Mumbai’s slum rehabilitation model was designed to replace informal settlements with formal homes on the same sites. But a study tracking projects and residents over 15 years finds that the system has often displaced people faster than it has rehoused them: for every eligible resident who moved into a formal home on site, 9.5 others moved to another slum.
That finding turns the central promise of the Slum Rehabilitation Act, introduced in 1995, into a question about the structure of redevelopment itself. The policy was meant to make Mumbai slum-free by combining free rehabilitation homes for eligible residents with additional development rights for private builders. In practice, residents can lose their homes immediately while the replacement asset remains dependent on a project’s financial viability, approvals, consent and construction progress.
The evidence comes from a working paper by Alishuba Philip, a PhD student of economics at the University of Zurich, presented in July at the National Bureau of Economic Research Summer Institute 2026 Real Estate conference in Cambridge, Massachusetts. The paper, titled ‘Slum Redevelopments and Evictions in a Developing Megacity’, examines slum redevelopment projects, residents’ locations and daytime satellite data over the period from 2007 to 2022.
Philip used Slum Rehabilitation Authority data covering projects, approvals and timelines. Her sample included 254 projects out of 592 slum projects during the period. She also digitised around 800,000 pages of electoral roll data and tracked residents using names and identifying details including fathers’ or spouses’ names, age and gender. Google Earth satellite data was used to verify whether a location had moved from a slum to a construction site and eventually to a building.
The research addresses a problem that is often missing from official measures of housing delivery: what happens to residents after redevelopment begins. The number of rehabilitation homes constructed can show output, but it does not necessarily show whether the original residents stayed near their jobs, schools and support networks, or whether they reached formal housing at all.
According to the study, only one-third of approved SRA projects tracked between 2007 and 2022 were completed. Twenty-nine per cent of approved projects never started, while 37 per cent had started but remained incomplete. The findings show a system in which approval does not reliably translate into construction, and construction does not reliably translate into completed rehabilitation.
The distinction matters because the redevelopment sequence is not symmetrical. Slum clearance generally happens before construction. As the paper puts it, “eviction is immediate, while the asset arrives only if and when construction is completed.” A resident therefore bears the immediate cost of displacement, while the benefit of a formal home depends on a chain of future events that may take years or may never occur.
The private-sector model is central to this gap. Under the SRA framework, builders are expected to rehouse slum residents without direct payment for the rehabilitation homes. In return, they receive additional floor space to sell either on the same plot or elsewhere through Transferable Development Rights. Whether construction begins depends on the expected economic return from the project.
Hussain Indorewala, assistant professor at the Kamla Raheja Vidyanidhi Institute of Architecture, told The Indian Express that builders can wait for bylaws and regulations to change if those changes would improve project profitability. This creates a system in which the delivery of a public housing obligation is linked to market conditions and the future value of development rights.
The delays are not only administrative. Projects also require the consent of 51 per cent of slum residents, down from the earlier threshold of 70 per cent. Conflicts between builders and residents, or between residents and their managing committees, can prevent a project from starting. The SRA has also pointed to conflicts involving residents who may be excluded from rehabilitation eligibility.
A senior SRA official said the original estimate of around five lakh homes needed to rehouse slum residents has increased to 13-14 lakh. The official attributed the increase to the proliferation of slums and the inability of agencies to prevent encroachments. The official also acknowledged that many approved projects do not take off.
The SRA official identified three reasons for stalled projects. The first was rising expectations as the government increased the size of rehabilitation homes over time. The second was opposition from residents who would not receive a free flat. Families in slums often build upwards, adding floors for their own use or to rent out. The official said residents on upper floors are generally not eligible for a free flat under the scheme, except in Dharavi, and may oppose redevelopment. The third was the effect of unforeseen disruptions, including Covid-19 delays that can push projects towards a decade-long timeline.
For residents, the practical response to delay is often temporary accommodation or transit rent. Neither necessarily provides a stable bridge to formal housing. Transit rent may stop, while transit accommodation can become a long-term residence. Shankar Patil, a resident of Lumbini Bagh Cooperative Society in Govandi, said a builder was offering Rs 14,000 a month in rent. He had chosen to move to a building in Panvel, while other residents preferred to remain near the original settlement, paying around Rs 5,000 in rent and using the balance for household expenses.
