Pulling off its third successive win in the Delhi assembly election, the Congress demonstrated that public dissatisfaction with its Sheila Dikshit-led government was not as overwhelming as supposed. The Bharatiya Janata Party did gain three more seats and more of the popular votes but it did not have enough in its armoury to upset the ruling party. The main gainer in the election was the Bahujan Samaj Party, which won two seats and attracted a large chunk of the traditional support base of the Congress and the BJP.
STATE ELECTIONS 2007-08Economic & Political Weekly EPW february 7, 200927Delhi Assembly Elections: 2008 Sanjay KumarPulling off its third successive win in the Delhi assembly election, the Congress demonstrated that public dissatisfaction with its Sheila Dikshit-led government was not as overwhelming as supposed. The Bharatiya Janata Party did gain three more seats and more of the popular votes but it did not have enough in its armoury to upset the ruling party. The main gainer in the election was the Bahujan Samaj Party, which won two seats and attracted a large chunk of the traditional support base of the Congress and the BJP. Delhi was the biggest “upset” in the latest round of assembly elections. The Congress in Delhi joined a small list of ruling parties or combines – the Left Front in West Bengal and Tripura, the Rashtriya Janata Dal (RJD) in Bihar, the Congress in Arunachal Pradesh, the Bharatiya Janata Party (BJP) in Gujarat and the Sikkim Democratic Front (SDF) in Sikkim – that have retained power three successive times or more. The victory was seenasa personal triumph for Chief Minister Sheila Dikshit who led the party to victory the second time around, a surprising achieve-ment in an era of “anti-incumbency”.The Congress secured a comfortable majority of 43 seats in the 70-member assembly. It won three seats less than in the last elec-tion and its lead in terms of popular votes came down to a mere 4 percentage points. Though theBJP managed to increase its tally of seats from 20 to 23 and wonmore votes in this election, its share of votesstill fell short of the majority mark. The surprise at the Congress victory was not without basis. Even before the elec-tions were held, Delhi was being counted as the one state where the BJP was sure to win. There was a popular perception of dissatisfaction with the ruling Congress Party in the city, a perception strengthened by the BJP’s victory over it in the Municipal Corporation of Delhi (MCD) polls in 2007.1 The confusion over sealing businesses in residential localities annoyed not just the business class but also the common people.2 Just when the Congress appeared to have negotiated the tide of popular anger with the Supreme Court allowing all commercial property to be reopened, the price rise brought fresh trouble. There was a sense that the Congress managedtostage a recovery after the BJP declaredVijay Kumar Malhotra to be its choiceforthe chief minister because popu-laropinion had it that he was no match for Dikshit. But the Mumbai terror attack, that took place less than 72 hours before the polling in Delhi, and the subsequent tide of popular anger against the central government again seemed to dim the pros-pects of the ruling party in the capital. What appeared to be turning theBJP’s advantage into near certain victory was the Bahujan Samaj Party (BSP) factor. After its victory in Uttar Pradesh, the BSP has been seriously thinking of expanding its base in other states. That it was taking the Delhi assembly elections seriously was clear from the very beginning when the party declared its candidates well in ad-vance. BSP supremo Mayawati addressed several rallies and campaigned extensively in the capital. The results revealed that, unlike in Madhya Pradesh and Rajasthan, theBSP’s efforts in Delhi succeeded: though it managed to win only two assembly seats, its share of the vote increased dramatically from 5.8% in 2003 to 14% this time. This paper examines these popular im-pressions in the light of the outcome of the election and evidence gathered by a post-poll survey carried out by the Centre for the Study of Developing Societies (CSDS), New Delhi.3 It argues that theBSP was a major factor that worked against the Congress. Also that the Congress made up for this loss by an overall posi-tive rating of its governance and the im-age gap between Dikshit and Malhotra. Popular opinion underestimated the im-pact of narrowing class divisions that helped the Congress. The party did suffer losses among the poor, but it maintained an upper hand over other parties even in this group. Its gains among the middle class and well to do voters proved deci-sive in making up for some of its losses. In ethnic and community terms too, the Congress managed to align itself with the changing social demography of Delhi much better than theBJP. At the same time, this achievement of the Congress is contingent, as the rise of the BSP opens the possibility that the structure of politi-cal competition is changing.Sanjay Kumar (sanjay@csds.in) is a Fellow at the Centre for the Study of Developing Societies in New Delhi. Table 1: Performance of Political Parties – Assembly Election 2008 VotesGain/LossSeatsGain/Loss Polled from 2003 Won from (%) (%) 2003Congress 40.3-7.843-4BJP 36.3+1.123+3BSP 14.0 +8.32 +2Independents 3.9 -0.9 1 0 Other smaller parties 5.4 -0.6 1 -1
