Scope conditions at the formal–informal interface

The material does not identify a distinct ‘informal worker’ type. Instead, it identifies recurrent situations in which small-scale undeclared earning became relevant: inaccessible formal employment, weak or opaque formal returns, unstable work–benefit transitions, and care-constrained paid work. These conditions made informal supplementation materially relevant, morally defensible, or explicitly rejected in favour of formal work; they shaped participation without determining it.

Entry into formal work was often blocked by institutional gatekeeping rather than by unwillingness to work (Brodkin and Majmundar, Reference Brodkin and Majmundar2010; Carpenter et al., Reference Carpenter, Knepper, Erickson and Ross2015). A 52-year-old Dutch single woman [104] wanted computer training but was told by the welfare office that she had to remain immediately available for work. A 36-year-old Cape Verdean man [207] had prior truck-driving experience but could not convert this experience into employment because employers required a Dutch licence that he could not finance. A 32-year-old Dutch welfare recipient [175] had completed taxi training but could not move into work after the welfare office refused a loan for the licence and vehicle. Together, these cases show that formal labour-market entry could remain blocked despite skills, plans, and clear work intentions.

Even where entry into formal work was possible, formal uptake did not always generate a reliable household gain (Cairo and Mahlstedt, Reference Cairo and Mahlstedt2023). Respondent [201], a 50-year-old Cape Verdean mother, described a subsidised job as financially disappointing and noted that she would not have accepted it had she known that her income would remain almost unchanged. Respondent [321], a 42-year-old Surinamese woman, made the same point more directly: Once rent-subsidy losses were included, ‘there is no incentive to work’. Here, formal employment appeared not as a clear improvement, but as an uncertain or negligible gain.

Administrative instability created a further constraint (Moynihan et al., Reference Moynihan, Herd and Harvey2015). Respondent [148], a 27-year-old Surinamese single mother, concealed a new job during a probationary period because a failed job match could trigger prolonged bureaucratic disruption before benefit payments resumed. Without a secure contract, movement from benefit into work created short-term exposure rather than stability. The central risk was transition failure followed by income uncertainty.

Care and time constraints shaped both formal and informal work through household organisation (Bettendorf et al., Reference Bettendorf, Jongen and Muller2015; Knoef and van Ours, Reference Knoef and van Ours2016). Respondent [121], a 32-year-old Dutch single parent, left a job because early departures, childcare demands, and negligible financial gain could not be sustained together. Respondent [231], a 32-year-old Surinamese single mother in a subsidised-work/job-training trajectory, described how formal childcare could not be arranged, making standard employment unworkable and shifting feasible earning into home-based arrangements. Respondent [354], a 33-year-old Surinamese mother, approached the same condition from the opposite direction: If childcare had to be arranged and financed either way, she preferred a legal job to undeclared side-work. Care constraints did not automatically produce informality, but they shaped which forms of earning were feasible.

These four conditions clarify when undeclared earning became relevant, but not how it was organised; under them, undeclared work appeared as a bounded supplement, temporary fallback, or practical adjustment.

Mechanism 1: moral norms and the boundary of legitimate informality

The first mechanism is moral classification (Boltanski and Thévenot, Reference Boltanski and Thévenot2006). Respondents evaluated undeclared earning in terms of fairness, harm, need, greed, obligation, and acceptable gain (Meuleman et al., Reference Meuleman, Roosma and Abts2020). These judgements did not simply mirror legal categories (Abid et al., Reference Abid, Bothello, Ul-Haq and Ahmadsimab2023). Instead, respondents distinguished between limited supplementation and abuse, necessity and opportunism, or acceptable side earning and morally excessive evasion.

Respondents drew sharply different moral boundaries (Bicchieri, Reference Bicchieri2017; Luttmer and Singhal, Reference Luttmer and Singhal2014). Respondent [105], a 21-year-old Surinamese single mother with young children and not yet obliged to work, approved of off-the-books work because ‘everybody is happy’ and ‘you do not hurt anybody’. Respondent [117], a 29-year-old Moroccan single mother available for the labour market, rejected undeclared work by describing people who worked off the books as ‘thieves’. A 58-year-old Dutch participant classified as medically unfit [110] called undeclared work ‘very unfair’ to those who tried to do things properly and linked it to greed. Informality could thus be normalised, condemned, or treated as morally ambiguous.

