The uneven plebeianisation of Madhya Pradesh politics


back to issue

IN the late 1980s, the very systematic survey edited by F. Frankel and M.S.A. Rao on ‘the decline of a social order’ (that was its sub-title) in India almost eluded Madhya Pradesh – and more or less rightly so because in this state the old ‘social order’ then showed a strong resilience.1 However, things have changed very rapidly since the late 1980s in terms of political mobilization of plebeian groups, including OBCs and Dalits.

According to the 1931 Census, the upper castes represented 12.9% of the state population of post-1956 Madhya Pradesh (5.7% Brahmins, 5.3% Rajputs and 2% Baniyas), the lower castes 42%, the Scheduled Castes 14% and the Scheduled Tribes an exceptionally high 22%. Yet, one must not be content to regard Madhya Pradesh as a whole; one must look at the level of the sub-regions comprising the state. The dominance of the upper castes was especially strong in Vindhya Pradesh (where one could find among the highest concentrations of Brahmins in India, about 14%) and Madhya Bharat (where the proportion of Rajputs is higher than in Rajasthan with about 9%).

Besides the demographic weight of the upper castes, Madhya Pradesh is marked by a fragmentation of the lower castes. With the exception of the Ahirs (or Yadavs), none of the OBCs accounted for more than 5% of the state population. The Kurmis came second with 2.6%. This fragmentation was also in evidence regarding the Scheduled Castes. The Chamars represented 9-10% of the population in Madhya Bharat and Vindhya Pradesh, but less than 5% of the total in the other sub-regions.


In addition to these factors, contrasted with Rajasthan and Uttar Pradesh where Jats form the dominant caste in several districts of the state, in Madhya Pradesh the intermediate (and locally dominant) castes who could have acted as the spearhead of anti-establishment movements are neither prosperous nor large or assertive enough. The Jats, who have always been classified as OBCs in Madhya Pradesh, represent only 0.1% of the population. And the Marathas, who are in equally negligible numbers, often belonged to the elite of the princely states, whose ruling families came from the same milieu, or related to them. They have, therefore, tended to emulate the Rajputs.

The strength of the upper castes does not rely on this demographic advantage alone but, more precisely, on the resilience of the princely elite because of the persistence of old kingdoms during the British Raj. In 1947, they were 35 in Vindhya Pradesh, 25 in Madhya Bharat and a dozen in today’s Chhattisgarh. In Madhya Bharat and Vindhya Pradesh, demographic weight and socio-political domination were concomitant (as mentioned above, the Rajputs represent, for instance, the largest caste in Madhya Bharat with more than 9% of the regional population). Rajas and Maharajas (sometimes with Maratha background in the western part of the state) headed networks of Rajput zamindars and jagirdars. In the post-1956 Madhya Pradesh, it was estimated that 170 out of 296 assembly constituencies and 20 out of 37 parliamentary constituencies were located in part or in totality on the territory of former princely states.2

The socio-political context presented in the previous section partly explains the domination of upper castes over the Legislative Assembly of Madhya Pradesh.


The elitist social profile of the state’s political class is a reflection of the domination of Congress and – more recently – the BJP. Indeed, the Congress till the last decade had over-dominated the political history of Madhya Pradesh. The party ruled pre-2000 Madhya Pradesh continuously since its formation in 1956 except for two short periods in 1977-79 (the Janata phase) and 1990-92 (when the BJP briefly took over). Nevertheless, the steady erosion of the Congress party – and of the socialists – gradually gave birth to a two party system with the Jana Sangh and then the Bharatiya Janata Party (BJP) gaining momentum from the 1960s onwards, without modifying substantially the conservative overtone of the state political scene.


The domination of the Congress party, to begin with, was partly due to its capacity to co-opt former princes, a process well illustrated by the case of the Maharajas of Gwalior, Narsingarh, Rewa, Sarangarh, Surguja and Kanker. Thus, the Congress established its domination over Madhya Pradesh by aggregating ‘vote banks’ owned by former princes or jagirdars. This ‘feudal’ brand of clientelism partly explains the over-representation of the upper castes among the Congress MLAs.



