Cacophony in conceptualizing and operationalizing ethnicity: the case of Roma in Hungary

ABSTRACT

Using secondary research from the political, education, and employment fields, this paper aims to demonstrate the consequences of confused and overlapping conceptualizations of the Roma in Hungary as an ethnic group, a racialized minority, a national minority, and a socially disadvantaged group. The resulting cacophony of operationalizing schemes blurs clarity and constrains efficient measures for inclusion policies. In social sciences and law, the purpose of classification is to help us understand the internal logic of concepts. Thus, classification has significant consequences, as it can imperil policy goals. Through examining the case of the Hungarian Roma, the article demonstrates how the confused conceptualization of ethnicity, race, and nationality and ill-applied methods of operationalization have vastly detrimental consequences. In addition, it is argued that many concerns regarding ethnic data processing that policy actors voice are legally unfounded, and pre-existing data protection regimes allow the processing of ethno-racial data.

KEYWORDS:

• Roma
• ethnic data protection
• categories of difference
• equality policies
• minority politics
• Hungary

1. Introduction

This article discusses the conceptualization and operationalization of race and ethnicity in Hungarian law, with particular interest in the use of legal categorizations, terminological clusters, and their relationship to policymaking. Our thesis is that conceptualization, operationalization, and policy-framing are connected and that the incoherent conceptualization and operationalization of race and ethnicity contribute to the ineffectiveness of policies geared toward ethnic groups in Hungary, especially with regard to the Roma population. We also identify the phenomenon of ethno-racial data “processing phobia” – a reluctance to “see” race and ethnicity by policy actors due to a false apprehension of European privacy regulations. We argue that this leads to Murphy’s law of data protection, whereby discriminators have no difficulty identifying their victims, but when it comes to remedial action or social inclusion measures, implementation is paralyzed by the misinterpretation or overly cautious application of the principles of processing ethnic data. The theoretical and general issue is addressed by focusing on the case of Roma in Hungary, where conceptual cacophony and practical inhibition paralyze efforts to promote social inclusion, equal treatment, and the exercise of minority rights. The case of Roma communities in Hungary provides a unique opportunity to reveal the political and ethical dilemmas associated with conceptualizing and operationalizing race, ethnicity, and nationality in the law, public policy, and related everyday practices. By presenting the above dilemmas, we intend to fill in a research gap left open by the Western and Northern European research traditions that mostly overlook how ethnicity (and race) function differently in Eastern Europe.

The case study of the Roma in Hungary is instructive in relation to this special issue in five ways. First, as one of the paper’s ambitions is to capture ethnicity as a category of the law, it shows how legal dimensions are entwined with categories of analysis and experience. Second, the article makes the point that policy objectives (such as collective rights recognition, protection from victimization, and social inclusion) and operationalizing techniques (such as self-identification or external identification by a third party) should also shape and determine conceptualization and legal terminology, that is, whether we talk about national-, or ethnic minorities. Third, the article argues that we need specific references to race and ethnicity, as well as to nationality: whereas race and ethnicity focus on victimization, the concept of national minorities accommodates cultural claims for some singled out, historical, or “deserving” new minorities in Hungary. Fourth, the article emphasizes the unique positionality of the Roma and, thereby, the historical and geopolitical contextuality of conceptualization (For the broader European context, see also Magazzini 2024). Roma were included in the 1993 minority law,¹ which set out the framework for the cultural recognition and political representation of 13 nationalities in post-communist Hungary. Even though the primary target of the legislation was the governments of the neighbouring countries where large Hungarian communities lived, Roma were included in the law because it would have been politically and morally unacceptable at the time to exclude the largest minority community. Finally, the paper’s illustration of the ambivalence of conceptualizing and operationalizing Roma ethnicity in three areas (minority self-government, education, and the labour market) helps to disentangle how the conceptual and practical frames as drawn by the law, by individual policy areas, and by individual policy actors intertwine in a confused way and result in a cacophony of concepts, principles, and practices.

As for the methodology, the article is based on an analysis of literature as well as on sources such as legal texts, political documents, and published research outputs. It is thus grounded on a qualitative reading and systematic analysis of secondary data. The legal part follows the customary and traditional methods of textual and structural analysis. The three case studies were selected because of their salience concerning the focal point of the project: illuminating the cacophony of conceptual and operational schemes concerning the Roma and the consequences of ill-fitted conceptualization in the everyday practices of policy actors. The case studies are rooted in the earlier work of the authors, as well as an analysis of scholarly literature on the subject matter and an assessment of other research outputs (i.e. working papers and reports) when these were available. The part on political representation relies on legal and literature analysis. The case study on education relies on a wide set of research on school segregation and its consequences. The part on employment is based on a systematic analysis of research outputs and publications related to employment policies and practices targeting Roma people.

