The recent California State vs. Cisco case has pressed a nerve of discrimination in the liberal, modern workplace, one that was viscerally experienced by lower-caste persons for a long time.
The key allegation of the Dalit employee was twofold – their caste identity was made public without their consent, and following their complaint, they, faced isolation and professional exclusion in the company. Such experiences at work are not new for lower caste individuals, but the lawsuit has brought to fore an important, under-examined form of caste relations: networks.
Reports emerging from this case point to a practice of exclusion enabled by upper-caste networks, which dominate high-paying professional circles and where implicit and explicit bias operates as a “virtual noose” around lower caste people. Who you are connected to, and how reciprocal these relationships are, determine how you access resources and opportunities in a community.
These network effects of caste relations are well understood among lower caste communities who face various forms of an embargo as a result of gatekeeping of valuable social capital. The discreet nature of these networks and interactions has led to these important issues being ignored, foregrounded by a burden of proof to “show” socio-legal exclusion based on caste.
The role and performance of networks are critical, especially in positions of power. One setting to observe caste-based fraternising, where individual actors can both act and enact power, is in the parliament. We studied Twitter networks of members of an institution which represents a diversity of caste identities – the Lok Sabha.
Also read: Cisco, Caste Discrimination and the Endurance of Denial in Overseas Indians
In the years since the 2014 general election, social media is increasingly the preferred means of primary communication by major political leaders – a place for them to manage their brands, communicate policy, and perform politics without the mediation of professional journalists. What politicians do online – who they communicate with, follow, or retweet, are integrally tied into the ways they perform and reciprocate relationships with their networks and their electoral base.
To understand caste-based networks online, we undertook a study using information on the castes of members of parliament (MPs) from the SPINPER project and compared it to their following and retweeting behaviour on Twitter. We found that caste has a significant relationship with the centrality, connectivity and engagement of an MP in the Lok Sabha network. The higher the caste of an MP, the more likely they are to be important in the network, the more likely they are to have reciprocal connections with Lok Sabha members, and the more likely they are to get retweeted by a higher caste Lok Sabha member.
Figure 1 shows two groups of MPs, BJP and non-BJP, with measures of importance between MPs of each caste. Those with green cells indicate a significantly higher score of importance for the caste on the row than that on the column. The importance here is determined by three things: how well followed an MP is in the network, how parsimonious they are with their connections, and how well connected one is with other important or well-followed MPs.
The more followed an MP is in the network by important MPs, and the less generous they are with whom they follow, the more important they become. For example, as can be seen in the left figure, upper caste MPs (Brahmins, other upper castes and intermediary castes) in BJP all have significantly higher importance on average than Scheduled Tribes or Scheduled Caste MPs in the network. Intermediary castes are traditionally upper caste but are fighting for a change of status to lower caste for access to reservations, like Jats, Patels, Lingayaths, etc.
Figure 1: Green cells indicate that the median value of caste1’s importance in the Lok Sabha network is higher than caste2.
When we looked at how “following” of MPs from different castes is reciprocated in the network, we found a trend favouring upper caste MPs over lower caste MPs. Figure 2 shows that if a scheduled caste MP from BJP “follows” another MP on Twitter, their chances of being followed back are significantly lower to an upper caste MP regardless of party. Lower caste MPs, particularly scheduled tribe MPs, are less likely to be heard by other MPs in the network. This offers a view into how exclusion can be performed in politicians’ social network, even at the highest levels.
Figure 2: Green cells indicate that the median value of caste1 being followed back by an MP on Twitter is higher than caste2.
We also found that the odds of lower caste MPs getting retweeted by upper caste MPs in the network were low. Figure 3 (right) shows that for Brahmins from non-BJP parties, the odds of being retweeted by other upper caste Lok Sabha members is about 12 times higher than scheduled caste MPs.
Figure 3: Odds of caste1 getting retweeted by an Upper-Caste MP on Twitter vs. caste2. Darker color means better odds.
These trends point to a form of ‘birds of a feather’ scenario where upper caste MPs not only follow each other more on Twitter but that they stay important by engaging each other more. For instance, Brahmin MPs are 3.5 times more likely to be retweeted by an upper caste MP in the network than a Scheduled Tribe MP.
Indeed, the BJP has a well-oiled social media machinery, thus connections and engagements may be better coordinated, and we see these numbers are much worse for other parties where scheduled caste MPs are 12 times less likely to be retweeted by an upper caste MP in comparison to a non-BJP Brahmin MP. One’s odds of getting followed back by another MP in the network as a Brahmin MP is almost always significantly higher than scheduled caste or scheduled tribe MPs, regardless of party affiliation.
These results indicate that equal opportunity of representation for scheduled caste and tribe communities through reservations in Lok Sabha has not ensured their integration with the upper-caste Twitter networks of MPs. Despite the messages of equality under a single “Hindu” banner, the way upper-castes choose to interact with lower-caste MPs on Twitter gives a window into how, in practice, the latter continue to face implicit bias. The significant advantage that Brahmins and other upper-caste MPs seem to have in BJP over almost all the lower caste MPs is indicative of a Brahmanical order within the party that effectively keeps the latter out of the circuits of information and importance in the network.
The findings for non-BJP MPs do not offer encouraging trends either, especially when it comes to how important and central an MP is within the network. While OBC MPs seem to do better in some non-BJP cases, the overall chilling effect on scheduled caste and scheduled tribe MPs is similar to BJP. While more nuanced analysis of these trends is necessary, we find that exclusion of lower caste MPs from upper caste networks of Lok Sabha seems to be a problem not just limited to BJP. Claims of equity-based on religious secularism in other parties, like Indian National Congress, too seem to fail lower caste MPs when it comes to inclusion in practice, albeit a little less than the ruling party BJP.
Also read: Now That Twitter Knows Just How Resilient Casteism Is, It’s Time to Act
Twitter may not be representative of how exactly these interactions take place offline, but it is a window into something that is otherwise difficult to capture. Politicians are “performing” their caste online through the networks they exhibit. While echo chambers of party or political ideologies are well known in the online world, the caste factor remains under-examined in these online interactions. Caste continues to be a consistent fault line in whom upper-caste MPs choose to interact and connect with online.
Caste discrimination can be covert and implicit in the way we interact and choose to engage with others. We tend to think of caste discrimination as casteist slurs or explicit casteist statements that exclude someone, but the biggest transformation of caste as a marker in the 21st century has been the process of invisiblizing it yet keeping it alive through tacit practices of ‘sticking to your kind’. Caste is relational, and looking at both our personal and professional networks can be a way to understand how caste plays a role in them. Representation of an underrepresented group is only the first step in the journey to right historical discrimination, it does not ensure inclusion.
The Lok Sabha is already more representative of caste diversity than private institutions by virtue of reservations. Yet, when we look at how people choose to fraternise within the network, their online coalescence is an uncanny enactment of the same cabalistic behaviour that has upheld caste-based exclusion for centuries. It’s a case of inviting someone to a party because you’re obliged to, but never actually talking to them.
We are grateful to the SPINPER project for the sociological information on LokSabha members that made this study possible.
Palashi Vaghela, Ramaravind K. Mothilal and Joyojeet Pal are researchers on society and technology.