Castes and Communities of the Madhya Pradesh MLAs, 1957-2003 (in %)

Castes and communities












Upper castes












Intermediary castes




























































Other minorities
















































From 1957 to 1967, the proportion of the Congress upper caste MLAs remained between 40-51%. The most important non-upper caste groups were the Dalits and the Tribals. But these MLAs, who were often uneducated, did not form powerful lobbies. In fact, they had been co-opted because of the reservation system and were part of a ‘coalition of extremes’ pattern, to use the phrase Paul Brass has coined to describe the situation prevailing in Uttar Pradesh.3 According to him the coalition supporting the Congress could be qualified in such terms because its constitutive groups were poles apart in the social structure. However, this terminology is somewhat misleading since it suggests that the groups in question might possess the same influence, while in fact, the Dalits in Congress depended on the upper caste leaders who were really in command – another indication that the Congress system essentially remains a clientelistic arrangement.


In 1969, the scission of the Congress prepared the ground for a major shift in the social profile of the ruling party: while the Congress (O – ‘O’ for ‘Organisation’ or for ‘Old’, as its detractors called it) represented the traditional notables, and Indira Gandhi’s Congress (R – ‘R’ for ‘Requisitionist’) pretended to epitomise a new, more people-oriented political programme. Indira Gandhi broke with the traditional Congress style during the 1971 mid-term election campaign. She tried to short-circuit the local notables at the helm of ‘vote banks’ by presenting her socio-economic programme directly to the people.

However, the 1972 elections showed that the Congress (R) did not break from the age-old collaboration between an upper-caste educated intelligentsia and the notables from the merchant and agricultural classes.4 The caste profile of the Congress (R) MLAs was even more upper caste-dominated than those of the 1967 Congress group with 51% upper caste MLAs and less than 10% OBC. The Congress (R) even relied on the princes during the 1972 elections. Of the 37 princes who contested in 1972 nation-wide, 24 stood on the Congress (R) ticket – and of them 17 were allotted Congress (R) tickets in MP.5

The elitist bias of the Congress was gradually put into question after the party returned to power in 1980. Already in 1977, while facing its worst defeat since independence, the party had a substantial number of OBC MLAs. But the decline of the upper caste MLAs was not evident before 1980 when their share dropped below 39%. Arjun Singh, the chief minister in 1980-1985, arguably tried to project himself as the spokesperson of the plebeians. At that time, the OBCs of Madhya Pradesh did not have a political organization of their own (the socialist movement, the main spokesmen for the OBCs in states like Bihar and Uttar Pradesh, was very much divided and dominated by upper caste leaders). The proponents of kisan politics were also very weak, largely because of an absence of a strong middle caste milieu, like the Jats in Uttar Pradesh.


There was, therefore, room for manœuvre for the Congress. In 1981, Arjun Singh appointed a commission named after its chairman, Ramji Mahajan, a former state minister and himself a Mali (OBC), to establish a list of the OBCs and to identify their needs in the state. The Mahajan Commission report was submitted in late 1983. It identified 80 OBC castes which represented 48.08% of the state’s population (including 2.08% Muslims). The commission’s report recommended that 35% seats be reserved for OBCs seeking admission in educational institutions6 and in all governmental, semi-governmental and public sector services. However, Arjun Singh did not really promote low caste leaders as assembly candidates; hence the low level of the Congress OBC MLAs which remained around 14-15% over the 1980s.


Things really changed in the 1990s, with the decreasing number of upper castes MLAs and the rise of the OBCs, except during the 1990 election, the worse election year for the Congress since the 1977 post-emergency rout, which suggests that its upper caste candidates were better entrenched in their constituency. The most striking evolution, though, was the sudden increase in the number of Congress OBC MLAs in the 1993 election and their subsequent relative stability. The 1993 shift of about seven percentage points compared to the 1980s was unprecedented. The percentage of the Congress OBC MLAs rose from about 14-15% in 1980-1990 to about 19-23% in 1993-2003. The largest group was made of the Kurmis who represented 7.6% of the Congress (I) MLAs in 1998, whereas they formed 3.7% of the state population according to the Mahajan Commission Report.

