How a slum is interpreted depends almost entirely on who is doing the interpreting and what tools they bring. A satellite analyst sees roof density and irregular street patterns. A public health official sees cholera risk and heat exposure. An economist sees unregistered businesses invisible to official data. A resident sees home. These interpretations are not just academic exercises; they shape which neighborhoods receive investment, which get demolished, and which are simply ignored. The gap between what outside observers measure and what life inside actually looks like is one of the defining tensions in urban planning today, and the tools available to close that gap are evolving fast.
Why There Is No Single Definition
The word “slum” carries enormous baggage. In policy documents, it typically refers to areas with inadequate housing, overcrowding, poor sanitation, insecure tenure, or some combination of these. But the boundaries of the concept blur quickly once you look at actual communities. Research across cities in different political, cultural, geographic, and economic contexts has found that settlements range from the urban fringe to the inner city, sit on farmland, post-industrial land, waterfronts, and hillsides, and develop at vastly different speeds and densities. The formal-informal distinction is not binary; there is a wide spectrum of informality even within a single city.1Habitat International. Towards a morphogenesis of informal settlements
This matters because the label itself changes what happens next. When a government classifies an area as a “slum,” that classification can trigger upgrading programs, but it can just as easily justify demolition or forced relocation. When international agencies track “slum prevalence” globally, they aggregate enormously different living situations into a single number, which can obscure more than it reveals. A corrugated-iron settlement on a South African hillside and a densely built but economically vibrant neighborhood in Mumbai may both count as slums under UN-Habitat criteria, but they face very different challenges and have very different needs.
What Satellites See and What They Miss
Satellite imagery has become one of the primary tools for interpreting slums at scale. From orbit, informal settlements tend to share visual signatures: small, irregularly shaped building footprints, narrow or absent road networks, high roof density, and limited vegetation. Over the past decade, deep learning and other artificial intelligence methods have added a new dimension to this work, allowing automated analysis of satellite images to identify complex spatial patterns associated with informal settlements.2arXiv. Deep Learning for Slum Mapping in Remote Sensing Images: A Meta-analysis and Review These approaches can segment high-resolution images into meaningful objects and then classify them using machine learning algorithms, combining techniques like object-based image analysis with geographic information systems.3arXiv. Deep Learning for Slum Mapping in Remote Sensing Images: A Meta-analysis and Review – Section: II-D Data Mining Techniques for Slum Identification
The promise is enormous: instead of relying on expensive and infrequent ground surveys, cities could track the growth of informal settlements monthly. But the interpretation of what the satellite “sees” is not straightforward. The standard evaluation framework for these models relies on metrics like accuracy, precision, and recall, assessing how well the model identifies slum areas against complex urban backgrounds.4arXiv. Deep Learning for Slum Mapping in Remote Sensing Images: A Meta-analysis and Review – Section: V-E Metrics Used A model might correctly identify dense rooftop clusters but still miss critical context: whether the structures are residential or commercial, whether residents have legal tenure, or whether piped water runs underneath.
Some cities are pushing the technology further. São Paulo’s Housing Secretariat has used drone surveys of favelas to create digital twins, three-dimensional models of slum areas reconstructed automatically from aerial point clouds and building footprints, providing a continuously updateable tool for upgrading projects.5Copernicus Publications / The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. DIGITAL TWIN CREATION FOR SLUMS IN BRAZIL BASED ON UAV DATA These digital twins offer a much richer spatial interpretation than a flat satellite image, but they still represent a view from above, filtered through the priorities of whoever commissioned the survey.
