Indonesia Is Burning Again in 2026. Why Does the State Keep Acting Surprised?

I am tired of writing this article. Not bored, tired. Tired the way you get when the same disaster arrives on the same schedule, the same officials say the same sentences, and the same people end up breathing the result. 2015, 2019, 2023, and now 2026. The numbers change, the ministers change, the agency names change. The smoke does not. This piece is opinion, and I will not pretend to be neutral about children walking to school in orange air. But the anger below is attached to evidence, and I have tried to be honest about where the evidence is weak. Figures are as reported up to November 2, 2026, drawn from official statements, local and international media and NGO analyses. They often disagree with each other, and I flag that where it matters.

Where things stand

Here is the snapshot, before the argument.

Indicator	Reported figure	Source
Official burned area, January to July	202,004 ha, up about 69% on the same period of 2025	Good News from Indonesia
Burned in July alone	roughly 94,500 ha	CNBC Indonesia
Burned area cited by AP on August 21	about 285,000 ha	AP
Satellite monitor, first nine months	more than 1.9 million ha	AFP via The Star
Largest provinces (official, January to July)	West Kalimantan about 38,300 ha, South Papua about 25,800 ha	AFU
Hotspots detected nationally in one day (August 28)	18,065	Metro TV
Hotspots in East Kalimantan in one day (October 30)	1,135	Media Indonesia / Metro TV
Acute respiratory infections (ISPA), six provinces, to October 12	277,075 cases in 63 districts	Rakyat Merdeka
Flights disrupted	2,458	Bisnis
Estimated economic loss	Rp 39.29 trillion to Rp 123.1 trillion	KLH

Burning has not stopped. East Kalimantan's one-day jump was reported on October 30, and Palangka Raya extended its emergency status to November 9, per Bisnis.

The numbers do not agree, and that is a problem too

Look at the burned-area rows. The official SiPongi figure is around 202,000 ha. AP cites about 285,000 ha. A satellite monitor cited by AFP says 1.9 million ha. These are not necessarily contradictions, because ministries verify burned area one way and satellite monitors measure it another. I cannot tell you which is closest to reality. What I can say is that a country facing a crisis of this size should not leave citizens guessing whether the burned area is 200 thousand hectares or nearly ten times that. An expert quoted by Katadata Green in August noted that El Niño years in his experience pass 1 million hectares, and that the 203 thousand then on the books was not the peak. Official figures can lag. The honest response is to publish the methodology and the gap, not to quote the smaller number and move on. One more caution: hotspots are not confirmed fires. A hotspot is a heat detection. It is a strong early signal, and groups like Walhi say their filtered data is high-confidence, but not every dot is a blaze.

Nobody can say this was a surprise

This is the part that makes me angriest, because fire season is the rare disaster that gives months of advance notice. Here is what was on the public record before the smoke reached cities.

February: KLH reported 32,637 ha burned in the first two months, with 3,609 hotspots, triple the same period a year earlier.
Late March: Walhi counted 1,351 hotspots in and around the concessions of 15 companies, and called the repetition proof that governance had not improved.
March 31: Auriga Nusantara reported 433,751 ha of deforestation in 2025, the highest since 2016, and warned that a repeat of 2015 could not be ruled out.
April: Officials promised there was no budget ceiling for fire response (Suara Kalbar). KLH formed a task force and said it was preparing civil suits against 35 companies.
June: A "super" El Niño officially arrived, per Mongabay.
End of July: BMKG counted 5,019 hotspots, a pace described by one outlet as faster than the 2015 El Niño year.
August and September: The haze reached Malaysia, Singapore and the Philippines, with Manila's air rated acutely unhealthy between August 30 and September 2.

When the haze was at its worst, a member of the House of Representatives rebuked the disaster agency in words I will paraphrase gently: you knew the dry season was coming, so where was the preparation? His name is Marwan Dasopang and the exchange is in IDN Times. That question is the whole article. I will just attach more evidence to it.

