Touch thee not: Group conflict, caste power and untouchability in rural India

TL;DR: In this article, the authors investigate the impact of community power on the practice of untouchability by upper and backward caste Hindus vis-a-vis scheduled castes (SCs) in rural India.

About: This article is published in Journal of Comparative Economics.The article was published on 2021-06-01 and is currently open access. It has received 5 citations till now. The article focuses on the topics: Caste & Untouchability.

Summary

1. Introduction

• This paper seeks to identify the determinants of an extreme form of social discrimination, viz., the practice of caste-based untouchability in India.
• Empirically, the village-level distributions of population and land ownership across caste and religious communities jointly turn out to explain much of the incidence of untouchability among upper and backward caste Hindus.
• Extensive, and often violent, conflicts between SCs and upper castes or OBCs, as well as between Hindus and Muslims, continue to constitute arguably the most salient features of the political landscape in India.

2. The structure of untouchability in rural India

• The India Human Development Survey-II (2011-12) is a nationally representative, multi-topic survey of rural and urban households drawn from across Indian states.
• The percentages in Table 1 can thus be interpreted as the likelihood of untouchability being practised, or its incidence, in the rural areas of the major Indian states.
• Unsurprisingly, Brahmins are most likely to engage in the practice, with almost 6 out of every 10 Brahmin households admitting to it.
• Table 2 further reveals that the caste blocs most susceptible to the practice of untouchability, viz., upper/general castes and OBCs, are also those who own most of the land in rural India.

3. A theoretical model of within-village group conflict over social norms

• The authors now develop a theoretical model to highlight how the inter-community distribution of resources within a village may jointly determine the prevalence of untouchability therein.
• The authors visualize a scenario wherein different caste and religious blocs within a village initially engage in multiple simultaneous contests over a matrix of social values and attendant behavioural norms.
• Greater success of Hindu ‘general’ or backward castes in these contests implies greater legitimacy for the norms of ritual purity traditionally espoused by such communities.
• All individuals take this normative structure as given and choose their behaviour according to a personal cost-benefit calculus – they trade off their benefit from indulging their idiosyncratic ‘taste’ for practising untouchability against the cost of doing so.
• The cost is determined by the collective value-system espoused by the village, arrived at through the initial process of inter-group contestation.

3.1. Group conflict and equilibrium determination of village norms

• Suppose the population of a village can be partitioned into four communities: Hindu general (or upper) castes (U), Hindu OBCs (B), Hindu SCs (S) and non-Hindus (M).
• They may both, therefore, potentially engage in a contest over the division of T with S and another over the division of R with M. Political contestations among communities may play out in many forms in practice.
• Together, (3.6) and (3.9) imply that the marginal utility from expenditure on material consumption must be greater than that on conflict of any kind for U in equilibrium if that community is significantly poorer than B in terms of its resource endowment.
• Proposition 2. Let Assumption 1 hold, and let 𝑔𝑔𝑆𝑆𝑀𝑀∗,𝑔𝑔𝑀𝑀𝑀𝑀∗ be the values of 𝑔𝑔𝑆𝑆𝑀𝑀 and 𝑔𝑔𝑀𝑀𝑀𝑀, respectively, in a Nash equilibrium corresponding to some initial configuration of 𝜌𝜌𝐻𝐻 ,𝜌𝜌𝑆𝑆,𝜌𝜌𝑀𝑀 and 𝜃𝜃.
• Thus, given the total resource endowment of the non-SC Hindu community, a reduction of the dominance of backward castes vis-à-vis upper castes has the effect of making villages norms less tolerant of untouchability.

3.2. Village norms, individual behaviour and the incidence of untouchability

• The greater the value of 𝑔𝑔𝑆𝑆𝑀𝑀, the lower the collective tolerance of untouchability, hence the greater the cost to upper caste and OBC individuals from its practice.
• The combined effect of such a redistribution of land and/or population share from upper castes and OBC Hindus to SCs and/or nonHindus must, by Corollary 1, reduce the proportion of the former that practises untouchability (𝜇𝜇).
• Note that, as can be easily seen from Proposition 1, the model becomes scale neutral under the additional assumption 𝐹𝐹(𝑛𝑛𝑖𝑖) ≡ ln𝑛𝑛𝑖𝑖.
• Thus, only the land and population shares of the different communities would matter, not the total land or population endowment of the village as a whole.

4. Empirical strategy and findings

• To this end, the authors convert IHDS 2012 rural household-level data into village-level, focusing on Hindu non-SC/ST households, to conform to their theoretical model.
• This gives rise to a sample of about 13000 households drawn from about 1100 villages.

