A Tale of Two Cities: the intrinsic spatial logic of courier protests

Abstract

Protests differ and so do protest outcomes. This is also the case in food delivery gig work, which is characterised by frequent labour unrest. Various cross-country analyses have pointed to the importance of the national context for the strategies and outcomes of courier protests. However, as the article shows, the protests already differ at the level of different cities. To analyse this, the study argues that the heterogeneity of protests in platform-mediated courier work is due to the spatially distinct logic of the respective cities. The research is based on two case studies of food delivery platforms in Germany, which were investigated with a mixed methods research design consisting of interviews, multi-sited ethnography and a survey. The findings show that the intrinsic logic of the two centres of courier protests studied (Cologne and Berlin) played a central role in the composition of the protesting groups, their strategies and subsequently the outcomes. However, it turns out that intrinsic logics are not homogeneous and in fact may exist in various forms, which can be complementary or in conflict with each other and are supported and realised by different social groups. In addition, the size of the cities also proves to be decisive for the dynamics of the protests.

Introduction

If you delivered food as a courier for the platforms Deliveroo or Foodora either in Cologne or Berlin during 2018, you would have largely identical experiences and at the same time noticed some striking differences. On the one hand, identical apps were used in both cities, the labour processes and remuneration were the same, contacts with the platforms were limited to face-to-interface contacts, and both cities were the centres of food courier labour unrest. On the other hand, in Cologne, in contrast to Berlin, no performance-based shift planning was established, there was a works council available to couriers committed to their interests, and different protest strategies were used during labour unrest. Neither of the two platforms are still active on the German market in this form. Nevertheless, key differences between the two cities are evident in the delivery services and couriers that follow them.

The different work experiences on identical platforms in a uniform national labour market and in an identical regulatory setting are the result of different protest strategies and actors, with more traditional union strategies in Cologne against direct action and a rank-and-file union in Berlin. But how these differences can be explained is less clear and is the focus of this article. The article follows the thesis that the spatially distinct and intrinsic logic of cities shapes the forms and strategies of labour unrest and thus also leads to different outcomes.

Food delivery gig work is a central facet of platform-mediated work. As such, it has already been widely analysed. According to these analyses, it is characterised by extensive technological control (Cant 2019; Griesbach et al 2019; Heiland 2021, 2022a; Shapiro 2018; Veen et al 2019). However, although platform labour is generally characterised by pronounced fragmentation and thus only a few opportunities and probabilities for workers’ voice exist (Heiland 2022b), platform labour also proves to be the ‘most vibrant and exciting areas of labour organising’ (Joyce et al 2020, 1). Food delivery workers are particularly committed (Bessa et al 2022; Heiland 2020; Tassinari and Maccarrone 2019; Woodcock 2021) and even transnational protest networks exist (Cant and Mogno 2020). In Germany, there were 69 registered protest events between 2017 and 2022, ranking Germany fourth after the UK, France, and Spain (Neumann 2023, 1).

These pronounced protest activities of food delivery couriers are on the one hand due to their special visibility in public space, which allows them to be perceived more widely than other platform workers both by the public and by their colleague couriers. In addition, different factors are cited in the literature. In many cases, rank-and-file unions in particular prove to be central actors who behave dynamically and with direct actions (Chesta et al 2019; Cini 2023; Heiland 2020; Trappmann et al 2020; Vandaele 2018). Moreover, couriers are shown to be far less atomised than assumed and establish lively communication networks or even ‘cultures of solidarity’ (Fantasia 1989) that provide the social basis for protests (Brinkmann et al 2022; Heiland and Schaupp 2021; Tassinari and Maccarrone 2019). In addition, political groups prove to be central support networks for labour struggles (Cini et al 2022; Cini 2023; Zamponi 2018, 11–12) and among couriers some make use of their political experiences (Cini and Goldmann 2020).

