WWW.INDIANEXPRESS.COM

THE INDIAN EXPRESS,THURSDAY, NOVEMBER 12, 2020 @ieExplained #ExpressExplained DECISION EXPLAINED 2020 If there are questions of current or contemporary relevance that you 18 E would like explained, please write to [email protected]

LOKNITI-CSDS POST-POLL ANALYSIS FOR THE INDIAN EXPRESS Decoding the close Bihar verdict Held against the backdrop of the pandemic, the Bihar contest was also one of the closest in recent times. In this post-poll survey, Lokniti-CSDS unpacks the many aspects of NDA’s narrow victory over the — caste & community, confidence in Modi, and women voting more for NDA

However, to be able to make a bid for power, SHREYAS SARDESAI, the MGB needed an MY+. The vote came SANDEEP SHASTRI, CHART 1: LAST-MINUTE DECISIONS AND VOTING TRENDS CHART 2: VOTE TRANSFER BETWEEN JDU AND BJP CHART 3: VOTE TRANSFER to the MGB in the first two phases,and the al- SANJAY KUMAR liance with the Communist parties was a cru- Phase 1 saw the most last-minute decision-making; NDA's biggest gains Voted Voted Voted Voted BETWEEN RJD AND CONGRESS & SUHAS PALSHIKAR cial factor. In the last phase, seem to have came among those who decided at the last minute in phase 3 MGB NDA LJP Others Voted MGB swayed towards the NDA, according to our JDU-HAM contested seats AS THE counting of votes in Bihar continued HOW DID THEY VOTE? I MGB I NDA I Others INC contested seats data. Within the Dalit community, support for Traditional JDU-HAM supporters 14 75 5 6 through the day on Tuesday, the close nature I Decided choice on the day of voting Traditional INC supporters 84 the MGB was restricted to the Ravidas com- of the battle became increasingly obvious. The Traditional BJP supporters 7 55 13 25 munity and the Dusadhs. Musahars, however, 37% 27% 26% 38% 2% 27% Traditional RJD supporters 83 fact that the NDA managed a narrow victory BJP contested seats mostly voted for the NDA. The NDA also got became the clear headline. Yet, the verdict of RJD contested seats four-fifths of votes from , the commu- the Bihar voter has multiple small stories that 32% 22% 22% Traditional JDU-HAM supporters 15 70 4 11 Traditional INC supporters 82 nity to which Nitish belongs, and nearly three- add up to the big picture. In this article, we 36% 36% 71% Traditional BJP supporters 6 75 2 17 Traditional RJD supporters 90 fifths of the EBC vote. hope to unpack these multiple strands. Phase 1 Phase 2 Phase 3 Note: Figures are percentages Note: Figures are percentages It may be useful to focus on the polarisa- was able to overcome the tion on religious lines, especially in the third inevitable fatigue of being in power for a phase. This is particularly true for all Hindu decade. This was partly on account of the CHART 4: MOST IMPORTANT VOTING ISSUE CHART 8: GENDER GAP IN VOTE communities (except for ) who con- Mahagatbandhan (MGB) not being seen as CHART 6: THE YOUTH VOTE CHOICE AMONG COMMUNITIES solidated behind the NDA in this phase. a viable option, and partly due to the positive 2015 2020 2020 Three-fourths of the Muslim vote in the fi- sentiment associated with the performance Post Poll Pre Poll Post Poll Contrary to the popular narrative, MGB did not end up getting any Dalit, EBC and upper caste women voted more nal phase went to the RJD-led alliance but in of the central government. The caste calculus Development 31% 29% 36% significant vote advantage among the youth for NDA than their male counterparts did the Seemanchal area it was the AIMIM that continued to offer important insights while Unemployment/jobs 9% 20% 20% I I I I I secured a high share of Muslim votes. MGB NDA LJP GDSF Others I MGB I NDA I Others the youth vote got divided on gender lines. Inflation 16% 11% 11% 4141 39 38

