Governing racial justice through standards and the birth of ‘White diversity’: a Foucauldian perspective

ABSTRACT

Drawing on a longitudinal qualitative approach to corporate diversity policies in France, based on more than 80 in-depth interviews (N = 86), this paper examines the paradox conveyed within these policies by the rise of ‘raceless’ diversity concepts. This is what I term White diversity, exploring its construction and appropriation in the light of technologies of normalisation, referred to in English as standards. Building on a Foucauldian approach to normalisation, the paper engages with the case study of the French Diversity Label. It explores the ways in which market-based mechanisms of regulation have shaped the management of race difference. It demonstrates in particular how normalising antidiscrimination through voluntary social certification has contributed to upholding whiteness in organisations.

KEYWORDS:

• Corporate diversity
• discrimination
• normalisation
• social certification
• race
• France

Introduction

Since the early 2000s and under the aegis of the European Union, France has gradually built up a legal and political antidiscrimination framework (Guiraudon, 2004, 2009). Among other actors, companies here have been the very first to translate these new legal provisions into categories of policy-making and organisational change, by (re)inventing the diversity framework (Bender, 2004; Bereni & Jaunait, 2009; Doytcheva, 2008). In a gesture of ‘mimetic isomorphism’ (DiMaggio & Powell, 1983), they followed the path set by multinational, American and Canadian firms (Chicha & Charest, 2009; Dobbin, 2009), and the recommendations of international actors and European authorities (European Commission, 2003, 2005). Infusing conceptions of law with managerial values, this shift from antidiscrimination to diversity management has been commonly construed in terms of ‘managerialization of law’ (Edelman et al., 2001; see also Edelman, 2016; Kelly & Dobbin, 1998; Dobbin, 2009; Gribling & Smith, 2014).

As I have shown elsewhere (Doytcheva, 2009, 2010), in French settings, but consistent with international research, this shift involved, on the one hand, the deployment of multi-criteria diversity definitions, labelled ‘inclusive’ or ‘generic’, which importantly expand upon formal legal anti-discriminatory concerns. On the other hand, it fostered the rise of an entrepreneurial or managerial prism, focused on rationales of profit, competitiveness, and ‘the well understood interest of companies’ (Bébéar, 2004) as the main drivers for organisational change. Following a now global path, the ‘transcoding’ (Lascoumes, 1996) of antidiscrimination into diversity management fostered an instrumental approach to differences (Zanoni et al., 2010), which not only distances the intrinsic normativity of law (Doytcheva, 2010) but also, more broadly, considerations of social justice and equality. In terms of policy-making, this approach translated into the implementation of so-called ‘soft-law’ measures and policy instruments, such as charters, statements of commitment, voluntary codes, and quality standards. Hailed as less coercive, more flexible and effective, these techniques of self-regulation have deliberately and further decentred the corporate agenda on equal opportunity towards an ever-increasing exposure to marketing rationales and strategies – built on gaining value, segmenting markets, branding, and preventing hostile campaigns.

Yet, these developments in framing diversity at the beginning of the 21^st century should be recast within a broader tendency of late neoliberal capitalism towards ‘governing by standards’ – technical rather than legal (Thévenot, 1997, 2009; see also Guthman, 2007; Brown & Getz, 2008). According to Thévenot, at the turn of the century, traditional normativity, ensured by national law and the judiciary underwent a major change. With the overwhelming position granted to market co-ordination, market mechanisms are expected to inscribe into the ‘certified properties of things’ concerns for fundamental goods, such as safety, health, the environment, and now racial justice and equal opportunity. The proliferation of global standards since the early 1990s has been a key feature of globalisation (Mutersbaugh, 2005; Nadvi & Wältring, 2002); while, according to many analysts, their profusion is far from being merely ‘coincidental’ with ‘rollout neoliberalization’ (Guthman, 2007; Peck & Tickell, 2002). Upstream, standards have been strategic in achieving the European single market, in the absence of political consensus and more binding instruments for change (Borraz, 2007; Majone, 1994).¹ However, under these new regulatory schemes of de facto private governance, the common goods for which citizens engage in political arenas are reduced to commodity properties. The normativity of law is transferred to the ‘normalisation of objects’, which become the carriers of a commodified and distributed authority. Following Thévenot (2009), a sharp distinction must be drawn between this new model of privatised governance and that of a public one, whereby social entitlements and citizens’ rights tend to be displaced by consumer choice.

In this article, I examine the effects induced by the growth of these new regulatory regimes, and standardisation in particular, on the shaping of equal opportunity policies in the corporate world. Based on the case study of the French Diversity Label, a national standard generated with support from the French government in the mid-2000s, I explore the ways in which techniques of normalisation have been applied to manage and reshape the boundaries of race difference – defining who is worthy of inclusion and who is not. Whereas purportedly designed to complement antidiscrimination law by certifying good practice of voluntary organisations, what transformations have these techniques brought about in the transition from legal to market realms? What kind of ‘sociotechnical co-ordination’ (Mallard, 2000) underpins these (self-)regulatory schemes? To what extent then have they contributed to the institutionalisation of specific diversity norms? – marked, as I will argue with regard to the French case, by their paradoxical whiteness, i.e. the eviction of race and ethnicity from the space of corporate diversity and the rise of ‘raceless’ diversity concepts.

To make the case for this, I take advantage of what Klarsfeld (2009, p. 363) considers a ‘French regulation school’, which views these problems as a ‘process of designing new rules’.

More precisely, I draw on an analytical framework that brings together scholarship on governance and policy instruments (Borraz, 2004, 2007; Halperm et al., 2014; Lascoumes & Le Galès, 2004), with science and technology studies (Cochoy, 2000; Mallard, 2000; Réseaux, 2000), and a conventionalist approach (Boltanski & Thévenot, 2006; Thévenot, 2009) to standardisation, in particular. According to these theoretical perspectives, far from an alleged neutrality, policy instruments and technical arrangements should be considered rather as a dispositif (Foucault, 1980²). Borrowing from a Foucauldian understanding of knowledge as practice, technological procedures are seen as embedded in contested interests and power relations.

In parallel with these perspectives, I draw here on a more thorough understanding of Michel Foucault’s (2007, 2004a, 2004b) concepts of norm, normation, normality (normalité), normalisation. This Foucauldian approach seems all the more relevant since, as of 2004, the almost unique public strategy against workplace discrimination in France has primarily centred on techniques of standardisation, referred to as normalisation in French (Doytcheva, 2018b). But, from a Foucauldian perspective, a much more complex relation should be envisioned between the ‘rule’ (of law) and its subsequently derived technical or managerial norms. This relationship goes well beyond, or far beneath, a purported ‘complementarity’. Borrowing from Foucault’s words, against the backdrop of late neoliberal governmentality, techniques of normalisation, which consist of bringing an ‘average normality’ into play by presenting it precisely as ‘normal’ – i.e. ‘suitable’, ‘acceptable’, ‘desirable’ – have led to the construction of what I refer to in this article as normalised diversity, one of whose remarkable characteristics is that of being predominantly White.³

As such, my approach falls within the area of critical diversity studies, a bourgeoning field that emerged from the mid-1990s onwards as a reaction to the re-appropriation of equal opportunities by business (Zanoni et al., 2010). It converges with critical management and critical marketing perspectives (Tadajewski, 2010) to underscore the importance of marketplace power relations (Johnson et al., 2017), and the ways in which history, social hierarchies, power, and privilege shape their participants. These questions acquire further significance as markets are today conceded as the locus of social regulation and are expected to achieve functions of wider social interest and utility. The case study stresses, in particular, the performative nature of ‘marketing knowledge’ (Roscoe, 2015; see also Cochoy, 1998) and ‘market subjectivities’ (Skålén, 2009; Varman et al., 2011) – that is their propensity to perform markets where there were (are) none. As such, it calls for a deeper scrutiny of the feedback or loop effects that these mechanisms exert on other governmentality forms and regulatory systems, notably legal and distributive ones. From this position, this paper shows how neoliberalising equal opportunity not only stands for managerialising antidiscrimination but equally, and not less importantly, for deracialising diversity. In what follows, I proceed by reviewing Foucault’s conception of normalisation as well as previous research that has drawn on notions of normalisation as a governance form. Then, I turn to the method used, followed by the case description, which is intertwined with an analysis of the relationship between normalising antidiscrimination and unfolding ‘White diversity’. I highlight how this process is achieved through three main functions – i.e. governance, instrumental rationality and valuation, and devolution of responsibility – that connect it to contemporary neoliberal doctrines.

Theoretical considerations

Framing diversity as the new business paradigm for differences should certainly be traced back to the mid-1980s and the US corporate world, where the publication of the Workforce 2000 Report urged firms and policy-makers to address these pressing concerns. For the first time within organisational fields, the notion of diversity portrayed differences as ‘strategic assets’, which, if well managed, could provide a ‘competitive advantage’ (Zanoni et al., 2010, p. 12). Coined by Edelman et al. (2001), the concept of ‘managerialization of law’ sought to address these managerial constructions of law (and especially of civil rights law), underscoring how they enabled managers to dissociate their efforts from legal ideals and social justice goals. From this perspective, ‘diversity is directly valuable to organizational efficiency and important in its own right’ (ibid., p. 1591), instead of being a ‘mere translation’ of legal or social concerns.