Platform-mediated courier work is locally bound, and the workers’ spatial flexibility is much more limited than the platforms promise (Heiland 2021), and therefore, it does not seem surprising that the corresponding protests also have a local distinctiveness, pursue different goals and strategies, are shaped by different actors, and therefore have different outcomes. Accordingly, some studies emphasise the importance of contextual factors for the emergence and establishment of the workers’ voice. Various cross-country analyses highlight the influence of national regulatory and trade union structures, political traditions, and labour markets (Bayurgil et al in press; Borghi et al 2021; Cini et al 2022; Tassinari and Maccarrone 2019).

These analyses provide valuable insights into the importance of the socio-political and cultural context. However, as will be shown in the following, not only do national variations exist but fundamentally different protest strategies and actors can be found in different cities, which also lead to different outcomes. While in the usual cross-country comparisons numerous intervening variables influence the variances of the couriers’ protests, it becomes apparent that not only a ‘time-space patterning of labour unrest’ (Silver 2003, 34) exists but that this is also and more importantly shaped by the respective urban context. For ‘the places in which workers live are not simply inert points on a map but active shapers of ongoing political practice’ (Carmichael and Herod 2012, 206). And as will be argued in the following, cities are not only the background and contextual variable of protests but represent a specific and spatially distinct structure.

The article makes three contributions: First, it emphasises the importance of local space for labour unrest, and second, it contradicts the assumption that the size of cities influences the likelihood of protests, whereas their dynamics are affected. Third, it uses the concept of the intrinsic logic of cities, yet demonstrates that these are not homogeneous, but diverse and contested. The remainder is organised as follows. First, the relationship between local context and labour unrest is discussed, (2.) and the theoretical framework of the article is reconstructed with a focus on the intrinsic logic of cities (3.). Then, the research design (4.) and the findings are presented (5.), which are discussed in the following (6.). Finally, conclusions are drawn and implications considered (7).

The locality of labour unrest and workers’ voice

The issue of local structures influencing labour unrest and workers’ voice has a long tradition. Friedrich Engels (1958), in his study from 1845 on the ‘Working Class in England’, examines the question of the ‘connections between the social structure of capitalism and the spatial structure of the town, and between this social geography and working-class formation’ (Katznelson 1992, 147). And Humphrey Southall (1988) emphasises how local cultures of workers shaped first unions at the beginning of the British trade union movements, but soon lost importance in the face of a more general loyalty to a craft. Despite this, ‘substantial interregional variation in the propensity to join and be active in national union organisations’ was and is evident (Sadler and Thompson 2001, 665). Dunlop (1948) already attributed these variations in trade union density to the specific character of regions, which influences the mentality of workers and thus their willingness to join a union. Others also emphasise that economic as well as political historical contexts differ between regions and that this influences trade union density (van der Laan 1993), a fact that has also been emphasised again in the recent past (McFarland 2019). This effect is described as ‘self-reproducing (sic) through a process of local institutionalisation and socialisation, thereby generating a regionally embedded set of industrial relations traditions and cultures’ (Martin et al 1996, 119).

Other empirical studies also reflect the importance of spatial and especially urban contexts. For example, Rizzo and Atzeni (2020) show how local factors and political traditions play a central role in organising workers in the transport and delivery sector of Buenos Aires and Dar es Salaam. And Darlington (2005) demonstrates the power of local contexts in showing how, despite industrial restructuring, political tradition and militancy persisted in the docks of Merseyside in England. Complementarily, Carmicheal and Herod (2012) emphasise the importance of spatial embeddedness for union organising in European ports.

And with regard to the role of cities, Doussard (2016) shows that protest strategies in individual cities cannot necessarily be transferred to other urban contexts and that, in particular, the different sizes of cities have to be taken into account. There are disparate positions on the relationship between the size of a city and workers’ voice (see McFarland 2019 for a discussion). While Engels (1958) assumes an increase in workers’ power as the size of the population expands, Gutman, for example, shows that workers come together primarily in smaller cities and that class and voice form more easily here (Gutman 1987).

