What the Dutch benefits scandal and policy's focus on ‘fraud’ can teach us about the endurance of empire

Abstract

In the benefits scandal, the Dutch state wrongfully accused at least 35,000 Dutch parents of young children of committing fraud with subsidies for childcare, based on ethnic profiling. Especially families with heritage in the former Dutch colonies in the Caribbean and Suriname were hit. Because they had to compensate for this supposed ‘fraud’, families were often thrown into severe debt and years of financial and legal struggle, a struggle that is ongoing for many. For the parents of more than 2,000 children, this even led to the loss of custody of their children. The Dutch state has formally acknowledged that ‘institutional racism’ was part of the problem, yet the scandal remains analysed by many, especially white scholars and journalists, as unintended, incidental and unrelated to Dutch colonial history. This commentary addresses the scandal as a case of an enduring colonial legacy of targeting racialised and gendered citizens in social policy. ‘Fraud’ is a social policy category through which this is made possible and violence continues to be inflicted upon racialised and gendered members of the Dutch population.

‘Fraudulence’ and the colonial cultural archive

‘Fraud’ has been a focal point of austerity and neoliberal welfare reforms for decades now in many welfare state contexts. The deservingness of citizens of certain social policy affordances is under much public and political scrutiny. In the Netherlands, years of extremely harsh anti-fraud policies were under investigation in a special parliamentary inquiry that was completed in February 2024 (Tweede Kamer, 2024). While the inquiry was meant to investigate all fraud policies, the concrete cause of the attention given was the benefits scandal (toeslagenschandaal). One of the largest public and political scandals in recent Dutch history, the benefits scandal concerns at least 35,000 parents of young children that were wrongfully accused of committing fraud with benefits that were to compensate for childcare costs (Dienst Toeslagen, 2024). The accusations, we now know, primarily targeted parents of colour and especially single mothers. In this commentary article, we consider what scholars may learn from the benefits scandal about the endurance of empire. Our question is how we may understand the racialised targeting of parents in this scandal and the use of the category of ‘fraud’ in this targeting.

A useful way to conceptualise how the Dutch empire endures is Gloria Wekker's work on white innocence and the Dutch colonial cultural archive (2016). Wekker builds on Edward Saïd's notion of a cultural archive to argue that there is ‘an unacknowledged reservoir of knowledge and feelings based on four hundred years of imperial rule' and that this has ‘played a vital but unacknowledged part in the dominant meaning-making processes taking place in Dutch society, until now' (Wekker, 2016: 3). The colonial cultural archive refers to the (power relations embedded within) racialised knowledge and affects that were ‘deposited within metropolitan populations' (Wekker, 2016: 19), and are institutionalised in contemporary policies, popular culture and everyday common sense (ibid.). In 2003, the State Secretary for Social Affairs and Employment, Mark Rutte (who later became Prime Minister), encouraged municipalities to target Dutch welfare recipients of Somali descent in their fraud investigations. This was considered discriminatory by the Dutch court (Rechtbank Haarlem, 2007). More generally, in welfare fraud investigations, municipalities often use ‘risk profiles’ and algorithms that include language skills, ‘birthplace outside the Netherlands’, ‘country of origin outside the Netherlands’ and ‘double nationality’ or ‘non-Dutch nationality’ (College voor de Rechten van de Mens, 2021; Lighthouse reports, 2023). Similarly, in the case of the benefits scandal, the Dutch colonial cultural archive and its ‘racialized common sense' (Wekker, 2016: 20) informed decisions about Dutch parents who were considered dishonest and fraudulent, undeserving of benefits and deserving of harsh penalties, as a result of their racialised ‘otherness’.

While formally acknowledged as a form of ‘institutional racism’ (Rijksoverheid, 2022), the benefits scandal is still routinely conceptualised as something other than racism, especially by Dutch white scholars and journalists. The scandal, then, becomes about algorithms-based ‘administrative exclusion’ (Peeters and Widlak, 2023), about ‘vulnerable populations’ that have a hard time navigating the Dutch state (Vollebergh, 2023), or even about ethical decision-making processes and forms of repair that apparently have little to do with racism (Claassens and Robeyns, 2022). While Dutch people of colour were the first to point out the racialised targeting in the benefits scandal, and continue to work to keep the issue on the public agenda as such (Achbab, 2021; Sitalsing, 2022), the rebranding of the scandal as about something other than racism seems a perfect example of what Wekker has termed ‘white innocence’: a Dutch self-representation that centres around the notion of ‘being a small, but just, ethical nation; color-blind, thus free of racism' (2016: 2) in which the Dutch are imagined as white and innocent – an often wilful and violent innocence.

