In Mexico, the Absent Inspector Is a Policy Choice, Not a Budget Accident

The contraction begins in the budget, and it is unevenly distributed. Between 2018 and 2023 the Dirección General de la Inspección Federal del Trabajo, which directs federal inspection policy, lost 12 per cent of its funding, while the Oficinas de Representación Federal, through which that policy reaches workplaces across the states, lost 79 per cent. The central body was trimmed. The offices that carry out the visits were cut to a fifth of what they had.

Personnel figures describe the same shortfall in another unit. Mexico has 1.1 labour inspectors for every 100,000 workers, against a global average of 5.8. The International Labour Organization's reference for industrialised economies is at least one inspector for every 10,000 to 15,000 workers, a ratio Mexico falls short of by a factor of six or more. At that density an inspectorate cannot sustain a presence across an industry. It can only ration a scarce resource, and rationing implies a rule for deciding where it will not go.

That rule shows first in what a visit is now for. The change is not in how many visits take place but in their purpose. In 2015, 60 per cent of federal inspections were ordinary, meaning scheduled, preceded by at least 24 hours' notice and part of routine supervision. For 2024 the projection was that 80 per cent would be extraordinary, conducted without notice in response to a complaint, an accident, an imminent risk or the need to verify compliance without delay. In the textile sector that transition has already taken place, with roughly 61 per cent of reported federal inspections extraordinary against 33 per cent ordinary.

The two categories do different work. An ordinary visit examines a workplace against which nothing has been alleged, which is how hazards and irregularities are corrected before they harden into violations of rights. An extraordinary visit answers something that has already happened, and in most cases something a worker has reported. Reorganised around the second model, inspection stops being a standing form of supervision. It becomes a mechanism triggered from outside, dependent on workers prepared to complain and on incidents severe enough to compel a response.

New instruments have been layered onto that diminished base. In March 2024 the Secretaría del Trabajo y Previsión Social published its Protocolo para la Inspección Laboral con Perspectiva de Género, applied across inspection visits and activated where violence, discrimination or harassment is detected. Its effectiveness depends on continuous specialised training and robust follow-up, both of them functions of the capacity the same period has been removing.

That removal is consistent with the economics the sector is held to. Mexican garment manufacture competes within global value chains on cost, and its position there is defended by keeping labour costs down, which places oversight among the burdens to be minimised rather than the functions to be resourced. What results is an apparatus responsible for an industry sustained by more than a million workers, its reach narrowing as the demands on it grow, and a probability of non-compliance ever being observed that falls year by year. The inspections that remain must still be directed somewhere. By someone.

Oaxaca illustrates how the boundary is used. Asked what inspections it had conducted in the clothing industry, the state's labour secretariat replied that it lacked competence over the sector and therefore held no information on it. Its response, issued as ST/SN/DJ/UT/13/2026, justified that position by citing Tesis XI.2°.32L, Registro 187185, a criterion of the Tribunales Colegiados de Circuito.

That criterion establishes the opposite of what the letter claims. It holds that while the textile industry, understood as the transformation of fibres such as cotton or wool, falls under federal authority, a factory making clothes is a matter of local competence, on the reasoning that garment assembly transforms no raw material and cannot be treated as textile manufacture for jurisdictional purposes. The authority declined the sector by quoting the ruling that assigns it. The vacuum rests on an administrative reading adopted to shed operational load, not on any silence in the law.

Oaxaca is not the exception but the legible case. Most institutions returned null or bureaucratic answers to the same request. Two states referred the matter to other areas of government, and two acknowledged having conducted no inspections in the sector at all between 2022 and 2026.

Four states reported substantive data. Baja California carried out 91 inspections indicating possible violations, of which 17 (18.6 per cent) reached a formal resolution, alongside seven concerning possible failings on gender equality and seven prompted by violence or harassment on grounds of gender. Querétaro conducted five extraordinary inspections and found a single workplace compliant, the anomalies in the rest referred elsewhere. Mexico City's 12 extraordinary inspections in the garment sector produced 10 sanctions and two acquittals. Tlaxcala found direct violations of training, safety and hygiene obligations in half of its 16 visits. Among the small minority that is watched, irregularity is the norm rather than the exception.

