How the grand alliance won


back to issue

ELECTORAL victories and defeats in India can rarely be attributed to just a few reasons. And when they are of the magnitude of the mahagathbandhan’s (MGB) victory and NDA’s defeat in the recently concluded Bihar assembly elections, a multitude of factors have to go right or wrong, and together, to produce the final outcome. Equally, to claim that everything mattered is to sin in the world of political analysis. The task of a political analyst is to tease out the important factors amidst the pile of reasons that may possibly have shaped the results. Why did the NDA lose so badly in Bihar? And what helped the MGB to score a landslide win in this seemingly difficult battle? An analysis of pre-poll and post-poll surveys conducted by Lokniti-CSDS in Bihar provides important insights into what changed within a month that produced a decisive verdict in favour of the mahagathbandhan.

In this essay we talk about the late surge in favour of the MGB which in our opinion was largely a result of NDA losing sight of its own campaign narrative. The NDA, and especially Prime Minister Modi, painted a picture of Bihar which was at odds with voters’ perception of the state and its political environment. We also discuss the role of communal issues and incidents that produced a counter-polarization among certain sections of voters which neutralized the NDA’s advantage among the upper castes and urban middle classes.

The BJP-led NDA seemed to have an advantage over the mahagathbandhan in the pre-poll survey which was conducted in the last week of September, a couple of weeks before the first vote was cast. The pre-poll survey estimated that had the election taken place in the last week of September, the NDA would have polled 42% votes and the MGB 38%. Further analysis of the data, however, cautioned us to qualify this finding as we believed that the pre-poll estimates may be misleading at a time when a bitter and long campaign had just started to gain momentum.1 The pre-poll survey asked respondents whether they were sure of turning out to vote on polling day, and would they vote for the same party they had indicated in the survey or might their decision change. Approximately three in every five voter said that they were sure of voting for the same party as indicated to our investigators. The data presented in Figure 1 points out that the NDA’s lead over MGB was primarily among those respondents who were unsure of their voting preferences.


A few weeks later, when we began the post-poll survey, we found that a significant turnaround had taken place with the mahagathbandhan taking a substantial lead over the NDA. We estimated a gap of about four percentage points between the two alliances, though eventually it turned out to be even bigger.2 So what changed in a matter of a few weeks that made the NDA shift from certain success to a grand failure in Bihar?

A comparison of our pre- and post-poll surveys suggests a late surge in favour of the MGB, with a large proportion of voters (26%) making up their mind about who to vote for only a couple of days before. Figure 2 shows that though the MGB had a big advantage among early deciders (non-swing voters) in our post-poll survey, the late deciders (swing voters) seemed to have given them a decisive lead over the NDA.3 It is important to highlight that contrary to popular perceptions, the analysis of survey data suggests that the late deciders are not overwhelmingly from the lower castes or the poor. Thus the popular commentary that the ‘silent voters’ had swung the election in favour of the MGB in the last leg is not borne out by the data.


What seems to have hurt the NDA most in the final stretch is that it overestimated the popularity of Modi and his government and underestimated the strength of the Lalu-Nitish alliance on the ground. Like many political observers, the BJP’s campaign managers thought that the mahagathbandhan may have got the arithmetic right but the chemistry between the allies was simply missing.4 The NDA completely lost its campaign narrative by the first week of October by simply harping on the ‘unholy’ alliance among MGB partners and the possible return of Lalu’s ‘jungle raj’ in Bihar. Their calculation behind this strategy was simplistic: attract Dalits and EBCs while consolidating urban middle class voters behind the NDA. At the same time, the RSS chief Mohan Bhagwat’s statement on reservations provided a new lease of life to MGB’s campaign which may have affected a marginal increase in their votes.5


The NDA from there on committed hara-kiri by making one mistake after another. An NDA campaign that could have focused on economic issues, infrastructure development, job creation, alongside the poor performance of Lalu Yadav as chief minister, instead focused on the rhetoric of Pakistan over performance. The gai (cow) triumphed over governance in its mobilization strategy, and instead of presenting a counter to Nitish Kumar’s model of development, Narendra Modi and the BJP took an easy route – announcing a financial package of over one lakh crore rupees to woo the voter. In fact, Prime Minister Modi and Amit Shah loomed so large in BJP hoardings that the state leadership was completely eclipsed, providing sufficient fodder for the Bihari vs Bahari campaign narrative.


