Worlds of Labour: Introducing the Standard-Setting, Privileging and Equalising Typology as a Measure of Legal Segmentation in Labour Law

Abstract

The Standard Employment Relationship (SER) in industrialised countries is associated with strong protection for employees who fulfil its criteria but tends to neglect those who do not. Although the theoretical concept of SER has had repercussions around the world, its global empirical incidence and the variation of regulatory patterns associated with it have not been scrutinised so far. Comparative quantitative research in labour law has mainly focused on the overall level of employment protection in the countries of the Northern Hemisphere. Against this background, we ask how legal segmentation in labour law, that is, the exclusion from and gradation in employment protection which seems to be connected with the SER, can be conceptualised and measured in a global perspective. Drawing on leximetrics, a method to measure and quantify norms, we make use of and extend existing data sets such as the Centre for Business Research Labour Regulation Index (CBR-LRI) and Employment Protection Legislation Index (EPLex) in order to grasp the nature of legal segmentation. We identify three main functions of individual employment law in the protection/segmentation context: the standard-setting (S), the privileging (P) and the equalising (E) functions. We develop the SPE-employment law model on the assumption that the three functions are mutually independent in normative terms. The SPE typology offers a genuinely new perspective for comparative labour regulation research, making it possible to see the differentiation of patterns of legal segmentation and their path dependencies in 115 countries. First findings on a global scale show that in 2013 no fundamental difference between the levels of regulation in the Global South and North can be found. Moreover, familiar patterns can be observed such as a tendency to stable and low protection levels in liberal welfare states, and a tendency to universalist types of regulation in former socialist countries.

1. Introduction

The spread of non-standard forms of employment and informal work has generated the current global debate on segmentation of labour markets. Increasing inequality, precarity and poverty provoke disputes on better labour regulation and universal protection for those dependent on labour-related income.¹ Beyond labour market processes, for the origin of segmentation can be found in legislation itself which neglects factual disparities and privileges ‘insiders’.² This has been a central issue in the critique of the Standard Employment Relationship (SER). The SER is considered to be built around a permanent, full-time remunerated employment relationship with a single employer and a single workplace. It has gained high scientific and public attention since the 1980s when deregulation and flexibilisation already had begun. Critical debates challenged the male breadwinner model and ethno-centrist constructions of labour markets closely related to the concept and practice of the SER.³

Although SER debates mostly focus on OECD countries, the nature of North-South relations and forms of labour in the Global South, which are related to aspects of the role and shape of SER in particular societies, have received growing attention. Here, informal employment and marginalisation of non-standard employees are especially detrimental phenomena. Therefore, research on informal employment, globalisation and labour standards in the context of the ILO has proliferated,⁴ and has explored the extent of exclusion from protective labour law resulting from the SER. Segmentation effects play a major role in the employment-related monitoring of the UN human rights treaty bodies.⁵ However, there is still a lack of data accessible for quantitative studies on rules concerning protection, segmentation and equalisation of workers. Systematic international comparison of segmentation by legal provisions has not been developed yet, and differentiated analysis of legal patterns inducing or contributing to variations of SER is still pending.

The use of the SER as reference point for analysis of global patterns of national employment legislation might cause surprise, considering the high impact of informal and non-standard work in the Global South. However, given that formal employment patterns all over the world follow certain paradigms of labour regulation, we can posit a variety of ‘SERs’ including in the countries of the Global South. Moreover, we may assume that European paradigms of employment law, although differing in kind from those in the Global South, have an influence beyond Europe as a result of the globalisation of trade and investment,⁶ colonial legacies⁷ and former European dominance in international organisations (like ILO).⁸ These factors mean that the SER has had a significant impact on labour law paradigms in the Global South—despite and even in contradiction with to the diverging forms of de facto labour relationships in those countries.

The SER results from the complex interplay of governmental and collective strategies in a wide set of arenas including family law, tax law, patterns of collective agreements, and childcare and migration policies. Nevertheless, we can identify a regulatory core to the SER, an element which remains present notwithstanding differentiation of the SER across countries. This regulatory core can be traced by a comparative analysis of individual employment law provisions and their segmenting effects. A quantitative analysis of individual employment law regulation therefore offers a first step towards a global mapping of the evolution of the SER.

Taking into account SER analyses and recent studies that investigate whether regulatory patterns go along with particular forms of empirical labour market segmentation,⁹ we ask how legal segmentation in employment law can be captured and measured at a global level. Our basic assumption is that national labour regulation can prevent, encourage and counteract labour market segmentation. We, therefore, identify three functions of national individual labour legislation which mark fundamentally different aspects of protection and segmentation. In functional terms, we call these the standard-setting (S), privileging (P) and equalising (E) function.

The differentiation between the three functions allows us to develop an approach to comparative employment law that offers a variety of genuinely new perspectives. We use the different functions to develop a typology of employment law systems—the SPE typology. This expands the scope of previous unidimensional approaches, allowing for a differentiation between eight ideal types of labour legislation. The typology captures not only different levels of employment protection, but also informs about the degree of active legal promotion of specific parts of the labour force. Furthermore, it considers the development of ‘reactive’ legal prohibitions of discrimination against others. Thus, not just exclusion from protection, but also gradation in levels of protection, that is, legal segmentation in labour law, become visible.

We expect that the SPE typology will enhance the debate on differentiation of SER models and the issue of universal versus particularistic protection. As we only refer to legal regulation, however, we cover governmental normative will rather than its implementation at the level of the workplace or wider labour market. The investigation of the relationship between legislation and de facto labour market segmentation, including the level of informal employment, requiring the usage of additional databases, is for future work.

In order to be able to cover almost all countries of the world, we developed a framework and collected data using leximetrics, a method for comparative research in labour law that quantifies norms.¹⁰ So far, while leximetrics has been utilised to measure employment protection,¹¹ its application to research on legal segmentation is new. We adapted indicators of the Cambridge Centre for Business Research Legal Regulation Index¹² that originally covered aspects of the standard-setting function and then added missing items for the privileging and equalising functions of employment law. We systematically gathered and analysed norms with respect to the three functions resorting to various sources including ILO’s Employment Protection Legislation Database¹³ and the International Encyclopaedia for Labour Law and Industrial Relations (IELL).¹⁴ Currently, our database Worlds of Labour (WoL) consists of 35 indicators (23 Centre for Business Research Labour Regulation Index [CBR-LRI] indicators, 12 indicators of own coding) and covers 115 countries from mostly at least 1970–2013.