These examples show that neither the labour context nor its institutional framing alone can explain the formation of workers’ voice. Instead, or at least in addition, contextual variables must be taken into account, and there is a need for a ‘systematic inculcation of an urban-geographical imagination into the analysis of working-class formation’ (Katznelson 1992, 141). Even though the literature referred above mainly analyses traditional labour, this is especially true for platform-mediated courier work. First, it is spatially bound and thus inevitably spatially embedded – particularly in the context of urban centres. Second, this sector is still young, so that labour relations have yet to be established, and, moreover, institutionalised power resources exist only to a limited extent. In the following, a specific theoretical approach is discussed that emphasises the distinct and intrinsic logic of cities and can thus contribute to explaining the different protest strategies and outcomes in different cities.

The distinct logic of cities

Obviously, cities differ. Faced with identical challenges, different cities react in different ways and identical situations can develop unequally in different cities. This leads to the assumption that the specific urban context makes a difference. This is shown by numerous studies: for example, Martyn Lee (1997) argues that cities develop a specific ‘cultural character’ that resembles a habitus. Ian Taylor, Karen Evans and Penny Fraser (1996), comparing the two cities of Sheffield and Manchester and their inhabitants, find different responses, structures, and practices in each with regard to post-industrial decay. Similarly, Henning Schridde (1997) shows how different perceptions, actor constellations, and milieus exist in four different major German cities, leading to different socio-political strategies and policies for analogous issues of poverty and exclusion. In a comparison of four French and British cities and their policies and governance structures, Alistair Cole and Peter John (2001) point to the particular relevance of local aspects: ‘Each city had its own culture and its own particular ways of understanding the external world’ (Cole and John 2001, 150). And, Janet Abu-Lughod (1999) shows that race has different denotations in the cities of Los Angeles, Chicago, and New York, so that even categories of social structures are shaped by a local logic (see also: Barbehön et al 2016; Dente et al 2005; Molotch et al 2000; Pizzi and Weiss-Sussex 2011; Wohl and Strauss 1958).

The approach of the ‘intrinsic logic of cities’ follows from this discussion (Berking and Löw 2005, 2008). Its aim is to analyse the characteristics of the city as such and the differences between miscellaneous cities. Similar to Lefèbvre (1974), the city is to be considered as an independent object of knowledge and not as a dependent variable of society alone. If the distinctiveness of cities is emphasised, this is usually understood as ‘local culture’ and only rarely theoretically conceptualised (Löw 2013, 895). In this context, terms such as identity or culture of a city are theoretically blurred and focus primarily on discursive aspects.

The approach of the inherent logic of cities, on the other hand, emphasises that the social world manifests itself in a specific way in each city: ‘The basic assumption is thus that all cities develop a constellation of specific, coherent stocks of knowledge and forms of expression by means of which they crystallise into symbolic subworlds’ (Löw 2012, 310). Cities are thus not a ‘derivative or miniature of wider societal or economic dynamics’ (Zimmermann 2012, 299) but distinct ‘entities of meaning’ (Löw 2012, 304).

As such, cities influence the actions of their inhabitants, because distinct meanings lead to a distinct ‘experiential space’ (Löw 2013) which leads to distinct practices. The intrinsic logic of cities is not to be understood as a deterministic social process, but as a hidden structure that is realised by the inhabitants and their practices. While the attribution of a specific habitus to different cities is criticised as ‘simplistic anthropomorphism’ (Bockrath 2008), others emphasise that cities can influence the habitus of their inhabitants and thus their practices (Löw 2008b, 89). Accordingly, both social structures and practices invariably have a local character and logic, which in turn ‘determine the character of cities, their atmospheres, but also their capacities for action and problem-solving’ (Löw 2008b, 63). Analogously, it is emphasised that cities are accompanied by ‘schemes of perception and feeling, of action and interpretation, which in their totality constitute what can be characterised as a ‘metropolitan doxa’ (Berking 2008, 23). They thus generate a spatially specific ‘natural attitude’ towards the world, a structure of meaning that frames the perceptions and actions of urban residents.