The case: The Dutch benefits scandal

The benefits scandal led to the resignation of prime minister Mark Rutte's third cabinet in 2021. It casts a long shadow over contemporary Dutch politics, in part because public trust in government was seriously hurt by it (Kanne and Driessen, 2021). The term ‘benefits scandal’ refers to a set of practices in which the state and its tax authorities wrongfully accused parents of young children of committing fraud with the tax benefits that compensate for childcare expenses. Scholars have termed this targeting ‘administrative exclusion’ (Peeters and Widlak, 2023), but it was precisely the parents’ inclusion into compensation schemes and subsequent targeting as undeserving, based on racist ideas informing policy decisions and algorithms, that was harmful and violent.

Approximately 35,000 parents (Dienst Toeslagen, 2024) that were thusly accused were often thrown into severe debt, because they were obligated to refund all the benefits that they had thus far received, which could amount to debts of tens of thousands or even over a hundred thousand euros, depending on how many years they received the benefits and for how many children. With claims made on their income by debt collectors that included state agencies, they were also often thrown into severe poverty, frequently unable to pay for necessities such as food and heating. Many accused parents are still, to this day, struggling in legal and financial battles to prove their innocence and claim compensation (Nationale Ombudsman, 2022).

After years of investigations, of journalists, researchers, ministries and parliament itself, it has become clear that the state mainly accused parents with a migrant background (College voor de Rechten van de Mens, 2022). Selection took place frequently, but not always, through the use of algorithms (Konaté and Pali, 2023). Families with roots in the former Dutch colonies of Suriname or the Dutch Caribbean islands were hit particularly frequently, as were single mothers. Roughly half of the families hit by the scandal were single parent families and 3105 out of the 10,270 households in the available data were from Surinamese or Caribbean descent, in addition to large groups with heritage in Turkey and Morocco. In total, in 7260 of the households either the parent themselves or one of both of their parents were not born in the Netherlands (CBS, 2022).

The Dutch state has, in 2022, formally acknowledged that its actions were an example of ‘institutional racism’ (Rijksoverheid, 2022). Racism could apparently no longer be denied, but the term ‘institutional racism’ offered the opportunity to state actors to claim that only the results of their policy were racist, not the intentions or the actors themselves. State secretary Marnix van Rij put it like this: ‘The examples (…) are unacceptable. There can and must be no room for institutional racism in our society. I am now (…) aware that certain groups in our population (bevolkingsgroepen) were put at a disadvantage. (…) Even as there has been no ideology that categorises people into races, the evident practices were very painful' (Ibid.). The separation between intent of harm and harmful outcomes here serves the protection of state actors like top administrators, that have to this day not been held accountable for their racialised targeting. It also serves the self-image of the Netherlands as an ‘ethically just nation' (Wekker, 2016: 5), where racism will only appear as unintentional. It is not for nothing that Sheila Sitalsing, a journalist and columnist of Surinamese descent wrote full of anger and sarcasm in response to Van Rij's letter: ‘after three years of investigations into racism in the tax authorities, we now have a letter full of legal bullshit (‘juridisch gelul’). (…) the institutional racism was tiny. Unintended. All in good faith' (2022).

The impact of the accusations of fraud and its legal consequences have been severe for the victims. Salama Konaté and Brunilda Pali's (2023) qualitative inquiry into the harms endured by victims lays out the seriousness of the harm done to 1) victims’ wellbeing, 2) their children's long-term wellbeing, 3) relationships with others, 4) their attitudes towards state institutions and 5) their hopes and needs for the future. For the families of more than 2000 children, the impact also included parents’ loss of custody of their children (CBS, 2022) and victims have in some cases committed suicide (Konaté and Pali, 2023). The Dutch ombudspersons for children also indicated that many children of wrongfully accused parents are and will continue to be severely impacted, for example because they left school early to provide for their families (Van der Does et al., 2022).