STATE ELECTIONS 2007-08february 7, 2009 EPW Economic & Political Weekly28Before turning to an explanation, let us first take a close look at the verdict itself (Table 1, p 27). Though the votes of the Congress went down considerably, and the party ended up with a lesser number of legislators, it still managed to register an impressive victory, winning 43 of the 70 seats and 40.3% of the votes. Though the BJP managed to increase its tally from 20 to 23 and polled more votes in this election than the previous election, its share of votes remained far short of the majority mark. The BSP managed to win only two seats, but its share of the vote increased dramatically from 5.8% in 2003 to 14% in 2008. Independents and candidates from other smaller parties together polled less than 10% of the votes. BSP Gains TheBSP’s votes came mainly from constit-uencies dominated by migrants from Ut-tar Pradesh and Bihar. As per a rough clas-sification, there are 16 assembly constitu-encies where these migrants play a deci-sive role and many among them seem to have voted for the BSP.4 Most of these con-stituencies are located on the periphery of the city5 and in them, the BSP won 19% of the total votes, 5 percentage points more than its average vote in the city. Both the constituencies where theBSP won had large migrant populations. The BSP polled 7.5% of the votes in constituencies where there were migrants in sizeable numbers even during the 2003 assembly elections. But its support base increased enormously this time around, to the detriment of the Congress. In constituencies dominated by the migrants, the Congress polled only 36% of the votes, 4% less than its average vote in this election (Tables 2 and 3).While Delhi’s politics is largely discussed in terms of the Jat vote and the Punjabi vote, nearly 17% of the capital’s voters be-long to dalit communities. Among the dal-its, the Jatavs are the largest caste and they voted for the BSP in large numbers. The Jatav vote for the BSP was not as overwhelm-ing as in Uttar Pradesh, but still 37% of them voted for it during this election. There was hardly any shift in the Jatav support for the BSP in the last five years. But the same period saw a shift of non-Jatav dalit voters to the BSP’s side. During this election, 22% of the non-Jatav dalit voters voted for the BSP, compared to only 8% in the previous assembly elections.Dur-ing the last five years, the Congress lost substantial support among these voters. The BJP was also affected to some extent. The BSP managed to increase its vote share and harm both the Congress and the BJP. Among those who voted for the BSP in this election, 55% were those who had voted for it in the previous elections but it man-aged to snatch 28% of its votes from the Congress and 12% from the BJP. But the success of the BSP did not depend only on the dalit vote. The party’s popularity increased among voters of other communities as well, thanks to its strategy of distributing tickets to non-dalits. Dur-ing this election, 17% of the voters from the Gujjar and Yadav communities voted for the BSP compared to only 4% during the previous assembly elections (Table 5, p 29). The party also increased its support among voters belonging to other back-ward castes (OBC). Among the OBC, 18% voted for the BSP in thiselection compared to only 5% in the lastone.TheBJP lost sup-portamongthe Gujjars and Yadavs. There was also a noticeable shift among Muslim voters towards the BSP. We do not know how enduring the non-dalit shift towards the BSP will be, but it is clear that the party’s rise has adversely affected both the Congress and the BJP and dented the bi-party nature of political competition in Delhi. How Did the Congress Win?In spite of losing support, the Congress managed to win 43 seats because it regis-tered smaller victories compared to its vic-tories in the previous assembly elections. The average victory margin in this election was 10,158 votes, smaller compared tothe the average victory margin of 10,946 votes in 2003. But if we compare the average vic-tory margins of the Congress in the last two assembly elections, we find that in 2008 it was 9,748 votes, much less than the 12,784 votes of 2003. On the other hand, in spite of increasing its vote share, the BJP man-aged to increase its tally by only three seats because the party registered bigger victo-ries compared tothe Congress. During the 2003 elections,the average victory margin of the BJP was 7,007 votes, which went up to 11,151 votes in 2008. Of six seats decided by a margin of less than 500 votes, the Congress won four and the BJP two. In this sense, the Congress was simply luckier. If it had lost more marginal seats, its majority may have been thinner. Luck, however, cannot explain the in-cumbent party’s return to power. Findings from surveys conducted during previous elections indicated a sharp class polarisa-tion in Delhi. The Congress was extremely popular among voters in the poor and low-er-income groups, those living in jhuggi-jhopari and lower income class locations (Table 4, p 29). During the 2003 assembly election, the Congress got 54% and 53% votes among the very poor and poor vot-ers. Compared to this, theBJP got only 18% and 24% votes among the very poor and poor. The class division among Delhi vot-ers was sharp enough to even influence the voting choices of voters from different classes within the same caste or communi-ty. One saw the divide in voting preferenc-es between the rich and the poor even among the Brahmins or the Punjabis, the traditional supporters of the BJP in Delhi. This was also seen among voters from the Vaishya community. The poor tended to Table 2: Region-wise Analysis of the Results – Delhi Assembly 2008 Seats Turn-Congress BJP BSP Others out(%) Seats Vote(%) Seats Vote(%) Seats Vote(%) Seats Vote(%) Regions Trans-Yamuna 16 59.0 10 42.2 5 35.6 1 15.5 0 6.7 Periphery 16 55.7 8 33.5 6 30.9 1 20.0 1 15.6 North-west 1759.6 9 41.18 40.80 11.0 0 7.1 Citynorth 11 59.1 8 44.1 2 39.5 0 8.6 1 7.8 Citysouth 10 53.7 8 44.6 2 36.4 0 11.4 0 7.5 Total 70 57.6 43 40.3 23 36.3 2 14.0 2 9.3 Source: CSDS Data Unit.Table 3: Performance of Political Parties in Different Types of ConstituenciesCommunity Type Seats Turnout(%) Congress BJP BSP Others Won Vote(%) Won Vote(%) Won Vote(%) Won Vote(%)UP-Bihar migrants 16 58.0 8 36.1 6 33.9 2 18.9 0 11.0Punjabi dominated 16 58.0 10 43.3 6 41.1 0 8.6 0 7.0Peasant community 11 55.8 7 34.7 3 31.8 0 17.7 1 15.7Rest 27 58.11843.9 8 37.20 12.41 6.4Total 7057.64340.32336.3214.029.3Source: CSDS Data Unit.
