The digitalisation of service work – social partner responses in Denmark,Sweden and Germany

Abstract

German

French

The discussion on the digitalisation of work has intensified in recent years. The literature points to two main trends accelerated by digitalisation – work automation that eliminates or changes job functions, and the creation of work without jobs via digital platforms. This article addresses the question as to how social partners define digitalisation of work and their perception of its consequences, while also looking at their recent responses to it. Drawing on interviews with unions and employers' organisations in Denmark, Sweden and Germany, it examines social partner initiatives in the unilateral, tripartite and bipartite arenas on various forms of neo-corporatist labour market regulation. The focus is on service work in the private sector, an area of the economy currently under pressure from both automation and the trend towards work without jobs. Whereas the social partners seem to be very active in the unilateral arena in all three countries, responses differ in the tripartite and bipartite arenas. The article concludes by discussing the strengths and weaknesses of the responses in the face of current digitalisation trends and existing models of labour market regulation.

Keywords

Digitalisation service work private sector social dialogue Denmark Sweden Germany

Digitalisation and work flexibility

Digitalisation of work and its consequences for labour markets and working conditions have been debated for many years in the media, politics and research, with discussions intensifying in Europe and the rest of the Western world in the 2010s. The European Commission has launched various digital agendas as part of its Digital Single Market Strategy (European Commission, 2014, 2016), and books and articles have addressed how digitalisation is now accelerating job losses and changes of work processes (Ford, 2015; Gray et al., 2016; Hill, 2015; Huws, 2014; Rifkin, 2014).

The literature highlights two main trends. First and foremost, digitalisation is accelerating work automation (Ford, 2015; Frey and Osborne, 2013), a process dating back to the early days of industrialisation. However, the invention of the computer and later the Internet have contributed to accelerating work automation in the form of increasingly sophisticated robots and software. Its pace and the resultant job losses are expected to be further intensified by the increasing use of Big Data in various professions.

Estimates point to automation eliminating many jobs in manufacturing but even more jobs in the service industries such as sales assistants and secretaries. Calculations based on job categories speak of a potential 47 per cent cut in the US workforce due to automation (Frey and Osborne, 2013). However, calculations based on skills and job functions are less pessimistic for both the US and Europe, underlining that specific parts of each job will become obsolete, but not the entire job (Arntz et al., 2016; Chui et al., 2015, 2016). Automation is not only eliminating certain jobs but is also changing the content of most jobs. The remaining jobs are likely to be characterised by new functions, competence requirements and workplace reorganisation. Furthermore, automation is helping to create new types of jobs such as e-commerce software development, etc. In sum, automation in both its early and later forms alters the functional flexibility of work – which job functions are needed or not and which combinations of job functions make a job (Benner, 2002).

Secondly, and more recently, digitalisation is creating work without jobs – work bought and sold on-demand as single assignments and without longer commitments or contracts, the so-called ‘gigs’ (Hill, 2015; Huws, 2014; Rifkin, 2014). Over the past decades, non-standard employment has grown in many Western countries, meaning that the standard full-time job on an open-ended contract is under pressure (Gautié and Schmitt, 2009). Developments include temporary contracts, zero-hour contracts, agency work, informal work, self-employment, freelance jobs, etc. (ILO, 2016; Manyika et al., 2016). In countries like the UK and the US, the growth in non-standard work has been associated with a more general trend of increasing work precariousness (Kalleberg, 2011; Standing, 2011), whereas in other countries such as Germany non-standard employment has been associated with a dualisation of the labour market distinguishing between standard and non-standard jobs (Palier and Thelen, 2010; Schulten and Buschoff, 2015). Digitalisation is contributing to this trend by allowing more work to be performed through digital platforms without the legal framework of a job and without a clear legal employer/employee identity.

It is important to distinguish between two different types of platforms. One group can be characterised as labour platforms, i.e. forums for discrete freelance tasks (Farrell and Greig, 2016) such as Upwork.com or Freelancer.com, where single work assignments or projects are traded. This category also includes micro-work forums such as Uber.com or TaskRabbit.com, where small tasks are bought and sold. Other digital platforms can be characterised as capital platforms. Here, participants sell goods or rent assets. Examples are Airbnb.com and Ebay.com (Farrell and Greig, 2016). The latter type does not directly form part of the labour market. Certain platforms were originally associated with the sharing economy (Schor, 2014). However, experience tells us that even platforms originally created with the intention of ‘sharing’ are now often used for buying and selling services. A prominent example of this development is Airbnb.com (Hill, 2015). Nevertheless, both labour and capital platforms offer new forms of economic activity outside the regulatory framework governing the labour market and welfare state, a framework built on the concept of full-time permanent jobs (Koch and Fritz, 2013).

The size of the platform economy is still relatively modest. Recent analysis of US banking data shows that, in any given month, 1 per cent of adults earn income from the online platform economy (Farrell and Greig, 2016). Although labour platforms are growing more rapidly than capital platforms, the capital platform market is larger: 0.6 per cent of adults received income from capital platforms, whereas 0.4 per cent of adults received income from labour platforms (Farrell and Greig, 2016). The limited use of labour platforms is also confirmed by analysis of US survey data (Katz and Krueger, 2016). Importantly, most participants used the platforms as a secondary source of income. We find a similar picture in Europe. Surveys in Sweden, the UK, the Netherlands and Austria indicate that most people use the platforms to supplement other forms of income (Huws and Joyce, 2016a, 2016b, 2016c, 2016d). Platforms rarely account for an individual’s full income today, but this segment is a growing market and expected to develop exponentially in the coming years. Therefore, digital platforms seem to contribute to the increasing and increasingly diverse forms of numerical flexibility of work – i.e. how many workers are needed and how these are employed (if at all employed) –, even though the platform economy is still just a small part of the total economy (Bernhardt, 2016; Bernhardt et al., 2016; Stone, 2004).

