India, April 23 -- Transit-oriented development (TOD) rests on three fundamentals: Density, diversity, and design. A good TOD neighbourhood is dense enough to support frequent transit, diverse enough to serve all people and activities, and designed in a manner that makes walking to the metro pleasant rather than punishing. Delhi's new Regulations for Transit Oriented Development and Charges, 2026, announced earlier this month, take an important leap forward on density. On diversity and design, however, the policy has serious gaps. The most important reform is the shift in the policy's approach, from being node-based to corridor-based. The earlier policy required a minimum of eight hectares of developable land for a TOD node - a threshold that effectively locked out most of Delhi's fragmented urban fabric. Switching to the corridor logic and reducing the minimum plot size from one hectare to 2,000 square metres means more landowners, more plots, and more locations now qualify for inclusion. Including 80 square kilometres of previously excluded areas - land pooling zones, low-density residential areas, and unauthorised colonies - is equally significant. These areas are already getting denser, though informally; dealing with that pressure within a planned framework is better than pretending it doesn't exist. The single-window clearance and unified TOD charge are also genuine improvements. The earlier policy's insistence on Influence Zone Plans for approval was a procedural bottleneck that killed momentum. One provision deserves particular attention: TOD charges are mandated to flow to a ring-fenced escrow account within a dedicated TOD fund. If this fund is directed towards public-realm improvements in the corridors, the policy could deliver on some of its promise. Now, let's discuss what is missing. The policy fixes the floor area ratio (FAR) mix at 65% residential below 100 square metres, 25% residential above 100 square metres or office use, and 10% commercial and amenities - and applies this formula uniformly across 207 square kilometres. There is no clarity on whether this particular mix is what these corridors actually need, because the question hasn't been asked at the level of individual stations or neighbourhoods. A station in a predominantly commercial district needs more office and retail space. A station in a low-income residential area may need more community facilities. A station near a hospital cluster needs allocations for different spaces entirely. The 65-25-10 FAR distribution seems calibrated to a generic high-density residential project, not for the full diversity of urban contexts Delhi's metro passes through. A more apt approach would set a minimum residential obligation and allow the remaining mix to be determined through local planning for each station catchment. Fixing the FAR distribution through a gazette notification makes it almost impossible to revise it without a fresh statutory process-locking in a one-size-fits-all formula for years. This is the diversity problem. Now, let's move to the design aspect of the policy. TOD essentially focuses on a walking-friendly commuter district. But the policy says project proponents "may, at their option, provide a pedestrian walkway" connecting the site to the transit station. In a document that mandates FAR splits to the last percentage point, walkability is left entirely to goodwill. Imagine a family moving into one of these new, affordable apartments. The metro station is 400 metres away. But the footpath is broken, lined with parked cars, and exposed to the Delhi sun. They use autorickshaws and the metro card sits in a drawer. This is transit-adjacent development, not TOD. TOD neighbourhoods need to be cooler, greener, and more pleasant, to encourage people to choose walking over driving. Without mandatory tree canopy, adequate shade, and street-fronting setback standards, the market delivers what it always has: Compound walls, security gates, and a ground floor given over to parking ramps. Streets lined with high compound walls will have no eyes on them - a safety concern for everyone. Under a well-designed TOD, living near a metro station should simply mean that you don't need a car. But Delhi's new TOD requires one equivalent car space per 100 square metres of FAR. This is not a parking cap, it is mandating parking provision - the exact opposite of what a transit-oriented neighbourhood requires. Every flat with a guaranteed parking spot will continue to make the most of its daily trips by car. The metro becomes a backup option, not a way of life. Indian cities are over-engineered and under-designed. This policy is meticulous about FAR, charge structures, and approval timelines, but almost silent on the quality of the streets, the comfort of the pedestrians/commuters, and the liveability of the neighbourhoods. Delhi's planning chronically under-delivers on the public goods needed to make private development work. TOD requires the State to lead - with walkable streets, first-and-last-mile connections, and green cover - so that private investment can capitalise on this. The affordability question sits within the same gap: A unit-size mandate is not an affordability guarantee. Land along Delhi's metro corridors is among the most expensive in the country. A 99 square metre flat built by a private developer recovering land costs and TOD charges will not be priced for a household earning Rs.20,000-30,000 a month. Affordable housing funding and income-targeted allocations could bridge that gap to bring low-income commuters near transit. Delhi is no longer scared of density, which is a good thing. The harder question - whether this density will be diverse, walkable, green, and genuinely affordable - depends on the work this policy has not yet done....