This is first of a series of articles on Lok Sabha elections in Kerala. It focuses on four Lok Sabha constituencies in south Travancore — Kollam, Pathanamthitta, Attingal and Thiruvananthapuram — where the BJP is expected to do well.
Kerala is turning out to be a tough state to forecast for this election season. One reason being most analysts’ inability to precisely quantify the rise of Bharatiya Janata Party (BJP) in vote share terms. Another reason being their inability to representatively gauge the shift of votes from either the Congress-led United Progressive Alliance (UPA), or the Left, to the other, or to the BJP.
All we know is that there is an electoral churning taking place, with votes moving from one camp to another in varied quanta, in different parts of the state. This haziness is also reflected in opinion polls which, perforce, take Kerala as one homogeneous whole, and predict about 15 seats to the UPA, around four to the Left, and perhaps one to the NDA or National Democratic Alliance.
But the truth is that Kerala is anything but homogeneous, and as this article will demonstrate, sizeable pockets of dedicated voters are often enough to swing a victory. This is especially true in a state where contests were traditionally straight two-way fights, with tight margins. However, the rise of the BJP means that the established groups — the UPA and the Left — have started to lose increasingly more votes, often at an alarming rate, to the new entrant. What this also means in electoral arithmetic is that even if the BJP doesn’t win a seat, the outcome gets decided by who lost more votes to the BJP.
This apparent absurdity manifested itself in assembly constituency after constituency during the 2016 Kerala assembly elections, and is expected to manifest itself in a more pronounced manner, during this election. In such a confusing situation, a little serious number crunching is useful for gaining some much-needed clarity on where political parties stand, vis-a-vis their electoral fortunes. To this end, we begin with four Lok Sabha constituencies of Kerala, in South Travancore: Kollam, Pathanamthitta, Attingal and Thiruvananthapuram.
The methodology adopted in this survey is as follows:
- Compare 2014 Lok Sabha results with 2009 Lok Sabha results
- Review 2014 results at assembly segment level (7 to each Lok Sabha constituency)
- Compare 2016 assembly results with corresponding assembly segment level results of 2014 Lok Sabha results
- Overlay latest opinion poll vote share forecasts on both 2014 and 2016 results to predict 2019 outcomes
- Factor in reality that the BJP vote base is presently more concentrated in certain regions, rather than being evenly spread across the state
- Factor in caste and religion demographics, because, shameful as it may be to state in the twenty-first century, many still do aggressively vote on community lines.
The broad demographics and historical voting patterns in Kerala are:
- Muslims (27 per cent) are largely concentrated in the northern Malabar province, and a majority traditionally votes for their own political party, the Indian Union Muslim League (IUML), or the IUML’s ally the UPA/United Democratic Front (UDF). Hardly any vote for the BJP.
- Christians (18 per cent) live mainly in Cochin and Travancore provinces (central and south Kerala). A majority vote for their own Christian parties like the Kerala Congress, or for their ally, the UPA/UDF. A small percentage has now begun to shift to the BJP.
- Hindus form 55 per cent of the state’s population, and are the majority in all districts except Malappuram in South Malabar (70 per cent Muslims). They are divided into four groups:
- Nairs (forward class, about 20 per cent) used to form the bedrock of the UPA/UDF, but since 2009, have started voting increasingly for the BJP. Today, they constitute the bulk of the BJP’s vote base.
- Ezhavas (OBC, about 25 per cent) form the core vote base of the Left parties. But, since the advent of Narendra Modi, they have begun to gravitate towards the BJP.
- Dalits (10 per cent). There is little quantitative evidence on precise voting patterns, but some recent electoral results and anecdotal reports suggest a gradual move towards the BJP.
- Tribals (2 per cent). They are concentrated in the hilly tracts of the Western Ghats, in two constituencies — Wayanad and Idukki. Voting patterns are unclear, with wild swings from election to election; possibly candidate dependent.