In this article, we will report first findings for 2013 based on the WoL data set and the theoretical assumptions related to the SPE typology. First, we sketch the discourse of the SER and how to capture it; then, the derivation of the three functions and their combination in the form of the SPE typology will be introduced; this is followed by a methodological section elaborating on the concepts, data collection and index construction; finally, first findings will be discussed, rounded off with concluding remarks.

2. HISTORICAL DEVELOPMENT OF THE SER AND HOW TO CAPTURE SEGMENTATION IN EMPLOYMENT LAW

A. The Emergence and Changing Relevance of the SER

Although the idea to quantitatively analyse the protecting and segmenting aims of employment legislation and built national systems on a global level is new, the underlying questions stem from the long-running debate about the SER. In our approach, normative concepts are drawn from this debate as it emerged in industrialised countries,¹⁵ but also taking account of its spill-overs to the Global South—partly through ILO conventions and recommendations,¹⁶ partly through particular epistemic communities.¹⁷

The SER has become a central point of reference in the field of labour, first and foremost in Europe. It plays a major role in analysis, critique and reform discussions concerning segmentation of labour markets and inclusion policies.¹⁸ Despite its prominent place in scientific and public debates worldwide, its legal foundations, conceptual differentiation and historical emergence have not been thoroughly discussed from a comparative perspective, particularly with a view to types of employment regulation in countries of the Global South.

In the nineteenth century, at the beginning of industrialisation, labour-related legislation laid the legal fundamentals for the development of labour markets and labour as a commodity. Its emergence in the Global North completed the transition from feudalism to capitalism by entirely including also the subaltern classes into the market economy. In this first phase of labour regulation, bonded labour and slavery were abolished, freedom of contract and freedom of movement guaranteed also for workers, and basic limitations set including restrictions on child labour and on payment of wages in-kind.¹⁹ Freedom of contract was a purely formal, legal construct, ignoring substantive aspects both regarding the low bargaining power of workers and their subordination within the employment relationship.²⁰

In the first half of the twentieth century, employment legislation with a protective character was successively introduced—marking the beginning of the second phase of labour regulation. It limited both freedom of contract in relation to employment and the managerial prerogative via time restrictions for full-time labour and different forms and aspects of dismissal protection, complemented by further norms only applicable to employees. Social security policies and labour policies began to be developed separately, as employment protection was being implemented. Social security legislation nonetheless directly related to the employment relationship in many countries. At the same time, collective labour law emerged and unfolded in different systems of industrial relations. It guaranteed the effective functioning of trade unions, including the right to strike, autonomous progressive improvement of working conditions via collective agreements and worker representation.

In the decades following World War II, legislation culminated in a successive supplementation of status-building legal standards for employees with a multifold incentive-giving, protecting and segmenting character and impact.²¹ The introduction of variants of the SER in different countries thus marked the transition from an employment-at-will system (in the Anglo-Saxon world ‘Manchester capitalism’) to a modern welfare state-oriented system of employee status.

The SER started to be discussed in industrialised countries in the 1980s, when its peak as uncontested model had passed, and non-standard forms of employment (NSER: fixed-term, part-time and agency work) and their regulation gained relevance.²² Despite its variants, the SER in the Global North revolves around a continuous, full-time remunerated employment relationship under the managerial prerogative of a single employer at a single workplace. Its fundamental relevance derived not necessarily from its incidence in practice, but from being the central point of reference and model for labour law and social security law. This conjunction shaped the differentiated mode of social protection of a given welfare state in general, including gender relations and the meaning of citizenship and nationality.²³

The predominance of neoclassical economics and neoliberal reforms of labour markets in the 1980s, however, resulted in a phase of re-contractualisation or ‘recommodification’²⁴ of labour. It reflected the introduction of a new, differentiated guiding principle and thus a third phase of labour regulation.²⁵ In the context of globalisation, tertiarisation and digitalisation of labour, the flexibilisation of employment and the privatisation of protection were emphasised, including the weakening of dismissal protection and collective representation rights.²⁶ Additionally, NSERs increased,²⁷ and informal work gained relevance, amplified by increasing labour migration. Against the background of a weakened, but not abandoned SER (both in the Global North and in formal labour in the Global South), the growing differentiation of society furthered de-standardisation of labour market participation and life course transitions. Consequently, the notions of the employment relationship and self-employment and their differentiation came on the agenda anew. In this context, emerging anti-discrimination legislation played a major role.

The three-phase development model of the SER we have sketched stems not only from law, but also from de facto labour market developments. Moreover, we agree that the entire protection provided by the SER—and the de facto flexibilisation of labour markets—result not only from employment law, but also from the complex interaction of different legal areas and policies including tax and family law, typical national content of collective agreements—areas beyond the explorative character of the index introduced here.

B. Capturing Employment Legislation Development via Leximetrics

In the current phase of our research, we are not yet able to reconstruct the worldwide variety of SERs and their historical development. We can, however, historically and comparatively reconstruct the existence and the degree of segmentation inherent in the employment laws of the world. The collection of a data set capturing protection and segmentation in labour law does not have to start from scratch. Both in terms of methodology and data, we profit from existing databases. Leximetrics allows us in a structured and transparent way to quantify legal norms. It can be defined as coding and numerically measuring labour legislation for comparative purposes. The attempt to quantify legal norms of employment protection has resulted in several data sets and indicators, albeit with different limitations according to transparency and verifiability. The most familiar data sets are OECD’s Employment Protection Indicators (EPI), World Bank’s Rigidity of Employment Index (REI), ILO’s Employment Protection Legislation Index (EPLex) and Cambridge University’s CBR-LRI.

The EPI measures the regulation of employment since 1985 for 76 countries.²⁸ As its name indicates, the EPI aims to measure procedures and costs involved in hiring and firing individuals or groups of SER and NSER workers. Although economic literature today has abandoned the undifferentiated view of labour regulation as harmful to employment creation,²⁹ the OECD focuses on investment-related aspects in the form of time and easily calculable costs while neglecting the aspects of protection and segmentation. Although recent changes in 2019 have introduced methodological improvements,³⁰ scored values often refer to averages which hide segmentation of protection between different categories of workers.

The REI ranked 190 countries in World Bank’s Doing Business (DB) reports and thus is the most comprehensive data set. When the DB project was launched in 2004, existing regulations in developing countries were predominantly assessed as rigid³¹ and to be blamed for poor labour market performance and low productivity. The ‘mission’ behind the data set is to measure ‘excessive or insufficient labor market intervention and investigate the state of social protection’.³² As a result of severe criticism and at least semantic changes in the Washington institutions, the REI was eliminated from the ‘ease-of-doing business’ index that measures the overall business conditions of a country. However, renamed as labour market regulation index, it is still part of the annex of DB reports. To guarantee the comparability of data, the World Bank uses a standard worker model, making several specifications concerning age, job duration, type of business etc. Thus, it is not possible, using this index, to specify different protection levels between different categories of workers. As in the case of OECD, normative sources used for the determination of indicators’ values are not published and therefore cannot be tracked.

