Worker Power in the Platform Economy

Trade union strategy in Germany in the age of platform capitalism

The response of unions in Germany to the growth of the platform economy can teach us important lessons about the challenges of ‘platform labour’ as an emerging category in the labour market, and in particular to the limitations of pan-national responses. Problems for national and international regulators include a lack of clarity on where work is commissioned or carried out, which regulatory frameworks are relevant, and where the burden of taxation or insurance lies. As a result, the formulation of either a supra-national labour market policy or the application of already-existing regulation becomes incredibly complex and onerous. In response to this, there has been a sea-change in the European Commission’s emphasis on who should be policing and monitoring the extent of platform labour in member countries, with an expanded role created for ‘social partners’ (including employers’ organisations, NGOs, and trade unions) since 20141.

This article focuses on one, I believe significant, example of trade unions’ responses to both the problems posed to workers by the platform economy, and to the opportunity for a greater role in planning social and legislative responses. In Germany, the two largest trade unions – IG Metall and Vereinte Dienstleistungsgewerkschaft (ver.di) – contain large numbers of platform workers as members, and have developed institutional strategies that work at two scales. First, they attempt to change the conditions for workers at labour market entry. Second, they apply pressure to platform owners to alter their behaviour within markets. In contradistinction to new-unionist interventions into platform labour in Britain (by the IWGB) or France (by Unsa-VTC), and to the German unions’ own prior interventions into agency contracts, I argue that this strategy is particularly ill-equipped to facilitate collective decision-making by workers, or to ensure the continued participation of workers in the planning and carrying out of campaigns.

While I have no desire to paint a crude dichotomy between ‘democratic’ campaigns of industrial action and ‘un-democratic’ policy and lobbying campaigns, nor to underplay the stultifying role of labour aristocracies or union bureaucracies on workers’ campaigns, it is the assumption of this article that organising along the lines of collective bargaining (balloting, setting demands, taking action) requires at least some collective agency from workers in a given sector or employment. In its absence, some analogous mechanism is needed to allow workers to influence union decision making, a criteria that I do not see evidence for in the case of IG Metall or ver.di’s platform labour campaigns.

German Trade Unions and Neo-Corporatism

Both IG Metall and ver.di organise predominantly outside of precarious work sectors, and the vast majority of union members and activists are subject to stable forms of employment in the formal German economy – largely in the metal industry for IG Metall, and in public and professional services and transport for ver.di. As such, the vast majority of union strategy and activity for both organisations is based around a model of extensive trade union integration into policy-making, including having seats on management boards at firm level, influence over individual workplace councils, and consultation rights for state and national-level policy decisions.

This level of integration into sectoral management encourages union strategies which rely on easy access to managers, HR departments, and workplace policy-making mechanisms. As Florian Schmidt points out, it is exactly these arms of workplace governance which are automated away in these new labour platforms2, making it impossible to extend pre-existing workplace strategies to work arranged through them. Further, as platform owners strenuously deny being either employers or labour agencies3, they are overwhelmingly resistant to taking on a negotiating position with unions analogous to that of either.

It is my position that in adapting to the challenges of having no ‘opposite number’ in the platform economy, both IG Metall and ver.di have adopted a model of social partnership that has compromised their ability to facilitate collective action and decision-making by platform workers at either the platform or policy level, and largely limits unions to being providers of individual services or informal labour market watchdogs. This is in stark contrast to their collective and militant interventions into temporary agency work and formal labour relations, which are based on the principles of collective bargaining.

Platform Labour

Defining the nature and scope of platform labour is plagued by technical and ideological problems, as well as the downright refusal of many platforms to share their own estimations of how many labour suppliers and customers they have in a given region4. Crucially, and especially for Germany, it’s important that the definition moves beyond the monopolies with which the concept of platform capitalism has become associated, particularly Uber. Not only do they represent a highly specific form of labour, but furthermore, if we are to take platform labour seriously, we must recognise that it has expanded way beyond a few well-known firms.