Generations of literature on workplace flexibility have drawn upon Atkinson’s famous concept of the flexible firm and its core and peripheral workers (Atkinson, 1984). Under it, workers with high skill levels are able to offer the firm a functional flexibility shielding them from numerical flexibility. This trade-off makes them core workers – i.e. functional flexibility is rewarded with job security. The other side of the coin is that workers with low skill levels have to deliver the numerical flexibility needed. These become peripheral workers with just a loose attachment to the firm (hiring/firing, non-standard terms, etc.). Digitalisation has the potential to alter these traditional trade-offs, accelerating both functional flexibility (the destruction and creation of tasks) and numerical flexibility (non-standard employment; work without jobs) and thereby increasing the risk of dual flexibility pressure. Core workers can easily become peripheral and vice-versa. This raises a number of important questions for the social partners – defined as trade unions, employers' organisations and the state – as to the future regulation of work and employment. If jobs disappear on a massive scale in certain areas of the economy and/or are replaced by robots, software or gigs, educational systems and labour market policies will be forced to adapt.

In this article, I examine how social partners are defining the digitalisation of work and perceiving its consequences, while also looking at their recent responses to it. Drawing on interviews with unions and employers' organisations in three European countries, I investigate initiatives in three arenas: the unilateral, tripartite and bipartite. The focus is on private sector services, the private sector currently being the most affected by both automation, and the trend towards work without jobs. Denmark, Sweden and Germany are compared, exploring similarities and differences in social partner responses. All three countries represent neo-corporatist models of labour market regulation, though the use of legislation as a regulatory instrument varies, being used more in Germany than in Sweden, and more in Sweden than in Denmark.

The article is structured as follows. After this introductory section on work digitalisation and flexibility, I briefly present the labour market models of Denmark, Sweden and Germany and possible political arenas of social partner responses. This section is inspired by the neo-corporatist literature. I go on to present the methods used in my analysis and my findings. I conclude with a discussion of the findings.

Neo-corporatism and social partner responses in Denmark, Sweden and Germany

Responses to digitalisation of work in Denmark, Sweden and Germany occur in labour market models characterised by coordination and cooperation between the social partners. Danish, Swedish and German labour market regulation has traditionally been described as a form of neo-corporatism where social partners cooperate and negotiate a regulatory mix made up of collective agreements, legislation and tripartite initiatives (Crouch, 1993; Visser, 1996). The different types of labour market regulation supplement each other, demonstrating a tradition of coordination covering different stakeholders and groups of interest.

However, there are also important country differences. Denmark and Sweden continue to demonstrate comparatively high union densities and employer organisation membership rates (albeit with a minor erosion), whilst these figures have dropped significantly in Germany (Visser, 2007, 2015) (see Table 1). Nevertheless, private sector services are under pressure when it comes to union density in all three countries – also in Sweden and Denmark (Bechter et al., 2012; Kjellberg and Ibsen, 2016). In Denmark, union density has dropped to a third in such service segments as retail, hotels/restaurants and office work, and this is also the case inter alia in hotels/restaurants in Sweden (Kjellberg, 2013; Toubøl et al., 2015). This background information is important when it comes to social partner initiatives in neo-corporatist models. Bipartite and tripartite initiatives depend on the legitimacy of the organisations participating in the coordination and/or negotiation. Declining membership rates put a question-mark over the legitimacy of an organisation’s bargaining mandate.

Table 1.

Union densities and membership rates of employers' organisations in Denmark, Sweden and Germany.

	Denmark	Sweden	Germany
Union density (% of employees)^a	67	67	18
% of employees working in companies belonging to an employers' organisation^b	68	82	58

^a Figures from 2012–2013.

^b Figures from 2010–2011.

Source: J Visser, ICTWSS Database. Version 5.0. Amsterdam: Amsterdam Institute for Advanced Labour Studies AIAS. October 2015.

Legislation has historically played a larger role in German labour market regulation than in Sweden and Denmark (Andersen, 2001; Ilsøe et al., 2007). While Germany recently introduced a statutory minimum wage, Denmark and Sweden are two of the few European countries still without one (Dølvik, 2016; Schulten, 2016). The introduction of the German minimum wage was a response to the growing area of the German economy not covered by collective bargaining. Bipartite regulation – i.e. collective bargaining – plays a dominant role in Sweden and Denmark in combination with tripartite coordination of the unemployment benefit funds, further training (CVET) and labour market policies (Dølvik, 2016; Mailand, 2008). Collective agreement coverage in Sweden and Denmark is still comparatively high, and social partners (especially in Denmark) have been hesitant to introduce more labour market legislation (Dølvik, 2016). However, employment protection legislation (EPL) is slightly stricter in Sweden than in Denmark, leaving Denmark as the least legalistic of the three countries (Andersen et al., 2014).

