Abstract
This article focuses on national public administration activities that relate to employment, social protection and industrial relations. The International Labour Organization (ILO) refers to these activities collectively as ‘labour administration’ and regards the bodies that conduct them within individual countries as together forming national systems of labour administration. This article explores the concept of ‘national system of labour administration’ and considers the potential contribution of comparative institutional analysis in understanding how national systems are organised and change over time. The article also compares the organisation of national labour administration systems in European Union (EU) countries and analyses how these systems have developed since the start of the economic crisis that erupted in 2008.
Introduction
This article focuses on national public administration activities that relate to employment, social protection and industrial relations. The International Labour Organization (ILO) refers to these activities collectively as ‘labour administration’ and regards the bodies that conduct them within individual countries as together forming national systems of labour administration (Heyes and Rychly, 2013). Although bodies that perform labour administration functions differ between countries, they typically include ministries of labour, public employment services (PES), labour inspection services and dispute resolution services. These bodies are responsible for devising and implementing labour laws and policies, providing support to unemployed workers, ensuring that firms comply with their obligations to their employees and providing advice to employers and workers. As such, labour administration substantially influences the operation of labour markets, the quality of jobs and outcomes relating to, for example, occupational safety and health (OSH) and the enforcement of labour standards.
Many of the issues that are of central importance to labour administration have received a substantial amount of research attention from academic researchers and the ILO. Much has been learned about, for example, the activities of ministries of labour (Rychly, 2013a, 2013b), measures directed at unemployed workers (e.g. King, 2013; Van Berkel, 2009), institutions of social dialogue (e.g. Papadakis and Ghellab, 2014), arbitration, conciliation and mediation services (e.g. Ebisui et al., 2016) and the role of enforcement agencies in encouraging employer compliance with labour laws (e.g. Páramo and Vega, 2017). Research has tended to focus on specific labour administration functions or policy areas. Recently, however, there has been interest in extending the scope of research to include comparative analyses of national labour administration systems (see Heyes and Rychly, 2013) and it is this that is the central concern of the present article. The aims of the article are twofold. The first is to explore the concept of ‘national system of labour administration’ and consider the potential contribution of comparative institutional analysis in understanding how national systems are organised and change over time. The second aim is to compare the organisation of national labour administration systems in European Union (EU) countries and analyse how these systems have developed since the start of the economic crisis that erupted in 2008. The article begins with a discussion of labour administration as a component of public administration. ‘Comparative capitalisms’ and comparative public administration scholarship imply that there will be important differences in the ways in which national labour administration systems are organised, particularly in relation to the degree of centralisation of labour administration, the regulation of the labour market and the extent to which social partners are involved in labour governance. The article therefore considers whether the characteristics of EU national labour administration systems differ between countries associated with different ‘varieties of capitalism’. It then moves on to examine changes in the organisation of different labour administration systems over time, paying particular attention to the period since 2008. The article concludes by mapping out elements of a research agenda for comparative research in labour administration.
Labour administration, public administration and varieties of capitalism
Labour administration is defined by the ILO’s Labour Administration Convention, 1978 (No. 150) as comprising ‘public administration activities in the field of national labour policy’. This includes institutions, activities and outcomes across the entire field of labour policy, including employment policy, labour law, social protection and industrial relations. The Labour Administration Convention and the accompanying Recommendation 1978 (No. 158) provide policy guidelines and objectives that are intended to be given concrete form in specific measures adapted to national conditions. Their premise is that ‘adoption in a country of labour laws and regulations might be ineffective if that country does not have at its disposal a competent and efficient labour administration, entrusted with the tasks of following the development of the social situation, of supervising the implementation of legislation and of ensuring the operation of existing machinery’ (Valticos, 1979: 214).
The ILO’s Convention No. 150
1
also refers to the term ‘system of labour administration’, which is defined as encompassing: … all public administration bodies responsible for and/or engaged in labour administration – whether they are ministerial departments or public agencies, including parastatal and regional or local agencies or any other form of decentralised administration – and any institutional framework for the co-ordination of the activities of such bodies and for consultation with and participation by employers and workers and their organisations.
National labour administration systems therefore involve labour ministries (or their functional equivalents), the bodies that are responsible for delivering services, particularly those that relate to the re-employment of unemployed workers, the enforcement of labour standards and dispute resolution, and methods of involving employer organisations and trade unions in the development of policies and their implementation. In practice, however, countries differ in terms of the bodies that are involved in labour administration, how services are organised, the resources that are devoted to labour administration and the extent and nature of employer and trade union consultation and involvement (Heyes and Rychly, 2013).
Although a number of comparative analyses of particular labour administration service functions, particularly public employment services (PES), have been undertaken (e.g. Borghi and Van Berkel, 2007; Finn, 2016; Páramo and Vega, 2017), differences between national labour administration systems in toto have rarely been studied in depth. The need for comparative studies has been emphasised by the ILO and researchers with an interest in labour administration (e.g. ILO, 2011; Heyes and Rychly, 2013) and a small number of studies have been undertaken on behalf of the ILO (for example, Heyes, 2004, 2011; Hastings and Heyes, 2016a). These studies tend to describe national systems at a particular point in time and pay less attention to how systems change over time and whether their development is path dependent. Issues related to system changes and path dependency are afforded considerable attention in the comparative political economy and public administration literatures. With regard to the former, the ‘varieties of capitalism’ (VoC) analytical framework, initially popularised by Hall and Soskice (2001), emphasises that the competitive advantages associated with different combinations of socio-economic institutions create incentives for governments and employers to preserve and enhance their specific characteristics. Hall and Soskice initially distinguished between ‘liberal market economies’ (LMEs) (such as the UK and Ireland) in which coordination of economic activity is supposedly achieved through formal contracting and market-mediated exchanges, and ‘coordinated market economies’ (CMEs) (such as Germany, Austria and the Nordic economies) in which coordination takes place through complementary institutions and networks that facilitate access to information, skills and other resources. Their analysis also emphasised the importance of cooperative industrial relations and social dialogue as underpinning characteristics of CMEs. Although Hall and Soskice did not directly discuss labour administration, their analysis nevertheless hints at ways in which national labour administration systems in CMEs and LMEs might differ. We might, for example, expect employers and trade unions to play a greater role in the governance of labour administration in CMEs than in LMEs. A ‘weakening of organized labor and a substantial amount of deregulation’ (Hall and Soskice, 2001: 57) accompanied by a ‘fundamental renegotiation of labor’s position … in the political economy as a whole’ (Thelen, 2001: 103) is predicted for LMEs, but in CMEs, where deregulation is a threat to comparative advantage, the maintenance of labour-supportive regulatory regimes, ‘resiliency in labor institutions’ and a strengthening of social partners are expected (Thelen, 2001: 71). LMEs might also be expected to make greater use of private and voluntary sector providers in the delivery of labour market services, given the greater emphasis they place on the use of market mechanisms. Furthermore, Culpepper (2001: 302) predicts a greater degree of decentralisation in LMEs when compared with CMEs and the development of associations aimed at encouraging private actors ‘to begin cooperating with each other’.