Löw (2012) distinguishes five structures of the inherent logics of cities (spatial, time, social, political and affective), of which social and political structures are particularly relevant for the context in focus, here. The former refers to the distribution of social characteristics and categories in the urban population, and the latter to both local power relations and political cultures of participation. Thus, the approach of the inherent logic of cities offers a suitable theoretical framework for analysing the different protest strategies and actors in platform-mediated courier work.

Methods

In order to adequately investigate the diversity of platform-mediated courier work, a mixed methods research design with interviews, ethnography, and an online survey was applied – approved by an ethical review board. Within this framework, two case studies of the platforms relevant in Germany at the time – Foodora and Deliveroo – were conducted from February to October 2018 (based on Yin 2018). In order to adequately reflect the spatial specificities of the labour processes and protests, the research was conducted multi-sited in different cities.

32 semi-structured interviews were conducted with couriers, 11 in Berlin and 9 in Cologne. The rest of the couriers interviewed were active in other cities. In addition, three interviews were conducted with managers.

Some of the couriers were active in the protests or even in works councils. Five of the interviewees were female, eight had no German nationality, the average age was 27, and most had at least a university entrance qualification. This social structure reflected the online survey findings. The interviews lasted an average of 80 minutes. A theoretical sampling guided the selection of interviewees (Glaser and Strauss 1967).

In addition, ethnographic methods were used in five different cities. More than 500 hours of courier work were carried out either as observational participation or in the form of participant observation accompanying couriers when working or attempting to mobilise and organise. Thus, it was possible to establish a trustful relationship with the couriers and at the same time to observe their practices in action. While interviews cannot capture the immediate practice, ethnography bears the danger of ‘going native’. By using both methods (as well as the survey), the biases were controlled reciprocally. At the same time, following Burawoy (2009), no neutral position was intended, but in the sense of an intervening and reflexive method, a position on the side of the couriers was taken.

As many of the couriers are known to exchange information online, six chat groups were also analysed in the form of a digital ethnography. Of these, three were regionally limited on specific cities and the other three were supra-regional. In all of them, the organisation of workers’ voice was a central theme. Only one of the chat groups was public. The author was invited to the others and his dual role as courier, and researcher was made transparent from the beginning. The data from the chat groups were collected, anonymised, and analysed qualitatively. In all methods used (interviews, ethnography, and chat analyses), the author disclosed his identity as a researcher. After the inquiries were completed, the results of the interviews, the ethnography, and the chat processes were coded and analysed by means of qualitative analysis software (Kuckartz 2016).

In addition to these qualitative methods, a quantitative online survey was conducted in which 252 couriers participated. Since the basic population of couriers is unknown, a purposive sampling strategy was used and couriers were actively recruited for the survey so that selection biases could be reduced.

The different survey methods were ‘fully integrated’ (Teddlie and Tashakkori 2006), so that the results were mutually referred to, with the goal to broaden scope of knowledge. The investigations were continued until there was sufficient saturation and thus ‘conceptual representativeness’ was achieved (Saunders et al 2018).

Findings

In the following, the divergent development of the protests will first be reconstructed (5.1), and then, the influence of the intrinsic logic of the respective cities will be analysed (5.2). Labour and platform labour are identically regulated (or non-regulated) in both cities. Exact figures on penetration in both cities are not known, although food delivery services are more prevalent in absolute terms in Berlin (3.8 million residents) than in Cologne (1.1 million residents).

Development of the protests

Shortly after the emergence and establishment of platform-mediated food delivery, the first courier protests emerged. After the prominent protests of London Deliveroo couriers in 2016, similar events took place in Germany from 2017 onwards. This first phase was mainly characterised by public demonstrations, which accounted for half of the protest events in Germany (Neumann 2023, 3). In the beginning, the protests were concentrated in Berlin, where couriers regularly attracted attention to their interests in front of the platforms’ headquarters or with joint protest rides. These actions were supported by couriers from both platforms, who met regularly and coordinated further action. In parallel, there were similar protests in several other cities, of which Cologne in particular stands out as the second centre. Here, too, early groups of couriers formed and working conditions were examined critically, accompanied by public demonstrations. The developments of the protests have been carefully examined by Denis Neumann (2023), whereby most of the protest events considered can be traced back to Berlin and Cologne.