The inferences are quite interesting, especially since forecasting from 2014 results can be misleading, because the BJP’s actual expansion first happened in 2016 during assembly elections (when it won one seat — Nemom). In 2014, the BJP polled only 10 per cent state-wide, and its surge was largely restricted to one Lok Sabha constituency — Thiruvananthapuram. This rose to 15 per cent in the 2016 assembly elections.
Today, opinion polls project 22 per cent for the BJP in 2019, with twice as many votes shifting from the UPA than from the Left. This is a rise of 12 percentage points over 2014, which may be split as +8 per cent from the UPA and +4 per cent from the Left (not at all a uniform assumption, I know, but this is how heterogeneous Kerala is).
However, the biggest problem is that the BJP votes are often concentrated in pockets, making sample surveys non-representative, and forecasts speculative. Nevertheless, the 2016 assembly surge, allied with recent opinion polls, offer some indicators, which we shall use in making constituency-wise predictions.
The demographics of this coastal constituency is about 65 per cent Hindus, 20 per cent Muslims and 15 per cent Christians. It is represented by N K Premachandran of the Revolutionary Socialist Party (the RSP, a UPA constituent), who won in 2014 with a narrow 4 per cent margin (RSP 46.5 per cent, Left 42.2 per cent, BJP 6.7 per cent). This is a UPA hold from 2009.
What is interesting is that the RSP actually trailed in three assembly segments, and only won in 2014 because of a handsome lead garnered in Chavara assembly segment (65 per cent of the total margin), courtesy the then sitting MLA who was also a minister in the Oommen Chandy government. The BJP was nowhere in the picture, polling its usual, desultory 6-7 per cent.
But in the 2016 assembly elections, the UPA (called the UDF in Kerala) lost all seven constituencies to the Left, by painful margins ranging from 15-28 per cent. Again, the only exception was Chavara, which they lost by a relatively narrower -5 per cent. Strangely, the BJP did not put up candidates in two of the seven constituencies, but where it did, its vote share doubled and trebled, mainly at the cost of the Congress. Indeed, in Chatanur, the BJP polled 25 per cent to push the Congress to third place. What this means is that the UPA enters 2019 in Kollam with a negative 14 per cent vote share deficit over the 2014 results, the Left at a healthy 51 per cent, and the BJP still at third place.
Applying current opinion poll trends to the 2014 results, and the vote-swing split, we get in 2019: UPA 38.5 per cent, Left 38.2 per cent, and BJP 18.7 per cent. Forecast: too close to call.
But, if we apply the latest opinion polls to the 2016 assembly election trends, the starkness of the vote shifts is dramatic in Kollam. Now, the forecast is: UPA 25 per cent, Left 47 per cent and BJP 23 per cent. The Left appears to have a clear advantage, not just because the chances of the UPA recovering the votes it lost to the BJP are extremely low.
To make matters worse for the UPA, the BJP has put up a Christian candidate, Sabu Varghese. Consequently, the usual exertions of painting the BJP as a fascist Hindutva force, hell bent on oppressing the minorities, will elicit no more than what dramatists like to call ‘a hollow laugh’. Add on the fact that the UPA’s 2014 victory was largely because of a single assembly segment, that it looks set to lose more votes to the BJP, and that it lost all seven assembly seats in 2016.
Forecast: Left wins in Kollam, with BJP making very strong gains primarily at UPA’s expense.
Pathanamthitta is a hilly inland constituency and home to Sabarimala temple. The district demographics are about 57 per cent Hindu, 38 per cent Christian and 5 per cent Muslim. The sitting Member of Parliament is Anto Antony of the Congress, who successfully retained his seat in 2014 with a 6.5 per cent margin (Congress 41.3 per cent, Left 34.8 per cent, BJP 16 per cent). He managed to maintain a decent lead in all seven assembly segments, even if his victory margin was down 10 per cent from 2009.
