The importance of Kuldeep Sengar and BJP’s social engineering strategy in Uttar Pradesh
BJP’s ability to stitch together identity groups outside of Dalits and Muslims has been crucial to its recent poll successes. In many states, BJP has created a “counter-moblization” against Dalits and Muslims by OBCs and upper castes.
Recent weeks have laid bare the ugly connections between criminality, violence, identity, and politics in India. On April 8, a young woman attempted suicide in front of Uttar Pradesh (UP) chief minister Yogi Adityanath’s house. She alleged that Adityanath was shielding Kuldeep Singh Sengar -- a Member of Legislative Assembly (MLA) from the Bharatiya Janata Party (BJP) -- from prosecution.
MLA Sengar allegedly raped the woman, then a minor, along with others, participating in a gang rape, but the police refused to take action. Instead, the police arrested the young woman’s father, and he was later beaten so severely, allegedly by the MLA’s aides, that he died. Despite public outrage, what explains this sort of response to allegations against MLA Sengar?
It is useful to look a little deeper into this situation to understand the intersection between politics, identity and crime. This is very much an alleged crime of power, and Sengar’s own words leave little doubt about it. When the media questioned him on the allegations on April 9, he dismissively responded, “Come on! They are just low-status people. (Arrey! Wo nimn star ke log hain)” Sengar is from the Thakur community, the same caste group as Adityanath. A CBI investigation has since been launched against the MLA .
If the young woman’s allegation that the chief minister was shielding the MLA is true, why would he want to do so? The MLA is not some committed party member — Sengar only joined the BJP before the 2017 state election after defecting from the Samajwadi Party (SP). In fact, he was elected on a Bahujan Samaj Party (BSP) ticket in 2002, an SP ticket in 2007 and 2012, and the BJP ticket in 2017 and has been an MLA from 3 different constituencies in Unnao district. But it turns out that Sengar is one of Unnao’s most prominent Thakur leaders in a district with a large Brahmin presence.
The ability to stitch together identity groups outside of Dalits and Muslims and a few other prominent caste groups (such as Yadavs) has been crucial to the BJP’s recent electoral successes. In many states, the BJP has effectively used Hindu-Muslim polarisation and caste appeal to create a “counter-moblisation” against Dalits and Muslims by other backward castes (OBCs) and upper castes. Given these electoral realities and Sengar’s importance in the Thakur community, the reasons for BJP’s steadfast support of Sengar become clear.
In order to understand the electoral impact of these strategies, I looked at the 2017 UP state election, which the BJP’s coalition, the National Democratic Alliance (NDA), won in a landslide, winning 325 out 403 seats. The analysis below looks at the strike rates (the percentage of seats won among those contested) for the NDA as a function of the caste and religious composition of an area.
The data on electoral outcomes is provided by the Trivedi Centre for Political Data (TCPD) at Ashoka University, and district-level social demographic data is available from the 2011 Indian Census.
Using this data, I characterise how the strike rates vary for the BJP by district social composition along two factors: the percentage of the population that is neither Dalit nor Muslim (a proxy for the population coalescing behind the BJP; and the percentage of the population that is Muslim (to understand the impact of Hindu-Muslim polarisation).
Chart 1 plots the BJP’s strike rate for the 2012 state election, 2014 general election, and the 2017 state election by percentage of the district population that is neither Dalit nor Muslim. (Note that this analysis makes use of data available at the level of assembly constituency for the national election.)
Naturally, as compared to 2012 (when the overall BJP strike rate was 12%), all of the strike rates are much higher in 2014 and 2017, 84% and 81%, respectively.
Interestingly, there is virtually no difference in strike rate as the percentage of the non-Dalit-Muslim population increases in 2012. However, as the BJP performs better, this cleavage materialises in a serious way, with 2017 displaying a 53% BJP strike rate when the non-Dalit-Muslim population is less than 40% of the population and an 84% BJP strike rate when the non-Dalit-Muslim is greater than 60% of the population.
Chart 2 plots the BJP’s strike rate for the 2012, 2014, and 2017 elections by the percentage of the district population that is from the Muslim community. Intriguingly, the BJP does slightly better in areas with more Muslims in 2012. More importantly, the BJP consistently reaches its highest strike rates in the intermediate category - populations with between 20% and 40% Muslims. Naturally, when the Muslim population gets much higher (>40%), the BJP’s strike rate drops significantly. This rise of BJP’s strike rate in the intermediate category is evidence of the electoral benefits of Hindu-Muslim polarisation. When the population of Muslims is significant, but not too large, it becomes easier to polarise the electorate against this population. As the population of Muslims drops or grows, polarisation is either less effective or infeasible.
The data suggest that there are noticeable electoral benefits (at least as far as UP is concerned) to creating a coalition in a counter-mobilisation against Dalits and Muslims.
At this point, I would be remiss to not mention a second brutal drugging and raping of a young girl in Jammu. Here too, the accused initially found political cover.
Electoral politics is typically based on certain ground-level identity-based calculations, generating a disjuncture between public outrage and the actions of a political party.
The slow response against the alleged perpetrators of a brutal crime — as opposed to a general change in law with regard to rape — reflects these cynical calculations.
The author is Senior fellow, Centre for Policy Research. Views expressed are personal.