Furthermore, nationwide networking of the couriers occurred in online discussion forums, which led to various coordination meetings and culminated in a nationwide protest action in December 2018 in which couriers took to the streets in numerous cities with Berlin and Cologne again as the centres. Over time, the protest goals and strategies, as well as the platforms in general, became increasingly institutionalised. Legal action was more frequently pursued and positive decisions in favour of the couriers were achieved in this way in numerous individual cases (only occasionally with collective effect). In addition, from 2017 onwards, couriers tried to establish works councils in various cities – which was increasingly successful in the following years. Works councils are only authorised to look after employees. However, in Germany, employment was the norm only for couriers at Foodora, while at Deliveroo couriers could choose between employment and self-employment as late as 2017. With the successful works council election at Deliveroo in Cologne, however, this platform let the couriers’ fixed-term contracts expire throughout Germany and only continued to operate with self-employed workers. As a result, the last works council member in Cologne left the company in spring 2018 and the committee was dissolved. In the autumn of the same year, Deliveroo abruptly left the German market. At the same time, a third platform, Lieferando, gained relevance in Germany and took over the German business of its competitor Foodora in December 2018, thus enjoying a quasi-monopoly in Germany to this day.

After the turbulent year 2018 with numerous protest events, these decreased somewhat, but were still and permanently present (Neumann 2023). In addition to multiple legal disputes, there were repeated demonstrations and contested works council elections. With the 2021 protests at Gorillas, a grocery delivery service, they reached a new peak with a clear focus on Berlin (Ewen et al 2022; Orth 2022).

The data of the online survey from the summer of 2018 show an average satisfaction among couriers in general: on an eleven-point scale from ‘very dissatisfied’ to ‘very satisfied’, the mean value is 5.74. However, this is well below the average 7.49 reported by German employees questioned in another representative survey (Heiland 2019, 302). Accordingly, only 14% of the respondents see no reason for strike or protest in their current work situation. Here, no regional differences can be noted, which also applies to the reasons for the high level of dissatisfaction – as is also expressed in the prevailing themes of protest (Neumann 2023, 2). Nevertheless, two centres of protest can be identified, Berlin and Cologne, which differ significantly from each other. In Cologne, the election of works councils was one of the central goals of the couriers from the beginning, which was successfully installed in 2017. The couriers sought support from the established trade union NGG (The Food, Beverage and Restaurant Trade Union). In addition, a Germany-wide Facebook page was organised from Cologne, which was crucial in managing the public discourse, networking among the couriers and the protests. This was later co-opted by the NGG union, and a union secretary was also recruited from the ranks of the page’s organisers to deal with courier issues. In contrast, protest activity in Berlin was continuously high but mainly focused on visible events, with works councils explicitly not being a target. A works council was not elected there until 2022. The background to this is that in Berlin the couriers cooperated with the anarcho-syndicalist Free Workers Union (FAU). Their strategy focuses on labour struggles based on direct action, while rejecting reformism which would in their eyes stabilise the existing relations of exploitation and oppression (FAU 2015). Thus, the starting point of the couriers’ discontent is identical, but their strategies of protest are divergent.

Logic of cities, logic of protests

In both Berlin and Cologne, there was a similar potential for organising and mobilising the couriers in a spontaneous and direct way, which was effective for individual occasions in the absence of institutionalised representation of interests – for example, successful resistance against dismissals of individual drivers in Berlin or against racist statements on the part of some restaurants in Cologne. However, through such individual direct actions, developments took different paths. In the following, first the protests and their development in Berlin and then in Cologne are outlined.