Another interesting figure concerns the Scheduled Tribes which represented the second largest group among the Congress (I) MLAs after the upper castes, with 26-30% of the MLAs in 1993-1998, whereas only 23.4% of the seats were reserved for them. As far as the upper castes are concerned, the weight of the Brahmins tended to decline whereas that of the Rajputs increased, probably because of the attractiveness of Digvijay Singh, the new Congress leader and chief minister over his castemen. One can, at last, see the Mandal impact in this social transformation. Indeed, these figures reflect the responsiveness of the party – and of its president, Digvijay Singh, whose mentor, at that time was Arjun Singh – to the Mandal phenomenon as evident from the social profile of the candidates nominated by the Congress in 1993-2003 since the share of the upper castes dropped from 38.1 to 34.5% whereas that of OBCs jumped from 17.8 to 23.9 between 1993 and 2003.


However, the social profile of the Congress governments is rather different from that of the party’s MLAs. Certainly, the share of the upper castes among the ministers tended to decline, but this change occurred very unevenly. For instance, while the 1985 elections sent a smaller number of upper castes candidates and a larger number of OBCs to the assembly, upper caste people massively dominated the 1985 Congress government. But here again things changed in the 1990s.

In fact the transformation of the caste profile of the Congress was especially pronounced in the governments formed by Digvijay Singh. In 1993 his cabinet was much less upper caste-dominated than the Congress governments of the 1980s. While the upper caste members represented 45-57% of the governments in 1980-1990, in 1993 their percentage fell to about 33% and remained below 40% in the 1990s (except in 1998). The Rajputs, once again, ‘resisted’ more effectively than the other upper castes, especially the Brahmins and the Baniyas who had been especially well represented in the BJP government of Sunderlal Patwa in 1990-1992: another indication that the Congress tended to be associated with the Rajputs while the BJP remained a ‘Baniya/Brahmin’ party.

In 1993-1999, the main beneficiaries of the decline of the upper castes were not primarily the OBCs (who represented about one fifth of the ministers, with the Kurmis still much more numerous than any other caste group) or the Scheduled Castes (who remained largely under-represented), but the Scheduled Tribes, whose representation jumped to 25-26%. These trends suggest, once again, that the traditional pattern of ‘coalition of extremes’ survives while being reshaped. An interesting variant of this pattern is emerging since the components of these vertical arrangements are not so much the Brahmins and the Scheduled Castes – as in the case of Uttar Pradesh – but the Rajputs and the Tribals.

In addition to diluting the domination of the upper castes over the state government, Digvijay Singh tried to promote non-elite groups within the machinery of the Congress party. However, upper caste Congress leaders retained their control over the party machinery.


In the Pradesh Congress Committee appointed in 1993, not only was the president, Digvijay Singh a Rajput, but the upper castes represented 55% of the members. Once again, the OBCs formed a small minority (15.5%) whereas the Scheduled Castes were almost as numerous with about 14%.

To sum up, the Congress in Madhya Pradesh has been slower than most of the large state units of the Hindi belt to become more plebeian – except Rajasthan. Even though Digvijay Singh’s strategy of ‘social engineering’ was designed to adjust to the post-Mandal context, it remained biased in favour of the Scheduled Castes and, even more, the Scheduled Tribes, initiating a new variant in the old pattern of ‘coalition of extremes’. In this framework the Rajputs tended to replace the Brahmins and the Scheduled Tribes the Scheduled Castes but the rationale of this arrangement was based on the logic of clientelism.

Digvijay Singh candidly admitted that his strategy of promoting tribal leaders had partly failed because none of them displayed statesman-like qualities.7 After the creation of Chhattisgarh, this strategy was put into question because of the decreasing influence of the tribals whose share had gone down from 23.27% according to the 1991 Census to less than 20% according to the 2001 Census. Digvijay Singh then promoted a ‘Dalit Agenda’ that was intended to counter the BSP which he considered to be more dangerous than the BJP.8 But this strategy failed as I have showed elsewhere in detail.9


The Jana Sangh was known in Madhya Pradesh as a ‘Baniya/Brahmin’ party, not only because its electoral strongholds comprised the urban upper caste middle class,10 but also because the party was over-dominated by leaders coming from this social milieu.11 Indeed, an overwhelming majority of the party MLAs had this social background. To begin with, the BJP was no different from its predecessor. However, the share of the upper caste MLAs – mostly Brahmins and Baniyas – underwent a steady, though limited, erosion over the years 1993-2003, from 42.3% to 35.8%. This change benefited the OBCs – who crossed the 20% threshold in 1998 and remained at this level in 2003, and to the STs, who jumped from 15.3 to 22.8%.


