Persistently non-compliant employment practice in the informal economy: permissive visibility in a multiple regulator setting

Abstract

The growing significance of non-compliant employment practice in the British economy has motivated scrutiny of the effectiveness of current regulation. In some markets, charges of labour exploitation, underpayment of the national minimum wage and associated ‘wage theft’ from workers are rife where business operations are characterised by academics, regulators and stakeholders as exuding ‘permissive visibility’. The current landscape of enforcement and regulation of informal business and employment practices features complex structural and operational issues for regulators subject to tight resource constraints. These enable permissiveness and offer scope for strategic regulatory tolerance of some violation types, possibly to raise compliance rates for other types of violations. Drawing on extensive empirical evidence and qualitative data sources in one market sector (hand car washes), this study investigates some key hypotheses focussing on compliance and responses by businesses and regulators to the extant regulatory regime. These inform a pragmatic institutional analysis considering the merits of some movement towards a single enforcement body instead of the existing arrangements featuring multiple regulatory institutions.

1. Introduction

In the British economy, non-compliant informal employment is not a residual category. Non-compliant businesses produce 10% of Britain’s Gross Domestic Product (Clowes, 2020; Johnson and Ghiglione, 2020) and sustain two and a half million workers, a number equal to 9% of the formal private sector working population generating £233 billion per year (ACCA, 2017). A prominently cited definition of informal business activity refers to ‘income generation’ that appears beyond regulation by the institutions of society in legal and social environments where similar activities are subject to regulatory enforcement (Castells and Portes, 1989, p. 12). The research reported herein takes this argument forward by suggesting that all businesses are subject to regulation but that some choose to remain non-compliant because of Britain’s piecemeal regulatory framework. Non-compliant informal businesses may be weakly regulated even though these business owners choose to deploy highly exploitative labour practices.

Exploitative labour practices centre on deliberate and systematic underpayment of Britain’s national minimum wage, therein underpayment represents wage theft from exploited workers. This in turn enables employers to avoid due taxation where many non-compliant workplaces are frequently unsafe for workers and customers regarding environmental and health and safety standards. The COVID-19 pandemic has exacerbated and reinforced this evidence base leading academics, stakeholders and some regulators to describe the regulatory framework as one that exudes ‘permissive visibility’ (Clark et al., 2020a, 2020b). This permissiveness stimulates a specific research question: how does the established regulatory framework enable many informal businesses to condone the use of informal non-compliant employment practices and to do so in plain view?

What prompts this study is the current policy consultation on creating a single enforcement body (SEB) for employment rights in the UK (BEIS, 2019). Movement to a SEB, which is a manifesto commitment of the current government, would entail a radical departure from the existing mosaic and hierarchy of regulatory institutions, each with different competencies, objectives, cultures and complementarities. Arguably, it would also risk falling into a ‘panacea trap’ (Ostrom et al., 2007). The single panacea in this case being the establishment of a single new all-powerful regulator/enforcement body that may potentially yield inferior outcomes and further adverse unintended consequences as compared to the outcomes arising from the existing multiple institutional arrangements. Accordingly, this study adopts a pragmatic institutional analytical perspective in order to (i) understand and account for the persistence of non-compliant employment practice in the current context and (ii) begin to explore the case for movement from the status quo to various gradations of cooperation among multiple regulators and perhaps onwards through to the establishment of a SEB.

The theorisation outlined in later sections suggests that the current landscape of enforcement and regulation of informal business and employment practices highlights several complex structural and operational issues in regulatory activity, leading to a permissiveness by regulators. To test this theorisation, the research empirically evaluates five hypotheses that help frame an investigation of the regulation of hand car washes in England and Wales. This sector was chosen because the Director of Labour Market Enforcement (DLME) and Parliamentary committees identify hand car washes in publications and policy announcements¹ as a sector where workers are significantly at risk of unlawful labour market exploitation (2018).

The article divides into four substantive parts to address the research question and contextualise the policy dilemmas for regulators. The next section outlines the current regulatory framework and theorises the dilemmas facing regulators operating independently or in concert with others. The third section outlines the research design and research methodology. The fourth section then reports on the empirical component of the study that identified 1,524 potentially poorly compliant hand car washes across 71 English local authority district areas in the East Midlands, Greater Manchester, North East, South Yorkshire, West Midlands, as well as specific cities such as Bradford, Bristol and Leeds. These localities represent 25.1% of all local authority districts or 31.5% of all neighbourhoods (at the Lower Super Output Area²) across England. The fifth section then discusses the research question in terms of the five hypotheses. A summary and concluding remarks are then presented in the final section.

2. The current regulatory framework: compliance, deterrence and dilemmas?

Here we preface our research question by outlining the current institutional structure for labour market enforcement, and then discuss the different enforcement approaches that the main enforcement bodies use in order to theorise each of these in the light of some economic scrutiny.

The dilemmas faced by regulators centre on the different methodologies they use that contrast compliance, deterrence and intelligence-based approaches to labour market enforcement. These differences seemingly build in the presence and scope for strategic regulatory tolerance for some types of violation in the light of resource constraints. In principle, there could be even more scope for regulatory dealing with a SEB (or a smaller number of enforcement agencies) as they would typically have oversight of more potential violation types or domains.