The EPLex (ILO, 2015) was created to replace the existing Digest on Termination of Employment Legislation. Starting in 2009, the EPLex Summary indicators provide information for 103 countries by 2019. Five areas are distinguished: (1) substantive requirements for dismissal; (2) maximum probationary period, including all possible renewals; (3) procedural requirements for dismissals; (4) severance and redundancy pay; (5) avenues for redress.³³ Furthermore, the database offers information on employment contract provisions concerning trial periods and fixed-term contracts. EPLex, linked to ILO’s law database NATLEX, extensively provides for the legal references used for coding, thus guaranteeing transparency for full verifiability.

The CBR-LRI covers labour laws of 117 countries over the period from 1970 to 2013.³⁴ It distinguishes five subindices measured by altogether 40 indicators: (1) the definition of the employment relationship and rules relating to NSERs; (2) working time restriction; (3) dismissal protection; (4) employee representation; and (5) industrial action. In contrast to the first two indices, the CBR-LRI does not purport to measure ‘costs’ or ‘rigidity’, but the level of protection. Its authors perceive it as a matter for inquiry whether laws and regulations impact labour market outcomes and economic performance, and thus impose costs on firms.³⁵ CBR-LRI sources the legislation and case law used to construct the index and thus also provides for transparency and verifiability of results.

Overall, the CBR-LRI and EPLex are two highly sophisticated leximetric databases that we were able to build on. The highly transparent CBR-LRI can be expanded to account for selective legal protection that leads to segmentation, and for legislation that counteracts discriminatory segmenting forces. Concerning selectivity of employment protection legislation, the EPLex database includes the ‘scope of regulation’ that enables an exploration of excluded categories of workers and enterprises (mainly by size).³⁶ Although relevant indicators can be found in these existing data sets, a more encompassing database that allows the systematic measurement of legal segmentation or universal legal protection of employees is still missing.

3. THEORETICAL BACKGROUND—THE FUNCTIONAL DIFFERENTIATION OF EMPLOYMENT LAW

Resulting from the focus on structural inequality between workers and employers, comparative research typically focuses on the protective function of individual labour law. Next to this, the function of market-making, referring to fundamental rules constituting labour markets, plays a significant role. This includes both prohibitions (child labour, slave labour, forced labour etc.) and certain formalities, and rights and duties marking the difference between contracts of service (employment) and contracts for services (self-employment). Functional differentiation of individual labour legislation as to its inclusive or exclusive (that is, segmenting) normative effects is usually neglected.³⁷ To fill this gap, we differentiate between the standard-setting (section The Standard-Setting Function), privileging (section The Privileging Function) and equalising functions (section The Equalising Function) of protective labour legislation.³⁸

A. The Standard-Setting Function

Legal protection via labour law in the proper sense refers to the counteracting of hazards and risks resulting from structural inequality, that is, rules formed by the ‘desire to protect persons who […] owing to inferior bargaining power, are liable to be exploited by others’.³⁹ For norms typically applicable to all employees in an equivalent or equal manner, we will speak of the standard-setting function of protective labour legislation. This includes norms that establish generalised rules for employment contracts which employers have to respect regardless of special circumstances of employees and in general cannot circumvent by contract terms. The standard-setting function stands at the heart of labour regulation designed to concretise the notion laid down in the ILO’s founding charter in the Treaty of Versailles and the Declaration of Philadelphia 1944 that ‘labour is not a commodity’ and thus should not be handled as an article of commerce.⁴⁰

The standard-setting function of individual labour legislation intervenes especially on two levels to limit employers’ power of imposing terms of contracts. These interventions limit employers’ power of disposition of their workers’ time, and they limit employers’ power of termination of a contract. Limitations of working time have been a central issue of labour legislation since industrialisation and aim to safeguard workers’ health and reproduction, but also their recreation and well-being, including sociocultural participation.⁴¹ Standard-setting relates to maximum hours of work and minimum rest periods on a daily, weekly, but also yearly basis including paid holidays, the right to paid leave and overtime restrictions, regulating the distribution of time between employer and employee. Limitations as to the termination of contracts are designed to reduce employers’ right to terminate employment relationships uncompensated and at free will, thus stabilising the status of employees. Regulation in employment law hence aims to reduce the level of insecurity regarding continuity of employment and income, but also to lower poverty risks due to lay-offs in times of crisis.⁴²

Central aspects of legislation in terms of the standard-setting function of individual employment law are thus concentrated in two dimensions: (1) regulation concerning working time, concerning employees’ rights to reproduction and free time or legal time limits to the time extension of work, and (2) dismissal protection, showing complexity and strength of employees’ job security.

B. The Privileging Function

Employment legislation can protect all employed workers equally. However, when the legislator exempts certain groups of employees from protective measures or when law introduces hierarchies of protection—and thus differentiates between employees in a selective manner—it seems appropriate to speak of a privileging function of labour legislation. This privileging function—which forms an essential part of criticism within the recent SER debate (see section 2A, above, The Emergence and Changing Relevance of the SER)—necessarily implies exclusion from protection of those segments of the labour force not covered by the privilege.

The application of the privileging function covers a wide range of norms. A typical set of privileging rules excludes certain groups of workers from special protection provisions. This can be for the size of the enterprise (typically the exclusion of small-scale enterprises), for the type of activity (for example blue-collar versus white-collar workers) or for the type and level of formal education. Another set of privileging rules augments protection for certain groups in comparison to the rest, especially on the basis of seniority.

Characteristically, the privileging function sets incentives for acquiring and maintaining certain positions in employment, thus normalising these positions and marginalising others. Seniority rewards length and continuity of service, thus setting strong incentives for continuity of gainful employment and loyalty towards the company. Seniority rules can occur both purely and in a mixed form, the level of protection being dependent on a mix of length of service, age, family status etc.⁴³

Hence, seniority rules usually exclude workers with discontinuous or interrupted work careers—often females with care roles. Accordingly, the segmenting effect of law can—and frequently does—counteract the principle of gender equality. It can also disadvantage certain ethnic groups, migrants or people from educationally disadvantaged backgrounds. Some authors consider precariousness, insecurity and informality of employment rather as the norm than the exception of the global labour relationship.⁴⁴

By definition, norms conceding special rights to workers for their higher vulnerability do not form part of the privileging function. Special protection due to disability, family-related maintenance obligations, age and other not work-related factors as well as for participative and collective labour law reasons (union officers, data protection officers etc.) is therefore not considered to be excluding or marginalising. The privileging function thus covers the dimensions of selectivity, concerning the non-applicability of protective legislation for certain groups of employees, and seniority, regulating the gradation of protection.

