A+| A| A-
Neighbourhood-scale Residential Segregation in Indian Metros
Residential segregation studies in Indian cities have relied on ward-level data. For a typical ward, the neighbourhood–ward dissimilarity index is greater than the ward–city dissimilarity index. Using 2011 enumeration block-level census data for five major cities in India—Bengaluru, Chennai, Delhi, Kolkata, and Mumbai—it is shown how patterns of caste-based urban residential segregation operate in contemporary India. The first visual snapshot of caste-based residential segregation in an Indian city is presented using geo-referenced enumeration block-level data for Bengaluru.
A previous version of this article was published as a working paper on the Indian Institute of Management Bangalore website.
Extant studies of residential segregation in urban India use the ward as the principal unit of empirical analysis (Sidhwani 2015; Singh et al 2019; Vithayathil and Singh 2012). Vithayathil and Singh (2012) use ward-level data (Census 2001) to find that caste-based residential segregation is more pronounced than segregation by socio-economic status. A similar study by Sidhwani (2015) showed that the percentage of Dalits and Adivasis in a neighbourhood was correlated to the quality of public good provisioning. A recent longitudinal study by Singh et al (2019) using the 2001 and 2011 Census data, showed that residential segregation persisted or worsened in 60% of Indian cities. One of the most serious limitations of all these studies is that they use ward as their unit of analysis, and this might under-report the intensity of residential segregation in Indian cities. Even when a ward is diverse in terms of caste composition, the communities are often highly segregated within a ward. In Bengaluru, for example, upper-caste neighbourhoods are often abutted by highly dense lower-caste settlements. Thus, even when spatially proximate, the social distance between these neighbourhoods can be very high (Shaw 2012). Clustering can happen even at a micro street level, with households from two different castes occupying adjoining streets, or even two different sides of the same street. Hence, the ideal unit of analysis must be still smaller, say a street. Given the compact spatial spread of an urban census enumeration block, we argue that they represent a good proxy for neighbourhoods.
We provide robust evidence for intra-ward segregation for five metropolitan cities of India—Delhi, Mumbai, Kolkata, Chennai, and Bengaluru. Of these five cities, we are also able to visually demonstrate such segregation for Bengaluru using geo-referenced enumeration block-level data. To the best of our knowledge, this article is the first ever attempt to use enumeration block data (released in 2015) to study caste-based residential segregation in India, and also provide a visual portrait of micro-level intra-ward segregation.
Dear Reader,
Or
To gain instant access to this article (download).
(Readers in India)
(Readers outside India)
