DISCUSSIONS surrounding the upcoming state assembly elections in Uttar Pradesh have tended, so far as the Bahujan Samaj Party (BSP) is concerned, to turn around that party’s concerted efforts to mobilize Brahmins, and to build a Dalit-Brahmin-Muslim alliance reminiscent of the Congress in its earlier days. Some commentators have pondered over what sort of impact this latest effort at coalition-building might have on the sectarian loyalties that have, despite various efforts at alliance construction, persisted in Uttar Pradesh over the past two decades. Others have wondered about the implications for the structure of the party system in the state, in so far as the coming together of an alliance such as that the BSP is attempting to forge would spell further trouble for both the Bharatiya Janata Party (BJP) and the Congress, and might help move Uttar Pradesh in the direction of a two-party system, with the Samajwadi Party (SP) as the BSP’s main contender. It may be noted that these debates do not question that the BSP has come a long way from being a Chamar, and indeed, even a ‘bahujan’ party; it is not therefore its transformation from an ‘ethnic’ to a ‘multiethnic’ party that is at issue here,¹ but the exact scope of that transformation, measured in terms of the success it might achieve amongst non-‘bahujan’ communities relative to what it has accomplished in the past.

 

It is this characterization of the BSP as a ‘multiethnic’ party, now common in the literature, that I wish to probe and problematize. ‘Ethnic’ parties are usually differentiated from ‘multi-ethnic’ parties in one of two ways: (i) by taking the distribution of support as the test; or (ii) by instead focusing on the party’s electoral appeals.² The BSP easily qualifies as a ‘multiethnic’ party on both counts. It has received votes not only from lower castes, but also from upper castes and from advanced sections of the ‘Other Backward Classes’ (OBCs) in 1998 and after (1998 being the year in which the party’s nomination of OBCs and upper castes increased significantly, relative to earlier elections). It has moreover, during this period, made explicit appeals to non-‘bahujan’ voters in an attempt to expand its support base beyond Scheduled Castes (SCs), Most Backward Castes (MBCs) and Muslims.

However, while the ‘distribution of electoral support’ tells us what party-mass linkages actually exist, and the ‘electoral appeals made’ tell us what party-mass linkages a party seeks to establish, neither of these indices is able to say much about the quality of existing linkages or the manner in which they are maintained. ‘Mobilization’, or the establishment of party-mass linkages, does not, in other words, guarantee ‘representation’, or the democratic functioning of those linkages;³ nor do indices of mobilization capture differences, if any, in the kind of representation benefits offered by parties to different sections of their constituencies. I argue that the notion of representation needs to be made part of the criteria by which we differentiate ‘ethnic’ parties from ‘multi-ethnic’ ones, and that a ‘multiethnic’ party, strictly speaking, ought to offer ‘symmetric representation’, or the same set of representation benefits, to the different ethnic constituencies comprising its support base.

 

It is this premise of ‘symmetric representation’, I further argue, that the BSP fails to meet. This is so because the BSP, post-1998, has provided patronage benefits to its non-‘bahujan’ voters while providing programmatic benefits to its ‘bahujan’ constituency. (I define ‘patronage benefits’ as those that individual legislators, exercising discretion in the allocation of goods and services at the disposal of the state, secure for individual voters. I define ‘programmatic benefits’, on the other hand, as those that accrue from policy legislation undertaken by a government to an entire constituency.)

Given the proclivity amongst legislators to channel patronage to members of their own castes, the BSP’s nomination of candidates from amongst upper castes and OBCs has deprived its Chamar base of the patronage benefits that Chamar legislators had secured for them in the past. The reason Chamars have suffered more than other ‘bahujan’ communities in this regard are as follows: (i) The BSP has continued nominating MBCs, non-Chamar SCs, and Muslims, while cutting back Chamar nominations specifically to accommodate upper castes and advanced OBCs. (ii) The fielding of upper castes, OBCs (both advanced and backward) as well as Muslims from non-reserved constituencies has made it necessary to accommodate and nominate non-Chamar SCs in reserved constituencies, which has meant that Chamars have fared no better in the reserved than in open constituencies in the event of BSP victory.