In this article, in addition to or in parallel with the US literature on ‘managerialization of law’, I thus ask what the consequences are of the ‘normalisation of law’? – that is its increasing exposure to market-driven governance forms. Joining a broader trend in the study of market dynamics towards shifting the lens from material markets devices to governmentality (Roscoe, 2015), in what follows, I return to Foucault’s conceptualisation of neoliberal governmentality to better explain: how technologies of normalisation (dis)embody the law; how they are involved with upholding privilege and discrimination (Crockett et al., 2003; Johnson et al., 2017) in the marketplace, and the justification of social and racial hierarchies elsewhere (Kehal, 2019).

Normation vs normalisation: a Foucauldian perspective

More precisely, I draw here on the distinction postulated by Foucault (2007; 2004a) in one of his first lectures at the Collège de France, in 1977–1978, between two different forms of normalisation – i.e. one of a set of techniques designed to exert power and control that are core to modern governmentality (see also Johnson et al., 2017; Roscoe, 2015; Varman et al., 2011). Incidentally, Foucault coins a new term to highlight the distinctive nature of two competing processes, namely normation, as opposed to normalisation. How does one normalise, he asks, in different societies and regimes? In disciplinary societies and rationality, which are based on the repression of any deviation from the norm (Foucault, 1977), normalisation should be understood above all as the search for an ‘optimum’. ‘Disciplinary normalisation’ consists of first positing a model considered as optimal in terms of a certain result, and then trying to get people, movements, and actions to conform with it. As Foucault writes:

I think it is indisputable or hardly disputable that discipline normalizes. But we must be clear about the specificity of disciplinary normalization […] In the disciplines, one started from a norm, and it was in relation to the training (dressage) carried out in reference to the norm that the normal could be distinguished from the abnormal. (Foucault, 2007, p. 57)

In this regime of governmentality, the norm comes first, with ‘the normal’ being precisely that which is capable of conforming to the norm. This is why disciplinary normalisation should be better understood as ‘normation’ – or the process of setting up a norm intended to be replicated – in order to underline the primary and fundamental character of the norm. In (neo)liberal governmentality, however, centred on ‘apparatuses of security’, how does normalisation work? In this governmentality regime, Foucault argues, normalisation is far less about issuing and imposing a general norm. Instead, it will primarily consist first, of an empirical study of the ‘different distributions of normality’, in order to establish a dynamic interplay between them, second. The norm is an ‘interplay of differential normalities’. In this regime, the detection of ‘the normal comes first and the norm is deduced from it’:

We have a system that is, I believe, exactly the opposite of the one we have seen with the disciplines […] Here, instead, we have a plotting of the normal and the abnormal, of different curves of normality, and the operation of normalization consists in establishing an interplay between these different distributions of normality and [in] acting to bring the most unfavorable in line with those that are more favorable […] So we have here something that starts from the normal and makes use of certain distributions considered to be, if you like, more normal than others. The norm is fixed and plays its operational role on the basis of this study of normalities. The norm is an interplay of differential normalities. The normal comes first and the norm is deduced from it. (Foucault, 2007, p. 63)

In his further discussion, Foucault points to the fact that the norm should not be confused with the ‘rule’ (of law): between the two, he argues, ‘there is and could not fail to be a fundamental relationship’, since ‘every system of law refers to a system of norms’. But the ‘intrinsic normativity of law’ has nothing to do with normalisation: ‘From and below a system of law, in its margins, and maybe even against it’ (Foucault, 2007, p. 57), procedures, processes and techniques of normalisation are developed. Governing by norms or standards thus takes over from reglementation, and the dirigiste state, replacing the overall imposition of an order with the internal play of ‘differential normalities’. Within apparatuses of security, mechanisms of power and governmentality are no longer based on territorial domination, ‘fixing’ and ‘demarcation’, but on movement, contact, exchange, and circulation. They do not function on the axis of sovereign-subjects relationship; and do not tend to a ‘nullification’ or a form of prohibition either, but ‘to the progressive self-cancellation of the phenomena by the phenomena themselves.’ (ibid., p. 66) Where disciplinary rationality ‘encloses, fixes and confines’, the neoliberal one organises instead the ‘circulation’ of commodities, consumers and production (Roscoe, 2015).⁴

The rise of this ‘smart regulation’ accompanies the neoliberal turn (Jobert, 1994; Lascoumes & Le Galès, 2004), as classic command and control interventions give way to essentially incentive-based and negotiated techniques. Praised for their supposed ‘efficiency’, these latter techniques flourish in a profusion of innovation, competitiveness, and creativity. Although deregulation and privatisation are understood to have created the conditions for this shift (Majone, 1994), the ‘regulatory state’ is said to ‘transcend privatisation debates’ (Levi-Faur et al., 2004). It plunges into a ‘regulatory society’ or ‘audit society’, wherein ‘inspectability'’ becomes the hallmark of an economy based upon state-enforced private accumulation’ (Mutersbaugh, 2005, p. 2048); and where reified and distributed or spacialised power and authority embody patterns of Bentham’s panopticism.

Normalisation as a governance form

As public policy instruments, technical norms or standards exhibit several characteristics: they involve various ‘interested parties’ or stakeholders (Borraz, 2004, 2007), and organise the co-ordination, possibly the convergence, of ‘sociotechnical networks’, composed of industrials, experts, NGOs, and consumer and user groups, in addition to public authorities (Mallard, 2000). Their legitimacy is therefore twofold: scientific and/or technological, on the one hand and democratic, on the other, insofar as they rely on consensus, understood here as a general agreement characterised by the absence of strong opposition (Borraz, 2004). Norms are therefore referred to as ‘technological diplomacy’ (Cochoy, 2000; Hawkins, 1995) or ‘technological democracy’ (Kessous, 2000). Finally, they remain of voluntary application.

The spread of these regulatory schemes in a wide range of social activities – from transnational market governance to increasingly environmental and social concerns – reveals the trend for public authorities to delegate their regulatory powers to private actors. However, technical norms are rather a re-regulation (Majone, 1994, p. 97). In theory, they form a particular category of regulatory instruments, whereby private actors operate as ‘government representatives’ to bolster regulatory goals and pursue the objectives of the latter (Egan, 2001).⁵

The political integration of normalisation tends to vary, however. At European level, where this strategy is highly valued,⁶ the partition of tasks between public authorities and private actors involves a clear distinction – objective-setting by the former, and then the means to achieve them by the latter. In the French case, however, the situation seems at least quite ‘entangled’ (Borraz, 2004, p. 158); although, as demonstrated by research, the making and enforcement of standards is also highly politicised elsewhere (Campbell & Liepins, 2001; Guthman, 2007). With the principles of consensus, participation and expertise being much less bold in France, the development of standards-based regulation is embedded into a broader strategy of ‘state reform’. According to De la Broise and Lamarche (2006), it seems thus closer to a process of responsibility devolution, whereby social actors are entrusted to enforce regulations, regulate their activity, and even to self-regulate.

Finally, these changing political configurations have an impact on the very nature of normalisation over the long term. During the 20^th century, normalisation has moved from industrial to marketable logics (Cochoy, 2000), in particular through the development of certification. At its origins, in 1926, AFNOR (Association Française de Normalisation), which is the French Standards Association, and a member of the International Standards Organisation (ISO), was designed above all as a technical tool for ‘good industrial entente’. Its mission is three-fold: ‘unify’, ‘simplify’, and ‘specify’. But the conformity to a norm (e. g. dimensional), which is central to the industrial logic of normalisation, will gradually give way to the ‘qualifying’ of services (ibid., p. 66; see also Mutersbaugh, 2005; Nadvi & Wältring, 2002), i.e. the recognition of specific quality (e. g. in health, safety or environmental matters). This quality recognition, often achieved by the affixing of a label or marking, acts first and foremost towards the marketplace, and consumers in particular.

Drawing on labelling or marking, standardisation has succeeded in leaving the confined space of industrial engineering and moving forward to the conquest of markets, and soon of public policies. Building on the example of EC marking, French certification underwent a similar process of differentiation, expanding its activities into new fields including the environment (since 1992), food production (1994), and the service sector (1994). In this article, I focus upon certification norms that target social justice and labour protections, and which I analytically distinguish from production norms and other, mainly consumer-oriented, labelling schemes (e.g., fair-trade, organics, protected origin labels). Although as the case study attests by the terminology choices at stake – those of labelling rather than standards-making – the boundaries between and within sub-categories of standards often remain blurred and unclear, especially since certification can also be used as a consumer marketing tool.⁷

Yet, the marketisation of norms eventually puts normalisation in a delicate position, namely ‘in the midst of a market for certificates’ (Cochoy, 2000, p. 85). With the withdrawal of state funding, which dates back to the early 1980s in France, standards bodies, and even more so certifiers, have increasingly become ‘enterprise service companies’. They are remunerated by business and operate within markets frames – set up, as stressed above by companies, shareholders, consumers and scoring agencies – according to approaches that do not necessarily concur with legality or equality aims (Doytcheva, 2010). From this, I will argue in the findings section, particular ways of constructing and categorising diversity will be derived.