C. The Equalising Function

Labour regulation can also counteract discrimination. Exclusion, marginalisation and segmentation of groups of workers can result from stigma related to gender, race and culture that are not linked to the employment relationship in any objective or functional sense. However, exclusion can also stem from vulnerabilities associated with a comparatively weaker employment status. Norms that serve to inhibit, prevent or overcome segmentation and discrimination of specific groups of workers can be described as comprising an equalising function. In this context, we differentiate between person-related and contract-related discrimination in relation to workers.

Discrimination is person-related, when the person is unable to influence the factors that are used as a base for differential, adverse treatment. Prejudices and stigma present in society do not stop at the gates of companies, although they may be independent of the profit orientation of capitalist companies. Nonetheless, they work as segmenting forces if they are not restricted. Groups thus affected may be disadvantaged or discriminated against because of sex and gender, race and ethnic background, age, religion and belief, disability, sexual orientation and gender reassignment, to name some of the protected characteristics identified in employment discrimination laws around the world.

Discrimination and disadvantaging treatment in labour can become manifest at all stages of the employment relationship. Here, wages as core elements of the employment relationship must be emphasised, since discrimination plays a fundamental role in the field of labour, especially in the gender dimension. Functionally, labour legislation can intervene with an equalising objective to neutralise or undo culturally based discrimination. Legal intervention can cover all types of discrimination or address specific discriminatory grounds. Legislation can be used to simply prohibit discrimination, or also allow for or even prescribe affirmative action or ‘special measures’ in order to overcome structural discrimination.

Disadvantage and discrimination are contract-related if differential or discriminatory treatment is rooted in the particular features of the contract. Structurally more vulnerable workers with working conditions differing from the ‘standard’ are susceptible to marginalisation beyond the mere contractual level. This may be linked to NSERs including part-time contracts, fixed-term contracts or temporary agency work,⁴⁵ but also to areas of responsibility or hierarchies in the pay scale. Those responsible for differential treatment and discrimination beyond mere contractual conditions may be employers, but also other workers or third parties. Legal reactions to these kinds of discrimination or disadvantageous treatment can focus on the outlawing of potentially discriminatory or disadvantaging contracts or especially discriminatory terms as such; or on prescribing equal treatment in some or all aspects which are not necessary side effects of contractual arrangements. The equalising function follows another logic than the privileging function since it addresses disadvantages conferred by social forces. In contrast, the privileging function reduces the applicability of norms set in the context of the standard-setting function to certain categories of workers. Both functions can therefore go together.

Thus, the equalising function covers the two dimensions of discrimination, concerning norms against differential treatment based on certain personal characteristics, and flexibilisation, covering norms which prohibit certain forms of differential treatment based on contractual arrangements in the context of non-standard forms of employment.

D. The Relation between the Three Functions

The three functions described above can be understood to be mutually independent in normative terms, and thereby to create a three-dimensional space within which they combine and interact. However, there is a close, though diverging normative relationship between the privileging and equalising function, since both relate to labour market segmentation, albeit differently. The equalising function relates to personal conditions which cannot be influenced by workers, and to secondary sociocultural effects of non-standard contractual conditions. Against this, the privileging function usually relates to objects that at least ideally involve an act of choice, such as the size of the company or enterprise to work at, loyalty to the company or under which legal regime the contract should be concluded. The aim of equalising norms is thus to eradicate de facto limitations on participation on equal terms in the labour market. The aim of privileging norms is to reward certain groups of workers for their choices, and in effect, steer the behaviour of workers in the long run. Equalising norms cover only prohibition and special measures to overcome differences, whereas privileging norms cover a wide range of possible types of norms.

It may be disputed whether the independence of the three functions in statistical terms can be fully maintained, particularly in research on global-level developments of national labour law. Moreover, in their normative meaning also, standard-setting norms might be regarded as privileging by covering ‘employees’ only (as most labour law norms do). Apart from deficits in enforcement, norms using the legal term ‘employee’ as a criterion of coverage, automatically exclude informal workers, the grey zone of precarious self-employed forms of work, and non-market-related home and care work.

The differentiation of functions is an act of abstraction: all three functions can be present in the same labour legislation, and even in the same norms. An example would be for a law to provide for standard-setting norms on employment protection (standard-setting) in a selective manner (privileging), but securing non-discrimination for female workers (equalising). This also highlights the point that the privileging and equalising function can easily co-exist, as they may address different topics or groups. And even if equalising norms may help to diminish differences of employment conditions between various groups of workers, they do not offset the privileging function unless being transferred into standard-setting norms which are equally valid for all workers. To illustrate their difference, Table 1 provides examples of weak and strong manifestations of the three functions.

E. SPE Typology of Employment Law

In terms of the three functions, national employment legislation is anything but uniform. Not all countries of the world have introduced the same measures, nor have they introduced them at the same time and with the same strength of regulation. The differentiation of the three functions allows us to identify different models or ideal types of employment law. We develop our typology essentially by deduction from the three functions, and although we do not expect all types to be historically relevant, we assume that it can capture all systems of employment law over time. When in the following referring to countries, we aim to make our types more vivid. They are nevertheless the result of deduction and normative analysis, not empirically found ideal types.

By differentiating between high and low strength of regulation of a given function, symbolising strong legislation by the capital letters ‘S’, ‘P’ and ‘E’ and no or weak legislation by the small letters ‘s’, ‘p’ and ‘e’ (eg high standard-setting = ‘S’ - low standard-setting = ‘s’ etc.), we come up with a typology of employment law that differentiates between eight ideal types (see Figure 1). The ideal type with ‘no’ or ‘weak’ regulation in all three functions is symbolised in three small letters as ‘spe’, whereas ‘strong’ regulation in all three functions is symbolised by three big letters as ‘SPE’. Next to these two, six further possible patterns exist. This logical set of patterns serves only as a matrix for the classification of any country in a specific historical stage and does not imply any empirical probability.