While this last development was not due to the fielding of a larger number of ST candidates, the party displayed a more voluntarist policy vis-à-vis the OBCs. It gave a ticket to 23.1% OBCs in 2003 – almost four percentage points more than in 1993. And after its electoral success – unprecedented in the annals of the state – the BJP MLAs voted to power an OBC chief minister, Uma Bharti (a Lodh). Though she did not stay long in office, she was succeeded by two other OBC chief ministers, Babulal Gaur (a Yadav) and Shivraj Singh Chauhan (a Kirar).

While the BJP has nominated an increasing number of non-upper caste candidates at the time of elections in the wake of the Mandal affair to benefit from the mobilization of the lower castes, the composition of Patwa’s government between 1990 and 1992 was a good indication of the persisting domination of the upper castes within the BJP in the early 1990s. As shown above, out of 31 cabinet members, 17 belonged to the upper castes, seven to the OBCs, three to the Scheduled Tribes, two to the Scheduled Castes – there was also one Muslim. Similarly, in 2003, when the BJP came back to power, the share of the upper castes in the government jumped from 39.5% to 61.2%: the chief minister may be from the OBCs, but his team was not at all plebeian. Things changed in 2004 when the proportion of the upper caste ministers fell to 48.6% – still about ten percentage points more than in the Digvijay Singh governments of the 1990s.

The BJP’s party machinery also remains over-dominated by upper caste dignitaries, with a significant over representation of the Brahmins and the Baniyas. Between 1991 and 2003, the share of the upper castes in the BJP Madhya Pradesh state executive committee oscillated between 57% and 67%, whereas the OBCs were never more than about 20%, the SCs never more than about 5%, and the STs never more than about 14%.

The over-representation of the upper castes among the BJP cadres stems from the conjunction of two factors. First, the party apparatus comprises many former pracharaks who have been seconded by the RSS to the BJP for organizational tasks; and most of the RSS pracharaks, till recently, came from the upper castes, particularly from the Brahmin jatis, often with a Maharashtrian background.


Second, the traditional elitist profile continues because the party establishment has not undergone any significant renewal in the last decades, which means that the traditional base of the Hindu nationalist movement among the Brahmins and the Baniyas remains over represented. Though the BJP tries to project a plebeian image by promoting OBC chief ministers and by appointing SC leaders within the party apparatus – today, for instance, the president of the state executive committee, Satyanarayan Jatia, is a Dalit – upper caste leaders remain very much in control, more than what an erosion of their share among the party MLAs may suggest.

To sum up, the two mainstream parties of Madhya Pradesh, the Congress and the BJP are still in the hands of upper caste politicians, even though they have pretended to plebeianise themselves, the former by revisiting the old coalition of extremes pattern, and the latter by promoting OBC leaders. But these two parties do not represent the same segments of the elite. While the BJP is still, to a large extent, a ‘Baniya/Brahmin’ party, the Congress, in the 1990s, has relied more on the Rajputs – something which has changed after Digvijay Singh left the scene.


In contrast to Uttar Pradesh and Bihar, Madhya Pradesh has no solid tradition of socialist politics. But the SP, the PSP and the SSP made some inroads in the state after independence, so much so that their parties became the second largest force in the Vidhan Sabha in 1962. Interestingly, the BSP has of late primarily developed pockets of influence in the districts of Vindhya Pradesh and Madhya Bharat bordering Uttar Pradesh where the socialists used to be strong.

