GaWC Project 61

GaWC logo
  Gateways into GaWC

Neglected Spatiality: Economic Geographies of Slums in Mumbai and Johannesburg

Funded by: National Science Foundation, USA (2007-2009)

National Science Foundation

Principal Investigators: Richard Grant and Jan Nijman, University of Miami, USA



Slums have dotted the landscapes of many large cities in the less developed world since early times. With increasing urbanization rates in the second half of the 20th century, they have in most cases proliferated. Presently, it is estimated that close to a billion people world wide live in urban slums. If current trends continue, the number of slum dwellers worldwide is projected to double over the next thirty years (UN-Habitat 2003:14), and some are now speculating about a future “planet of slums” (Davis 2006).

This proposal focuses on slums but it does so in an unconventional manner. Our study recognizes the severity of problems of inadequate housing and poor living conditions – we have addressed these issues elsewhere – but that is not the subject of inquiry in this proposal. Instead, we direct attention to the economic geography of slums - slums not as places of residence but as sites of economic production. There are plenty of examples from descriptions in the popular news media and in more serious writings, from pottery or shoe ware in Dharavi (Mumbai) to curios or plastic containers in Klipspruit Valley (Johannesburg). Examples of such economic activities seem to be characteristic of major slum areas all over the less developed world (Chen 2004; Rogerson 2006).

We advance a general theoretical argument that economic activity is an important function of slums – and we are strengthened in our estimation by our observations in the field over the past decade or so. To date, there are many impressionistic accounts and some scholarly ethnographic studies but no systematic study has been undertaken of the nature of economic activity in slums, its rootedness in the slum and its connections to the wider urban economy. Hence, we propose a new line of theoretical and empirical inquiry that addresses this neglected spatiality and that is intended to contribute a more complete understanding of slum-spaces.

In the proposed study, we first, elaborate a theoretical argument about the nature of economic production in slums. We define production broadly to include manufacturing and services but we exclude retailing (for research on the latter, see, e.g., Dierwechter 2004; Ligthelm 2005). Our theoretical argument is that certain slums located in proximity to urban centers provide local milieus for slum firms to engage in specialized economic production. Global cities of the South affected by economic globalization and liberalization and accompanying trends of flexible production provide a new context for linking firms to particular external economies (city, region, the state, abroad).

Second, we will conduct extensive and parallel empirical studies in Dharavi and Klipspruit Valley, two well-known and large slum areas in Mumbai and Johannesburg. Mumbai and Johannesburg are global cities and as such they highly connected to the world economy. The research design will involve household surveys to document socioeconomic characteristics and particularly issues related to work (including place(s) of work in relation to home). On the basis of this data and field site reconnaissance, we will implement a second round of surveys, this time not of households but of firms that are based in the slums. The second survey will focus on the nature of economic activities, the significance of local milieu, and linkages to wider urban economy. This information will be enriched through a series of in-depth interviews with selected firms. Our analysis will also involve tracking the production of slum firms through value chain analysis.


Mumbai and Johannesburg are the main world cities in, respectively, South Asia and Sub-Saharan Africa (see, e.g., Taylor 2004: 74). These cities embody what is sometimes referred to as the ‘global south’: situated in the less-developed world but strongly connected to the world economy in terms of the global division of labor, capital circuits, movement of people, information technologies, and commodity flows. Their global connections and gateway functions are reflected in the space-economy, with global city functions clearly articulated in the urban landscape (Grant & Nijman 2002).

At the same time, Mumbai and Johannesburg have large slum populations (around 50% in Mumbai and about 25% in Johannesburg). Indeed, the globalization of these urban economies during the past couple of decades has been accompanied with absolute and relative increase of slum populations. Both cities form the backdrop to interesting and sometimes heated public debates about the role of slums and slum dwellers. These debates center on the tension surrounding the coincidence of globally competitive business and extreme economic marginalization in the same city. The “Cities Without Slums” initiative (Cities Alliance 2001) represents a shared vision among the global and local policy elites about the future of these cities but urban boosters argue for slum eradication for entirely different purposes. Some portray slums dwellers as ‘parasites’ for not contributing meaningfully to the urban economy while others espouse the view that without the slums, the urban economy would grind to a halt.

Much has been written about Mumbai’s slums, especially with regard to housing and living conditions. One of the best historical accounts of Mumbai’s slums in the city’s overall spatial development is by Dossal (1991) and recent works include Sundaram 1989; Mahadevia 1998; Deshpande 2004; Swaminathan 2003; Das 2003; Shaw 2004; Desai 1995; Verma 2002; Vora & Palshikar 2003; and Seabrook 1987, 1996. Useful appraisals of recent government policies regarding slum rehabilitation can be found in Ruiter 1999 and Mukhija 2003. For Johannesburg, one of the best accounts of slums in the context of spatial development within a divided city is Beall et al. 2002. Other important works are Beavon 1998; Bremner 2004; Tomlinson et al 2003; and Gordon & Nell 2006. The most comprehensive review of slum rehabilitation policies in South Africa is provided by Huchzermeyer (2004).

