Diversity, solidarity and the construction of the ingroup among (post)colonial migrants in The Netherlands, 1945–1968

ABSTRACT Concerns about the impact of immigration (‘diversity’) on welfare states (‘solidarity’) are widespread among political economists. This article presents an alternative theoretical framework for understanding their relationship. Using social and cultural theory, I argue that it is tautological to suggest that diversity and solidarity covary; both emerge out of the same ideological and material efforts to construct an ‘ingroup.’ I probe this theory with a historical case study of the inclusion of (post)colonial migrants in the Netherlands from 1945 to 1968. Complementing secondary literature with original archival research, I show key state and non-state agents of the emergent Dutch welfare state constructing racial categories by fixing ‘Westernness’ or ‘rootedness’ as a salient determinant of ingroup membership, locating (post)colonial migrants in relation to it, and distributing entry, citizenship and social rights accordingly. An elusive metric of cultural proximity, ‘Westernness’ was under construction at the same time as it was in use, as state officials, social workers, and private charities negotiated its meaning with the public. The article not only compels European welfare scholarship to acknowledge race and racism in its recent past, but also builds theory regarding the influence of identity on redistributive outcomes.


Introduction
At the nexus of migration and welfare state studies, anxieties about the impact of diversity on welfare are neither new (Freeman 1986) nor going anywhere (Cavaillé and van der Straeten, 2022, Soroka 2016, Kymlicka and Banting 2006).Assumed is that immigrants introduce 'diversity' into an otherwise homogeneous system, and solidarity suffers because the public prefers members of the cultural or racial 'ingroup' (van Oorschot 2000;2006, van Oorschot andRoosma 2017).However, decades of scholarship has failed to reach consensus on the mechanisms of this thesis (Cavaillé and van der Straeten 2022).Additionally, it has been accused of exaggerating European homogeneity (Pontusson 2006, p. 322) and ignoring the impact of exploitative colonial practices and other blatant 'failures of solidarity' on welfare expansion (Bhambra andHolmwood 2018, Bhambra 2022).
My own concern with this thesis is the inchoate notion of 'identity' upon which it rests.For immigrants to introduce diversity, their cultural, racial, and/or religious identities need to be distinct and exogenous.Yet a growing body of work in social and cultural theory suggests that social identities are historically contingent (see e.g.Du Bois in Meghji 2022, p. 83, Hall 2021, Appiah 1998).Any impact of immigration on welfare states is mediated by a process of identity construction, through which the terms of ingroup belonging are fixed and subjects are assigned identities that relate to these criteria in specific ways (see Meghji 2022).Diversity and solidarity ensue.
This article puts these theoretical claims through their paces with a case study of (post)colonial migrants from the former Netherlands Indies.Using original archival research and secondary literature I explore their inclusion in the Dutch welfare state from 1945 to 1968.This influential case sheds light on boundary-making dynamics during a critical juncture in the demarcation of the European body politic and the emergence of social citizenship.The paper unpacks the ways that identity affects redistributive efforts, a noteworthy blind spot in political economy which has mostly viewed political, social, and cultural dynamics as secondary to the economic (but see Gamble 2021, Best 2021).In addition, it elaborates on the importance of race and racism for welfare states, a topic that has received considerable attention in the US context (Lieberman 2001, Fox 2017, Gilens 1999, Alesina and Glaeser 2004) and has been largely ignored across the Atlantic (but see Bhambra and Holmwood 2018).
In the next section I build the theoretical foundations of my argument.I then introduce my research design and provide historical background.My analysis shows how state officials, social workers, and charitable (religious) organisations racialised different (post)colonial migrants differently by redefining, and making distributive decisions with reference to, biological, psychological, and cultural 'rootedness' in an imagined Dutch nation.The emergent welfare state, by materially privileging some groups over others, reinforced the appeal of the racialised ontology upon which it relied and the solidarity to which it gave rise.A final section discusses and concludes.

Diversity and solidarity as jointly constructed
The uneven character of inclusion Social rights are the right to and realisation of economic welfare at a level commensurate with prevailing standards, alongside recognition as a moral equal (Marshall 1950, p. 11, Somers 2008, p. 6).Despite the temptation to view citizenship as carving a fortified crevasse between insiders and outsiders (Brubaker 1992, Geddes 2000, Marshall 1950), existing literature reveals an unequal distribution of social rights within the citizenry.Bismarckian welfare systems entrench existing class (Esping-Andersen 1990) and gender (Lewis 1992) stratifications.Meanwhile, racial exclusions to welfare are well-documented in the US (Somers 2008, Lieberman 2001).For example, in her study of progressive era relief patterns, Fox (2017) finds that Black American citizens were summarily excluded from relief services, and Mexican immigrants were more likely to receive relief from private (rather than public) funds compared to their white European counterparts who generally received the federal relief to which they were entitled and whose reliance on welfare was defended by researchers and social workers.