The fires are not a mystery: look at who holds the land

Every year the first public narrative is about careless farmers, and every year the data points somewhere less comfortable. President Prabowo's own government has now said as much. The forestry minister described the devastation as largely man-made, per Reuters, and the environment ministry's own verification found:

almost 85,000 ha of concession area burned and 335 concession holders with hotspots as of August 22 (Katadata);
in the concessions of 57 companies alone, 34,636 ha burned, about 95% of it oil palm plantations and most of the rest sugarcane (Katadata).

Outside analysts point the same way. Counter Point's analysis, reported by Mongabay Indonesia, counted 120,638 fire points in Kalimantan since August 1 inside 276 forestry concessions. Then there is the detail I cannot get past. Forestry ministry officials, as reported by Katadata, said peat hotspots in the priority provinces were often found on land with no active permit or on former concessions, effectively open access. The ministry's enforcement chief added that fires also hit dormant concessions where no activity was under way, so nobody fought the fire. A licence on paper, no guard in the field, no water on the peat, and a burning landscape. That is not nature. That is administration. Now the response from the industry side. Indonesia's forest concession association agreed that hotspots can occur in members' areas but argued, in the chairman's words to Katadata, that nobody should be blamed. I read that sentence three times. Nobody should be blamed for 85,000 hectares of burning concessions and 283 thousand sick people? Fine. Then who carries the licence? Enforcement has also run aground before. Mongabay's July reporting on Riau found civil society groups saying corporate enforcement was not working, and the September Mongabay report cited above notes that cases against five forestry concession holders over the 2025 fires were still stuck with the police.

Singapore is doing the work

On October 16 and 19, Singapore's National Environment Agency announced preventive notices to ten Indonesian companies under its haze law, seven in Sumatra and three in Kalimantan (AP). It also asked ten Singapore-registered companies for information (Bernama). Two caveats, because fairness matters. A notice is not a finding of liability. And the Straits Times, via Asia News Network, reports that three of the firms are linked to APP's supply chain, the same three flagged in 2015, while APP says it keeps a strict zero-burning policy across its supply chain. Singapore's own minister has also admitted that enforcing across borders is hard (Bernama). My opinion: when a neighbouring state is naming suspect concession holders in a legal notice, it tells you the home system has left a gap that someone else is filling.

The state's own projects are on fire too

This is the section the government least wants to hear, so I will be careful with attribution. Walhi and Greenpeace Indonesia reported large hotspot clusters on and around former and current food estate land. As Katadata reported in September:

about 77 thousand hotspots, 55% of the national total in that analysis, were in Central Kalimantan, and about 41% of those sat in former food estate areas;
in South Papua, of about 4,800 hotspots, almost 91% fell inside the food and energy project area.

A later Katadata piece on Merauke cited Walhi's count of 807 hotspots between January and August in the food and energy zone, with 674 in areas administratively under direct government management. Zoom out. Auriga's 2025 deforestation analysis found about 78,000 ha cleared for food production, and the programme has allocated 20.6 million ha of forest land, of which Auriga identifies 8.8 million ha as natural forest (Arab News, The Star). The government's side deserves its hearing. A deputy chair of House Commission IV from Golkar argues the Merauke food estate is a long-term answer to global food insecurity (Kontan). And hotspots in a project zone do not prove the project lit them. But here is my argument. If a state decides to clear land at industrial scale, on drained and fire-prone ground, in the middle of a forecast super El Niño, it has chosen the risk. It does not get to treat the resulting smoke as an act of God. The part of the burn that sits on government-managed land is the part where the government cannot say "somebody else's concession."