Considering the key role of unions and employers' organisations in Danish, Swedish and German labour market regulation, this article will focus on these two players and their responses to the digitalisation of work. Neo-corporatism theories distinguish between three different political arenas where social partners can seek influence: the unilateral arena, the tripartite arena and the bipartite arena (Ebbinghaus, 2002; Mailand, 2008). I will use these three arenas to analyse union and employer responses in Denmark, Sweden and Germany.

The unilateral arena is characterised by government regulation. In many ways this is the arena with the weakest influence of unions and employers' organisations. The state rules by legislation and policy. Unions and employers' organisations can try to influence policy and legislative processes at national level (or at EU level) through lobbying initiatives (Mailand, 2008). They can engage the media, publish and present analyses and reports, or leverage their formal and informal interactions with politicians and government officials. Furthermore, they can take solo initiatives and offer services to their members to cope with challenges experienced with existing regulations.

The influence of unions and employers' organisations is potentially larger in the tripartite arena (Mailand, 2008) where unions and employers' organisations collaborate with the government. Here, the three players coordinate or cooperate on policies and revisions of legislation in existing formal forums or negotiate tripartite agreements. These formal forums exist at both national and local level. Sometimes this arena will include other players and organisations with an interest or competence in the topic under discussion.

Unions and employers' organisations have the highest level of influence on labour market regulations in the bipartite arena (Mailand, 2008), conducting self-regulatory collective bargaining and concluding and implementing collective agreements. Negotiation and collaboration take place at both national/sector level and local level. However, influence in the bipartite arena is very much dependent on the bargaining power of unions and employers' organisations. If membership rates are dropping, this will affect collective agreement coverage quantitatively and qualitatively. High coverage and the efficient implementation of agreements are important factors when governments evaluate self-regulation via collective agreements.

The choice of arena has an impact on the degree of influence and the content, form and implementation of any regulation. When responding to the digitalisation of work, unions and employers' organisations in Denmark, Sweden and Germany potentially have the highest degree of influence in the bipartite arena, via for instance collective agreements. This is especially true for Denmark and Sweden, where bipartite regulation plays an even more dominant role than in Germany. Considering the declining union density in large parts of private sector services in all three countries, this might, however, not be their first choice. In the case of digital platforms which neither consider themselves to be employers nor are members of employers’ organisations, it can be also difficult for unions to find a bargaining partner. This may lead the social partners to give priority to government initiatives – commission work, legislation, etc. – and work in the tripartite and unilateral arena. In the unilateral arena, organisations can for instance influence legislative processes, whereas the tripartite arena can be important with regard to overall policy-making. In many cases, organisations make use of more than one arena to influence future regulation. Forum shopping is a known phenomenon in neo-corporatist models – if organisations cannot gain influence in one arena, they try the others until they get their message over (Due and Madsen, 2009).

Methodology

This article is based on desk research on the digitalisation of service work, background interviews with researchers working on the topic, followed by interviews with unions and employers' organisations on their responses to digitalisation. All empirical data were collected during 2016 and included a total of 14 interviews. I started by doing desk research to formulate this article’s main research questions. I then conducted four background interviews with researchers in the UK and the US working on digitalisation of service work. These background interviews were used to design semi-structured interview guides for social partners in the service sector in Denmark, Sweden and Germany. Six interviewees from the social partners in the three countries were carefully selected using the following criteria. First, they had to be at similar levels (sector level in the case of private sector services) to facilitate a valid comparison of answers across unions and employers' organisations and across countries. Secondly, I used an information-oriented selection strategy (Flyvbjerg, 1996), selecting the sector-level representative with the greatest knowledge and experience within the field of digitalisation of work. This resulted in two interviews at federation level in Denmark and two at EU level to identify the relevant interviewees. Finally, I interviewed the sector-level representatives. Each interview lasted 1½–2 hours and was recorded. The results of each interview were then condensed to summarise the conclusions. The analysis addressed the following three research questions:

How do unions and employers' organisations in private sector services define digitalisation of work? Do they include both the automation and the work without jobs debate?

What are – in their view – the consequences of digitalisation of service work?

Which social partner initiatives have they been part of? Do initiatives have a unilateral, tripartite or bipartite character?

Answers were coded according to these three questions in all six sector-level interviews and displayed in four tables, which were then sent to the interviewees for comments and revised accordingly. Although the representatives from the sector-level organisations in all three countries answered all questions from a service sector perspective, it quickly became clear that some of the initiatives were not restricted to private sector services, as was the case for some forms of commission work. Furthermore, the answers from the unions and the employers' organisation in each country were not always identical in the bipartite and tripartite arena. Sometimes, one side of the table would mention certain initiatives not mentioned by the other side. Such initiatives were nevertheless included in the analysis, as the two sides did not always participate in the same initiatives (for instance company-level agreements, certain forms of commissions at federation level, etc.).

Analysis

In this section the results of our analysis are presented. We start by looking at how social partners define digitalisation of work and its consequences in each country – and the differences and similarities across the two sides of industry. These analyses are then compared across the three countries. Next, social partner initiatives in three political arenas – unilateral, tripartite and bipartite – are analysed in each country, with the findings again compared.

Digitalisation of work – definitions and consequences

Union and employer representatives seem to define digitalisation quite similarly in each country – and across the three countries. However, when asked about their views on the consequences of digitalisation, the two sides of industry in all three countries differed.