2.1 The institutional structure of labour market enforcement

The current institutional structure is headed by Britain’s Director of Labour Market Enforcement. Created by the Immigration Act of 2016, the DLME has a remit to produce an annual enforcement strategy that sets priorities for the Gangmasters and Labour Abuse Authority (GLAA), Her Majesty’s Revenue and Customs (HMRC) and the Employment Agency Standards Inspectorate (EAS). The Director heads an arms-length public body co-sponsored by the Home Office and the Department for Business, Energy and Industrial Strategy (BEIS). The DLME provides oversight and strategic direction to the work of three labour market enforcement bodies: GLAA, HMRC (Minimum Wage compliance and enforcement) and EAS.

The GLAA is the most active of these three bodies in the specific area of protecting vulnerable and exploited workers (DLME, 2018) and is a Non-Departmental Public Body (NDPB) that operates using intelligence that its inspectors receive from inspections of businesses, from the public, from industry and from other government departments/public service agencies. The GLAA investigates reports on all aspects of labour exploitation and illegal and unlawful activity such as human trafficking, forced labour and unlicensed labour provision.

In addition to these, the GLAA investigates other offences under the National Minimum Wage and Employment Agencies Acts, often working with the police, the National Crime Agency and other government law enforcement agencies to target, dismantle and disrupt serious and organised crime across the UK. The GLAA operates a licensing scheme to regulate businesses that provide labour to the fresh produce supply chain (agriculture, horticulture, shellfish gathering and associated processing and packaging sectors) to ensure they meet the employment standards required by law. Employment agencies, labour providers (sometimes termed intermediaries or network actors) or gangmasters that provide workers to these sectors require a GLAA licence. Labour users in these sectors can also face criminal prosecution if they use unlicensed gangmasters under The Gangmasters (Licensing) Act 2004.

The Employment Agency Standards Inspectorate works alongside the GLAA and HMRC with employment agencies and businesses to ensure that they comply with the law and, where necessary, investigates complaints received from agency workers. The EAS can take enforcement action that includes prosecutions and unlimited fines, recovering unpaid wages or money owed to temporary workers and unlawful fees charged to workers. More strategically, the EAS works with other organisations, for example, regional organised crime units (ROCUs), to identify agency workers at risk of exploitation.

Multiple regulators have emerged in a historically piecemeal fashion in response to specific concerns, each with different specialist technical and human capital skills that address different types of regulatory violations or issues. For example, for some (Meadowcroft and Blundell, 2004), the contemporary regulation of informal labour began in 2004 with the creation of the Gangmaster Licensing Authority (GLA) that was launched because of market failure in the UK when at least 21 undocumented Chinese ‘cocklers’ drowned in Morecombe Bay. For Strauss (2013), the creation of the GLA represented the state’s historical return to gangmaster licensing that had existed from the mid-nineteenth century until it was eventually repealed by the Employment Agencies Act (1973). Strauss, therefore, argues that ‘the [2004] Gangmasters and Licensing Act represents less a de- or re-regulatory model than a socially and politically contingent re-regulatory choice’ (2013, p. 164—emphasis in original). In 2014, the GLA moved to Home Office jurisdiction and was renamed the GLAA, where its remit now covers the whole economy.

It is also the case that the GLAA, HMRC and EAS, whilst overseen by the DLME, operate with significant degrees of relative autonomy. For example, DLME policy endorsed by the government recommends a locally enforced national licensing system for hand car washes and nail bars (DLME, 2018). However, initially, the GLAA sponsored a licensing scheme for the 8% of hand car washes that operate on supermarket car parks, whereas licensing schemes for nail bars are confined to a few boroughs in London and Manchester city centre (Responsible Car Wash Scheme, 2019 and, for example, Barking and Dagenham, 2021). Therefore, it is unsurprising that labour market enforcement suffers from narrow enforcement between multiple unrecognised brands.

2.2 Fragmentation, deterrence and non-compliance

Theoretically and empirically relations between regulators and enforcement agencies are currently the antithesis of a coordinated SEB with poor unified support for businesses and uncoordinated, competitive and counterproductive enforcement action. Intelligence is dispersed into regulatory silos creating segmented work relations between enforcement partners. Whilst separate enforcement bodies operate with some effectiveness in amalgamation the labour market enforcement system is very fragmented for three reasons.

Firstly, as Figure 1 illustrates, several institutionally separate enforcement bodies have some jurisdiction over specific areas of labour market enforcement, for example, holiday pay (Advisory, Conciliation and Arbitration Service [ACAS], EAS, GLAA, Health and Safety Executive [HSE] and employment tribunals). Secondly, the different enforcement bodies have different theoretical approaches to labour market enforcement that contrast compliance, deterrence and intelligence-based approaches. Thirdly, a hierarchy of influence and ability to take direct action (before the formulation and adjudication of legal due process) exists. This is headed by HMRC that possesses significant operational powers to close and inspect businesses, including freezing business assets. HMRC is followed by GLAA in terms of operational influence, then the HSE. The police have significant operational powers too but frequently defer to specialist enforcement agencies such as the GLAA that is populated by many ex-police officers.

The wider literature on non-compliance and labour market regulation examines different approaches to the structure of labour market enforcement. Therein, Howe et al. (2013, p. 88) provide a comparative study of the Fair Work Ombudsman (FWO) in Australia to contrast approaches with wide discretion and those which divide responsibility for different labour issues among separate inspectorates. Similarly, Hardy’s (2021, pp. 158–59) study of deterrence-based theories conclude that improving labour market compliance requires non-compliant employers to perceive that the likelihood of being caught is high. In Australia and in Britain, greater compliance is more likely to be achieved by an integrated FWO or SEB operating on a project management basis rather than silo approaches. However, as Mustchin and Martínez Lucio (2022, p. 7) argue for Britain at least the context of austerity, de-regulation and ‘light-touch’ regulation result in little engagement between atomized enforcement agencies. These contributions build on Piore and Schrank (2008) who suggest that multiple inspections by different inspectors (as in the British case) who often have competing priorities may increase non-compliance.