Inequalities are functional.Sociologists, historians and political theorists have argued that boundaries around and within a territorial unit consolidate and stabilise the political unit (Ferrera 2005, p. 20, Anderson 1991, Walzer 1983, Peterson 2021).Gender and racial inequalities in the household enable the performance of 'socially necessary but unpaid work' (Peterson 2021, p. 292).Meanwhile Rajaram (2018, p. 628) posits that capitalism has a front stage, characterised by free relations of exchange, and a backstage, featuring a 'surplus population' who 'work [s] the dark underbelly of capitalism.' Inequalities enable the expropriation of backstage players, who are trapped in unfree, dependent or unwaged labour (Klein 2023).This form of labour has historically fuelled capitalist accumulation (Bhambra 2022, Beckert 2014).

The origins of identities
Why does stratification occur along lines of social identities like race, culture, gender, or religion?Some suggest that discriminating against outsiders is natural.For Putnam (2007, p. 142), this explanation finds support in 'most (though not all) empirical studies.' Cavaillé and Van der Straeten (2022) find the evidence less convincing.Either way, there are several issues with this theory.At a basic level, researcher and research subject alike face the insurmountable constraint of being historically and socially situated, making it near impossible to discern which behaviours derive from 'human nature.'Additionally, discrimination has never assumed the exact same form universally: 'there have been many different racisms' (Hall 2021, p. 59).
A more immediate challenge is that discrimination against an 'outgroup' relies on this group's existence.However, group membership is not predetermined.On the one hand, the impressive genetic diversity with which humans are endowed does not cluster neatly into racial categories (Appiah 1998).On the other hand, even if a social identity operates independently of a biological referent (like religion or culture), it is still logically insufficient for group constitution.As Sen has argued, humans possess multiple, if not infinite, affiliations, the importance of each determined by context and choice.Speaking of groups involves collapsing the multiplicity of human associations into one 'dominant system of classification ' (2006, p. 175).
If identity is not a fixed attribute, then what is it?Identity is a positioning (Hall 2021); a symbol signifying one's location in a given system of meaning.Identification happens as real individuals assume and/or are ascribed a value (Appiah 1998) on a dimension rendered salient by (ideological) context.For example, identifying as a man or woman has traditionally meant taking up residence on one side of a binary constructed on biological sex differences.Meanwhile, racial identities rely on the 'transformation of phenotypical variation into concrete systems of differentiation based on 'race' and 'culture.''(Gilroy 1987, p. 38) Both the rules of differentiation and the location of a given group within the system are subject to change.Hence, different groups, from nineteenth century Irish immigrants in the US to poor council estate residents in the UK, can be said to have acquired whiteness over time (Roediger 1999, Shilliam 2018).
Social identities offer state and market actors an easy template for structuring resource allocation, which in turn, imbues fictional categories with meaning.A vast body of literature illustrates this point by underscoring how race, though fictional, becomes 'material over time' (Tilley and Shilliam 2018, p. 534).Racialisation involves powerful actors directing material and symbolic benefitsincluding group membershiptoward some racial identities and away from others, giving those identities material relevance (Meghji 2022, p. 20).This is why Du Bois defined 'the black man' as 'a person who must ride "Jim Crow" in Georgia' (cited in Meghji 2022, p. 81): segregation and exclusion constituted, rather than accompanied, his blackness.It also explains how welfare states can affect social identities.van Staalduinen (2022) shows how social investment policies reinforce inequality by training immigrants for employment in sectors with limited upward mobility.
Identities legitimate the existing social order.For Peterson, states 'produce and sustain the structural violence' of inequality only through the 'cultural codification … of differently valorised social categories ' (2021, p. 292).By propagating ideas about the natural division of humankind into specific groups, some are positioned as more deserving than others (see van Oorschot andRoosma 2017, van Oorschot andRoosma 2000 for more on deservingness in welfare studies).

Recovering 'diversity'
What are the implications for diversity?Constructed identities lose their permanent, essential character.Diversity can then no longer be understood as a property of a given community.Instead, it expresses a historical struggle over the salient dimension(s) of variation, the meaning of similarity and difference in a given ideological context, and the identities of individuals in relation to this dimension.Ideological and material efforts to construct an 'ingroup' result in solidarity, ideological and material efforts to construct an 'outgroup' create diversity.Diversity disappears when differences lose their salience.Solidarity disappears when similarities do.