More common was bounded tolerance of limited supplementation (Luttmer and Singhal, Reference Luttmer and Singhal2014; Wenzel, Reference Wenzel2005). Respondent [119], a 73-year-old Dutch pensioner, accepted undeclared work as ‘an addition to the benefit’ because she understood that people might otherwise fail to get by, but rejected ‘a full-time job next to the benefit’. A supplementary passage shows the same boundary logic: Respondent [141], a 40-year-old Dutch former nurse classified as medically unfit, drew a similar line after taking a small off-the-books cleaning job herself. Occasional supplementation was understandable, while people working “on the side all week” crossed the line. Here, the relevant moral question was not only whether income was declared but also whether the scale and purpose of the activity remained defensible.

Respondent [251], a Turkish single mother available for the labour market, expressed this distinction explicitly. Although she defined undeclared work as work not reported to the welfare office or tax authorities, she stated: ‘If someone has little income and cannot make ends meet, then I find it completely normal that they work off the books. What else should they do?’ At the same time, she rejected such behaviour among those with sufficient income. Tolerance therefore depended on perceived necessity, not on legality alone.

Respondent [335], a 46-year-old Dutch woman available for the labour market, personally opposed undeclared work. Yet she treated it as understandable under hardship (Feld and Frey, Reference Feld and Frey2007; Wenzel, Reference Wenzel2005). She defined it as work over which ‘no tax is paid’ and said she would feel ‘incredibly guilty’, yet added: ‘If you have to live on such a minimum… I can completely understand it. If I could not make ends meet, I would be forced to do it myself’. Personal rejection and contextual justification coexisted within the same legitimacy judgement. For the previously discussed single mother [231], undeclared home-based work became morally defensible only after a formal job opportunity was blocked by the welfare office’s refusal to provide childcare support. She knew that off-the-books work was ‘not right’ and ‘not allowed’, yet added: ‘I saw it, at that moment, as the only way out’. Necessity did not erase the illegality of the act, but it changed its moral interpretation.

Respondent [163], a 65-year-old Turkish man in an extended household, classified as medically unfit/disabled, rejected undeclared work categorically: ‘Working off the books is wrong. Our religion does not allow it’ (Bicchieri, Reference Bicchieri2017; Tyler, Reference Tyler2006). In this case, moral norms excluded informal earning regardless of circumstance.

Across these cases, moral norms bounded informal earning: Limited supplementation could be tolerated, while greed, durable evasion, or full-time undeclared work alongside benefits was condemned. Where reasoning shifted to detection, fines, repayment, or benefit loss, the operative logic became sanction expectation rather than moral classification (Kirchler et al., Reference Kirchler, Hoelzl and Wahl2008).

Mechanism 2: trust relations and social access

The second mechanism concerns access. Trust mattered first at entry and then in the coordination of repeated, low-supervision exchange. It was not a general stock of social capital. It was a relational filter through which actors were admitted to small jobs, occasional tasks, or informal earning opportunities (Hardin, Reference Hardin2002; Marin, Reference Marin2012; Smith, Reference Smith2005). Respondents had to be known, recommended, or familiar enough to be trusted outside formal recruitment (DiMaggio and Louch, Reference DiMaggio and Louch1998; Pallais and Sands, Reference Pallais and Sands2016).

Respondent [204], a Cape Verdean-Dutch man in his mid-twenties living alone, described access through brokerage rather than anonymous exchange (Stovel and Shaw, Reference Stovel and Shaw2012). He positioned himself between people and opportunities: ‘People call me up and say, can you arrange this or that?’ His role connected information, people, and opportunities across separate circuits. He also distinguished between partners with whom arrangements could be made reliably and those who did not pay or took too long. Access depended on selective inclusion based on perceived reliability.