The paper proceeds as follows: first, we map out the legal landscape of classification clusters of race, ethnicity, nationality, and class. We make the theoretical point that most legal regimes do not provide guidance regarding the relationship between legal and social aspects of identity. We argue that an overzealous attitude toward data protection leads to colour-blind racism (see Bonilla-Silva 2006) and impedes the implementation of social inclusion and minority protection. In the third chapter, we present the unique case of the Roma in Hungary, a community that is located at the intersection of all four concepts: race, ethnicity, national minority, and class. The fourth chapter provides a case study of three policy areas (political representation, employment, and education) to show how inconsistent conceptualization is paired with ill-suited operationalization, resulting in faulty policy implementation. In the concluding chapter, we offer some normative guidance on how to match conceptual approaches with operational categories.

2. Theoretical, conceptual, and practical dilemmas and claims

This section first contextualizes the legal landscape of classification clusters for race, ethnicity, and nationality. We begin with an inventory of legal benchmarks for classification and argue that the distinction between race and ethnicity is practically redundant.² Following this, we delineate the five legislative and policy strategies and techniques for operationalizing ethno-racial group membership: self-identification; identification by other members or elected, appointed representatives of the group; ascribed classification by outsiders, relying on perception; “objective” criteria; and using proxies such as names or residence (Pap 2023). Finally, we present the phenomenon we refer to as data-processing phobia.

2.1. The legal landscape of classification clusters for race, ethnicity, and nationality

We begin with an assessment of how the law conceptualizes and operationalizes race, ethnicity, and nationality.³ In this section, our focus is on legal categorization, leaving broader social science approaches for later discussion. Questions of recognition are relevant for numerous fields, for example, in debates on census categories, especially concerning the enumeration of national minorities; in defining anti-discrimination and affirmative action eligibility; identifying hate crimes, and even in asylum law (where refugee status can be granted based on the fear of persecution due to race or ethnicity) (Pap 2023). Most international and domestic legal documents on minority rights and human rights only provide vague descriptions of race, ethnicity, and nationality without defining the groups or affiliation requirements (see Pap 2023).

In many jurisdictions, as well as by international organizations, race is often used as synonymous with ethnicity. For example, under Article 1. of the 1965 International Convention on the Elimination of All Forms of Racial Discrimination, “the term ‘racial discrimination’ shall mean any distinction, exclusion, restriction or preference based on race, colour, descent, or national or ethnic origin  … ” The common feature in provisions referring to race and ethnicity is that for most both concentrate on protection from maltreatment such as discrimination or victimization by hate crime. Ethnicity, however, also “implies a more complex set of claims [than race]. As well as being grounded in the anti-discrimination discourse, some ethnically defined groups have also set out cultural claims similar to those of national minorities” (Pap 2023, 3). The core of differentiating between national and ethnic minorities by international legal terminology is thus that the former have nation-states as national homelands, whereas ethnic minorities do not (Hannum 2000). Consequently, ethnic minorities have claims that blend those made by racial and national communities (Pap 2023).

An outstanding example of avoiding defining relevant categories is the 1995 Council of Europe (CoE) Framework Convention for the Protection of National Minorities which does not provide a definition of its targets. It is only the Recommendation 1201 (1993) of the CoE Parliamentary Assembly for an additional protocol⁴ that gives some guidance as to how the category of national minority is conceptualized:

the expression “national minority” refers to a group of persons in a state who: (a) reside on the territory of that state and are citizens thereof; (b) maintain longstanding, firm and lasting ties with that state; (c) display distinctive ethnic, cultural, religious or linguistic characteristics; (d) are sufficiently representative, although smaller in number than the rest of the population of that state or of a region of that state; (e) are motivated by a concern to preserve together that which constitutes their common identity, including their culture, their traditions, their religion or their language. (Section 1 Definitions; Article 1; for more see Pap 2023)

According to this framework national minorities’ claims for collective rights go beyond an anti-discrimination logic and include claims for cultural and political rights (Pap 2023). While the national minority rights framework has arguably well-settled conceptual and policy boundaries, the distinction between race and ethnicity is practically superfluous in this sense. A thick conception of a single term and concept, either that of race or, preferably, ethnicity, would bring more clarity to theory as well as practice. It also needs to be added that a more inclusive concept could also incorporate further related but similarly inconsistently used and ambiguous concepts such as colour, cast, and indigeneity.

The operationalization of ethnic/racial/national group affiliation can emerge in five ways: through self-identification; classification by outsiders; classification via “objective”, externally defined criteria; authority given to elected or appointed members of the group and the construction of proxies (For more see Pap 2023). Let us elaborate on some examples. Ethno-racial self-identification is the most frequent but not the only operationalizing model that legal regimes apply. Community acceptance and classification by the respective group is the basic principle of indigenous tribal membership. Anti-discrimination law and practice recognizes a number of markers as proxies for ethnicity or race: skin colour, name, citizenship, place of birth, parents’ place of birth, language (mother tongue or language habitually used), or even customs and religion, etc. The perception of outsiders is yet another method of conceptualization applied for example in ethno-racial data collection, which regards the census taker’s classification. Similarly, medical and educational documents often rely on doctors’ or teachers’ assessment of group membership (see, for example, Chopin, Farkas, and Germaine 2014; Farkas 2017 and Pap 2023).

