Reclassifying gig workers as employees will likely raise wages and grow unions’ membership, but it will not delay the next round of class struggle over who controls gig workers’ labour. To shift back or maintain the focus on the control side of gig work, gig workers are continuing to to assert self-organised models of struggle.
Self-organised struggle in Europe
Rather than being channeled into failed organising models, gig workers are adapting their tactics and strategies to address the conditions of their work to find new ways to connect and take collective action. One of the most impressive efforts at self-organising can be seen in the Deliveroo and food delivery rider strikes in Belgium, Germany, Spain, France, Italy, The Netherlands, Hong Kong, South Africa and India beginning in 2016. Callum Cant, in his book Riding for Deliveroo: Resistance in the New Economy (2020), counted 41 strike actions over 18 months between 2016-17 involving 1493 workers in these countries. With assistance by independent militant unions like the IWGB in the UK, Deliveroo worker organisers engaged in workers’ inquiries to reverse engineer the algorithm to identify new tactics and strategies for organising to undermine and defeat algorithmic management strategies.
For example, Deliveroo and other food delivery workers organising around issues of low pay, unsafe working conditions, stolen bikes, and the lack of bathrooms find ways to overcome the fragmentation built into algorithmic management by connecting with other riders outside key pick up spots and nearby parks where they meet and share phone numbers, coordinate actions, distribute paper newsletters, and develop relationships that leads workers to engage in off- and on-line forms of collective action and mutual aid on What’s App, forums, chat rooms, and even mass de-activation from the apps as strike action at critical times when demand is high. Storefronts and parks where large numbers of workers are mapped out for purposes of organising become “a stand-in for a more traditional workplace”, according to Jamie Woodcock and Mark Graham, determined by the algorithm allowing workers to build in person “cultures of solidarity”, according to Heiner Heiland and Simon Schaupp.
In coordination with other self-organised unions in Europe, Deliveroo undertook public strikes by refusing to log on at times of peak demand, blocked popular restaurants, and filled the streets in moving festivals on wheels. Organisers used “Web 2.0” tools as a “virtual watercooler” combined with established tactics of in person one on one meetings and newsletters determined by the algorithm allowing workers to build what Heiner Heiland and Simon Schaupp call in person “cultures of solidarity”.
Gig drivers as well have coordinated their struggles across borders forming the International Alliance of App Based Transport Workers in 2020 with representation of unions in 23 countries on five continents. Such cross border networks and federations of independent unions recognise the global reach of the app companies, investors, and the influence of the business model and are adapting their tactics to take interconnected local actions on a global scale to amplify their disruptive impact.
Self-organised struggle in the US
Similar tactics and strategies are being devised by Los Deliveristas Unidos (United Delivery Drivers), a self organised group of gig workers in New York City. With backing from 32BJ SEIU and the Workers Justice Project (WJP), immigrant riders, some of whom are related, from Guatemala and Mexico are organising in local parks, at soccer games, and at key pick up spots and resting areas where riders take breaks while connecting on line using WhatsApp. They engaged in mutual aid by mapping available bathrooms for riders to use and parks to meet and organise. Using their shared language and ethnicity to build a network of organisers, they have organised strikes, rolling protests, and marches on city government to win access to bathrooms for “the simple right to piss” and more police investigations into stolen ebikes. Their organising provoked a response from Doordash which arranged for 200 restaurants to provide bathroom access. While only representing a small number of the restaurants served by the riders, the action demonstrated their power to move a multinational company through organising, issue framing, and media savvy.
As WJP recognised, the organisation began supporting these workers after they had already self-organised. “They’ve put in a ton of the work already, when we began working together they were already fired up politically, with a list of demands and everything. It’s almost, like, they’re already organised.” Rather than manufacture a campaign and focus it on the tactics and strategies that meet the prerogatives of union leadership, the Deliveroo and NYC gig workers first self-organised and then sought support from existing unions and worker organising non-profits to amplify the campaigns they launched while maintaining control over crucial decisions of tactics, strategies, and objectives.
What is also impressive about these campaigns is that they began to undertake their own workers’ inquiries into the the technical composition of the tech companies’ use of algorithmic management to control and exploit their labour, understand how the app was used to segment and stratify the workers, and then use what they learned to identify tactics and strategies to organise fellow gig workers to apply pressure at critical points of vulnerability where they could choke operations and extract concessions.
For example, Deliveroo workers realised that riders were divided by ethnicity and immigration status which also corresponded with whether they rode bikes or mopeds. Although they attempted to apply more focus on reaching the mostly immigrant moped riders they were less successful at initially recruiting them to takings strike action because they were also more likely to be the main earners for their family and found collective action too risky.
