Wired | China’s Gig Workers Are Challenging Their Algorithmic Bosses

PHOTOGRAPH: QILAI SHEN/BLOOMBERG/GETTY IMAGES

Food delivery drivers are using platforms’ data-powered systems, mass WeChat groups, and unofficial unions to fight unfair conditions.

HUANG HUI CAN sum up his six-month stint as a gig worker in Shanghai in one word: “exhausting.” A PhD candidate at King’s College in London researching the lives of couriers, Huang soon realized that he could not keep up with his colleagues, many of whom delivered food on their electric motorbikes for 12 hours a day.

Working as a courier was a risky business. He says he saw six traffic accidents involving couriers during his stint, one of which was fatal. “It was really shocking for me,” he says. The danger caused him anxiety, Huang says, and completing 50 to 60 orders a day was a “huge task.”

Huang is among a cohort of researchers studying how algorithms control gig workers' lives. Over the past years, China's platform economy has engulfed almost a quarter of the country’s labor force, with an estimated 200 million people working in “flexible” employment. The industry has also attracted criticism, not just on labor rights but also because of the data-driven pressures imposed by major food delivery platforms Meituan and Ele.me to keep the 469 million people ordering food online happy.

Companies leveraged algorithms to force couriers to work faster and more efficiently, squeezing them with tighter delivery times. Since these changes were introduced, reports of traffic accidents involving couriers have grown, and couriers have started to rebel.

Some are gaming the platforms to gain higher wages; others are banding together to turn the data-driven approach to performance assessment against their bosses. Many more are forming unofficial unions with the help of social media platforms such as WeChat and Douyin, China’s version of TikTok.

Couriers have set up WeChat groups of up to 500 people to exchange information on places in the city where it’s difficult to deliver—such as large buildings with multiple lifts or gated communities. These are treated as “no-fly zones,” where couriers refuse to go, according to researchers Tiziano Bonini from the University of Siena and Zizheng Yu and Emiliano Treré, from Cardiff University, who have been studying gig work in China.

The couriers “know it is impossible to deliver in the time expected by the platform,” says Bonini. “So they organize these kinds of collective rejections until that order comes back with a higher price.” 

In busy Chinese cities such as Shanghai, some delivery work is organized around “stations.” These are hubs for independent subcontractors that organize delivery work from different restaurants, guaranteeing couriers a steady stream of work but taking a fee from the platforms they work for.

“There are two types of delivery drivers,” says a courier at China’s largest food delivery platform, Meituan, who asked to remain anonymous. “Some are part of stations, and they will get more orders. Then there are drivers who go it alone, and they are freer but get fewer orders.”

This situation gives gig workers bargaining power—if they refuse to work for lower wages and a station loses orders, especially at peak lunch hours, the station’s rating plummets and receives less business from platforms as a result. In the meantime, they can continue working independently.

“Even minor disruption in the form of these very small-scale collective actions can bring the station-level managers to the bargaining table,” says Eli Friedman, an assistant professor of international and comparative labor at Cornell University.

Although China bans independent unions and labor strikes, that hasn’t stopped gig workers from organizing unofficially. Many food delivery riders find opportunities to band together thanks to other types of algorithms—those that help them find like-minded people on Douyin, where gig workers share experiences and advice.

“We do support each other,” says another gig worker at China’s second-largest catering platform, Ele.me, who requested anonymity. “It definitely helps some, but not that much.”

The most famous informal union is Knights League, which was set up in 2018 for riders to share tips with each other. The most famous gig activist, Chen Guojiang, also known as Mengzhu, reportedly managed 16 WeChat groups reaching over 14,000 delivery drivers. Chen was arrested by Chinese authorities in March of last year on charges of “provoking trouble,” after he tried to mobilize strikes among fellow delivery couriers in Beijing. It is unclear who is running the protests since his arrest and subsequent release.

He would tell gig workers “how to support each other, because everyone is weak, but if they can form a link, some kind of solidarity, then maybe they can ask more from the platforms,” says Yu.

Time constraints are among the common complaints among couriers. Between 2011 and 2020, China’s online food delivery industry ballooned from $3.4 billion to $105 billion. But as the industry expanded, more couriers joined, driving down prices, while platforms kept lowering delivery times.

At the same time, reports of couriers involved in traffic accidents as they raced to make deliveries became more common. One report from Shanghai’s traffic police showed that in the first six months of 2017, a courier was involved in a deadly traffic accident roughly every 2.5 days.