4.1. Empirical Strategy

• Section 3 defined the measure of a community’s power (i.e., its normalized resource endowment) simply as the population share weighted land share of the community (recall Remarks 2 and 3).
• Thus, for each village in their sample, the resource base or power of the OBC community, 𝜌𝜌𝐵𝐵, is empirically measured by the population share of OBC households in that village multiplied by the proportion of the total village land owned by OBC households therein.
• The authors key outcome variable is a measure of the likelihood of practising untouchability by the upper caste (i.e., Brahmin and Forward caste) and OBC Hindu households within a village.
• 15 According the latest (2011) Indian Population Census, Sikhs, Jains, Parsees and Buddhists collectively constitute only about 3.5% of the Indian population.
• The set of variables X includes the village’s distance from the nearest town, the presence of outside workers, all-weather roads as well as government and private primary schools.

4.2. Empirical findings

• The authors key regression results are presented in Table 4.
• Thus, reduced power of non-SC/ST Hindus within the village (i.e., lower 𝜌𝜌𝐻𝐻), when associated with a corresponding increase in the collective power of either SCs or non-Hindus (or both), implies significantly lower propensity of Hindu upper caste and OBC households living in the village to practise untouchability (Corollary 1(i)-(ii)).
• Thus, the broad patterns revealed earlier by Table 2 are confirmed by their regression analysis.
• Recall now the finding that having a larger proportion of households where the head has studied at least till class 5 does not reduce H households’ propensity to practise untouchability.

5. Policy considerations and concluding remarks

• Instead of cataloguing the central findings of this paper all over again, the authors close with a brief discussion of some policy questions brought to light by their analysis.
• Any increase in the community power of SCs reduces the propensity of upper caste and OBC Hindus to practise untouchability (Table 4).
• Measures which encourage cross-village migration by SCs, to foster greater locational consolidation of that community, may reduce untouchability directly, by increasing the population share of SCs, in the villages of SC consolidation.
• These features indicate greater market integration, and consequently greater exposure to outside influences and modern value systems.
• Nor have the authors addressed untouchability in urban settings.

Figures

Table 4. OLS estimates of untouchability practised by Hindu non-SC/ST Hindu household

Table 1. Incidence of untouchability in rural areas of Indian states

Table 3. Variable definitions and summary statistics in the sample villages

Table 2. Community-specific incidence of practising untouchability (rural all-India)

Frequently asked questions

Q1. What have the authors contributed in "Touch thee not: group conflict, caste power, and untouchability in rural india" ?

The authors investigate the impact of community power on the practice of untouchability in rural India. 

Q2. What future works have the authors mentioned in the paper "Touch thee not: group conflict, caste power, and untouchability in rural india" ?

Empirical examination of the relations between rural connectivity and urban economic growth on the one hand, and the incidence of untouchability on the other, would constitute an important line of future research. Future work may usefully focus on these aspects as well. 

Q3. What are the main factors that influence sanitation in rural India?

Beliefs in purity and pollution contribute to the acceptability of widespread open defecation and the rejection of inexpensive latrines in rural India. 

Q4. What is the effect of greater market integration on the incidence of untouchability?

Since greater market integration typically implies greater competitive pressures, it is also likely to increase individuals’ costs of satisfying their ‘taste’ for discrimination and thereby reduce discrimination, as suggested by Becker (1957). 

Q5. What are the key determinants of caste power in India?

Since land ownership is the principal source of economic power in rural India, and numbers matter in India’s electoral democracy, numerical strength and land share constitute the key determinants of caste power in this definition. 

Q6. How do the authors minimise the bias of their estimates?

In order to minimise the extent of the omitted variable bias of their estimates the authors also include a vector of village characteristics X to minimise the omitted variable bias of their estimates. 

Q7. What is the role of the determinants of untouchability in the study?

Empirical examination of the relations between rural connectivity and urban economic growthon the one hand, and the incidence of untouchability on the other, would constitute an important line of future research. 

Q8. What is the effect of caste contests on the outcome of untouchability?

The outcomes of such caste contests are however affected by a simultaneous process of religious conflict between Hindus and Muslims (or Christians). 

Q9. What is the effect of ggMMMM on the extent of untouchability?

This suggests that, in villages dominated by Muslims or Christians, greater resource acquisition by them may increase the extent to which the practice of untouchability is12 

Q10. How do the authors abstract from the issue of preference differences and coordination problems within each community?

The authors abstract from the issue of preference differences and coordination problems within each community, by assuming that each community can be modelled as an individual. 

Q11. How many Brahmins are likely to engage in the practice?

Brahmins are most likely to engage in the practice, with almost 6 out of every 10 Brahmin households admitting to it. 

Q12. What is the main reason why Brahmins and Forward castes are assumed to be?

in their theoretical model presented in Section 3, the authors shall model Brahmins and Forward castes as constituting one unified group, while OBCs will be assumed to constitute another.