Various attempts have been made to extend and refine the VoC typology. Amable (2003), for example, substituted ‘continental’ (e.g. Germany) and ‘social democratic’ (e.g. Denmark) models of capitalism for Hall and Soskice’s CME category and also outlined Asiatic (e.g. Japan) and Mediterranean (Spain, Portugal, Greece and Italy) models. King (2007) sought to insert the economies of Central and Eastern Europe into the VoC framework by defining them as ‘liberal-dependent’ economies that resembled the LME type while being more reliant on foreign capital.
Similar typologies are to be found in the comparative public administration literature. Page (1995), for example, has distinguished between German Continental, French Continental, South European, British-Irish and Scandinavian types of administration, while Tepe et al. (2010) have claimed that public administration systems within the OECD cluster into three regimes, which they define as Anglo-American, French/German and Scandinavian public management. These regimes are distinguished according to differences in their degree of public employment regulation (e.g. public sector collective bargaining) and politico-administrative regulation. Furthermore, Tepe et al. claim that, with the partial exception of the Scandinavian regime, there is a complementary relationship between public administration types and different ‘varieties of capitalism’ (e.g. LMEs have Anglo-American public administration systems).
As has been the case for the comparative political economy literature, public administration researchers have been concerned with the question of whether national institutions and policies are converging or diverging. Particular attention has been given to the extent to which practices associated with the New Public Management (NPM) have been adopted and whether they have been adopted in similar ways in different countries. Some studies (Bevir and Rhodes, 2003; Pollitt, 2007) have found that NPM reforms have been strongly affected by national contextual factors (Pierre and Rothstein, 2011), such as inherited institutions, prevailing traditions of governance and domestic politics and, as a consequence, reform processes and outcomes have varied among countries associated with a specific public administration regime type (Halligan, 2011; Kickert, 2011). However, while public administration regimes remain diverse, focusing on administrative systems as a whole may result in similarities relating to specific elements of public administration being overlooked, given national complexity. As Brans (2007: 271) has argued, comparative research is most likely to prosper if attempts are focused on relatively narrowly defined subjects rather than public administration systems as a whole (Brans, 2007: 271). A focus on labour administration, as opposed to public administration in toto, may therefore serve to provide clarity.
An important concern for this article is the impact of the economic crisis that erupted in 2008 for labour administration. Researchers in the field of comparative political economy have noted that liberalisation of labour markets since the start of the crisis has not been confined to LMEs and, moreover, that the crisis accelerated processes of liberalisation that were already underway (Thelen, 2014). Liberalisation in relation to labour market policy has included a weakening of employment protections and a dilution of entitlements in relation to social protection (Hastings and Heyes, 2016b). While such developments suggest a common direction of change across varieties of capitalism towards an ‘unravelling of CMEs and the further liberalization of LMEs’ (Baccaro and Howell, 2017: 10), it does not imply ‘that the CME category has been emptied so that the advanced capitalist world is populated solely by varieties of LME’ (Baccaro and Howell, 2017: 15). For instance, on German capitalism, Hall (2015: 48) recently wrote that, labour market liberalisation notwithstanding, ‘the basic features of the German political economy continue to distinguish it from many others’, notably ‘its Anglo-American counterparts’. According to Thelen (2014: 202), ‘CMEs increasingly find themselves pulled in one of two directions’. The first is a Nordic CME model of ‘embedded flexibilization’ in the Scandinavian countries, which combines marketisation with a collectivisation of risk, and the second is a Continental CME model of ‘dualizing liberalization’ in countries such as Germany, which remains coordinated but has become less egalitarian (Thelen, 2014: 202).
Analysis of the effects of the crisis on labour administration in different countries can, as the article demonstrates, shed further light on these reform trajectories. In the VoC literature, economic crises are generally treated as ‘exogenous shocks’ that trigger responses that are closely related to the different varieties of capitalism, with each type of country acting to protect its respective institutional complementarities and derived comparative advantage (Iversen, 2007). CMEs are expected to make greater efforts to minimise labour market disruptions resulting from the crisis in order to ensure continued access to a highly skilled workforce, whereas LMEs are expected to continue to favour more fluid labour markets. Consequently, the implied VoC prediction is that crises are path-reinforcing, therefore leading to increased divergence between the different varieties of capitalism. The public administration literature (see e.g. Kuhlmann and Wollmann, 2014; Randma-Liiv and Kickert, 2017), however, considers an alternative possibility, namely that economic crises could be path-breaking rather than path-reinforcing and should therefore be treated as critical junctures or turning points. This notion is borrowed from historical institutionalist theory, according to which national institutional trajectories are characterised by ‘punctuated equilibrium’ – institutions that are relatively stable due to historical legacies, but which nonetheless may undergo change as a result of an exogenous shock (crisis). These two alternative theoretical underpinnings mean that public administration literature on the effects of economic crises on administrative reform is centred around a dichotomy between institutional continuity and change (Peters et al., 2011), which however draws on a common background that views crises as exclusively exogenous, neglecting capitalist dynamics and endogeneity.