For theorisation, compliance-based approaches to labour market enforcement occur where employer violation of labour market regulations are rationalised as flowing from ignorance and incompetence. Therefore, to improve enforcement, regulators prioritise information and compliance assistance to enable employers to become law-abiding actors; for example, the EAS frequently takes an educative approach to non-compliance other than in cases of severe or pre-meditated non-compliance. In contrast to compliance-based approaches, the GLAA, HMRC and the Employment Tribunal service (that enacts enforcement on behalf of individual workers), prioritise deterrence. Theoretically, deterrence assumes that labour market violation is intentional where employers strategise the avoidance of labour market regulations as they determine the use of informalised business and employment models specifically based on labour market exploitation (Clark and Colling, 2018). Therefore, enforcement approaches aim to alter employer behaviour by raising the risk of inspection and prosecution where the penalties of strategic avoidance of regulation are severe.

The HMRC penalty regime applies to all non-compliance irrespective of intent and scale; for example, at hand car washes irrespective of whether an entrepreneur owns one or ten businesses. The GLAA combines deterrence and compliance approaches utilising the former in cases of modern slavery and strategised exploitation beyond its areas of licensing, where in contrast, compliance orders are imposed on unlicensed labour providers or gangmasters in the first instance. More controversially, intelligence-based approaches to enforcement focus on the intelligence gathering and ‘bigger pictures’ that at times appear permissive in the acceptance of routine labour market non-compliance. Therein, enforcement agencies and ROCUs often focus on modern slavery and labour trafficking. These are beyond immediate issues of routine labour market coercion and exploitation such as underpayment of the minimum wage, denial of holiday pay and/or rest time that are much more prevalent than modern slavery and labour trafficking (Davies, 2019; Kenway, 2021). HMRC, ROCUs and sometimes the GLAA make strategic choices between tolerating routine exploitation and monitoring businesses of interest if there is hard intelligence and informed suspicion that these businesses are fronts for organised crime or counterfeit goods and services.

We argue that a further limitation of the regulatory system is the dominance of ‘law and order’ inspired compliance approaches, often at the behest of politicians, such as the ‘hostile environment’ policy, instituted when Theresa May was Home Secretary in 2012. This describes a set of administrative and legislative measures designed to make staying in the United Kingdom as difficult as possible for those without the right of leave to remain. The policy aims to persuade those in this category to voluntarily leave the country but does so by criminalising them whilst they remain in Britain (Kenway, 2021, pp. 48–50)—an approach that remains in place and one that confuses unfree labour with a preference for the UK’s highly flexible labour markets rather than the application of labour law to reduce the presence of non-compliant informal practice.

Moreover, the dominance of this approach focuses the problem on workers and ignores the employer motives for non-compliance. Motives can be material and economic utility maximising where normatively and socially, entrepreneurs can have no commitment to respect the rights of others or to obey the law. Nielsen and Parker (2012, p. 431) who develop this argument, suggest that regulatory dissenters do not want the approval of regulators. Rather, as our empirical material outlines non-compliant dissenters are influenced not by formal regulatory practice but by networks and informal self-generated institutional practice. This preference reinforces rather than curtails the tendency to informal practice and in part undermines the work of the DLME (see also Fudge, 2018 and Brown and Wright, 2018). For example, primary research for this project found failed asylum seekers who were prepared to wash cars for £2–5 per day or in return for food and shelter.³

2.3 The regulatory environment

The complex landscape of enforcement and regulation of non-compliant enterprises highlights many structural and operational concerns. These are rendered more acute where the ambient propensity to comply without active enforcement amongst a specific population of enterprises is low. This is a feature of the findings in section four and other related studies where evidence on the scale of enterprises that use informalised unlawful business and employment practice suggest that non-compliance is a stubborn problem. Case study research reports informalisation as an embedded presence in small firms across a diverse range of sectors. For example, in the British context at least, the extant literature demonstrates that the fragmented regulatory system leads to permissive visibility of informal business operation in the UK in several markets, but most notably, car washes, small unit garment manufacturing (in and beyond workplaces) and in nail bars.

For these markets, detailed survey data suggests that coercion associated with labour market non-compliance and informalisation is a generalised presence (Hammer and Plugor, 2016; Silverstone and Brickell, 2017; Clark and Colling, 2018). Many small unit garment manufacturers re-produce capitalist production relations where non-compliance is endemic, with much work ‘put-out’ beyond centralised workplaces (Hammer and Plugor, 2019). Moreover, the informalisation of employment is a continuum where some otherwise lawful businesses condone the use of informalised employment practice. In sectors such as construction, parcel delivery, restaurants and independent supermarkets, employers degrade employment and worker status to that of self-employment to casualise and informalise work (Moore and Newsome, 2018; Ram et al., 2020; Mustchin and Martínez Lucio, 2020).

Comparatively, evidence on street food vendors from India suggests that informal business practice is embedded and interacts with formalised business and employment practice but non-competitively (Narayanan and Véron, 2018). Similarly, evidence on street traders in Barcelona and Italian tomato harvesters in Foggio suggests whilst they are exploited and excluded from formally regulated employment, both traders and harvesters possess agency where this work is acceptable to them and better than other alternatives (Alford et al., 2019; Howard and Forin, 2019). The presence of agency among workers is itself a form of permissiveness where the ‘lived experience’ of some workers may belie the exploitation, potential coercion and modern slavery they endure (see Kenway, 2021, pp. 163–65). However, ‘put-out’ homework by the garment manufacturing sector in the UK, Barcelona’s street traders and Foggia’s tomato harvesters may unbeknown to them represent the bottom rung on a global value chain that runs all the way to national brand businesses.