38 Assessment of Nitish & Modi 36 37

The Lokniti-CSDS Post poll, on which this 37 36 36 Poverty/hunger/financial problem 6% 6% 5% 35 Women 37 38 25 analysis is based, got the vote shares of the Education/school/college-related 2% 7% 3% In the run-up to the campaign, a lot of at- Men 38 36 26

two major alliances reasonably close. We 16

16 tention was focused on anti-incumbency 14 14 Condition of roads/bridges 3% 4% 2% 14 7 projected a vote share of 39% for the MGB 6 against the Nitish government. The survey in- 5 5 5 5 5 4 4 Government's work/performance - 4% 6% 3 Upper caste women 12 59 29 and 36% for the NDA with an error margin of dicates that there certainly was a significant +/-3%. Eventually, the NDA got 37.3% and the Electricity/water 7% 1% 1% 18-29 30-39 40-49 50-59 60-69 Upper caste men 16 49 35 degree of fatigue with the government, but Others issues 16% 16% 13% years years years years years MGB 37.2%. All numbers reported here have women 82 5 13 no strong sentiment among voters to get rid been weighted by the final outcome. Can’t say/No response 10% 2% 3% of it. The jury was equally divided when it CHART 7: YOUTH VOTE BY GENDER Yadav men 84 6 10 came to whether the JD(U)-BJP government Late swing in last phase CHART 5: UNEMPLOYMENT VOTE Young men for MGB, young women for NDA Other OBC women 18 63 19 should get another chance (Chart 10). One of The post-poll survey indicates a substan- every 10 voters said that although the gov-