This is an identical model used successfully by RDU in California. RDU began in 2017 by organising drivers at designated waiting areas at Los Angeles International Airport (LAX) for ride-hailing car services. As large numbers of drivers congregated in these lots they came to build a community of workers who engaged in mutual aid and designed tactics to defeat the algorithm so that “the airport’s waiting lot fostered the makings of collective worker power.” Ultimately, they found that reaching large numbers of these majority people of colour and immigrant drivers depended on worker organisers building networks outwardly from committed workers.
RDU spent about $3,200 to organise the 2019 strikes buying targeted social media advertising, building an app for tracking organising, phone calls, texts, and in-person meetings to identify drivers who were uninvolved, supporters, activists, or potential organisers. It ran the campaign out of offices donated by the labour council, a labour backed group and the UCLA Labour Center. Over the course of four months from 2018 to 2019, they recruited and trained an organising committee, and doubled their membership with nearly 1,150 new members to organise two strikes and expand with two new chapters in San Diego and San Francisco.
Although RDU didn’t win its demands, its protests and strikes began to self recruit new drivers to join. According to Brian Dolber, who was involved in the organising campaign, “RDU’s production of a media spectacle influenced investor behaviour. RDU demonstrated that gig worker organisations can exercise considerable power over gig companies through strategic collective action.” The strike certainly got Uber and Lyft’s attention, both of which warned of the risk of unionisations, strikes and other forms of collective action in both prospectus filings for their IPOs. According to Ivan Pardo, who designed the app used in the campaign and was a key organiser, RDU’s “primary power is people power by having 1,000s of drivers supporting us by going on strike or making noise somewhere else”.
The app was a critical tool for their organising. Worker organisers used it to carry out a workers’ inquiry, collect contact information, and communicate directly using social media. Establishing their own “mass self-communication network” was critical to counter and drown out the daily push alerts Uber sends to drivers that must be swiped away to log into the app and begin working. According to Brian Dolber, RDU’s use of on-line and in person organising provided an organising model that “can help overcome the obstacles endemic to building a democratic organisation of a massive, unidentified, disaggregated, and fluid workforce that can exercise real power.” RDU received temporary backing from TWU in late 2019 to fund the social media ads and help with the AB5 lobbying campaign. RDU later worked in a tense coalition with the SEIU backed Gig Workers Rising in the unsuccessful No on Prop 22 campaign.
The experiences of RDU and Los Deliveristas Unidos in the US, and the IWGB Deliveroo and ARCU workers in the UK, demonstrate a vast capacity of workers to self-organise and unionise despite lack of status as employees and exclusion from labour law protections. With sufficient positional, strategic and disruptive power, workers who self-organise as an independent union, even when it is illegal to do so under threat of sanction under US anti-trust law, have moved city, provincial, state and national governments to change their legal classification. This demonstrates that self-organised workers’ struggles drive changes in the law and policy not the other way around, as is commonly assumed.
The current focus of much of union led gig worker advocacy, using the legislature and courts, has been to make gig workers normal employees while also normalising the very technology that controls their work. Why should we be concerned with this? Algorithmic management technologies are not unique to this sector but have become engineer Frederick Taylor’s digital stopwatch. In the UK and EU independent unions such as the App Drivers and Couriers Union (ADCU) are struggling to control the algorithm that controls their work. ADCU is using UK and EU data rights laws to fight for access and control of the data collected about their work, how it is used, stored and shared, and how it is used to discipline and control their work. Data control, transparency, and ownership is critical for workers to develop tactical forms of defiance and intransigence to defeat the algorithm as a strategy to assess, fragment, control, and discipline workers. The outcome of gig worker organising struggles over control of their work has implications that effect all workers as algorithmic management spreads like a virus into all types of work and and areas of life.
Prop 22 and self-organised struggle
There are several lessons to be learned from the Prop 22 defeat. First, despite being limited in scope and locally focused, gig worker organising has knocked the companies back on their heels. AB5 and other similar reform efforts, such as the stalled PRO Act, are a direct hit to the very survival of the gig companies, which like the publicly traded Uber and Lyft, had yet to post a profit and faced increasing pressure from shareholders and investors.
In fact, both Uber and Lyft warned of several risks from new laws, court rulings, and regulations. Their efforts to fight such reforms subjected them to continued legal losses, unionising, strikes, collective bargaining. Uber explicitly warned about international workers struggles including “driver dissatisfaction,” “protests” costing them hundreds of thousands of lost customers, “work stoppages or slowdowns” that create “interruptions to our business” and “difficulties in managing, growing, and staffing international operations.”