Enforcement and regulation pans across employment and labour market matters, health and safety for workers and customers, potential disturbance to nearby residents and other businesses, environmental protection, corporate governance and tax affairs. Entrepreneurs have their own objectives and resource constraints and, in the light of these, can choose to respond to different levels of enforcement in the short, medium or long term by one or more of:

1. Active compliance

2. Attempts at camouflaging violations by mimicking legitimate practice.

3. Obfuscation and delay in responding to violation enforcement and subsequent notices/proceedings.

4. Reliance on ‘hit and run’ market entry and exit to avoid enforcement and violation proceedings. If this occurred, one would expect to observe relatively high entry/exit flux.

5. Regulatory arbitrage (from a perceived high enforcement area to a perceived low enforcement area, that is, spatial relocation).

6. Establishing compliance priorities and adjusting efforts in the light of perceived differences in enforcement levels among different regulators.

In turn, the multiple regulators operating in this setting have their own objectives and performance targets and may choose, in light of 1–6 above, to either operate independently or engage in various levels of cooperation (including tip-offs to other regulators, enforcement visit date sequences and scheduling) or active coordination in enforcement.

For those regulators operating either independently or in concert with others, there may be scope for some strategic regulatory tolerance depending on their own resource constraints. Following Harrington (1988) and focusing on compliance, a regulator may categorise firms utilising information from recent inspection outcomes. Group 1 would contain firms found to be compliant at the last inspection and group 2 would be those found to be non-compliant. The regulator may choose to levy no penalty upon a group 1 firm caught violating but, focusing on deterrence, apply the maximum penalty upon a group 2 firm caught violating. In this way, a ‘representative firm’ can be induced to comply some of the time (especially a group 2 firm). The suggestion is that higher compliance rates can be obtained over those that might be obtained from a policy of full deterrence pursuit with limited penalties (the ‘Harrington Paradox’). Heyes and Rickman (1999) provide an alternative rationalisation premised on the exercise of some ‘regulatory dealing’. By this route, a regulator with responsibility for enforcing over more than one violation category or domain allows a firm to comply in one domain in exchange for non-compliance in another and may still, in overall terms, achieve higher compliance rates in an enforcement regime with restricted penalties.

What renders the Harrington Paradox and the Heyes and Rickman (1999) characterisation potentially more difficult to discern, and complex to investigate in this particular context (as opposed to Harrington’s (1988) and Heyes and Rickman’s (1999) original single regulator study focus) is the multiple regulator setting of hand car washes (HCWs). Each enforcement body has different but, in some cases, overlapping spheres of competence and typically follow different diary schedules of inspection and enforcement with little or no explicit coordination. In principle, with some liaison or degree of cooperation in a multiple regulator setting, the possibility exists for the Harrington-type categorisation to take place sharing and utilising compliance status history from other regulators or enforcement bodies (i.e. harnessing the potential benefits from institutional complementarities).

Similarly, in principle, regulatory dealing could also operate across the domains of different regulators and enforcement bodies and could support arguments for retaining some pattern of multiple regulator arrangements. Effectively, they would be offering a mutually reinforcing mosaic of regulatory oversight across safety, environmental and work practice whilst sustaining specialist human capital in each area of competence. That said, even if such liaison and cooperation was not readily identified, there might be opportunities to be revealed that can incentivise or nudge regulators towards such practice.

Drawing on this body of work, five hypotheses are outlined which guided the mode of enquiry and the interpretation of findings. Hypotheses 1 to 3 relate to HCW owner responses and Hypotheses 4 and 5 pertain to the actions of regulators.

 

The implication here is that a simple and very light-touch regulatory regime is unlikely to successfully meet objectives related to achieving specific violation-level targets.

 

The implication here is that HCW owners will vary their compliance outlays in relation to their perception of the expected remediation and penalty costs from (further) inspection visits where no response action by the HCW has been taken.

 

Consideration of this hypothesis embraces the possibility of both hit-and-run entry and exit (sometimes termed ‘phoenixing’) and regulatory races as potential responses to enforcement action by regulators.

  

3. Research design and methodology

The methodology builds through a three-part mixed methods research design focusing on England and Wales. Firstly, interviews with the DLME, the DLME’s head of secretariat, head of policy and the DLME economist, the CEO, the head of strategy and the head of prevention at the GLAA enabled the research team to understand the sector and the focus of regulators each of whom we interviewed at successive stages as the project developed. Further interviews with the head of the Car Wash Advisory Service and the chief executive of the Downstream Fuel Association that represent petrol station retailers provided the views of key stakeholders in the established car wash industry.

In the operations field, it was less easy to conduct in-person interviews, so these were sometimes held as telephone interviews with strategic analysts and police officers attached to the East Midlands and South East ROCUs, the Greater Manchester organised crime unit—Programme Challenger and the GLAA North-East operations team. The regulator, stakeholder and field interviews, as of May 2021 total 110.

The second stage of the research design was a two-city pilot study of 45 hand car wash workplaces. The source material centred on interviews with 42 car wash workers, 3 entrepreneurs, trade union officials from the General Municipal and Boilermakers (GMB) and Union of Shop, Distributive and Allied Trades (USDAW) and telephone interviews with estates managers at branded supermarkets that had trolley washes on supermarket car parks. To gather information on the ‘lived experience’ of workers, interviews with a further 25 hand car wash workers were facilitated by a community self-help group for migrant workers in the East Midlands—the Romanian Society for the East Midlands (ROSOC).