Peat: the agency that ended, the agency that returns

Indonesia built a Peatland Restoration Agency after the 2015 disaster. It later became BRGM, and its legal term ended on December 31, 2024, as Fortune Indonesia noted. Restoration duties were split among ministries. Within months the warnings came. Madani Berkelanjutan found 26,761 ha of peatland burned in July and August 2025 and asked who now owned the job (Kompas). That was without a major El Niño. Then 2026 happened. On September 18, according to The Star, President Prabowo accused unnamed critical experts of being "so smart that they become idiots". On the following Thursday he held a session with 22 experts, and out of it came a decision to create a new peat and mangrove management body, BPGMN, under the president (Asia News Network, Katadata). Is a new agency bad? No. Peat is the core of the problem and I would rather have a body than a vacuum. Pantau Gambut estimates that around 16 million of Indonesia's 24 million ha of peat are becoming more vulnerable, per Good News from Indonesia, though estimates of total peat vary by source. But look at the pattern Madani described: create an institution in a crisis, dissolve it when the pressure fades. Pantau Gambut notes the earlier restoration target was cut from 2.6 million ha to 1.2 million ha, and Greenpeace says BRGM's results were far from a success (Greenpeace Indonesia). Experts quoted by Mongabay say the real test is whether the new body can enter concessions and government projects, and warn the fight over turf between ministries will be hard. A body that cannot enter the burning land is a logo. I will say the quiet part out loud. Disbanding the peat agency and, within two years, rebuilding it in the middle of a haze emergency is not strategy. It is amnesia that has been given a letterhead.

Money: numbers that do not add up, and a corruption allegation

Budget claims in this crisis have been loud and mutually incompatible. On one side, the disaster agency's chief said in September that efficiency cuts only touched routine spending and that disaster funds had no ceiling. On October 20 the finance ministry said BNPB's budget was not cut, citing Rp 2.23 trillion plus about Rp 750 billion in ready-use funds. On the other side, academics and media citing Kompas described BNPB's allocation falling from Rp 4.91 trillion in 2024 to Rp 2.01 trillion in 2025 and about Rp 491 billion in 2026 (UMY, IDN Times). I cannot reconcile these, possibly because they measure different things. But I notice that citizens in a smoke emergency should not need a forensic accountant to learn how much the state spends to protect their lungs. Two further points. A member of House Commission VIII said most of the agency's budget leans towards recovery rather than prevention (IDN Times). And Walhi's Uli Arta Siagian lamented how thin the emergency funds for fire mitigation were in 2026; the Straits Times report set that against Rp 335 trillion earmarked for free meals out of a Rp 3,843 trillion budget (The Star). Free meals are a legitimate programme. The comparison simply shows what was treated as a priority. Then the worst detail. Investortrust reported that the Corruption Eradication Commission uncovered an alleged corruption case involving BNPB's disaster ready-use funds in the middle of the haze (Investortrust). I will stress the word alleged and I do not know its scope. The coordinating minister said unfinished fire handling was not only due to it. The agency also received a new chief on October 15, per Sindonews. I found no reporting linking the two events and I am not suggesting a link. But the combination of a contested budget, an alleged fund scandal and a leadership change is a bad moment for anyone to ask the public to simply trust.

Enforcement: loud on individuals, quiet on bosses

To be fair first: this year's enforcement is larger than I expected.

The police had 341 fire-related reports by the end of September, and corporate suspects numbered 4 on September 30 (iNews). By October 29 Kompas reported 20 corporate suspects, four of them foreign-capital companies (Kompas).
The forestry ministry had frozen the permits of 32 companies by October 21 and was processing 120 cases (RMOL).
KLH had sealed 21 businesses by early September (Investortrust).
The president said there would be "no mercy" for corporate offenders (Metro TV), and the forestry minister said enforcement must not be sharp downward and blunt upward (RMOL).

Now the cold water. First, timing. As of Mongabay's reporting on August 30, no company had yet been named a suspect (Mongabay). The corporate numbers rose only after the smoke crossed borders. That looks like a response to embarrassment more than a deterrent. Second, the ratio. Mongabay noted hundreds of individuals had been named suspects while corporate cases lagged, and Greenpeace's Asep Komarudin argued that success should be measured by whether corporate fines are actually executed, burned areas are restored and repeat fires stop, not by how many individuals are arrested (Mongabay). Third, history. In 2019, the authorities sealed 42 company sites, and Walhi complained at the time that sealing had no deterrent effect (Suara). Seven years later the headline is a bigger number of freezes. I have not found reporting that the 35 civil suits announced in April have produced judgments. If readers know of one, I would genuinely like to see it. Fourth, the legal theatre. The president asked ministers to draft a proposal to classify these fires as "ecological terrorism" (KLH). A judge writing in Dandapala argued that existing Indonesian law already gives a strong basis to prosecute corporations (Dandapala). Meanwhile Walhi warned that the new presidential instruction on burning could squeeze customary practices faster than it constrains corporations (NU Online). A tougher label does not fix a weak habit of using the laws already written.

