Mexico’s platform-work reform reached 1,772,296 people with at least work-risk protection in August. Only 199,454 earned enough in the month to qualify for all five branches of the Mexican Social Security Institute, or IMSS.
That is roughly one in nine. It is also the denominator that matters.
The government’s larger numbers describe activity accumulated over time. The Labor Ministry reported 9,853,538 platform-worker registrations from January through June and 1,052,653 recognitions as subordinate workers with social security. Those are period totals. They do not establish that 1.05 million distinct people held comprehensive coverage at once, or kept it from one month to the next.
The useful test is persistence: how many regular platform workers clear the income threshold repeatedly, keep contributions flowing and use the benefits attached to formal coverage.
A monthly stock, not a cumulative victory lap
The reform is built around a monthly reset. Under IMSS’s operating rules, platforms register people when they start activity. After each calendar month, the platform determines who met the legal test. Within the next five days, it cancels the registration of those who did not qualify and adjusts the contribution salary for those who did.
A person can therefore appear in records across several months. A worker can leave the full-coverage group, re-enter it, or work through more than one app. The government has a unique social-security number for each person, but its first-half release does not publish a deduplicated cohort or explain how repeated platform records were consolidated.
August’s monthly stock is clearer. The IMSS employment release says 199,454 people crossed the net-income threshold through 18 platforms. That was 48,446 fewer than in July, a 19.5% monthly fall. The reported change implies 247,900 people had full coverage in July. IMSS had described about 237,000 in June.
The fall is real. Its cause is not yet identified.
The threshold separates two kinds of protection
The labor-law amendment treats a platform worker as a subordinate employee for the month when net earnings reach at least one monthly minimum wage in Mexico City. Net income excludes tips and applies a cost factor to gross platform earnings.
The Labor Ministry’s calculation rules set maximum exclusions of 36% for a four-wheel motor vehicle, 30% for a motorbike and 12% for non-motorized transport or no vehicle. The same gross billings can therefore produce different qualifying income depending on the equipment used.
Workers above the threshold enter all five IMSS insurance branches: sickness and maternity, occupational risk, disability and life, retirement and old age, plus childcare and social services. Workers below it are classified as independent for that month. They can join the voluntary scheme, but they do not receive the same automatic comprehensive package through the platform.
They are not unprotected. The August IMSS release says all 1,772,296 people covered by the reform retained occupational-risk protection while working. Below-threshold workers are eligible for medical, surgical, pharmaceutical and hospital care after a work accident or occupational illness, plus incapacity pay at 100% of their registered wage while the incapacity lasts.
That distinction matters. The reform has widened accident cover across a large, intermittent workforce. The open question concerns comprehensive protection, not whether the system provides anything at all.
August could be churn, seasonality or design
The three-month path offers several explanations. IMSS’s director said in July that June platform activity had likely received a boost from the World Cup. July’s higher full-coverage count and August’s reversal fit a post-event slowdown. They do not prove one.
Workers can also cross the threshold in one month and miss it in the next because their hours, demand, bonuses or choice of app changed. The law explicitly calls the work flexible and discontinuous. It automatically ends the employment relationship after 30 consecutive days without activity, then starts a new relationship if the worker later qualifies again.
August still had 66,276 more fully covered platform positions than a year earlier, an increase of 49.8%. That comparison is encouraging but carries a low-base problem: August 2025 sat near the start of the pilot. The monthly drop and annual rise can both be true without telling us whether established workers retained protection.
Aggregate totals cannot separate seasonal demand from worker churn. They cannot show whether the same people stayed covered in June, July and August. They do not reveal how many people had simultaneous registrations across platforms, whether earnings from several apps were combined, how many contribution months reached individual accounts, or whether eligible workers registered their families and used health, disability, childcare or pension benefits.
Publish the cohort, then judge the reform
The next useful release needs a cohort table, not another cumulative headline. IMSS can report unique social-security numbers with risk-only and full coverage, then show how many retained full status for three and six consecutive months. It should disclose exits, re-entry, multi-platform registration, paid contributions and benefit claims.
Platforms and policymakers would then see different failure modes. A high re-entry rate after August would support the seasonality explanation. Persistent loss among workers with regular hours would point toward unstable earnings or a threshold design that withholds broad coverage too often. Large gaps between registrations and paid contributions would expose an enforcement problem. Low benefit use despite stable eligibility would suggest an access or awareness problem.
Mexico has already answered the first question. The reform can place a wide platform workforce inside work-risk insurance. It has not yet answered the harder one: whether regular workers can keep comprehensive social security when a busy month ends.
Discussion
Sign in to join the discussion.
No comments yet. Be the first to share your thoughts.