2.2. Data-processing phobia

Operationalizing ethnicity in continental Europe is often associated with sensitivity and an over-cautious approach to privacy. Some European privacy stakeholders take an extremely broad approach to personal data processing, interpreting as the given person’s sensitive personal data “any sort of statements about a person, covering ‘objective,’ as well ‘subjective’ information, opinions or assessments”,⁵ and even individuals’ perceptions of another person, such as their ethnicity. This privacy-fundamentalist approach has been endorsed in policy implementation by administrators and HR professionals. The particular Murphy’s law that stems from this is the following: perpetrators of discrimination or hate crimes never have problems identifying their targets, while authorities and benevolent stakeholders face privacy dilemmas in their efforts to tackle direct and even more complex forms of institutional marginalization and discrimination related to ethnicity or race. Thus, authorities and policy actors’ reluctance to recognize and register ethnicity obstructs efforts to combat discrimination in education, employment, policing, and other social fields. For example, employers committed to the idea of applying hiring preferences for minorities find themselves paralyzed by ethno-racial data processing concerns. Data protection arguments are also used as an excuse not to monitor ethnic profiling (over-policing), as well as for not prosecuting racially motivated hate crimes as such (under-policing). This approach not only obstructs combatting discrimination but also limits the potential of extending equality and diversity measures.

It is important to state, however, that this ethnic data-processing phobia is often legally unfounded, as a number of circumstances offer a wide array of exemptions from data processing restraints in the field of employment, law enforcement, social care, etc. Even the strictest data protection regimes allow the processing of ethno-racial data under certain circumstances. The European framework known as the General Data Protection Regulation, or GDPR,⁶ defines “personal data” as any information relating to a natural person (“data subject”). Article 9⁷ classifies “racial or ethnic origin” (among political opinions, religious or philosophical beliefs, genetics, and information pertaining to health, sex life, and sexual orientation) as “extra-sensitive data” in relation to fundamental rights and freedoms. The processing of such data is allowed only under ten narrowly defined circumstances⁸, creating the grounds for processing ethnic data when this is in the vital interests of the data subject or about the establishment or defense of legal claims, including discrimination. Authorization to collect ethnic data is corroborated by various further international documents, as Patrick Simon’s (2007) study on the relationship between ethnic statistics and data protection, published by the European Commission against Racism and Intolerance (ECRI) shows (Simon 2007).

3. Roma: class, race, ethnicity, nationality – in Hungary and beyond

In Hungary, the Roma are the only visible ethno-racial minority aside from recently arrived immigrants. They have been present in the country for centuries. The Roma are linguistically assimilated – practically all speak Hungarian, with most having Hungarian as their mother tongue and few being bilingual. They live a sedentary lifestyle and do not differ significantly from the majority in terms of religious affiliation. Still, they have been targets of discrimination and social exclusion for centuries. Although the category “Roma” includes a wide diversity of cultures and traditions (Romungor, Boyash, Vlah, just to name the three largest subgroups), the Roma community is homogenized by the “othering” of the majority and by the political concept of the Roma as constituted by state policies and the international Romani movement to some extent (see, e.g. Fosztó 2003).

According to the Council of Europe, about 7 per cent of the European Roma population lives in Hungary (CoE and ERICarts 2010), constituting the largest minority group in the country. In the 2022 population census, about 2.2 per cent of the population, 209,909 people, identified as Roma (Hungarian Central Statistical Office 2013), but the Council of Europe (2010) suggests that the actual number might be closer to 700,000 people.

What makes the case of the Roma special when talking about the complex relationship between categories of difference? (van der Tol and Becker 2024) The Central-East European socio-demographic scene is quite different from that in Western Europe. There is no history of colonialism, slavery (except for the enslavement of Roma in Romania), or the large-scale immigration of visible “ethnic” others, thus there is no other visible, “racialized” or “racializable” minority⁹ or (at least in most parts of Central-Eastern Europe) a visible, “othered” Muslim community. This region has no experience with a Western type of multiculturalism, nor, for instance, claims attached to ethno-religious otherness as made by Western European Muslim communities. Thus, culture – be it the accommodation of collective rights or a source of othering – is only meaningful in the context of national minorities. It also needs to be added that the very concept of the protection of national minorities is rooted in Central-East Europe’s specific context and history, emerging after WWI and intrinsically related to the positionality of populations with challenged and conflicted nation-building histories. In Hungary, the context includes the ambition to compensate for territorial losses, and the reference points are populations who find themselves in the diaspora. In sum, in countries like Hungary, the Roma are the group most clearly seen by the ethnic majority as an “ethno-racial other”. Othering is not only produced by being conceptualized as a national minority but also by how new cultural racism securitizes, ethnicizes or racializes poverty and how far-right discourses create and weaponize the discourse of “Roma/Gypsy criminality”. Roma in Hungary fit the conceptual frame regarding the complex relationship between categories of difference described in the introduction to this journal Special Issue (van der Tol and Becker 2024). Their visibility involves the intertwining of four items: “racializable” features, such as (skin) colour, as well as “ethnic culture” (like clothing), national minority status (as in law), largely overlapping with class (i.e. a marginalized social position).¹⁰