When asked how they would define digitalisation of work, both the union and the employer representative in Denmark quickly pointed to the two main trends mentioned in the literature – how digitalisation leads to work automation and creates work without jobs. However, they saw these trends from different perspectives. The union representative underlined how automation led to job losses and demands for new competences and further training, whereas the employer representative stressed the importance of ensuring an attractive business environment in Denmark to attract economic activity (see Table 2). This also included providing employees with new competences. When it came to the creation of work without jobs, the union representative was concerned with how decent wages and working conditions could be guaranteed, whereas the employer representative was concerned with how to get rid of (or avoid) excessive regulation and welcome new business models to Denmark.

Table 2.

Definition and consequences of digitalisation of service work – responses from unions and employers' organisations in Denmark, Sweden and Germany.

	Denmark		Sweden		Germany
	Union	Employers' organisation	Union	Employers' organisation	Union	Employers' organisation
Digitalisation accelerates work automation	Job losses, new competences	Securing economic activity, new competences	Job losses, reskilling	Eliminates tasks – not jobs	Job losses, new qualification requirements	Job losses, evolution
Digitalisation creates work without jobs	Securing pay and working conditions	Welcome new business models	New ways of organising work – how to unionise and create portable benefit systems?	New forms of aggregating tasks – more employment opportunities, regulation challenges	Freelance work, no employer, no workplace and no employee, digital precariat	Test cases, unfair competition
Digitalisation facilitates control and surveillance*	-	-	-	-	Digitalisation facilitates employer control and calculation of worker tasks	Digitalisation questions data security and privacy for firms (SMEs)

*The topic of how digitalisation facilitates control and surveillance was mentioned by both the union and the employers' organisation in Germany, whereas this was not the case in Denmark and Sweden.

Although the two parties defined digitalisation of work in a similar manner, they had differing views on its consequences. Whereas the Danish union looked at them through the lens of labour market policy (how to regulate wages and working conditions), the employers' organisation viewed them through the lens of business policy (how to attract and retain economic activity).

In Sweden, both the union and the employer representative were also quick to address both issues – how digitalisation leads to work automation and creates work without jobs. However, they did not disagree as much as in Denmark over its consequences. The union representative mentioned (as did his Danish colleague) job losses and the need for reskilling, whereas the employer representative emphasised that automation would eliminate tasks – not jobs (see Table 2). Therefore, he expected overall job losses to be lower than anticipated. Furthermore, the Swedish employer representative stressed that the creation of work without jobs via online platforms was a new form of aggregating tasks that actually created more employment opportunities. However, such platforms also produced regulatory challenges. The Swedish union representative was less optimistic about these new forms of work organisation and their implications. How can unions organise workers performing tasks via platforms? How can these workers participate in the unemployment benefit systems? Is it possible to create a portable benefit system? As in Denmark, the Swedish employer representative was more concerned with job creation, while his union counterpart was more concerned with job destruction. However, both sides viewed digitalisation as an important issue in labour market politics.

Last but not least, the respective representatives in Germany also addressed how digitalisation leads to work automation and creates work without jobs. However, both representatives also mentioned a third aspect: how digitalisation facilitates control and surveillance (see Table 2). This may reflect historical reasons – since the Second World War, Germany has been very cautious about data privacy. However, they raised this topic for different reasons. The German union representative feared that digitalisation of work would lead to more employer control and to the calculation of a monetary value for every individual task, including small breaks during work. Digitalisation allows employers to record everything by the second, facilitating a detailed control of worker behaviour. The employer representative feared that digitalisation would lead to a loss of data security and privacy – especially for SMEs who cannot afford to buy the competence needed to secure their data. In conclusion, both sides feared the potential consequences of data recording facilitated by digitalisation.

With regard to work automation, the answers from the German representatives were very similar to those of their Swedish and Danish colleagues. The German union representative mentioned job losses and new qualification requirements as a consequence of automation, whereas the employer representative accepted that there would be job losses but that this would happen slowly and incrementally. Again, he was less pessimistic than his union counterpart. As regarded work without jobs, the German representatives were very similar in their views, in contrast to their Nordic colleagues. The German union representative assessed the situation rather pessimistically, fearing that digitalisation would lead to the creation of a freelance labour market with no employers, no employees and no workplaces. In his view, there was a real danger that this would lead to a digital precariat. The employer representative viewed the current platforms as test cases to be addressed and regulated over time. However, he considered any expansion of the platform economy as a challenge, as this could trigger unfair competition between digital and physical firms.

In sum, union and employer representatives in Denmark, Sweden and Germany defined digitalisation of work in rather similar ways. All interviewees mentioned how digitalisation accelerated work automation and how digital platforms created work without jobs. The German interviewees added a third dimension, highlighting how digitalisation also facilitates the increased control and surveillance of employees and of firms. Focusing on the consequences, there was disagreement among the respective representatives in all three countries. Whereas the German, Danish and Swedish union representatives mainly perceived the consequences of digitalisation as a labour market policy issue, fearing job losses and increasingly precarious wages and working conditions, the employer representatives mainly considered the consequences as a business policy issue, fearing that excessive regulation would lead to fewer jobs and less economic activity. This difference was most visible in the Danish case, whereas the differences were more moderate in the Swedish and German ones.

Social partner initiatives on the digitalisation of work – unilateral, tripartite and/or bipartite responses?

Our attention now turns to the three possible political regulatory arenas in neo-corporatist labour market models. Here, the analysis focuses on the interviewees’ answers to the question of which initiatives their organisation had participated in. In general, the unions and employers' organisations in all three countries seem to have been very active in the unilateral arena, whereas activities in the tripartite and the bipartite arena differ.