Research context, methods and data collection

I draw here on a qualitative longitudinal approach to corporate diversity policies in France, based on three surveys that I conducted between 2006 and 2014 (in 2006–2008, 2011 and 2014). Mainly relying on in-depth sociological interviews (N = 86), this research also includes ethnographic work, based on observations, documentary study and analysis (Doytcheva, 2008, 2015). The first of these surveys (2006–2008), launched shortly after a series of pioneering corporate initiatives on diversity, was the most extensive. Whereas the subsequent ones (2011, 2014) were primarily designed to update and complement the initial dataset tracking fresh developments in the field. Overall, one of the main objectives of the research design was to include a multi-stakeholder, multi-site and multi-temporal approach in order to address these concerns over the medium term. This represents the decades 2000 and 2010, from the inception of the corporate social movement in favour of workplace diversity (Alaoui & Doytcheva, 2010) to its latest achievements. By doing so, another important aim of the research was to confront companies’ commitments, and other aspirational policy statements, with the practical outcomes of their diversity work (Ahmed, 2012).

As I will further elaborate in the concluding section, antidiscrimination emerged in the early 2000s, under EU influence and legislation, in many European countries, including France, which until then had a rather thin legal and political framework for equal opportunity.⁸ Predominantly enforced in criminal proceedings, a very small number of race discrimination cases were litigated in court, resulting in an average of about ten convictions per year (Stasi, 2004, p. 37; CNCDH, 2016, pp. 185, 250), and even less in employment cases (Doytcheva, 2018c, p. 21). In parallel, although possible in theory, civil action had remained inoperative (Latraverse & Doytcheva, 2018; Stasi, 2004), stumbling upon the imposition of the burden of proof on plaintiffs. Even though statutes prohibit discrimination, the law fails to successfully remedy the widespread and attested practices (Suk, 2008).

However, this situation was undergoing a potential change in the early 2000s with a novel antidiscrimination law,⁹ which transposes the EU legal provisions – especially burden-shifting provisions – designed to make civil litigation of discrimination more friendly to courts. At the same time, French public authorities, eager to recast their previous highly integrationist approaches, albeit for a variety of political reasons (Guiraudon, 2009), started incentivising major companies to become aware of the new ‘legal risk’.

Already targeted by several multi-actors schemes (such as the 2000 EU initiative EQUAL), major business leaders had not been slow to embrace the rationale of diversity management, in a move of ‘mimetic’ rather than ‘normative isomorphism’ (DiMaggio & Powell, 1983). Stressing the need for ‘soft action’ – but quite ‘efficient’ at the same time – they rallied to the initiative of a Diversity Charter that was launched in 2004, amid great fanfare, by a national, employer-led corporate social movement in favour of workplace and marketplace diversity in France (Alaoui & Doytcheva, 2010).

To explore this entirely new and evolving landscape of public and private corporate diversity actors, I draw on data collected through qualitative fieldwork with a sample of participants stratified by the various profiles of stakeholders at play Table 1: 1/companies and companies’ professionals (managers, personnel professionals, diversity officers n = 25); 2/key business and political leaders (national and local key business and professional networks and associations, political leaders, state representatives n = 18); 3/intermediaries and experts (consultants, recruitment agencies, public and private workforce intermediaries); and 4/last but not least, ‘diverse candidates’ (n = 16), modestly referred to as ‘diversity applicants’ (candidats issus de la diversité), i.e. a racialised workforce trained and monitored through specific schemes, and a service offered to firms eager to diversify their recruitment pools.

Table 1. Breakdown of the interviews

Datasets	Time Period
	2008	2011	2014	GL
(1) Companies^1 (executives, HR, and diversity practitioners)	20	2	3	25
(2) Political leaders and public officials	18		2	20
(3) Diversity ‘brokers’ and intermediaries	20		5	25
(4) Diversity ‘applicants’		10	6	18
TOTAL	58	12	16	86

1.These included: Auchan, Boulanger, Décathlon, Gaz de France, Jules, Okaïdi, Redcats (PPR), La Redoute, Finaref, CIC-BSD, BNP-Paribas, Schneider, Coca-Cola France, Michelin, Renault, Manpower, ADIA, L’Oréal, two SMEs.

Laid down from the outset, these four relatively autonomous but nevertheless connected datasets were constructed and updated incrementally, eventually merging in a global corpus of more than eighty (N = 86) in-depth interviews (see Table 1 for breakdown). Conducted face to face, the interviews lasted on average between an hour and a half and two and a half hours. Their protocols consisted of open-ended, semi-structured questions, adapted to each organisational field. The case study also builds on a twofold, regional and national, component. First, a territorially focused approach, carried out in Northern France, and the Lille metropolitan area, in particular, aims to engage contextually and relationally with the various networks of corporate diversity actors. Mapping their situated interactions, it intends to compare discourse with action, and uncover potentially conflicting or aligning perspectives; it targets companies and business networks, primarily but not exclusively. As for the national component, institutional and political actors, as well as intermediaries and experts are the main participants. In this article, I draw mostly on fieldwork with companies and business leaders and associations, and, to a lesser extent, with diversity intermediaries and experts.

Fully transcribed, these data have been enriched through observations (carried out at hiring forums, job fairs, professional meetings, conferences, and seminars), documentary research and analysis of grey literature (official reports, practical guides) and communication materials (brochures, booklets, flyers, internet sites). To analyse these sets of discourses, in both speaking and writing diversity, I favoured a Critical Discourse Analysis (CDA) method, whose ambition is to consider language as a social practice at the interface of structures of cognition and action (Fairclough, 2013). CDA thus assumes a dialectical relationship between particular discursive events and the structures in which they are embedded (Johnson et al., 2017). In addition, the fieldwork received two thematic anchors: the making of and implementation of the Diversity Charter, since 2004, and the setting of the Diversity Label or standard in 2008.

As noted above, corporate commitment to diversity in France had from the outset a clear preference for ‘flexible means’, in a prevailing climate of low regulation and resistance to additional ‘constraints’ (Gribling & Smith, 2014). As part of a broader trend of manifest support at EU level for ‘soft law approaches’ to economic and employment policies, the tension between voluntarism and (enforced) regulation has therefore been central to debates. Converging to produce discourses that S. Ahmed (2007a) terms ‘aspirational’ – suggesting they might be a sign of the lack of commitment to stronger organisational change – soft-law approaches to workplace and marketplace diversity have materialised into measures such as charters, statements of commitment, trophies, and awards.

The Diversity Charter – a short document, built around six points – is a flagship project in this field. It broadly commits managers to ‘reflect the diversity of French society, particularly in its ethnic and cultural dimensions, at every level of [the] workforce’ (Diversity Charter, section 3¹⁰). Initially designed for companies, the initiative is also opened for signature by public entities, and has gathered more than 3,900 signatories to date.¹¹ Since 2004, it has been emulated by other texts, also prompted by companies, such as the Parenthood Charter (Charte de la parentalité), promoted by L’Oréal in 2008, or the LGBT Charter, sponsored by Accenture in 2013.¹² Whereas across the EU, 13 national Diversity Charters have been established adapting the French initiative, and gathering since 2010 within a Diversity Charter Platform, funded by the Commission (European Commission, 2014). Together, these various charter texts and statements of commitment – some of which can, for instance, be endorsed directly online without further formalities – have laid the foundations for an ever-expanding corporate and marketplace diversity frame.

Faced with harsh criticism about their open and non-prescriptive nature, notwithstanding the warm welcome and enthusiastic sponsorship by the Commission (Alaktiff & Doytcheva, 2018), as of 2008, these initiatives were supplemented in France by the setting of a diversity standard. Spearheaded by public authorities with the support of personnel professional and business organisations, the Diversity Label (Label Diversité) is based on third-party monitoring and certification, and forms in theory a more binding instrument for change. In what follows, I engage with the fabric of these projects and especially the Diversity Label. Making use of Foucauldian conceptualisations of norm, normation, and normalisation, I trace the involvement of these methods and new regulatory schemes in reproducing racial orders through the ‘whitewashing’ of discrimination. Then, I discuss in the concluding section how these techniques have become the vehicles of an ‘exclusionary protection’ (Guthman, 2007) that reinscribes discriminatory practices (Berrey et al., 2017; Doytcheva, 2018a; Ray, 2019) within the very procedures designed to pursue equality, particularly on the grounds of faith, origin, and race.