Voting for BSP candidates from upper castes and advanced sections of OBCs, apart from entailing the sacrifice of patronage benefits for Chamars generally, have also imposed additional costs on the Chamar poor. The protection accorded by legislators from these landed castes to members of their own communities has made it difficult for Chamar agricultural labour, and BSP cadres, to report instances of abuse and exploitation involving upper caste and OBC farmers. All of which leads to the question: why have Chamars continued voting for the BSP post-1998?

 

 

I argue that BSP governments have provided programmatic benefits to Chamars, the most important of which have been: (i) government jobs flowing from the BSP’s forceful implementation of quotas (it may be noted here that the BSP has, by its own account, as well as those of the bureaucrats I interviewed in Lucknow, accomplished the fullest implementation of quotas ever achieved in the state); (ii) amenities and welfare benefits secured through the Ambedkar Village Programme; (iii) land plots culled out from fallow village commons and distributed to landless SCs and MBCs (though this has hardly been a thoroughgoing effort on account of the limited availability of land, and the difficulty of implementation in the face of resistance from landed castes and sections of the lower bureaucracy); and (iv) psychological empowerment secured through (a) the strict implementation of the ‘SC and ST (Prevention of Atrocities) Act’, and (b) the provision of an administration responsive, relative to other regimes, to the needs of the underprivileged.

 

A caveat, however, is in order, to the effect that the most important of material benefits provided by BSP governments, namely government jobs, have accrued overwhelmingly to the Chamar middle class. This, taken together with the fact that the Ambedkar Village Programme has targeted only selected SC villages, and the effort to distribute land plots has benefited only a minute proportion of the landless poor, has meant that BSP rule has failed, by and large, to materially empower the Chamar labouring class. It is no doubt this same understanding that has led many scholars to posit that the Chamar poor’s support for the BSP is explained by the psychological, rather than material, empowerment secured by BSP governments for this constituency.⁴

I, however, make a different argument here, namely, that it is in good part on account of their belief that the BSP remains committed to their material empowerment over the long term that Chamar labour vote for the party. This belief arises not so much from the actual efforts undertaken by the BSP in the material realm, which have been quite minimal, but from the perception that the failure of those efforts that were undertaken are to be explained partly in terms of the fierce resistance their implementation faced from landed castes and an upper caste bureaucracy, and partly in terms of the efforts made by the BSP’s conservative, upper caste, coalition partners to subvert its social justice agenda. The understanding, in other words, is that it is not for lack of right intention or political will that BSP governments have failed to deliver material benefits to the Chamar poor. And in so far as its coalition partners have been in part responsible for this failure, the BSP may be counted upon to deliver material empowerment as and when it is able to form a majority government in the state.

 

It needs to be emphasized here that while the BSP has made every effort to encourage such perceptions to take root amongst voters, my fieldwork provides strong ground for the argument that these perceptions also arise in an important way from Chamar labour’s own experience of BSP rule⁵ (see footnotes for details). In so far as the Chamar poor’s electoral support for the BSP proceeds from trust, it needs to be remembered that ‘trust is a perception or a cognitive act’,⁶ a ‘bet on the future’ that people place on the basis of their knowledge of the past.⁷ To conceive of trust in this way is to place it squarely within the realm of the rational; my point, in other words, being that to argue for trust-based voting by the Chamar poor is not to argue for a ‘false consciousness’ or ‘elite manipulation’-centric interpretations of political choice.

 

How does all this relate to my argument about ‘asymmetric representation’? I argue that faith in the BSP’s commitment to their long term material empowerment induces committed voting by the Chamar poor in the immediate term, and it is this that explains their readiness to vote for BSP candidates irrespective of caste, and their consequent sacrifice of patronage benefits. And it is this committed voting on the part of its Chamar base that has, most importantly, enabled the BSP to nominate candidates from amongst upper castes and landed OBCs. For the BSP simply cannot capture power without Chamar support, on the basis of the combined votes of its other ethnic constituencies alone.