The BSP of Madhya Pradesh has tried to broaden its base beyond the Scheduled Castes in a systematic manner. First, the party has nominated an increasing number of OBC candidates, from 23.7 in 1993 to 50.5% in 1998. In 1998, the BSP OBC candidates ‘represented’ 27 different castes. Among them, the Kurmis stood prominent with 12.4% of the party candidates – the largest group. The decision to nominate more than half OBC candidates was not made at the expense of the Scheduled Castes (whose percentage was stable around 30% with the Chamars and the Satnamis remaining the largest groups), but of the Muslims and the Scheduled Tribes. Acknowledging their strength in a state where they represented one fourth of the population, the BSP had nominated almost 26% tribal candidates in 1993 but reduced this percentage to 16% in 1998.

Even though the BSP registered a setback during the 1998 election – its percentage of valid votes dropping from about 8% in 1993 to 6.3% – the party’s strategy of attracting OBC voters, beyond its Dalit base, as is evident from its nomination of OBC candidates in large numbers, might have posed a threat to the mainstream parties and obliged them to further ‘democratize’ themselves. However, the BSP changed its strategy in 2003 when it gave tickets to a larger number of upper caste candidates at the expense of the OBCs in order that its list of candidates should reflect the general social structure. Out of 146 candidates, 17.8% were from the upper castes, 34.2% from the OBCs, 34.9% from the Scheduled Castes, 19.2% from the Scheduled Tribes and 7.5% from the Muslim community.12 This strategy bore fruit since the party polled 7.6% of the valid votes – interestingly, its two MLAs were both returned in non-reserved constituencies.

In contrast to other states of the Hindi belt where non-upper castes have been in a position to take the lead against the savarna – like the Jats in UP and Rajasthan or the Yadavs in UP and Bihar – Madhya Pradesh politics has been characterized for decades by the domination of upper castes. Things began to change in the late 1970s after the victory of the Janata Party and more evidently during the following decades after the Mandal affair when the rise of the OBC MLAs was especially obvious on the BJP side and that of the STs on the Congress side. However, both parties have remained content to give tickets to plebeian groups at the time of elections, even as upper castes leaders retain power within the state governments and the party apparatuses. They may now be forced to further democratize themselves under pressure from the BSP and also, possibly, from the Samajwadi Party, both of which are importing a different political culture from the North.



1. F. Frankel and M.S.A. Rao (eds), Dominance and State Power in Modern India, Oxford University Press, Delhi, 1989, p. 422.

2. For more details, see C. Jaffrelot, The Hindu Nationalist Movement and Indian Politics, Penguin India, New Delhi, 1999, p. 215.

3. Paul Brass, ‘Politicization of the Peasantry in a North Indian State’, Journal of Peasant Studies 8(1), October 1980, pp. 3-36.

4. See what happened in Haryana, for instance P. Singh, ‘Haryana State Assembly Polls of 1968 and 1972’, Indian Journal of Political Science 7, April-September 1973, pp. 143-164.

5. Iqbal Narain and M.L. Sharma, ‘The Fifth State Assembly Elections in India’, Asian Survey 13(3), March 1973, p. 325.

6. See Ramji Mahajan, Madhya Pradesh Rajya Picchra Varg Ayog – Antim Prativedan. Antim Prativedan – Bhag ek, Bhopal, December 1983, chapter 15 (Hindi).

7. Interview with Digvijay Singh, November 1997, Bhopal. According to Jamuna Devi, one of the tribal leaders sidelined by Singh, his pro-adivasi move had failed because he had not promoted the right persons, lest they took his place (Interview with Jamuna Devi, 20 February 2004, Bhopal).

8. Interview with Digvijay Singh, March 2007, New Delhi.

9. C. Jaffrelot, ‘The Uneven Rise of the Lower Castes in Madhya Pradesh’, in C. Jaffrelot and S. Kumar (eds), Rise of the Plebeians? Routledge, New Delhi, 2009 (forthcoming).

10. See B.D. Graham, Hindu Nationalism and Indian Politics: The Origins and Development of the Bharatiya Jana Sangh, Cambridge University Press, Cambridge, 1990, chapter 6.

11. see C. Jaffrelot, The Hindu Nationalist Movement and Indian Politics, op. cit., pp. 172-178.

12. Central Chronicle, 7 November and 6 December 2003.