As far as the international legal and policy conceptualization goes, European Roma (referred to as Roma, Romani, Gypsies, and Travellers) are at the intersection of being classified as a race, an ethnicity, a national minority and (a proxy for) a marginalized social class. When conceptualizing the Roma, courts and other legal fora seem to use “race” and “ethnicity” with similar ease. For example, the UN Committee on the Elimination of Racial Discrimination, the monitoring body of the treaty (on the elimination of racial discrimination) accepted complaints concerning the treatment of Roma,¹¹ thereby recognizing the community to fall under the auspices of a racial group (Guglielmo 2004, 10). Likewise, the Grand Chamber of the European Court of Human Rights (ECHR) identified a school segregation targeting members of the Roma minority in the Czech Republic as a form of racial discrimination¹² (Pap 2019). In another judgment,¹³ the very same ECHR accepted that Gypsies constituted a distinct ethnic group in Britain. In Ireland, Travelers have formally been recognized as an ethnic minority, and Romani and Gypsy (used interchangeably) and Irish Travelers have been coded as “ethnic” groups under the Race [sic!] Relations Act in the UK.¹⁴ The House of Lords also held¹⁵ that a local council’s refusal to let public rooms to a Gypsy family for a wedding amounted to discrimination on racial grounds under the Race Relations Act.

In Hungary, in addition to the international conceptualization framework, which is relevant insofar as it creates policy-making obligations, in the national political realm, Roma are also conceptualized as a national minority. As mentioned above, in 1993, Hungary adopted a comprehensive law on the rights of national and ethnic minorities. In 2011, this was replaced with a slightly modified version, amending its scope to refer only to national minorities. The Roma are among the 13 recognized communities for which the Act guarantees preferential seats in parliament, cultural and linguistic rights, and establishes a unique Hungarian institution, the minority/nationality self-government. These self-governments are elected bodies that operate at the local, regional, and national levels and have special competences with regard to protecting cultural heritage and language use, fixing the calendar for festivals and celebrations, fostering the preservation of traditions, participating in public education, managing public theatres, libraries, and science and arts institutions; awarding study grants; and providing services to the community (Darquennes et al. 2012).

Thus, Roma are conceptualized simultaneously as an ethnic group, a racialized minority, a national minority, and a socially disadvantaged group. Also, all of the operationalizing techniques coexist: the realm of politics (electing minority self-governments and MPs) is largely shaped by self-identification. Third-party (external) identification is applied in several policy arenas, including the educational sphere. Sociological surveys frequently apply third-party identification or a combination of self- and third-party identification of Roma ethnic origin (Chopin, Farkas, and Germaine 2014; Messing 2014a). NGOs and law enforcement bodies (courts, equality bodies, and the ombudsman) also collect perceived ethnic data when adjudicating discrimination complaints (Chopin, Farkas, and Germaine 2014). External classification is also used in adoption cases in Hungary, where prospective parents can refuse to adopt Roma children based on the informal classification of the child’s ethnicity by professionals in the child protection system (Pap 2023). Endorsements by community organizations are a widely used practice, for example in Roma scholarship programs, where Roma organizations validate information about ethnic belonging of scholarship applicants (Pap 2020; Pap 2023). Designing proxies for ethnicity is yet another strategy of identifying Roma group membership. This was done for the purposes of strategic litigation concerning educational desegregation, where residence along with first and last names have been used to identify Roma pupils (Chopin, Farkas, and Germaine 2014; Farkas 2017). This method has been approved by the Data Protection Commissioner (see Miller et al. 2008).

There is intense debate among academics, community representatives, and stakeholders about classification schemes in statistics and demography (Horváth, Landau, and Szalai 2000; Messing 2014a). For example, the above-noted difference in the estimated size of the Roma community depends on whether counting is based on self-identification or external/third-party classification (Messing 2014a; Rughiniş 2010), which, on the other hand, has significant consequences for redistributive public policies. Debates about ethnic statistics used in equality policies go far beyond the case of the Roma in Central-East Europe and are an issue in most ethnically diverse societies. In the introductory article of the ERS Special Issue “Accounting for ethnic and racial diversity: the challenge of enumeration”, Simon and Piché (2012) claim that “rising political and social demand for evidence-based policies […] require[s] precise definitions of group boundaries”. The latter maintain that “there is as yet no consensus on the best ways to collect such data” since competing paradigms about ethno-racial diversity – i.e. “assimilationist”, “multiculturalist” and “anti-discrimination” – call for different approaches to data collection. This is very much the case for the Roma in Hungary, too (Krizsán 2012).