Within the two cities, we concentrate the empirical analysis on two slum areas: Dharavi in Mumbai and Klipspruit Valley in Johannesburg (for general overviews, see Morris et al 1999, Cranksaw et al 2000; Mears 2004; Rogerson & Bevon 1992; Dossal 1991 and Sharma 2000. Both are large, dense, and mostly contiguous areas with long histories dating back to the first half of the 20th century. Both have a stable milieu for economic production compared to newer slums (we elaborate more on this argument in the theory section). We have observed all kinds of economic production at both sites (e.g. sorghum beer brewing facilities in Klipsruit and pottery in Dharavi). The exact slum population in both areas is unknown: estimates range upwards from 300,000 in Klipspruit and from 500,000 in Dharavi. Both consist of a large number of slum communities with more or less visible boundaries. These communities are often formed on the basis of the regional origins of the slum dwellers when they migrated to the city and they are also often associated with ethnicity. Sometimes, certain smaller areas are specifically known for the type of in situ economic activity/production. Both Dharavi and Klipspruit are located in relative proximity to their respective central business district, likely to be of importance for labor as well as for production linkages. But little is known about the nature of production, its embeddedness in the slums, importance for local labor, and connections to the urban economy.


We identify four strands of literature that are important to our research. The first two literatures, on global cities and on the new economic geography, are theoretically relevant but neglect the places of slums. The other two literatures both focus on slums and are important in understanding the local milieu of our research, but they neglect (theoretical) questions about their economic geography.

Global Cities in the Less Developed World

Considerable research efforts have been made to understand the spatial form of cities in the less developed world (Chakravorty 2000; Grant & Nijman 2002; Marcuse & Van Kempen 2000) as well as to extend the boundaries of research on ‘global cities beyond the West’ in the context of recent economic globalization processes (Gugler et al 2004; Short et. al 2000; Brown et al 2002). This literature shows that global cities of the less developed world have come to share common spatial patterns (Murray 2004).

In our previous research, we argued the importance of research from the ground-up in selected world cities in order to capture the complexity of economic relations and a city’s position in global networks. This becomes an even more pressing matter when studying the role of slums in globalizing urban economies in the South. We acknowledge anthropologists’ bottom-up research on slums in both cities (e.g., in Soweto: Ashford 2005; Kramer 1985; in Dharavi: Fuchs 2003; Sen 2004) that examines the broader cultural realm but this work lacks an explicit urban-economic focus. Our aim, therefore, is to contribute to the global cities literature by shedding light on the geographical reach of globalization processes deep into these urban economies. We want to understand better how slum economies fit or are excluded in global cities of the South.

The New Economic Geography, Industrial Location and Agglomeration

We draw on some of the theoretical arguments of the literature on “the new economic geography’ and their application to the economic geography of slums. Recent developments in economic geography have received various labels including the “cultural turn,” the “institutionalist turn,” or, more generally, the “new economic geography.” (e.g., Amin & Thrift 1992; Scott 2000; Thrift 2000; Yeung 2003; Amin 2004). These trends comprise a large and varied research efforts but what they have in common is a renewed emphasis on what Scott (2000:31) refers to as the “stubborn locational rootedness” of production. Put differently, there is renewed interest in the question of the geographic concentration of economic activity. Most important, for our purpose, is the emphasis in some of this literature on cultural and institutional conditions that are specific to places or regions and that shape the presence of production and agglomeration.

Thrift (2000: 689) referred to the cultural turn in economic geography as “the most important event to impact on the sub-discipline in the last ten or fifteen years.” So important, indeed, that “economic geographers have become some of the leading exponents of cultural geography.” (692). Slums, in our view, represent a spatiality that is often at once cultural and economic. They are place to live but they are often also workplaces. To understand slums as sites of production, it is necessary to study their economic geography in cultural and institutional context.

Slums: Housing and Living Conditions

This literature deals with urban slums in the less developed world, has a specific spatial focus, and concentrates on housing, living and environmental conditions, and slum upgrading/rehabilitation. Attention often centers on the quality, affordability and availability of housing. Generally, this literature draws on the more common physical conception of a slum in terms of poor housing and residential infrastructure and insecurity of tenure. Important early contributions were based on detailed empirical descriptions of slums in a variety of ‘third world’ cities (e.g., Abrams 1964; Clinard 1966; Juppenlatz 1970; Turner & Fichter 1972). These kinds of studies set the stage for a broad range of prescriptive and policy-oriented work in more recent times (e.g., Van der Linden 1987; Huckzermeryer 2004). Public policy scholars are well-represented in these works, and a smaller purely academic tradition has continued in terms of the detailed ethnographic studies, mainly by anthropologists (e.g., Peil 1976; Pellow 2002). Geographical contributions have been sparse throughout (e.g., Gilbert & Ward 1985; Eyre 1990; Oldfield 2002; Olds et al 2002) and, at any rate, much of the existing literature on slums predates recent theoretical developments in (economic) geography.