When Global Numbers Contradict Local Reality
One of the more striking findings in recent slum research is that global-level estimates and satellite-level evidence can tell opposite stories. A 2024 study applied a minimally supervised deep learning model to sub-meter satellite imagery across 12 cities in low- and middle-income countries over a ten-year period. The model found an average increase in slum area ratios from about 21.6% to 21.8%. During the same period, official UN-Habitat estimates reported the opposite trend: a decrease from roughly 36.5% to 36.2%.6Communications Sustainability. Minimally supervised learning on sub-meter satellite imagery reveals slum expansion during the COVID-19 pandemic – Section: Results
That discrepancy is not a rounding error. It reflects fundamentally different interpretive frameworks. UN-Habitat relies largely on household survey data collected at intervals of several years, aggregated nationally, and smoothed through statistical models. Satellite-based methods capture physical change on the ground at much higher temporal resolution. The satellite study found that nine out of 12 cities experienced an increase in slum area during the COVID-19 pandemic specifically, with an average rise of about 1.1 percentage points. In Cape Town, existing settlements expanded by 3.7% and new ones emerged in a leapfrog pattern across the city.7Communications Sustainability. Minimally supervised learning on sub-meter satellite imagery reveals slum expansion during the COVID-19 pandemic The pandemic, in other words, appears to have accelerated informal settlement growth in ways that global headline figures did not capture.
This gap has real consequences for policy. If decision-makers rely on global estimates showing a slow decline in slum prevalence, they might reasonably conclude that current strategies are working. The satellite evidence suggests the situation on the ground is more complicated.
Residents as Mapmakers
A radically different interpretation of slums emerges when residents produce the maps themselves. In Caracas, Venezuela, a web-based participatory mapping project using Google My Maps trained residents to create online maps for all 13 of the city’s slum corridors, each containing several sectors. For the first time, maps were produced showing all sectors within a single corridor, which had previously been mapped independently of one another, preventing planning at a wider scale.8Habitat International. Web-based participatory mapping in informal settlements: The slums of Caracas, Venezuela – Section: Results: the creation of spatial knowledge in informal settlements The result was not just a technical product but a shift in who held spatial knowledge about these communities.
In Rio de Janeiro, a university-community collaboration in the Santa Luzia favela took participatory mapping a step further, producing a low-cost cadastral survey using drone photogrammetry at two-centimeter precision. Volunteer students and community residents co-produced the cartography, meeting legal requirements for land regularization without expensive professional surveys. The approach integrated counter-cartography, which aims to correct power imbalances in who gets to draw the map, with technical standards required by Brazil’s Social Interest Regularization framework. The result was a legally compliant pathway to tenure security that embedded local knowledge within the official system.9Geographies. Counter-Mapping Informal Settlements: Participatory Cadastral Surveys and Land Governance in the Santa Luzia Community, Rio de Janeiro, Brazil
The contrast with satellite interpretation is instructive. From above, a favela is a dense cluster of irregular structures. From within, mapped by residents, it is a network of named streets, property boundaries, businesses, and social ties. Neither view is complete on its own, but only one carries the lived experience needed to design interventions that residents will actually accept.
Interpreting Risk From the Outside
Environmental hazard assessment represents another interpretive lens, and it often frames informal settlements primarily in terms of vulnerability. In eThekwini, South Africa (the municipality containing Durban), geospatial analysis found that most informal settlements are located in moderate to high flood-prone areas, with about 65.5% in moderate zones and 31.6% in high zones. The key drivers are low elevation, gentle slopes that hinder drainage, and proximity to rivers. Landslide susceptibility followed a similar pattern: roughly 68% of settlements fell in moderate susceptibility zones and 29% in high zones, with steep terrain, unconsolidated soils, and heavy rainfall as the main risk factors.10The Egyptian Journal of Remote Sensing and Space Sciences. Informal settlement vulnerability to multi-hazards in eThekwini, South Africa: An AHP and geospatial approach – Section: Results
These findings are important, but the interpretive framing matters. A hazard map can be used to justify upgrading drainage infrastructure, or it can be used to justify relocating people. Research in KwaZulu-Natal has highlighted that risk and vulnerability to climate-induced disasters like flooding are made worse by socioeconomic factors, especially housing shortages that leave people with no choice but to build in flood-prone areas.11Journal of Asian and African Studies. Understanding Socioeconomic Risk and Vulnerability to Climate Change–Induced Disasters: The Case of Informal Settlements in KwaZulu-Natal, South Africa The risk is real, but interpreting it purely as a feature of the landscape, rather than a consequence of exclusion from safer land, leads to very different policy conclusions.