Interviews with stakeholders were recorded, transcribed and coded, but recording interviews with car wash workers was not possible and to secure access beyond individual workplaces, and via ROSOC, this was a condition of the interviews and in accordance with University research ethics approval. Similarly, some research findings from interviews with regulators are reported in a general manner, that is, they are not attributed to specific role holders, for example, compliance and enforcement units. This is the case because these interviews provided access to live field operations and covered live operational intelligence and policy issues.

The third stage in research design represents a major innovation in methods and moves beyond a case study approach towards the use of digital sources to map the location and location types of hand car washes. To identify the presence of hand car wash businesses within a specific location, application of Facebook and Google Maps identified those hand car washes advertising within particular local authority areas.

To supplement this data, virtual fieldwork on a street-by-street basis across all neighbourhoods within each local authority area used Google Street View imagery to identify those hand car washes that do not appear within the Ordnance Survey’s Points of Interest dataset and/or operate without advertising or a social media presence. Finally, outside pandemic conditions, fieldwork drives around these neighbourhoods recorded additional car washes that lacked a Google Street View presence. This innovation created a more accurate picture of the scale and nature of the hand car wash sector within each local authority area and enabled the development of a hand car wash risk classification index that classifies workplaces in terms of physical (e.g. state of repair, signage), worker (e.g. clothing and protection, equipment and evidence of on-site accommodation) and environmental issues (e.g. drainage, storage of chemicals, water storage, site surface). For each dimension of the index, hand car washes were graded on a scale between 1 (non-problematic) and 10 (highly problematic).

The scaling of the index was determined by an inductive methodology where researchers were each given 10 examples of hand car washes and asked to individually assign a score using a risk classification scoring sheet. The scoring sheet broke down the scoring into the three sub-categories but did not suggest any scoring for certain conditions relating to specific aspects of the site such as the drainage provision or use of personal protection equipment (PPE). These elements of interest were identified through preliminary research into how hand car washes were non-compliant to operating requirements in the UK. Once completed, the individual reviewer scores were then compared and discussed in order to determine scoring motivations and the weighting of visual elements of the sites. This reflection produced an agreed reference for which elements warranted inclusion and an approach to calibration that produced a uniform research grading to allow for comparable scoring across subsequent hand car washes which were risk evaluated by individual researchers.

The calibration process was often refreshed with discussions raised over hard-to-score sites. This process resulted in 6 and 7 being the most common scores, rather than 5 which would be the middle point in a more constructed scaling. Some elements of a hand car wash would clearly push the score one way, with the presence of a caravan on the site generating scores greater than 8 and the use of full uniform with a clearly marked staff room doing the opposite.

Multiple regulators exist because of the potential specific forms of non-compliance (deliberate wage underpayments, coercive and exploitative terms and conditions) and components of non-compliant informal activity (a general presence of health and safety violations). Therefore, to improve the classification of HCWs, it was necessary to chart and map the extent to which the distribution of non-compliance and informality across HCWs is homogenous or heterogeneous. That is, do most problematic HCWs exhibit non-compliance and informal practice in relation to specific environmental and worker welfare issues—or do the most problematic and non-compliant workplaces appear disproportionally across a general arena such as health and safety violations? Use of this design enables an assessment of whether there is an argument that the multiple enforcement agency approach (created on the expectation of multiple areas of non-compliance) was necessary on the basis of the empirical evidence for a priority sector of concern to regulators.

4. Findings

The study found lawful businesses, unlawful businesses, ‘pop-ups’ or itinerant trolley wash businesses in supermarket car parks where wage underpayment in terms of the national minimum wage and therefore wage theft from workers was evident at every business in terms of the national minimum wage of between 15% and 40%. Application of the three elements in the research design identified 1,524 potentially weakly compliant hand car washes across 71 English local authority district areas. In summary, the figures show four broad findings that summarise the condition of hand car washes:

1. The longer a hand car wash is in operation, the more likely the site’s physical and environmental condition is considered worse.

2. A large majority of the sites do not exhibit significant improvement or degradation.

3. Of the sites that did change in the East Midlands, slightly more improved than degraded.

4. City-based local authorities have the most hand car washes, but they are more likely to exhibit degradation than improvement.

Figure 2 shows the average risk classification scores of hand car wash sites by observed years of operation. The overall average score is based on the three risk subcategory scores of Physical Place, Worker and Environmental. Figure 2 demonstrates the first broad finding that there is a clear trend of increased risk scores over time for both the place and environmental categories rising from 6.56 to 6.93 and 6.74 to 7.02, respectively. The worker risk score peaks at the year 6–7 banding at 6.98 after rising with time and then falls again with the oldest year banding of 10–11 averaging 6.86. The overall average score has its highest scores in the 6–7 and 10–11 year banding, with both scoring 7. The average score may have reflected a decline over time if not for the less orderly worker risk score. Figure 2 also supports the second broad finding of overall stability. This is the case because the second broad finding requires significant change to be recorded; for example, the deterioration of sites over time can be gradual enough to not trigger a move to the significant change category.

Figure 3 establishes that the scores across different regions are relatively consistent. Regions with fewer hand car washes tend to have more varying averages than the four regions with the most. Figure 3 provides some narrative for hypothesis 3 in that if the quality of sites across England and Wales are relatively similar where the standard of regulatory pressure and participation are also similar, it is unlikely that a hand car wash operator will move to a different region to escape harsher regulation.