Normalising antidiscrimination: the case of the French diversity standard

From the perspective of sciences and technology studies, normalisation should be jointly understood as the production of a text and the ‘converging of sociotechnical networks’ (Mallard, 2000, p. 39). Based on Latour’s (1985; see also Lynch & Woolgar, 1990) work, Mallard points out the specific ‘sociological depth’ of ‘the negotiated construction of norms’. Addressing the development of normative texts (charters, technical specifications, compliance referentials) thus implies not only analysing their writing but also reporting on how the agreement reached in the closed cenacles of standardisation institutes is likely to prevail outside. In other words, the process of writing a norm ‘gradually constitutes the framework for a sociotechnical co-ordination’ (Mallard, 2000, p. 39), which reflects the subsequent dynamics of diffusion and implementation of the norm. A rapid incursion into the networks involved in workplace diversity normation, at the beginning of the 2000s in France, reveals three main profiles of stakeholders, in addition to public authorities.

What sociotechnical co-ordination for a diversity norm?: the actors at play

1 – These stem firstly from employers’ unions or employer-led organisations Table 2. Historically positioned in the field of ‘social patronage’ (mécénat social), these groups have repositioned themselves in the 1990s on issues of ‘corporate citizenship’, before embracing issues of corporate social responsibility in the early 2000s. In the same move, they hailed the notion of diversity management as a new finality of their engagement, as well as a development opportunity.¹³

Table 2. Summary table of organisations surveyed (chronological list)

Organisation	Creation	Major Achievements	Role in the field
(1) IMS | Institut pour le Mécénat Social	1986 by C. Bébéar	- Association and corporate network - More than 250 member companies - Committed to ‘social patronage’ - In charge of the Steering Committee of the Charter - Expert at the Normalisation Commission (AFNOR)	Leading
(2) Alliances	1993 by B. Liber	- Regional employer led business network - More than 120 member companies - MEDEF’s ‘showcase for social responsibility’ - Committed to ‘corporate citizenship’ - Regional promoter and sponsor of the charter	Important
(3) FACE Fondation Agir contre Exclusion	1993 by M.Aubry	- Foundation, recognised public utility - Backed by 15 founding corporations - Network of business-oriented ‘regional clubs’ - Promoting locally repertoires of diversity management and servicing corporate plans trough: - Awareness campaigns, training, promotional events (plays, sports matches), and the monitoring of ‘diverse applicants’	Important
(4) Institut Montaigne	2000 by C. Bébéar	- A liberal think-tank - Launched several major reports: Blivet (2004), Sabeg and Méhaignerie (2004), Bébéar (2004) - From which stems the Charter initiative	Leading
(5) ANDRH	1947 2007	- The French national association of HRDs - Co-leading the Label – at the request of public authorities (2007) - Chair of the Normalisation Commission at AFNOR	Leading
(6) AFMD	2007	- A non-profit organisation - A ‘trade association’ - Leading the task force set by the government to ‘revitalise’ the Diversity Label (2013)	Important
(7) Mozaïk RH	2007 S. Hammouche	- Consultancy and recruitment agency - Based on ‘social entrepreneurship’ - Committed to diversity hiring	Important
(8) L’Autre Cercle	1998 2013	- Network of local associations, composed of HR professionals - Committed to fight employment discriminations against LGBT communities - Founder and promoter of the LGBT Charter sponsored by Accenture (2013)	Complementary
(9) CDJE	1938 2004	- First French Employer Movement - Charter partner (2004)	Complementary
(10) E & P	1964 2004	- For-profit organisation and national network of major companies focused on HR challenges and ‘solutions’ - Charter partner (2004)	Complementary

2 – Human resource professionals organisations, secondly, of directors and human resource managers, such as the National Association of Human Resource Managers (Association Nationale des Directeurs de Ressourses humaines, ANDRH), L’Autre Cercle (committed to LGBT issues) or Enterprise et Personnel (E & P, see Table 2 for presentation). As highlighted by US scholarship on antidiscrimination (Dobbin, 2009; Edelman et al., 2001), human resource practitioners have played a huge role in upholding these concerns across the Atlantic, as well. As they were at the heart of equal opportunities programmes during the 1960s and 1970s, personnel professionals became also actively involved in their redefinition, from the mid-1980s onwards. The reasons for this were partly corporatist but, as Edelman et al. (2001) argue, as a professional group with precarious legitimacy in business, their recognition often depends precisely on their capacity to translate social concerns into a managerial language. In France, the involvement of personnel professionals associations was first less important but continually increased, often at the request of public authorities, especially with the emphasis placed on ‘technical’ aspects in developing the Label.

3 – Thirdly, ‘service providers’ acting as ‘relay structures’ (Crozier & Friedberg, 1980), to whom companies selectively outsource some aspects of their commitments – in particular the monitoring of race discriminations and the recruitment of minorities (Doytcheva, 2015). As I have shown elsewhere, in the French, purportedly colour-blind settings, these diversity brokers and intermediaries (Doytcheva, 2011) play a key role in managing uncertainty, in Crozier’s terms. Notably, they help circumvent the very critical issue of dealing with ethnoracial categorisations. Among them, some are consultancy or recruitment firms (e.g., Mozaïk RH, see Table 2), while the most numerous are non-profit organisations (AFMD see Table 2, APC Recrutement: Doytcheva, 2015, p. 121); backed, however, by businesses or public authorities (and often a mix of both). Rather than being mere ‘contractors’ or executors of corporate action plans, diversity intermediaries have moved to quickly become the driving force behind these plans, shaping and actively positioning themselves into emerging diversity markets.

Yet, at the beginning, the entrepreneurial and even more so employer-driven impulse was essential. Although backed politically by the centre-right French government,¹⁴ the corporate social movement for workplace diversity was primarily orchestrated by the involvement of emblematic big business leaders – among whom, first and foremost, Claude Bébéar, CEO of AXA and dubbed ‘godfather of French capitalism’.¹⁵ The movement also received support from employers’ unions (MEDEF, CPME, see Table 3 for acronyms), employer networks (Club des Jeunes Dirigeants d’Entreprise, Young Company Leaders Club, CJDE), as well as their local chapters and relays.

Table 3. List of acronyms and abbreviations

AFEP	French Association of Private Corporations (Association Française des Entreprises privées)
AFNOR	French Standards Organisation (Association Française de Normalisation)
AFMD	French Association of Diversity Managers (Association Française des Managers de la Diversité)
ANDRH	French Association of HR Managers (Association Nationale des Directeurs de Ressources Humaines)
CJDE	Young Business Leaders Club (Club des Jeunes Dirigeants d’Entreprise)
CPME	Employer Federation of Small and Medium-sized Enterprises (Confédération des Petites et Moyennes entreprises)
MEDEF	Movement of The Entreprises of France (Mouvement des Entreprises de France): largest employer federation
FACE	Foundation Acting Against Exclusion (Fondation Agir Contre l’Exclusion)
E&P	Company & Staff (Entreprise & Personnel)
IMS	Institute for Social Patronage (Institut du Mécénat Social)
HR	Human Resources
HRD	Human Resources Director

The co-ordinating pattern (Mallard, 2000) that emerges from these very pioneering actions could thus be described as dual. It typically crystallises around two functions: 1/a political dimension of ‘representing’ (big) business and business interests, and even more so employers’ interests; and 2/a more ‘technical’ dimension, based on specific expertise, gained either in the field of public policies (against poverty and exclusion, e.g., FACE) or in professional areas (human resource management, e.g., ANDRH). It is noteworthy how this dual structure – which leans more strongly on business interest groups than has been observed in the US (Edelman et al., 2001) – is now emulated by a developing offer of expertise and services in the field. This is the case for Mozaïk RH – a recruitment agency, turned foundation – which has become the right arm of French government and public policies against employment discrimination; the AFMD (French Association of Diversity Managers), a small non-profit organisation, functioning as a ‘trade association’, which has been entrusted by public authorities the mission of ‘revitalising’ the Diversity Label in the perspective of its review (see below).

As evidenced by the case study, in examples like these, functions of expertise, service supply, and ‘representation’ of business interests often tend to conflate, however. Most players now try to exhibit the whole set of these aspects of legitimacy, ranging from shared values and references with business to technical expertise in HR management or social welfare. But this is not without raising pressing questions about the potential for collision between public, collective and private interests and goods. Although normalisation is said to be of public interest, public authorities tend often to get marginalised in the process, while still continuing to fund certain activities, as are the legal ideals of antidiscrimination. By contrast, as we will see in the next section on settling a highly disputed ‘ethnic question’, established and well-organised economic and professional interests are even reinforced, as companies are used to sitting through normative processes as ‘both judge and jury’.¹⁶

The norm as textual mediation: arbitrating the controversial ‘ethnic question’

As Mallard (2000, p. 42) points out in his aforementioned study, the textual dimension and documentary nature of normative processes are of key importance. In writing and visualisation, as norms are cited, referenced, and interpreted, they acquire a performative character that goes beyond rhetoric and illustration. By emphasising the materiality of scientific or technical productions, documents offer a specific mediation between structures of meaning and action. Other works have also pointed out, though in a more practical manner, the importance of documentation (specifications, diagrams, drawings, statistics) in doing diversity work (Ahmed, 2007b, p. 599).

