Public Housing as Social Infrastructure: The HDB Story

Ravi Subramaniam: Good morning everyone. I want to start with a question rather than a slide. When Singapore's Housing Development Board was established in 1960, it had a mandate to house 1.5 million people who were living in overcrowded slums and kampongs. It had very little money, almost no skilled staff, and no track record. How does an institution in that position deliver? That's the paradox I want us to sit with for the next two hours.

Ravi Subramaniam: The answer, I think, is that HDB succeeded not because of its resources but because of its mandate clarity. It had one job: build flats, as many as possible, as fast as possible. That singularity of mission, before it expanded to quality, to community formation, to asset accumulation, is what allowed it to move at the speed it did. By 1965, five years after establishment, HDB had completed 54,000 units. By 1970, another 100,000. The quality was basic. The estates were austere. But they were dry, they had piped water, and they were better than what people had left.

Ravi Subramaniam: Now, the second phase is where it gets interesting. Once you've solved the emergency, once you've moved people out of kampongs and into public housing, you face a different problem. You've created a vertical city where people who used to live on the ground in extended family compounds now live in separate units on separate floors. The social support structures of kampong life, the grandmother who watches the children, the neighbours who share meals, the communal spaces where community forms, those don't transfer automatically to a 12-storey block.

Ravi Subramaniam: HDB's response to this in the 1980s and 1990s was what we'd now call community infrastructure embedding. The void deck, the open ground floor of every HDB block, was deliberately left unenclosed to create community space. The Multi-Purpose Hall was mandated in every new town. The town council system, established in 1988, gave residents democratic ownership over the management of their estates. These are not architectural features, they are social policy instruments embedded in physical design.

Ravi Subramaniam: Then the ethnic integration question. By the late 1980s, a troubling pattern was emerging in the resale market. Chinese households were preferring to buy into blocks with more Chinese residents. Malay households clustered. Indian households clustered. Not because anyone told them to, but because of natural preferences and social networks. Left to market dynamics, Singapore's public housing estates were quietly re-ethnicising. The government made a very difficult decision: to intervene at the point of resale with ethnic proportion limits.

Ravi Subramaniam: The Ethnic Integration Policy, introduced in 1989, was politically controversial. There were legitimate questions about whether the government had the right to tell someone they couldn't sell their flat to a buyer of their own ethnicity. The government's argument, which I think history has largely vindicated, was that Singapore's social compact depended on daily inter-ethnic interaction, not just formal tolerance. You can't achieve that through aspirational values alone. You need to engineer it structurally. The EIP is that structural engineering.

Ravi Subramaniam: Now I want to open this to discussion. I know several of you are from cities dealing with urban segregation, not necessarily ethnic, but income-based, which creates many of the same dynamics. Nguyen, your hand was up.

Nguyen Thanh Hung: Thank you. In Ho Chi Minh City, we see significant spatial segregation forming along income lines as the housing market has liberalised. High-end gated communities in District 2, and informal settlements in the peripheral districts. We don't have the public housing stock to use allocation as a policy lever. What we do have is zoning authority. But our experience with inclusionary zoning requirements, requiring market developers to include affordable units, has been weak, because developers find workarounds. What made the EIP effective where inclusionary zoning tends not to be?

Ravi Subramaniam: That's the right question. The EIP is effective for two structural reasons that most inclusionary zoning doesn't replicate. First, it operates at point of title transaction, it's not a building requirement that can be met at construction and then dissolved, it runs with every sale of every unit indefinitely. Second, it covers 80% of the housing market, not 10% or 15%. When you require inclusionary units in market developments, you're affecting a minority of the housing stock. The EIP governs the dominant tenure type. The scale is what makes it socially significant.

Ravi Subramaniam: The third generation of HDB policy, what we're living through now, is estate rejuvenation. Our earliest blocks were built in the 1960s and 1970s. They are structurally sound but their facilities, layouts, and energy systems are obsolete. The Home Improvement Programme and the Selective En-bloc Redevelopment Scheme are the two main instruments. The SERS, selective demolition and rebuilding, is the most powerful and most politically sensitive, because it requires displacing residents who have lived in a place for decades.

Ravi Subramaniam: Why do residents accept SERS? Because the compensatory offer, a new flat in the same general area, at a subsidised price, with priority allocation, has been credible every time it's been made. That credibility was built in the 1960s when the first kampong residents were resettled. They were told they'd get better housing. They got it. Every subsequent resettlement exercise inherits that credibility. It's an institutional trust asset that took 60 years to build and would take perhaps one broken promise to destroy.

Ravi Subramaniam: Let me address the fourth generation directly: what I call the 'longevity challenge.' Singapore is one of Asia's fastest ageing societies. The population cohort that moved into public housing in the 1970s and 1980s is now 70, 80, 90 years old. They own their flats, they have 99-year leases, and many of those leases have 30 to 40 years left. The lease decay problem is this: as the lease shortens, the flat's value declines, because at lease end the flat reverts to HDB with no compensation. For many elderly households, the flat is their primary or only asset.

Ravi Subramaniam: The government has introduced the Lease Buyback Scheme, which allows elderly households to sell a portion of their remaining lease back to HDB to monetise the asset while retaining the right to live in the flat for life. It's an elegant financial instrument. But it doesn't solve the underlying policy question of what happens to blocks whose leases expire, and whether the government will selectively extend leases or mandatorily redevelop. That decision hasn't been made publicly yet, and it's the most contested frontier of HDB policy today.

Ravi Subramaniam: Kampung Admiralty, which some of you visited last week, represents the fourth-generation design response to demographic ageing. It's a vertical village: a hawker centre and supermarket at ground level, then a community plaza and polyclinic, then childcare facilities, then 104 units of senior public housing at the top, all stacked above an MRT station. The idea is that everything an elderly person needs, medical care, daily groceries, community interaction, childcare for grandchildren, is within the same building. You eliminate the mobility barrier without institutionalising the resident.