Other cases showed the same access logic in smaller jobs. Respondent [151], a 43-year-old Dutch mother in a subsidised-work trajectory, entered side-work through a friend’s shop, where she cleaned for six weeks before looking elsewhere. Respondent [166], a 38-year-old Dutch woman in a subsidised-work trajectory, described her husband’s informal odd jobs as coming through ‘friends, acquaintances and family’, because people already knew what he could do and called when such work was needed. In these cases, prior recognition of competence created access to informal work (Pallais and Sands, Reference Pallais and Sands2016; Smith, Reference Smith2005).

Kinship mattered when it connected respondents to wider exchange circuits. A 50-year-old Dutch man classified as medically unfit [122] obtained off-the-books work through his brother-in-law, who was already ‘in the circuit’ and connected him to a workshop via broader pub-based contacts. The mechanism was referral into an existing opportunity circuit, not kinship intimacy itself (Kramarz and Skans, Reference Kramarz and Skans2014; Stovel and Shaw, Reference Stovel and Shaw2012).

Trust also operated through routine and entrusted repetition. Respondent [225], a 66-year-old Dutch pensioner, cleaned for a man across the street, held the key to his house, and entered daily for a short clean and weekly for a longer one. A supplementary routine-trust trace comes from respondent [133], a 26-year-old Dutch single mother classified as working poor, who described entering with a key, knowing what had to be done, and finding the money ready in the house. In these cases, familiarity substituted for formal contracting and made low-supervision exchange possible (DiMaggio and Louch, Reference DiMaggio and Louch1998; Hardin, Reference Hardin2002; Safuta and Camargo, Reference Safuta and Camargo2019). Across these cases, need alone did not create access; access depended on brokerage, referral, recognised competence, and entrusted routine.

Mechanism 3: community-based enforcement and reciprocal governance

Community-based enforcement regulates repeated informal exchange among people who meet, observe, remember, and deal with one another. It is narrower than general social control or welfare-sanction fear: observed conduct in a repeated relation must affect future exchange, turning visibility into inclusion, tolerance, withdrawal, or exclusion.

It resembles private ordering: Where formal contracts, courts, or official enforcement are absent, unusable, or risky, cooperation can be sustained through repeated dealings, reputational information, ostracism, and exclusion (Bernstein, Reference Bernstein1992; Ellickson, Reference Ellickson1991; Greif, Reference Greif1993; Kandori, Reference Kandori1992; Milgrom et al., Reference Milgrom, North and Weingast1990). Here, enforcement is fragile and less formalised: not stable sanctioning, but traces of repeated relations, visibility, conduct evaluation, reputational memory, and consequences for work, credit, information, or tolerance.

Respondent [136], single, around forty, childless, formally unemployed and on social assistance, is clearest. His informal economic life moved through a neighbourhood meeting place, local bar, informal work broker, lottery mediation, borrowing relations, and small work crews. He described his weekly routine as ‘the same little circuit’, and work for the broker as something he did ‘usually’, not once. Repetition made conduct observable: People knew who was available, who had money, who borrowed, who repaid, who worked hard, and who avoided effort.

Visibility was central. After an earlier undeclared newspaper round, [136] was confronted by the welfare office with ‘an anonymous tip’, after which he felt as if ‘there was an enemy behind every tree’. This marks the local-formal interface. Horizontally, he had to ‘watch who you tell things to’, because information could be passed on, and ‘as soon as they see that you are doing a bit better than someone else, then there is trouble’. Being seen created opportunity and risk.

Visible conduct was evaluated. A poorly performing co-worker was criticised as someone who ‘would rather do nothing than do a little’. It was ‘shameful to hold out your hand for money in the evening’ after contributing so little. By contrast, [136] presented his own value through punctuality and competence: ‘I am never late for anything’, and ‘if you are a bit handy, that makes a difference’. Those who consumed without contributing were ‘freeloading on others’. Reciprocity transfers not only resources but also obligations and moral assessments (Molm, Reference Molm2010).

Evaluation became enforcement when remembered and carried into future dealings. In [136], reputational memory circulated through talk. Others might label him ‘the borrower’, and negative stories were ‘talked around’. Reputation could also be positive: People told his mother he was ‘always proper, always tidy’. The strongest evidence concerns the informal work broker. At first, the broker ‘did not want to see’ that another worker contributed little; later ‘he had seen it’. Once underperformance was recognised, future access changed. The poorly performing co-worker was no longer taken along, while [136] was preferred: ‘now he has seen it; when he asks someone along, he asks me’. Reliable conduct was rewarded with renewed recruitment: ‘if you sign up for something, you have to give yourself to it. That is why people often ask you again — a job here, a job there’.

