit without changing their position on it: they claimed they would abide by the Declaration, which is legally non-binding, as long as it is not in contradiction with their pre-existing national laws (Lightfoot, 2012).

The Second International Decade of the World's Indigenous Peoples (2004–14) ran alongside the first Roma Decade of 2005–15. There was no second Roma decade, but the EU instead formulated the EU NRIS Framework, which concluded in 2020. In the penultimate year of the EU NRIS Framework, under the leadership of the Romani MEP Soraya Post, the European Parliament introduced a Resolution (2019/2508) that would acknowledge not only that the EU NRIS Framework did not fully succeed in its goal but also that it failed to address the underlying cause of Romani marginalisation: antigypsyism. The Resolution showed some awareness that Member States – like those states in the case of the UN Declaration on the Rights of Indigenous Peoples – are willing to go along with multicultural protection as long as it does not question the very fundamentals upon which the states are based, even if those fundamentals instil the marginalisation of some of its citizens.

Conclusion

The analysis of law and policy documents as well as public discourses and practices shows that there are parallels in how states position marginalised minorities on the fringes of citizenship both in settler colonial contexts and in the context of the expanding EU. States are willing to support benevolent policies towards marginalised minorities but only if they do not address the original structures (which may well be racist) upon which these states themselves are built. It is these invisible edges of citizenship, which position some groups as a visible problematic minority rather than shining the light on the problematic policies that create inequality among citizens. Despite existing liberal democracies and fundamental rights, it is precisely within such systems that civil marginalisation occurs. The FRA (2018) itself compared the position of Roma in Europe to that of individuals living in Third World countries. However, this masks the fact that their position is not a product of underdevelopment but arises in the most developed states in Europe. Similar observations have been made about alleged underdevelopment of Indigenous people in Australia, New Zealand, the US and Canada (Cooke et al., 2007). In this chapter, I have analysed (inter)national legislation and policies introduced for the protection of minority rights of Roma. I have shown how plans to improve the position of Roma in Europe as well as other marginalised minorities around the globe were anchored by the invisible edges of citizenship and put these minorities on the fringes of citizenship: while legislation and policies gave the appearance of their minority rights being over-protected, at the same time the rights of these minorities as citizens remained under-protected despite the legislation on minority protection that was introduced. In the next chapter, I look at a different angle: I show how states willingly violate their own laws and international legislation in respect of marginalised minorities in order to serve their preferred greater interests.