That watched minority is not the only place responsibility goes missing. The sector externalises production through intermediate suppliers, independent workshops and micro-units, a chain that blurs the legal obligations attaching to hiring, social security and working conditions, and simulated subcontracting persists despite the 2021 prohibition. An inspector who reaches a registered workplace may still find responsibility for the workers producing its goods distributed across parties who were never visited.

Beneath the dispute over competence lies a limit that no authority contests and none is empowered to waive. Inspection extends by statute only to registered economic units, which places informal workshops and home-based production outside its jurisdiction whichever government claims the sector. The withdrawal at state level meets a structural exclusion built into the rule itself, and together they put more than 80 per cent of the industry beyond any inspecting authority. Whichever level is answerable, the floor is the same.

Of the women who make up two-thirds of the clothing industry's workforce, 48.92 per cent work in micro-units without an establishment. That is the tier no inspecting authority is empowered to visit. A further 25.86 per cent work in micro-units that do have premises, which leaves roughly three-quarters of the sector's female workforce in the two smallest categories of economic unit. The pattern holds in the informality figures. In the fourth quarter of 2025, 76 per cent of women in the industry were in informal employment, against 53 per cent of men.

The overlap between the least protected workplaces and the most feminised ones is not coincidental. Women enter the fragmented segments of production in part because those arrangements can be combined with domestic and care responsibilities, which fall disproportionately on them, and home-based or intermittent work is what makes that combination possible. The same flexibility that allows the work to be done removes it from the registers through which the state identifies a workplace.

That removal repeats inside the workplaces that are visible. Men hold 10.13 per cent of managerial, supervisory and coordinating posts across the industry against 3.83 per cent of women, while the occupations requiring manual precision and repetition are heavily feminised: for every 10 men working as tailors, seamstresses and garment makers there are 34 women, and in embroidery and drawn-thread work, 82. The tasks assigned to women are the ones concentrated in the smallest units. That is where the sexual division of labour and the boundary of inspection coincide.

What the absent inspector would otherwise supply is an external check on conditions that workers are poorly placed to contest themselves. Women in the sector earn a monthly average of 5,387.38 pesos against 9,282.15 for men, and 89 per cent of those working without pay at all are women, figures that describe weak bargaining positions before any question of enforcement arises. Where a complaint is the only trigger for a visit, the complaint has to come from somewhere, and workers report withholding it out of fear of reprisal, out of experience with internal grievance channels that produce transfers rather than remedies, and out of distrust of the institutions notionally charged with protecting them.

Formal registration does not settle the matter either. In the visible tiers, wage under-registration understates the contribution base, strategic rotation interrupts the accumulation of seniority-linked entitlements, and workers frequently lack copies of their own contracts. Recorded working weeks reach 69 hours against a sector average of 44. Formality here reaches only as far as the paper it is recorded on.

The result is that a mechanism intended to correct the sector's inequalities operates in the same direction as them. Its remaining capacity is concentrated where employment is documented, wages are higher and men are proportionally better represented, and it withdraws from the units where women are concentrated and conditions are least observed. The workers with the least protection face the lowest probability that any authority will ever record what they are working under. A shortfall this size is not merely large; it has a direction.

On Paper and Out of Reach

The same design appears in the state's care infrastructure. The CECI model has been announced and specified, yet no centre operates; its daytime schedules do not match the night and rotating shifts women in the sector work, and its locations sit off their commuting routes. It exists on paper while its intended users cannot reach it, as inspection exists in law while the sector lies outside it. Rebuilding inspection around the industry's real shape would have to happen as its capacity contracts, which leaves open whether the distance between a right and its exercise narrows or widens.