4. Consequences of the confused and overlapping conceptualization and operationalization of the Hungarian Roma

This section will show that the multiplicity of categories of race, ethnicity, and national minorities that intersect with socially disadvantaged positions results in inadequate policies and, therefore, deficient execution. Often, legal tools developed as instruments for minority protection and social inclusion are, in practice, abused, mostly by members of the majority community (Pap 2015; Pap 2023; Pap András 2018). In this section, we provide three examples from different areas – political representation in (national) minority self-governments, education, and employment – to highlight the consequences of conceptualization and operationalization choices and the further complications caused by the widespread misinterpretation of data protection law that we refer to as data-processing phobia.

4.1. Political field: minority self-government

The first case study will show that conceptualizing the Roma as a national minority is just as problematic as relying on an uncontrolled self-identification operationalizing scheme for the election of the unique minority self-governments because it allows for and gives rise to widespread abuse by non-Roma.

In Hungary, in order to contextualize minority policies, we need to look back a century, to 1920, when due to the Treaty of Versailles that concluded the First World War, Hungary lost two-thirds of its territory and the corresponding population to its neighbouring states. Ever since a sense of responsibility for ethnic kin in the neighbouring countries has been a cornerstone of politics, and after the political transition in 1989, this was further seen as a constitutional responsibility and a foreign policy priority. Bilateral agreements on minority rights were signed with neighbouring states, and Hungary joined Council of Europe minority rights treaties and conventions on national minorities (such as the Framework Convention for the Protection of National Minorities and the European Charter for Regional or Minority Languages) (See Pap 2019).

The comprehensive law on the rights of national and ethnic minorities was adopted in 1993 and enumerated 13 recognized minorities, the Roma among them. The Act guarantees cultural and linguistic rights for these groups and establishes a unique Hungarian institution, the minority self-government (MSG). Funded by the local authorities or by the State when national-level bodies are concerned, MSGs are elected bodies that operate at the local, regional, and national levels. The function and design of MSGs are generally seen as controversial: they involve a combination of institutions fostering political representation, empowerment, and cultural autonomy.

While Roma MSGs arguably serve as “training schools” for up-and-coming Romani politicians, this model is widely (Dobos 2020; 2022) understood as inappropriate for addressing the social marginalization of the Roma, which many Roma see as their most important goal. Hungarian Roma leaders have repeatedly called for redistribution-oriented, rather than recognition-oriented, minority policy. According to Molnár and Schaft (2003, 41), in practice, “Roma self-governments see as their main objective the improvement of social conditions in their community rather than the preservation of minority culture and the strengthening of minority identity”. Kóczé and Rövid (2017) refer to this contradiction of recognition- and redistribution-oriented politics as the “double discourse of Roma politics”. In sum, conceptualizing Roma issues as identity politics fails to comprehend the complexity of the position of Roma communities and cannot address either anti-discrimination or social inclusion challenges.

The operationalizing scheme applied for MSG elections is equally problematic, as self-identification is the dominant method, which has led to widespread abuse. If Hungarian citizens are willing to spend some time navigating the bureaucracy, regardless of their ethnic origin or identity, they can run as and vote for minority self-government candidates. Such schemes for operationalizing minority group membership and passive and active voting rights for MSGs may corrupt the very system of representation since – as some examples show – it may lead to the abuse of elections by (ethnic) majority politicians. Carstocea (2011) has stated that about 40 per cent of the Romanian self-governments in Hungary were reported to be headed by non-Romanians. Another example is that of the National Ukrainian MSG, the members of which requested that no Ukrainian be spoken during official sessions because they did not understand it. Yet another case showing how the system of minority self-government may be abused is that of the Roma minority self-government in Jászladány, a village notorious for segregating Roma, including in its schools. Here, the non-Roma wife of the mayor successfully won elected office in the local Roma minority self-government due to the votes of local non-Roma. To demonstrate the fallacies of this legal framework and the loopholes that often lead to the abuse of the system of representation, some Roma politicians publicly decided to run under different labels (in most of the reported 17 cases, they ran as Slovakians) (Pap 2015).

4.2. Education

Education is another field in which confused conceptualizations that shift between the categories of nationality, ethnicity, race, and the socially disadvantaged create everyday dilemmas and have considerable long-term consequences for children and adults with a Roma background. Education policies, as well as actors in education (principals, teachers), struggle with simultaneously addressing minority cultural needs and the socially disadvantaged position of children and meeting the special educational needs of students that often stem from experiences of discrimination and social deprivation. These three issues call for different conceptualizations and consequent classification strategies.

The performance gap between Roma and non-Roma children has multiple, complex, and intersecting causes, including ethno-racial discrimination, the ethnic gap in social position, and the lack of recognition of Roma culture (Kende and Szalai 2018; Kertesi and Kézdi 2011). Depending on the identification of the core cause of educational disadvantages, policymakers need to address Roma students as members of a racialized group suffering discrimination, as members of an ethnic community with a need to maintain cultural heritage, as socially disadvantaged students, or as children with special educational needs. In reality, a cacophony of approaches and practices are at work, some addressing social disadvantages, others recognizing cultural minority rights, while imprecise or incomplete conceptualizations become grounds for ill-designed measures.