In Berlin, the protests were not unionised at the beginning, but self-organised, as one of the interviewees describes: ‘In the beginning, there were not many of us and we organised the first actions on our own. That attracted a lot of attention. But nobody from the trade unions approached us. They didn’t even agree on which of them was responsible for us’. Soon after, the anarcho-syndicalist trade union FAU was brought on board:

And then there was someone in our group who had contact with the FAU. And they reacted immediately and first had a hands-on mentality, wanted to do something directly, and second, they were closer to most of us in terms of ideas. So it wasn’t a cooperative ‘we’ll make the world a better place together with the platforms’, but a ‘go into the conflict and oppose the platforms and squeeze something out of them’.

As expressed by various interviewees, there was a distinct political fit between the couriers and the FAU: ‘I really believe in self-organisation in grassroots movements, even before the struggles at Foodora, but even more so now’, expressed one courier. And another courier said: ‘That’s what we do here. We don’t put on a fancy shirt and sit down with the platforms to discuss. We go to the streets and force them to act’. And in one chat, one wrote: ‘Yeah, Berlin is showing the platforms the ropes’. Moreover, in some cases the protests themselves attracted more couriers. For example, one of the interviewees described that working as a courier only came into question for him when a friend told him about the FAU struggles there: ‘And then I found the FAU-struggle. And I thought to get in to this job to also be part of this struggle campaign. With the old 70s idea to work within the factories to organise the factory workers’.

Another courier described the clientele of his colleagues accordingly: ‘The people who ride for Deliveroo are all dropouts. They are not career people. They are all politically left-wing’. This is also confirmed in the participant observations where an alternative clientele was strikingly represented among the couriers in Berlin. And various other local political groups and milieus were also represented at the protests. In addition, a significant proportion of the protesting couriers were migrants, so that the preparations for the protests were conducted in English. Most of these couriers were from southern European countries, which they mostly left as a result of the economic crisis in 2008 and the high youth unemployment that followed. They often had a high level of formal education and came to Berlin specifically because of an interest in the distinct alternative community. Accordingly, many of these couriers already had experience within organising and participating in protests. Accordingly, the meetings and protests in Berlin were mainly characterised by young and international couriers from a left-alternative milieu.

However, the protests and the groups organising them were not characterised by homogeneity. There were objections from other couriers to the cooperation with FAU and the corresponding actions: ‘Nobody from our group is in contact or agrees with what the FAU is doing. Not even one. It’s sort of two worlds’. Others spoke out in favour of either cooperative dialogue with the platforms or the traditional trade union route of electing works councils and criticised the other couriers and their strategies. Despite this, the FAU cooperation and strategy remained dominant in Berlin for a long time.

In Cologne, too, the initiative for the protests came from the couriers. The first protests emerged from chat groups and meetings after the end of the shifts. Attempts to approach the established trade unions were not successful at first: ‘They told us that we couldn’t be organised. Too precarious, too distributed, too irrelevant’. At the same time, there was also contact with the FAU:

Yes, we also met with them. And we also noticed that they wanted to support us. But they couldn’t really do that. They were a few people who did union work from time to time. And they had somewhat different ideas than we did. Our goal was a works council. And that was somehow not high on their list.

Here, the different resources of FAU in the various cities become apparent. While in Berlin FAU represented a large group, from which someone regularly appeared at the courier meetings or initiated them, in Cologne there were only individual meetings with a few people.

From the beginning, the Cologne couriers aimed to elect works councils. The demonstrations were a means to an end, on the one hand to increase the pressure on the platforms regarding the improvement of working conditions and on the other hand to establish an institutionalised representation of interests with works councils. The differences compared to Berlin were explicitly identified: ‘We are not in Berlin here’ was one chat comment. And one interviewee explains: ‘Not everyone here is left-wing per se and supports our project. We have to work for solidarity’.

The differences between the cities were also expressed in the couriers’ social structure, which in Cologne was more student-oriented and less characterised by a left-wing alternative milieu. There were similar numbers of migrants, but most of them were not from Europe, had less formal education and many were refugees with few alternative job options. This group was clearly underrepresented in the protests: ‘There are a lot of students here, almost all of them “organic” Germans’, describes one courier. The couriers tried to organise their colleagues – especially those with a migration background – with social activities such as barbecues, repairing bicycles and bicycle races.