Different researchers and institutions have adopted incommensurate definitions of the ‘platform economy’ or its near synonyms (‘gig economy’, ‘sharing economy’, ‘collaborative economy’, etc). Many of these definitions suffer from subordinating the labour relationship to an analysis of the platform as an economic technology (analysing labour alongside capital and commodity exchanges rather than other wage relationships5), the substitution of an effect of the labour relationship for a definition6, or a reliance on terms like ‘sharing’ or ‘collaborative’ which are, at best, already value-laden7.

In an attempt to avoid these problems, I will be using the definition proposed by Lanaerts et al8, that labour (or, ‘online talent’) platforms are:

digital work platforms that provide online peer-to-peer intermediaries through which users can have temporary access to other users’ services and the payment is made through the platform.

This definition is far from exhaustive, but it provides a basic conceptual framework for understanding platforms as not only a passive marketplace for labour, but as being ingrained within the wage relationship as a payer of, and commission-taker on, wages. It also has the benefit of clearly differentiating platforms from job advertising boards, while allowing for the integration of those platforms who do take on some of the responsibilities of labour agencies9.

By applying this definition to existing European studies, Lanaerts et al hold that the likely size of the platform workforce across Europe is about 100,00 people, or 0.05% of the total EU workforce10. According to qualitative surveys among platform workers undertaken by Huws and Joyce11, the vast majority of workers use platforms to supplement their main income, with 18-24 year old males vastly overrepresented, alongside a comparative overrepresentation for new labour market entrants – particularly migrant workers.

While the size of the platform labour workforce is growing rapidly, particularly in Germany and the UK, the vast majority of growth to date has been in sectors which have traditionally involved large amounts of self-employed workers, a trend predicted to continue in the next few years with demand for freelance software development fuelling growth12. Peculiarly, Germany’s platform workforce appears to have a much higher proportion of ‘crowdworkers’ (those who bid for online work that doesn’t necessitate face-to-face contact or being in a given location) than the estimates for the rest of Europe or North America, where location-based ‘gig’ work, especially in the transport sector with Uber-style services or in deliveries, predominates.13

This point is significant for labour organising. Where campaigns against Uber and Deliveroo have been successful in London and Paris, they have had access to networks of drivers who know each other in the real world and can plan campaigns together at meetings, and to important physical space that can be contested by protests and direct actions (most notably in the case of Unsa-VTC, Paris’s CDG airport). When organising workers who are geographically dispersed across cities, and where the anonymity of clients and the lack of platform-owned property makes even defining relevant protest locations difficult, it’s necessary to discover new forms of leverage in order to have a campaign at all, never mind a successful one.

German Unions and Platform Workers

Evidence that the growth of the platform labour workforce directly affected the membership of either IG Metall or ver.di before their intervention into the platform economy is difficult to come by. Not only there been no evidence that platforms have ‘disrupted’ the metalworking industry, the most overrepresented age group in the platform workforce appears to be underrepresented in the whole of German engineering, with young people on apprenticeship schemes equivalent to just under 0.6% of those employed in the sector in 201114. In the case of ver.di, which has always contained self-employed artists and journalists as part of its Medien branch, it is conceivable that design work platforms did have an impact on their members15. As Medien is but one of the five previously independent unions that founded ver.di, and as the branch also includes print workers and workers in paper manufacturers, it seems unlikely that those directly or indirectly affected by the platform economy made a sizable proportion of ver.di membership.

This implies that, for both unions, the motivation to expand into the small but growing platform workforce was not the result of an organic need arising from normal trade union activity. Instead, it was more likely a policy decision made either to respond to the perception of an emerging social problem, or as a strategy believed to directly benefit the union through profile-raising or future recruitment. As I will show in the next two sections, this kind of outside intervention into an emerging form of labour has a much more ambiguous relationship to collective action and decision-making than that found in more common forms of labour organising.