A salient example is how the Hungarian state and actors in education address ethno-social segregation in schools. The segregation of Roma children, especially if they come from socially marginalized families, is one of the most appalling processes associated with contemporary education and has remained largely unresolved over the past three decades. Segregation is the most widespread form of discrimination against Roma children, inflicting life-long consequences; as such, it is a violation of their human rights (Durst et al. 2022; Kertesi and Kézdi 2014; Messing and Árendás 2022). At the same time, it creates immense economic costs (Kertesi and Kézdi 2006) and is detrimental to social cohesion (Szalai 2020). At the core of this process lies what we have referred to as Murphy's law of data protection: while actors in education (school personnel, school management, and parents) claim that they do not and cannot know individual children’s ethnic background and thus differentiate between children, still, they do have “subjective knowledge” (perceptions) about who the Roma children are. When it comes to allocating children between schools or classes, they do not hesitate to group children perceived as non-Roma into schools of higher quality or classes with higher prestige (maths, science, or foreign language specialization), while Roma children are left in lower quality and low-prestige educational units. This occurs in primary schools during the selection and placement process when parents exercise their right to the free choice of schooling and generate “white flight” based on their perception of where children of Roma families will be schooled (Kertesi and Kézdi 2014; Messing 2014b, 2017; Neumann 2017; Zolnay 2018).

Paradoxically, practicing minority rights related to the ethnic culture of self-identified Roma families became another less significant but prevailing technique of segregating Roma children in schools. There have been several instances when Roma parents were pressured by school management to request specialized minority education because this would provide additional funding for the school (Balogh 2012; Balogh et al. 2013; Messing 2017). Most frequently, such minority education means low-quality Roma folklore classes once a week in separate, segregated, and often worse-equipped classrooms with inferior conditions and infrastructure; in less common but worse scenarios, it becomes a means of entirely segregating Roma children within the educational setting (Lakatos 2010).

Summing up the above cacophony of approaches and practices, we can identify the parallel existence of self-declared and perceived belonging to the Roma minority, both leading to ethno-social segregation in schools. When parents ask for minority education, they declare that their child belongs to the Roma minority. If they do not practice their right to minority education, the child’s “Romaness” is still observed and plays an important role in everyday practices of ethno-racial differentiation and segregation. Although on an individual level, schools and school personnel do not collect data on or record the ethnic background of children, they make subjective diagnoses of the ethnicity of individual children, which play a role in placement practices. A further element of the cacophony of identification schemes is described by Farkas (2017). She explains how the yearly competence assessment of students incorporates an institutional (school level) questionnaire in which – in addition to a number of institutional information – principals need to give an informed estimation about the proportion of Roma students. Finally, and most importantly, social status is the most important proxy of Roma background in everyday practices of schools. The “socially disadvantaged” category refers to the social needs of the parents (claiming social assistance), while the “multiple socially disadvantaged” label refers to children brought up in families in social need whose parents have very low educational attainment (ISCED2 or lower). In the everyday practices of schools, this latter category is used as a proxy for Roma (Kertesi and Kézdi 2011; Messing 2014b; Zolnay 2018).

4.3. Employment

The labour market offers yet another illustration of how the above-described cacophony of categories and related puzzles of operationalizing the target for equality measures obscures stakeholders’ focus and results in a lack of efficient action to promote Roma’s access to employment.

In the labour market, being Roma is relevant because of their disadvantaged social position, mainly due to structural, institutional, and often also individual-level discrimination. Thus, here, nationality plays no role, while racialized difference-making as well as a disadvantaged social position are more prominent framings of Roma positionality. A broad review of studies that discuss the issue of approaches to Roma people in labour market policies highlights that even though policymakers acknowledge the multiple disadvantages that people of Roma heritage suffer when trying to access the labour market, they do not target Roma explicitly but instead apply a colour-blind approach (Messing and Bereményi 2017).

Promoting colour-blindness (i.e. policies insensitive to ethno-racial belonging) is the dominant approach of the Hungarian state. Active labour market programs (ALMP) operated by the state usually target people in vulnerable positions (for example, unemployed youth or those in marginalized economic positions) irrespective of race or ethnic background and thus address the consequences of ethnic or racial discrimination only to the extent to which they contribute to the vulnerability of an individual’s position. This approach assumes that a vulnerable position in the labour market stems from other disadvantages (age, education, regional inequalities in the economy) and should not be perceived as related to race or ethnicity. Consequently, LMP's application of such an approach fails to address disadvantages stemming from discrimination. Research assessing labour market policy impact proves that colour-blind policies run the risk of excluding unemployed Roma from the pool of beneficiaries because they disregard ethno-racial discrimination and the hierarchical power relations this stems from (Adamecz et al. 2013; Messing and Bereményi 2017). This process is often referred to as colour-blind racism (Bonilla-Silva 2006).