This overview of the comparative capitalisms and public administration literatures raises the question of whether differences between national labour administration systems correspond to the different ‘varieties of capitalism’. We explore this in a European context, assessing the veracity of the following assumptions about labour administration that are implicit in the VoC literature:
Labour administration is expected to be more centralised and coordinated in CMEs than LMEs.
Marketisation of labour administration is expected to be more in evidence in LMEs than CMEs.
A higher degree of involvement of social partners in labour administration is expected in CMEs than in LMEs.
Economic crises are expected to be path-reinforcing in both LMEs and CMEs, thereby leading to increased administrative divergence between the two varieties of capitalism.
Hypotheses 1–3 are analysed in the following section of this article, while the effects of the economic crisis (hypothesis 4) are examined in the section ‘Developments in labour administration since the crisis’.
The current state of national labour administration systems in Europe
In this section we examine the organisation of national labour administration systems in selected European countries, in comparative perspective, against the VoC predictions outlined above. Our sample contains 11 countries: the UK and Ireland, classified as LMEs in the VoC literature, plus Germany, Austria, the Netherlands, Denmark and France as CMEs. For purpose of comparison we also include the Czech Republic and Hungary as liberal-dependent economies which are expected to follow a broadly similar trajectory to LMEs (King, 2007) and Spain and Italy as examples of the Mediterranean model. Table 1 summarises key labour administration indicators for our sample.
Labour administration indicators – selected European countries.
To evaluate the first hypothesis (that labour administration will be more centralised and coordinated in CMEs than LMEs), we begin by analysing labour ministries as the central institutions in national labour administration systems. One way to assess the degree of centralisation is to ask whether the ministry is holistic – i.e. responsible for all matters related to work and employment as well as social policy (Rychly, 2013a). This is the case for six countries in our sample: Ireland (Department of Employment Affairs and Social Protection), Germany (Federal Ministry of Labour and Social Affairs), the Netherlands (Ministry of Social Affairs and Employment), Denmark (Ministry of Employment), Italy (Ministry of Labour and Social Policy) and the Czech Republic (Ministry of Labour and Social Affairs). In the UK, France and Spain, responsibilities are shared across ministries: between the Department for Work and Pensions and the Department for Business, Energy and Industrial Strategy (UK); between the Ministry of Labour, the Ministry of Solidarity and Health, and the Minister of State for Gender Equality (France); and between the Ministry of Labour, Migration and Social Security and the Ministry of Health, Consumption and Social Welfare (Spain). In Hungary, there is no de facto labour ministry; since 2010, responsibilities for labour and social welfare issues have been shared between the Ministries of the National Economy (renamed Ministry of Finance in 2018), Interior and Human Capacities. In Austria, labour issues are the domain of a multi-portfolio ministry, the Federal Ministry of Labour, Social Affairs, Health and Consumer Protection (created in January 2018). Thus, there is diversity in the way responsibilities for labour matters are distributed across government, with no evidence of a consistent difference between CMEs and LMEs.
The veracity of the first hypothesis can also be assessed by examining labour inspectorates and PES – specifically, whether they are part of their respective labour ministry’s organogram, or instead autonomous. Labour inspectorates are on the organogram in six of the selected countries, suggesting a high degree of centralisation: Austria (Arbeitsinspektion), the Netherlands (Inspectorate SZW), Denmark (Arbejdstilsynet), France (Inspection du travail), Spain (Inspección de Trabajo y Seguridad Social) and the Czech Republic (the State Labour Inspection Office). Although Ireland’s Workplace Relations Commission (WRC) and the Italian National Labour Inspectorate are quasi-autonomous (not on their respective labour ministries’ organogram but periodically supervised by the ministry), they nonetheless represent examples of centralisation, given that they had been created to merge functions previously carried out by a range of state agencies. Germany presents a mixed picture: labour inspection is not part of the ministry organogram, but in 2009 many formerly independent labour inspectorates were integrated into the general state administrative apparatus. At present, there is great variety between Länder, with autonomous labour inspectorates in some states, and labour inspectorates that are part of the administrative structure of the Land in other cases. Hence, it is Germany’s federalism that limits centralisation. In the UK, the Health and Safety Executive (HSE) is sponsored by the Department for Work and Pensions but remains quasi-autonomous (not on the organogram). In Hungary there is no labour ministry organogram to speak of: safety at work and overall labour inspection activities are currently the domain of the Ministry of Finance, while the Ministry of Human Capacities presides over the Chief Medical Officer (ÁNTSZ), which carries responsibility for occupational health and work hygiene issues (Eurofound, 2015).
PES are autonomous or quasi-autonomous bodies (not on the organogram) for most countries in our sample: Ireland (the independent Intreo service), Germany (Bundesagentur für Arbeit [BA]), Austria (Arbeitsmarktservice [AMS]), the Netherlands (Werkbedrijf, a division of UWV, the Employee Insurance Agency, an autonomous body), France (Pôle emploi) and Italy (Agenzia Nazionale Politiche Attive Lavoro). PES appear on the organogram in the case of Denmark (the Agency for Labour Market and Recruitment), Spain (Servicio Público de Empleo Estatal) and the Czech Republic (Employment Services Administration). The UK’s PES have been part of the Department for Work and Pensions since Jobcentre Plus ceased to exist as a separate entity in 2011. In Hungary, the PES is currently governed by the Ministries of Finance and the Interior. There is evidence that county labour centres receive conflicting instructions from the two ministries (Eurofound, 2015).