In this sense, we have labelled the ideal type with low regulation in all three functions of employment law (spe) as the laissez-faire model. It refers to the case of ‘collective laissez faire’, characterised by strong institutions of collective bargaining, used by Otto Kahn-Freund⁴⁶ to describe the British post-war model, and also fits Scandinavia before the 1970s. From the legislative perspective, the same result can be observed for countries and periods prohibiting or not accepting strong trade unions, where negotiation of employment conditions is left to the individual, or, in effect, to the market. Three more models equally weak in terms of the standard-setting function can thus be discerned. The type with strong anti-discrimination legislation but neither strong standard-setting nor privileging function (spE) we have named market-egalitarian model. As for the elitist model (sPe), we identify a type where rules of employment protection are highly selective and only apply to small groups of workers, possibly to be found in some states with strong authoritarian tradition. The ideal type combining strong privileging and equalising functions with a weak standard-setting function is called the individualising model (sPE), since the setting of generalised standards valid for all is hardly present here.

Four ideal types with a strong standard-setting function can be contrasted to these models. If only the standard-setting function is strongly developed (Spe), we speak of the proto-socialist model. Here, protective norms do not differentiate within the working class, but distinctions based on other aspects cannot be ruled out. This model reflects early working-class aspirations concerning employment legislation which considered gender and racial discrimination as secondary and did not reflect on NSERs. This model might be found in countries influenced by the Soviet Union. In difference, we call the ideal type adding strong anti-discrimination legislation to the strong standard-setting function the universalist model (SpE). This can be found in some Scandinavian countries since the late 1970s, but also Latin American countries with progressive labour legislation. The ideal type showing strong regulation in the standard-setting and privileging functions (SPe) we term the paternalist model. Here, the state does not only protect workers from exploitation by standard-setting but also provides incentives in a selective manner, aiming to steer career planning via labour legislation. As this ideal type does not incorporate anti-discrimination law, it promotes paternalist male breadwinner family models. Here, one might think of the authoritarian states in southern Europe before democratisation. Finally, we name the non-universalist, highly protective, but selective ideal type the ordre public social model (SPE). Strong worker protection is combined here with anti-discrimination legislation on one side, but paternalistic incentive-based steering on the other hand. Examples may be found in contemporary states with authoritarian leadership.

4. METHODOLOGY

Traditional methods of comparative law are hardly suitable to compare legal norms on the standard-setting, privileging and equalising functions to classify countries, trace legal developments in historical terms and detect supranational patterns and changes on a global basis. Hence, the method of choice was leximetrics, meaning the quantification of legal norms. For this aim, existent indicator-based comparative research on employment legislation has been taken into account. In a first step, we defined the concept of legal segmentation, then developed a measurement concept, and examined existing databases for suitable indicators. In a second step, we collected data for indicators that were missing, using various sources. In a third step, we calculated indices for each function and categorised countries concerning the SPE typology.

A. Leximetrics

Leximetrics, or the quantification of legal norms, is a suitable method for capturing the central aspects of protecting and segmenting norms, in the context of discussions on SERs and legal segmentation on a global level. On the one hand, this method reduces complexity to inform about the introduction, changes and potential relations of norms. On the other hand, it is expandable to cover a range of types of norms, and can be used to encode or summarise information that serves as a reference point for the understanding of legal concepts and subsequent qualitative studies. Based on the research question being considered, a set of variables apt to identify the sought-for legal conceptions on the level of comparison can be developed and applied to the legislation to be analysed. The numerical results (index values) can be used to enhance visibility and simplify the comparison of legal concepts utilising time lines, clustering, graphic representations and statistical analysis for a large number of countries.⁴⁷ Leximetrics has been applied to measure, among other things, the foreign influence of statute law, differences in substantial law and the protective strength of legal rules,⁴⁸ and is apt for comparing the functional design of employment laws.

However, leximetric approaches face a number of critiques.⁴⁹ Since numerical quantification abstracts from concrete sociocultural backgrounds, leximetrics fails to fulfil one core task of comparative law, namely the observation of the environment of norms in terms of their doctrinal, legal-cultural and non-legal embedding.⁵⁰ Moreover, a purely numerical approach to norms detracts from the hermeneutic or interpretive dimension of legal thought and practice. In addition, leximetrics differ from traditional aims of comparative law as they do not serve as a method for legal transfers or legal interpretation. These critiques do not detract from the potential value of leximetrics for the kind of exercise we are conducting.

B. Concept Development and Scaling

We followed the guidance on measurement theory outlined by Adams et al.⁵¹ and so developed a construct for the concept of legal segmentation in employment law. This meant identifying its main functions, that is, standard-setting, privileging and equalising, the main dimensions of each function, and, where necessary, central aspects of each dimension. Then, we searched for indicators for each dimension and aspect, that is, observable, tangible manifestations of the abstract phenomena. To establish validity, in other words ensure that we measured what we aimed to measure, we looked for at least three indicators for each dimension or aspect. For a number of categories we could make use of existing indicators of the CBR-LRI, which in some cases had to be rearranged to meet our aims; the remaining indicators were developed and collected by us.

Each indicator was informed by concrete legal norms that needed to be classified on a scalar basis. We followed Adams et al.⁵² in using standardised scales ranging from 0 to 1, where low values represent low levels and high values high levels of a function, dimension or aspect. The nature of each indicator’s scale depended on the number of possible states that we could envisage for that variable; in the end, all indicators that we developed had either binary (true and false) or graduated scales, some of which, drawing on the CBR, were ordinal in nature, and some cardinal (in the case, for example, of certain working time standards). Regarding the assignment of values to our ordinal scales, we hierarchically ordered the possible states according to the respective function and allocated values to points on the scale in equal parts. The inclusion and rearrangement of CBR-LRI indicators sometimes meant that scales needed to be reversed, for example when the privileging function opposed the protecting function measured by the CBR-LRI.⁵³

C. Measurement Concepts

Concerning the measurement of the three functions standard-setting, privileging and equalising, each function consists of two relevant dimensions and—in the case of equalising—two aspects that are measured by various indicators based on legal norms. Our measurement concepts build on 23 individual employment law indicators of the CBR-LRI,⁵⁴ partly putting them into a broader, partly into a different context. The CBR-LRI indicators are supplemented by 12 indicators that we designed and collected ourselves. In the following, we give an overview of the dimensions, aspects and indicators of each function and highlight their relevance.

The measure of the standard-setting function consists of 13 indicators and bears a great resemblance with previous employment protection measures as it captures the level of general worker protection. The function refers to active norm-setting that defines general standards of legal protection. As with previous measures, it focuses on the two core dimensions of individual employment law that shape the SER: working time restriction and dismissal protection (see Figure A1 in the Online Supplement). The measure solely relies on CBR-LRI indicators, and the operationalisation of the dimensions is almost identical to that of the CBR-LRI dimensions regulation of working time and regulation of dismissal.⁵⁵ Our dismissal protection measure differs from that of Adams et al.⁵⁶ by excluding three indicators that capture privileging features which were moved to the privileging function or replaced by a more specific item.

