The money bought presence

On the night of 22 August 2026, there were coordinated incidents at thirty-four points across Yala, Narathiwat and Pattani.

Phone masts burned. CCTV cameras destroyed. Local authority vehicles set alight. Caltrops scattered on roads to slow the response. ISOC Region 4 Forward recorded twenty-two cases of arson, six bombings, three harassment shootings and three of property destruction or theft, with two more under investigation. Yala had fourteen points, Narathiwat thirteen, Pattani five. A later count put the night at fifty-one points across twenty-one districts. A martial law curfew followed in Narathiwat.

Since 4 January 2004, the running total for these provinces stands at 23,536 incidents, 7,790 people dead and 14,794 injured.

And the question everybody asks, in Thai, every time this happens, is where the money went.

Three totals, and why there are three

The first thing to establish is that there is no single figure, and the reason matters.

The Parliamentary Information Repository’s series for fiscal years 2004 to 2026 gives ฿340,706.76 million.

The same source notes that if other integrated development subsidies are counted, the figure may reach ฿500,000 million.

Thai investigative reporting has put the twenty-two-year total at ฿523,363.48 million.

These are not three claims about one quantity. They are three different quantities. The narrow figure counts the dedicated southern budget lines. The wide figure adds development money that flows to the same provinces through other channels. Which you use depends on whether you think a road built in Narathiwat under a national programme is southern conflict spending or simply a road.

Both answers are defensible and the difference is ฿180 billion. Anyone quoting one number without saying which is quoting a scope, not a fact.

Three plans, and their names are the story

The Repository’s series divides into three phases, and the names are worth reading closely because they record how the state kept re-categorising the problem.

Phase one, 2004 to 2009: the Plan to Resolve the Unrest. ฿102,750.60 million. It opened at ฿13,451.00 million in 2004 and reached ฿20,991.60 million by 2009. A security framing, and a budget that roughly doubled inside six years.

Phase two, 2010 to 2016: the Plan to Resolve and Develop. ฿153,254.20 million — the largest of the three — peaking in 2016 at ฿30,512.80 million, the highest single year on record. The word develop has entered the title.

Phase three, 2017 to 2026: the Integrated Drive Plan. ฿84,701.96 million, declining continuously. By 2020 it was ฿10,641.91 million. For 2026 the line stands at ฿1,473.43 million.

From ฿30.5 billion to ฿1.5 billion. The dedicated line has fallen by more than ninety-five per cent from its peak.

⚠️ And here the article has to be careful with its own argument, because this is precisely the trap article 158 Produced, not received identified. That collapse is in the integrated plan line. It is not total state spending in these provinces. ISOC’s own budget runs separately and has been publicly discussed at around ฿5,000 million. A ninety-five per cent fall in one line is not a ninety-five per cent fall in the money. Anyone — including this site — who reads the first as the second has changed the denominator mid-sentence.

Now the answer to the actual question

Where did it go?

The Royal Thai Army has published the answer, in its own defence, and it is the most useful figure in the whole subject.

About 75% of ISOC’s budget goes to the southern border provinces mission. And of that, close to 70% is personnel compensation — pay for soldiers, police, administrative officers and staff of other agencies, amounting to roughly 49,000 posts.

The money went on wages. For about forty-nine thousand people, to be there.

The Army’s further explanation is that ISOC’s apparent headcount is largely secondment — personnel borrowed from the three services, the Army most of all, assigned to help. The division of labour is that ISOC Region 4 Forward holds security, public protection and law enforcement, while SBPAC holds development.

That is not money melting into a river. It is a payroll.

And a payroll is a much harder thing to argue with than a scandal, which is why it is the more important finding.

What a payroll can and cannot purchase

The paper prepared for Parliament that accompanies these figures sets out five dimensions to the conflict’s roots.

History, including the Anglo-Siamese Treaty of 1909. Governance, specifically the centralisation of authority and national-identity policy as it bears on local distinctiveness. Identity and religion — a Malay Muslim way of life requiring cultural understanding. Economic structure, where resource allocation and development have not always matched how communities actually live. Education, where curriculum and resources have not adequately answered the area’s linguistic and cultural context.

Now ask which of the five a budget line can buy.

History cannot be purchased. It can be acknowledged, taught differently, or apologised for, and none of those is an appropriation.

Centralisation cannot be purchased. It is a constitutional and administrative question about where decisions are made, and article 159 Every province already has an election established that every province in Thailand except Bangkok already sits under a governor it does not elect.

Identity and religious recognition cannot be purchased. They require political and legal accommodation.

Education can be partly purchased — buildings, teachers, materials — but the paper’s complaint is about curriculum and language, which are policy choices, not costs.

Economic structure is the one that maps onto money, and even there the paper’s objection is about the fit between development and community life rather than the volume of it.

So the state’s own analysis identifies five causes, of which roughly one is purchasable — and the instrument it has deployed for twenty-two years is a budget, seventy per cent of which is wages.

That is not a claim that anybody was corrupt or foolish. It is a claim about what an instrument can do.

Why this keeps happening

This archive has now arrived at the same shape thirteen times, and this is the version in public finance.

Article 165 The rule is about the road found a building regulation constructed entirely from quantities verifiable with a tape measure, silent on every question requiring judgement. Article 162 The fine has a ceiling and the damage does not found the anti-corruption commission asking for sensors to remove officials’ discretion from a weighbridge. Article 161 None of them is about the building found a venue test sorting by observable social conduct while the danger was structural.

Here: the state addressed the dimension of the problem its machinery can process.

A budget is what a state allocates. Posts are what a budget fills. Presence is what posts produce. Historical grievance, constitutional arrangement, religious recognition and curriculum are not things an appropriations process has an instrument for — not because anyone decided they did not matter, but because the annual budget cycle has no line for them.

And article 142’s finding turns out to run the other way here. That piece found a small market where presence itself is the product — somebody has to be in the chair, and that is worth paying for. The southern budget bought presence at a scale of forty-nine thousand people. Presence was the deliverable there and it is the deliverable here. The difference is that in a queue, presence is the whole of what is wanted. In a conflict with five root dimensions, presence addresses the symptom.

The dispute this article will not settle

There is an active argument about whether the accounting is complete, and it is on the record on both sides.

A proposal has been made in Parliament to dissolve ISOC, and the security establishment has argued it remains necessary. A minister has raised the question of concealed budget; the Army has publicly rejected it, stating that its allocation follows regulations, that nothing is hidden, and that its work does not duplicate SBPAC’s.

This article does not resolve that. The published figures it relies on are the parliamentary series and the Army’s own statement of composition, and neither establishes anything about the other’s completeness. Reporting that a dispute exists is not the same as taking a side in it, and the argument above does not depend on which side is right. Seventy per cent of a security budget being wages is a fact about what the money does, not about whether all of it is disclosed.

---

Common misconceptions

“Half a trillion baht was spent and nothing happened.” Three totals circulate for different scopes — ฿340bn, roughly ฿500bn, and ฿523bn. They count different things.

“The budget is still pouring in.” The dedicated integrated line has fallen from ฿30.5bn in 2016 to ฿1.47bn in 2026. ISOC’s own budget runs separately, so this is not a fall in total spending.

“The money disappeared.” By the Army’s published account, roughly 70% of the security budget is pay for about 49,000 posts.

“It was all military.” SBPAC holds the development mandate; ISOC Region 4 Forward holds security. The second phase of the plan, 2010–2016, was the largest and carried development in its title.

“More money would fix it.” The parliament’s own paper lists five root dimensions, of which about one is purchasable.