STATE ELECTIONS 2007-08Economic & Political Weekly EPW february 7, 200929vote in favour of the Congress while the rich voted for the BJP (Kumar 2004). This time, the Congress suffered a set-back in its traditional support base of poor and the lower-class voters. The BSP suc-cessfully managed to make inroads among them. Since the poor and the lower-class people can hardly afford to live in the city, they tend to settle down on its outskirts. They either live in jhuggis, small huts, semi-pucca houses or in Janata flats. Many assemblyconstituencies on the periphery of the city have a sizeable migrantpopula-tion who generally belong to the poor or the lower-income class. In the 16 assembly constituencies located on the periphery of the city, the Congress suffered heavy loss-es. The party polled nearly 7% less in these constituencies compared to its average vote share.6 TheBSP was the major gainer in these constituencies. However, despite losing some of its huge lead among the working classes, the Congress managed to still finish ahead of the BJP in this section.The Congress managed to make up for the loss of its support in poor constituencies with a reasonably better performance in constituencies where there were sizeable numbers of upper class voters. These con-stituencies are mainly located at the two ends of the city, in the south and in the north. In these constituencies, the Congress did better compared to the rest of constitu-encies. In the nine assembly constituencies of South Delhi, with a large number of up-per middle class and rich voters, it polled 45% votes and in the 11 assembly constitu-encies of the North Delhi, it polled 44%. Though the Congress lost about 8% of the votes compared to the 2003 assembly elections, it still remained an umbrella party, getting support from voters across castes and communities. Though the BJP remained more popular among upper caste voters, especially those from the Punjabi Khatri, Vaishya and other commu-nities, the Congress managed to win votes even among them. The anger of the Sikhs against the Congress also seems to have come down. Though theBJP led among Sikh voters, 41% of them voted for the Congress. A majority of the Jat voters vot-ed for the BJP, but still, a sizeable number of them voted for the Congress. The unhappiness of Saj-jan Kumar, the Jat leader whose son was denied a party ticket, did damage the Congress, but it still did well and managed to get numerous votes from among the Jats. A sizeable number of voters from the Gujjar-Ya-dav communities also voted for the Con-gress. Though there was some shift among Muslim voters towards the BSP, the Con-gress remained the first choice of Muslim voters in Delhi. Its biggest worry should be the shift among dalit voters who have moved away to the BSP in sizeable num-bers during the last five years. Why Did the BJP Fail?The choice of Malhotra as the candidate for the chief minister affected the chances of the BJP. When asked who they would like to see as their next chief minister, 32% preferred Dikshit while only 13% men-tioned Malhotra’s name (Table 6, p 30). Though in small numbers, there were those who preferred the BJP’s Sushma Swaraj or Madan Lal Khuranamorethan Malhotra. There were a few others who did not mention Malhotra’s name but said they would like any one from BJP. So while even traditionalBJP supporters were di-vided on their preference for the chief minster, Dikshit was the undisputed lead-er among all Congress leaders for the post of chief minister. Parties opposed to the Congress in the electoral race, mainly the BJP, misjudged what the common people in Delhi felt about the performance of the Dikshit-led govern-ment. The belief that the people were gen-erally unhappy with the ruling party and that it would be difficult for the Congress to control the tide of anti-incumbency after being in power for 10 years, negatively af-fected the BJP in two ways. One, it resulted in some complacency among BJP leaders. Two, its assessments about what people re-ally felt about the Congress government were grossly incorrect. The findings of the post-poll survey conducted by the CSDS in-dicate, in varying degrees, that two-thirds of Delhi’s voters felt satisfied with the per-formance of the ruling Congress govern-ment while only 30% felt dissatisfied. There has not been much change in the level of satisfaction within the state. In the 1998 and 2003 surveys, 54% and 65%, respec-tively, mentioned they were satisfied with the performance of the state government. If we compare the level of satisfaction of the people with that in other states, this is rea-sonably high. Not surprisingly, among tra-ditional Congress supporters, 88% were satisfied with the performance of the gov-ernment (Table 7, p 30). Even among tra-ditional BJP supporters, in varying degrees, 44% felt satisfied (8% fully satisfied, 36% somewhat satisfied) with the performance of the Dikshit government. The level of sat-isfaction with the performance of the state government was also reasonably high among traditional BSP supporters. So, people from different economic classes, up-per, middle and lower, all felt highly satis-fied with the performance of the Congress government and of Dikshit as the chief min-ister.Even on the Bus Rapid Transport corridor issue,thegovernment was not negatively evaluated. While 30% supported the view that the overall traffic situation had dete-riorated due to the BRT system, 36% be-lieved that it had madeitconvenientfor people to use public transportandthatit was only those using private vehicles who felt affected. One-third of the people did not express an opinion on this issue. Clear-ly, this did not reflect a mood of great dis-satisfaction with the ruling Congress. Though many did not approve of the poli-cy of sealing domestic property used for commercial purposes, its later reversal helped the Congress recoverlostground. Table 4: Shift Among Voters from Different Economic Classes (2003-08) CongressBJPBSP 2008 2003Change2008 2003Change20082003ChangeUpper class 42 41 +1 44 51 -7 6 1 +5Middle class 41 47 -6 38 37 +1 13 5 +8Lower class 39 53 -14 32 26 +6 19 10 +9All figures are in %.Table 5: Shift Among Voters from Different Communities(2003-08) CongressBJPBSP 2008 2003Change 20082003 Change 2008 2003ChangeJatav 3642-62313+103740-3Other dalit 40 63 -23 32 20 +12 22 8 +14Gujjar + Yadav 42 41 +1 27 30 -3 17 4 +13Other OBC 37 51 -14 39 34 -5 18 5 +13Muslims 63 67 -413 9 +410 3 +7All figures are in %.