In Denmark, union and employer initiatives on digitalisation of work have been mainly in the unilateral arena (see Table 3). The two have published analyses/reports, commented in the media, conducted political projects, initiated dialogues with government departments and political parties, and participated in dialogues in European organisations/forums. Furthermore, the union reported having responded to EU strategies on the topic. It had also created a new pension scheme for union members not covered by a collective agreement (including freelancers) and a freelancer network. Initiatives in the tripartite arena were limited at the time of writing. Unions and employers' organisations had exploited their existing tripartite cooperation on education and further training to discuss and create further training opportunities within the field of digitalisation. Unions had organised a conference on the platform economy, and the government had initiated a strategy writing process on the platform economy. However, the latter had been a relatively closed and delayed process. Other tripartite initiatives were scheduled for 2017. Initiatives in the bipartite arena were even more limited and mainly informal. Unions and employers' organisations had had informal talks with each other, and the union had contacted new digital employers for informal talks.

Table 3.

Social partner initiatives on digitalisation of work in Denmark – unilateral, tripartite and bipartite arenas in private sector services compared.

Denmark	Union	Employers' organisation
Unilateral arena	Analyses/reports, media appearance Political project Dialogue with government departments and political parties Responses to EU strategies Pension scheme for union members – including freelancers; freelancer network Dialogue with European and international unions	Analyses/reports, media appearance Political project Dialogue with government departments and political parties Dialogue with European forums and employers' organisations
Tripartite arena	Tripartite cooperation on education and further training Roundtable at Copenhagen municipality Government-led strategy process on sharing economy (open ended) Union-led conference on platform economy Union-led expert panel on platform economy scheduled for 2017 Government-led disruption panel scheduled for 2017	Tripartite cooperation on education and further training Roundtable at Copenhagen municipality Company forum Debate at Denmark’s Political Festival on Bornholm Tripartite meetings on unemployment benefits Tripartite negotiations on further training scheduled for 2017
Bipartite arena	Informal contacts to employers' organisations Contacts to new digital employers	Informal contacts to unions

In Sweden – as in Denmark – there have been a number of initiatives on digitalisation of work triggered by the union and the employers' organisation in the unilateral arena (see Table 4). The Swedish union has also (together with other white-collar unions) participated in an attempt to build a private unemployment insurance for white-collar workers, able to deliver portable benefits. Unlike Denmark, there have been several initiatives in the tripartite arena. The Swedish government has set up at least eight commissions on digitalisation of work: a digitalisation commission, a taxi commission, a commission on workplace safety in the new economy, a commission on future work, etc. Furthermore, many new tripartite initiatives on digital platforms, lifelong learning, etc. are in the making. As in Denmark, bipartite initiatives in Sweden are rare.

Table 4.

Social partner initiatives on digitalisation of work in Sweden – unilateral, tripartite and bipartite arenas in private sector services compared.

Sweden	Union	Employers' organisation
Unilateral arena	Analyses/reports, media appearance Dialogue with government departments and political parties Dialogue with European and international unions Responses to EU strategies Website on automation Attempt to build a private unemployment insurance for white-collar workers	Analyses/reports, media appearance Think tank which deals with the issues of education in a digital context Dialogue with government departments and political parties Dialogue with European forums and employers' organisations
Tripartite arena	Digitalisation commission (2013) Taxi commission (2015) Workplace safety in the new economy (2015) + five more commissions since 2015 Many new tripartite initiatives in the making on platforms, lifelong learning and social security	Commission on future work (Arbetet i framtiden)
Bipartite arena	Informal contacts with employers' organisations Contact new digital employers	Informal contacts with unions

In Germany – as in Denmark and Sweden – there have also been several unilateral initiatives on digitalisation of work by the union and the employers' organisation (see Table 5). However, there have been more bipartite initiatives. Alongside informal dialogue between unions and employers' organisations on the topic, collective agreements have also been concluded, most at larger companies within the framework of ‘Collective agreements to protect employees from redundancies as a result of technical progress and automation’ (Rationalisieringsschutzverträge). The union recently concluded an agreement with Telekom addressing job losses and reskilling needs as a consequence of digitalisation. The German union and employers’ organisation have been very active in the tripartite arena. As in Sweden, the German government has set up several commissions on digitalisation (at least eight, and including Arbeiten 4.0 and Platform competitiveness) made up of government representatives, unions, employers' organisations and other stakeholders. It has also launched a large research fund entitled ‘Innovations for the production, services and labour of tomorrow’ that supports research into digitalisation of work.

Table 5.

Social partner initiatives on digitalisation of work in Germany – unilateral, tripartite and bipartite arenas in private sector services compared.