Recent years have brought increased attention to slums in the news media, popular books, and academic writings (e.g., Neuwirth 2004; Mitlin & Satterthwaite 2004). A landmark publication was the UN-Habitat report The Challenge of Slums: Global Report on Human Settlements of 2003. The Report that has been already labeled “historic” and the most comprehensive document to date on the global problems of slums. The authoritative warning in the Report about the worldwide catastrophe of urban poverty has been equated with the reports of the Intergovernmental Panel on Climate Change that represented an unprecedented scientific consensus on the dangers of global warming (Davis 2006).

Livelihoods, Labor, and Informality

The fourth literature, finally, pertains to urban livelihoods, labor, and issues of informality. Here, the focus is not so much on the slum as a place but rather on the economic plight of the poor (slum dwellers, very often) and on their role as providers of labor in the urban economy – this literature usually lacks a geographical focus.

There is a myriad of studies about urban informal economic activities following its “discovery” in 1973 by Keith Hart. Research on the informal economy has been extensive, mirroring the growth of the phenomenon itself, and it is not possible to provide a comprehensive review on these pages. We can say that the largest effort in informal economy research has been to uncover the size, determinants and characteristics of this so-called “elusive socio-economic phenomenon” (Alderslade et. al 2006). Even more important, a broad consensus appears to be emerging that the informal economy is changing in complexity, expanding, becoming more segmented, and perhaps more exclusionary in a “new poverty” than what prevailed in the twentieth century (Special issue of Latin American Research Forum 2004; UN-Habitat 2003; Huckzermeyer 2007).

Our study seeks to contribute to new thinking about the interrelationships and interdependencies between the informal economy and “official” economic activity (Rogerson & McCormick, 2004). As flexibilization becomes more widespread, the media is reporting how subcontracting networks of formal companies are now “extending deep into the misery of urban slums” (Business in Africa 2006:3). Rogerson (1997:340) invoking Storper’s (1992) industrial geography framing suggests that slums may “form a necklace of localized production agglomerations strung out around the world.”


We are interested in the nature of economic production in slums and in the manner and extent of integration of this production in globalizing urban economies. The research question and study design are based in existing theory, but modified to fit the context of slums. The premise is that slums as economic sites are an important “neglected spatiality” in urban and economic geography. Thus far, slums have been studied one-sidedly as places of residence but they are better understood as places of living and working.

Moreover, the conventional liberal economic assumption of slums as an economic category of dysfunctional exclusion is an unsatisfactory explanation. Instead, our thinking is informed by recognizing a myriad of uneven relationships between slums and the wider space economy. The ways in which the production in slums is integrated in the urban economy is likely to shed light on questions of formality/informality, integrated modes of production, and the overall spatial organization of the urban economy.

We present a theoretical argument as to why it is to be expected that production takes place in slums, i.e., why production inside slums makes sense; and why this production is likely to be differentiated as well as stratified in terms of connectivity. Our reasoning is informed by value chain research on small/informal enterprises (Kaplinsky & Morris 2000; McCormick & Schmitz 2001) that specifies the interconnectedness of individual enterprises and links within a chain of production. To our knowledge there has been no research (to date) on examining the production in slums through the value chain lens. The distinction between links which are enterprise-to-enterprise and enterprise-to-many enterprises and between vertical (up and down the chain) and horizontal links are useful starting points to our research effort. Firms in slums are also likely to vary in terms of their relationship with intermediaries that shape their connections to markets. Firms may hold a more or less subordinate position in the value chain. It is not clear to what extent production is low-profit, non-core function that is externalized to a competitive and decentralized system of subcontractors (Gibbon & Ponte 2005). It is also possible that production in slums is highly specialized and involves relatively scarce skills, suggesting significant vertical linkages and relatively high added value. We expect our research to shed light on these and related matters.


The comparative empirical analysis focuses on economic activities inside slum areas in Mumbai and Johannesburg. We document the nature of production, how it ‘fits’ the milieu of the slum, and how this production is connected to the wider urban economy. The analysis concentrates on Dharavi in Mumbai and Klipspruit Valley in Johannesburg. Given the new directions of our research, the study design is largely of an exploratory nature. Our methodology is purposely eclectic to fit the needs of a new economic geography that acknowledges the social embeddedness of economic action and that is sufficiently reflexive (Yeung 2003). It combines surveys, quantitative analysis, qualitative methods, attention to discursive practices about slums, network analysis, and mapping. We will reference all relevant existing data, but the essence of the study is built around primary data collection. The scope of this kind of primary data collection, as far as we know, is unprecedented.

The analysis is driven by two sets of questions:

1. What is the nature of economic activity in slums? What kinds of sectors/products? Where does production take place (home/workshops)? Who controls production? What explains the location of production? What is the nature and extent of agglomeration and clustering?
2. How is this activity connected to the wider economy? What are the relations to suppliers, buyers, associates, partners, producer services, owners? Where do the products end up? How important is this production site to this particular sector, slum and the overall urban economy?