I am interested in understanding the economics of slums. UN statistics suggest that nearly 1 billion people live in slums. In India, a country whose economy I study, nearly 100 million people live in slums. The topic is so important, aside from COVID-19 research the last 2 years, it is the topic I allocate the most time to. My U. Chicago Law School colleague Adam Chilton and I began a multi-year ethnographic project to understand the economics of slums. This effort produced a documentary, Basti, directed by the talented Parinaz Jal. We are now working on a book project. In addition, I have worked with a number of graduate students to gather and analyze quantitative data on slums. Hopefully you see some articles soon.
In the interim, I am going to write some Substack posts on the economics of slums. These posts will preview some of the points I will make in our book and articles, and cover some topics were cut from those pieces due to space constraints.
One of the first questions I get when discussing slums is what I mean by the term. How do I define a slum? Anyone who has been to a lower-income country has likely seen one. The first image to come to my head is the cluster of slums I would see every time I would fly into Mumbai to visit my family. They lived in the neighborhood of Dadar, but it does not matter where they lived. The airport was surrounded by slums that you had to drive through to get anywhere from the airport in Mumbai. But “I know it when I see it” is an unsatisfactory definition of a slum. At the very least, it does not help the many people who have not actually seen a slum understand the object of the conversation.
There are three reasonable definitions of a slum
Slums are communities with low-quality housing
Slums are communities where residents have no formal right to live where they do.
Slums are communities officially recognized as slums by the local government
Of these definitions, I prefer the second one. It explains why slums have low-quality housing. It also explains why communities with low-quality housing that are legally recognized by the government as slums often begin to grow faster than communities with similar housing that are not recognized by the government. It is also a definition that clarifies why slums are more than a product of poverty and require different solutions than the usual solutions to poverty. I elaborate on these points in the remainder of this post.
1. Low-quality housing
The first, and likely most popular, definition of a slum is a dwelling that falls below some minimal bar of quality. For example, it may be a residence without solid (concrete or brick or wood) walls or an adequate roof. It may be a residence without running water, without a private toilet, or without electricity.1
The problem with this definition is that it is overbroad. Historically, non-solid walls, inadequate roofs, and lack of electricity might characterize the majority of homes. It doesn’t seem helpful to think of housing before 1900 as mainly slums. There are many homes that do not have running water, a private toilet, or working electricity. Many rural cabins do not have running water. My college dorm had communal toilets. And electricity is a problem across economic classes in low and middle-income countries. Heck, even in my reasonably well-off neighborhood (Wicker Park) in Chicago there is a mansion on the corner that was abandoned, has boarded-up windows, is not heated, leaks, and probably has non-functional sewers. I don’t think people think of any of those dwellings as slums.
A savvy response is to use a combination of these criteria to identify slums: e.g., you have to have 4 out of 5 of poor walls, poor roofs, poor water, poor toilet, and poor. You could even add in a criterion line income of the resident, though that is difficult to measure. The problem with such multifactor definitions is that the lines are arbitrary and where you draw them can dramatically change the number of homes that qualify as slums.
2. No formal rights
A second definition — the definition I think makes the most sense — is that a slum is a community of squatters, i.e., a community of dwellings occupied by people who have no formal right to live in those dwellings. More precisely, a slum resident is one who has neither formal ownership of the land under their dwelling nor a formal leasing contract to live in their dwelling.
I stress formal rights rather than informal rights for two reasons. First, a formal right provides greater security than informal rights. The biggest threat for a slum resident — and the threat that makes a slum resident’s life much worse than that of an equally poor person living in a non-slum — is that of arbitrary eviction and perhaps home demolition. A formal right to land or an apartment would allow residents to block eviction and demolition, or obtain damages for wrongful slum clearance. An informal does neither, creating a risk of loss that deters investment in the home or community.
Second, slum residents often have informal rights to stay on the land. By informal rights, I mean rights that are enforced not by the government, but by the community. These are important to understanding why slums grow despite the risk of demolition.2 But informal rights provide qualitatively less protection than formal ones. This explains why slums have housing markets, but those markets produce both low-quality housing and misallocation of that housing.
To summarize, this definition of slums has the functional advantage that it can explain why slums have lower quality housing than non-slums. Slums are worse because they lack well-defined property rights, which in turn inhibits investment in housing and efficient trade of housing.
This rights-based definition is better than the low-housing-quality definition for two reasons. First, it explains why slums have low-quality housing. Second, it shed light on why a house in a slum should be treated differently than an equal-quality house in a non-slum. The latter will fetch a higher price and be improved at a faster rate than the former because it faces less risk of dislocation.
3. Recognized slums
A third possible definition of slums is communities that are legally recognized by the state as a “slum”. There are three problems with this definition. But before we get to them, we need to understand the process by which governments recognize communities as slums.
Consider all the communities where residents do not have formal rights to their land or dwellings. As I explained, these communities have low-quality housing because they lack property rights and therefore don’t invest in their homes.
The government will recognize some — but not all — of these communities as slums. Which ones? Those that live on public land and whose residents are sufficiently organized to lobby the government for recognition. The government has no right to modify the rights of private landowners. And governments are responsive to lobbying.
Why do communities lobby for recognition? Because recognition comes with commitments from governments to provide some degree of compensation in case the government demolishes homes in those communities. This compensation is less than the compensation that the government would have to provide if residents had formal rights. But it is better than nothing. Importantly, the promise of compensation increases the cost of demolition to the government, and thus discourages the government from demolishing recognized communities.
This dynamic makes clear the shortcomings of the government-recognition definition of slums. First, it omits communities on private lands. Second, it omits the worst communities. Unorganized communities not only fail to lobby successfully for recognition but also are not organized enough to overcome internal collective action problems and improve their communities. Third, recognition by the government, because it brings some degree of protection against demolition, leads communities to invest in improving their homes. This again causes this definition of slums to identify better-off rather than worse-off communities as slums.
Bigger lesson
I am done justifying my preferred definition of slums. But this discussion has a second purpose: it explains what makes a good sociological definition. A definition should be functional, i.e., it should be one that gets at root causes and helps clarify causal pathways. Looking instead at outcomes just requires defining another term that captures the proximate or root cause. Delegating labeling to the government may lead to misclassification error because governments and government officials have political interests or career concerns rather than research interests.
One could go even further and suggest that a slum is a housing unit that does not meet local building codes. The building code definition has even bigger problems. Building codes vary across countries. A home that barely fails Chicago’s requirements might be considered a great home in Chennai. Surely we want constant standards across locations. Further, code violations do not always get at quality. For example, many cities like Mumbai place limits on how tall buildings can be, partly for safety. These codes were written when city populations were small and without regard to the latest building technology. They don’t change because burdensome regulations offer opportunities for officials to get bribes for not enforcing them. But surely buildings that are taller are not slums even though they violate codes.
A related reason to use the formal-rights definition is that resolves the problem of what to do about certain, slum-like low-quality housing on private land. In India perhaps up to 50% of slum-like communities are on private land. Why would private landowners allow such communities on their land? Because residents pay rent. But then why should one think of them as slums under the second, no-formal-rights definition? Because these landowners do not sign formal rental agreements with residents of these communities, for two reasons. First, if they did they would trigger a range of tenants’ rights laws that make it difficult to evict residents, even if they stop paying rent. Second, formal rental buildings have to comply is certain housing quality standards. By leaving these rental agreements informal, private landowners can have their cake and eat it too: generate additional revenue without additional inflexibility or cost. These communities on private land qualify as slums because residents have only informal rights to dwellings. And those informal rights also explain the low quality of housing in these slums.