Consequently, not only does this overview of labour inspectorates and PES invalidate the VoC assumption that decentralisation is more extensive in LMEs, but it also shows that (de)centralisation in one aspect of labour administration (labour inspection) does not necessarily mean (de)centralisation in other aspects (PES). On the other hand, we will see later in the article that centralisation has been an important trend in labour administration since 2008.
A further potential indicator of the ‘fit’ between varieties of capitalism and national labour administration systems is the extent to which labour administration bodies and activities are coordinated. To assess this, we focus on the degree of coordination between employment services, social assistance and social services, drawing on Bouget et al. (2015), who collated responses from national experts who are part of the European Social Policy Network. These experts were asked to rate the degree of coordination in their country as either very good, medium, or weak. The degree of coordination was rated as ‘weak’ in the case of the UK, Italy, the Czech Republic and Hungary, and medium in all other cases. Administrative fragmentation, limited scope for formal coordination and the lack of common databases were all cited as barriers to coordination (Bouget et al., 2015).
The second hypothesis is that marketisation will be more extensive in LMEs than CMEs. We assess this by using the ‘extent of adoption of new public management tools’ as a proxy. NPM is a trend in public administration that prioritises efficiency, quantifiable results, targets, as well as limits on spending. Values for this indicator are taken from Jeannot and Bezes (2016), who calculated these based on a survey with thousands of top civil servants across Europe. Participants were asked to rate the extent to which NPM tools are used in their organisation on a scale from 1 to 7, where 1 means ‘not at all’, and 7, ‘to a large extent’. Country figures represent means of responses from that particular country. Amongst the countries in our sample, the UK appears to have adopted NPM tools to the greatest extent (5.19 out of 7), but it is closely followed by the Netherlands (5.01) and Denmark (4.82). The lowest degree of NPM adoption seems to have occurred in Spain (3.28) and Hungary (3.49), with Italy, Ireland, Germany, Austria and France somewhere in the middle. This suggests that marketisation is not confined to LMEs. Jeannot and Bezes’ (2016) data on NPM refer to all policy fields, not just labour issues. However, with regard to the adoption of NPM tools in European public employment services specifically, Weishaupt (2010) has shown that this trend started in the UK, then spread to Denmark and Sweden, and was later internalised by most other European PES as a result of a diffusion of ideas via EU institutions and the OECD. In practice, this resulted in ‘a common focus on performance, quality and case management as well as contestability in service delivery’ (Weishaupt, 2010: 480) across European PES.
Our final indicator reflects whether social partners are involved in labour governance. The information in Table 1 refers specifically to the governance of PES. A high degree of involvement of social partners in the governance of PES is apparent in Germany, Austria, France and Spain, where social partners are members of governing boards. In these countries, not only are social partners directly involved in the governance of PES, but they also comprise the majority of the governing board members (Weishaupt, 2011). In Italy, the Steering and Surveillance Council of PES is bipartite, including unions’ representatives but not employers’ representatives (European Commission, 2016a). These cases seem to confirm the VoC notion that CMEs are characterised by strength of social partnership. However, in the Netherlands and Denmark there is little social partner involvement in the governance of PES (although, as discussed in the following section, this has not always been the case). The liberal-dependent countries also score poorly on this indicator: in the Czech Republic, the state engages in negotiations with social partners only in informal settings (European Commission, 2016a), whereas in Hungary the role of social partners has been further weakened following the closure of the National Labour Office (NMH) (Eurofound, 2015).
Changes in labour administration over time
The discussion thus far has suggested that the distinctions employed in the VoC framework are of little explanatory value where analysing national differences in the organisation of labour administration systems. In this section we consider common tendencies in the development of national labour administration systems over time.
There is broad consensus among public administration researchers that the economic crisis of the 1970s was a turning point which gave rise to common ‘mega trends’ of decentralisation, privatisation and marketisation in public administration reform across Europe and North America, which marked a rupture with the immediate postwar period (see e.g. Kuhlmann and Wollmann, 2014; Randma-Liiv and Kickert, 2017). State–market relations were redefined towards ‘enterprising’ the state (Considine, 2001), with reform discourse focused on the notion of ‘efficiency’. These processes involved a shift away from traditional organisational principles of hierarchy and functional specialism towards new forms of governance that involved both a change in the way employees of public administration bodies were managed and the development of new relationships involving public and private sector agencies and those who consume their services, including an enhanced emphasis on performance management, entrepreneurial actions by public organisations, principal–agent separation, the development of quasi-markets and an emphasis on ‘citizen responsibility’ (Considine, 2001; Kuhlmann and Wollmann, 2014: 270).
These developments in public administration coincided with a widespread shift towards the adoption of supply-side employment measures that placed emphasis on ‘gainful employment as the principal channel to achieve effective citizenship’ (Ferrera and Hemerijck, 2003: 123) and political rhetoric that stressed the importance of workers enhancing their ‘employability’ through skill acquisition and work experience. This in turn influenced the delivery of employment services. A study of operational reforms in Europe (Van Berkel, 2009) identified five common trends: first, the creation of local one-stop agencies that provide job search assistance and administer benefits; second, decentralisation intended to encourage greater responsiveness of service provision to local and individual circumstances; third, outsourcing and the creation of quasi-markets in relation to activation services; fourth, efforts to diminish the involvement of employer organisations and trade unions in policy making or implementation; and finally, the use of targets and performance management systems. Efforts to strengthen the link between welfare entitlements and participation in the labour market have also led many European governments to locate responsibility for social protection and employment policy within a single ministry or agency. Examples include France (Pôle emploi), Ireland (the Intreo service), the UK, through the creation of Jobcentre Plus in 2001, the Czech Republic, where social benefit administration was added to the employment service functions of Labour Offices in 2012 (Lehmann and Kyzlinková, 2015: 8), and Germany, where administrative responsibility for unemployment assistance and social assistance payments was transferred to job centres in 2005 as part of the Hartz labour market reforms. The merging of employment and benefit services is thus evident in both LMEs and CMEs. These developments coincided with the imposition of stricter conditions on eligibility for unemployment benefit and stricter obligations for ‘job-seekers’. For instance, the Hartz reforms in Germany entailed also a reduction of long-term benefits, in terms of both amount and duration, as well as the replacement of means-tested flat-rate benefits with earnings-related unemployment assistance (Weishaupt, 2011). In France, the creation of Pôle emploi in 2009 coincided with a reduction in the duration of short-term benefits and an increase in the contributions requirement for long-term benefits (Coquet, 2015). Pôle emploi, Intreo and Jobcentre Plus are all examples of agencification of the state. While ‘agency fever’ is assumed to have originated in the UK (Kuhlmann and Wollmann, 2014) this trend has extended across Europe, and, as the French example suggests, is not confined to LMEs.
