Shanghai’s eight departments jointly issued implementation opinions on protecting workers in new employment forms, requiring platform companies to optimise their algorithms and banning the use of the strictest algorithm as a performance bar with fines standing in for management. The debate over ‘the algorithm that traps delivery riders’ flared up again.
Su Yong, a professor at Fudan University’s School of Management, argues the issue underneath is whether high-tech, information-age management can hold on to people-centred principles.
The problem is not the algorithm
China’s flexible workforce is nearly 200 million, about a quarter of them couriers attached to platforms — platform companies are now enormous employers. The form creates jobs and convenience, but it also creates protection gaps workers have never faced before.
Riders bound to delivery systems run faster and faster to meet targets, traffic violations pile up, accidents follow, and exhaustion is normal. As algorithms get more precise, some platforms set assessment bars with almost no slack, leaving riders no margin for a red light, a refused handover, a spilled order.
Algorithms themselves are a genuine advance: management by algorithm is modern, scientific, efficient, and it cuts waste. The problem appears when a company applies the algorithm to people and calculates to the bone — every second strict, no allowance for the accidents of the road. At that point it is Taylorism with better tools.
Taylorism, rebuilt
In 1911 Frederick Taylor published The Principles of Scientific Management, which revolutionised management by intuition. Efficiency rose, and so did a problem: workers’ behaviour alienated under time-and-motion analysis, which allowed only the few motions the analyst deemed necessary.
Chaplin’s worker in Modern Times, driven half-mad by endless screw-turning, ends up tightening traffic policemen and fire hydrants on the street. Behavioural science — Mayo and after — overturned the theory: managing people is not managing machines, and human needs are part of the system.
A rider on the street is no different from a worker on the line, and faces more risk and uncertainty. Platforms, as the employing entity, should use algorithms for what they are good at and leave margin for the human cases: optimise the algorithm, then add human management on top — value rationality alongside instrumental rationality.
Large platforms all publish fine-sounding values. The test is whether those values reach the algorithm, or stay on the wall.
Surveillance software and where to draw lines
The questions are not new. More than a decade ago, Su discussed employee-monitoring software in MBA classes: one student’s company installed a tool that let the boss read every email sent through the company mailbox.
The tool itself is neutral; the use is not. Su’s three points: a company mailbox is a work tool, so authorised managers may monitor it; if the company uses such software it must tell employees clearly, in the handbook or on the mailbox itself; and leaders should not read the records except when truly necessary — frequent checking puts staff on the other side of a trust line and corrodes the culture.
Flexible work itself deserves defence: it suits platforms, and younger workers genuinely value the control over their own time. What matters is completing the basic protections for a new employment form, and finding, beyond the algorithm, the management methods where human judgement and human care do the work the math cannot.
The deeper question is the point of it all: is the goal a workforce that is efficient and frazzled, or one that lives better and creates more? If the latter, then efficiency tools and human margin both belong in the design.
Republished for information. The original text has not been independently verified and does not necessarily reflect our views.
Want to talk through what you just read?
We work end to end, from product design to launch. Tell us about your scenario and we can start there.
Shenzhen XZS Technology Co., Ltd. · Tel 0755-82494862 · 13760424861