This is because the vast majority of the non-Chamar sections of the ‘bahujan samaj’, as well as of upper castes and advanced OBCs, vote BSP only on those occasions when it nominates candidates from their own communities, and even on those occasions, do not always vote for the party. It is this special relationship that the BSP has with its Chamar constituency that the use of the term ‘multiethnic’ obfuscates, in so far as it flattens the nuances, and unevenness, of the ties that bind the party to its different ethnic bases.

 

Let us now consider the ties that bind the BSP to its upper caste and advanced OBC voters. I argue that the BSP provides these communities precisely those benefits that have been sacrificed by Chamars, namely, political representation, secured through the election of candidates from their own castes, and patronage benefits and services, delivered by these candidates, if elected, to members of their own communities. It would also be instructive to delve briefly into how upper castes in Uttar Pradesh have arrived at a situation where they need to approach the BSP, or alternatively the SP, rather than the BJP or the Congress, for benefits of the above sort.

While the demise of the Congress in Uttar Pradesh began with the onset of caste-based mobilization and the consequent emergence of a tripartite structure of party competition involving the BJP, SP and BSP in the early 1990s, the more recent decline of the BJP, and its fading popularity amongst upper castes in the state, has had much to do with that party’s alliance with the BSP, and the policies of the coalition governments that have resulted from this alliance.

 

Mayawati, as the BSP chief minister presiding over BSP-BJP coalitions, has undertaken to implement the ‘Prevention of Atrocities Act’ in the strictest possible way, making it difficult for upper caste landowners to abuse SC agricultural labour with the impunity to which they had hitherto been accustomed. The enforcement of the Act has also brought SC farm labour some indirect material benefits, with the immunity from harassment afforded by the Act facilitating strike action over the raising of agricultural wages in parts of eastern Uttar Pradesh.⁸ The BSP government’s crackdown on upper caste criminals, and the consequent imprisonment of Raghuraj Pratap Singh, a Thakur leader of considerable notoriety, has further alienated upper castes from the BJP, which has come to be seen as incapable of protecting the interests of its vote bank, its participation in the governing coalition notwithstanding.

It may be noted here that the above measures have served to antagonize Thakurs rather more than Brahmins, for the reason that there is a much higher proportion of Thakurs than Brahmins amongst large landowners, and hence, amongst employers of agricultural labour. And the alienation of Thakurs from the BJP has, by contributing to the demise of that party, impacted the BSP in a significant way. The movement of Thakurs towards the SP has led Brahmins to increasingly perceive the BJP as a losing party, causing them to shift towards the BSP. However, in view of the fact that the BSP has so far provided only political representation and patronage benefits to its Brahmin constituency, it is the ‘bahujan samaj’ alone that continues to receive programmatic benefits from BSP rule.

 

It is with its non-Chamar ‘bahujan’ constituency that the BSP has the most interesting and challenging relationship. The majority of non-Chamar SCs, MBCs and Muslims vote for the BSP on the lines of non-‘bahujans’, supporting the party only when it nominates their fellow ethnics. The BSP’s treatment of this constituency, however, differs from its treatment of upper castes and landed OBCs, as well as from its treatment of Chamars. The BSP has continued to nominate candidates from amongst non-Chamar ‘bahujans’ post-1998, thereby providing them representation and patronage, while ensuring that they also receive programmatic benefits. Non-Chamar SCs have received the same set of programmatic benefits that have gone to Chamars; MBCs and Muslims, on the other hand, have benefited from the BSP’s efforts to distribute land, and from the psychological empowerment that has accrued to the underprivileged in general from the BSP’s attempts to provide an administration responsive to the poor.