"We can still cope": the national disaster question

The Coordinating Minister for Political and Security Affairs, Dudung Abdurachman, said on October 12 that the fires were not a national disaster because the government could still handle them. He repeated it on October 19, saying the country was still able, and added that a national declaration could bring foreign assistance that would need technical coordination, including fuel distribution (Rakyat Merdeka). He also asked the public not to be swayed by claims that the government was doing nothing (Investortrust). The logistics worry is not absurd. Foreign help does need coordination. But look at what is being weighed. Against it sit hundreds of thousands of respiratory infections, schools closed in several countries, and a crisis that has already travelled across a region. The University of Indonesia international law professor Hikmahanto Juwana argued the government should declare it, and a House member called for a national health emergency (Sindonews via RCTI+). Walhi's Uli Arta Siagian argued that national status would shift the burden from provinces with limited capacity and unlock emergency budgets (Mongabay). The courts may weigh in too. A coalition including AMAN's West Kalimantan chapter filed a class action against the president and other officials over negligence in the fires, case number 292/Pdt.G/2026/PN Pontianak, with national disaster status among its demands (Katadata). I do not know how it will end. I do know the precedent: in the 2016 Palangka Raya case, the court found the government had acted unlawfully, and the ruling was upheld on appeal (Mojok). A decade later, citizens are in court again about the same thing. And the neighbours are not asking politely anymore. Singapore's 24-hour PSI hit 230, Johor closed schools for five days from October 19 (Rakyat Merdeka), Timor-Leste closed schools for three days (Rakyat Merdeka), and Singapore's Coordinating Minister for National Security, Shanmugam, called it "disappointing" that an Indonesian official had belittled the crisis with a remark that Singapore gets oxygen from Indonesian forests (AP). Indonesian outlets also ran headlines about a remark by a senior communications official on smoke and oxygen drawing foreign media attention; I could not confirm from the sources I found that this is the same remark. Either way, a remark about oxygen while people across the region were choking on smoke is exactly the kind of sentence I mean when I say the tone is wrong.

What the haze costs people

Policy debates blur into abstractions, so here are the concrete things the smoke did.

Illness: ISPA cases reached 277,075 across 63 districts by October 12, led by South Kalimantan with 73,392, South Sumatra with 72,335 and West Kalimantan with 52,201 (Rakyat Merdeka). IQAir reported that more than 12.5 million people in seven provinces had been exposed to haze by early autumn (IQAir), and that public discussions highlighted a double burden on women, people with disabilities and Indigenous Peoples.
School: Pontianak moved to remote learning from August 24 (Mongabay Indonesia), and Palangka Raya was considering online exams in mid-October with about 10,400 ISPA cases (Kompas TV).
Daily life: Palembang visibility fell to 50 to 100 metres on October 15, and Palangka Raya's airport briefly suspended operations on October 14 (Rakyat Merdeka).
Wildlife: More than a dozen proboscis monkeys were found dead in a fire in Hulu Sungai Selatan, South Kalimantan, on September 13 (Mongabay Indonesia).
Climate: Indonesia recorded its highest August fire emissions in about 24 years (Health Policy Watch).

What I could not find is an official count of premature deaths from the 2026 haze. For scale only: a Harvard study linked the 2015 haze to more than 100,000 premature deaths across Indonesia, Malaysia and Singapore, as cited by World Press Photo. That is not a projection for 2026, but it is why I refuse to treat sick-leave statistics as the whole story.