Increasing Perception of Development Among Bihari Voters

Voters who saw Bihar as…

2005 Oct



Highly developed (1-3)




Moderately developed (4-7)




Underdeveloped (8-10)




Don’t know




Source: Bihar Post-poll Survey, 2005, 2010, and 2015. Lokniti-CSDS.



Communal Incidents Gave No Advantage to the NDA


Total Seats


(in %)





Vote (in %)


Vote(in %)


Districts with more than


20 communal incidents







Districts with less than


20 communal incidents







Source: Lokniti-CSDS Data Unit.

While Nitish and his alliance gained likes, as the comparison of our pre and post-poll surveys show, the popularity of Modi and his government declined among voters. In September, only 42% of respondents said Nitish’s government should get another chance; in the post-poll survey, this had increased to 52%. In the pre-poll survey, only 27% of respondents said they would like Nitish to be chief minister again; in the post-poll survey, this was 40%. Moreover, barring the upper castes, voters’ initial unease with Nitish having tied up with Lalu also subsided as the campaign progressed. While only 32% of voters had approved of the alliance during Lokniti’s pre-poll study in late September, the figure rose to 42% by the time the post-poll was conducted.


The MGB, particularly Nitish Kumar, made price rise, and especially the soaring price of pulses, a big election issue. The MGB supporters mocked the PM for having gone from ‘har har Modi’ to ‘arhar Modi’.6 Thus, it seems that during the campaign, the Modi-Shah combine somehow became the incumbent and Nitish-Lalu the opposition. Nitish and Lalu left no stone unturned to explain to voters the failure of the Modi government to deliver on promises made during the 2014 elections.

However, to be fair, the NDA entered the election lagging behind the combined vote base of the grand alliance of the Congress, RJD, and JD(U). In the 2014 Lok Sabha elections, the NDA had secured 39% of the vote, whereas the combined total of the grand alliance was over 43%. The BJP leadership was perhaps aware that 2014 was a high point for the party and it would be difficult to repeat this performance, as its vote share had declined by six to eight percentage points in most assembly elections since then. Thus, it also inducted Jitan Ram Manjhi in its alliance. The BJP’s social arithmetic was simple: build a coalition of upper castes, lower OBCs and Dalits to counter the Muslim-Yadav-Kurmi consolidation behind the grand alliance.


While there was indeed an upper caste polarization in favour of the NDA, a comparison of our pre-poll and post-poll survey clearly indicates that their social arithmetic did not click on the ground. The pre-poll survey indicated that the NDA would win massive support among Dalits and make serious inroads among the OBCs. The analysis of post-poll data shows that while the NDA managed to largely retain its vote base among the lower OBCs and Kurmi-Koeris, its vote share among Dalits (Paswans and other Dalit communities) and Yadavs declined significantly. The mahagathbandhan’s social coalition on the other hand stood firmly behind it. In a state where approximately two-thirds of the population is poor and lower middle class (based on assets owned, house type, and household income), the survey data suggests that the grand alliance had a massive advantage among poor and lower class sections of the electorate. The NDA, on the other hand, received significant electoral support primarily from the middle and upper middle classes.7

Prime Minister Modi and the BJP did everything to project that the formation of an NDA government in Bihar would herald an era of hope and change. They painted a bleak prospect of Bihar under Nitish Kumar. Voters, however, clearly had a better image of Bihar than what Modi painted. Not only did a large proportion of respondents claim that their household economic condition had improved, they also appreciated the performance of the Nitish government in the provision of electricity, roads, among many other indicators.


The data presented in Table 1 clearly shows that in the last decade, a majority of Biharis thought that their state had developed. On being asked to rate the development of their state on a scale of one to ten, with one being most developed and ten least developed, most chose to place Bihar between 4 and 7. Only about 18% rated it as underdeveloped, a drastic decline from the figure of 65% that the 2005 Lokniti-CSDS survey in Bihar had indicated when the same question was asked. This popular assessment is also backed by more formal economic and performance related data released by various government agencies.


Higher Religious Polarization in Districts With More Communal Incidents


Districts with less than 20 communal incidents

Districts with more than20 communal incidents


Vote for MGB (in %)

Vote for NDA (in %)

Vote for MGB (in %)

Vote for NDA (in %)

Upper caste















Lower OBC




















Source: Bihar Post-poll Survey, Lokniti-CSDS, 2015.