In the 2016 assembly elections, the UPA/UDF lost five of the seven assembly seats of Pathanamthitta to the Left, often losing many of their votes to the BJP in the process. This is important: the Left won in some seats not because of a dramatic surge in their votes, but because far more UPA/UDF votes shifted to the BJP. Totalling the seven seats, we get: UPA/UDF 36 per cent, Left 42 per cent and BJP 19 per cent.
In 2019, the Left have nominated popular sitting MLA Veena George as their candidate against sitting MP Anto Antony of the Congress. This means that in Pathanamthitta, there will be no ‘adjustment’ between the two groups to jointly sink the BJP (as they have done earlier elsewhere on numerous occasions).
The BJP has sent in K Surendran – their big-hitter, who lost in the 2016 assembly elections by a mere 237 votes. He was at the forefront of the Sabarimala agitation, and had to suffer the ignominy of being arrested by the Kerala Police on orders, in a churlish and self-defeating show of strength by the Left. He is extremely popular, equally acceptable to all castes, and a powerful orator. The rousing reception he received at Thiruvalla railway station, following his nomination as the BJP’s candidate for Pathanamthitta, is a clear indicator of the local mood. This will be a tight fight, and the BJP is confident of its chances. The numbers indicate exactly this: applying our formula to the 2014 numbers, we get in 2019: UPA 33 per cent, Left 31 per cent, BJP 28 per cent. Far too close to call, since an error margin of 3 per cent could put any party in front.
Applying the same to 2016 numbers, we get in 2019: UPA 28 per cent, Left 38 per cent and BJP 31 per cent. But, a further 3-4 per cent vote-shift from the Left to the BJP would just as easily put the BJP in front. This cannot be ruled out because Pathanamthitta was at the heart of the temple agitation, and a vote shift from the Left to the BJP, especially from amongst the Left’s core Ezhava vote base, is only to be expected. Remember, Surendran is an OBC who has crossed the caste-divide, and hugely popular. The BJP may thus want Narendra Modi to hold a rally here, because he has the ability to further consolidate votes above traditional caste divides.
Also, there is a piquant problem for the Left. If George wins, then the assembly seat she vacates will require a by-election, giving the BJP one more opportunity to consolidate their votes, in the long run up to the 2021 assembly elections. The Left might not want that, because Kerala is all they have left in India, and the BJP is now starting to eat into that vote base too. Either way, it’s advantage BJP.
Forecast: K Surendran of the BJP wins in Pathanamthitta.
This coastal constituency bordering the state capital has been a Marxist bastion for long. Dr A Sampath of the Left won here in 1996, 2009, and again in 2014, sweeping all seven assembly segments with an 8 per cent margin. Vote shares: UPA 38 per cent, Left 46 per cent, BJP 11 per cent. Constituency demographics are roughly 65 per cent Hindu, 20 per cent Christian and 15 per cent Muslim.
The Left retained their dominance in the 2016 assembly elections, winning six out of seven constituencies. But interestingly in Attingal, the BJP registered a 3.5 per cent vote swing from the Left, countering the state-wide trend of the BJP gaining more votes from the UPA/UDF. Also, the BJP did surprisingly well in Kattakada (27 per cent) and Nedumangad (23 per cent), making it a lottery for the other two sides. In Attingal (SC) assembly seat, the BJP polled 20 per cent, and the RSP (a UPA/UDF member) was reduced to 23 per cent. Aggregating the seven seats, the vote shares in 2016 were: UPA/UDF 37 per cent, Left 42 per cent and BJP 18 per cent.
Considering the above, it would be more prudent to tweak the vote-split formula to 6 per cent-6 per cent. With this, a forecast on the 2014 results gives us in 2019: UPA 32 per cent, Left 40 per cent, BJP 23 per cent.
But, the gaps narrow when we use the 2016 numbers. Now, the projected vote shares become: UPA 31 per cent, Left 36 per cent and BJP 30 per cent. The Left looks to retain its lead, but if either the Modi factor, or supra-caste consolidation, or a good campaign, or a good candidate, comes into play, then it is anyone’s game except the UPA’s. The BJP could then be in with a faint chance.