Which criteria for belonging are likely to determine perceptions of diversity?First, path dependency matters.Like theory building, ideological work is tiring.It requires joining together 'different elements,' like symbols, practices, and interpretations, into a 'chain of meanings' (Hall 2021, p. 100).If historical efforts have already furnished an internally coherent narrative, actors will find it convenient to draw from these.
Second, ideas are always attached to their carriers, who function as 'cognitive baggage handlers' (Haas 1992, p. 27).Conceptual schema compete with one another in the public sphere via their carriers (Hall 2021, p. 320).The more powerful the carrier, the more likely the classification will take.Finally, the utility of the proposed boundariestheir ability to offer a solution or explanation for a problematic or puzzling empirical realitymatters.When Fanon says, 'it is the settler who has brought the native into existence,' (Fanon 2001, p. 28), he means that the 'native' is impossible as a conceptual category without a coloniser whose life depends on the existence of immutable differences between themselves and the people they want to conquer.

Materials and methods
This article aims to build theory on ingroup and outgroup construction by welfare states using an original case study from a historical moment in which the meaning and boundaries of political membership in Europe were revisited.After the Second World War, central governments assumed ever greater responsibility for their citizens' material welfare: by the 1970s, social spending in many European countries had doubled (Ferrera 2005, p. 77).Meanwhile, after having been conscripted to fight in European armies, colonial citizens and subjects increasingly demanded the same rights as their metropolitan compatriots (Cooper 2013).The five to six million (post)colonial migrants (Buettner 2018) who arrived in Europe during the four decades after World War Two brought the conflict over political boundaries to policymakers' doorsteps.
I study the roughly 300,000 former subjects and citizens from the former Netherlands Indies (present-day Indonesia) who arrived in the Netherlands between 1945and 1963(Willems 2003, p. 34), p. 34).I selected this case for several reasons.First, it is an influential case (Gerring 2007, p. 94).The Netherlands was the third biggest empire after Britain and France (Colley 2002, pp. 172-3).Migration from the (former) Netherlands Indies, its most prized colonial possession (Borch 2017, p. 6), represented the lion's share of immigration flows into the Netherlands at the time (Obdeijn and Schrover 2008, p. 230).Additionally, it predates other (post)colonial migrations by over a decade, and there is evidence that Dutch policy attracted the attention of its French and British neighbours (Vacher-Desvernais 1962, Schuster 1999, p. 81).
Second, the group exhibits within-case variation in historically assigned racial identities.Several scholars have argued that race lacks analytic clout in the Dutch case on the grounds that 'Europeanness' supplanted explicit reference to skin colour or that religion, class, and education mattered for social status (Luttikhuis 2013, Verheijen 2021, Bosma 2009).This argumentation relies on defining race as skin colour.However, race is a mode of classification (Shilliam 2018, p. 537).It involves submitting the vastness of human genetic difference to a reductionist logic from which discrete categories of human subjects emerge.During Dutch colonial rule, 'Europeanness' and 'civilisation' constituted dominant modes of classification (Verheijen 2021, p. 453).'Foreign Oriental,' 'native' and 'European' were the formally sanctioned categories to which this schema gave rise.The categories continually failed to fully stabilise the differences they encountered.For example, 'Europeans' included totoks (white Dutch people), Australians, Japanese and US Americans (Beets et al. 2003, p. 61).Still, using racial arguments about differences in 'mental capacities' (Stoler 2008, p. 84), the colonial state concentrated citizenship rights, representation, social status and career opportunities among 'Europeans' and a select number of Indische Nederlanderspeople of 'mixed' heritage (Schuster 1999, p. 83).In 1930, both categories combined represented only 0,4% of the population (Beets et al. 2003, p. 60).At Indonesian independence, those with Dutch citizenship retained it and received two years to opt for Indonesian nationality if they met certain criteria (Schuster 1999, p. 83).Those with subject status became Indonesian citizens.