Germany	Union	Employers' organisation
Unilateral arena	Analyses/reports, media appearance Dialogue with government departments and political parties Dialogue with European and international unions Responses to EU strategies (Digital Single Market Strategy, Refit strategy)	Analyses/reports, media appearance Dialogue with government departments and political parties Dialogue with European forums and employers' organisations Internal political project: Digitalisation Responses to EU strategies (Services Directive) Participation in High Level Group
Tripartite arena	Federal Ministry of Labour: Arbeiten 4.0 (Green Paper 2015/White Paper 2016) Ministry of Economic Affairs: Platform competitiveness (Green Paper/White Paper) + six other government- initiated procedures with tripartite elements Ministry of Education and Science: Innovations for the production, services and labour of tomorrow (research funding)	Participation in several commissions initiated by ministries Conference 2015: Digital services
Bipartite arena	Informal contacts to employers' organisations Collective agreements to protect employees from redundancies as a result of technical progress and automation (Rationalisieringsschutzverträge) Telekom agreement (2015): early retirement, reskilling.	Informal contacts to unions

In conclusion, both unions and employers' organisations in all three countries seem to have resorted primarily to the unilateral arena when it comes to overcoming the challenges posed by digitalisation of work. However, the Danish and the Swedish unions differ from the German union in the unilateral arena, as they have tried to create concrete benefits for freelancers and other non-standard workers not covered by a collective agreement. The Danish union has successfully created a favourable pension scheme for union members (not covered by collective agreements), whereas the Swedish union unsuccessfully attempted to create a private unemployment insurance. The tripartite arena has been used very differently in the three countries. While the Danish social partners have discussed digitalisation of work more or less informally in the existing forums for tripartite cooperation on education and further training as well as in a couple of new forums and roundtable discussions, in Sweden and Germany, the tripartite approach has been much more formal and substantial. Both the Swedish and the German governments have set up at least eight commissions on digitalisation of work in which the unions and employers' organisations have participated. On top of that, the German government has also launched a large research fund on the theme. The bipartite arena has been used the least in all three countries. Most initiatives here are little more than informal dialogues. However, the German union stands out, having negotiated collective agreements at company level addressing the challenges of work automation, primarily in larger companies.

Conclusion and discussion

The discussion on digitalisation of work has intensified in recent years. The literature points to two main trends accelerated by digitalisation – work automation which eliminates or changes job functions and the creation of work without jobs via digital platforms. These trends are especially strong in private sector services. This article raises the question as to how social partners in such work define digitalisation of work and its consequences and what they are doing to overcome the challenges posed. Drawing on interviews with unions and employers' organisations at sector level in private sector services in three European countries, our analysis explores social partner initiatives in the unilateral, tripartite and bipartite arenas.

Our analysis suggests that the unions and employers' organisations in all three countries agree on how to define digitalisation and what the important trends are (automation and work without jobs). The German social partners are also concerned about how digitalisation of work facilitates control and surveillance. However, the social partners differ when it comes to how to address these trends. Whereas the union representatives mainly view digitalisation as a labour market policy issue, fearing job losses and precarious wages and working conditions, the employer representatives mainly see digitalisation as a business policy issue, fearing that excessive regulation will lead to a decline in jobs and other forms of economic activity. This difference is most visible in the Danish case, whereas the difference is more moderate in the Swedish and the German case.

The varying views on digitalisation between unions and employers' organisations might relate to their respective use of political arenas. In all three countries, both sides of industry boast many initiatives on digitalisation of work in the unilateral arena. In Denmark, there are few initiatives in the tripartite and bipartite arenas, meaning that the unilateral arena plays the dominant role. However, it is also the arena where the influence of unions and employers' organisations is potentially weakest. Moreover, in this arena, the two sides do not meet directly to create common perspectives and policies. This can help explain why the Danish union and employers' organisation address the consequences of digitalisation from very different perspectives – and disagree more than their Swedish and German colleagues. They have not to the same degree been forced to engage in social dialogue on the topic in tripartite forums, allowing common concepts and perceptions to be developed. Government-initiated tripartite initiatives might be especially important in private sector services in Denmark, since union densities are lower than in other parts of the private sector and most companies are small or medium-sized. The bargaining mandate for unions is thus relatively weak and bipartite initiatives not easy to establish.

The Swedish and the German governments have set up several large commissions on the digitalisation of work. This might be a consequence of the larger extent of labour market legislation in the two countries. The commissions allow unions and employers' organisations to combine their unilateral initiatives with coordinated ones in the tripartite arena, giving them greater potential to create a common understanding of digitalisation of service work and influence future legislation thereon. However, one may question whether all these commissions have had any significant impact in practice. Have their recommendations been implemented at workplace level, and what are their effects? In view of declining union densities in private sector services in Sweden and Germany, the bargaining power of unions to implement recommendations via collective agreement is limited. Enforcing legislation can also be a challenge. This is especially true with regard to work via digital platforms when such work is not registered by the public authorities in one way or another. This brings us back to the German discussion on data control – who owns and has access to data? This will be a decisive aspect in any future regulation of digital work. In addition to this, one can question whether the organisations and the government have the capacity to process the data. It does not help to have data access if data cannot be analysed.

Our analysis suggests that only Germany has made formal use of the bipartite arena in the field of digitalisation, with collective agreements on the consequences of digitalisation concluded at enterprise level. Examples mainly cover larger, unionised companies like Telekom. The question is of course whether bipartite initiatives will work also for SMEs which make up the majority of private sector services in Germany, but are often less unionised.

In general, the unilateral arena seems to be the main arena for initiatives on digitalisation of work by unions and employers' organisations in all three countries. Tripartite coordination in Sweden and Germany is initiated by the governments – unions and employers' organisations are participants. As yet, bipartite coordination only seems to function in Germany in larger companies. These conclusions might reflect the declining union densities in private sector services in all three countries. Union density is lowest in German private sector services – with the exception of certain large companies –, though significant drops in union densities have also been observed in Danish and Swedish private sector services over the last decade. Faced with weakened bargaining mandates and resources, the unions might prioritise lobbying initiatives to influence legislative processes, as well as accepting invitations to participate in commissions.