Nguyen Thanh Hung: How do you prevent such facilities from becoming isolated elderly ghettos? If all the residents are seniors, you lose the inter-generational interaction that also matters for wellbeing.

Ravi Subramaniam: The design at Kampung Admiralty addresses this by making the community plaza and childcare facilities available to the surrounding non-senior residential population, not just the building residents. The hawker centre is a public draw that brings in the whole neighbourhood. So the building is intentionally porous, residents of senior housing share common spaces with a much broader cross-section of the community. Whether that's sufficient is a research question we're still studying, but early social wellbeing surveys are positive.

Ravi Subramaniam: Let me shift to the policy question I want to leave you with, because I think it's the most transferable. What we've learned from 65 years of public housing is that the hardest moments in housing policy are not technical, they're political. The decision to resettle kampong residents in the 1960s was politically hard. The Ethnic Integration Policy was politically hard. SERS is politically hard. Every generation of housing policy has required decisions that were unpopular with some constituents in the short term but necessary for the long-term social outcome.

Ravi Subramaniam: What allowed Singapore's government to make those decisions is a combination of things: long political tenure that reduced electoral vulnerability; a technocratic civil service that could develop and implement policy without constant political renegotiation; and critically, an early track record of delivering on its promises that generated the public trust needed to sustain hard subsequent decisions. You can't manufacture that track record quickly. But you can start building it today with every promise you keep on a smaller decision.

Ravi Subramaniam: For those of you from cities with shorter electoral cycles and more contested political environments, which is most of you, I want to offer this reframing. You may not be able to make Singapore's category of decisions. But you can identify the decisions in your own context that are the equivalent: the ones where the short-term political cost and the long-term social benefit are both high, and where your city's leadership needs to absorb the short-term cost to deliver the long-term gain. The housing policy question is not 'how do we replicate Singapore', it's 'what is our equivalent of the 1966 land acquisition act in our context, and do we have the political will to do it?'

Ravi Subramaniam: We have about 45 minutes remaining and I want to use it in dialogue rather than lecture. I've put four questions on the screen, they're also visible in the session page on the platform. I'd like each of you to take two minutes with the person next to you to discuss one question, and then we'll bring it back to the full group. The questions are about your own city's housing situation, mapped to the HDB generations framework we've just discussed.

Ravi Subramaniam: While you're discussing, I want to make one more observation that I think is important for understanding Singapore's housing success. The 99-year lease structure was a policy choice, not a technical necessity. HDB could have sold flats in freehold. The decision to use leasehold was deliberate: it preserves the state's ability to reclaim land for redevelopment when planning needs change, without having to go through a politically costly compulsory acquisition of freehold private property. The lease is both a housing instrument and a long-term land planning instrument. Those two functions reinforcing each other is what makes it so powerful.

Nguyen Thanh Hung: That's a really significant point for Vietnam. We have a similar leasehold system for state land, long-term land use rights rather than freehold, but we haven't used it as an active planning instrument. The lease terms are set and then largely forgotten. What Singapore's experience suggests is that actively managing lease duration and reversion terms, making the lease an ongoing governance relationship rather than a one-time transaction, is where the real planning leverage lies.

Ravi Subramaniam: Exactly right. The lease renewal decision, whether to extend, at what price, and with what conditions, is a planning decision. Singapore hasn't fully resolved how to make those decisions transparently and equitably for the current generation of ageing leases, which is why I describe it as the most contested frontier. But the architecture for making those decisions actively exists. You have a similar architecture in Vietnam if you choose to use it that way.

Ravi Subramaniam: Let me bring the full group back together and hear from some other cities. Claire, Kigali has a fast-growing public housing programme. What's the biggest difference you see between your situation and Singapore's 1960s context?

Ravi Subramaniam: And as others respond, I want us to keep in mind the framing from this morning: housing policy is not primarily about buildings. It's about what kind of society you want to build and what physical infrastructure is needed to enable that society. Singapore's genius was in understanding that housing allocation and estate design are governance instruments as much as social services. The question for each of your cities is: what social outcomes do you need housing policy to deliver, and are your current instruments, your legal frameworks, your design standards, your allocation systems, aligned with those outcomes?

Ravi Subramaniam: I'll take two more questions and then we'll move to the site visit briefing. The bus for Punggol departs at 13:45 from the hotel lobby. Please have your name tags, you'll need them for the Digital District access gates.

Nguyen Thanh Hung: One final question from me. You mentioned the political cost of hard housing decisions. Singapore is often described as a context where political costs are lower because the ruling party has governed continuously. For those of us in more contested political environments, what is your honest advice about how to protect housing programmes from political reversal between administrations?

Ravi Subramaniam: Honest answer: put everything you can in law rather than policy. Policy can be reversed by a minister's minute. A statutory scheme requires legislative amendment. Every housing programme, your allocation rules, your integration requirements, your subsidy entitlements, should be embedded in primary or at minimum secondary legislation to the greatest extent your constitutional framework allows. The second line of defence is making the beneficiaries politically visible: when a programme visibly benefits a large constituency, reversing it becomes electorally costly for any administration. Singapore's housing programme has 80% of the population as its constituency. That's a significant political protection.

Ravi Subramaniam: Thank you all for a very engaged session. The materials, including the Urban Solutions housing issue and the Kampung Admiralty case study, are all on the platform. I'd particularly recommend the Fifty Years of Public Housing policy brief as a reading for tonight, as it frames the generational evolution we discussed in more detail. See you at the bus at 13:45.