Few labour market measures/programs in Hungary that had an explicit aim to include Roma. One of the largest was launched in 2009 following the Ombudsman for the Rights of National and Ethnic Minorities’ argument that targeting members of the Roma minority using employment programs is legitimate because the Hungarian Roma have been experiencing discrimination in the labour market, so they have greater difficulty accessing quality jobs in the public and private sector. The program, entitled “Roma’s Employment in the Public Administration and the Judiciary”, aimed to include self-identified Roma with a tertiary education. The program, however, failed to include Roma due to the conceptualization scheme that allowed for the uncontrolled self-identification of applicants. According to a report by the Ombudsman, it was largely members of the majority who benefited from the program due to the inadequate operationalization of the target group and the lack of proper cooperation with and outreach to Roma communities (Ombudsman 2010). One of the six suggestions the Ombudsman made in its report was elaborating on the legal procedures associated with the operationalization and conceptualization of Romaness related to affirmative action in employment. There has been no follow-up to this.

Honing in on the hiring practices in firms, we see the dominance of blindness to ethnicity and race. A recent study in Hungary about including ethnicity and/or race in firms’ diversity strategies and practices revealed that ethnicity and race are not high on organizational agendas (Árendás, Messing, and Zentai 2018). When talking about efforts to hire Roma employees, HR professionals emphasized meritocratic principles and claimed that they evaluated applicants exclusively based on their performance; ethnicity, race, and socially disadvantaged background were not considered. The research identified two important obstacles to the hiring of Roma by employers willing to support this in their firms: (1) misinterpreted data-protection legislation, i.e. the same data-processing phobia that we described earlier in this paper (chapter 2.2), and (2) misinterpretation of the principle of non-discrimination. Some company representatives emphasized that recognizing Roma ethnicity and applying any supportive measures would breach the rule of non-discrimination (Árendás, Messing, and Zentai 2018).

5. Concluding remarks and normative guidance

The case of the Hungarian Roma arguably demonstrates the consequences of their confused and overlapping conceptualization as an ethnic group, as a community experiencing racialization, as a national minority, and the confusing cacophony of operationalizing schemes that operate in a climate of ethnic-data-processing phobia. This leaves the Roma between the Scylla of the abuse of ethnic identification (in political, educational, and labour market institutions designed for the national minority) and the Charybdis of Murphy’s law of data protection (the phenomenon that discriminators never have difficulty identifying their victims, but when it comes to remedial action, stakeholders are often unable or unwilling to navigate privacy regimes). Furthering earlier discussions, we claim that normative guidance should be followed and specific operationalizing methods should be paired with specific policy and legal instruments. The noted paradox that despite the (often) broad set of preferences and privileges awarded to the respective groups, many legal regimes will not provide guidance for conceptualization or definitions of membership in the latter has severe consequences. This puts policy implementation at risk and leads to a spiral of discrimination and marginalization, making the instruments susceptible to misuse.

Regarding how to navigate out of the conceptual and practical terminological labyrinth, two points should be made. First, a useful set of terminology should center on the substance of legal and policy claims and frameworks. In line with this approach, there are three clusters. First, minority rights that both recognize and accommodate the cultural claims of groups and individuals. The second array of legal and policy frameworks are oriented towards individual rights and focus on non-discrimination. The latter term should be understood broadly to include protection from hate crime and hate speech and cover several other related individual-based human rights claims (including consequences directly associated with experiences of discrimination). The third set include those diverse social inclusion measures that “ethnicize” social policies or, when endorsing multiculturalism, recognize other forms of group-based, collective claims. “National minorities” and “nationalities” are adequate terminologies for the first cluster, “racial” and “ethnic” minorities are adequate for the second, while the third approach institutionalizes a curious mix of all three clusters.

In this sense, racial-ethnic duplication is redundant because here – i.e. in Central East Europe – “ethnicity” as a sign of cultural difference is only relevant if it is a basis of maltreatment, exclusion, or institutionalized marginalization. However, “class” emerges as a relevant category due to how it is racialized, i.e. how poverty is transformed into ethnicity or race by making it (again) visible.

Consequently, each policy goal and legal instrument will require a specific operationalization scheme. The need to define membership criteria arises in a completely different way when group formation is based on claims for different kinds of preferences and privileges. In the domain of discrimination or hate crime, the perpetrator’s perception of the group membership of the victim should lead the classification strategy, whereas concerning affirmative action (remedial preferences), self-identification combined with community identification or endorsement is the adequate approach (Pap 2023).

The second point pertains to privacy/data protection concerns. The over-cautious approach to tackling and operationalizing Romaness is present in all areas of policies and leads to a (legally unsubstantiated) reluctance to tackle institutional marginalization and discrimination. This reluctance by authorities or policy actors to recognize and register ethnicity obstructs efforts to combat educational desegregation or focused labour market support (as the system fails to see Roma students or Roma unemployed as a target group for policies). Data protection arguments are also used as an excuse for not monitoring ethnic profiling (over-policing), as well as for not prosecuting racially-motivated hate crimes (under-policing) or the lack of recognition of Roma ethnicity in the context of labour market discrimination. Concurrently, data protection arguments significantly limit the potential for extending equality and diversity measures in the labour market to the Roma.