This importance of the ‘letter’ of the normative text has been meaningfully illuminated in the case study through the negotiations held at AFEP (Association française des entreprises privées), at an early stage of drafting the norm, part of the Diversity Charter. The Charter project stemmed directly from the Sabeg and Méhaignerie (2004) and Bébéar (2004) reports, but three versions were drafted at Institut Montaigne (see Table 2), before it was discussed at AFEP, where some twenty CAC 40 leaders and top managers were convened. As Huët and Cantrelle (2006, p. 13) put it, the text eventually agreed to on that occasion was ‘the result of the homogeneity of this social group’. Decided self-referentially and in closed fora, it obeyed a ‘random mode of production, where particular interests were likely to prevail over the principle of the common good’.

One point, in particular, captured the attention of attendees, namely section 3, stating the corporate commitment to ‘represent the diversity of French society and in particular its cultural and ethnic diversity […] at the different levels of qualification’.¹⁷ As the excerpts reproduced below attest, the term diversity itself initially appeared problematic, referring alternately to a nationality, a culture, or a religion. Beyond indeterminacy, it was above all its ‘ethnic’ and ‘cultural’ dimensions that raised questions. The ethnic label, in particular, was perceived as ‘embarrassing’ by many who wished that diversity should not be given a specific qualifier, arguing for a ‘broader’ definition, considered more consensual:

For example, someone said: ‘Ethnic diversity, I have a real problem with it, I don’t even know what it is.’ And then everyone picked up, you know, a rather sheepish tendency, saying: ‘Yes, indeed, ethnic is a worry … ’ (AFEP Committee member and Charter leader)

Although after lengthy discussions, the terms were eventually agreed to (in spite of some clauses of the original proposal being definitively removed, such as the commitment to some degree of affirmative action¹⁸), the AFEP negotiations clearly prefigure subsequent dynamics of implementation of the initiative by companies. While for its proponents (Sabeg & Méhaignerie, 2004), the commitment to address segregation and integrate ‘visible minorities’ was critical, AFEP meetings mainly emphasised arguments of ‘added value’ and ‘competitiveness’ as the main reason for large companies’ support. On the other hand, while the originality of the initiative was to target ethnic and racial discrimination, very quickly, in just a few months and even weeks, a shift arose within these initial conceptions towards what has been referred to as a ‘global’ or ‘comprehensive’ diversity frame.

As I have argued elsewhere (Doytcheva, 2009, 2010), grassroots corporate diversity initiatives were marked from the outset by a twofold movement. First, it stems from a rationale of decategorisation, following the path set by EU and French legislations to enact the ‘legal universality’ of non-discrimination, with no less than 25 protected classes under current French law (17 in the 2000 EU Charter of Fundamental Rights). Second, in this context of swift and almost unlimited expansion, ‘tactical choices’ were discretionarily endorsed by organisations, as to the ways of categorising diversity and discrimination.

Formal discourses around a ‘global’ or ‘generic diversity’ norm (Caradec & Doytcheva, 2008; Clarke, 2011) have thus been effectively coupled with informal practices, based on the statement of ‘priorities’ by each organisation. While proving highly selective, these tactics have contributed de facto to excluding issues of race, faith and ethnicity from corporate diversity procedures. Since these concerns are commonly construed as a matter of ‘lower priority’, or which must be ‘pushed’ and ‘should follow’, they give direct advantage to other status categories, perceived as more socially meaningful – mainly sex, age, and disability.

Over time, however, based on a longitudinal analysis (Doytcheva, 2015, 2018d), the variable-geometry commitments and à la carte diversity tactics have crystallised into more binding visions about ‘good’ and ‘bad’ diversity; with the latter – such as race but also faith diversity – being definitely banned from corporate plans. As I will argue in the following sections, it was eventually this practical conception of a comprehensive, then selective and finally prescriptive diversity, that was ‘normalised’ by AFNOR, in 2008. Using a Foucauldian lens, the outlines of experimental procedures launched by companies in the early 2000s, against a backdrop of voluntarism and lack of ‘constraints’, have ‘serve[d] as the norm’, or at least have provided the operational basis for the ‘study of normalities’ from which the norm shall be inferred (Foucault, 2007, p. 64).

Manufacturing the norm: the origin and development of the diversity standard

From 2007 onwards, and in order to mitigate these somehow flawed efforts by companies, public authorities and the government-mandated central state administration via the ministry of immigration, in particular, wanted to re-enter the normative process. Shortly after failing to take over the national leadership of the Charter, which was actively resisted by business organisations and interest groups, the government entrusted the French Association of Human Resource Directors (ANDRH) with the setting of a diversity standard. Third-party social certification for the purpose of employment policy-making had been already tested upstream, with the introduction of the Equality Label in 2004. Accredited by AFNOR, it certifies organisational achievements in gender equality at work. It was thus through the creation of an ‘independent’ quality standard – i.e. third-party assessed and monitored – that public authorities planned to extend control over those activities.

The Label is indeed expected to provide a more binding regulation. As such, it does not invalidate previous initiatives but is rather said to represent a ‘later stage’, or their ‘natural extension’.¹⁹ To initiate the standards-making process and prepare the normative work, an extended committee is convened at the Ministry of Immigration (which ironically becomes involved based on its competencies on antiracism and race discrimination), composed of human resource professionals, public officials, trade unions, and ‘experts’ (e.g., business organisations, such as IMS). Following its work, which resulted in the drafting of an early set of specifications, the French government entrusted ANDRH with the development and the testing of the norm.

At that time, ANDRH already had its own ‘Professional Equality and Diversity Commission’, chaired by the Director of Human Resources at Eaux de Paris, and the Association’s vice-president. From mid-2007 onwards, several working groups were set up to prepare the draft requirements of the norm. In October, the Normalisation Commission (Commission de Normalisation), also led by ANDRH’s vice-president, is held at AFNOR, setting the foundations for the standard and testing it with some pilot companies. By way of a public tender, the administration entrusts its monitoring to AFNOR Certification, which is the trade branch of the French standards association.

As in previous normative work around the Charter, the standards-making process is primarily based on for-profit and market co-ordination: no activist or civil society organisations are expected to join in, apart from business interest groups and organisations. As stressed above, unlike labels which are mainly consumer-oriented schemes, standards are supposed to provide firms with more generic communication, transmitted to their wider stakeholders, including employees, suppliers, clients, and shareholders. In French settings, they imply significant institutional communication, as voluntary social certification is intended in theory to equip companies with ‘social best bid policies’ (mieux disant social, that gives some degree of priority in the context of public tenders). From a corporate perspective, they are used as internal communication tools, and are expected to enable more efficient management of employees and candidates for recruitment in tight (low-skilled) job markets (Doytcheva, 2008). However, as one interviewee, a senior administration official, puts it, issues of branding and preserving reputation feature thus far among the most prominent concerns: ‘Indeed, there is some legal risk at play [around discrimination] but, above all, it is all about branding the image of the company’.

The outcome is the normalisation of both an expanded and ‘flexible’ diversity concept, which strongly matches the one already emerging from grassroots corporate action plans. According to a member of the extended standards-setting committee, convened at the Ministry of Immigration in 2007, the definition of a ‘generic diversity’ norm is clearly much more attractive to businesses: ‘Rather than isolating this or that aspect of diversity, a global standard seems more effective!’. Similarly, at that time, the Institut du Mécénat Social (IMS), in charge of the National Steering Committee of the Charter, ‘changed its communication’:

The Charter first focused on racial discrimination. Now, we will never be able to promote it if we stick to this aspect alone […] As time went by, we realised that with companies, it is much easier to come up with something global: we are dealing with seniors, gender equality, disabled workers, sexual orientation, work-life balance … Our message is not the same anymore. (Diversity Department, IMS)

Yet, on the other hand, the Label also endorses the selective approaches to (good) diversity. Under a generic standard, indeed, companies remain free to select, as if from a menu with quite a variety of items, the categorical axes of diversification they are willing to endorse. The articulation of these axes remains at the exclusive discretion of companies. According to another participant in the extended committee:

In fact, the idea, the important thing, I’ll tell you, is that people tackle the subject from any angle they want, but that they do so! And by experience when you start with a particular issue, you will soon get caught up in the stream of diversity. It’s true that if you narrow the entry-point too much, you can discourage. If you chose to work on this through disability, so be it! The diversity stream will still drive you. Because, I think, diversity leads to diversity. (Chair of the Normalisation Commission, emphasis added)

The requirements of the standard, therefore, enshrine the variable geometry of à la carte diversity norm. A first draft, dated July 2008, explicitly discards sex discrimination (which is targeted by the Equality standard). But again, under pressure from companies, gender equality is gradually reintegrated into the grounds of discrimination, based on which companies can achieve certification. In a new version, dated May 2010, the reference to the exclusivity of the two marks (Equality vs. Diversity) is removed: many examples in the appended practical guides deal with gender equality as an integral part of corporate diversity policies. According to AFNOR’s Normalisation Commission President himself:

We are not going to ask a company to work on all categorical axes from the beginning, that’s impossible! We are not going to ask everyone to be clean on all kinds of discrimination. For a start, if the company has already been working on disability … [that’s well enough]. Because when your human resource management process becomes objective, it’s quite naturally that it will open up the realm of possibility. (Chair of the Normalisation Commission)

Yet, according to fieldwork, the very opposite happened. While corporate diversity stakeholders advocate ‘tactical approaches’, aimed at addressing these issues incrementally, by first dealing with ‘the most numerous categories’ (i.e. gender and age), or those that benefit from specific, normative-based, provisions (often prior to antidiscrimination law, e.g., disability, sex, and age), racial diversity, which supposedly ‘must be led’ or is ‘meant to follow’, very often ‘falls flat’ (Doytcheva, 2015, p. 184). These findings significantly converge with the most recent work on Athena SWAN and Race Equality Charter Marks in UK higher education (Bhopal & Henderson, 2019), showing how, within organisational diversity policymaking, gender – among many other status-categories in France – ‘has taken precedence over race’.