Muslims, it may be noted, had received programmatic benefits from BSP governments in the early 1990s as well. Mayawati had, over the 1994-1995 period, granted Muslims 8% of police officers’ posts, and 8.44% of the 27% quota due to OBCs in the state administration. Muslims have also benefited from the BSP’s refusal to allow the Vishwa Hindu Parishad (VHP) to organize a public event celebrating Krishna’s birthday at Mathura in September 1995, fearing that the VHP would use the opportunity to mobilize Hindus to ‘reconquer’ a Muslim space, a historic mosque being situated at the site in question.⁹

 

It may also be noted that it is in good part the failure of non-Chamar ‘bahujans’ to deliver the BSP a committed vote that has led the party to approach upper castes and landed OBCs.¹⁰ While the possibility that the nomination of non-‘bahujans’ is simply a temporary route to power, that may be revoked once the ‘bahujan samaj’ has come together, may seem far-fetched and unrealistic, the BSP’s decision to appeal to Brahmins rather than Thakurs indicates that the party is unwilling to impose more burdens on its ‘bahujan’ constituency (from which agricultural labourers employed by landed Thakurs are overwhelmingly drawn) than is absolutely necessary. This once again points to the asymmetric relationships that the party has with its different ethnic bases. Not only does the BSP discriminate between its ‘bahujan’ and non-‘bahujan’ constituencies, but the party leadership is also supremely aware of the debt the BSP owes to its Chamar base. It is in this light that we need to interpret Mayawati’s recent announcement that her successor, and the next party leader, would also be a Chamar.

By way of conclusion, I refer briefly to a strand of analysis that argues that ethnic voting driven by patronage benefits leads ethnic voters to support candidates from their own ethnic communities, political party notwithstanding, and that this, by acting as an incentive for parties across the spectrum to nominate ethnic candidates, serves to make the vast majority of parties multiethnic in terms of the ethnic background of legislators, consequently weakening the monopoly position of ethnic parties as vehicles of representation, voice and empowerment for ethnic constituencies. The argument further goes that in so far as this reduces possibilities of radical and intolerant claims-making by ethnic parties, it helps further prospects of ethnic peace and democracy.

 

I argue, on the contrary, that the relationship between patronage-based voting and democracy need not hold in cases where a prime issue sought to be addressed through ethnic mobilization is the deepening of material justice. For patronage-based voting generates no mandate for socioeconomic change, and hence fails to apply pressure on parties to deliver material empowerment to subaltern ethnic constituencies on a systemic basis. This last, I argue, can be accomplished by ethnic constituencies voting for ethnic parties in expectation of programmatic benefits. In so far as Chamar support, or at least Chamar working class support, for the BSP is largely based on expectations of future material benefits of a programmatic kind, it may be concluded that the BSP remains under pressure to deepen its delivery of socio-economic justice in the not- too- distant future. Whether or not the party will prove equal to the task is a question the answer to which will need to take into account a gamut of factors. But that has to be the subject of another paper.

 

* The author thanks Narendra Subramanian for his comments on the article, and Sudha Pai for her response to some early versions of the arguments made here.

The arguments made in this paper are based on ethnographic research conducted in Uttar Pradesh over a period of nine months (December 2003-August 2004). The fieldsites chosen were Meerut and Muzaffarnagar districts in western Uttar Pradesh, and Azamgarh and Jaunpur in the east. Amongst those interviewed were (i) BSP voters from all castes resident in the selected fieldwork villages; (ii) upper castes and OBCs living in these villages; and (iii) BSP cadres and party officials active at the level of the district and below. Interviews were also conducted with BSP and BAMCEF officials, as well as bureaucrats, based in the state capital of Lucknow.

 

Footnotes:

1. For literature arguing for such a transformation, see Kanchan Chandra, ‘Post-Congress Politics in Uttar Pradesh: The Ethnification of the Party System and its Consequences,’ in Paul Wallace and Ramashray Ray (eds.) Indian Politics and the 1998 Election. Sage Publications, New Delhi, 1999, pp. 55-104, and Christophe Jaffrelot, ‘The Bahujan Samaj Party in North India: No Longer Just a Dalit Party?’,Comparative Studies of South Asia, Africa and the Middle East, 18:1, 1998, pp. 35-52.