The 300,000 migrants to the Netherlands varied in their socioeconomic status, occupation, citizenship rights, and historically inherited racial identities (Oostindie 2012, p. 103).The initial wave included mostly totoks evacuated from Japanese prisoners-of-war camps (Oostindie 2012, p. 106, Willems 2001, p. 19).Over time, more Indische Nederlanders of lower socioeconomic status made the journey.In 1951, 12,500 members of a minority group known as Moluccans arrived.Because Ambon and the other islands in the Moluccas from which they hailed were among the earliest to be colonised and the most influenced by early Protestant missions, Moluccans had been targeted for recruitment by the colonial army (KNIL, Koninklijk Nederlands Indische Leger) (Jones 2007, p. 57, van Amersfoort 2004, p. 154).After KNIL was dissolved in 1950, the Dutch government reluctantly granted Moluccan soldiers and their families entry.Meanwhile, free movement between Indonesia and the Netherlands was suspended on short notice on 1 December 1952, ostensibly to guarantee 'the safety of the Netherlands.'By December 1957, most Dutch citizens had left Indonesia (Laarman 2013, p. 65, Ellemers andVaillant 1985).Yet from 1957 to 1964, about 25,000 'regrettants' (spijtoptanten)or former Dutch citizens who had opted for Indonesian nationality at independencearrived under a special admissions scheme (Beets et al. 2003, p. 61).
Although I share Paul Gilroy's (1987, p. 19) scepticism about the possibility of isolating race and racism from the complex social processes in which they are embedded, the within-case variation afforded by this case facilitates cautious inquiry into the 'nature, location, and social significance' of racial boundaries in Europe (Fox and Guglielmo 2012, p. 220).I study access to economic welfare and security by evaluating entry, citizenship, and social rights in tandem.(Legal) entry was the basic precondition for access to the Dutch welfare state, which, up until 1965 consisted of (a) an old Poor Law administered locally by charitable and religious associations and funded by municipalities, (b) contributory insurance schemes jointly administered by employers, unions and the state, and (c) from 1956 onward, the General Old Age Act (Algemene Ouderdomswet, or AOW): a flat-rate, old-age pension financed by worker contributions.However, after Indonesian independence, citizenship mediated entry rights, and independently conferred specific privileges, like entitlement to targeted welfare schemes and to the transitional rules of the AOW, which allowed those with an inadequate contribution history to benefit from a full pension at retirement (Sociale Zaken en Werkgelegenheid 1956).
Archival work began in October 2020 in the records of the Ministry of Internal Affairs, the Ministry of Social Work, the Ministry of Justice, and the Office for War Victim Care, housed in the Dutch National Archives in the Hague.Secondary literature from Dutch authors supplements my analysis, which starts in 1945, when the first evacuation from the Netherlands Indies begins, and ends in 1968, when the CCKP held its last meeting.My sampling strategy involved searching for documents using key words like 'gerepatriëerden' (repatriates).I subsequently used a snowballing technique, pursuing references to legislation or events.Key sources included legislative drafts, memos, correspondence, and meeting minutes.As 'technologies' designed to 'serve the needs and interests' of empire and state (Ghaddar and Caswell 2019, p. 79), state records typically exclude the viewpoints of marginalised communities.Future research recovering the perspectives of migrants themselves during this period is urgently needed.

Relative generosity for repatriates
Early migrants from the Netherlands Indies were classified as repatriates, a category reserved for citizens and given meaning through the institutionalisation of public and private agencies devoted to their care.In keeping with the Dutch tradition of dualism, religious and non-governmental institutions played a central role in their reception, but the central government's involvement was important and surprising given the 'laggard' character of the Dutch welfare state at the time (Oude Nijhuis 2018, p. 76).Their combined efforts mark an unmistakeable, if not uniformly successful, attempt at designating repatriates as insiders.