Second, the successes achieved by gig workers have been achieved by gig worker self-organisation. Their ability to reverse engineer changes to apps to devise new tactics and strategies to defeat and sabotage corporate strategies have many lessons for workers facing algorithmic management elsewhere. While their victories have been localised and small scale, they have successfully used these small victories to establish their credibility to other uncommitted drivers and bring them into organising to win concessions against giant multinational tech companies.
Third, the involvement of small militant unions and worker organising non-profits have played a constructive role when they allow the workers to lead rather than steer or use their members as a source of dues. They proved most valuable when providing organising expertise and resources to facilitate the building of local, national, and global networks in other cities, regions, states and even countries to take simultaneous action and amplify the impact of their strikes.
Lastly, the strategy of attempting to codify their small victories in the form of state and local laws and rulings by labour and courts have had mixed results at best. Such reforms have categorically improved the working conditions and pay of gig workers in the short term but present four long term risks and dangers.
Four risks of advocacy and lobbying strategies
The current focus on advocacy and lobbying shifts the terrain from a struggle over work to a struggle over law and policy. The power to disrupt work and withhold their labour provides workers their most effective form of power in the workplace, virtual or otherwise. Once they leave the workplace to fight on the terrain of representative democracy they are immediately at an unsurmountable disadvantage as corporations can unleash their flood of money to buy and shape laws, regulation, and court rulings, and failing that make laws directly by initiative as with Prop 22. Even in instances where the legislature, regulators, or courts come down on the side of workers, these short-term gains can immediately be sabotaged, delayed, defied, or transmogrified into their complete opposite.
The second risk is that gig workers are disempowered at work, where their strength ultimately lies. Because a disproportionate number of gig workers are low waged, work long hours or for multiple apps, immigrants, and people of colour, they are forced to fight on a terrain in which they are severely disadvantaged to achieve, defend, and expand these reforms. As a result, self-organised gig worker struggles have either been harnessed, usurped or redirected away from organising for power, which is what provided the leverage which was needed to achieve these reforms.
Many of these unions have very different, often incompatible, tactics, strategies, and objectives. In the US, TWU, SEIU, and IAM’s focus is on engaging in advocacy rather than organising campaigns with the objective of expanding their ranks of dues paying members. When that fails, those resources and the membership are channeled away from the workplace struggle into campaigns to win through establishment politics what they cannot or have failed to win at work. When this strategy proves unsuccessful, they pivot to sweetheart deals with employers to “recognise” them as the primary bargaining agent in exchange for “non-interference” and coalitional efforts to lobby local and state governments to expand subsidies for the sector or look the other way on regulation infractions.
The last risk is that regulatory and judicial victories can be turned into the means by which workers are ultimately defeated. This can be seen in the final days before AB5 passed when Governor Newsom facilitated an effort of the unions to settle with the gig companies on amendments to the law while leaving groups like RDU out in the cold. Similarly, after Foodora workers won at the labour tribunal in Canada the company launched a capital strike by disinvesting in the entire country and laying off the workers. The passage of the New York City laws was followed by the IAM collaborating with Uber to set up a company union. In Seattle, the city settled a lawsuit it was winning in exchange for gutting the law.
Workers will need to continually focus on beating back an endless series of roadblocks and impediments to be placed in the way of enforcing the law. The gig companies continue to refuse to comply with AB5 mandates.
The UK Supreme Court which found that because Uber controlled the driver’s work and pay the drivers are workers with employment protections and not self-employed. Following the ruling, Uber proceeded to change which hours of work are paid and entered into a recognition agreement with the GMB union opposed by the independent unions ADCU and UPHD. Contesting their obvious attempt to sabotage the authority of the court will be costly and time consuming. The companies can also retaliate with a capital strike, such as Foodora did when it simply disinvested from Canada to avoid the ruling of a provincial labour court. If such spatial fixes are not possible, then Uber’s effort to rush self-driving cars to market demonstrates that they are also pursuing technological fixes to the class struggle as well.
Relying on reform means relying on politicians
Relying on judicial, legislative and regulatory reform means placing workers’ power into the hands of elected and appointed officials whose interests lie with the companies rather than the workers. Whether it be changing classification of gig workers to employees in the US or enforcing EU treaty law to create a “rebuttal presumption of employment”, such regulatory changes do not alter the fundamental balance of power on the virtual shopfloor.
Relying on government officials brings with it all the dangers of the revolving door. My interview respondents all noted the support of the California Assemblywoman Lorena Gonzalez who sponsored AB5. There has also been much celebration of President Biden’s new Secretary of Labour Marty Walsh, a former building and construction trades council union official, and David Weil as appointee to head the department’s Wage and Hours Division, his second stint in the position, for declaring that gig workers should be considered employees. Expectations for Weil are high following his first term when he issued a directive classifying most gig workers as employees. But that directive was just as easily reversed by the Trump administration.