It is important to reiterate that even the strictest data protection/privacy regimes allow “talking about” race and ethnicity, i.e. processing ethno-racial data under certain circumstances. Thus, ethnic-classification phobia is legally unfounded. In 2009, the Hungarian data protection and the minority rights ombudsmen published a joint recommendation on ethnic data collection. The document points out that the right to equal treatment takes a priority over the protection of personal data when investigating hate crimes and sanctioning discrimination (Chopin, Farkas, and Germaine 2014). The recommendation holds that discrimination and hate crimes revolve around external classifications of ethnicity (as perceived by the perpetrator), thus here self-identification is of little relevance (Pap 2020).

Finally, by bringing together the perspective of social sciences and the law in this paper, we argue that the consistency of the conceptualization of “what the Roma are” is crucial in terms of how policies for Roma are designed. Conceptualizations must be distinct when referencing rights holders of minority (cultural) rights and when addressed to beneficiaries of social inclusion policies or victims of discrimination. The inconsistent labeling of the Roma as an ethnic, racial, or national minority reflects a lack of consistent conceptualization and determination of whether social inclusion, anti-discrimination, or a cultural rights-oriented approach should dominate policies.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by European Commission, DG Employment Social Affairs and Inclusion [grant number: VS/2016/0236]; Hungarian Research and Innovation Grant [grant number: 134962]; European Commission, Horizon Programme. Grant agreement ID: 101094684 “RAISE. Recognition and Acknowledgment of Injustice to Strengthen Equality”.

Notes

1 Act LXXVII of 1993 on the Rights of National and Ethnic Minorities.

2 A disclaimer is in order: despite a streamlined effort to use precise terminology throughout the text, occasionally, when making distinctions would obstruct conveying the message, “race” and “ethnicity” are used interchangeably, and a conflated “ethno-racial” designation is used. This is also justified by the fact that in legal text and practice, legislators and judges, as explained below, apparently often treat the terms as synonymous.

3 Due to spatial constraints, the discussion of related concepts and terms like colour and caste (also widely used in various domestic and international legislative and policy documents) remains outside the scope of analysis.

4 https://assembly.coe.int/nw/xml/XRef/Xref-XML2HTML-en.asp?fileid=15235. The document was also endorsed by the European Parliament’s 2005 resolution on the protection of minorities and anti-discrimination policies in an enlarged Europe.

5 http://ec.europa.eu/justice/policies/privacy/docs/wpdocs/2007/wp136_en.pdf

6 Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation)

7 https://www.gdprsummary.com/gdpr-definitions/article-9/?amp

8 (i) if the data subject provides explicit consent for specific purposes; (ii) for the carrying out of obligations under employment, social security or social protection law, or a collective agreement if it is authorized by law; (iii) if processing is necessary to protect the vital interests of the data subject or of another individual when the data subject is physically or legally unable to give consent; (iv) under the course of legitimate activities by a foundation, association or another non-profit body with a political, philosophical, religious or trade union aim in relation to members of such bodies or persons who have regular contact with it; (v) if data is manifestly made public by the data subject; (vi) if processing is necessary for the establishment, exercise or defence of legal claims or whenever courts are acting in their judicial capacity; (vii) under “reasons of substantial public interest with a basis in law”, appropriate to the aim pursued and contain appropriate safeguards measures. (Here examples include journalism, academia, art, literature, anti-doping in sport or statutory or government purposes); (viii) health or social care with a basis in law, say, occupational or preventative medicine or for assessing the working capacity of the employee; (ix) public health with a basis in law, say, to ensure quality and safety of health care and of medicinal products or medical devices; and (x) archiving, research and statistics with a basis in law.

9 From a historical perspective, an exception would be Jewry, who experienced ethnicization, racialization, and prosecution in the region.

10 It is important to add that the essentializing and homogenizing legislative classification of the “Roma national minority” also disregards the plurality and non-essentialism of the Roma community.

11 Koptova v. Slovakia (OHCHR 2000) and Lacko v. Slovakia (OHCHR 2001).

12 Case of DH and Others v. The Czech Republic (2007).

13 Chapman v. The United Kingdom (2001).

14 See, for example, Commission for Racial Equality v. Dutton (1988), McVeigh (2007), O’Leary v. Allied Domecq (2000), Wrexham County Borough Council (Appellants) v. Berry (Respondent), South Bucks District Council (Appellants) v. Porter and another (FC) (Respondents), Chichester District Council (Appellants) v. Searle and others (Respondents) (Consolidated Appeals) (2003, paragraph 41).

15 Hallam v. Cheltenham Borough Council and Others (2001).