When the state and its agencies accused mostly mothers of colour of fraud, they were often denied the opportunity to defend themselves, found little legal support (in part because of austerity measures on publicly available legal assistance in the same years and in many cases because they were flat-out denied legal assistance (Verhoeven et al., 2023)) and otherwise became estranged from many people in their lives because it was so difficult to show their innocence. While the benefits scandal may appear as an ‘administrative’ measure for some, it became a life and death issue for the victims involved and one that was directed specifically and deliberately at racialised and gendered citizens.

The category of ‘fraud’

This case shows how the category of ‘fraud’ operates in contemporary social policy to target and inflict violence upon racialised and gendered citizens. The category ‘fraud’ has a long history in social policy. With neoliberal welfare reform, characterised by ruinations of social protection and increasing punitive interventions into the lives of people who are deemed unproductive, irresponsible and undeserving, the focus on benefit fraud intensified, resulting in new policies and larger, further reaching anti-fraud bureaucracies across many well-developed welfare states. While a focus on welfare fraud has always been part of welfare policies, the neoliberal turn marks a break from the post-war ‘Keynesian’ period and entails a particular logic of ‘the welfare state in crisis’ that needs protection from ‘fraudulent’ and ‘undeserving’ ‘others’ in order to retain its legitimacy and protective functions for those who are considered ‘deserving’ and ‘belonging’ to the nation state. ‘Fraud’ came to encompass a wider range of actions, increasingly replacing alternative notions of ‘abuse’ and ‘error’, enabling more punitive interventions (Gaffney and Millar, 2020; Vonk, 2014; Wilcock, 2024).

In the case of Ireland, this has led to the following list of behaviours that are considered to constitute ‘fraud’: ‘concurrent working & claiming', ‘nondisclosure of means', ‘multiple claiming or personation', ‘life events' (e.g., not reporting changes in marital status), ‘cohabitation', ‘social insurance and employer noncompliance' and ‘non residency in the state' (Gaffney and Millar, 2020: 79). In the case of Sweden, the Benefit Fraud Act ‘penalizes welfare recipients for deliberately or by negligence attempting to receive undue compensation, with a maximum sentence of six years imprisonment' (Stenström, 2021: 775). In the Netherlands, ‘fraud’ entails the provision of incorrect and/or incomplete information, or the concealment or failure to provide (timely) information relevant to the determination of the right to benefit and the duration and amount of the benefit, with the result that a benefit is wholly or is partly provided incorrectly (Fenger and Voorberg, 2013: 14). Penalties imposed can entail a temporary or permanent lower amount or exclusion of benefits, repayment, a fine, community service or imprisonment with a maximum sentence of five months (Goudswaard et al., 2013; Tollenaar, 2019).

These example show how ‘fraud’ is a wide and vague, yet important policy category. It can be a tool in institutional violence, as it causes physical and psychological harm through legitimate, ‘ordinary’ bureaucratic practices (Cooper and Whyte, 2017). Moreover, it enables welfare states to target marginalised, gendered and racialised populations ‘in ways that reproduce and increase the likelihood of other, ongoing, violent circumstances occurring' (Cooper and Whyte, 2022: 208). In the Netherlands, sanctions resulting from a welfare fraud conviction are explicitly intended to inflict harm (‘leedtoevoeging’) on the offender in the interest of deterrence (Goudswaard et al., 2013; Tollenaar, 2019). The logic is, often, that those worthy of protection need to be protected against fraudulent ‘others’. Contemporary welfare states use automated decision-making (ADM) systems based on ‘risk profiling’ and algorithms to detect fraud, which reinforces the discriminatory and stigmatising effects of fraud detection (Appelman et al., 2021; College voor de Rechten van de Mens, 2021; Lighthouse reports, 2023).

A growing literature shows how European welfare states have always been harmful, offering social protection to some citizens, while excluding, exploiting and inflicting violence on classed, racialised and gendered members of the imperial state. From the onset, those living in overseas parts of the empire, people without formal employment status and/or those who were not granted legal status (mainly women and people of colour), were excluded from social rights (Bhambra, 2022; Hemels, 2019; Jones, 2014a; Orloff and Laperrière, 2021). Yet, this does not mean that welfare state violence was only inflicted on ‘excluded’ populations. Precisely inclusion, for many, meant to be met with disciplining or even eugenic measures, both in the metropole and in the colonies (Lucassen, 2010; Schrauwers, 2001; Van Wel, 1992). Moreover, as Bhambra (2022) argues for the case of Britain, although those living in the colonies were excluded from relations of redistribution, they were included in relations of extraction, contributing to the development of the welfare state by paying taxes and fighting wars, while not reaping the benefits. Active targeting of classed, racialised and gendered populations never disappeared and revived under neoliberal welfare reform and austerity (Cooper and Whyte, 2017; Tyler, 2022; Wilcock, 2024, for the case of the Netherlands, see Jones, 2014b; Schields, 2024).