STATE ELECTIONS 2007-08february 7, 2009 EPW Economic & Political Weekly30The Congress had an edge over the BJP because people rated it better in various as-pects of governance, the exceptions being the party’s ability to control prices and ter-rorism. While a quarter of Delhi’s voters believed that the Congress was better in controlling prices than the BJP, 35% people held the opposite view. On controlling ter-rorism, 30% believed the BJP was better than the Congress while 26% thought it the other way round. Still, the Mumbai terror-ist attack failed to turn people away from the Congress in big numbers. With live cov-erage of the Mumbai attack on television channels, 94% of the people knew about the incident when the election took place, but it affected the voting choices of only 9% of the voters. Though a majority of those influenced by the Mumbai incident shifted from the Congress to the BJP, there were a few who shifted from smaller parties to the Congress. The BJP did benefit from the Mumbai attack, but not to a large extent – its net gain was only 0.6%. There was the popular belief that the BSP would harm the Congress by eating into its support base. Findings of surveys con-ducted during various elections do suggest that the support bases of the BSP and the Congress overlap, both the parties attracting support from among the dalits and other poor voters. But Delhi presents a some-what different picture, more on the lines of Uttar Pradesh. Though dalits remained the backbone of theBSP, it performed so well mainly because it managed to draw some support from among Brahmins, Punjabi Khatris and other upper-caste voterswho had been traditional supporters of the BJP. Of the 38 constituencies where theBSP played spoiler (that is, it polled more votes than the difference between the winner and the runner-up) the BJP finishedsecond in 20 constituencies while the Congress did so in only 12. In that sense, the BJP suffered more from the new popularity of the BSP. Conclusions Migration to the city of Delhi has changed the social composition of its voters and its politics can no more be discussed in terms of the Jat vote and the Punjabi vote. Rapid mi-gration has added large numbers of dalit and OBC voters, who are mainly poor or belong to the lower-income groups. The success of the Congress in the last few elections has been mainly because the poor and lower-class voters have always been its natural allies. However, the last five years have seen a breaking up of old social divisions. The poor are no more as firmly for the Congress as before and the BSP has successfully managed to lure many of them away. This signals that the Congress cannot be com-placent with its third successive victory in Delhi. The decline in its share of the vote should also be a worry for the party. These results do not presage a comfortable victory for the Congress in the next Lok Sabha elections. If these results are converted intoparliamentary seats, the Congress has an edge over the BJP in all the seven LokSabha seats, but only marginally. A slight shift of voters away from the Congress could make the outcome very different. TheBJP faces a much deeper challenge. Though it managed to increase its share of the vote, it has lost popularity among its traditional supporters, the urban, the rich and the young. The party has never been popular among poor voters and if its tradi-tional supporters are deserting it, it is a matter of serious concern. Mayawati has reasons to feel satisfied with the reasonably good performance of the BSP. But the party cannot afford to be complacent. The upper caste and OBC support for the party is fragile. Dalit voters are divided with the BSP having failed to mobilise them the way it managed to do in Uttar Pradesh. The party still has a long way to go.NOTES1 During the municipal elections held in Delhi in 2007, the BJP won 164 seats while the Congress secured 64 seats and the BSP 15.2 Even one year after the sealing, in the recently conducted post-poll survey, though a majority be-lieved that it was incorrect to run businesses in residential localities, they disapproved of the policy of sealing existing businesses. 3 The CSDS conducted the post-poll survey in as-sembly constituencies randomly selected using the probability proportionate to size sampling technique. In each assembly constituency, four polling booths were randomly selected using the systematic random sampling technique. A sample of 1,674 respondents randomly selected from the electoral roll was interviewed using a structured questionnaire. The respondents were interviewed at their home and not at the polling station. The sample of respondents included 55% men, 23% dalits, 14% Muslims, and 2% Sikhs. About 28% of the respondents were interviewed in rural locations. The sample represented voters from all age groups and from all educational backgrounds. The 2003 survey was conducted in all 70 assem-bly constituencies of Delhi. In each assembly con-stituency, 20 locations were randomly selected using the systematic random sampling technique. A total of 14,460 respondents were interviewed at their homes using a structured questionnaire. All the respondents were selected from the electoral rolls. The sample of respondents included 53% men, 19% Dalits, 10% Muslims, and 4% Sikhs. It was also representative of voters from all age groups and from all educational backgrounds.4 There are 16 assembly constituencies that have migrants from Bihar and Uttar Pradesh in sizeable numbers. So much so they can play a decisive role in elections. These constituencies are Burari, Adarsh Nagar, Wazirpur, Uttam Nagar, Dwarka, Palam, Ambedkar Nagar, Sangam Vihar, Tugh-lakabad, Badarpur, Trilokpuri, Kondli, Seemapu-ri, Gokulpuri, Mustafabad and Karawal Nagar. Similarly, there are 17 assembly constituencies that have Punjabi voters in sizeable numbers. These are Timarpur, Badali, Shalimar Bagh, Shakur Basti, Wazirpur, Patel Nagar, Madipur, Rajouri Garden, Tilak Nagar, Hari Nagar, Janak-puri, Jangpura, Malviya Nagar, Kalkaji, Vishwas Nagar, Krishna Nagar and Shahdra. 5 As per the CSDS data unit’s regional classifica-tion, Delhi can be divided into five broad regions, the periphery, the north-west, the trans-Yamuna region, City South and City North. The periphery has 16 assembly constituencies. Trans-Yamuna has 16 assembly constituencies. North-west Delhi has 17 assembly constituencies. City North has 11 assembly constituencies and City South has 10 as-sembly constituencies.6 I have compared the vote share of the Congress be-tween 2003 and 2008 assembly elections in 16 con-stituencies located on the periphery of Delhi. It should be noted that a new delimitation of assem-bly constituencies has taken place since the 2003 assembly election and the boundaries of the con-stituencies have changed. So these 16 constituen-cies may not be exactly the same in terms of locali-ties and the vote share should be treated as an ap-proximation. References Kumar, Sanjay (2004): “Increasing Fluidity in Elec-toral Contest: Is This Mere Anti-Incumbency?” in Rajendra Vora and S Palshikar (ed.) Indian Demo-cracy: Meaning and Practices (New Delhi: Sage). – (2004): “A Tale of Three Cities”,Seminar 534, Feb-ruary, pp 33-36.Table 7:Level of Satisfaction with the Performance of the Congress Government Fully Satisfied Somewhat Somewhat Fully SatisfiedDissatisfiedDissatisfiedAll voters 28 35 11 19Traditional Congress supporters 55 33 4 6Traditional BJP supporters 8 36 14 37Traditional BSP supporters 5 37 21 24Upper class voters 29 37 9 22Middle class voters 28 38 10 20Lower class voters 27 32 12 18All figures are in %.Table 6: Preferred Chief Minister – Sheila Dikshit Leads the Race Sheila Dikshit Vijay Kumar MalhotraAll voters 32 13 Traditional Congress supporters 62 2Traditional BJP supporters 6 32Upper class voters 34 18 Middle class voters 32 15 Lower class voters 30 9 All figures are in %.