In an initial phase of emergency evacuation between 1945 to 1948, the Dutch Red Cross supplied medical assistance aboard repatriating ships and, together with local NGOs, packages of foodstuffs, blankets, bedlinens, and toiletries for reception centres (Willems 2001, p. 31, 36).Public assistance, coordinated by the Central Agency of War Victim Care (CBVO) of the Ministry of Internal Affairs, included giving repatriate evacuees extra coupons for rationed food and firewood (Willems 2001, p. 68).The Ministry of Foreign Affairs advanced the costs of transport to the Netherlands. 1 From 1948 to 1950, as Dutch authority in the Netherlands Indies waned, public and private efforts entered a new phase of institutionalisation. Formal responsibility for repatriates shifted to Social Assistance Services (DMZ), a different subdepartment of the Ministry of Internal Affairs (Willems 2001, p. 67).Within the Council of Ministers, prime minister Drees established a subcommittee devoted to repatriate policy known as RAVI (Raad voor Aangelegenheden van Indonesië).An interministerial committee was entrusted with coordination and advisory functions ( van Winkel 2006).Finally, at Prince Bernhard's bidding, 2 the Central Committee of Clerical and Secular Initiatives for Social Care of Repatriates (CCKP) was established.The CCKP united around twenty private organizations concerned with repatriates, including the Red Cross and Nederland Helpt Indië, as well as the Roman Catholic church, Protestant churches and trade unions ( van Winkel 2006).Its most important source of financing was a 'modest' subsidy from the Ministry of Internal Affairs (Willems 2001, p. 185, Ellemers andVaillant 1985, p. 51).Occasionally, it cooperated with public agencies, for example, in sensitising Dutch employers toward 'the labour force from Indonesia.' 3  Now formally sharing responsibility with private actors, government offered benefits, both in-kind and cash, to the newcomers.Repatriates could request targeted assistance under four different schemes: Evacuatieregeling Nederland (1945)(1946)(1947)(1948)(1949); Regeling Hulpverlening Oorlogsgetroffenen (1950)(1951)(1952)(1953)(1954)(1955); Regeling Hulpverlening Gerepatriëerden (1956)(1957)(1958)(1959)(1960); Rijksgroepregeling Gerepatrieerden (1961)(1962)(1963)(1964). 4 The latter included periodic transfers at fixed benefit levels, as well as one-off loans for housing or furniture, with repayment plans determined by the municipal council. 5In 1950, housing loans assumed the form of guesthouse contracts (contractpensions): agreements in which hotels received a tax-funded allowance for providing room and board to repatriates.Officials reported that those living in guesthouses were so content that they feared moving out. 6In practice, anxieties may have reflected more uncertainty about the future than satisfaction with the present (Molemans 2014).Either way, residents received a flat-rate allowance and advance payments on their rent. 7An official brochure suggested flexible repayment by reassuring repatriates that 'when the debt is collected, account will be taken of your financial security.' 8 Promoting take-up of these schemes, CCKP organized educational evenings all around the country. 9If ineligible, repatriates could make recourse to the Poor Law, with central government subsidizing municipalities for the cost. 10Eventually, the generosity of social provisions for repatriates raised eyebrows in official circles, prompting a gradual transition out of targeted assistance. 11

Constructing and discouraging 'Eastern' citizens
The above provisions were not equally available to all newcomers from the Netherlands Indies.Instead, repatriates were assessed on their geographical and cultural 'rootedness,' a racialised variable with colonial origins.This dimension produced two distinct categories: the 'Western' citizen and the 'Eastern' citizen.Depicted as cultural outsiders, the latter would struggle to access the same rights as the former.
Immediately after the Second World War, evacuation guidelines refashioned racialised criteria for national belonging by stipulating that potential evacuees were to be differentiated according to whether the Netherlands was their 'mother country' (Jones 2007, p. 147).In 1948, Minister without portfolio Lubbertus Götzen openly worried that a 'larger number of individuals 'rooted in the Indies' had arrived in the Netherlands than what we thought desirable' (Schuster 1999, p. 86).The meaning of 'rootedness' took shape in the early 1950s, after the jurist Philip Werner was commissioned to investigate the conditions of Indische Nederlanders in Indonesia.Werner and his team returned in 1952 with a polished distinction between 'Western' and 'Eastern' Dutch citizens, describing the latter as 'culturally, physically and psychically fixed in their natural environment,' with their children 'in more urgent need of being raised at home in the tropics' due to their 'inherently slow pace of work and other specific Eastern characteristics and behavioural traits' (Schuster 1999, p. 101).Werner's committee recommended they stay in Indonesia where they would feel more 'socially at home' (ibid: 99).Though the report was initially confidential, its recommendations reached RAVI (Willems 2001, p. 127).
The government subsequently attempted to shape entry rights in line with the 'rootedness' criterion.Because it was technically illegal to forbid entry to citizens, various informal tactics fell under discouragement policy (ontmoedigingsbeleid), which centred on keeping 'Eastern-oriented' Dutch citizens in Indonesia.The Ministry of Foreign Affairs conditioned advance transport payments on the applicant's Westernness, determined in part by 'contact officials' (contactambtenaren) aboard repatriating ships who reported on their passengers' 'orientation' (Jones 2007, p. 160).One contact official reported that when their ship docked in Ijmuiden on 30 December 1959, Minister Beerman of Justice boarded to ask about the passengers' integration prospects. 12She reassured him that 'this group of repatriates is very certainly not more Eastern oriented than the group that is already staying in [the Netherlands],' adding that at the captains' dinner, they took their place at the table in 'full regalia' (ibid.).