Figure 4 also illustrates the frequency that a site experiences a significant change within the East Midlands region. Of the 310 sites observed, 82% (254) were in the no significant change classification, 4.5% (14) were considered improved and 3% (10) were degraded. The remaining 10% (32) could only be observed at one point, so change over time observations were not possible. These results align with broad findings 2 and 3 in that most of the sites did not experience a significant change in condition, and that of those that did change, 14 improved and 10 degraded. Figure 4 also shows the average risk classification score for each county for comparison to the long-term state of the sites.

Figure 5 breaks the East Midlands region down into the local authority areas and provides additional evidence for the broad findings 2 and 3. In addition, Figure 4 demonstrates that city-based local authorities exhibit the greatest likelihood of downgraded hand car washes rather than ones that improved. Across these authority areas, 93% (68) maintained, 4% (3) declined and 3% (2) improved their site condition.

5. Discussion

 

In relation to Hypothesis 1, if hand car washes ignore or fail to respond to enforcement directives fully, we would expect a greater number of sites to move towards the ‘degraded’ category over time. It is evident that many sites with high scores in degradation and poor practices have been open for several years. Thus, if there was a wide-scale response to compliance and deterrence drives, then it might plausibly have been expected that more sites should have made noticeable changes to improve. If sites are being issued with enforcement directives to improve, it is unlikely that they are observed and therefore not likely for any improvements to have occurred. It might be assumed that typical business behaviour would be to respond to enforcement directives to maintain viable commercial activity. As hand car washes often change legal identity, it seems that the cost to of becoming compliant is often greater than the punishment for non-compliance.

The punishment often takes the form of fines and prohibitive orders against the current owners which can be side-stepped by creating a new company with a new owner, transferring ownership of the site or the tools of the trade and finally dissolving the original fined company. Beyond the data set, case study evidence suggests that non-compliance is universal among HCWs. HMRC compliance enforcers stated this clearly in interviews. Therefore, this insight combined with the scores for non-compliance and deterioration increasing over time is supportive of Hypothesis 1.

 

In relation to Hypothesis 2, what is now termed PPE provision is a relatively cheap and visible form of compliance. A company that wishes to move towards legitimacy would be expected to be able to provide the relatively cheap uniform of high-visibility clothing and protective gloves as a first, immediate step. Other site requirements such as re-surfacing for correct drainage or construction of non-temporary buildings require planning permission, time and investment and can present a much greater challenge to becoming compliant. However, there is a clear falling trend in PPE provision over time that is not reflected in the apparently improving worker scores. Moreover, before the onset of pandemic conditions in March 2020, fieldwork observations for the workplace study found that the use of PPE varied widely across different workplace types; for example, trolley wash workers on supermarket car parks rarely wore protective gloves or waterproof footwear or trousers. The ‘Wash while you shop’ system meant workers would move around the supermarket carpark with a trolley carrying the cleaning apparatus and larger items or those that required access to utilities were often absent, such as pressure washers. The bulk of the storage was often a small cargo container on the edge of the site where bulk chemicals and water were stored.

At car washes on former or open petrol stations, the use of PPE was also variable, with some operatives wearing waterproof boots and trousers and others not so; similarly, very few operatives wore protective gloves. Interviews with car wash workers suggested that the use of PPE was a personal choice that many confined to footwear. It is also unclear from interviews with regulators at the GLAA if there is a specific focus on PPE. The absence of appropriate PPE may be a component element within workplace coercion and exploitation but does also demonstrate some worker agency. Similarly, enforcement teams at the GLAA frequently saw the application or absence of PPE as an issue for the Health and Safety Executive. However, enforcement officers did suggest that the absence of PPE at a HCW workplace represented a worker welfare signal that was their main concern. The data set and qualitative material from workplace studies support Hypothesis 2, where entrepreneurs and workers feel there is little chance of inspection. It is also the case that non-compliant workplaces use mimicry of compliant businesses as a central factor in compliance avoidance and inspection avoidance. Mimicry takes the form of professional-looking price boards, accurate stencil copies of branded cleaning products and the development of on-site customer waiting areas.

 

Hypothesis 3 relates to continued business operation at the same location. Many sites have clearly remained open for several years despite high scores for degradation and deterioration. Figure 2 indicates that most regions have relatively similar classification scores; therefore, closure and relocation elsewhere to find potential business locations subject to fewer enforcement measures is unlikely to be a necessary action followed by many HCW entrepreneurs. There is a greater concentration of HCWs that exhibit deterioration in urban centres.

This suggests that HCW entrepreneurs can remain open because the likelihood of inspection and enforcement measures is low. This is the case even as a workplace exhibits deterioration where conditions across all measures worsen. The falling standard of sites overtime also suggests they are not facing demands to improve; for example, virtual fieldwork via Google Street View timeline functions and in-person fieldwork found little evidence of investment to improve the infrastructure at many HCWs.

There are, however, instances of ‘static phoenixing’ whereby a workplace does close—potentially because of enforcement action only to re-open a short time later under a new name but with many of the same workforce and sometimes the same owner. Therefore, there may be occurrences of enforcement action against HCWs, but it is unclear from which agencies.