40 Other OBC men 18 59 23 39 39 39 39 tive last-minute swing in favour of the NDA. 36

34 ernment should return, it should be without

NDA trailed MGB massively among the ‘unemployment vote’, but 33 This was a vital factor in explaining the direc- NDA won the ‘development vote’ comfortably Dalit women 24 33 43 Nitish at the helm. Such voters, the survey 18 17 tion of the final verdict. One of every four re- 12 12 found, eventually ended up voting for the 7 6 5 5 5 5 5 Dalit men 31 27 42 spondents said they decided their vote only on NDA’s vote lead over MGB 4 NDA. What is also important is that the MGB the day of voting. Close to half of them voted for Post-Poll Pre-Poll 18-29-year 30-39-year 18-29-year 30-39-year Muslim women 74 6 20 was unable to consolidate the sizeable chunk an NDA candidate. This trend was stronger in Among those for whom Unemployment -27% -1% -old men -old men -old women -old women Muslim men 79 5 16 (40%) of voters who wanted the government the last phase of voting, when more than two- was most important voting issue Note: Figures are percentages to go. Only two-thirds of them voted for the Note: Figures are percentages thirds of those who decided on the day of vot- Among those for whom Development +15% +27% MGB. The survey also found a majority of vot- ing, chose NDA candidates (Chart 1). ers quite positive in their overall assessment was most important voting issue Among the one-third who decided their CHART 10: DIVIDED OPINION ON WHETHER NITISH GOVERNMENT SHOULD REMAIN of the Nitish government’s performance, al- choice either after the declaration of the can- Among those for whom Inflation was -10% -17% though these satisfaction levels were much didates or the start of the campaign, it was most important voting issue I State govt. should get I State govt. should get another I State govt. shouldn't I No response below what was recorded in the previous the MGB that did better. This implies two another chance chance but without Nitish at the helm get another chance elections (Chart 11). The dissatisfaction with clear trends. The whirlwind campaign by CHART 9: HOW CASTES AND COMMUNITIES VOTED the Nitish government was not of a sufficient (247 rallies in 20 days) 60% 30% 52% 42% 10* 40% high intensity to dislodge it. helped the MGB put up a stiff fight against MGB NDA LJP GDSF Others The levels of satisfaction with the Modi- the NDA. The Lokniti-CSDS pre-poll survey 15% 52% 7% 1% 25% 2010 10% 2015 6% 36% 2020 14% led central government were much higher in October had found the MGB trailing the 19% 51% 3% <1% 26% than with the Nitish-led state government NDA by quite a distance. Secondly, Prime (Chart 12). This may have helped the NDA off- 9% 55% 11% 4% 20% Note: Figures are percentages *This response was spontaneous and recorded as such only this time. Minister ’s rallies also seem Other Upper castes 16% 59% <1% <1 % 24% set some of the negative sentiment associ- to have made a difference in galvanising NDA ated with the state government. The survey Yadav 83% 5% 2% 3% 6% voters, particularly in phase 3, thus taking CHART 11: SATISFACTION WITH THE CHART 12: STRONG SATISFACTION WITH found that two of every 10 of those dissatis- the alliance past the majority mark. 11% 81% 3% <1 % 5% fied with the Nitish government’s perform- 16% 51% 6% 8% 18% NITISH KUMAR GOVERNMENT MODI GOVERNMENT WAS HIGHER THAN 2015 ance, but satisfied with the Modi govern- Vote transfer within alliances Other OBC/EBC 18% 58% 4% 3% 18% 4 3 10 5 ment’s work, ended up voting for the NDA. 13 12 23 30 It is clear that many BJP traditional voters Ravidas 34% 27% 9% 13% 18% 35 36 13 17 6 stayed away from the NDA in JD(U)-con- Dusadh/ 22% 17% 32% 3% 27% 6 Migrant households 8 6 tested seats, and this election may not have Musahar 24% 65% 1% 1% 8% How critical was the manner in which 46 ended up being so close but for the LJP dam- Other Dalits 24% 30% 4% 7% 34% 42 44 42 people perceived the actions of the central aging the prospects of the JD(U). The survey Muslims 76% 5% 2% 11% 6% and state governments in handling the lock- indicates that in BJP-contested seats, three- 2010 VS 2015 VS 2010 VS 2015 VS down? The survey indicates this issue may fourths of traditional BJP supporters voted have hurt the NDA’s prospects but was not 2 for the NDA, whereas in JD(U)-contested CHART 13: RATING GOVTS ON ATTITUDE TOWARDS 10 6 3 5 strong enough to deny it victory. Athough a 32 30 14 32 41 seats, the support shown by BJP supporters MIGRANTS (PRE-POLL) 12 26 20 very small fraction of voters cited the ap- for the NDA candidate was just over 50% State government's Modi government's 7 palling treatment of migrants during the lock- (Chart 2). In such seats, the LJP seems to be down as their most important voting issue, attitude towards migrants attitude towards migrants 11 have swung a segment of the traditional BJP the survey also indicates that the NDA did far 44 13 25 voters in its favour (around 13%), thus dam- during lockdown during lockdown 44 23 worse among migrant households than 5 5 2019 LS 2020 VS 2019 LS 2020 VS aging the prospects of the JD(U) candidates. 14 24 among non-migrants. Around four of every Many of these traditional BJP voters who 20 10 households said they had a member/s liv- 26 I Fully satisfied I Somewhat satisfied I Somewhat dissatisfied I Fully dissatisfied I No response shifted to LJP were Dusadhs and, to some ex- 31 ing outside Bihar and among voters from such tent, those from the upper castes. 18 Note: Figures are percentages Source: Lokniti-CSDS Post Poll Surveys in Bihar at the time of elections households, the NDA secured around four Interestingly, despite the MGB losing, the 24 33 percentage points less votes as compared to survey shows that vote transfer among MGB I Very Good I Good I Bad I Very Bad I No response non-migrant households (Chart 15). The MGB alliance partners was more or less smooth CHART 15: HOW DID MIGRANT CHART 16: TOP CHOICES FOR CM appears to have got 4 percentage points more with Congress and RJD supporters voting for votes among migrant households than non- HOUSEHOLDS VOTE? Since mid-October pre-poll, Tejashwi gained much more than each other’s candidates in almost equal CHART 14: RATING GOVTS ON HANDLING OF migrant households. In our pre-poll survey Nitish, but not enough, perhaps measure (Chart 3). The Congress not doing PANDEMIC (PRE-POLL) I MGB I NDA I Others in mid-October, while we had found many well, therefore, cannot be attributed to a poor voters dissatisfied with the way the state gov- Migrant Non-migrant transfer of votes from RJD supporters. I State government’s handling of pandemic Post-poll % Pre-poll % ernment had handled the situation, they had I Modi government’s handling of pandemic households households Tejashwi Yadav (RJD/MGB) 39 27 not been as critical of the central government Unemployment as key issue (42%) (58%) Nitish Kumar (JDU/NDA) 38 31 — another indication that Modi’s appeal may To the MGB’s credit, the strategy to put un- Fully Satisfied 17 29 have been at work (Charts 13 and 14). employment at the heart of its campaign did Somewhat Satisfied 31 30 (LJP) 5 5 The Bihar electorion was one of the clos- work in its favour. While more than one-third Somewhat Dissatisfied 16 14 25 25 36 Sushil Modi (BJP/NDA) 3 4 est in recent times. The community consol- of voters (36%) identified development as the Fully Dissatisfied 30 20 40 Lalu Prasad(RJD/MGB) 2 3 idation is nothing new and cannot ade- most important poll issue, one-fifth (20%) fo- quately explain the competitiveness. The No response 7 6 cused on lack of jobs (Chart 4). Among voters 35 39 Upendra (GDSF) 1 5 pandemic and consequent complications for whom the job crisis mattered the most, the provided a unique backdrop and yet these MGB ended up securing over half the votes Note: Figures are percentages Note: the rest took other names could be neutralised by the NDA. The com- and leading the NDA by a massive margin Source: Lokniti-CSDS Pre Poll Survey in Bihar between Question asked: After the election, who would you prefer to see as the petitiveness emerged from an unstated dis- (Chart 5). Thus, by promising 10 lakh jobs to October 10 and 17, 2020 Note: Figures are percentages next Chief Minister of Bihar? tancing from Nitish, a continued confidence voters, Tejashwi did sway the vote of these in Modi and a pattern of women voting more voters. However the attractiveness of that in favour of the NDA. promise got eclipsed by the grand narrative of vikas — those voters for whom develop- young voters on the whole as the MGB got by a good 5 percentage points (Chart 7). Older such major difference was, however, noticed competing consolidations — Yadavs and (Shreyas Sardesai is Research Associate at ment was the key issue mostly voted for the nearly the same proportion of votes from the women voters, however, voted for the MGB among Yadavs and Muslims with respect to Muslims on the side of the MGB, and upper Lokniti-CSDS. Sandeep Shastri is the Pro Vice NDA. Thus the MGB raised the right issues but 18-39 age group as the NDA (Chart 6). This hap- more. While the gender advantage for the NDA the NDA, although women from the two com- castes, Kurmi- and EBCs on the side of Chancellor of Jain University, Bengaluru and was not able to convert it into results. pened because of a gender divide in voting was just two percentage points overall, we no- munities were less likely than men to vote for the NDA, with Dalits being the swing vote the National Co-ordinator of the Lokniti preferences. The youth swing for Tejashwi was, tice major differences in how women and men the MGB. Given that women turned out more (Chart 9). Muslims and Yadavs, RJD’s tradi- network. Sanjay Kumar is Co-Director of the Youth vote and gender in a sense, mostly limited to young men and from certain castes and communities voted than men, this may have made the difference. tional voters, consolidated in a major way be- Lokniti programme at CSDS, Suhas Palshikar Although joblessness was a major poll is- not so much young women. While the RJD led (Chart 8). Upper caste, Kurmi, Koeri and EBC hind the MGB, at least in the first two phases. taught Political Science, and is currently the sue, particularly for the youth, it did not lead among young men in the 18-39 age group, it women were far more likely to vote for the Two competing consolidations Close to nine of every 10 Yadavs and three- Co-Director of the Lokniti programme and to a significant advantage for the MGB among trailed among women in the same age group NDA than men from these communities. No The elections seem to have witnessed two fourths of Muslims voted for the MGB. chief editor of Studies in Indian Politics)

