Indonesia suspends 1,999 free-meal kitchens over missing hygiene certificates

Indonesia’s National Nutrition Agency has taken 1,999 kitchens off the country’s free-meals programme because they never registered for mandatory hygiene and sanitation checks. Agency chief Sudaryono announced the suspension on 8 September. As of 7 September the programme still listed 27,022 kitchens.

The kitchens must apply to the health ministry for a hygiene certificate before they can cook again. Sudaryono, who uses one name, said the certificate “is a mandatory requirement for operating kitchens that provide the free meals programme.” The agency will inspect each site before it decides whether to reopen it, shut it for good, or hand the case to another authority.

The order is the sharpest operational cut since President Prabowo Subianto launched the scheme in January 2025 as his main campaign promise. The programme, known in Indonesian as Makan Bergizi Gratis, was built to feed schoolchildren, pregnant women and toddlers at a pace few countries have tried. The kitchen count climbed past 27,000 in under two years. That speed is now the problem.

Charles Honoris, a lawmaker who oversees the programme in parliament, said health-department figures put the number of people sickened by free meals at least 50,000 by 1 September. Last week alone more than 2,000 students and teachers in several provinces fell ill after eating the food. The agency has blamed two recurring faults: meals stored too long, and meals delivered late.

Those numbers sit on top of an earlier leadership crisis. In June the Attorney General’s Office detained then-agency head Dadan Hindayana and two deputies. A moratorium on new kitchens followed. The finance ministry had already trimmed the 2026 budget from 335 trillion rupiah to 268 trillion rupiah. Officials inside the agency later discussed a further cut of about 40 trillion rupiah, or more than $2 billion, and said only about 21,000 of the contracted kitchens were actually needed.

Sudaryono’s 1,999-kitchen list is therefore both a food-safety step and a way to shrink a network that grew faster than inspectors could visit it. Seven percent of the fleet is now idle. In districts where six kitchens already cover the school list, a seventh that never filed paperwork is an easy target. In districts with one kitchen and a long dirt road, a suspension leaves children without a meal until a neighbour kitchen can take the load.

The hygiene rule itself is not new. Indonesian law already requires a sanitation certificate for large kitchens. What changed is enforcement. After the June arrests and the poisoning clusters, the agency stopped treating registration as a formality. Kitchens that never entered the health-ministry queue are now treated as non-compliant rather than merely late.

Prabowo has not cancelled the programme. He has also not restored the original budget. That leaves the agency in a squeeze. It must keep feeding a national constituency that voted on the promise, and it must show that the food will not send children to hospital. Suspension without a published reopening timetable risks both outcomes at once: empty plates in some schools, and a political argument that the state cannot run a kitchen.

Honoris has asked for a public list of the 1,999 sites. The agency has not released one. Until that list appears, parents cannot tell whether their school’s kitchen is among the shut sites or merely late with a form. District education offices are already calling neighbouring kitchens to ask if they can absorb extra trays.

The investigation Sudaryono promised will decide more than certificates. If inspectors find storage rooms without refrigeration, or cooks without basic training, some kitchens will not come back. If they find only missing paperwork, the 1,999 could shrink quickly. Either result will reset the map of who eats under the programme this term.

For a scheme sold as the largest free-meal effort of its kind, the test is now simple. Can the state certify a kitchen as fast as it can open one? The 1,999 closures are the first hard count of that gap.