Moreover, this enforcement action is insufficiently disruptive enough to stimulate permanent closure and relocation elsewhere. So, it could also be the case that enforcement or short-term closure may follow on from specific individual incidents that are serious enough to stimulate regulators to act, but with no lasting effect. For example, beyond our data set but within the qualitative fieldwork, the following finding is instructive on this issue:

…at a particular hand car wash in (name of city)—at the XXXXXX (name of business) enforcement officers noted an entrepreneur using ‘illegals’ (undocumented workers). We proceeded to enforcement action but not on the business; we let this remain open, and we made no checks on other workers who all claimed to be European Union nationals with the right to work in the UK, some of whom produced national insurance cards. Instead, we arrested the two and took them to the central police station where they remained overnight; we released them the next day and told them not to work unlawfully. In theory, they should have been detained at an immigration enforcement centre. We chose not to take this course of action for a mixture of resource reasons; several centres were full, necessitating them remaining in detention cells in the police station or securing appropriate detention facilities elsewhere and transporting them to the facility. (Anonymised, merged, and redacted treatment of an interview with a former Chief Constable and an Enforcement Officer)

Where sites do close permanently, they do so due to a better opportunity for the landowner, for example, property development. Accordingly, support for Hypothesis 3 is relatively weak and physical relocation as a response to enforcement may be substituted by temporary closure and re-opening at the same location.

 

For hypothesis 4, the discussion points flow from interviews conducted at an in-person stakeholder event hosted by the Responsible Car Wash Scheme that all regulatory bodies attended. It soon became clear that the multiple members of the regulatory pool operate independently without cooperation.⁴ Moreover, in the case of the Health and Safety Executive (HSE) team who attended the event, it was made clear that enforcement decisions in lower priority areas are influenced by the extent to which the recovery of costs were possible, particularly where overall the HSE saw HCWs as a low priority for enforcement operations. This view developed because the HSE policy saw health and safety issues as important at HCWs but less so than issues of worker exploitation in terms of pay and conditions—issues that were not their responsibility.

Moreover, in extending hypothesis 4 further, it became noticeable that problems of independent action and lack of coordination were also occurring internally within regulatory agencies themselves because of bureaucratic dysfunctions, particularly the ‘silo mentality’ (Diamond and Allcorn, 2009). In a similar vein to Blau’s (1963, p. 170) classic study of a US federal business law enforcement agency, the compliance officials were caught in the conflicting role and function ambiguity between either acting as ‘policeman’ or as public servants who sought a conciliatory role of educating businesses so they can obey the law. Our interviews with compliance officers from HMRC were also informative as they revealed their own internal frustrations. These centred on what they saw as the counterproductive ‘educative processes’ that attempted deterrence action brought to HCWs. Apparently, every time HMRC sought to enforce regulations at a HCW, these educational deterrence actions merely helped the entrepreneurs to find new ways to either hide funds or cheat the system, for example, via manipulation of self-employed status for workers.

Another issue was the difficult, time-consuming and costly nature of successfully challenging non-compliant entrepreneur claims. Therefore, if a car wash entrepreneur was known to be, or suspected to be, involved in organised crime or trade in counterfeited goods such as alcohol, cigarettes and running shoes, HMRC compliance sometimes chose to accept their claims but continued to monitor specific workplaces. Here it became clear that gathering and maintaining intelligence sources is often more important than winning small victories over VAT payments, where for this sector, HMRC sometimes saw national minimum wage compliance as an issue for the GLAA. The reverse view was sometimes cited too by the GLAA, hence the term ‘permissive visibility’ appears to generate and/or align with some regulatory dealing.

Another example of regulatory dealing became evident and followed on from a specific piece of information that was made available to one ROCU. Data on the ten worst offending HCWs in the region where the organised crime unit operated was made available to the unit. They reported that they were aware of three of the workplaces, but whilst they appeared to be non-compliant, they chose to take no action against these workplaces. Three other workplaces were no longer operative as non-compliant businesses, but four workplaces that had high scores for non-compliance were unknown to the unit, and they began to engage in intelligence monitoring on each workplace. That is, the ROCU was interested in the intelligence on four workplaces they did not know about but were unconcerned about three non-compliant workplaces they did know about.

In both cases, a strategic choice decision was made where being influenced by the generation of ‘intelligence’ appeared as a crucial determining factor. The evidence thus supports Hypothesis 4—there is little active cooperation between regulators. This seems to be the case because regulators are looking at different issues and have different competencies and priorities. The outcomes may be an enforcement action for some bodies but contributions to intelligence gathering for others to achieve higher impact outcomes in the future. The key actor in the institutional hierarchy is HMRC; they do have the power to move regulatory norms and sometimes state that they are keen to do so. However, the compliance unit within HMRC is, like all regulators, subject to Government budgetary oversight and austerity measures and is also relatively slow-moving.

Moreover, in the absence of research data such as that reported herein, HMRC intelligence-gathering generally precedes enforcement action. Further, enforcement action is confined to intelligence coming into the compliance unit from elsewhere, rather than them undertaking geographically concentrated enforcement activity in a city, a neighbourhood or on specific roads. That is, the multiple regulatory bodies do not actively cooperate in a systematic manner and have different objectives such that regulatory laxity or tolerance of violations remain prevailing market features of this sector, in a way that broadly aligns with the Harrington (1988) Paradox and the practice of regulatory dealing (Heyes and Rickman, 1999).

In addition, qualitative evidence indicates that some regulatory tolerance is strategic in its specific intent to support intelligence gathering on organised criminal groups to service higher impact enforcement actions in the future.

By association, in terms of the wider literature on non-compliance and different approaches to labour market enforcement, the research data presented herein aligns with the argument of Howe et al. (2013) and Hardy (2021). That is, the data suggest that regulatory approaches which divide responsibility for different labour and workplace issues among separate inspectorates may provide employers who actively choose non-compliance as a business model with a low likelihood of being inspected or caught. Therefore, as Hardy (2021) argues in this context, deterrence tends to be lower than that achieved by more integrated inspectorates.