Heat, Health, and the Urban Environment
Urban heat is another dimension where interpretation of slums diverges depending on the analytical lens. In Dar es Salaam, Tanzania, researchers found surface urban heat islands of up to 6.3°C in informal settlements, compared to 9.2°C in long-urbanized parts of the city. Green and blue spaces protected some informal settlements from the worst heat exposure, and settlements further than four kilometers from the coast with low vegetation cover showed modeled temperature effects of more than 5°C. Surface temperature measurements tracked closely with actual indoor heat exposure measured in a subset of 12 settlements.12PubMed. Spatio-temporal development of the urban heat island in a socioeconomically diverse tropical city
Separately, research comparing informal and formal housing areas found that informal housing contributes to elevating urban heat island intensity more than formal housing, with statistically significant spatial clustering of high surface temperatures around informal structures.13Environment and Urbanization ASIA. Urban Informal Housing and Surface Urban Heat Island Intensity The materials used in informal construction, typically metal sheeting and concrete block without insulation, absorb and re-radiate heat differently than planned development. But again, the question is what you do with the interpretation. Preserving green space near settlements, as the Dar es Salaam research suggests, is a very different response from condemning the structures themselves.
Health risks beyond heat follow a similar interpretive pattern. In Dar es Salaam, analysis of cholera incidence found roughly a one percent increase for every percentage point increase in informal residents within a ward, and about a two percent increase for every additional thousand people per square kilometer. Interestingly, even modest household assets, equivalent to owning a radio, correlated with about 50% lower cholera incidence, suggesting that economic marginalization rather than informality alone drives disease risk.14PLoS Neglected Tropical Diseases. Informal Urban Settlements and Cholera Risk in Dar es Salaam, Tanzania – Section: Results
The Economic Life That Official Data Misses
One of the most persistent misinterpretations of slums is that they are economically inert, places where people live but do not work or produce. Spatial analysis of commercial activity in informal settlements tells a different story. Research applying spatial analysis to visible commercial establishments found a polycentric structure with distinct business clusters in both established centers and peripheral areas. Comparing the density of visible establishments with officially registered firms revealed that informal economic activity concentrates in poor but densely populated areas, and there is a large gap between what official data captures and what is actually happening on the ground.15PubMed Central. Uncovering commercial activity in informal cities – Section: Results
This finding reframes the standard interpretation. If official economic data show no significant commercial activity in a settlement, the usual reading is that the area is economically dependent on the formal city around it. But if the gap exists because informal businesses simply are not registered, the interpretation flips: these are economically active neighborhoods whose productivity is invisible to the systems that allocate public investment. A government that interprets a settlement as a drain on city resources, based on official data, might be looking at a neighborhood that generates substantial economic value but receives almost nothing in return.
What Policy Interpretation Gets Wrong About Tenure
A longstanding interpretation in development policy holds that giving slum dwellers secure land tenure will unlock investment in housing and drive improvements in living conditions. The logic is intuitive: if you know you will not be evicted, you will invest in your home. But research from Nairobi complicates this story. A conjoint experiment with slum residents found that improving land tenure security had almost no influence on residents’ ability to afford upgraded housing. Rent cost was the most important factor determining affordability. Over half of respondents who preferred housing on clearly owned land could not actually afford it. Improving land tenure alone, the researchers concluded, would have limited success in helping residents transition to better living conditions.16Habitat International. Secure land tenure for urban slum-dwellers: A conjoint experiment in Kenya
This matters because it challenges a widespread policy interpretation. If tenure security is treated as the primary lever, other interventions, like rent subsidies, infrastructure investment, or income support, get deprioritized. The evidence suggests that the relationship between tenure and housing quality is mediated by affordability in ways that the standard interpretation underestimates.
Upgrading Versus Relocation and Their Spillover Effects
The two dominant policy responses to slums, upgrading settlements in place or relocating residents elsewhere, rest on fundamentally different interpretations of what the problem is. If the problem is the physical location (hazardous terrain, lack of infrastructure), relocation makes sense. If the problem is underinvestment in an established community, upgrading is the more logical response.