The street is noticing

Residents are not just breathing it, they are marching. AFP reported protests in several cities on October 20, with one sign reading that haze and the president's regime equalled a national disaster (Kuwait Times / AFP). On October 28, students protested in Jakarta under a banner saying that if they burn, you burn with them (Jakarta Post). The protests also bundled in other complaints, including opposition to some welfare and food programmes, so I would not claim they are only about fire. The government's response was that it welcomed the protests and was open to input (bdnews24 / Reuters), and the state secretary said it was doing everything it could and would impose severe sanctions on responsible companies (Asia News Network). I think the more telling fact is that it took smoke over several countries and protests in the street for the language to sharpen.

The strongest case for the government, and why it does not settle it

In fairness, here is the best version of the other side.

The weather really was extreme. A very strong El Niño, a positive Indian Ocean Dipole and record emissions made this season brutal (Health Policy Watch).
The state did mobilise. The president visited West Kalimantan on August 22 and put senior military and police figures in charge across the worst-hit provinces (Investortrust). About 70,000 personnel were deployed (Bisnis), and weather modification and water bombing were used.
Sanctions escalated. From a handful of frozen permits in early September to 32 by late October, plus a new peat agency, a presidential instruction and a ban on planting oil palm on burned land (Katadata).
By official data, the burned area was still below the 2015, 2019 and 2023 extremes as of July (CNBC Indonesia).
Not every fire is a corporate fire. The ministry itself found much burning on unpermitted land, and community burning exists, which is the stated reason for the new burning ban.
The finance ministry says disaster money was not cut, and the foreign-aid logistics concern is real.

Why that does not settle it for me: nearly all of this is response, and the question in the title is about preparedness. A mobilisation that scales up once the haze reaches Malaysia is a late rescue, not a preparation. Rising sanction counts are not deterrence until fines are collected and land is restored. Burned-area totals look low only if you trust the lowest of three incompatible estimates. And a very strong El Niño explains the dryness, not why drained peat sat unguarded in dormant concessions or why government-run project zones lit up. Critics also have blind spots. Hotspot counts overstate fires, some NGO overlays involve judgment calls, and no one has shown the government sat idle. The accurate charge is narrower and, I think, worse: it was slow, reactive and unwilling to confront the people who hold the licences.

What I would demand

This is my opinion, not a technocratic consensus.

Decide the status question openly. Either declare a national-level emergency for the affected provinces or publish the criteria for why not. "We can still cope" is not a criterion.
Publish concession-level fire data. Names, boundaries, burned area and enforcement status, in machine-readable form, updated weekly.
Show the money. One clear table of BNPB and ministry fire budgets, prevention versus response, and an independent audit of the disaster-fund allegations.
Give the peat agency teeth before the next dry season. Public mandate, budget and powers to enter concessions and government projects, with a rewetting target it can be held to.
Stop clearing fire-prone land in El Niño years. Pause clearing on peat and drained land inside food and energy estates until hydrology is fixed.
Convert freezes into outcomes. For every sealing, suit and criminal case, publish the status: collected, restored, dismissed. Include the 35 civil suits promised in April.
Support fire-free villages, not only prosecutions of smallholders. Prevention costs less than a haze season.

Closing

BMKG expects rain to build from early November, with 50 to 300 mm per ten-day period forecast for the six priority provinces in the first ten days (RCTI+), and El Niño effects may last until the first quarter of 2027 (Reuters via NST). Earlier in the season, two days of heavy rain in Pontianak finally cleared the haze after nearly a month (iNews). The smoke will thin. The skies will clear. And then comes the part I dread: the pressure drops, the headlines move on, the freeze orders sit in a drawer, and in two or three dry seasons someone will write this article again. The test of 2026 is not whether the government can put out a fire in November. It is whether, by March, the peat is wet, the concession map is public, the names of the companies are still being spoken, and the people who burned and the people who allowed it are paying for it. I would love to be wrong about my cynicism. Show me.

What I could not verify

A reliable count of premature deaths linked to the 2026 haze.
The outcome of KLH's 35 planned civil suits announced in April.
Which burned-area figure (official, AP-cited or satellite-based) is closest to reality.
How the conflicting BNPB budget figures reconcile.

Sources: the articles and statements linked inline above.