Patil’s concern was shaped by the experience of Bhim Wadi, a nearby settlement that entered redevelopment around a decade ago. He said the project stopped midway after 190 families had been displaced, with residents dispersing to different locations. The project is nearing completion only now. Such delays do more than extend a construction schedule: they can break the local networks through which residents access employment, schooling, childcare and informal support.
Philip found that slum redevelopment projects took an average of six years to build. Residents who successfully moved from a slum home into formal housing on the same site generally did so when projects were completed within four years. The longer the delay, the less likely it becomes that displacement will result in a transition out of slum living.
The study estimates that approval of a redevelopment project increases the probability of a resident moving into formal housing by only six percentage points. The probability of living in another slum increases by 22 percentage points. In practical terms, this produces the ratio of 9.5 residents displaced to another slum for every one eligible resident moving into formal housing on the original site.
Philip tested whether duplicate electoral-roll entries could be driving the results. Around 2 per cent of entries appeared to be duplicates even after matching names, fathers’ names and ages. She recalculated the findings in three ways: removing possible duplicates, counting each person only once and treating possible duplicates as less important. The findings did not change, leading her to conclude that duplicate names were unlikely to explain the results.
The research does have a defined limit. It does not cover residents who move into in-situ rehabilitation buildings but do not remain there for long. Indorewala said some residents may rent out or sell their units because they lose livelihoods or face higher costs in formal buildings, including maintenance expenses. If families experience financial distress, they may move elsewhere. This means that even a completed rehabilitation building may not represent a permanent end to housing insecurity.
That limitation points to a broader weakness in how rehabilitation is evaluated. Counting completed homes measures construction, but not necessarily housing stability. A project can be classified as successful when a building is delivered even if residents cannot afford to remain there, lose access to nearby work or move out because the building’s social and economic conditions do not match their livelihoods.
Mumbai’s policy response is now moving towards larger-scale redevelopment. The Maharashtra government has turned to cluster redevelopment for areas larger than 50 acres, relying on private development capacity. When such projects are taken up by a government body such as the SRA or the Mumbai Metropolitan Region Development Authority, they require government approval rather than the consent of individual slum dwellers, according to the report.
The SRA has also increased action against inactive developers. More than 450 developers have been terminated in the past three years after receiving approvals without progressing with work on the ground. These measures address the problem of dormant projects, but the evidence from the study suggests that the deeper issue is the time between displacement and delivery, and the dependence of rehabilitation on project economics.
The paper identifies several possible ways to reduce that gap, including stricter enforcement of construction timelines, interim housing support and phased demolition and construction. These are not simply construction-management choices. They determine whether residents retain access to their communities during redevelopment and whether the risk of a failed project is transferred from the developer to households with limited savings.
The numbers also reveal the scale of the challenge. Census 2011 data cited in the report showed that 42 per cent of Mumbai’s population of around 12 million lived in slums. From the start of the SRA programme until October 2025, the authority had built 2,79,307 rehabilitation homes. At the same time, the SRA’s estimate of the total housing requirement had risen from around five lakh homes to 13-14 lakh.
These figures show why the question cannot be reduced to whether Mumbai needs more redevelopment. The city needs a system that can deliver homes without making displacement the first guaranteed outcome and rehabilitation the uncertain one. The existing model has produced substantial construction, but the study suggests that its performance must be judged against residents’ actual movement through the housing system, not only against approvals or buildings completed.
Thirty years after the Slum Rehabilitation Act, the central urban question is therefore not whether informal settlements can be cleared, but whether redevelopment can protect residents during the long interval before a promised home exists. The available evidence confirms a significant displacement and completion problem. It does not establish that every project fails or that in-situ rehabilitation is unworkable in every location. But it shows that without reliable timelines, interim protection and resident tracking, a policy designed to move people into formal housing can instead move many of them from one slum to another.