Isolated by Caste: Neighbourhood-Scale Residential Segregation in Indian Metros

Abstract

We present the first ever neighbourhood-scale portrait of caste-based residential segregation in Indian cities. Residential segregation studies in Indian cities have relied on ward-level data. We demonstrate in this paper that wards cannot approximate an urban neighborhood, and that they are heterogeneous. For a typical ward, the neighbourhood-ward dissimilarity index is greater than the wardcity dissimilarity index. Using 2011 enumeration block (EB) level census data for five major cities in India -Bengaluru, Chennai, Delhi, Kolkata, and Mumbai -we show how patterns of caste-based urban residential segregation operate in contemporary India. We also present the first visual snapshot of castebased residential segregation in an Indian city using georeferenced EB level data for Bengaluru. Besides implications for policy, our analysis also points to the need for publicly available, geospatially-linked neighborhood-scale census data that includes data on economic class for a spatial understanding of economic and social stratification within Indian cities.

A 'Ghetto' of One's Own: Communal Violence, Residential Segregation and Group Education Outcomes in India

“…Susewind (2017) provides a description of segregation across religious lines in India using electorate rolls at the ward level for 11 cities 6 . However, Bharathi et al (2018) demonstrates that ward level variation does not fully capture concentration of social groups in various parts of a city, and therefore, document segregation along caste lines at neighbourhood levels.…”

“…I have used quasirandom variation in communal violence to measure community (or religion) based segregation, as this is a displacement shock that affects Muslims only. In order to verify this, I check if the events of 1990 have any import on residential segregation of Scheduled Castes, who have also been documented to live in segregated enclaves (Bharathi et al, 2018). This group is not expected to get displaced due to communal conflicts.…”

“…To check the validity of this correlation, I then look for treatment effects on 'untreated' groups. Displacement on account of communal riots between Hindus and Muslims should not affect residential patterns of the Scheduled Caste (henceforth, SC) population, which has also been documented to live in segregated neighbourhoods in Indian cities (Bharathi et al, 2018). Therefore, SCs are being held as the untreated group.…”

Intergroup contact and its effects on discriminatory attitudes Evidence from India

“…This implies that caste may be an important factor in an individual's housing decisions in this city. Recent work by Bharathi et al (2018) provides evidence of higher levels of segregation at the intra-ward level than at the inter-ward level in Indian cities, which increases the need for more reliable neighbourhood-level segregation measures in urban India. The policy experiment I use allows me to define a neighbourhood at a precise and granular level, which can contribute to the discussion on intra-ward segregation.…”

Bibliography

“…Miklian and Sahoo (2016) look at segregation in three Indian cities, focusing on both Muslims and SC/STs. Systematic quantitative measurement of segregation in India has been elusive, however, because measures of segregation require comprehensive neighborhood-level data across the country Bharathi et al (2018). Data on settlement patterns for small neighborhoods have not previously been widely available.…”