The standard review and the role of diversity experts

It is thus very much in the light of this appalling observation that in 2013 a new initiative was launched by public authorities, for another ‘multi-actors committee’, aimed at ‘boosting the Label within a republican approach’ (Alaktiff & Doytcheva, 2018, p. 158). In the administration’s terms, the expected ‘revitalisation’ involves ‘expanding the scope of the norm’ beyond the age-gender-disability triptych, in order to ‘respond to the need for professionalising antidiscrimination in areas that are not yet very well addressed.’ (ibid.) Indeed, five years after its introduction, the Label is reportedly effective only with regard to the three criteria mentioned above, i.e. sex, disability, and age. For state actors, ‘revitalising’ it therefore clearly implies extending its application to other situations among those initially targeted, including ethnic and racial discrimination.

But, on the other hand, while shortly after its introduction in 2008 the Label ‘sold rather well’,²⁰ the figures ‘stagnated’ in a second phase. By 2013, there were reportedly some 250 accredited organisations, compared to only about 50 recipients of the Equality Label, although introduced earlier. A few years later, this growth stalled, however, with around 300 certified organisations in 2015²¹ and just over 100 in 2019²². Hence, the objective of ‘dynamisation’, understood as a relaunch rather than an overhaul of the norm. The concerns of public authorities were thus twofold: somewhat split between extension and relaunch. In the Committee’s work, however, the former objective of extension will rather be lost to a technical discussion aimed at further normalising the already normalised certification procedure. Taking the shape of ‘a quarrel between normalisers’ (Mallard, 2000; see also Doytcheva, 2018b), discussions focus on issues of ‘concurrence’ between Diversity and Equality Labels, and the opportunity of integrating them into a single certification in order to ‘stay attractive for companies’. Against the backdrop of weak political will, this dispute among normalisers prevailed in the Committee’s work, once again giving greater significance to a logic of technical ‘rationality’ and social acceptability or normality, which is in practice detrimental to ethnic and racial minorities. Based on fieldwork, this move appears to be largely informed by ‘experts’ leading the Committee’s work.

Indeed, although multi-actors in design, the task force is mainly populated with diversity and human resource managers from major accredited companies, joined by two researchers and, intermittently, key contacts (lawyers, certifiers). Its coordination is entrusted to the Association of Diversity Managers (AFMD see Table 3). As pointed out above, AFMD is a hybrid structure – blending ‘expertise’ with political legitimacy, based on the support and funding from a cluster of member companies. Its position as a de facto trade association makes it an attractive ‘policy entrepreneur’ (Kingdon, 2002) vis-à-vis the political. Jointly with the lead association, personnel managers and diversity practitioners manage to strongly shape the Committee’s priorities. Shifting away from extension concerns, issues of ‘attractiveness’ and ‘shared costs’ widely dominate the year-long meetings and activities.

Two topics, in particular, draw sustained attention: the integration of the Equality and Diversity Labels into a single certification, on the one hand, and the ‘tensions’ said to arise between companies, certifiers, and AFNOR’s Commission for Normalisation, on the other. Cross-cutting to all sessions, these topics become more explicit in two workshops: ‘The multiplicity of norms, labels and certifications’ and ‘Audits’. The vested interests of companies are to reduce the costs of monitoring and inspection, which are recurring and high (Nadvi & Wältring, 2002). Companies are also pushing to ‘adapt’ (allege) inspections, especially in the case of renewal, in order to ‘maintain dynamism, without exerting too great a pressure on companies’ (Alaktiff & Doytcheva, 2018, p. 161). Voluntary social certification, which for them is an interesting means to leverage management and communication, is also costly in both material and human resources terms. Investments required by compliance undermine direct business profitability. According to technical officers, they are currently faced with ‘too much auditing’.

Further difficulties are said to arise from what companies consider as ‘discrepancies’ and ‘tensions’ between AFNOR’s inspectors and its own Commission for Normalisation. Indeed, under a highly centralised French system, normalisers, certifiers and inspectors are part of the same holding company. Nevertheless, they apparently have (slightly) divergent interests in practice: whereas auditors’ goals are to ‘share good practice’ and to foster business relations with client companies (who pay them), those of the Commission are primarily to preserve its own legitimacy and that of the standard. The ‘good cooperation’ reportedly experienced by companies with certifiers and inspectors can thus contrast with ‘frustrations’ arising at the hearing before the Commission stage.²³ Some threaten to withdraw from the process and ‘make challenges on their own’, in a logic of self-valuation and a desire to ‘re-enchant professional relations’, in De la Broise and Lamarche (2006) terms.

All in all, dynamics of rationalisation, or what would be the normalisation of the already normalised procedure, prevail over the objective of extending the application of the norm. As companies take control of the Committee, this becomes ‘a much-awaited opportunity to voice “frustrations” with the label and take the advantage to further adapt it for their own benefit’ (Alaktiff & Doytcheva, 2018, p. 161). Concerned with the problem of ethnic and racial discrimination, public authorities lack the means and political will to impose this agenda over private interests. Hence, the recommendations issued by the task force’s report ‘How to revitalise the Diversity Label. Feedback from accredited companies’ primarily tackle rationalisation, merely to the advantage of companies. The review process became effective shortly afterwards with the introduction of the Alliance Label in 2016, which merges the diversity and equality marks into a single certification, thereby still further advancing the agenda of deracialising antidiscrimination.

Discussion

In this paper, I have focused on how techniques of normalisation, which are key to late neoliberal governmentality, have been applied to the management of race difference. Based on the case study of the Diversity Standard – and before, a Diversity Charter – I have shown the contradictory outcomes and lack of transparency of voluntary social certification as a strategy to achieve racial justice and equal opportunity. As a privatised and individualised governance form, certification largely accepts neoliberal imperatives for flexible and market-driven frameworks. Hence, ‘neoliberalising’ (Peck & Tickell, 2002) equal opportunity has added ‘normalisation’, in a Foucauldian sense, to the process of ‘managerialisation’ of antidiscrimination law. This paper has engaged with one of the most compelling outcomes associated with this path, namely deracialisation and the rise of ‘White diversity’ concepts.

From my position here, these developments come at a later stage in the global rollout of corporate diversity policies, which goes beyond their infusion with managerial rhetoric to reflect a wider political, social and spatial dynamic based on the marketisation of these frames. That means, the way in which they are increasingly governed through market mechanisms of regulation, patterns of legitimacy or ‘rationality’ (Boltanski & Thévenot, 2006; Thévenot, 1997, 2009). Converging with a broader trend in the study of market dynamics towards shifting the lens from material markets devices to governmentality (Roscoe, 2015), I have highlighted how a close engagement with Foucault’s understanding of neoliberal governmentality, and normalisation, in particular, can offer valuable insights. This paper thus adds to extant understanding of marketisation processes by highlighting their propensity to reconfigure and reinscribe privilege and hierarchies (Johnson et al., 2017), in addition to those already emphasised towards individual self-disciplining and responsibilisation (Varman et al., 2011), or the neglect of social concerns. This is achieved I argue, on the one hand, by extending logics of rating and market valuation – where there were (are) none – which ultimately leads to the replacement of the entrepreneurial self famously described by Foucault in the late 1970s with the ‘investee’ or ‘self-capitalising’ one (Doytcheva, 2020; Feher, 2018); the embeddedness of ‘marketing knowledge’ and ‘market subjectivities’ in wider social relations (Roscoe, 2015; Varman et al., 2011), on the other hand, increasingly understood to play a performative role in the construction of a variety of social fields and participants. I have highlighted this performative role in constructing diversity around three functions: instrumental rationality, first, commodification and valuation, second, and privatisation and devolution of responsibility, third. Converging with critical perspectives and moves to (re)write markets back into marketing (Araujo et al., 2010 as cited in Roscoe, 2015), this paper set out to shed light on the highly controversial outcomes of governing racial justice through standards.