A further notable development in a number of countries were attempts to subordinate labour policy to economic policy, associated with reforms aimed at labour market liberalisation. Again, there were consequences for the organisation of labour administration. Some countries merged their economics and labour ministries into a single ministry. Examples include Hungary (1998–2002) and the UK. In 1995, the UK’s de facto Labour Ministry, the Department of Employment, was merged with the Department of Education and responsibility for industrial relations was transferred to the Department for Trade and Industry (now the Department for Business, Energy and Industrial Strategy), a ministry whose primary responsibility was promoting the competitiveness of British industry. Germany, the ‘ideal type’ of CME in the VoC framework, took a similar path in the early 2000s when a Ministry of Economics and Labour (Bundesministerium für Wirtschaft und Arbeit: BMWA) was created through a merger of the Federal Ministry for Labour and Social Affairs and the Ministry of the Economy. The creation of the BMWA, which lasted until 2005, eased the introduction of the liberalising reforms associated with Agenda 2010 and the Hartz Commission, in that it brought about a diminution in the voice of organised labour in policy debates at cabinet level and led to the orthodox economic analyses associated with the former Ministry of Economics becoming dominant within the BMWA (Heyes, 2004).
Both the spreading of responsibilities for labour issues across different ministries and the lumping together of such responsibilities with other policy areas into one ministry suggest a downgrading of labour issues. The VoC literature would expect such downgrading to occur in LMEs, but our sample indicates instead that this occurs in both LMEs and CMEs. The frequency with which ministries are renamed and responsibilities reshuffled and the confusion this creates places a significant strain on the implementation of labour policy. Innocuous as they may appear at first, such administrative reforms are indicative of deeper changes in the balance of social forces within a country. For example, the case of the German BMWA, mentioned above, represented the culmination of a process of ‘business mobilization for neoliberal reform’ (Fleckenstein and Lee, 2017: 166) in the face of globalisation pressures, which had started soon after German reunification. Against this backdrop of power resources shifting towards business, a scandal involving inflated job placement statistics in the German PES in 2002 provided the then-Chancellor Schröder with an excuse for excluding trade unions and social democratic ‘traditionalists’ from policy making, on account of them being ‘defenders of the status quo’ (Fleckenstein and Lee, 2017). Thus, the perceived failure of a labour administration body (the German PES) led to the sidelining of organised labour within German society, with repercussions for another labour administration body, the Ministry of Labour, which became subordinated to the Ministry of Economics in this process. The evolution of the Spanish Ministry of Equality is similarly best understood in a broader context of changing power relations. Created in 2008 as a standalone ministry, it was merged two years later with other ministries to form the Ministry of Health, Social Policy and Equality (later Ministry of Health, Social Services and Equality). This change coincided with the apogee of Troika power and pro-austerity discourse in Spain, which turned equality ‘into a luxury that could no longer be afforded’ (Campillo Poza, 2018: 21). The ministry was restored in 2018 as a result of a new change in the balance of social forces, this time in the form of a new socialist government.
A weakening of the voice of organised labour in labour administration has also occurred in many European countries. In the Netherlands the principle of tripartism in relation to PES was abandoned in 2002 (Hoogenboom, 2011), and the social partners have retained only an advisory role at the national level as members of the Social and Economic Council whose recommendations are non-binding for the government (Guardiancich and Molina, 2017). Denmark has followed a similar trajectory: the social partners’ active role at the regional level in governing regional PES offices ended with the creation of the Regional Employment Council in 2007 (Weishaupt, 2011), although they continue to play an advisory role at the national level. Even in Germany, where the social partners ostensibly still play an important role in the governance of PES, this role has been weakened since Schröder’s labour market deregulation reforms in the early 2000s. Since 2005 the social partners have not been involved in supporting the long-term unemployed, even in an advisory capacity (Weishaupt, 2011). Prior to that, unemployment assistance payments made to long-term unemployed persons after their unemployment benefits expired were administered by the BA, and hence indirectly by the social partners who make up the governing board of the BA. But following the Hartz reforms, responsibility for unemployment assistance payments (merged with social assistance payments) was transferred from the BA to newly-established job centres, in which the BA governing board plays no function and has no rights to information (Weishaupt, 2011). Within the LME category, social partner involvement in the governance of PES ceased in Ireland after 2007–2008 while in the UK employer organisations and trade unions ceased to be involved in the governance of PES in 1987, when the Conservative government dismantled the Manpower Services Commission (Weishaupt, 2011).
The developments discussed in this and the preceding section suggest convergence in labour administration in the sense of a gradual side-lining of labour issues in all countries, as evidenced by the limited prevalence of holistic labour ministries, the diminished role of social partners, and the high rate of adoption of NPM tools, instead of growing divergence between CMEs and LMEs as predicted by VoC scholars. The recent economic crisis appears to have further reinforced this trajectory. The effects of the crisis on labour administration are now discussed in more detail.