One challenge facing the Danish and Swedish models of labour market regulation is that the bipartite arena has not been used more, particularly when considering that the two are mainly regulated by collective agreements. This is especially true for Denmark, where no government-led commission on digitalisation has been set up, and where the social partners still talk about the topic in quite different ways. Joint participation in commissions on digitalisation might therefore enhance social dialogue in Denmark.

The creation of work without jobs via digital platforms is still in the making in the sense that this part of the labour market only constitutes a small share of Western economies. However, if this trend grows, it will challenge all sorts of regulation in the Western world – including legislation. The German labour market is regulated by legislation to a larger extent than in Denmark and Sweden, and it has a statutory minimum wage, meaning that in principle it has a safety net for new forms of digital work. The question is, however, whether such regulation can be enforced for work via online platforms. Many questions remain unanswered when it comes to implementing the statutory minimum wage in a manner encompassing the platform economy in an efficient and ethically sound way. This calls for future research into enforcing regulation of work via digital platforms.

Footnotes

Acknowledgements

This article has been presented to a number of colleagues at the following conferences: The 8th Annual TURI conference, Sesimbra, Portugal, 11–13 May 2016, Shaping the new world of work. The impacts of digitalisation and robotisation, Brussels, Belgium, 27–29 June 2016, Delingsøkonomien møter den nordiske modellen, Fafo, Oslo, Norway, 27 September 2016, FAOS research seminar, Nice, France, 28–30 September, the DSE 40th Anniversary Conference, 6–7 October 2016, Copenhagen, Denmark and Fri fugl eller frit fald – konference om platformsøkonomi, Danish Parliament, Copenhagen, Denmark, 25 October 2016. The author thanks audiences for their fruitful comments that have helped improve the final result. Furthermore, I would like to express my gratitude to Mikkel Mailand and Trine P Larsen who read and commented on earlier drafts as well as the two anonymous referees.

Funding

The empirical study presented in the article was funded by The Danish Chamber of Commerce (Dansk Erhverv), The Union of Commercial and Clerical Employees – Retail (HK Handel), The Union of Commercial and Clerical Employees – Private sector (HK/Privat) and FAOS programme funding from 2014 to 2018. The author would like to thank all funding bodies for their generous support of the empirical research.

References

Andersen

(2001) Mellem politik og marked. Copenhagen: Jurist- og Økonomforbundets Forlag.

Andersen

Dølvik

Ibsen

(2014) De nordiske aftalemodeller i åbne markeder -udfordringer og perspektiver. Oslo: Fafo.

Arntz

Gregory

Zierahn

(2016) The Risk of Automation for Jobs in OECD Countries: A Comparative Analysis, OECD Social, Employment and Migration Working Papers, No. 189. Paris: OECD Publishing. Available at: https://dx-doi-org.web.bisu.edu.cn/10.1787/5jlz9h56dvq7-en (accessed 20 March 2017).

Atkinson

(1984) Manpower strategies for flexible organisations. Personnel Management, August 16: 28–31.

Bechter

Brandl

Meardi

(2012) Sectors or countries? Typologies and levels of analysis in comparative relations. European Journal of Industrial Relations 18(2): 185–202.

Benner

(2002) Work in the New Economy: Flexible Labor Markets in Silicon Valley. San Francisco: Wiley-Blackwell.

Bernhardt

(2016) It’s Not All About Uber. Perspectives on Work. Champaign, IL: LERA, pp. 14–16.

Bernhardt

Batt

Houseman

. (2015) Domestic Outsourcing in the U.S.: A Research Agenda to Assess Trends and Effects on Job Quality. IRLE Working Paper No. 102-16. Berkeley: University of California, Berkeley.

Chui

Manyika

Miremadi

(2015) Four fundamentals of workplace automation. McKinsey Quarterly, November. Available at: mckinsey.com. (accessed 20 March 2017).

10.

Chui

Manyika

Miremadi

(2016) Where machines could replace humans—and where they can’t (yet). McKinsey Quarterly, July. Available at: mckinsey.com. (accessed 20 March 2017).

11.

Crouch

(1993) Industrial Relations and European State Traditions. Oxford: Oxford University Press.

12.

Due

Madsen

(2009) Forligsmagere og forumshoppere: Analyse af OK 2008 i den offentlige sektor. Copenhagen: Jurist- og Økonomforbundets Forlag.

13.

Dølvik

(2016) The Danes’ uneasy relationship between collective bargaining and legislation in implementing EU workers’ rights: The case of minimum wage setting. In: Larsen

Ilsøe

(eds) Den danske model set udefra. Copenhagen: Jurist- og Økonomforbundets Forlag, pp. 103–126.

14.

Ebbinghaus

(2002) Varieties of Social Governance: Comparing the Social Partners’ Involvement in Pension and Employment Policies. Cologne: Max Planck Institute for the Study of Societies.

15.

European Commission (2014) Grand Coalition for Digital Jobs. Brussels: European Commission.

16.

European Commission (2016) Digital Skills and Jobs Coalition Initiative. Brussels: European Commission.

17.

Farrell

Greig

(2016) Paychecks, Paydays, and the Online Platform Economy. Big Data on Income Volatility. Washington D.C.: The JPMorgan Chase Institute.

18.

Flyvbjerg

(1996) Rationalitet og magt. Bind 1: Det konkretes videnskab. Copenhagen: Akademisk Forlag.