Any union strategy in Germany needs to be understood against the background of overall union density. Since the mid-nineties, union membership in Germany has been steadily declining, falling to just 18% of the workforce in 2013 – 49% lower than in comparable European neo-corporatist countries16. Given the high level of unions’ integration into policy-making bodies at both a local and national level, this dramatic fall in membership creates a problem of legitimacy, as trade unions begin to occupy a position of power in relation to industries where they represent a shrinking minority of workers. This is a problem which Gunter Haake, director of the ver.di-run company Mediafon, is refreshingly honest about: not only openly claiming that existing German trade unions are losing their social legitimacy, but expressing the fear that new social organisations are likely to emerge as rivals for traditional unions perceived as failing to represent growing swathes of the workforce.17

Options to stave off this legitimacy crisis involve either a concerted attempt to recruit greater numbers in the industries where the union has traditionally been strong and break into new sectors where they can predominate, or to emphasise the policy credentials of the trade union on wider economic matters – supplementing their claim to represent workers at large with the kind of cachet that policy think tanks or research bodies command. IG Metall and ver.di have attempted utilise both of these tactics, with varying or unclear results, and with organisational disjuncts between expanding normal union activity and specific policy interventions emerging over time.

IG Metall’s recruitment drive into precarious work began in earnest following the Hartz reforms of 2003, which relaxed labour regulation for the hiring of temporary workers by large companies. After initial misgivings by union leadership that organising temporary workers would legitimise a ‘conscious creation of a cheap workforce’18, the union launched a series of campaigns to include temporary workers within existing collective bargaining arrangements by exercising leverage over contracting firms and hiring companies in a strategy described by one union spokesman as ‘recruitment orientated, participation orientated, and conflict orientated’.19

While some scholars read IG Metall’s intervention into platform labour as a continuation of this initial orientation towards temporary workers, it is notable that there is no corollary in the platform economy to the points of leverage used. Hirers of platform work remain largely anonymous, and most platforms themselves reject outright that they are hiring companies or agencies of any sort to avoid the costly responsibilities towards workers implied in the employment relationship.

For ver.di, which has included self-employed workers within its membership since its formation in 2001, there appears to be a similar disjunction between its organisation of temporary workers and its strategy for platform labour, with the former (as seen in the 2013 campaigns amongst Amazon and Hermes logistics workers20) characterised by large-scale strikes and direct action that appear to be absent from accounts of the latter.

In the absence of offering direct leverage for workers against platforms and those who hire on them, IG Metall and ver.di have focussed their recruitment drives amongst platform workers on the provision of benefits and services that can be accessed through membership – including unemployment and medical/incapacity insurance, status-based counselling, and access to legal and regulatory advice for individual workers.21 While such initiatives are likely to provide vital services for platform workers, who are often forced to forgo the benefits included within the employment relationship when they agree to work through a platform, they do not go beyond the level of individual problem-solving, and often replicate the duties of an employer under German labour regulations. Whether or not this strategy proves to be sustainable, it is unclear to what extent it promotes either the collective decision-making of platform workers in spite of their geographical separation or differences in their status, or their integration into the union’s own decision-making process.

Both unions have attempted to supplement the lack of a cohesive collective bargaining strategy by designing initiatives to influence legislation and the relative competitiveness of different platforms in the labour market. These priorities are reflected in the participation of both unions in German government commissions’ reports on the platform economy, in single-union attempts to pressure platform owners to adopt their recommendations for best practice, and in unilateral or cross-union statements and policy recommendations which, due to their international character, fall outside of the scope of this piece.

In recent years, there have been eight separate government reports on the platform work or digitalisation that German unions have contributed to – including the much publicised white paper Arbeit 4.0. It is not evident how decisions within unions about the content of their contribution was made and, in compliance with the usual terms of such enquiries, unions are not the only bodies which provide input to the commission, with employer’s organisations and nebulously-defined ‘social partners’ also permitted to contribute their perspectives on the issue at hand. In the case of the platform economy, recommendations arising from this process have been rather vague. In the case of Arbeit 4.0, the final report has not contained any substantive recommendations whatsoever22.