 

In terms of Hypothesis 5, the collected data suggest that it is rare for a major change in condition to occur once a site is established. Therefore, major enforcement interventions by regulators that aim to make a site more compliant are also rare, stimulating a non-compliance-non-enforcement dialectic. If sites propagated the Harrington Paradox, lower risk classification scores should prevail when HCW businesses enter the market. Similarly, the condition of these workplaces should be stable, that is, display some maintenance and upkeep as entrepreneurs seek to comply with regulations. Sites may relish and condone the absence of regulatory enforcement—the non-enforcement side of the paradox. Yet site conditions also suggest that HCW businesses do not effectively self-regulate—reinforcing the non-compliance side of the paradox. Regulatory dealing may occur as the average decline in scores shows that HCW entrepreneurs condone falling workplace standards but seek to mimic compliant businesses. Mimicry aims to fulfil the minimum expectations of regulators by improving peripheral aspects of the business workspace whilst allowing the condition of the site overall to decline, for example, in the corrosion and break-up of the tarmac on the deck of the workplace. The paradox seemingly prevails though the findings herein do seriously question its possible status as a key explanation for regulatory behaviour. It is found that regulatory dealing is manifest both formally (e.g. HMRC’s lesser concern with health and safety issues) and informally or unstated as individual regulators choose to ignore non-compliance for specific reasons, for example, because they cannot recover costs, for fear of making the problem worse, or to enable intelligence gathering.

6. Summary and concluding remarks

Non-compliant employment practice in the UK is increasingly pervasive across the informal economy, with some sectors sliding more appropriately into the category labels where endemic wage underpayment and elements of coercion associated with the spectrum of modern slavery appear embedded. Most observers would argue that this situation, termed permissive visibility, warrants greater regulatory intervention and enforcement. The current body of economic theory regarding such intervention is typically framed in terms of single regulator scenarios (Harrington, 1988; Heyes and Rickman, 1999), but regarding labour market enforcement in the informal economy, multiple regulators are relevant features with each focusing on different aspects of the employment experience and business setting. Therefore, the research question posed by this study is a legitimate one that asks whether the status quo of the multiple regulator setting is providing the best possible outcomes for all relevant stakeholders, but especially those subject to poor employment conditions and underpayment of wages. Empirical survey data and qualitative evidence were assembled for one exemplar sector where abuse of the national minimum wage is acknowledged to be rife but maintains geographically widespread and seemingly permissive visibility—hand car washes.

Poor working conditions and no significant improvement in site conditions seem to be persistent features of the site data collected. Many sites served with enforcement actions seem not to actively comply or do so perfunctorily, addressing some concerns but neglecting to attend to others. In the light of tight resource constraints, the likelihood of regular site visits and enforcement actions are seemingly deemed by HCW owners as persistently low, and so the need to close and relocate following enforcement actions seems generally unnecessary.

The empirical evidence suggests the theorisation that HCW entrepreneurs fail to respond to enforcement directions by active compliance is normal practice. Similarly, entrepreneur responses to enforcement by regulators align to perceptions of the expected cost levels of remedying the range of violation types supported, most of which are low, encouraging non-compliance and mimicry of lawful best practice. Conversely, where the expected cost of violation remedies and penalties are greater than the expected cost of business liquidation and potential re-establishment elsewhere, a HCW owner will actively consider termination and possible relocation of the enterprise. The empirical evidence fails to support this hypothesis. Rather the evidence suggests that action by a given regulator amongst the pool of multiple regulators of HCWs operates primarily independently with little or no cooperation but also has little or no effect on location and operation decisions by HCWs even where a business improves some aspects of compliance. Therefore, in a multiple regulator setting for the regulation of HCWs, enforcement patterns that broadly align to ‘Harrington Paradox’ type motivations or ‘regulatory dealing’ type motivations do exist across the sector.

Unfortunately, the fallout from such a mosaic of regulatory actors with different objectives and guiding timescales is a default business landscape where ongoing worker welfare is persistently inadequately monitored and regulated. Despite such workers doing jobs in full view of their town and city location residents, this is a hallmark feature of permissive visibility. Accordingly, if the underpayment of the national minimum wage and worker welfare are deemed important, then it is important to acknowledge the presented evidence highlighting persistent failings in the current arrangements. Accordingly, some movement to new policy experimentation seems warranted. This could be at various points along the spectrum ranging from systematic ongoing collaboration among multiple regulators to more radical policy re-design comprising SEB formation. Whilst the outcomes from some pattern of systematic collaborative working may yield greater returns in terms of more successful elimination of wage underpayment and worker welfare improvement, this would not be likely to be a revenue-neutral undertaking and would involve greater time and resource commitments among regulatory bodies.

A more radical movement to a SEB would arguably require new agent operatives trained or literate across a diverse and challenging body of knowledge in policing, fiscal, trading, environmental, health and safety domains. Similarly, it is essential that a SEB moves to a project management mode of operation so as to avoid the transfer of silos into a new body and overcome the limitations of isolated light touch regulation. This might prove a significant stumbling block to reasonably swift activation of a SEB in this arena. Yet meaningful experience from a substantial period of greater systematic collaborative working may well inform, help scope, and improve the outcomes from such a radical development in a shorter timescale.

We are grateful for support from the Arts and Humanities Research Council/Modern Slavery Policy and Evidence Centre [Grant no: AH/V009036/1] and also a grant from the Home Office Modern Slavery Fund

Footnotes

Bibliography