Contrary to the dominant idea in Dutch political and public debate that institutional racism in the case of the benefits scandal was unintended, we contend that it is the result of the institutionalisation of the Dutch colonial cultural archive in public policies and government practices that (implicitly) prioritise white citizens, constructing them as ‘entitled’, while problematising racialised ‘others’; those who are considered to be ‘out of place’ in the Netherlands (Jones, 2014b: 66) and pose a threat to the legitimacy of social services, because of their perceived inherent and fundamental (moral and cultural) difference and inferiority. Through the category of ‘fraud’, these racialised ‘others’ are not only excluded from social rights they are entitled to as citizens – thereby becoming ‘semi-citizens’ (Jones, 2014b) – they are also sometimes actively ruined by the welfare state. The gendered and racialised targets of the benefits scandal are, thus, excluded from relations of distribution (Bhambra, 2022), yet included in relations of violent domination. The targeting of gendered and racialised citizens through the category of ‘fraud’ is, therefore, an example of an enduring Dutch colonial legacy of institutional violence. However, even in the special parliamentary investigation, racism disappeared from the analysis: rather than investigating the deep roots of the racialised targeting, the report speaks in the universalising language of ‘people’ and the ways in which anti-fraud policies ‘had violated (their) fundamental rights and cast aside the rule of law’ (Tweede Kamer, 2024).

Conclusion

The case of the Dutch benefits scandal shows how colonial history lives on in social policy and the lives of Dutch citizens, in the claims they can make and the violence they still endure. ‘Fraud’ appears as a social policy category through which targeted exploitation and harm is made possible and violence continues to be inflicted. The benefits scandal is not an exception, but an example of a pattern. In the Netherlands, the targeting of perceived ‘fraudulent’ racialised and gendered citizens also happens in the case of welfare benefits, where predominantly young single mothers receiving welfare benefits, while not having a good command of the Dutch language, are targeted (Lighthouse reports, 2023). In the case of student allowances, the Dutch study financing provider Duo mainly accused students with a migrant background of fraud (Belleman et al., 2023). And the Dutch tax authorities continue to use their discriminatory algorithms in their fraud-detection practices, despite the outrage about the benefits scandal (Davidson and Brommersma, 2023). The Netherlands is not an exception either, as scholars have shown similar patterns in, for example, Australia (Wilcock, 2024), Canada (Chan, 2011) and Denmark (Jørgensen, 2018). Although the colonial cultural archive and its racialised common sense is not static or essential and differs across time and place, in many welfare state contexts, it shapes the thinking, feeling and acting toward perceived ‘fraudulent’ citizens.

Importantly, in the context of the Netherlands, where race is predominantly considered a non-issue and institutional racism perceived as unintended, incidental and unrelated to ‘persistent imperial patterns of thought and affect' (Wekker, 2016: 25), it is difficult to convince various audiences (including scholars) of the appropriateness of analysing the benefits scandal in terms of institutional racism and its relation to Dutch colonial history. Whereas mainly people of colour argue for such an analysis, it is striking how white scholars and journalists often refrain from mentioning race and racism at all when analysing this case. Similarly, the report resulting from the first parliamentary inquiry into the scandal, does not mention racism (Tweede Kamer, 2020) and outsourced the task to scrutinise the role of racial discrimination to the Dutch Data Protection Authorities, who concluded that the tax authorities had made illegal use of personal data, namely parents’ nationality, and considers this ‘discriminatory’ and ‘unseemly’ (Autoriteit Persoonsgegevens, 2020), yet this has not resulted in a focus on racism in the investigation of anti-fraud policies (Tweede Kamer, 2024). The investment in ‘colour-blindness’ or ‘smug ignorance: (aggressively) rejecting the possibility to know' (Essed and Hoving in Wekker, 2016: 18) reproduces the dominant way in which ‘the Dutch’ think of themselves, ‘as being inherently on the moral and ethical high ground, thus a guiding light to other folks and nations' (Wekker, 2016: 2), exactly while they ruin the lives and livelihoods of gendered and racialised citizens.