METHODOLOGY

The data presented here is based on a post- ate to size method, and then four polling sta- started a day after voting had taken place by the four major political fronts – the NDA, Hilsa); Dr Md Irshad Ali (Asst Prof, BN by Prof Sanjay Kumar of CSDS. poll survey by Lokniti-CSDS in Bihar be- tions within each of the sampled con- and went on for about 4-5 days except for MGB, LJP and GDSF. College, Bhagalpur University); Dr Rajnish Field investigators were provided with tween October 29 and November 9, 2020. stituencies were sampled using the system- the last phase when the survey was com- The survey was supervised by a team of Kumar (Asst Prof, Jamuni Lal College, masks and sanitisers for their work and in- The survey was conducted among 3,612 atic random sampling method (selecting pleted in 2 days. The interviews were con- 7 supervisors associated with Lokniti’s team ); and Rakesh Kumar (Chainpur, terviews were conducted following proper voters at 148 polling stations spread across polling stations at fixed intervals). Finally, ducted face-to-face at the voters’ homes us- in Bihar — Dr Rakesh Ranjan (Assoc Prof, Kaimur). The survey data has been analysed physical distancing and mask rules. Field in- 37 Assembly constituencies. The sampling the voters to be interviewed were also sam- ing a specially designed phone application. University); Dr Kumar Rakesh Ranjan by Shreyas Sardesai and Himanshu vestigators were made to sign an undertak- design adopted was multi-stage random pled using the systematic method from the The final data set was weighted by Census (Asst Prof, LND College, Motihari); Prof Vijay Bhattacharya from Lokniti under the guid- ing that they were not showing any Covid- sampling. First, the constituencies were electoral roll of the sampled polling station. 2011 figures for age-group, gender, locality Kumar (VM College, Pawapuri); Dr ance of Prof Suhas Palshikar and Prof 19 symptoms and that they would take all sampled using the probability proportion- In every constituency, the fieldwork and community and the vote shares secured Mukesh Kumar Rai (Asst Prof, SU College, Sandeep Shastri. The survey was directed the necessary precautions in the field.