Attempts to shape the nature of platform employment at the level of the labour market have resulted in a number of substantive agreements between unions and platform owners, and have created - to some degree - a voice for platform workers within broader union policy, but suffer from the limitation of there being no collective unit of platform workers able to enforce the terms of agreement at the level of transactions. IG Metall have been much more active in this sphere than ver.di, whose own approach appears rather ad hoc and is described by Haake as the union and Mediafon prioritising campaign goals based on how regularly a certain issue emerges in individual counselling or mediation23.

Fair Crowd Work, which is funded by IG Metall, extends the scope of previously existing forums for platform workers which allowed them to ‘rate’ customers on a particular platform on the basis of their treatment of workers, wage theft, harassment, or late payment. Forums such as Turkopticon for Amazon’s Mechanical Turk platform (AMT) provided a corollary to the worker rating systems of platforms that gave customers information with which to discriminate between workers when commissioning tasks, by allowing workers to warn each other of unscrupulous hirers with the aim of changing their behaviour or forcing them off the platform. After Turkopticon co-founder Michael Six Silberman was hired to the IG Metall staff, a cross-platform forum was created to allow workers to rate the working conditions on platforms themselves, and to provide legal evaluations of individual platforms by lawyers commissioned by IG Metall. Just as Turkopticon provided the opportunity for workers to affect the actions of hirers on AMT, Fair Crowd Work provides a mechanism to allow consumers and workers to choose between platforms where workers report being better or worse treated, creating the possibility of market share shifting to platforms rated more favourably by their workers. So far, Fair Crowd Work has gained the agreement of eight platforms to abide by a shared Code of Conduct and submit to an arbitration process overseen by an Ombudsman (Verband) made up of workers’ and platforms’ representatives and labour judges24.

While the creation of an ombudsman and a broader forum for workers to evaluate platforms is an achievement, doubts remain about the extent to which these encourage collective and democratic decision-making by workers. While a worker is free to say what she likes about a given platform, her concerns and grievances are reduced to a simple thumbs-up-or-down rating in five pre-determined categories, rather than integrated into a debate about what the strategy of the workers’ movement should be. Furthermore, if she were to be on the receiving end of mistreatment by one of the Verband signatories, her only recourse is to a lengthy arbitration process over what is likely to be a small amount of money, rather than seeking support through a collective bargaining unit that could police the agreement before arbitration and negotiate on aggregate cases simultaneously.

Conclusion

While it would be disingenuous to suggest that the tactics adopted by German unions were not responses to the specific difficulties involved in organising crowdworkers, I do not believe that these are sufficient explanations for the lack of evident collective decision-making and worker participation in the German campaigns.

The geographical dispersion of crowdworkers, the anonymity of both they and their clients, and the relative ease with which customers can get around online industrial action by hiring crowdworkers from abroad makes it difficult to coordinate the kinds of strike and direct action campaigns seen by organisations of location-based gig workers in Britain and France. Despite this, at least one attempt to utilise direct leverage on crowdworking programs has been made outside of the traditional trade union movement. Under the moniker We Are Dynamo (WAD)25, a collective of crowdworkers on AMT decided to directly target academic hirers with a campaign of lobbying and the threat of informing ethics boards in the case of underpayment for research work. As academic researchers were believed to be around 20% of hirers on AMT at the time of the campaign, this constituted a display of force against a significant fraction of the platform’s trade, requiring a collective division of labour amongst multiple crowdworkers26.

Whether any established trade union adopts an approach which relies upon creative collective action by crowdworkers is yet to be seen. Campaigns like WAD’s do not sit easily with the principles of the Frankfurt Paper signed by IG Metall, Sweden’s Unionen, or the Austrian OeGB – which outlines that platforms, not hirers, are appropriate negotiating partners and promotes the voluntary agreement of platform providers to codes of conduct. The WAD example, however, does prove that collective bargaining and direct action can be used by crowdworkers regardless of the attitudes of the mainstream labour movement.