Developments in labour administration since the crisis
The crisis exposed some of the weaknesses of decentralised approaches to public administration, in particular the growing inability of governments ‘to monitor effectively and to regulate effectively’ (Peters et al., 2011). Consequently, the crisis is considered to have marked a turning point towards (re)centralisation. On the other hand, despite this new centralisation trend, other aspects of the pre-crisis model of public administration are shown to have in fact been reinforced post-2008, for instance in the form of a new wave of privatisations following the crisis, especially in Southern Europe, as part of Troika-imposed bailout conditions (Kuhlmann and Wollmann, 2014). As Randma-Liiv and Kickert (2017) argue, this suggests that the crisis had a complex effect upon administrative reform that goes beyond the path-reinforcing/path-breaking dichotomy: it triggered new reforms in some countries or policy areas, boosted reforms already underway in others, and cancelled previous reforms in yet others. 2 Centralisation along with fiscal retrenchment are listed as two main forms of response to the crisis common to all European countries (Randma-Liiv and Kickert, 2017). However, there is variation in terms of the extent of such reforms, which is closely linked with the extent of the crisis itself, more vulnerable countries being associated with more extensive reforms (Randma-Liiv and Savi, 2016; Raudla et al., 2015). For example, while budget cuts happened everywhere, they were most extensive in countries that required financial assistance, especially given that international creditors provided financial aid to these countries on condition that they implement structural reforms. Another development linked to vulnerability is that in the countries hardest hit by the crisis the legitimacy of social dialogue was called into question most extensively, with public opinion increasingly viewing the social partners as at least partially responsible for the crisis (Guardiancich and Molina, 2017). In what follows, we examine centralisation and fiscal retrenchment as responses to the crisis in relation to labour administration per se and assess also the impact of the crisis upon social dialogue as it pertains to labour market governance.
Centralisation
The recent trend towards centralisation can be explained by two related factors. The first is that centralisation helps to improve coordination and allows public administrators to better influence outcomes compared to a situation where the public sector is fragmented into myriad units (Raudla et al., 2015). The second is that centralisation can serve as a strategy for achieving spending cuts (Levine, 1979; Raudla et al., 2015), and thus fits in with the wider austerity drive since the crisis. Consequently, centralisation since the start of the crisis has tended to take two main forms: mergers of agencies or ministries, including reabsorption of agencies into their parent ministries, and a strengthening of central financial control over ministries and agencies, i.e. an increase in the power of finance ministries (Randma-Liiv and Kickert, 2017).
The survey evidence presented by Randma-Liiv and Kickert (2017) suggests that mergers of ministries and agencies have been the most widespread type of reform in European public administration since the crisis, occurring in LMEs (UK, Ireland, also East European liberal-dependent economies such as Hungary and Romania), as well as CMEs in their different variants (the Netherlands, Austria, France, Italy). Examples of this in the field of labour administration include the creation of the WRC in Ireland, National Labour Inspectorate (INL) in Italy, Pôle emploi in France, the DWP’s reabsorption of Jobcentre Plus in the UK, the merger between the Labour Inspectorate and the Social Inspectorate in Romania, and the aforementioned (see above) merger between the Ministry of Labour and the Ministry of Health in Austria. The Irish, Italian and Romanian cases are examples of centralisation of labour inspection services, bringing formerly fragmented services together under one central administrative body. The French and British cases represent examples of PES centralisation. While these mergers had ostensibly been designed to simplify and streamline labour administration processes while reducing costs, in many instances the resources allocated to the merged bodies have been insufficient, resulting in these institutions being understaffed and overburdened (De Comarmond, 2010).
The increase in the power of ministries of finance has also been an overarching aspect of centralisation across Europe (Raudla et al., 2015). This trend is evident even in countries such as Sweden or Norway, which had not been severely affected by the crisis (Randma-Liiv and Savi, 2016). This increased influence of finance ministries carries potentially profound implications for labour administration, particularly in terms of resource availability. For instance, PES budgets are generally decided as a result of consultations between the governing board of the PES, its sponsoring ministry and the Ministry of Finance (Heyes, 2011); increased influence of the last mentioned changes the balance of power in these negotiations and hence also the amount of resources allocated to the PES. Another obvious implication is that other ministries – including labour ministries – become less powerful.
Fiscal retrenchment and its impact on labour administration
As mentioned above, fiscal retrenchment was introduced across most of Europe as a response to the crisis, although the harshness of these measures tended to depend on how affected each country had been by the crisis, and on whether it had requested financial assistance from international institutions, which demanded structural reforms in exchange (Randma-Liiv and Savi, 2016). Cuts have been most extensive in the areas of welfare, pensions and healthcare (Randma-Liiv and Savi, 2016). A further consequence has been a reduction in the staffing and capacity of labour administration bodies. In a recent survey, 7000 public administrators across Europe were asked to rate the performance of public administration in their respective policy field over the previous five years. Employment policy received the lowest rating and was also the only policy field rated significantly below the mean (Huxley et al., 2016). The diminution of resources coincided with intensified challenges to labour administration as a result of the crisis, which significantly increased the workload of labour administrators (EPSU, 2012; Rychly, 2009a). In particular, increases in unemployment led to substantial increases in PES caseloads (the ratio of unemployed persons per employment advisor). Between August 2009 and August 2010, the caseload of Pôle emploi in France increased from 92 to 103 (De Comarmond, 2010). In Finland, the caseload doubled from 80 to 160 between the onset of the crisis and 2016 (OECD, 2016). Increases in PES caseloads have persisted in more recent years as well: for instance, between 2014 and 2016, caseloads increased by 23.9% in Finland, 19.5% in the Netherlands, 10.5% in Cyprus, 10% in Greece and 4.3% in Austria (European Commission, 2016b). Even where caseloads declined in recent years, high caseload numbers persist as a result of the crisis: for example, while Spain experienced a reduction in caseloads similar to that of Germany between 2014 and 2016 (8.3% and 8.6%, respectively), the outcome is a caseload of 596 in Spain versus 48 in Germany, with Spain remaining the country with the highest caseload in Europe (European Commission, 2016b). High caseloads make it more difficult for advisors to provide individualised support, calling into question the effectiveness of PES in minimising social hardship. Fiscal retrenchment also resulted in reduced training budgets for PES.
