What thirty years in four cities taught me about master plans

I have lived in Delhi, Washington DC, Singapore, and now Nairobi over the past thirty years. Four cities, three continents, every point on the planning spectrum from Singapore’s legendary order to Delhi’s pollution and traffic, its magnificent Lutyens boulevards and its great food, its beauty and its dysfunction living side by side.. What I have learned is that the thing called a master plan is almost never the thing that makes a city work. And in India, where I grew up watching Delhi grow, the results have been sobering. The planning system has often generated the very problems it was designed to solve.

Start with Singapore, because everyone does, and almost everyone draws the wrong lesson.

When I lived there I kept waiting to find the master plan that explained the city. The MRT arriving before the housing estates it served. The drainage canals wide enough to take a monsoon. The streets laid before the buildings went up. What I found instead was less glamorous and more instructive: a government that spent decades acquiring land ahead of demand, servicing it with trunk infrastructure, and releasing it in a sequence calibrated to what the city could actually absorb. The Housing & Development Board built apartments, yes, but the prior act was always land: assembled early, held in public hands long enough to be planned around, serviced properly before anyone built on it. Singapore works because the state did the unglamorous infrastructure work first. It is a story about land readjustment and sequencing, not master planning.

I moved to Delhi in 1997.

Delhi has had master plans since 1962. Each one is thicker than the last, more elaborate in its land use categories, more precise in its vision of the city it wants to become. The city has become something else: unauthorized colonies spreading across the fringe, mixed-use ground floors operating outside their sanctioned categories, buildings that routinely violate their approved plans, millions of people living in settlements whose legal status has been disputed, regularized, re-disputed, and deferred across decades of political negotiation. Delhi was shaped by its master plans. The shape is visible in every blocked drain, every missing footpath, every family living in a structure that does not exist within the “official” planning system.

Now I live in Nairobi, which has had master plans since 1948. Every rainy season the water finds the drainage corridors that were built on across decades, and the roads flood, and the settlements that grew in the places the plans didn’t cover fill with water. The plan said don’t build there. People built there because they had nowhere else to go. The plan lost.

I love Nairobi: Karura Forest sitting improbably in the middle of five million people, the pole pole warmth of the people, the food. But I watch it flood during the rains and I think about Delhi, and I think about the 400 million people India will add to its cities by 2050, and I think: we know exactly how this goes wrong. We have watched it go wrong, in city after city, for seventy years.

This pattern is well documented. A World Bank study of seven secondary cities in Tanzania found that master plans offered broad strategic direction but had weak effects on actual land use and infrastructure. Satellite imagery showed the gap plainly: the mapped city and the built city were two different places. The same gap exists across Africa and South Asia, and it is not a failure of ambition. It is a failure of design.

When plans are detached from the city that exists, three structural failures follow.

The first is informality at scale. India’s planning system made formal compliance so expensive, so slow, and so disconnected from what people actually need to build that informality became the rational response for most of the city. The unauthorized colonies and buildings that violate their sanctioned plans are not aberrations. They are the system’s predictable output.

The second is chronic undersupply of public realm. Streets, parks, drainage corridors, trunk infrastructure: Indian cities are starved of all of them. Plans that allocate land use without mechanisms to actually assemble and service that land produce cities where public space exists on paper and nowhere else. The infrastructure that should have preceded growth arrives, if it arrives, decades late.

The third is wasteful private land use. Suburban setbacks, minimum plot sizes, FSI limits calibrated for cities that no longer exist: India’s development regulations systematically prevent the density that would make urban land affordable. A planning system designed to impose order has produced sprawl, fragmentation, and land use patterns no planner would have chosen.

The standard response to all three has been targeted programs: sites and services, cross-subsidized housing, slum upgrading, regularization drives. These help at the margins but leave the system intact. You cannot solve a supply problem with interventions for the poor while the regulatory system continues to suppress supply for everyone.