Key pillars of the Dutch welfare state upheld discouragement policy.In 1960, the Ministry of Social Workbudgeted for social assistance to Dutch citizens in Indonesia to prevent relocation. 13Meanwhile, local level officials responsible for the implementation of public assistance discriminated based on orientation.Social workers paying home visits could deny a repatriate permanent housing if they failed to pass as Western, due, for instance, to eating rice rather than potatoes, furnishing their house in an 'Indisch style' (this was not further elaborated), or wearing a sarong instead of European clothes (Jones 2007, p. 175, Willems 2001, p. 189).A 1954 brochure of Pro Patria, an organisation representing the interests of Indische Nederlanders, protested that 'Western' families received twice as much in financial support as 'Eastern' (Schuster 1999, p. 108).
In 1952, the Werner report was leaked (Laarman 2013, p. 61).Civil society and parliament were incensed.Jan van Baal, MP for the Protestant party (who would later govern Netherlands New Guinea), argued that the Werner committee 'proceed[ed] from a racial doctrine which [did] not differ much from that of Hitler and Alfred Rosenberg' given its belief in 'unfavourable inherited traits which stem from Indonesian ancestry' (Willems 2001, p. 128).Despite the upheaval, discouragement policy remained intact until 1956, when Minister Marga Klompé of social work removed the 'orientation' criterion for advance transport payments.However, the Western-Eastern distinction endured; a survey from 1963 asked social workers to report whether families were 'predominantly Eastern or Western.' 14   Were regrettants 'not too well-cared for?' The previous section suggests that the demarcation of the ingroup proceeded without regard to formal citizenship status, but under the preponderant influence of a vague dimension of 'rootedness.'If 'Eastern' citizens scored low on this criterion, regrettants were constructed as rooted, Western, deserving insiders.
Regrettants were former Dutch citizens who had opted for Indonesian nationality at Indonesian independence, in part due to the Dutch government's encouragement (Jones 2007, pp. 147-149). 15 Without citizenship, entry rights, as well as rights under targeted welfare schemes and AOW transitional rules, were in question.However, a powerful, cross-partisan coalition successfully protested their exclusion, prompting the introduction in 1955 of a special admissions policy for regrettants (Ellemers and Vaillant 1985, p. 43).The coalition reached peak organisational capacity in 1960 when William Lemaire -MP for the Catholic party and himself an Indische Nederlanderfounded the National Action Supporting Regrettants of Indonesia (NASSI). 16NASSI aimed at '(a) expressing the conviction of the Dutch people that our fellow kinsmen in Indonesia and in misery deserve our help, and (b) improving the actual help offered' (Roessingh 1974).Together with the CCKP, the cross-partisan NASSI formed an 'active and successful lobby' for regrettants, bombarding different ministries with memos detailing their plight (Laarman 2013, p. 72, Willems 2001, p. 162).They managed to expand entry possibilities by increasing the maximum quota under the admissions scheme from 1200 to 1800 and finally to 2000 in 1959 (Willems 2001, p. 162). 17Testifying to the political clout of the protest, on 21 September 1960 the Liberal Senator de Vos van Steenwijk asked the Minister of Social Affairs about the ineligibility of Indonesian citizens for the AOW transitional rules, asking whether this 'injustice [ought not] be removed.' 18Nine days later, the answer came in the affirmative: the Royal Decree would be modified at the shortest delay (ibid.).Earlier that year, at a Ministerial Council, the question was raised of whether regrettants 'were not too well-cared for?' 19  An anonymous minister responded: 'as a rule, the standard is indeed higher than that of the Poor Law' (ibid.).
Key to lobbyists' strategy was racialising regrettants as insiders.A series of 1960 radio speeches testify.A Liberal MP based the Dutch 'moral obligation' toward regrettants on the 'inhumanity' of 'let [ting] so many Dutch men, women and children perish' (emphasis added). 20The actress Nel Oosthout invoked the hardship of those 'born and raised under the tropical sun; of 'mixed blood' if you will and slightly darker skinned than you and I, but who have Dutch names.' 21This last subclause was crossed out and replaced with 'but who are Dutch, regardless of their origin, like you and I.' The fact that, with the stroke of a pen, a group could be moved from merely 'hav[ing] Dutch names' to '[being] Dutch, regardless of their origin,' bespeaks not only the discretion of local actors in shaping national belonging at the time, but the flimsy, unfinished fabric out of which this belonging was fashioned.The sensitivity of race in this process is reflected in Oosthout's decision not to mention regrettants' 'mixed blood,' but to instead emphasise that they had 'remained Dutch in their appearance' (emphasis added).'In their clothing' was struck through; fashion, apparently, too weak a drywall on which to hang national identity.