These findings add to, and closely resonate with the critical literature in diversity studies in showing how market and organisational applications of these frames have increasingly subverted not only civil rights ideals but also, more broadly, notions of social justice and equity. Alongside critical race theorists (Bonilla-Silva, 2018; Mueller, 2020; Ray, 2019), they demonstrate the enduring institutionalisation of racial orders by organisations. From this perspective, ‘White diversity’ undoubtedly pertains to the repertoires of ‘epistemologies of ignorance’ and other epistemic manoeuvres by hegemonic whiteness to produce everyday organisational and structural dominance through colour-blind racism. However, unlike colour-blindness, ‘White diversity’ has the peculiarity of reinscribing racist structures and power relations within purportedly race-conscious procedures, designed precisely to right their wrongs. As such, it calls for fresh explanatory strategies, whereby a Foucauldian perspective on dispositifs of governmentality appears as particularly productive. From this perspective, this paper adds to extant understanding and provides a novel account of how markets appropriate legal ideals and goals, favouring a definitely multi-site, historical and de-nationalised lens.

Certainly, from the literature consulted above, framing diversity as the new business paradigm for differences should be traced back to the mid-1980s and the US corporate world. While a first interpretive gesture construed this movement as simply ‘old wine in new wineskins’ (Kelly & Dobbin, 1998), soon there were more critical assessments of how it carried forth the legacy of civil rights. Following Edelman et al. (2001), the concept of ‘managerialization’ has been widely adopted to account for this shift, first undergone by US companies, from antidiscrimination, and in particular, affirmative action, to diversity management (Dobbin, 2009; Edelman, 2016; Edelman et al., 2001; Kalev et al., 2006; Kelly & Dobbin, 1998). Based on a twofold hypothesis, the managerialisation model posits first, the ‘endogenization of law’ (Edelman, 2016) through the work of compliance professionals; and second, the reframing of ideas inherent in civil rights legislation through the ‘business case’ for diversity management. However, parallel to the belated ‘Europeanisation’ of these concerns, this paper discusses how some significant trends are left unaddressed and may even conflict with important assumptions of the managerialisation framework.

Indeed, decades after civil rights and race relations laws in the US and UK, respectively – but also the prohibition of sex discrimination in the 1958 Treaty of Rome – the European Union has committed to antidiscrimination and ‘valuing diversity’ as a means to address migrants’ integration and citizenship. For the first time, in 1997, Article 13 of the Treaty of Amsterdam granted the Union new powers to combat discrimination on the grounds of sex, racial or ethnic origin, religion or belief, disability, age or sexual orientation. Implemented by two 2000 Council Directives on race and employment equality,²⁴ this legislation has been gradually transposed into national laws, some of which had until then only a very modest legal framework. Similar to other ‘continental law’ countries, France saw these developments trigger a dose of judicial and political uncertainty (Doytcheva, 2010), with regard to the breadth and scope of such new legal ideas and policy frameworks. Refraining from the semantics of racism, as well as from the perspective of administrative sanctions and ‘constraints’, major French companies soon initiated a move towards being proactive by embracing rationales of diversity management. Their endeavours were therefore commonly construed within the framework of managerialisation (Alaktiff & Doytcheva, 2018; Djabi‐Saïdani & Pérugien, 2020; Gribling & Smith, 2014, 2017; Klarsfeld, 2009; Safi, 2017), stressing the shift beyond legal constraint to the economic value of, and the ‘business case’ for workforce diversification.

However, as I have argued in this article, the prism of managerialisation, and correlatively the endogenisation of law, should be supplemented in the French settings by underscoring patterns of normalising antidiscrimination through diversity management. This is all the more relevant in the context where voluntary social certification stands for the sole and unique strategy for public policy-making against workplace and marketplace discrimination. The significance of a Foucauldian approach to governmentality is therefore reinforced not least by a shared lexical field – around norm, normalisation – which, in French, qualifies the policy instruments used to implement antidiscrimination (referred to in English as standards).

Indeed, the framework of managerialisation fails to account for at least two critical features of corporate diversity in France, and more broadly in the context of the late Europeanisation of these concerns: first of all, their gradual deracialisation, based on the eviction of race and ethnicity from the space of diversity politics; and second, the also widespread phenomenon of outsourcing or ‘externalising’ diversity plans, which, once again, is particularly relevant when it comes to the monitoring of race discrimination. Instead of ‘endogenization’, which eventually helps ‘legalize organizations’, although the managerial logic is said to alter the spirit of laws, corporate diversity in France relies heavily on external advocacy and services by diversity brokers and intermediaries.

In the same vein, even though in the US the concept of diversity has been expanded to include a wide array of personal characteristics not explicitly covered by law – diversity of thought, culture, lifestyle or dress – the spread of these concerns does not erase, the significance of race. Yet, it has been used to minimise it. In European and French settings, however, the tropism towards ‘universalisation’ (Guiraudon, 2004, 2009), which goes within the very legal system of antidiscrimination, in the space of only a few years or even months, has contributed to almost completely dismissing issues of race and ethnicity from the space of corporate diversity.

In this paper, I have hypothesised that this situation is at least partly performed through technologies of normalisation. From a Foucauldian perspective on normalisation, stressing mechanisms of ‘normality’, ‘acceptability’, and ‘desirability’ behind the enactment of purportedly neutral or ‘only technical’ arrangement and tools, the move to ‘White diversity’ is all the more intelligible. As becomes clear, the latter overlaps with still dominant conceptions of the (good) society. The interplay of variable-geometry commitments and à la carte diversity – although mostly White tactics – has brought about the reinforcement of invidious hierarchies within the very procedures set up to pursue equality (Berrey et al., 2017; Ray, 2019). Normalising antidiscrimination, in both conceptual and practical terms, is therefore accountable for the processes of its paradoxical ‘whitening’ (Doytcheva, 2018a). Based on the case study of the Diversity Label, I have shown how these processes are performed through valuation – or the primacy given to marketing mechanisms and subjectivities – and privatisation, inducing the collision between public, collective, and private goods.

Concurring with previous research on standardisation as a governance form (Borraz, 2004, 2007; Brown & Getz, 2008; Campbell & Liepins, 2001; Cochoy, 2000; Guthman, 2007; Mutersbaugh, 2005), the case study sheds light on how governing racial justice by standards coalesces around three tenets. These are governance; instrumental rationality and valuation; and privatisation or devolution (Heynen & Robbins, 2005). On the one hand, norms are depoliticising policy instruments; in relying on purported neutrality, they conceal democratic debates, hide unspoken objectives, and manufacture consensus to reform – while forclosing avenues for collective action, at least temporarily. On the other hand, by delegating regulatory powers to private actors, public authorities have entrusted the latter to regulate their activities, and even to self-regulate, in a logic of self-production of ‘certificates of good conduct’ (De la Broise & Lamarche, 2006, p. 17). Both these aspects converge to reinscribe market mechanisms and subjectivities within interstitial spaces, supposedly designed to ‘protect’ against and resist them; yet still circumscribing ‘the realm of possibility’ for more substantive change (Brown & Getz, 2008, p. 1184).

What must be appreciated with equal acuity, hence, is how the emergence of third-party social certification as a strategy for performing antiracism and equal opportunity does not only yield dissociation with social causes, social movements and legal concepts. Through a loop or a feedback effect, it is increasingly understood to have contributed to their current assessment in terms of lack of ‘effectiveness’ or ‘operability’ (Inghilterra, 2018). Rather than the common understanding of a ‘managerial response’, a subtler relationship should then be envisioned between normative-based regulation, on the one hand, and corporate and market activism through voluntary certification, on the other. Instead of considering market devices as substantial and material, only passive backgrounds, the emphasis placed on their performative nature invites to scrutinise their practical outcomes (Skålén, 2009; Varman et al., 2011). From this perspective, instead of being solely managerialised, concepts inherent in antiracism and the right to equal protection run the risk of being dismantled and jeopardised (Hirschman, 1991); as they might be used as vehicles for a ‘predatory inclusion’ (Seamster & Charron-Chénier, 2017) or ‘exclusionary protection’ (Guthman, 2007) practices.

To conclude this paper, this research has offered an understanding of how equal opportunity is influenced by marketisation under neoliberalism, in the context of what I have termed ‘the reverse sequence of antidiscrimination Europeanisation’ (Doytcheva, 2018b; Helly & Doytcheva, 2011; see also Klarsfeld, 2009). That is, how corporate diversity management, resting on market-based incentives and (self-)regulatory practices, has been deployed in the EU and France prior to any firm legal commitment over rights-based litigation and/or a public antidiscrimination strategy. My hypothesis is that these specific temporal or sequential and historical factors have been reinforced by structural ones and the unprecedented legitimacy granted to self-regulatory practices within emerging EU governance. Together, they have been working to further enhance the paths already illuminated in scholarly discussions around corporate diversity procedures, towards both overflowing and circumventing the legal system and the social causes of antiracism and antidiscrimination. On the one hand, this bolsters the dominance exerted by market mechanisms and economic actors through self-referential, provisional and revocable procedures, wherein they sit as ‘both judge and jury’. On the other hand, and subsequently, it accounts for the certification of variable geometry commitments, one of whose most relevant features, within normalised French diversity conceptions, is the eviction of ethnicity and race.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Funding

This work was supported by the University of Lille (CeRIES ULR 3589) as part of its Interdisciplinary Research Programme ‘Theories and Politics of Recognition: A Sociological-Philosophical Exchange’ (TheoPOL, 2019-2022). The author also acknowledges support from Institut Convergences Migrations (Collège de France-CNRS) under grant n° ANR-17-CONV-0001.