The first form of legal segmentation—the privileging function of employment law—refers to legal norms that actively promote specific parts of the labour force (legal promotion). These norms link employment protection to specific conditions that are covered by the two dimensions seniority and selectivity (see Figure A2 in the Online Supplement), and high scores reflect a high level of norm-related privileging. Although three of the seven indicators stem from the CBR-LRI, the privileging measure supplements the CBR-LRI concept used by Adams et al.⁵⁷ The dimension seniority refers to norms that specify a gradation of employment protection based on length of service. Three of the four indicators have been newly designed; the last indicator stems from the regulation of dismissal subindex of the CBR-LRI.⁵⁸ In contrast, the dimension selectivity covers norms that regulate the applicability of employment protection and decide on the inclusion and exclusion of various parts of the labour force from the scope of protection. One of the three indicators was newly designed, the two others were taken from the different forms of employment and regulation of dismissal subdimensions of the CBR-LRI.⁵⁹ Since the CBR-LRI measures the strength of protection and the privileging or excluding character of norm-setting is accorded a low value, the scales of the indicators have been reversed.

The second form of legal segmentation arises from passivity, or non-regulation, meaning that contractual freedom allows employers to circumvent employment protection resulting in inequalities between groups. The equalising function is measured by 15 indicators that refer to norms that aim to rectify actual differences between groups (legal inhibition) and thus have a reactive nature. This measure captures norms relating to the two phenomena discrimination and flexibilisation and focuses on two aspects each (see Figure A3 in the Online Supplement). High values refer to a high level of norm-related equalising. While the operationalisation of flexibility is almost identical to that of the subdimension different forms of employment of the CBR-LRI,⁶⁰ the dimension discrimination consists of eight newly designed indicators. The dimension discrimination captures norms that aim to rectify inequalities based on person-related characteristics such as gender and race. It focuses on two aspects—access to work and working conditions—and two groups that are especially vulnerable at the labour market: women and ethnic minorities. The dimension flexibility refers to norms that aim to tackle contract-related disparities. By this we mean differential treatment rooted in the particular aspects of the contract which are used as hierarchical features that have effects beyond functional hierarchies. This dimension focuses on the aspects restriction and equal treatment of three forms of non-standard employment: fixed-term contracts, part-time work and agency work. The dimension flexibility differs from the CBR-LRI subdimension different forms of employment by excluding the indicator on the definition of the employment status,⁶¹ which is shifted to the selectivity dimension of the privileging function.

D. Rationale of the Coding Process and Sources

The construction of the SPE data set was challenging and contained potential for error. Whereas for countries of the Global North current legislation was easily accessible via the ILO’s NATLEX database, many national databases and secondary literature, difficulties were encountered with respect to research relating to periods of the past and for many countries less analysed in international publications, especially of the Global South. Language barriers, but also political borders like the Iron Curtain as well as changing borders in the context of wars, decolonisation and revolutions imposed additional challenges. Whereas certain mistakes could be avoided by quality assurance, a certain degree of inaccuracy and gaps in its pioneering stage may have been inevitable given the dimensions of the project. We developed a set of principles and techniques, which alongside the avoidance of unnecessary errors, allowed for transparency of results and hence the possibility of error correction.

Research, data collection and coding were designed to realise the following five principles which in practical coding were interconnected: (1) reliability of used sources, (2) correctness and accuracy of data and coding results, (3) transparency, traceability and verifiability of results and processes, (4) accessibility of data and results and (5) usability of data and results.⁶²

Wherever possible, next to the relevant laws and excerpts of the applicable norms secondary literature was retrieved and analysed. Here, systematical use was made of the publications of the ILO and human rights treaty bodies’ committees of experts’ periodic reports, observations and direct requests in the context of relevant international conventions. The ILO’s EPLex⁶³ was systematically consulted, as was the IELL⁶⁴ with monographs on 78 countries. Values by year and indicator were coded with a traffic-light system, indicating the quality/reliability of the source.

The resulting WoL data set consists, as of 2021, of 35 indicators covering 115 countries for the period from 1970 to 2013. It contains 23 indicators incorporated from the CBR-LRI and 12 new indicators coded by us. It will be extended both historically and geographically to 151 countries for the period of 1970–2018, while data for the period from 1880 to 2018 will be available for England (UK), France and Germany and 73 of their former colonies and alike dominated countries. The data set will be publicly available through the interactive, web-based information system on global dynamics of social policy: the Global Welfare State Information System (WeSIS). Specific underlying norms and sources are documented in the CBR-LRI coding template⁶⁵ and the SPE country documentation.⁶⁶

E. Index Calculation and SPE Categorisation

To measure the standard-setting, privileging and equalising function, we calculated three indices to identify SPE types of employment law. Following the measurement concepts described above, we chose the following method of weighting and index calculation: We used the mean of all indicators of one aspect, again the mean of all aspects of one dimension, and again the mean of all dimensions of one function, thereby assigning equal weights to each aspect or dimension, and thus equal weights to all indicators of one aspect or dimension. However, we made one adjustment: In contrast to the other dimensions, the dimensions working time, dismissal protection and selectivity showed empirical maxima below ‘1’ over the period of 43 years (0.87, 0.95 and 0.72, respectively). We therefore normalised their scales, that is, we divided them by their empirical maximum. This decision was based on the consideration that the theoretically derived scales in these cases might be too rigid, and that the highest observed value represented the highest empirically achievable value, which would result in a stronger weighting of the other dimensions inherent to a measurement error. We are aware that this is a time-varying transformation that might cause problems in the future; in other words, future maxima of the three dimensions might increase, although the likelihood is rather small. The correlation between the three indices above the whole sample are 0.05 for S and P, 0.37 for S and E, and 0.13 for P and E. The clear positive relation especially between S and E leaves enough unexplained variance to consider them to be separate concepts.

To categorise countries according to the SPE typology, we tried various approaches and examined the results. We decided for the method that according to legal content analysis bore the best outcome. First, we established two thresholds that differentiated between low, medium and high levels of each function. We chose relative thresholds that fitted the distribution of each function better than absolute ones, but avoided the problem of time variation by anchoring them in specific years to allow comparisons across different samples and periods. As anchors we selected years when the respective functions of labour law were more or less fully developed, i.e. the year 1980 for S and P, and the year 2006 for E. The middle category of each function was defined as the standard deviation around the mean of the anchor year which captures around 38.2% of the distribution, while ‘low’ covers all values below the mean minus half of the standard deviation, and ‘high’ all values above the mean plus half of the standard deviation (for S: mean = 0.49, lower threshold = 0.41, upper threshold = 0.57; for P: mean = 0.29, lower threshold = 0.21, upper threshold = 0.36; for E: mean = 0.51, lower threshold = 0.41, upper threshold = 0.62). Based on this categorisation, ideal SPE types are countries where all three functions showed low or high levels. Furthermore, we identified relatives of ideal types, i.e. settings that resembled ideal types. First-degree relatives are countries where two of three functions are either high or low, and the third function ranges in the middle, but shows a tendency towards being high or low according to its location above or below the mean. In contrast, second-degree relatives are countries where one function is clearly lowly or highly developed, and the other two functions are in the middle, but show respective tendencies. Countries, where all three functions are in the middle category, make up a middle type.

