One interesting trend in election 2016 is the increased localisation of opinion polls. There is now a much higher incidence of local polls than national polls. Secondly, only one national pollster polled all States, an unusual trend. Secondly, with the increased engagement of voters widespread of dissemination of election information (TV and Online) and large number of opinion polls, it is becoming increasingly difficult to predict election results. Still I undertake this hazardous activity more to guage how well we were able to read the incoming information from the States. 

Opinion Polls: First, an analysis of recent history.  In Bihar, 2 polls had predicted a Grand Alliance win, 2 had predicted an NDA win and 1 had predicted a tie. The range of seats predicted was MGB(137-108), NDA(128-95). In Delhi, 4 opinion polls had predicted a win for AAP, 3 had predicted a win for BJP, 1 had predicted a tie. The range of seats predicted were AAP(53-23), BJP(45-17). One interesting trend in election 2016 is the increased localisation of opinion polls. There is now a much higher incidence of local polls than national polls. Secondly, only one national pollster polled all States, an unusual trend.

Kerala

I believe it will be a comfortable win for the LDF. The local elections last year provided the first evidence of this trend. The corruption scandals (Bar Bribery specifically) had damaged the image of the UDF. The aggressive campaigning by the LDF ensured that almost all of their core voters were retained while gaining some from the UDF. The BJP’s clever alliance making and campaigning ensured that it gained both upper and lower caste hindu votes from the UDF. The only segment that held out in favour of UDF was women. This segment too may have shifted in favor of LDF (even if marginally) after the Jisha rape incident.

Opinion Polls: Out of the 5 opinion polls released in the last one month of the campaign, 3 are predicting a win for LDF, one for UDF and one is a tie. The range of seats for LDF are (89-63) and for UDF (75-50). In other words, the opinion polls are also a predicting a clear win for the LDF.

Assam

I believe this will be a comfortable win for the NDA. The dissatisfaction with the Congress party is high because it is suffering from severe incumbency having been in power for 15 years. This is primarily due to issues around economic growth and age old issues of immigration. The Congress sufffered a severe setback during the Lok Sabha election in 2014 and it is likely that the trend may have only worsened for it. I anticipate that some more hindu votes may have shifted towards the BJP and its allies AGP. With the addition of the BPF in the alliance, the NDA alliance has become a formidable one. The one last hope for the Congress party was that the Bengali muslim vote might be open to voting for it to stop the BJP. This may have indeed happened in many constituencies but the AUDF will still retain bulk of the vote it won in 2014 given that it is seen as a better representative of Bengali muslims (migrants). In the overall context, the BJP is likely to win more than 40% of the vote giving it a landslide.

Opinion Polls: Out of the 3 opinion polls released in March and April, the NDA was ahead in all the 3 with a seat range of 78-48. The Congress was given a seat range of 53-36 seats. The opinion polls indications suggest a comfortable win, may be even a landslide for the BJP

Bengal

Bengal is a complicated exercise because of two reasons. One, the elections were spread over a month period with intense campaigning during that period. Second, voters in Bengal are less likely to reveal their true voting intentions when compared with other states. Capturing the mood is quite difficult. Still, I believe this will be a narrow to comfortable win for the TMC. One, the rural economy in Bengal has done much better than the urban economy. Farmers and agriculture workers appeared to be quite happy with the TMC specifically on Incomes, Roads and a host of schemes that directly affected them. The Left front did not appear to be able to offer a positive alternate vision. Secondly, on the leadership question, Mamata was way ahead when compared with the rest. This meant neutral voters were more open to voting for the TMC. Thirdly, while the TMC regime was seen as corrupt, nearly half of voters did not appear to be open to the Left Front given the long rule from 1977 to 2011. Last but not the least, in these circumstances, many voters who voted for the BJP in 2014 may have shifted to Mamata instead of moving to the Left-INC coalition. The Left-INC coalition was a much more formidable combination than the Left alone and was able to attract many TMC voters who were not loyal to the TMC. However, at the same time, the alliance may have led to some traditional Congress voters and BJP voters of 2014 shifting to the TMC . In sum, the TMC vote share would certainly have increased from 2014 while it is also likely that the Left-INC vote may have increased from the 2014 election. However, the TMC strength is spread across a much larger number of constituencies unlike the Left-INC combination. In this context, I would give the election to the TMC

Opinion Polls:  Out of the 4 opinion polls carried out close to the first phase, all 4 predict a win for TMC. The range of seats predicted are 201-160 for TMC and 127-64 for Left-INC.

Tamil Nadu

This is the most intriguing of all the 4 elections. Unlike other States, it appeared that the incumbent enjoyed a huge advantage over the rest. Also, unlike other States, the opposition launched a very scientific campaign some 9 months ago connecting to young voters. Unlike previous elections, opposition alignment against the ruling party was much weaker and there is a clear contest between more than 2 alliances in many seats. At the same time, the ruling party ran a poor campaign during the final stretch of the campaign. The floods in North Tamil Nadu and general levels of unemployment in urban Tamil Nadu offer the opposition a great opportunity to win votes in that region and amongst youth voters. Further, the muddied position of the AIADMK on prohibition gives the DMK alliance a great opportunity to win some of the women votes that were strongly in favor of the AIADMK in the past.The other complexity is the performance of both the PWF and BJP. Both have run spirited campaigns and along with the likely caste base of some of the PWF parties, some of the anti-AIADMK vote may not accrue to the DMK but to the PWF and BJP. Further, the intensity of the attacks around corruption on both the AIADMK and DMK may have damaged the AIADMK and slowed down the momentum in favour of DMK. So in sum, a very difficult election to call. Given the final momentum and trends, I am leaning in favor of a hung assembly. But I won’t be suprised to end with egg on my face. 

Opinion Polls: Out of the 8 opinion polls released in May, 6 have reported a win for the DMK alliance with a range of 141-51 seats. 2 have reported a win for AIADMK alliance with a range of 182-73 seats. All the pollsters are local and many of them have never been heard before. The range of forecasts are extraordinarily high and make them less reliable than in other States

My overall Forecast summary

2016 Forecast