Stateless soldiers
If civil society contested the outsider status of some (post)colonial migrants, they did so selectively.In 1960, ten years after KNIL was dissolved, Minister Luns of Foreign Affairs promised Parliament that Moluccans be treated 'no worse than refugees.' 22This minimal promise formally established what a decade of practice had already demonstrated: unlike regrettants, Moluccans were outsiders rather than moral equals.Their status as such did not follow inevitably from exogenously given cultural distance.Instead, it responded to political expediency.As late as 1949, Dutch MPs considered Moluccans 'much closer to the Dutch citizen than many other peoples of Indonesia' (Jones 2007, p. 97), legitimising their extensive involvement in Dutch military efforts.After Indonesian independence, when this involvement was no longer needed, their cultural proximity vanished.
Moluccans acquired entry rights only after considerable foot-dragging on behalf of Dutch officials.Although totok KNIL soldiers were transferred to the Royal Netherlands Army (Willems 2001, p. 108), Moluccans were asked to serve in the Indonesian army (their former enemy) or demobilise on enemy-controlled territory.They understandably refused.Only after a court ruled against demobilisation were they transferred to the Netherlands, after which they were immediately decommissioned (van Amersfoort 2004, p. 154).The Minister of Union Affairs and Overseas Territory saw their arrival as the 'worst imaginable solution,' lamenting that Moluccans were inassimilable due to their 'physical condition,' 'fitness for work,' and 'habits of life and social views' (Laarman 2013, p. 110).Top officials like Minister van Thiel, the first Minister of Social Work, suggested deporting them and framed this as repatriation (van Amersfoort 2004, p. 154), betraying an attempt to bar them not only from entry but from the repatriate identity.Social workers actively participated in their racialisation as outsiders.One report described a white Dutch woman adapting to the Moluccan 'lifestyle' through cooking and language.In the margins, an official retorted, 'A bit fanatic, no?I think it's a little creepy' (Laarman 2013, p. 114).Two years later, the woman appeared more 'Dutch', and the official wrote, 'thank goodness.Blew over.It just goes to show: nature is always stronger than nurture!' (ibid.) Contrasted with regrettants, Moluccans struggled to access welfare and no comparably powerful lobby emerged to contest this.Citizenship rights were squarely off the table, even when the alternative was statelessness.The 1949 Citizenship Agreement held that Indonesian nationality would expire if an individual was outside of Indonesia for five years without filing for extension.On principle, most Moluccans did not do this.Consequently, over three quarters were stateless by the early 1970s (Jones 2007, p. 129).Meanwhile, juridical ambiguity prevailed.In 1968 a local police agent queried the chief of police regarding a man listed in one register as Indonesian and in another together with his children as stateless. 23The agent pleaded urgently for clearer guidelines: 'I have to work on [cases like] this daily.' 24Regarding social rights, the CCKP did not consider them within its remit.In 1953, its chair wrote to the Ministry of Social Work complaining that an office for Moluccans had been established in repatriate housing centres. 25They worried that it would have 'far-reaching psychological consequences' for other residents and lamented that the Dutch public already could not distinguish between the repatriates and Moluccans. 26 Where private assistance faltered, the state offered a modicum of in-kind assistance, initially covering paid Moluccans' housing, clothes, food, and an allowance (Jones 2007, p. 120).From 1952 this responsibility fell under the Commissioner of Ambonese Assistance (CAZ, Commissiaraat Ambonezenzorg) at the Ministry of Social Work.However, housing consisted of wooden barracks in isolated camps, including in two former Nazi transit camps (Westerbork and Vught, renamed Schattenberg and Lunetten; Laarman 2013, p. 104).Residents had to report to the camp administration if they wanted to visit relatives (Steijlen 2016, p. 240).The conditions were harsh enough that CAZ officials suggested shutting them down. 27However, residents reportedly feared falling 'under the surveillance of the alien police' if they moved, and a CAZ official accordingly requested leniency from the Head of Police. 28The latter forwarded the request to his superior at the Ministry of Justice, scribbling in the margins: 'I find the feelings expressed by the Ambonese concerned grossly exaggerated.' 29Yet he ultimately acquiesced after carefully considering the alternative possibility of forcibly interning 'these difficult people.' 30

Discussion
In recent decades, interest in the impact of 'diversity' on redistribution has exploded (Putnam 2007, Cavaillé and van der Straeten 2022, Hooghe 2009, Taylor-Gooby 2005, Tolbert and Hero 2001, Kymlicka and Banting 2006, Larsen 2011).Changes in diversity are viewed as a secular, demographic trend: exogenous and impervious to politics.Using insights from social and cultural theory (Meghji 2022, Hall 2021, Shilliam 2018, Fields and Fields 2012, Sen 2006), I offered a different take.In a universe of incalculable human difference, changes in diversity reflect changes in dominant modes of classification and in the identities to which these give rise.As new dimensions of variation compete with the old, the meanings of similarity and difference, cultural proximity and distance, and, crucially, diversity and solidarity, take shape.