A 20-year study across Chile compared these approaches using satellite imagery, census data, construction permits, crime reports, and property-tax records. Both policies reduced the share of land devoted to housing within the original settlement perimeter, but through different mechanisms: upgrading reallocated land toward neighborhood infrastructure, while relocation simply reduced the number of households. Only in-situ upgrading produced durable improvements in housing quality and residents’ socioeconomic status. Upgrading also generated substantial positive spillovers in nearby neighborhoods, increasing formal housing investment and reducing crime. On top of that, upgrading was roughly a third less expensive per household than relocation.17National Bureau of Economic Research. In-situ Upgrading or Population Relocation? Direct Impacts and Spatial Spillovers of Slum Renewal Policies
The interpretation of a slum as a problem to be removed versus a community to be invested in leads to measurably different outcomes. Relocation, in this evidence, achieves demolition but not improvement. Upgrading achieves both, at lower cost.
Infrastructure Beyond the Grid
Standard frameworks for measuring access to water, sanitation, and energy tend to classify households based on a single primary system: your water is either piped or not, your electricity is grid-connected or not. Research in urban informal settlements has found that this binary framing misses the reality. Residents draw on a range of mostly off-grid water sources and sanitation facilities to cope with inconsistent service provision, creating a complex “service ecosystem” that formal monitoring systems like the WHO-UNICEF Joint Monitoring Programme service ladders do not capture well.18PLoS One. Beyond the grid: Navigating water supply and sanitation service ecosystems in informal settlements – Section: Discussion
Energy access tells a parallel story. In Cape Town’s informal settlements, where state engagement in service provision is minimal or absent, a diverse range of non-state actors, including NGOs, donor agencies, and research institutions, actively develop and operate decentralized solar energy systems to address chronic energy poverty.19Energy Research & Social Science. Who powers the off-grid city? Non-state actors, ethics, and the politics of solar infrastructure in Cape Town’s informal settlements The standard interpretation of “no grid connection” as “no service” misreads a situation where people have cobbled together functional alternatives, sometimes more resilient than the centralized systems that serve wealthier neighborhoods.
How Residents Interpret External Attention
Perhaps the most overlooked form of slum interpretation is how residents themselves read the attention directed at their communities. Slum tourism, which has grown into a visible industry in places like Mumbai’s Dharavi, provides a useful window. Research on Dharavi residents’ perceptions found that most respondents were happy and proud that outsiders were interested in their community. Tours there focus on the commercial versatility and industriousness of the inhabitants, and the narrative is framed to leave visitors with a more positive picture. Residents saw foreign interest in their way of life as a positive thing, believing the tours highlighted the strengths of their community rather than dwelling on poverty.20Tourism Review International. “Breaking the Silence”: Local Perceptions of Slum Tourism in Dharavi – Section: Dharavi residents’ perceptions of tourism
This complicates the instinct, common among outside observers, to view slum tourism as inherently exploitative. The residents’ own interpretation is more nuanced: they distinguish between attention that frames them as objects of pity and attention that acknowledges their agency and productivity. The same neighborhood can be interpreted as a symbol of urban failure or as a functioning community with a dense economic and social fabric, and residents are well aware of which framing is being applied.
Historical Roots of the Interpretation Problem
The tension between these competing interpretations is not new. Research on nineteenth-century British urbanism has traced how town halls became involved in removing substandard dwellings, redesigning neighborhoods, and building housing estates in the suburbs. Environmental conditions improved and mortality rates fell, but the reform effort hit its limits when upgrading projects collided with the private housing market. The slums, the historian notes, remained in place at the outbreak of the Great War.21Journal of Urban History. From Hygeia To the Garden City The pattern of interpreting slums as a sanitary problem, intervening just enough to reduce the most visible harms, and then stopping short of structural change has repeated across continents and centuries. What has changed is the range of tools available to measure and contest the dominant interpretation, from satellites and drones to participatory maps drawn by the people who live there.