After the first successful actions, the traditional trade union NGG started to support the couriers with advice. After two successfully elected works councils in Cologne (see above), an online campaign of the couriers and a public meeting with the German Minister of Labour, the NGG made the courier movement one of its flagship projects. As a result, Cologne was also the starting point for organising couriers in other cities. Quite frequently first groups already existed in these cities, but now the NGG targeted its local secretaries on the issue and couriers from Cologne regularly travelled to other cities to tell those couriers about their experiences with protests and, above all, the election and work of their works councils. These efforts were effective insofar as people tried to establish works councils in many cities – often successfully. When Foodora was integrated into the new monopolist Lieferando, the latter had to take over the existing works.

As a result, the different political strategies led to a comparatively important role of the traditional union NGG in the courier protests. Between 2017 and 2022, NGG was involved in 62% of the protest events. The FAU played a role in only 30% and workers collectives as autonomous forms of organising were involved in 43% of the protests (Neumann 2023). The relevance of the NGG becomes even more evident if one takes into account that the online campaign ‘Deliver at the Limit’, which was co-opted by the NGG at a later stage, is also regarded as a workers’ collective. In contrast, mainstream unions were only involved in 28% of protest events worldwide and in 37% in Europe (Stuart et al 2023, 4). A new phase began with the industrial action at the Gorillas delivery service, particularly in 2021. This was almost exclusively confined to Berlin where they again recruited their participants mainly from a migrant and left-alternative milieu. But unlike before, the activists in Gorillas, while using the resources of the various unions, did not cooperate exclusively with any specific union. In contrast to its usual principles, the FAU supported the goal to establish works councils and at the same time distanced themselves from the traditional trade unions by demanding a general and political right to strike.

Discussion

As outlined, the couriers in Cologne and Berlin were faced with an identical starting situation and shared an identical discomfort with the status quo of their work. The prevailing variances in the strategies and outcomes of the subsequent protests can be explained with reference to the intrinsic logic of cities, which is able to ‘influence action independently of the specific actors’ (Löw 2008a, 40). The cultures of participation and the social structure were particularly important in this context (Löw 2012). Couriers in both cities formulated explicit spatial demarcations or references and thus emphasised an existing local style of dealing with labour conflicts. In other words, different political cultures of participation prevailed in both cities, which led to different political strategies and institutional connections and thus established path dependencies. While in Cologne the connection to traditional trade unions was sought and with them the establishment of works councils, in Berlin an anarcho-syndicalist trade union was predominant in the courier protests and with it the focus on direct action. In addition, the couriers in Berlin found support from left-wing political groups and in Cologne, for example, the traditional trade union demonstration on May Day was a point of reference.

Additionally, the different intrinsic logic of the cities is expressed in specific social structures. Thus, the participant observations revealed striking differences between the couriers and their associated milieus, with, for example, a more pronounced radical left orientation in Berlin. Moreover, the different structure of migrants influenced the protests and the composition of their participants.

Thus, it can be stated that the intrinsic local logic of cities is relevant. But at the same time, it becomes clear that this logic is not a fixed entity, but is in continuous development, interacts with other actors and institutions and is subject to conflicts. The intrinsic logic establishes new path dependencies (e.g., with the choice of cooperation with trade unions) and reinforces existing ones (e.g., with the influx of milieu-congruent people). Furthermore, with regard to the courier protests, the resources of the trade unions are relevant. Especially, the organisational network of the traditional trade union NGG as well as the touring of individual organisers through the cities – analogous to early trade unionists (Southall 1988) – resulted in enforcing the institutionalising of conflict strategies with a focus on works councils. At the same time, the risk of a ‘militant particularism’ (Harvey 2008, 24) that comes with the Berlin couriers’ focus on direct action and confining struggles to place-based ones is not only a result of their strategy but also of the limited resources at their disposal.