Let us not forget that the BJP have nominated their senior-most lady, Sobha Surendran, from Attingal. She is a firebrand orator who has contested elections previously. Indeed, in the 2016 assembly elections, she came second polling 29 per cent in Palakkad, and might even have polled more, if manpower, funds and resources hadn’t been diverted by the BJP to adjoining Malampuzha seat (where too, the BJP came second, against Marxist stalwart and former Kerala chief minister, V S Achuthanandan).
Forecast: Left wins in Attingal unless the BJP surprises everyone with a spirited campaign.
The state capital is very interesting, because it presents the BJP with its best chance of winning in Kerala. In 2009 and 2014, Dr Shashi Tharoor of the Congress was returned to Parliament from this constituency, albeit the last time, by a proverbial whisker.
In 2014, he defeated the BJP’s O Rajagopal by just 15,470 votes (1.8 per cent). The vote share split was UPA 34.1 per cent, BJP 32.3 per cent and the Left at 28.5 per cent. Interestingly, the BJP led in three out of seven assembly segments. It performed poorest in Neyyatinkara segment, where the Christian vote is around 40 per cent.
In the 2016 assembly elections, the UPA/UDF won only three seats. The Left won three, and the BJP, one. The aggregated vote share of the seven seats was: UPA/UDF 32.5 per cent, Left 38.2 per cent and BJP 26.8 per cent. An important observation is the shift of votes from the BJP to the Left between 2014 and 2016. This is a pointer to the electoral worth of the Modi factor in 2014. Still, in four seats in 2016, the BJP got nearly 30 per cent of the vote share or above, and in two, above 20 per cent. Only in one seat, Neyyatinkara (again), did it get a poor 11 per cent.
Using the 2014 numbers as a base, our forecast formula gives us: BJP 44 per cent, UPA 26 per cent and Left 25 per cent. Forecast: BJP wins.
But, the BJP is already at 32 per cent, and there exists the very real possibility of a minority vote consolidation under the UPA. Accordingly, it is probably more prudent to change the vote-shift formula, to better reflect the impact of tactical voting. Although, as has been the case so often of late in many parts of India, every consolidation engenders a counter-consolidation, which means that a part of the Left vote will then necessarily move towards the BJP.
In electoral math, this means that if there is an 8 per cent vote shift from the UPA to BJP, and a 6 per cent shift from the Left to the UPA (a net loss to the UPA of 2 per cent), there will still be a minimum 2 per cent shift from the Left to the BJP as well (a net loss to the Left of 8 per cent). Total vote gain for BJP would then be 10 per cent. This actually comes back to the opinion polls which suggest a 12 per cent overall rise in the BJP’s vote share in 2019, with the effect manifesting itself more intensely in selected seats. Thiruvananthapuram appears to be one such seat. Readers must also not forget the visible rise in popularity of the BJP’s candidate for this seat – Kummanam Rajashekaran.
Thus, running this modified math on the 2014 results, we get for 2019: BJP 42 per cent, UPA 32 per cent, Left 21 per cent. Forecast: BJP wins.
Just to make sure, let us discount the Modi factor, the impact of Sabarimala, Kummanam’s popularity, and run this modified formula on the 2016 assembly results. We still get: UPA 30 per cent, Left 30 per cent and BJP 37 per cent.
Forecast: BJP’s Kummanam wins in Thiruvananthapuram.
The million dollar questions then are: will the shift of minority votes from the Left to the UPA (if that happens in any material fashion), be enough to offset the shift of votes from the UPA and Left to the BJP? Will the index of opposition unity, if implemented, be strong enough to make the BJP lose? Can the Marxists afford to compromise thus? Only time will tell. Of course, if Modi holds a rally in Neyyatinkara, then the arithmetic won’t matter, and the BJP will win this seat hands down.
In the interim, here is a probabilistic summary of the forecasts for these four, electorally important, seats:
References: all electoral data from Election Commission of India website.