5. PATTERNS OF LEGAL SEGMENTATION AROUND THE WORLD

Our findings indicate that the grading of standard-setting, privileging and equalising functions of labour law and their composition to different SPE types can be applied to employment regulation in countries all over the world. They thus provide a central component for a worldwide reconstruction of the SER. Data for 2013 display a large variety of paradigms of labour regulation. As we deduced eight ideal types of legal protection and segmentation theoretically based on the three historically developed functions of worker protecting labour law, we could not expect to find empirical equation of all types for the present state of development. We empirically identified five of them, namely laissez-faire (spe), individualising (sPE), paternalist (SPe), universalist (SpE) and ordre public social (SPE). For the other three, namely elitist (sPe), market egalitarian (spE) and proto-socialist (Spe), we only found related models. We cannot provide the results of the descriptive analyses for all countries for reasons of space.⁶⁷ We focus on general impressions concerning the empirical (non-) clustering of countries in different regions of the world based on the SPE typology.

We start with the conventional perspective of quantitative comparative labour law research, in other words, the level of protection captured by the standard-setting function. Map 1 displays a map that shows the standard-setting index for the 115 surveyed countries, where a darker tone indicates a higher level of protection. At first sight, only few countries display a low level of dismissal protection and working time regulation. Low levels can be observed for liberal welfare states or market economies⁶⁸ such as the USA, the UK and Australia, but also for Japan or the former British colony Nigeria.

Map 1 shows that most countries have high standard-setting values. On the European continent, this is true for Western European countries such as France, Italy, Spain and Portugal, most of the Scandinavian countries and almost all former communist countries in Central and Eastern Europe, including Russia. In Asia, emerging economies such as India and China, Bangladesh, Pakistan and Malaysia are in this group. Latin American countries such as Argentina, Bolivia, Chile, Paraguay, Peru and Venezuela as well as many African countries such as Algeria, Angola, Botswana, Sudan, South Africa also show high protection standards. In contrast, Germany, Norway and Brazil display only medium levels of employment protection and working time regulation—similar to countries such as Saudi Arabia, South Korea, Singapore, Kenya or Uruguay.

Map 2 provides a glimpse on the extended perspective associated with the differentiation between different functions of labour law. Here, the distribution of the privileging function is mapped, showing to what extent specific groups of employees are advantaged by norms that emphasise seniority or selectivity in regard to employment protection, thus creating active legal segmentation. Again, the darker the tone, the higher the index value. A first view indicates that there are only some countries with high privileging values. Among them are Western European countries such as Germany, Belgium, Finland, Italy and Portugal, but also Brazil, India and China. Some more Asian countries such as Bangladesh and Pakistan, or Honduras and Paraguay in Latin America as well as Botswana and Tanzania in Africa, among others, are in this group. This group is complemented by some of the liberal market economies such as UK and Canada.

A medium level privileging function has been measured in OECD countries such as USA, Australia and France, several African countries such as South Africa and Nigeria and many Latin American countries such as Ecuador and Bolivia.

Countries with low privileging values include Denmark and Spain, most Central and Eastern Europe (CEE) and other former communist countries such as Russia, but also Japan, Chile, Venezuela, Algeria or Ghana. As in the case of standard-setting, there is neither a clear North-South divide, nor a differentiation according to continental proximities.

Extending this line of analysis but omitting a discussion of the equalising function,⁶⁹Map 3 shows how the SPE types are distributed around the globe.⁷⁰

In line with the small number of countries that have low standard-setting values, the laissez-faire (spe) and market egalitarian type (spE) of labour regulation are rather rare. The pure spe pattern exists only in Japan and Myanmar, while related patterns are to be found in Georgia, New Zealand and Zimbabwe. A market egalitarian (spE)-related type is to be found only in Ghana. Similarly, the elitist type (sPe)—combining low protection with high selectivity of labour law protection—only exists in the form of related regulation patterns in Cameroon, Nigeria, Switzerland and the USA. However, it seems remarkable that there is a certain clustering of liberal market economies (UK, Ireland, Australia and Canada) according to the individualising type (sPE) and its relatives—meaning that these countries are combining low protection and high selectivity with strong equalising regulations.

In contrast, the paternalist type of regulation (SPe) and its relatives—lacking strong equal treatment regulations—are predominant in Asian countries, as, for example, in the countries formerly colonised as British India (Bangladesh, India and Pakistan) and China, but also in some African countries such as Tanzania, DR Congo and Egypt. Moreover, Latin American countries that are members of the Mercosur tend to be found in this group.

Proto-socialist (Spe)-related types of regulation that—without strong regulation of both privileging and equalisation—seem to be spread around the Global South, resuming, for example, some African (eg Uganda, Zambia), Asian (eg Thailand, Singapore) and also a Latin American country (the Dominican Republic).

Some European countries including France, Germany, Italy, Portugal, Finland and Sweden, but only a few CEE countries, have combined high or at least medium level protection with medium or high privileging and strong egalitarian regulations, thus displaying the ordre public social type of labour regulation (SPE) and its relatives. Moreover, we find, for example, Brazil and Mexico, but also countries such as Cambodia, South Korea, Morocco and Turkey in this group.

The universalist type (SpE) and its relatives suggest high protection, little selectivity and high equity regulations and are predominantly to be found in CEE and other former socialist countries, plus some other EU member states such as, for example, Denmark, the Netherlands, and Spain. Beyond Europe this type that indicates the least legal segmentation is to be found, for example, also in the Andean States of South America, Indonesia and South Africa.