2. I owe this categorization to Kanchan Chandra. Chandra differentiates between ‘ethnic’ and ‘multiethnic’ parties on the basis of the nature of electoral appeals made, and distinguishes her approach from that of Donald Horowitz, who adopts the distribution of electoral support as the criterion of distinction. See Chandra, Why Ethnic Parties Succeed: Patronage and Ethnic Head Counts in India. Cambridge University Press, Cambridge, 2004, pp. 3-5 (especially footnote 3).

3. For this distinction between ‘mobilization’ and ‘representation’, and the argument that ‘representation’ is the normative dimension of ‘mobilization’, see Samuel H. Barnes, Representation in Italy: Institutionalized Tradition and Electoral Choice. University of Chicago Press, Chicago and London, 1977, pp. 4-5.

4. The argument that lower caste mobilization in North India has been driven mainly by a politics of psychological empowerment has been made by Myron Weiner, Ashutosh Varshney, and Yogendra Yadav. See Weiner, ‘The Struggle for Equality : Caste in Indian Politics’, in Atul Kohli (ed.) The Success of India’s Democracy. Cambridge University Press, Cambridge, 2001, pp. 193-225; Varshney, ‘Is India Becoming More Democratic?’, Journal of Asian Studies, 59: 1, 2000, pp. 3-25; and Yogendra Yadav, ‘Reconfiguration of Indian Politics: State Assembly Elections, 1993-1995’, Economic and Political Weekly, 13-20 January, 1996, pp. 95-104. See also Ashis Nandy, ‘Sustaining the Faith’ and D.L. Sheth, ‘Prospects and Pitfalls’, in India Today, 31 August 1996.

5. An important facet of this experience has been Chamar labour’s witnessing at first hand the attempts made by upper caste and OBC farmers, at times working in tandem with the lower bureaucracy, to (i) block the construction of link roads linking SC villages to main thoroughfares (which figured as an item on the agenda of the Ambedkar Village Programme), and (ii) prevent landless SCs from taking possession of plots that BSP governments had made out in their names. Upper caste farmers were in both cases in danger of being dispossessed of land they had been cultivating illegally, and consequently resorted to means ranging from threats of violence to the actual use thereof, and occasionally in eastern Uttar Pradesh, to the murder of BSP cadres active in the relevant village. While the above served to persuade Chamar labour that the implementation of any agenda of material justice was bound to be a difficult task, the tendency of non-BSP governments to reverse material gains made by ‘bahujans’ during BSP rule served to persuade them that the BSP was the only party that remained committed to their material empowerment, even if it had been unable, on account of circumstances beyond its control, to deliver them any significant material gains so far.

6. Mabel Berezin, ‘Emotions and the Economy’, in Neil J. Smelser and Richard Swedberg (eds.) The Handbook of Economic Sociology. Princeton University Press, Princeton and Oxford, 2005, pp. 109-127, at 111.

7. For this interpretation of trust, see James Coleman, The Foundations of Social Theory. Harvard University Press, Cambridge, 1990.

8. My fieldwork confirms Jen Lerche’s observation that the movement of dalit cadres from the CPI to the BSP in eastern Uttar Pradesh has enabled SC agricultural labour in these parts to take advantage of the BSP’s capture of the state to organize strike action over agricultural wages. See Lerche, ‘Politics of the Poor: Agricultural Labourers and Political Transformations in Uttar Pradesh’, Journal of Peasant Studies, 26: 2/3, 1999, pp. 182-241.

9. See Jaffrelot, op. cit., p. 47.

10. This perspective was voiced by some senior BAMCEF (Backward and Minority Communities Employees Federation) and BSP officials whom I interviewed.