Meituan's 2026 Winter Care Program covers more than 1 million delivery riders across 186 cities, according to China Daily HK on 5 November 2026. The company funds cold-weather medical coverage from 1 November through the end of March 2027. It is also introducing self-heating uniforms and opening service stations in 40 cities, offering hot water, meals, winter supplies and e-bike maintenance. The timing matters. The Ministry of Human Resources and Social Security (MOHRSS) draft rules on gig-worker protection closed public comment on 8 November. The question is whether voluntary welfare and binding regulation complement each other or crowd each other out.
The case for binding rules
The strongest argument for regulation is that goodwill is not an entitlement. A winter uniform can be withdrawn next season, and a medical scheme can be redesigned when margins tighten. China Daily quotes Meng Quan of the China University of Labor Relations saying rider protection is evolving "from the distribution of welfare supplies toward a more systematic approach." The scale also makes the case. Official data cited by Global Times estimated that workers in new forms of employment passed 200 million by the end of 2025, nearly 30 percent of total employment. A workforce that large cannot depend on the discretion of a few firms.
Meituan's own numbers show both the strength and the limit of that discretion. The company says it has paid a cumulative 3.6 billion yuan in occupational injury protection contributions, covering about 19 million riders as of end-August. That is a large, real commitment. But it is a company-reported figure, and it sits inside a broader public pilot rather than replacing one.
What the draft actually does
MOHRSS released the draft on 8 October with comment open until 8 November, per Sixth Tone. China Daily counts 54 articles and says six other departments joined MOHRSS. Xinhua describes it as the first time workers who do not fully fit an employment relationship have been brought into labor-law protection through a ministerial rule.
The design is graduated rather than a blanket reclassification. Workers who meet the legal test for employment get labor contracts. Platform workers who do not meet it get written agreements, with a three-party agreement where a partner company also manages them. The core provisions are concrete:
- Pay must be settled at least monthly, with a cycle of no more than one month.
- Workers on four continuous hours should get appropriate rest, and holiday work should be paid above the normal rate.
- Fines and other punitive measures may not be abused.
- Companies must respond to worker challenges within 48 hours, and decisions such as stopping dispatch or banning an account require human review.
- Where a partner company lacks legal operating qualifications, the platform bears employer responsibility, and multi-layer subcontracting is banned.
Several of these are sensible. A human-review requirement for account bans is a proportionate response to opaque algorithmic discipline, and it does not require any firm to disclose its source code. The anti-subcontracting rule targets a real accountability gap in which the party that profits is not the party that is liable.
Where regulation should be careful
The risk is not in the direction of the draft but in its calibration. Three points deserve attention before the final text.
First, holiday pay. The draft says it should be "higher" than normal pay but, as Xinhua's text summary shows, it names no multiplier. Vagueness here invites uneven local enforcement, which is the worst outcome for both riders and firms. A defined floor with room for platform-specific schemes would be more predictable.
Second, the line between employee and non-employee. The draft's middle category is pragmatic, since it avoids forcing millions of people who value flexible hours into fixed employment. It works only if the criteria stay clear. If courts or local regulators stretch the employment test, platforms will cut flexibility, shorten shift options or reduce the number of riders they take on. The riders who lose income first would be those who use delivery as a second job.
Third, algorithm oversight. Xinhua quotes commentators who say the scope is limited to labor rules. That limit is wise. Requiring explanation and human review for consequential decisions is workable. Broad algorithm audits would add compliance cost without clearly improving rider outcomes.
Keep the floor, preserve the ceiling
The stronger insight from the Meituan announcement is that competition has already pushed platforms beyond what any floor would require. Cold-weather medical coverage and heated uniforms are not mandated by the draft. A good regulatory framework should not turn such benefits into a compliance checkbox that firms provide at the legal minimum and stop improving. It should instead set a clear baseline and leave room above it, so that employers compete on conditions.
The occupational injury scheme shows how a shared floor can work. Global Times reports that the pilot began in July 2022 and by the end of June 2026 had reached 17 provinces and 11 platform companies, with cumulative enrolment above 29.9 million. Individual workers pay nothing, and platforms pay a per-order charge settled monthly. That structure spreads cost across order volume, avoids penalizing small platforms disproportionately and gives workers coverage that does not vanish when they switch apps.
The courts have also tightened the baseline. The Supreme People's Court's Labor Dispute Interpretation (II), effective 1 September 2025, holds in Article 19 that agreements or promises not to pay social insurance are invalid. For platforms that use partner companies, that raises the stakes of getting contracts right.
What to watch
The draft's value will be decided in the final text and in enforcement. Three signals matter: whether the final version quantifies holiday pay, whether the employment test is stated in terms workers and platforms can apply in practice, and whether the 48-hour response and human-review rules are enforced consistently across provinces. If they are, China will have moved from ad hoc welfare to predictable rules without freezing the flexibility that made platform work attractive. If the final text is vague, firms like Meituan will keep filling the gap with voluntary programs, and workers at smaller platforms will be left without them.