The solution is to reform and liberalize planning so that the formal city becomes accessible to ordinary people. The instruments exist and have been tested.

Ahmedabad’s Town Planning Schemes are the clearest Indian example. Rather than waiting for a metropolitan plan to produce results, Ahmedabad used repeated micro-plans: one to two square kilometers at a time, up to 250 landowners per scheme, land pooled and reconfigured, roads and public space carved out, infrastructure financed as land values rose. Two hundred and fifty landowners, one negotiated deal, a road and a park at the end of it. This happened dozens of times across the city, in different neighborhoods, at different moments. Built city, not imagined city. The instrument exists in most state planning acts. It is used almost nowhere at the scale current growth demands.

Thailand’s Baan Mankong program demonstrates a parallel logic for existing settlements. The formal system was never going to replace 2,000 informal settlements across 200 cities. So the Community Organizations Development Institute accepted the city as it was and improved it where people lived. Communities surveyed their own settlements, mapped them, negotiated solutions, planned their own upgrading. Three hundred thousand households. The premise was that residents knew their neighborhoods better than any planning office. They did.

Four reforms would make the biggest difference in India.

Scale land readjustment for urban expansion. Town Planning Schemes and similar instruments can convert agricultural fringe land into serviced urban land with streets and drainage built in, without coercive acquisition. Secondary cities growing at five percent a year need this running continuously, not as an occasional project when a development authority gets around to it.

Liberalize FSI in serviced areas with development charges that fund infrastructure. Where density is permitted and infrastructure is financed together, supply increases and prices moderate. The evidence from cities that have tried this, including Indian cities, is consistent. Suppressing density to manage congestion while avoiding infrastructure financing produces the worst outcome: low supply and poor services simultaneously.

Simplify development regulations so formal compliance becomes realistic for ordinary builders. Performance-based standards for daylight, street-to-height ratios, and fire safety can replace suburban setback rules and minimum plot sizes that make legal construction unnecessarily expensive. Legalizing small-scale mixed use and home-based work as of right across most urban zones would bring millions of economic activities into the formal city without a single rupee of public spending.

Upgrade existing informal settlements with guaranteed minimum standards for streets, drainage, and services, while creating realistic paths into better formal housing as supply expands through liberalization. The goal is better land use and better services over time, not the preservation of current layouts.

How would we know this is working? A small number of measurable outcomes over ten to fifteen years: new informal growth slows, existing informal areas get better streets and public realm, households have realistic paths into formal housing as supply expands, and redevelopment increases land use intensity rather than freezing today’s patterns.

Those are testable propositions, and they are now politically timely.

India’s Budget 2025-26 announced a Rs 1 lakh crore Urban Challenge Fund, and the Cabinet has since approved it. The fund runs to 2030-31, covers cities from major metros down to Tier-II and Tier-III towns, and explicitly ties funding to planning and spatial reforms. Continuation of reforms is a prerequisite for further fund release. At least 50 percent of project financing must come from market sources, which means cities that want the money have to be creditworthy enough to raise it, which means they have to be functional enough to generate revenue, which means they have to reform.

This is a rare alignment of political will, fiscal incentive, and reform pressure. The Urban Challenge Fund rewards implementation. This is exactly the shift this argument has been making.

State governments deciding how to access that fund over the next five years face a practical question: which planning reforms are high enough leverage to unlock the money and actually change what gets built? The answer, from Ahmedabad, from Singapore, from every city that has managed to grow with streets and drains already in place, is the same. Protect the corridors before they are built on. Run land readjustment at the pace of growth. Liberalize density where infrastructure can be financed. Make the formal city legal and affordable for the people who actually live in it.

Streets before walls. Drains before floods. Rules that guide growth instead of pretending it isn’t happening.

The fund is there. The instruments are known. The question is whether states will use this window to reform the system, or use the money to build more infrastructure inside a regulatory regime that will continue to produce the same outcomes.

Views are my own.