This election also saw the stoking of communal issues such as a legal ban on cow slaughter, reservations for Muslims, and an increase in the number of communal incidents in the state.8 Appu E. Suresh of The Indian Express in a report on 21 August stated that since June 2013 there had been a threefold increase in the number of communal incidents.9 We classified all the 243 assembly constituencies into two categories based on the district-wise data on communal incidents in Bihar since June 2013: constituencies that fell in districts that experienced more than 20 communal incidents and constituencies that fell in districts which witnessed less than 20 communal incidents. The former witnessed 2.5 percentage point lower turnout during the 2015 assembly elections in comparison to districts with a lower number of incidents. Similarly, the proportion of urban population in districts with a higher number of incidents was almost double in comparison to districts with fewer incidents. These two points are well established in the literature on Hindu-Muslim violence.


Table 2 shows that the NDA seemed to have done relatively better in areas that saw more communal incidents. However, even in these districts the NDA trailed behind the mahagathbandhan both in terms of votes and seats. The data presented in Table 3 shows that in communal incident prone districts, the NDA managed to attract a larger portion of non-upper caste Hindu voters. But these districts also witnessed a counter-polarization among Muslims, thus neutralizing the advantage the NDA could have had in these constituencies.


The lesson for the BJP from this election lies in a simple Newtonian formulation – every action has an equal and opposite reaction. Polarization of any kind generates its own opposition and incentivizes counter-polarization.10 The BJP won a historic mandate in 2014, largely aided by voters on the middle of the ideological spectrum. Unfortunately, the party seems to have been in denial about their role in its success. The NDA paid the price of excessively relying on Narendra Modi’s popularity and acceptance. As with Delhi, Modi seems to have lost out to an equal if not more charismatic opponent. Nitish Kumar won this seemingly difficult battle of Bihar and he was ably supported by Lalu Yadav in this herculean task.


* This essay draws heavily from the articles written by Lokniti-CSDS team members for The Indian Express. The authors thank all the contributors for graciously allowing us to use their analysis for this essay.


. For details on these caveats, see, ‘Advantage NDA as Bihar Gets Ready: Lokniti-CSDS Pre-Poll Survey, The Indian Express, 7 October 2015.

2. See, ‘NDA Loses Lead After Late Surge in Favour of Nitish Kumar-Lalu Prasad in Bihar: Post-Poll Survey, The Indian Express, 7 November 2015.

3. What distinguishes swing voters from non-swing voters is a matter of degree. All voters are potentially open to change their vote up until the moment it is cast, but voters differ in their degree of uncertainty about who they will vote for. The big question however is whether swings are a response to events that take place during the campaign or are merely reversions to predictable positions as voters become more informed about the candidates.

4. See, for example, Rahul Verma and Pranav Gupta, ‘Good Arithmetic, But No Chemistry, The Hindu, 23 September 2015.

5. The post-poll survey indicated that 37% of voters had not heard of Bhagwat’s statement, and 7% were ambivalent over the issue. The remaining 56% were equally divided in their support and opposition to the statement, and their voting intention for the NDA and the MGB was equally intense.

6. In the post-poll survey, inflation emerged as the second most important poll issue behind development, unlike the pre-poll survey where unemployment was the second most important concern. The proportion of those blaming the central government for price rise also went up.

7. The polarization along caste and class lines is also reflective of deeper tensions in the state, a point highlighted by Yogendra Yadav in his analysis of Bihar election results. See, Yogendra Yadav, ‘Political Vacuum in Bihar Win’, The Hindu, 23 November 2015.

8. Though a large percentage of the Bihari electorate principally agrees with the beef ban, as over three-fourths of electorate indicated, the post-poll data clearly suggests that the BJP did not have a large electoral advantage on this issue. While the NDA had a mere three percentage point advantage among those who supported the beef ban, the MGB’s advantage among respondents with other opinions was massive. Simply put, the cow issue turned its back on the NDA.

9. Appu E. Suresh, ‘Three-Fold Surge in "Communal Incidents" in Bihar After BJP-JD(U) Parted Ways’, The Indian Express, 21 August 2015.

10. For a similar point see, Pratap Bhanu Mehta, ‘A Vote Against Hubris’, The Indian Express, 9 November 2015.