STATE ELECTIONS 2007-08Economic & Political Weekly EPW february 7, 200927Delhi Assembly Elections: 2008 Sanjay KumarPulling off its third successive win in the Delhi assembly election, the Congress demonstrated that public dissatisfaction with its Sheila Dikshit-led government was not as overwhelming as supposed. The Bharatiya Janata Party did gain three more seats and more of the popular votes but it did not have enough in its armoury to upset the ruling party. The main gainer in the election was the Bahujan Samaj Party, which won two seats and attracted a large chunk of the traditional support base of the Congress and the BJP. Delhi was the biggest “upset” in the latest round of assembly elections. The Congress in Delhi joined a small list of ruling parties or combines – the Left Front in West Bengal and Tripura, the Rashtriya Janata Dal (RJD) in Bihar, the Congress in Arunachal Pradesh, the Bharatiya Janata Party (BJP) in Gujarat and the Sikkim Democratic Front (SDF) in Sikkim – that have retained power three successive times or more. The victory was seenasa personal triumph for Chief Minister Sheila Dikshit who led the party to victory the second time around, a surprising achieve-ment in an era of “anti-incumbency”.The Congress secured a comfortable majority of 43 seats in the 70-member assembly. It won three seats less than in the last elec-tion and its lead in terms of popular votes came down to a mere 4 percentage points. Though theBJP managed to increase its tally of seats from 20 to 23 and wonmore votes in this election, its share of votesstill fell short of the majority mark. The surprise at the Congress victory was not without basis. Even before the elec-tions were held, Delhi was being counted as the one state where the BJP was sure to win. There was a popular perception of dissatisfaction with the ruling Congress Party in the city, a perception strengthened by the BJP’s victory over it in the Municipal Corporation of Delhi (MCD) polls in 2007.1 The confusion over sealing businesses in residential localities annoyed not just the business class but also the common people.2 Just when the Congress appeared to have negotiated the tide of popular anger with the Supreme Court allowing all commercial property to be reopened, the price rise brought fresh trouble. There was a sense that the Congress managedtostage a recovery after the BJP declaredVijay Kumar Malhotra to be its choiceforthe chief minister because popu-laropinion had it that he was no match for Dikshit. But the Mumbai terror attack, that took place less than 72 hours before the polling in Delhi, and the subsequent tide of popular anger against the central government again seemed to dim the pros-pects of the ruling party in the capital. What appeared to be turning theBJP’s advantage into near certain victory was the Bahujan Samaj Party (BSP) factor. After its victory in Uttar Pradesh, the BSP has been seriously thinking of expanding its base in other states. That it was taking the Delhi assembly elections seriously was clear from the very beginning when the party declared its candidates well in ad-vance. BSP supremo Mayawati addressed several rallies and campaigned extensively in the capital. The results revealed that, unlike in Madhya Pradesh and Rajasthan, theBSP’s efforts in Delhi succeeded: though it managed to win only two assembly seats, its share of the vote increased dramatically from 5.8% in 2003 to 14% this time. This paper examines these popular im-pressions in the light of the outcome of the election and evidence gathered by a post-poll survey carried out by the Centre for the Study of Developing Societies (CSDS), New Delhi.3 It argues that theBSP was a major factor that worked against the Congress. Also that the Congress made up for this loss by an overall posi-tive rating of its governance and the im-age gap between Dikshit and Malhotra. Popular opinion underestimated the im-pact of narrowing class divisions that helped the Congress. The party did suffer losses among the poor, but it maintained an upper hand over other parties even in this group. Its gains among the middle class and well to do voters proved deci-sive in making up for some of its losses. In ethnic and community terms too, the Congress managed to align itself with the changing social demography of Delhi much better than theBJP. At the same time, this achievement of the Congress is contingent, as the rise of the BSP opens the possibility that the structure of politi-cal competition is changing.Sanjay Kumar (sanjay@csds.in) is a Fellow at the Centre for the Study of Developing Societies in New Delhi. Table 1: Performance of Political Parties – Assembly Election 2008 VotesGain/LossSeatsGain/Loss Polled from 2003 Won from (%) (%) 2003Congress 40.3-7.843-4BJP 36.3+1.123+3BSP 14.0 +8.32 +2Independents 3.9 -0.9 1 0 Other smaller parties 5.4 -0.6 1 -1