  1. For recent formulation of this orientation, see The EU Directorate General for Internal Policies, The Social Protection of Workers in the Platform Economy; EMPL Committee Study 2017 (pp. 107-8) 

  2. Florian A. Schmidt, Digital Labour Markets in the Platform Economy; Friedreich Ebert Stiftung (2017) (p.11) 

  3. Kilhoffer, Lenaerts, Beblavy, The Platform Economy and Industrial Relations: Applying the Old Framework to the New Reality; CEPS Research Report: August 2017 (p.27-8) 

  4. ibid., (p.18) 

  5. For an example, see Measuring the Gig Economy; Office_ for National Statistics, 2016 

  6. For a critique, see Drahokoupil and Piasna: ‘Work in the platform economy: beyond lower transaction costs’ in Intereconomics: Issue 52 (2017) (p. 335-6) 

  7. For a critique, see Drahokoupil and Fabo: ‘The platform economy and the disruption of the employment relationship’ in ETUI Policy Brief: No 5 (2016) (p.3) 

  8. Lanearts et al: Online Talent Platforms, Labour Market Intermediaries and the Changing World of Work; CEPS/IZA independent report: May 2018 (p.i-ii) 

  9. Platforms like Ploy, Temper, and Book a Tiger have integrated screening processes for workers, part payment for vocational training, or minimum shift wages into their contacts with workers 

  10. ibid., pp, ii, 15-6, 20 

  11. Cited in ibid., pp, 20-1 

  12. ibid., p. 14 

  13. Huws and Joyce: ‘Size of Germany’s “gig economy” revealed for the first time’; Uni Europa Crowd Working Survey November 2016 

  14. Paul Ryan et al: Finanzielle Aspekte der betrieblichen Ausbildung in Deutschland, Grossbritannien, und der Schweiz;f Hans Boeckler Stiftung, Arbeitspapier 241 (2011) (p.39) 

  15. although I have been unable to come across any account of this in my research 

  16. Ilsoe, ‘The digitalisation of service work – social partner responses in Denmark, Sweden, and Germany’ in Transfer – European Review of Labour and Research: Issue 23, Volume 3 (2017) 

  17. Haake, ‘Trade unions, digitalisation and the self-employed – inclusion or exclusion’ in Transfer – European Review of Labour and Research: Issue 23, Volume 1 (2017) (p.66) 

  18. Benassi and Dorigati, ‘Straight to the core – explaining union responses to the casualization of work: the IG Metall campaign for agency workers’ in British Journal of Industrial Relations 2014: Issue 7 (p.13) 

  19. Wetzel, quoted in ibid (15) 

  20. Ver.di, Anpacken 2013, NR 1 (Englische Vesion) (pp. 1, 3-4) 

  21. Haake, p.66, IG Metall recruitment leaflet 2008 (EN) https://www.igmetall.de/0161455_Leistungen_Aug2008_gesamt_Englisch_28c4f7286cd82fa03678012a851a6f0c8131a413.pdf 

  22. EN edition, p.55-7 https://www.bmas.de/SharedDocs/Downloads/EN/PDF-Publikationen/a883-white-paper.pdf;jsessionid=1168DC4FFF15C493F497515C11A63338?__blob=publicationFile&v=3 

  23. p.64 

  24. While Fair Crowd Work asserts that the platforms involved in the Verband (‘Obudsman’) account for more than two-million workers, only three of the signatories have more than 15 reviews from workers or customers on any ratings site beyond Fair Crowd Work at the time of writing. 

  25. A group arising from the forum of the same name 

  26. Johnson and Land-Kazlauskas, Organising On-demand: Representation, Voice, and Collective Bargaining in the Gig Economy; ILO Conditions of Work and Employment No. 94 (2018) (pp. 13-4)