This shows that intrinsic logic is not the only explanatory variable. It ‘is the complex of rules and resources that gives events a locally specific shape’ (Löw 2012, 313). However, first, its influence can be superimposed. In the case of most other cities where courier protests took place, for example, the traditional trade union route was deliberately driven by the NGG union by means of its use of resources, limiting the impact of an intrinsic urban logic. Second, intrinsic logics are not homogeneous but contested and characterised by conflict. Cities are not uniform spaces of experience (Kemper and Vogelpohl 2013, 12) and urban logics are not determinisms, but shaped by conflictual interactions of different groups competing for interpretive supremacy, strategies and resources. If they are understood as a homogeneous essence, their explanatory potential gets lost and the interplay of conflicts as well as the constitution of intrinsic urban logics is neglected. Thus, in the case of Berlin, although the protest development worked according to the intrinsic alternative logic of the city, at the same time conflicts existed over the protest strategies and a few years later the couriers changed their protest strategy and likewise realised a works council. And in the new wave of protest of groceries couriers of Gorillas, different strategies, militant and institutionalising, were combined. It follows that there is not one intrinsic logic in cities, but different ones that can be complementary or in conflict with each other and are supported and realised by different social groups.

Furthermore, and third, the primary focus on intrinsic logics neglects social and economic structures and relations. Speaking with Bourdieu, intrinsic urban logics are thus not only structuring structures (Frank et al 2013, 206), but at the same time structured structures. The political economy influences the built urban environment and the social structure of cities and thus their intrinsic logics (Harvey 1982).

It should also be mentioned that the size of the city has an effect on the protests. Analogous to Engels’ assumption (1958), the pool of couriers willing to protest proved to be larger in Berlin, but at the same time it was easier to achieve a community in the form of a bigger group of connected couriers in the smaller city of Cologne, as Gutman (1987) already emphasised. Hence, size does not necessarily affect the likelihood of protests, but it certainly affects their dynamics and characteristics.

Conclusion

Courier work is spatially embedded. Accordingly, it must be analysed as such. Since this service is limited to urban metropolises, it is important to consider cities as distinct socio-spatial contexts in this analysis. And as shown, cities have intrinsic logics that influence the practices of their inhabitants. Thus, there is not only an ‘industrial atmosphere’ (Marshall 1919) but also a specific urban ‘atmosphere’. This affects the capabilities and strategies of the workers, so that, as in the case analysed here, despite similar starting points, different forms of protests and outcomes result. Although in Cologne protests were increasingly institutionalised and traditional, in Berlin they were characterised by autonomous and militant strategies. While the influence of platforms on cities is discussed in the context of platform urbanism (Barns 2020), so that a ‘new form of urban infrastructure’ is stated (Moore and Rodgers 2018), the analysis shows that cities also have an impact on platforms the other way around.

First, the article points out the importance of space and especially of intrinsic urban logics for labour unrest. Unionisation and workers’ voice are ‘a process of coming together, of organising over space’ (Southall 1988, 467). The role of the local is often neglected. When space is taken into account, the focus is usually on national differences. But scale and place matter when it comes to workers’ voice. Protests are not place- or even spaceless. This is especially true in the case of locally bound platform labour, which usually lacks established labour relations and industrial culture, making local political cultures and structures even more relevant. As shown, the intrinsic logic of cities is an appropriate approach to analyse the characteristics of places and their influence.

Second, the size of cities does not necessarily determine the likelihood of protests, but their dynamics. However, third, the findings show that intrinsic logics are neither self-evolving nor homogeneous. Contrary to what is often conceived, they are diverse, contested, and developing in practice. They are both dependent and independent variables, both product and constituent of conflicts and negotiated practices. As such, their influences and outcomes are not determined but, as the development of protests in Berlin shows, changeable. However, it should be mentioned that intrinsic logic is one explanatory variable among others. It is a piece of the puzzle that has different effects depending on the situation. Moreover, it should be clear that it is also a dependent variable and is inevitably linked to the local, national, and global political economy.