On 13th October 2020, I had written an article on this platform before any opinion polls came out. After my article, there were as many as six different opinion polls all of which either gives NDA a big victory or makes it a tough contest between NDA and MGB with an advantage to NDA. However, I will stick to two opinion polls of national pollsters like C-Voter and Lokniti-CSDS.
In my analysis, I always consider that the data of opinion polls since the vote shares is accurate. I just differ in their predictions. I am really not aware of how the pollsters predict the seats. I always believe that arithmetic addition of vote shares never happens in real politics. This fact has been proved by me in various articles published on this platform also. I would like to analyze based on these two opinion polls. I have taken both the survey raw data and will do the analysis based on those data.
As per C-Voter data, BJP and JDU are getting 33.8% and 14.4% vote share whereas RJD and Congress are getting 24.3% and 11.7% respectively. LJP and others are getting 6.7% and 9.1%.
Just refer to my previous article where I have mentioned the core vote base of BJP, JDU, RJD, Congress, and LJP as 24.69%, 15.4%, 19.3%, 10.3%, and 6.75% through excessive data mining. Just compare the opinion poll data. BJP is able to get (33.8-24.69) 9.69% extra vote share beyond its base vote share of 24.69%. JDU is in fact losing (15.4-14.4) 1% of its vote share. That speaks how anti-incumbency has an effect.
Parties, when they become very unpopular, lose even their core vote banks. RJD is getting (24.2-19.3) 4.9% extra vote share thanks to anti-incumbency to Nitish Kumar. Similarly, Congress is getting (11.7-10.3) = 1.4% extra vote share beyond its base vote share. LJP’s vote share remains intact as I have calculated LJP’s base vote bank as 6.75% and the opinion poll shows as 6.7%.
Now, the pollster had just added the vote share of BJP and JDU and then find the numbers for NDA as 160 as a comfortable margin. But in practice, can it be done?
All opinion poll says that Nitish Kumar is suffering huge anti-incumbency and is highly unpopular at the moment. In fact, the same C-Voter survey says that only 31% of people want Nitish as CM whereas 26% of people say no, and 34% of people want a new leader. Isn’t it a contradiction between the predicted number and these survey figures?
First, all BJP candidates will get JDU votes. So, BJP is definitely performing well. Can the same be said about BJP voters for JDU candidates? The point that must be noted here is that BJP doesn’t have a strong face at the Bihar polity level. Thus, most BJP voters are in fact fans of Narendra Modi. No doubt, Narendra Modi will do aggressive campaigning. But does that make BJP voters vote for Nitish Kumar?
We have seen in Rajasthan, MP, and Karnataka how the voters vote differently in Parliament and assemblies. For the General election, Narendra Modi was able to win almost all the seats of the state. But in assembly election? Don’t forget the ‘Modi Tujhse bair nahin, Raje teri khir nahin’ slogan which sunk Vasundhra Raje’s BJP government in Rajasthan. The same has happened in MP. Also, in Karnataka despite being the largest party, BJP couldn’t get a majority.
Here comes LJP’s role. The displeased BJP voters may not vote for MGB but may not also vote for JDU! They may either skip the voting or give their vote to LJP. In this process, MGB might be a gainer and in some seats or LJP can win handsomely. Point is that I still believe that JDU, despite BJP’s popularity, may suffer huge electoral loss and get sub-30 seats. ‘Modi Tujhse bair nahin, Nitish tera Khair nahin’ may be a new slogan for this Bihar assembly election. The problem is that pro-BJP voters generally are not bound by the dictate of their favourable leaders. That’s always proven. So my conclusion is despite the pollster’s prediction, Nitish may lose this assembly election. Let me re-establish the same with the Lokniti-CSDS survey data.
This also endorses my above analysis of C-Voter opinion poll data.
In this poll-survey, I couldn’t find the party-wise vote share rather it gave the vote share of alliances. The pollster perhaps collected data in the name of alliances instead of individual parties. According to this NDA (BJP+JDU+HAM+VIP) is getting 38% vote share, MGB is getting 32% vote share, GDSF and LJP are getting 7% and 6% respectively while others are getting 17% Vote share.
In the first instance, the data would look very different from that of C-Voter. But as I said earlier, I always consider that the data is valid but predictions aren’t. According to Lokniti-CSDS, NDA is wining 133-143 seats whereas MGB winning 88-98 seats. But let’s keep aside the seat projections and consider the vote share data.
As I said earlier, you just can’t add vote-shares arithmetically. I have explained above how the BJP votes may not go to the JDU. That’s why such vote share comes from this survey when asked about voting preference as an alliance. The NDA vote share as an alliance reduced from (33.8+14.4) 48.2% (as per C-Voter) to 38% vote share when asked their preference for the NDA alliance.
Similarly, the MGB vote share too reduced from (24.3+11.7) 36% (as per C-Voter) to 32% as GDSF ate away its vote share which C-Voter didn’t consider at all. This survey also mentioned 17% of votes for other mean non-decided voters in preference to alliances. As I have explained, this means that no vote share can ever be added arithmetically because all parties have different dynamics of base vote bank.
Here in the Lokniti-CSDS survey, the difference between the vote shares of MGB and NDA is just 6%, and how can we forget that the margin of error of such surveys is 3%. That means if the error works in favour of MGB, then the vote share maybe 35% each for NDA and MGB! Now imagine, in such a case, who would be the loser? MGB alliance vote share will be uniform whereas NDA vote share will be concentrated on BJP seats (due to popularity of BJP) and will be fragmented in JDU seats because of anti-incumbency.
According to Lokniti-CSDS, LJP is not gaining although GDSF is getting some MGB votes. But the non-decided voters of 17% may play a spoilsport for JDU and even NDA if they vote for LJP or MGB.
The point remains that with all these data I still believe Nitish is losing Bihar as per the above logical analysis.
The only caveat is that sometimes, pro-BJP voters in a dissent mode don’t vote. In such a case, there could be low turn-out and because GDSF (Grand Democratic Secular Front, which is an alliance of many small parties) may eat-away sizable vote shares of MGB making NDA win the majority. This would also enable Nitish Kumar to retain power albeit with a lesser mandate which could officially make JDU as a smaller partner to BJP in Bihar.