Two housing crises, opposite directions

Article 43 established that Bangkok is carrying more than 60,000 unsold condominium units, worth over ฿200 billion, with roughly six years of inventory nationwide.

Thailand also runs a national programme whose twenty-year plan targets 1,053,702 households who lack secure housing.

Two housing crises, in the same country, pointing in opposite directions. Too many homes nobody can buy, and too many households living somewhere they have no legal right to remain.

That’s not a paradox so much as a description of how housing markets fail — they produce what’s profitable, not what’s needed. But the scale of the mismatch here is unusual, and the response to the second crisis is genuinely interesting.

What Baan Mankong is

บ้านมั่นคง — “secure housing” — began in 2003, with an initial Cabinet allocation of ฿146 million to the Community Organizations Development Institute, CODI.

Its premise is the opposite of what most countries do with informal settlements.

The standard approach is clearance: demolish, relocate residents to public housing somewhere cheaper, redevelop the land. It’s fast, it’s legible to planners, and it has a long record of destroying functioning communities and producing estates people don’t want to live in.

Baan Mankong instead gives the community the money and the legal security, and lets it rebuild itself.

The mechanics: residents form a housing cooperative. The cooperative negotiates land tenure — purchase, long-term lease, or an arrangement with the landowner. CODI provides low-interest loans with long repayment terms, plus subsidies toward construction and utilities. The community decides the layout, the design and the allocation.

The state finances. The community builds. That inversion is the whole idea, and it is the reason the programme gets studied internationally.

Why it’s a better idea than it sounds

Three reasons worth stating for anyone who assumes community-led means slow and amateurish.

It preserves the thing that actually works about informal settlements — the social network. People in these communities have childcare arrangements, credit relationships, work contacts and mutual obligation of exactly the kind described throughout the parallel economy series. Clearance destroys all of that and the estate that replaces it has none.

It produces housing people will maintain, because they designed it, chose it and are paying for it.

And collective tenure is more robust than individual title for low-income households. An individual with a title can be pressured or bought out one at a time. A cooperative holding land collectively is considerably harder to dismantle, which matters in a city where land values move fast.

Now the honest criticism

Here is where I’d be doing propaganda rather than journalism if I stopped.

The pace is the problem.

Cumulative results reported as of March 2024 show 133,382 households assisted under Baan Mankong since 2003. CODI’s broader housing work reaches larger totals — its own anniversary reporting cites figures in the hundreds of thousands across all its programmes — and the counting definitions differ, so treat the numbers as a range rather than a single figure.

But take the twenty-year master plan target of 1,053,702 households by 2036 and set it against roughly 133,000 in twenty-one years.

At the observed rate, the plan takes over a century.

That is not a criticism of the model. It’s a criticism of the funding and the political priority. A programme that works and is resourced to reach a fraction of its own target is a policy success and an implementation failure at the same time, and both halves are true.

The juxtaposition

Put article 43 next to this one and the picture is uncomfortable.

60,000 empty condominium units in Bangkok, built for a demand that didn’t materialise, in a market with six years of inventory and take-up that halved in a quarter.

A million households targeted for housing security, on a programme moving at a fraction of the necessary pace.

I’m not going to pretend those two things can simply be connected — the empty units are in the wrong places, at the wrong price points, owned by individual investors rather than available to policy. Housing doesn’t work that way and anyone claiming otherwise is selling something.

But the mismatch is worth sitting with. A country capable of financing 350,000 unsold units nationwide is not a country short of construction capacity, credit, or land. What it’s short of is a mechanism for directing any of that toward the households who need it most — and it has a good mechanism, running slowly, in the same cities.

Why this belongs alongside the healthcare piece

Article 36 argued that Thailand did something remarkable with universal healthcare, and that a publication cataloguing gaps without acknowledging it would be inaccurate.

Baan Mankong is the same category of thing. A genuinely innovative, internationally studied, community-led approach to a problem most countries handle badly — designed and run here, for two decades.

It’s just, unlike the health scheme, never been resourced to match its ambition.

Which may be the more general pattern in Thai policy: the ideas are frequently excellent and the follow-through is uneven. Healthcare got both. Housing got one. Old-age provision, per article 38, got neither.

Common misconceptions

“Thailand demolishes slums.” Clearance happens. But it also runs a two-decade programme explicitly designed to do the opposite, and that programme is internationally recognised.

“Informal settlements are lawless.” They have dense internal organisation — which is precisely what the cooperative model builds on.

“It’s charity housing.” Residents take loans and repay them. It’s finance and legal security, not a gift.

“Community-led means low quality.” Communities that design and pay for their own housing tend to maintain it. The failure mode of this model is slowness, not shoddiness.

“The empty condos could just be used.” Wrong locations, wrong prices, privately owned by individual investors. It isn’t a supply that policy can redirect.