Decoding data: Varying degrees of support across class and caste

The Muslim vote, making up about 16 per cent of Bihar’s electorate, has also tended to be segmented on the basis of class.

An examination of Lokniti-CSDS survey data for Bihar elections from 2000 onwards indicates that class can be a significant factor in voting behavior within various caste groups. That is, upper class members of the same caste may vote differently than lower class members of the same caste. However, this interaction between class and caste is stronger among some groups than others.

Where class matters most: Lower OBCs and Muslims

Lower Other Backward Classes (OBCs) or Extremely Backward Classes (EBCs), making up close to a quarter (24 per cent) of Bihar’s electorate, are a heterogeneous group whose vote has not firmly aligned behind any party. Results from the past three Assembly elections and the most recent Lok Sabha election show that there are significant differences in voting patterns between upper class lower OBCs and lower class lower OBCs. In the two most recent Assembly elections, the BJP performed more strongly among upper class members of the lower OBCs.

It received 39 per cent of the upper class votes among the community as compared to 18 per cent of lower class votes in October 2005, and 42 per cent of upper class votes as compared to 21 per cent of lower class votes in 2010. The BJP snagged a similar shares of the EBC vote across class in the 2014 Lok Sabha elections, but JD(U) notably performed more strongly among the upper class (26 per cent) than the lower class (14 per cent) of EBCs. Undoubtedly, the BJP will be contesting strongly for the EBC vote. The question is whether they will be able to consolidate it across class, as they did in 2014, or whether their support within EBCs will come more from its upper classes, as in 2005 and 2010.

chart.jpg759The Muslim vote, making up about 16 per cent of Bihar’s electorate, has also tended to be segmented on the basis of class. But this segmentation has not been as consistent as that of the upper castes. Lalu’s RJD — for whom Muslims have been a critical base of support — performed more strongly among lower class Muslims (53 per cent) than upper class Muslims (38 per cent) in 2000, but observed the opposite trend in 2010 (32 per cent among upper class Muslims and 23 per cent among lower class Muslims). Congress mirrored these trends, winning a larger bloc of upper class (27 per cent) than lower class (11 per cent) Muslim voters in 2005, but did better among lower class (16 per cent) than upper class Muslim voters (7 per cent) in 2014. Class then appears to matter among Muslim votes, though its relationship with the RJD and Congress share of vote has not been consistent.

Where class sometimes matters: Yadavs, Kurmis/Koeris

Class has sometimes been a significant variable in the share of the Yadav vote, making up 15 per cent of Bihar’s electorate. As with the Muslim vote — the other cornerstone of Lalu’s base — the RJD performed better among the lower class (63 per cent) than the upper class (54 per cent) in 2005. But Lalu performed better among upper class Yadavs in 2014, winning 56 per cent of their vote compared to 42 per cent of lower class Yadavs. In the 2000 and 2010 Assembly elections, Lalu won roughly an equal share of Yadav votes across class. It is therefore difficult to predict whether the Yadav vote will be fragmented on the basis of class in the upcoming Bihar polls.

The Kurmi/Koeri vote, making up 11 per cent of Bihar’s electorate, too has been segmented on the basis of class in some elections and not in others. In the 2000 Assembly elections, the JD(U) took more than one in four (27 per cent) upper class Kurmi/Koeri votes while receiving just a nominal share (3 per cent) of lower class votes.

Conversely, the RJD won 25 per cent of lower class Kurmi/Koeri votes but only 12 per cent of upper class votes. The JD(U) saw the opposite trend in 2010, when it did better among lower class Kurmi/Koeris (41 per cent) than upper class members of the same castes (21 per cent). In both 2005 and 2014, however, JD(U) won a similar share of Kurmi/Koeri votes among upper and lower class voters. As with the Yadavs, there is not a consistent enough pattern here to speculate about whether the Kurmi/Koeri vote will split along class lines.

Where class matters the least: Upper castes

As Lokniti colleagues Nitin Mehta and Pranav Gupta noted in a previous article in this series, the BJP enjoyed even greater consolidation of the upper castes, which makes up 15 per cent of Bihar’s electorate, among lower class members (71 per cent) than upper class members (53 per cent) in the 2014 Lok Sabha elections. However, this gap is the exception that reveals the norm. In the 2000, October 2005, and 2010 Bihar Assembly elections, the upper caste vote was not segmented on the basis of class.

These findings indicate that class is likely to matter among some caste groups only. This reflects the reality that caste predetermines class in much of Bihar. However, previous elections illustrate that caste vote can divide along class, and the above analysis suggest that the lower OBC and Muslim vote are likely to differ along class lines; the addition of Asaduddin Owaisi’s Majlis-e-Ittehadul Muslimeen to the electoral mix in Bihar may also scramble traditional voting patterns among Muslim voters. With opinion pre-polls showing a tight race, how the NDA performs among lower class OBCs or how the Muslim vote splits could prove decisive factors in the battle for Bihar.

The writer is a Fulbright-Nehru student researcher working with Lokniti-CSDS