Concerning a global pattern of SPE ideal types, in 2013, a general disparity between Global North and Global South could not be found. Liberal market economies of the Global North and Japan stand out among the regulation types with lowest values of protection in combination with diverging levels of selectivity and equity regulations along the laissez-faire (spe), the elitist (sPe), the market egalitarian (spE) and the individualising (sPE) type of labour regulation. In contrast, proto-socialist (Spe)-related and paternalist (SPe) types and their relatives are rather to be found in the Global South, particularly in Asia and Africa—indicating a weak relevance of equalising regulations in these regions. Moreover, the ordre public social (SPE) type is most common within European countries, reflecting the history of high, but most selective labour protection in combination with equalising regulation. However, this group is joined by numerous countries of different regions of the Global South indicating similar types of employment regulation. Finally, the least segmenting labour regulation along the universalist type (SpE) and its relatives are often to be found in countries with some degree of a communist past. The high incidence of EU Member States—or those applying for membership—among the last two types may be influenced by the fact that the EU has promoted equalising regulations during the last few decades.⁷¹ All this indicates significant discrepancies between de jure labour regulation and de facto labour market structures, pointing towards the ‘governance issue’ of the informal economy and the ‘the lack of good governance for the proper and effective implementation of policies and laws’⁷² in many regions of the Global South.

6. SCIENTIFIC AND POLITICAL PERSPECTIVES

Unidimensional approaches of former analyses that focused on the protection of workers, principally through dismissal protection, controls over working time and the regulation of non-standard forms of employment, were able to benchmark labour law by reference to social justice agenda of protecting labour vis-à-vis capital. Outcomes based on such unidimensional measurements may produce results which can be used in correlations of labour protection with unemployment or employment rates, productivity or investment—in other words, to scientifically assess whether, or not, labour protection is an obstacle to the free market or economic growth. These analyses, however, are not able to shed light on legal segmentation, as they remain indifferent concerning the just legal treatment of working people vis-à-vis other working people (other sex, other age, other social strata, other ethnicities etc.). They ignore regulatory patterns which selectively protect male breadwinners and discriminate against women or workers not holding a SER. Thus, they ignore possible consequences of law with respect to the de facto segmentation and polarisation of labour markets, and fail fully to address the quest for equity along these dimensions, which has been gaining ground enormously in the global labour movement within the last five decades.⁷³

As against that, the WoL database introduces a necessary differentiation, combining different functions of employment legislation and making visible or traceable different forms of legal segmentation around the world. The newly developed three-functional measurement concept we have presented here does not only allow us to research the variations in regulation across the different selective, protective and equalising functions of labour law, but also provides new indicators of different types of regulation concerning the selectivity and equity of labour law around the world.

Admittedly, in this article, we have only provided a snapshot of our findings for the year 2013. Our comparative database, however, is historically orientated, and will in due course cover the time period starting with the origin of labour law in most countries of the world. The three-functional analysis will therefore allow not only for modelling the current relationship of protective, privileging and equalising labour standards. In addition, it will shed light on how the three functions developed globally over time, and in particular which of the functions gained momentum in which historical period, allowing for explanations related to colonialism, world-trade, UN or ILO decisions. It will provide foundational findings, enabling us to discover and elaborate which variants of the western SER model were diffused to other world regions over time. The SPE typology, thus, may contribute to and enable the analysis of social justice with respect to labour, as it brings to the fore the ways in which normative inequalities are either legally aggravated into social disparity, or moderated and cushioned into legally recognised diversity.

The typology of employment laws based on the three functions we have set out here (the SPE typology) provides global comparative information about legal segmentation, focused on the interplay between standard-setting and privileging components of employment laws in the world. It can therefore be regarded as an important step towards a SER typology. Building on the outcomes of SPE types discovered by the WoL project, we will be able to identify correlations between typical indicator constellations (regulatiory ‘patterns’) of individual employment regulations in Europe and their counterparts abroad. This will allow us to research how global interdependencies (on issues including trade, migration, colonialism, war, ILO standards, emulation and epistemic communities) influenced the spread of labour regulations into selected countries and regions of the world. Moreover, it will enable us to analyse, how legal segmentation influences de facto labour market structures—since we know that in many countries of the Global South high shares of informal employment go along with high standards of legal protection and/or privileging regulations.⁷⁴

With respect to policies and politics there are obviously strong relationships between the development of SPE types we found, and current controversies both among different national actors, but also among global players on the perspectives of future social and labour policy.

Each of our eight types of SPE-employment law models represents a particular way how societies are dealing with the ‘social question’ with respect to norm-setting—and whether they choose either inclusive (covering all those who are in a social situation of ‘need’) or exclusive (abstaining from universal coverage and limiting standards to selected groups) forms of regulation, or something in between. An exclusive type is the laissez-faire model (spe) where low standard-setting coincides with low privileging and equalising. This type of regulation may relate to quite different states—those entirely rejecting the responsibility for vulnerable groups and leaving them to market forces, or those executing this responsibility in cooperation with organisations of the social partners.⁷⁵ The point of culmination for an inclusive employment law policy is the universalist model (SpE) where low privileging coincides with high levels of standard-setting and equalising. Behind this type of regulatory pattern, promising overall inclusion, we envision a state proactively caring for all those who are in need and therefore providing for a high amount of protecting labour legislation. However, the effects of such regulation highly depend also on effective enforcement by the state that indicates a great variety between Global North and Global South.⁷⁶

The ideas inspiring these two opposite ideal types correspond to current, if controversial global social policy developments. The market-led laissez-faire type of employment regulation is the legal equivalent of ideas that the OECD and the World Bank share with neoclassical economists.⁷⁷ It is driven by economistic assumptions which treat employment regulation as such as an obstacle to economic performance—although this has long been criticised for both theoretical and empirical shortcomings.⁷⁸

In contrast, the universalist model instantiates ideas of the constituting documents of the ILO based on ‘inclusive’ labour and social policy. The preamble to the 1919 ILO Constitution contains the important statement affirmed in the Philadelphia Declaration of 1944 that ‘lasting peace can be established only if it is based upon social justice’.⁷⁹ This demands an improvement in those working conditions that continue to perpetuate injustice, hardship and privation for many people today, and, as the UN-International Covenant on Economic, Social and Cultural Rights puts it, the creation of ‘conditions [...] whereby everyone may enjoy his economic, social and cultural rights, as well as his civil and political rights’.⁸⁰

This article has focused on the methodological concepts and tools needed to identify, measure and compare legal segmentation patterns around the world, generating on a quantified, differentiated data set. We do not intend here to enter into a substantive debate about the justification of this or that regulatory pattern. But we are confident that our research approach, as well as future analysis based on the new data set, may prove relevant to researchers and to national and international political actors. It may thus provide an incentive interest to evaluate the consequences of the different regulation models we have identified and to re-regulate them with the objective to eliminate or reduce de facto labour market segmentation and inequality in the future.

We are grateful for funding from the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) -Projektnummer 374666418 - SFB 1342.

Footnotes