Since the start of the crisis, there has been a significant decline in the number of labour inspectors in European countries associated with different varieties of capitalism: the UK and Ireland as LMEs, Germany and the Netherlands as CMEs, Italy as a Mediterranean country, Romania and Latvia as Eastern European states, etc. (EPSU, 2012). In the UK, the HSE budget was cut by 35% in 2010, with an additional 28% cut to the budgets of local authorities responsible for occupational health and safety; this meant job losses for many frontline staff, including labour inspectors (EPSU, 2012). Ireland’s National Employment Rights Authority (the predecessor of the current WRC) had employed 84 inspectors; by late 2016, the WRC employed only 53 (Heyes and Hastings, 2017). Between 2013 and 2014, the number of labour inspectors employed by the Inspectorate SZW in the Netherlands fell from 1119 to 1024; this is expected to decline even further given that the inspectorate will see further cuts in its budget between 2016 and 2020 (Heyes and Hastings, 2017). In Germany the number of labour inspectors declined partly because of the non-replacement of retiring staff (EPSU, 2012). In Italy, the number of inspectors was reduced from 3761 in 2007 to 3479 in 2009 (EPSU, 2012) while training days for civil servants were reduced by 29.3% over 2008–2011 (DiMascio and Natalini, 2014). The number of inspectors employed by the Romanian labour inspectorate was reduced by half between 2009 and 2016 in 2009 (Heyes and Hastings, 2017), partly as a result of the merger between the Labour Inspectorate and the Social Inspectorate in 2009, and partly as a result of further restructurings in the Ministry of Labour (EPSU, 2012). The Latvian inspectorate experienced a 20% staff reduction in 2009 (Heyes and Hastings, 2017), which caused several regional offices to shut down as well (EPSU, 2012). Even in France, where the number of labour inspectors initially increased at the beginning of the crisis (from 488 inspectors and 1053 sub-inspectors in 2007 to 767 and 1423, respectively, in 2009 [EPSU, 2012]), this fell again after 2009, when all labour inspection services were merged and sub-inspector positions were eliminated altogether.
The fact that the crisis has made the work of labour inspectors more complex and demanding, but the number of inspectors went down in most countries, poses risks for the safety of workers, for it implies fewer inspection visits, a greater risk of accidents at work, laxer enforcement of safety laws, as well as fewer resources to tackle discrimination and undeclared work (EPSU, 2012). There are also concerns that insufficient staffing will make it difficult for individual countries to honour pledges made to the European Commission in the field of work and employment: for example, Romanian labour inspectors stated publicly in 2010 that it is becoming increasingly difficult to comply with the number of controls the inspectorate had pledged to undertake under a memorandum signed with the European Commission to tackle undeclared work (EPSU, 2012). The problem of understaffing of labour inspectorates is further compounded by the downgrading of labour issues in general following the crisis, manifest for instance in the placing of responsibilities for labour issues under other ministries: according to Eurofound (2015), the quality of labour inspection services has been steadily declining in Hungary in recent years, not least because such services had been placed under the supervision of ministries whose portfolios do not naturally encompass labour issues (Ministry of Finance).
The exclusion of social partners from labour administration
In most European countries and North America, the first crisis management phase consisted of neo-Keynesian demand stimulus policies. Countries with established institutions and traditions of social dialogue involved trade union and employer organisations in discussions concerning these policies, on the principle that cooperation would help reduce social tensions and conflict (Guardiancich and Molina, 2017; Rychly, 2009b). Negotiations between the government and social partners in 2008–2009 led to the adoption of the ‘Towards 2016’ agreement in Ireland which contained provisions vis-a-vis the rights of temporary workers (the first category of workers to be affected by the crisis), to the introduction of shortened work hours to avoid crisis-related dismissals in the Netherlands, and to the formulation of the tripartite ‘Declaration of principles for the promotion of the economy, employment, competitiveness and social progress’, containing anti-crisis measures, in Spain (Rychly, 2009b).
However, this was only a short-lived phase; the crisis soon began to be portrayed as one of fiscal profligacy and lack of competitiveness in peripheral Europe, and hence used as justification for fiscal consolidation measures. The implementation of austerity reduced the time-frames for consultation as well as room for manoeuvre in national policy making, severely curtailing or even annulling the participation of social partners in governance altogether. In countries where social dialogue was maintained during the crisis, this had a positive effect, facilitating growth, peaceful industrial relations and increases in productivity and wages (Rychly, 2009b); but this was the exception rather than the rule, with most countries in Europe experiencing exclusion of social partners from labour market governance instead (Guardiancich and Molina, 2017).
The greatest degree of discontinuity in social dialogue following the start of the crisis occurred in Ireland (an LME) and the Mediterranean countries (Spain, Portugal, Italy, Greece). The exclusion of social partners from governance is particularly striking in the case of Ireland and Spain, both of which had enjoyed strong social partnership prior to the crisis (Guardiancich and Molina, 2017). In Ireland, social dialogue continues only in the National Economic and Social Council (NESC), which has not been involved in any of the crisis or post-crisis reform negotiations (Guardiancich and Molina, 2017). Similarly, formal social partnership broke down in Spain in 2011, despite the fact that Spain had been one of the first countries to involve the social partners in designing policies to tackle the crisis back in 2008–2009 (Rychly, 2009b). In the absence of formal social partnership, the only forums left for social dialogue in Spain are bipartite consultations between trade unions and employers (Guardiancich and Molina, 2017). In spite of higher growth rates in both Ireland and Spain since 2014, there have been no attempts to restore institutionalised tripartite social partnership, and ‘no expectations’ of this in either the short-run or the long-run (Guardiancich and Molina, 2017). The fact that in Ireland ‘the trade unions, for the first time in over 20 years, have no direct access to government’ (Hardiman, 2010: 26) makes it increasingly similar to Britain, the other European country classified as an LME, which has no institutionalised role for the social partners.