STATE ELECTIONS 2007-08february 7, 2009 EPW Economic & Political Weekly28Before turning to an explanation, let us first take a close look at the verdict itself (Table 1, p 27). Though the votes of the Congress went down considerably, and the party ended up with a lesser number of legislators, it still managed to register an impressive victory, winning 43 of the 70 seats and 40.3% of the votes. Though the BJP managed to increase its tally from 20 to 23 and polled more votes in this election than the previous election, its share of votes remained far short of the majority mark. The BSP managed to win only two seats, but its share of the vote increased dramatically from 5.8% in 2003 to 14% in 2008. Independents and candidates from other smaller parties together polled less than 10% of the votes. BSP Gains TheBSP’s votes came mainly from constit-uencies dominated by migrants from Ut-tar Pradesh and Bihar. As per a rough clas-sification, there are 16 assembly constitu-encies where these migrants play a deci-sive role and many among them seem to have voted for the BSP.4 Most of these con-stituencies are located on the periphery of the city5 and in them, the BSP won 19% of the total votes, 5 percentage points more than its average vote in the city. Both the constituencies where theBSP won had large migrant populations. The BSP polled 7.5% of the votes in constituencies where there were migrants in sizeable numbers even during the 2003 assembly elections. But its support base increased enormously this time around, to the detriment of the Congress. In constituencies dominated by the migrants, the Congress polled only 36% of the votes, 4% less than its average vote in this election (Tables 2 and 3).While Delhi’s politics is largely discussed in terms of the Jat vote and the Punjabi vote, nearly 17% of the capital’s voters be-long to dalit communities. Among the dal-its, the Jatavs are the largest caste and they voted for the BSP in large numbers. The Jatav vote for the BSP was not as overwhelm-ing as in Uttar Pradesh, but still 37% of them voted for it during this election. There was hardly any shift in the Jatav support for the BSP in the last five years. But the same period saw a shift of non-Jatav dalit voters to the BSP’s side. During this election, 22% of the non-Jatav dalit voters voted for the BSP, compared to only 8% in the previous assembly elections.Dur-ing the last five years, the Congress lost substantial support among these voters. The BJP was also affected to some extent. The BSP managed to increase its vote share and harm both the Congress and the BJP. Among those who voted for the BSP in this election, 55% were those who had voted for it in the previous elections but it man-aged to snatch 28% of its votes from the Congress and 12% from the BJP. But the success of the BSP did not depend only on the dalit vote. The party’s popularity increased among voters of other communities as well, thanks to its strategy of distributing tickets to non-dalits. Dur-ing this election, 17% of the voters from the Gujjar and Yadav communities voted for the BSP compared to only 4% during the previous assembly elections (Table 5, p 29). The party also increased its support among voters belonging to other back-ward castes (OBC). Among the OBC, 18% voted for the BSP in thiselection compared to only 5% in the lastone.TheBJP lost sup-portamongthe Gujjars and Yadavs. There was also a noticeable shift among Muslim voters towards the BSP. We do not know how enduring the non-dalit shift towards the BSP will be, but it is clear that the party’s rise has adversely affected both the Congress and the BJP and dented the bi-party nature of political competition in Delhi. How Did the Congress Win?In spite of losing support, the Congress managed to win 43 seats because it regis-tered smaller victories compared to its vic-tories in the previous assembly elections. The average victory margin in this election was 10,158 votes, smaller compared tothe the average victory margin of 10,946 votes in 2003. But if we compare the average vic-tory margins of the Congress in the last two assembly elections, we find that in 2008 it was 9,748 votes, much less than the 12,784 votes of 2003. On the other hand, in spite of increasing its vote share, the BJP man-aged to increase its tally by only three seats because the party registered bigger victo-ries compared tothe Congress. During the 2003 elections,the average victory margin of the BJP was 7,007 votes, which went up to 11,151 votes in 2008. Of six seats decided by a margin of less than 500 votes, the Congress won four and the BJP two. In this sense, the Congress was simply luckier. If it had lost more marginal seats, its majority may have been thinner. Luck, however, cannot explain the in-cumbent party’s return to power. Findings from surveys conducted during previous elections indicated a sharp class polarisa-tion in Delhi. The Congress was extremely popular among voters in the poor and low-er-income groups, those living in jhuggi-jhopari and lower income class locations (Table 4, p 29). During the 2003 assembly election, the Congress got 54% and 53% votes among the very poor and poor vot-ers. Compared to this, theBJP got only 18% and 24% votes among the very poor and poor. The class division among Delhi vot-ers was sharp enough to even influence the voting choices of voters from different classes within the same caste or communi-ty. One saw the divide in voting preferenc-es between the rich and the poor even among the Brahmins or the Punjabis, the traditional supporters of the BJP in Delhi. This was also seen among voters from the Vaishya community. The poor tended to Table 2: Region-wise Analysis of the Results – Delhi Assembly 2008 Seats Turn-Congress BJP BSP Others out(%) Seats Vote(%) Seats Vote(%) Seats Vote(%) Seats Vote(%) Regions Trans-Yamuna 16 59.0 10 42.2 5 35.6 1 15.5 0 6.7 Periphery 16 55.7 8 33.5 6 30.9 1 20.0 1 15.6 North-west 1759.6 9 41.18 40.80 11.0 0 7.1 Citynorth 11 59.1 8 44.1 2 39.5 0 8.6 1 7.8 Citysouth 10 53.7 8 44.6 2 36.4 0 11.4 0 7.5 Total 70 57.6 43 40.3 23 36.3 2 14.0 2 9.3 Source: CSDS Data Unit.Table 3: Performance of Political Parties in Different Types of ConstituenciesCommunity Type Seats Turnout(%) Congress BJP BSP Others Won Vote(%) Won Vote(%) Won Vote(%) Won Vote(%)UP-Bihar migrants 16 58.0 8 36.1 6 33.9 2 18.9 0 11.0Punjabi dominated 16 58.0 10 43.3 6 41.1 0 8.6 0 7.0Peasant community 11 55.8 7 34.7 3 31.8 0 17.7 1 15.7Rest 27 58.11843.9 8 37.20 12.41 6.4Total 7057.64340.32336.3214.029.3Source: CSDS Data Unit.