The need to secure the boundaries of the ingroup was particularly acute during the Golden Age of welfare expansion.The Netherlands was no exception: it had lost 200,000 lives in the Second World War, with a further 400,000 displaced persons scattered around Europe (Schrover and Walaardt 2018).Less than 5 years after the war's end, the imperial kingdom lost its crown jewel after deploying 200,000 troopsand committing hundreds if not tens of thousands of war crimesin its name (Oostindie 2012;2017, p. 176, Lijphart 1966)).Facing the reality of (post)colonial migration from the newly independent Indonesian republic, Philip Werner and his colleagues reached for the annals of Dutch colonial law, pulling a tired, scantily clad racial criterion for ingroup belonging from the shelves.In official parlance and at the street level, geographical and cultural 'rootedness' in the West became a central determinant of insider status.
Unsurprisingly, given existing research on the 'unruly and piecemeal' character of classificatory efforts (Stoler 2008, p. 1), ascertaining the Westernness of migrants and welfare recipients was no simple feat.Social workers and 'contact officials' examined the eating, dressing, furnishing, and speaking habits of Dutch citizens to locate them within racial categories constructed on either side of a fictional Western-Eastern binary.At stake was not only the cultural proximity of individual groups, but also the specific features that mattered for cultural proximity.For example, when the Minister of Social Work and the leading private organisation for repatriate care refused to identify (Christian) Moluccans as repatriates, they were severing the links between religion and cultural proximity.The plasticity of these features and habits threatened the racial logic underpinning the classification exercise and forced officials to come to its defence.State officials portrayed changing cultural tastes as temporary deviations from a predetermined biological essence, for example by proclaiming 'nature' as 'stronger than nurture' when exalting a white client's renunciation of their 'Moluccan lifestyle' (Laarman 2013, p. 114).
Perpetually defending its rationale against an uncooperative reality is just one of many demands that racialisation makes of its practitioners.To maintain relevance, its rules need regular updates.The success of the coalition advocating for regrettants depended on reverse-engineering the meaning of Western to fit their profile.In February 1960, the contact official on board the SS Zuiderkruis highlighted fashion as relevant for its passengers' orientation.One month later, Nel Oosthout crossed out a line in her speech about fashion, opting to focus instead on regrettants 'remain[ing] Dutch in their appearance.'The process resembled what Shilliam (2018, p. 154) has identified as the whitening of the British working class.In both cases, racialisation (as white or as Western) conferred respectability, heightened the cognitive dissonance associated with their social position, and drew public sympathies for their grievances.In the wake of these efforts, the diversity that specific newcomers represented receded.
Welfare played a central role in bringing the 'rootedness' criterion to life by lending material relevance to fictional categories (Tilley andShilliam 2018, Meghji 2022).In this case, assistance available to repatriates classified as Western not only matched prevailing standards, but meaningfully surpassed them.Alongside religious and charitable initiatives, a 'laggard' welfare state (Oude Nijhuis 2018, p. 76) drew up targeted assistance schemes, subsidised repatriates' recourse to the Poor Law, and fronted the costs of their private accommodation despite pronounced housing shortages.Resource allocation proceeded in this manner even when it proved juridically complicated, as special admissions schemes, exemptions to the transitional rules for old-age pensions, and the unconventional tactic of 'discouraging' the exercise of rights all testify.The effect was the promotion of specific cultural practices.That the Dutch government no longer listed Indische Nederlanders as 'minorities' by the 1980s is no accident (Jones 2007, p. 178).Decades of cultural work had subdued the diversity that they threatened to introduce.
The point, then, is not that race or even diversity matter for inclusion in the welfare state.My main contribution is to recast identity as foundational to solidarity, but, crucially, as pliable, relational, and subject to change.In short, race has mattered, but largely via collective decisions that social actors made about the constitution of ingroup identity.Furthermore, welfare states actively shape what membership in the national community means, not only in terms of the rights it confers, but also the people it encompasses.Racial ideology has supplied a painfully resilient template for this project.Future research, besides retelling this story from the perspective of the migrants themselves, could explore the conditions under which different factorslike civil unrest, international advocacy, and alternative imaginingscurb its appeal.