These algorithmic changes have even involved making couriers compete against each other, as if they were in a game. Couriers on Meituan and Ele.me are ranked by performance, similar to the rankings in China’s most popular mobile game, Honor of Kings. Riders are awarded titles such as Bronze, Silver, Gold, and Kings, which impact their status and income. Companies often organize competitions to encourage couriers to take more orders—inspiring some riders to game the algorithms right back, by faking orders to improve their standing.

But perhaps the most damaging trick a platform can pull is deleting a rider from the system entirely. An investigation by Beijing Zhicheng Migrant Workers Legal Aid and Research Center in January found cases in which couriers’ data would inexplicably vanish after an accident involving a rider. Gig workers need to use order data on the app to prove that they were injured at work, but if they cannot access the app they are unable to provide any evidence to back up their claims.

Public scrutiny over gig-work conditions was heightened after a slew of viral incidents, some of them deadly. In 2019 a driver carrying meals for Meituan, the country’s biggest food delivery platform, stabbed a clerk during a dispute, sparking debate on couriers’ time constraints.

Other high-profile cases followed, including the dramatic protest of a courier working for Alibaba-owned Ele.me, who set himself on fire over $770 (US dollars) in docked payments.

Meituan spokesperson Xiang Xi told WIRED that it has been improving the openness and transparency of its order dispatch system and easing its delivery time limits. Ele.me did not respond to a request for comment, but it has been introducing similar measures in the years since the public backlash against the two companies started.

Delivery drivers are often hired by a labyrinth of contractors in different jurisdictions, making it difficult to find a boss to help or to pay for damages if something goes wrong.

The majority of food delivery drivers are migrant workers from poor and rural areas, saving money to send back home or for personal development. As migrants, they also lack access to social welfare such as health, unemployment, and work insurance, which in China are tied to residency, or hukou.

Many of them were once employed by the country’s shrinking industrial sector, where long hours of repetitive work are the norm. For these former factory workers, gig jobs offered better pay but also flexibility and autonomy. There’s no boss to control them except the algorithm in the palm of their hand, so it feels like entrepreneurship, says Huang.

“I work on my own, and I am free to work overtime if I want to or work less when it suits me,” the Meituan courier says.

This might soon change. On March 1, China introduced an algorithmic law, affecting not only how ecommerce platforms recommend products and social media serves up content, but also how platforms allocate orders, pay salaries, and hand out rewards and penalties for gig workers. Over the next three years, authorities are also planning to set up evaluation teams tasked with making algorithms “fair and transparent.”

Pressure to ensure better conditions for “delivery brothers,” as they’re often referred to in China, increased during the Covid-19 pandemic in 2020, as they became essential workers.

“Delivery drivers suddenly became heroes, and we were treating them like garbage,” says Kendra Schaefer, a partner at Trivium China, a policy analysis consultancy. “There was this public outcry to treat them better.”

The government signaled its stance on gig work with a viral two-minute video, where a local Beijing bureaucrat was shown working as a delivery driver, earning just 41 yuan ($6), enough to pay for a modest meal but not much else.

Labor issues are just one aspect that regulators are hoping to fix. That same week the government video was released, Meituan joined the ranks of tech companies under scrutiny that hit after years of unregulated growth. Like Alibaba, which received a record $2.8 billion fine in April last year, Meituan was fined $530 million for abusing its market position in October. The crackdown has so far wiped out trillions in market value and imposed heavy anti-monopoly penalties on companies operating in the tech sector.

Companies received guidelines last year calling for better compensation and benefits for China’s flexible workforce.

The Chinese government has also asked gig workers to join China’s only legal union, the All-China Federation of Trade Unions. Unlike most unions, the ACFTU is not organized bottom-up. It answers solely to the Communist Party of China. And while it doesn’t help workers with collective bargaining or strikes, it may play a role as a feedback mechanism for platform algorithms. Calls have been mounting across the country, including in ACFTU branches, to give more voice to platform workers on how algorithms are made. 

Although the Meituan courier says that the algorithm hasn’t affected their work much, the changes could bring increased costs for food delivery platforms.

“Given the fact that the operating profit margin of the food delivery platform is only 3.3 percent, this will be a significant challenge for Meituan,” explains Jamie Chen, an analyst at research firm Third Bridge. Some of the costs will be transferred to consumers and suppliers, but the platforms will have to shoulder the rest.

Companies are likely to make the smallest adjustments possible, Schaefer adds. This would mean not only keeping gig worker employment precarious but also controlling algorithms. 

The main issue is not how algorithms control workers, says Huang. It’s about who writes the algorithm rules, protocols, and policies. “Algorithms are just a tool used by people,” he says.

Additional on the ground reporting by Kyle Mullin.