Notes on contributors

Milena Doytcheva

Milena Doytcheva is Professor of Sociology and Philosophy at the University of Lille (France); a Fulbright laureate and Distinguished Visiting Scholar (2021) at the ARC Institute (The Graduate Center, CUNY). Her research focuses on processes of regulating race and ethnicity through politics and public policy, including issues of diversity and non-discrimination in the marketplace. Her work has appeared in top journals, including SociologieS, Revue des Droits de l’Homme, Revue Française des Migrations Internationales, Revue Française de Sociologie.

Notes

1. Council Resolution of 7 May 1985 on a new approach to technical harmonisation and standards (85/C 136/01).

2. That is ‘a thoroughly heterogeneous ensemble consisting of discourses, institutions, architectural forms, regulatory decisions, laws, administrative measures, scientific statements, philosophical, moral and philanthropic propositions- in short, the said as much as the unsaid […] The apparatus (dispositif) itself is the system of relations that can be established between these elements.’ (Foucault, 1980, p. 194).

3. Here I confer to the concept of ‘White diversity’ a dual sense: first, it is intended to highlight the paradox of a certain eviction of race and ethnicity from the space of corporate diversity politics, as has been notably but not solely documented in the case of France (Doytcheva, 2009, 2010); second, the progressive mainstreaming of these ideals and their embeddedness into a majority-centred discourse. Both aspects concur to reinscribe, rather than mitigate, race hierarchies and power relations within the very procedures, designed to pursue equality (Doytcheva, 2018a, 2020; see also Berrey et al., 2017; Ray, 2019; Mueller, 2020).

4. To make the case for this, Foucault elaborates on the historical phenomenon of the town (17^{th –} mid-18^thCentury). On the one hand, the town is the ‘free town’, i.e. an exception within a territorial system of power based on domination (e.g., feudalism). On the other hand, the town ‘turns on the problem of circulation.’ How to reconcile therefore ‘legitimate sovereignty’ with the assurance that ‘things are always in movement, constantly moving around, going from one point to another, but in such way the inherent dangers of this circulation [i.e. scarcity, contagion, and revolt] are cancelled out. No longer the safety (sûreté) of the Prince and his territory, but the security (sécurité) of the population’ (Foucault, 2007, p. 65). New mechanisms of power – the apparatuses of security – do not make use of obedience and do not function on the axis of sovereign-subjects relationship. They do not tend to a ‘nullification’ or a form of prohibition either, but ‘to a progressive self-cancellation of the phenomena by the phenomena themselves’ (ibid., p. 66), resembling what physiocrats called ‘physical processes’, which could also be called, Foucault adds, ‘elements of reality’.

5. According to the French standards-setting body (AFNOR): a standard or, in French, a technical norm ‘is a public policy tool that acts as a supplement to regulation and a reference point for opening up public markets and promoting their transparency.’ (Borraz, 2007, p. 59).

6. According to Borraz (2007), at EU level, standards were initially promoted as a means to quicken the pace towards the single market, before becoming an instrument of wider governance for the Commission. Standards-based regulation was thus first a response to the lack of political consensus and constraining means at the disposal of the Commission, before embodying a broader trend towards the rise of ‘regulatory states’, based on new regulatory, meta-regulatory, and self-regulatory techniques (Levi-Faur et al., 2004).

7. ISO 9000 is one of the most well-known examples here. According to Nadvi and Wältring (2002, p. 7), ‘within the universe of standards there are a number of sub-categories [which] adds to the confusion, especially where the boundaries between, and within, these sub-categories are unclear.’ This extends in particular to the distinction between standards, labels, and codes of conduct. Labels are said to ‘provide consumers with a simple way to rapidly and easily acquire information about product characteristics (the woolmark label […] the kite mark label […]), or about conditions of production (the fair-trade label)’ (Nadvi & Wältring, 2002, p. 7). Labels tend to be sector-specific and concentrate on particular themes, while codes of conduct are even firm-specific. In contrast, standards have a more generic dimension; they are transmitted to a company’s employees, suppliers, and its wider stakeholders, including its clients and shareholders.

8. In the French case, the first legal provisions against racism date back to the 1939 Marchandeau law, which amended the 1881 free press law to include the prohibition of racial defamation and incitement to racial hatred. Repealed under the Vichy government in 1940, these provisions were reintroduced by the Pleven law of 1 July 1972 ‘on the fight against racism’, the first legislation to explicitly tackle employment discrimination. Its promulgation followed France’s ratification in 1971 of the International Convention on the Elimination of All Forms of Racial Discrimination (ICERD, adopted in1965). Civil provision to litigate employment discrimination was first introduced only ten years afterwards, by the Auroux law of 4 August 1982, transposing the 1976/207/EEC Directive on equal treatment between women and men.

9. French Law No. 2001–1066 of 16 November 2001 on the fight against discrimination; French Law No. 2008–496 of 27 May 2008 on various provisions adapting to Community law.

10. http://www.diversity-charter.com/french-charter/actions/(accessed on 6 July 2020).

11. https://www.charte-diversite.com/charte-de-la-diversite/(accessed on 30 June 2019).

12. The Corporate Parentality Charter dwells primarily on issues of work-life balance: https://www.observatoire-qvt.com/charte-de-la-parentalite/presentation/. Sponsored by Accenture, the LGBT Charter is led by The Other Cercle (L’Autre cercle), a network of HR professionals and organisations dedicated to combat employment discrimination against LGBT communities:https://www.autrecercle.org/page/l-autre-cercle (accessed on 30 June 2019).

13. But this is also a more general trend: according to a survey of the European Business Test Panel (EBTP), employer organisations and networks are the main source of information on diversity in the workplace (European Commission, 2005).

14. Following a report from Institut Montaigne (a liberal think tank, in the French sense of this term) entitled Les oubliés de l’égalité des chances (The left-behind in equal opportunities: Sabeg & Méhaignerie, 2004), the Prime Minister J.-P. Raffarin entrusted Claude Bébéar, CEO of AXA, ‘to shed light on the specific interests of companies committed to diversity and equal opportunities and to develop tools to enable them achieve these objectives effectively’ (Cf. mission statement of 28 May 2004 and Bébéar, 2004). The Diversity Charter, launched in October 2004, is a direct outcome of both of these reports.

15. Claude Bébéar, le ‘parrain’ du capitalisme français, L’Express, 28 February 2008 https://lexpansion.lexpress.fr/actualite-economique/claude-bebear-le-parrain-du-capitalisme-francais-quitte-axa_1335782.html (accessed 10 July 2019).

16. The test case here is that of AFNOR, at the same time client and assessor of the accredited organisations. Through its trade subsidiaries such as ‘AFNOR Certifications’, the standards organisation is in charge not only of the normative process but also of monitoring and certification, which are essentially market-driven and for-profit activities (AFNOR, 2018, p. 4 and here https://www.afnor.org/en/about-us/who-we-are/(accessed on 6 July 2020)).

17. Section 6, which states the inclusion in the annual report of ‘a chapter describing our commitment to non-discrimination and diversity: measures implemented, procedures and results’ is also the topic of hot discussions. The other Charter sections, totalling six, provide for actions such as ‘awareness raising’ (1), communication (4), collective bargaining (5). http://www.diversity-charter.com/french-charter/actions/(accessed on 6 July 2020).

18. For example, section 2, which initially proposes to insert a non-discrimination clause in all documents related to hiring and promotion, as well as to commit to a certain degree of affirmative action, finally reads a much vaguer ‘respect and promote the principle of non-discrimination at every stage of the HR management […]’ http://www.diversity-charter.com/french-charter/actions/(accessed on 6 July 2020).

19. Chair of the Normalisation Commission at AFNOR.

20. Indeed, one should be reminded of the certification costs, which are high and are charged by AFNOR Certification (the trade branch of the standards organisation) for the initial inspection of the company and the subsequent monitoring, which takes place every two years (AFNOR, 2018; see also footnote 16 here).

21. https://travail-emploi.gouv.fr/IMG/pdf/Flyer_Diversite_v2.pdf (accessed on 4 July 2019).

22. https://www.fonction-publique.gouv.fr/remise-label-diversite-au-ministere-de-la-culture-ainsi-qua-11-etablissements-publics-places-sous (accessed on 10 June 2019).

23. With a multi-stakeholder composition – including state representatives from the Ministries of the Interior, Labour and Urban Development, trade unions of employees and employers, and a college of experts appointed by the ANDRH – the Commission is intended to be a guarantee of ‘good practice’ and to uphold the legitimacy of the Label (AFNOR, 2018).

24. Council Directive 2000/43/EC of 29 June 2000 implementing the principle of equal treatment between persons irrespective of racial or ethnic origin; Council Directive 2000/78/EC of 27 November 2000 establishing a general framework for equal treatment in employment and occupation.