STATE ELECTIONS 2007-08Economic & Political Weekly EPW february 7, 200929vote in favour of the Congress while the rich voted for the BJP (Kumar 2004). This time, the Congress suffered a set-back in its traditional support base of poor and the lower-class voters. The BSP suc-cessfully managed to make inroads among them. Since the poor and the lower-class people can hardly afford to live in the city, they tend to settle down on its outskirts. They either live in jhuggis, small huts, semi-pucca houses or in Janata flats. Many assemblyconstituencies on the periphery of the city have a sizeable migrantpopula-tion who generally belong to the poor or the lower-income class. In the 16 assembly constituencies located on the periphery of the city, the Congress suffered heavy loss-es. The party polled nearly 7% less in these constituencies compared to its average vote share.6 TheBSP was the major gainer in these constituencies. However, despite losing some of its huge lead among the working classes, the Congress managed to still finish ahead of the BJP in this section.The Congress managed to make up for the loss of its support in poor constituencies with a reasonably better performance in constituencies where there were sizeable numbers of upper class voters. These con-stituencies are mainly located at the two ends of the city, in the south and in the north. In these constituencies, the Congress did better compared to the rest of constitu-encies. In the nine assembly constituencies of South Delhi, with a large number of up-per middle class and rich voters, it polled 45% votes and in the 11 assembly constitu-encies of the North Delhi, it polled 44%. Though the Congress lost about 8% of the votes compared to the 2003 assembly elections, it still remained an umbrella party, getting support from voters across castes and communities. Though the BJP remained more popular among upper caste voters, especially those from the Punjabi Khatri, Vaishya and other commu-nities, the Congress managed to win votes even among them. The anger of the Sikhs against the Congress also seems to have come down. Though theBJP led among Sikh voters, 41% of them voted for the Congress. A majority of the Jat voters vot-ed for the BJP, but still, a sizeable number of them voted for the Congress. The unhappiness of Saj-jan Kumar, the Jat leader whose son was denied a party ticket, did damage the Congress, but it still did well and managed to get numerous votes from among the Jats. A sizeable number of voters from the Gujjar-Ya-dav communities also voted for the Con-gress. Though there was some shift among Muslim voters towards the BSP, the Con-gress remained the first choice of Muslim voters in Delhi. Its biggest worry should be the shift among dalit voters who have moved away to the BSP in sizeable num-bers during the last five years. Why Did the BJP Fail?The choice of Malhotra as the candidate for the chief minister affected the chances of the BJP. When asked who they would like to see as their next chief minister, 32% preferred Dikshit while only 13% men-tioned Malhotra’s name (Table 6, p 30). Though in small numbers, there were those who preferred the BJP’s Sushma Swaraj or Madan Lal Khuranamorethan Malhotra. There were a few others who did not mention Malhotra’s name but said they would like any one from BJP. So while even traditionalBJP supporters were di-vided on their preference for the chief minster, Dikshit was the undisputed lead-er among all Congress leaders for the post of chief minister. Parties opposed to the Congress in the electoral race, mainly the BJP, misjudged what the common people in Delhi felt about the performance of the Dikshit-led govern-ment. The belief that the people were gen-erally unhappy with the ruling party and that it would be difficult for the Congress to control the tide of anti-incumbency after being in power for 10 years, negatively af-fected the BJP in two ways. One, it resulted in some complacency among BJP leaders. Two, its assessments about what people re-ally felt about the Congress government were grossly incorrect. The findings of the post-poll survey conducted by the CSDS in-dicate, in varying degrees, that two-thirds of Delhi’s voters felt satisfied with the per-formance of the ruling Congress govern-ment while only 30% felt dissatisfied. There has not been much change in the level of satisfaction within the state. In the 1998 and 2003 surveys, 54% and 65%, respec-tively, mentioned they were satisfied with the performance of the state government. If we compare the level of satisfaction of the people with that in other states, this is rea-sonably high. Not surprisingly, among tra-ditional Congress supporters, 88% were satisfied with the performance of the gov-ernment (Table 7, p 30). Even among tra-ditional BJP supporters, in varying degrees, 44% felt satisfied (8% fully satisfied, 36% somewhat satisfied) with the performance of the Dikshit government. The level of sat-isfaction with the performance of the state government was also reasonably high among traditional BSP supporters. So, people from different economic classes, up-per, middle and lower, all felt highly satis-fied with the performance of the Congress government and of Dikshit as the chief min-ister.Even on the Bus Rapid Transport corridor issue,thegovernment was not negatively evaluated. While 30% supported the view that the overall traffic situation had dete-riorated due to the BRT system, 36% be-lieved that it had madeitconvenientfor people to use public transportandthatit was only those using private vehicles who felt affected. One-third of the people did not express an opinion on this issue. Clear-ly, this did not reflect a mood of great dis-satisfaction with the ruling Congress. Though many did not approve of the poli-cy of sealing domestic property used for commercial purposes, its later reversal helped the Congress recoverlostground. Table 4: Shift Among Voters from Different Economic Classes (2003-08) CongressBJPBSP 2008 2003Change2008 2003Change20082003ChangeUpper class 42 41 +1 44 51 -7 6 1 +5Middle class 41 47 -6 38 37 +1 13 5 +8Lower class 39 53 -14 32 26 +6 19 10 +9All figures are in %.Table 5: Shift Among Voters from Different Communities(2003-08) CongressBJPBSP 2008 2003Change 20082003 Change 2008 2003ChangeJatav 3642-62313+103740-3Other dalit 40 63 -23 32 20 +12 22 8 +14Gujjar + Yadav 42 41 +1 27 30 -3 17 4 +13Other OBC 37 51 -14 39 34 -5 18 5 +13Muslims 63 67 -413 9 +410 3 +7All figures are in %.