19.

Ford

(2015) Rise of the Robots: Technology and the Threat of a Jobless Future. New York: Basic Books.

20.

Frey

Osborne

(2013) The Future of Employment: How Susceptible Are Jobs to Computerization? Oxford: University of Oxford.

21.

Gautié

Schmitt

(eds) (2009) Low-Wage Work in the Wealthy World. New York: Russell Sage Foundation.

22.

Gray

Suri

Ali

. (2016) The Crowd is a Collaborative Network. Computer-Supported Cooperative Work and Social Computing. Proceedings of the 19th ACM Conference on Computer-Supported Cooperative Work & Social Computing. San Francisco: CSCW, pp. 134–147.

23.

Hill

(2015) Raw Deal: How the “Uber Economy” and Runaway Capitalism Are Screwing American Workers. New York: St. Martin’s Press.

24.

Huws

(2014) Labor in the Global Digital Economy: The Cybertariat Comes of Age. New York: Monthly Review Press.

25.

Huws

Joyce

(2016a) Size of the UK’s “Gig Economy” Revealed for the First Time. Hertfordshire: University of Hertfordshire.

26.

Huws

Joyce

(2016b) Size of Sweden’s ‘Gig Economy’ Revealed for the First Time. Hertfordshire: University of Hertfordshire.

27.

Huws

Joyce

(2016c) New Estimate of the Size of Dutch ‘Gig Economy’ Revealed for the First Time. Hertfordshire: University of Hertfordshire.

28.

Huws

Joyce

(2016d) Character of Austria’s ‘Gig Economy’ Revealed for the First Time. Hertfordshire: University of Hertfordshire.

29.

ILO (2016) Non-standard Employment Around the World: Understanding Challenges, Shaping Prospects – Overview. Geneva: International Labour Office.

30.

Ilsøe

Madsen

Due

(2007) Impacts of decentralization - erosion or renewal?: The decisive link between workplace representation and company size in German and Danish industrial relations. Industrielle Beziehungen 14(3): 201–222.

31.

Kalleberg

(2011) Good Jobs, Bad Jobs: The Rise of Polarized and Precarious Employment Systems in the United States, 1970s–2000s. New York: Russell Sage Foundation.

32.

Katz

Krueger

(2016) The Rise and Nature of Alternative Work Arrangements in the United States, 1995–2015. Boston: Harvard University.

33.

Kjellberg

(2013) Kollektivavtalens täckningsgrad samt organisationsgraden hos arbetsgivarförbund och fackförbund. Department of Sociology, Lund University: Studies in Social Policy, Industrial Relations, Working Life and Mobility. Research Reports 2013:1 . Lund: Department of Sociology, Lund University.

34.

Kjellberg

Ibsen

(2016) Attacks on union organising. Reversible and irreversible changes to the Ghent systems in Sweden and Denmark. In: Larsen

Ilsøe

(eds) Den danske model set udefra. Copenhagen: Jurist- og Økonomforbundets Forlag, pp. 279–302.

35.

Koch

Fritz

(eds) (2013) Non-Standard Employment in Europe: Paradigms, Prevalence and Policy Responses. London: Palgrave Macmillan.

36.

Mailand

(2008) Regulering af arbejde og velfærd: mod nye arbejdsdelinger mellem staten og arbejdsmarkedets parter. Copenhagen: Jurist- og Økonomforbundets forlag.

37.

Manyika

Lund

Bughin

. (2016) Independent Work: Choice, Necessity and The Gig Economy. San Francisco: McKinsey Global Institute.

38.

Palier

Thelen

(2010) Institutionalizing dualism: Complementarities and change in France and Germany. Politics & Society 38(1): 119–148.

39.

Rifkin

(2014) The Zero Marginal Cost Society: The Internet of Things, the Collaborative Commons, and the Eclipse of Capitalism. London: Palgrave Macmillan.

40.

Schor

(2014) Debating the Sharing Economy. Boston: Great Transition Initiative.

41.

Schulten

Buschoff

(2015) Sector-level Strategies against Precarious Employment in Germany. Düsseldorf: Hans Böckler Stiftung (WSI).

42.

Schulten

(2016) Danish, German and European perspectives on (statutory) minimum wages. In: Larsen

Ilsøe

(eds) Den danske Model set udefra. Copenhagen: Jurist- og Økonomforbundets Forlag, pp. 85–102.

43.

Standing

(2011) The Precariat: The New Dangerous Class. London: Bloomsbury Academic.

44.

Stone

KVW

(2004) From Widgets to Digits: Employment Regulation for the Changing Workplace. Cambridge: Cambridge University Press.

45.

Toubøl

Ibsen

Jensen

(2015) Det mobile danske arbejdsmarked og organisering af lønmodtagere. LO-dokumentation. Copenhagen: LO.

46.

Visser

(2007) Trade union decline and what next: is Germany a special case? Industrielle Beziehungen: Zeitschrift für Arbeit, Organisation und Management 14(2): 97–117.

47.

Visser

(2015) ICTWSS Database. Version 5.0. Amsterdam: Amsterdam Institute for Advanced Labour Studies AIAS. October.

48.

Visser

(2016) Variation in decentralization; articulation and legal culture. In: Larsen

Ilsøe

(eds) Den danske model set udefra. Copenhagen: Jurist- og Økonomforbundets Forlag, pp. 53–84.