The same logic appeared more weakly in [136]’s credit relation with a local bar owner. Debt was tolerated as long as he visibly brought money back: ‘at least I come and bring my money’. Enforcement operated through conditional tolerance rather than immediate exclusion; actors remained included while their conduct remained minimally credible.

Other cases provide partial and boundary evidence. Respondent [225], 66, Dutch, living alone on an old-age pension after long-term social assistance, shows repeated relations, trust, and selective monitoring. She cleaned undeclared for a man across the street, held his key, and had ‘free rein’: routine trust enabling low-supervision exchange. Her wider neighbourhood account shows visibility: card games for money outside, police passing by, observed drug premises, and underletting visible through ‘different faces’ in the building. She reported drug premises, but not underletting, because she was ‘not a traitor’. This shows distinctions between violations to report, tolerate, or ignore, not [136]’s work-circuit enforcement.

Respondent [114], a Dutch pensioner living alone, adds softer reciprocal governance: neighbourhood babysitting, club contacts, weekly coffee, small gifts, repair help, and later a packet of rolling tobacco as a return gesture. These exchanges lacked formal pricing and strict sanctioning, but small gestures maintained local balance. She also drew moral boundaries: A welfare mother earning ‘a little on the side’ was understandable, but people earning substantially while on benefits and ‘bragging about it’ were resented. This is adjacent evidence: conditional tolerance and reciprocal balancing, not a complete enforcement sequence.

A final boundary appears when neighbourhood knowledge becomes administratively actionable rather than socially sanctioned. Respondent [305], a Dutch pensioner living alone, marks the closest boundary between community-based enforcement and formal welfare control. She suspected that a small cleaning arrangement had ‘probably been reported‘, after which the Social Service appeared; after earlier inspection, she ‘would not dare’ again. Local visibility may therefore have exposed the arrangement to formal intervention, but the sanctioning force came from the welfare office. Respondent [207], mentioned earlier, marks the stricter boundary: his withdrawal from side work followed administrative detection and repayment, not exclusion or loss of trust within a repeated exchange relation. These cases show how neighbourhood information can activate formal welfare control, while clarifying that formal sanction risk remains distinct from community-based enforcement proper (Gilliom, Reference Gilliom2001; Soss et al., Reference Soss, Fording and Schram2011).

Community-based enforcement is not general neighbourhood policing, mere secrecy, or formal welfare surveillance. It is repeated relations making conduct visible, evaluated, and consequential for future exchanges. In [136], reliable actors are asked again, and unreliable actors lose access; in [114] and [225], local actors maintain reciprocal balance, tolerate some violations, and mark others as reportable or excessive; in [207] and [305], the mechanism is formal sanction risk. The effect is selective but important: Access, tolerance, credit, and cooperation depend on being seen as reliable, discreet, and minimally reciprocal. In constrained livelihoods where earning, security, and social obligation intersect (Carr et al., Reference Carr, Napolitano and Keating2007), formal rules are joined by reputational consequences.

Synthesis: the triadic governance of informal exchange

The findings show that formal constraints explained when informal supplementation became relevant, but not how it was governed. Table 3 summarises the proposition-level synthesis and shows how the three mechanisms interact under the identified scope conditions. Informal exchange became governable when three analytically distinct mechanisms aligned: Moral norms bounded the acceptable scale and purpose of informal earning, trust relations opened access to opportunities, and community-based enforcement stabilised repeated exchange when conduct became visible and consequential. The full triad was not present in every case. Some cases showed moral evaluation, others trust-based access or routine exchange, and only a smaller subset showed repeated interaction in which conduct affected future dealings. The contribution therefore lies not in claiming that all three mechanisms were always present, but in showing how they could align under particular conditions. Informal economic activity was coordinated through repeated relations and adjusted under constraint. It did not form a stable alternative to formal labour-market participation.

Table 3.

Proposition-level synthesis of findings