STATE ELECTIONS 2007-08february 7, 2009 EPW Economic & Political Weekly30The Congress had an edge over the BJP because people rated it better in various as-pects of governance, the exceptions being the party’s ability to control prices and ter-rorism. While a quarter of Delhi’s voters believed that the Congress was better in controlling prices than the BJP, 35% people held the opposite view. On controlling ter-rorism, 30% believed the BJP was better than the Congress while 26% thought it the other way round. Still, the Mumbai terror-ist attack failed to turn people away from the Congress in big numbers. With live cov-erage of the Mumbai attack on television channels, 94% of the people knew about the incident when the election took place, but it affected the voting choices of only 9% of the voters. Though a majority of those influenced by the Mumbai incident shifted from the Congress to the BJP, there were a few who shifted from smaller parties to the Congress. The BJP did benefit from the Mumbai attack, but not to a large extent – its net gain was only 0.6%. There was the popular belief that the BSP would harm the Congress by eating into its support base. Findings of surveys con-ducted during various elections do suggest that the support bases of the BSP and the Congress overlap, both the parties attracting support from among the dalits and other poor voters. But Delhi presents a some-what different picture, more on the lines of Uttar Pradesh. Though dalits remained the backbone of theBSP, it performed so well mainly because it managed to draw some support from among Brahmins, Punjabi Khatris and other upper-caste voterswho had been traditional supporters of the BJP. Of the 38 constituencies where theBSP played spoiler (that is, it polled more votes than the difference between the winner and the runner-up) the BJP finishedsecond in 20 constituencies while the Congress did so in only 12. In that sense, the BJP suffered more from the new popularity of the BSP. Conclusions Migration to the city of Delhi has changed the social composition of its voters and its politics can no more be discussed in terms of the Jat vote and the Punjabi vote. Rapid mi-gration has added large numbers of dalit and OBC voters, who are mainly poor or belong to the lower-income groups. The success of the Congress in the last few elections has been mainly because the poor and lower-class voters have always been its natural allies. However, the last five years have seen a breaking up of old social divisions. The poor are no more as firmly for the Congress as before and the BSP has successfully managed to lure many of them away. This signals that the Congress cannot be com-placent with its third successive victory in Delhi. The decline in its share of the vote should also be a worry for the party. These results do not presage a comfortable victory for the Congress in the next Lok Sabha elections. If these results are converted intoparliamentary seats, the Congress has an edge over the BJP in all the seven LokSabha seats, but only marginally. A slight shift of voters away from the Congress could make the outcome very different. TheBJP faces a much deeper challenge. Though it managed to increase its share of the vote, it has lost popularity among its traditional supporters, the urban, the rich and the young. The party has never been popular among poor voters and if its tradi-tional supporters are deserting it, it is a matter of serious concern. Mayawati has reasons to feel satisfied with the reasonably good performance of the BSP. But the party cannot afford to be complacent. The upper caste and OBC support for the party is fragile. Dalit voters are divided with the BSP having failed to mobilise them the way it managed to do in Uttar Pradesh. The party still has a long way to go.NOTES1 During the municipal elections held in Delhi in 2007, the BJP won 164 seats while the Congress secured 64 seats and the BSP 15.2 Even one year after the sealing, in the recently conducted post-poll survey, though a majority be-lieved that it was incorrect to run businesses in residential localities, they disapproved of the policy of sealing existing businesses. 3 The CSDS conducted the post-poll survey in as-sembly constituencies randomly selected using the probability proportionate to size sampling technique. In each assembly constituency, four polling booths were randomly selected using the systematic random sampling technique. A sample of 1,674 respondents randomly selected from the electoral roll was interviewed using a structured questionnaire. The respondents were interviewed at their home and not at the polling station. The sample of respondents included 55% men, 23% dalits, 14% Muslims, and 2% Sikhs. About 28% of the respondents were interviewed in rural locations. The sample represented voters from all age groups and from all educational backgrounds. The 2003 survey was conducted in all 70 assem-bly constituencies of Delhi. In each assembly con-stituency, 20 locations were randomly selected using the systematic random sampling technique. A total of 14,460 respondents were interviewed at their homes using a structured questionnaire. All the respondents were selected from the electoral rolls. The sample of respondents included 53% men, 19% Dalits, 10% Muslims, and 4% Sikhs. It was also representative of voters from all age groups and from all educational backgrounds.4 There are 16 assembly constituencies that have migrants from Bihar and Uttar Pradesh in sizeable numbers. So much so they can play a decisive role in elections. These constituencies are Burari, Adarsh Nagar, Wazirpur, Uttam Nagar, Dwarka, Palam, Ambedkar Nagar, Sangam Vihar, Tugh-lakabad, Badarpur, Trilokpuri, Kondli, Seemapu-ri, Gokulpuri, Mustafabad and Karawal Nagar. Similarly, there are 17 assembly constituencies that have Punjabi voters in sizeable numbers. These are Timarpur, Badali, Shalimar Bagh, Shakur Basti, Wazirpur, Patel Nagar, Madipur, Rajouri Garden, Tilak Nagar, Hari Nagar, Janak-puri, Jangpura, Malviya Nagar, Kalkaji, Vishwas Nagar, Krishna Nagar and Shahdra. 5 As per the CSDS data unit’s regional classifica-tion, Delhi can be divided into five broad regions, the periphery, the north-west, the trans-Yamuna region, City South and City North. The periphery has 16 assembly constituencies. Trans-Yamuna has 16 assembly constituencies. North-west Delhi has 17 assembly constituencies. City North has 11 assembly constituencies and City South has 10 as-sembly constituencies.6 I have compared the vote share of the Congress be-tween 2003 and 2008 assembly elections in 16 con-stituencies located on the periphery of Delhi. It should be noted that a new delimitation of assem-bly constituencies has taken place since the 2003 assembly election and the boundaries of the con-stituencies have changed. So these 16 constituen-cies may not be exactly the same in terms of locali-ties and the vote share should be treated as an ap-proximation. References Kumar, Sanjay (2004): “Increasing Fluidity in Elec-toral Contest: Is This Mere Anti-Incumbency?” in Rajendra Vora and S Palshikar (ed.) Indian Demo-cracy: Meaning and Practices (New Delhi: Sage). – (2004): “A Tale of Three Cities”,Seminar 534, Feb-ruary, pp 33-36.Table 7:Level of Satisfaction with the Performance of the Congress Government Fully Satisfied Somewhat Somewhat Fully SatisfiedDissatisfiedDissatisfiedAll voters 28 35 11 19Traditional Congress supporters 55 33 4 6Traditional BJP supporters 8 36 14 37Traditional BSP supporters 5 37 21 24Upper class voters 29 37 9 22Middle class voters 28 38 10 20Lower class voters 27 32 12 18All figures are in %.Table 6: Preferred Chief Minister – Sheila Dikshit Leads the Race Sheila Dikshit Vijay Kumar MalhotraAll voters 32 13 Traditional Congress supporters 62 2Traditional BJP supporters 6 32Upper class voters 34 18 Middle class voters 32 15 Lower class voters 30 9 All figures are in %.