CMEs may not have experienced a total breakdown in social dialogue as Ireland did, but they too have been facing strains in regard to social partnership. Germany and Denmark in particular have witnessed a decline in the influence of social partners over labour market governance (Weishaupt, 2011). As noted earlier, in Denmark this is a direct consequence of restructuring during the crisis, when regional PES offices governed by the social partners were dissolved, while in Germany it is a process that dates back to the Hartz reforms. The Dutch case also mirrors the German one, in the sense that social partnership had been weakened there due to labour market flexibilisation and a decline in unionisation that preceded the crisis, but which intensified tensions during the crisis, especially in consultations over pensions reform (Guardiancich and Molina, 2017). Due to these changes, the social partners in Denmark, Germany and the Netherlands alike acted more like advisors than proactive co-decision makers during the crisis (Guardiancich and Molina, 2017; Weishaupt, 2011). Weishaupt (2011) illustrated this distinction with reference to the Danish case. Prior to 2007, social partners in Denmark had been directly involved in the implementation of active labour market policies at the regional level, with substantial funds at their disposal, but lost their influence once the Regional Employment Council was created, and now mainly engage in ‘re-active criticism when Jobcentres fail to perform’ rather than ‘pro-active planning’ (Weishaupt, 2011: 18). Moreover, even as advisors, social partners exerted limited influence. Despite the fact that social dialogue is institutionalised at the national level in Denmark in the National Employment Council, the demands of Danish unions for fiscal stimulus and improving qualifications were largely ignored, with the government instead choosing to tighten the benefits regime as part of its flexicurity strategy (Weishaupt, 2011). On the other hand, in Germany social partners were consulted in an ad hoc manner during the economic crisis summits held at the Federal Chancellery and a formal institutionalisation of social dialogue at the national level was deemed unnecessary (Guardiancich and Molina, 2017). Austria is the only strong exception in this respect, as a CME that ‘has retained a modus operandi of strong social partnership and mutual trust’ (Weishaupt, 2011: vii). The UK and Austria are thus the only two countries in our sample characterised by continuity in social dialogue, albeit as polar opposites, with Austria continuing to exhibit a strong institutionalised role for social partners and the UK none.
Conclusion
This article has shown that differences and similarities in how labour administration systems are organised across countries and their evolution over time do not correspond to the different ‘varieties of capitalism’ outlined in comparative institutional analysis. Evidence from Europe on indicators covering the extent of (de)centralisation in labour administration, the degree of marketisation in labour administration, as well as the degree of involvement of social partners in labour governance, invalidates the central assumption of a clear CME–LME divide as pertains to labour administration, which is implicit in the VoC literature.
In terms of the impact of the recent economic crisis on labour administration across Europe as a common challenge, the path-reinforcing/path-breaking theoretical divide is likewise of little explanatory value. The crisis appears to have reinforced trends towards marketisation in the form of budget cuts, placing further strain on already overburdened labour inspectorates and PES offices, whose resources have been slashed while challenges have increased. It has, on the other hand, reversed other trajectories, as evidenced most notably by the recent turn away from decentralisation to centralisation, with mergers between ministries and agencies as the most prevalent type of reform in labour administration since the crisis. The gradual side-lining of labour issues appears to be an element of convergence across Europe, but the timing of particular stages in this process varies: for example, the merger between employment and benefit services, accompanied by stricter eligibility criteria, occurred before the crisis in some countries (e.g. Germany, the UK), and during the crisis in others (e.g. Ireland, France). The same is true of the dissolution of holistic labour ministries or of the exclusion of social partners from labour governance – such trends have been precipitated by the crisis in some countries, but date back to earlier restructurings of the state–market relationship in others. These (radical, in some countries) reforms of national systems of labour administration over the past two decades have reflected wider changes in public administration systems, which have involved the adoption of new management practices and organisational forms and a reconsideration of the interface between the public and private sectors and the role of the latter in delivering services.
Reality is therefore more complex than the models envisioned by VoC theorists; indeed, it seems to invalidate the idea of coherent and continuous models altogether. In this context, the very concept of a national labour administration system becomes problematic, as it implies coherence and stability. Comparative research in labour administration would thus benefit from transcending the VoC framework, not least also because labour administration represents an arm of the state, and the state itself is under-theorised in VoC. A new research agenda would need to move beyond a strictly comparative analysis to one that is also relational; as increases in the power of the Ministry of Finance vis-a-vis those of Labour, under the pressure of external conditionality, show, interactions between countries (in this case, the wider EU policy framework under which national policies in individual member states operate) also matter. Labour administration would be best understood by using power and conflict as theoretical anchors, reflective of ‘the wider social relations in which [labour] institutions sit and are constituted by’ (Bruff and Ebenau, 2014: 3). Doing this would allow for a ‘double comparative perspective’ (Bieling, 2014), helping to explain differences in labour administration between states, as well as within states through time, on the other hand accounting also for commonalities in capitalism. It would also aid our understanding of the effects of the crisis beyond the path-breaking/path-reinforcing dichotomy, by treating the crisis as endogenous, with power and conflict as triggers of dynamism in capitalism, but also as potential sources of path dependency, in the sense of policy conditionality that arises from the way in which past conflicts had been resolved in a particular space.
Footnotes
Acknowledgements
We are very grateful to the journal’s referees for providing very helpful comments on an earlier version of this article. We are responsible for any errors that remain.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
