Google Analytics

Friday, March 11, 2011

Battle for Tamil Nadu: Just How Formidable is the DMK Alliance?

Topline Findings:

1. With the DMDK (Vijay Kant) the third force in the state aligning with the AIDMK, the political battle in Tamil Nadu reverts back to its bi-polar character from a tri-polar one as seen in 2009. The Index of Opposition Unity (IOU) correspondingly rises sharply.

2. Normally a higher IOU bolsters the opposition chances to victory. This advantage is however swamped by electoral arithmetic being stacked heavily against the opposition AIDMK alliance. If 2009 Lok Sabha performance is used as a benchmark, then the AIDMK alliance appears considerably less formidable than their DMK rivals, trailing by a whopping 9-11%. A two digit vote share margin can result in landslide win, similar in scale to what the AIDMK achieved during the 2001 Assembly elections. The AIDMK alliance then won 197 out of the 234 seats, with the AIDMK gaining a majority on their own.  This raises the hope of DMK to replicate the AIDMK’s landslide win of 2001 but factoring anti-incumbency, the battle could still be a cliff hanger, if not ending in a stunning surprise.

3. If the start up vote share holds, the Indian National Congress (INC) is poised to be the second largest party in the state. In the eventuality of any DMK split due to family feuds post the death or incapacitation of DMK patriarch, Karunanidhi, the Congress positions itself to play the role of a key partner of one DMK faction in a scenario where the AIADMK begins entering into a phase of terminal decline.

4. While both have a pan Tamil Nadu presence, the DMK core vote base  as per their showing in 2009 is estimated around 28%, while those of the AIDMK is only around 22%. This difference of 6% between the respective poles of the two alliances is what gives the DMK alliance a platform to build a formidable alliance in addition to offering greater flexibility in picking and choosing its partners. At the level of constituents, the top 3 parties within the DMK alliance have a combined vote that decisively crosses 40% while the AIDMK struggles to cross this mark even having 14 political parties within its umbrella. To give DMK a run for its money, the AIDMK should either have the PMK or the Congress, preferably the latter. The entire story of this election maybe encapsulated here. But is it? Once pre-poll studies are published in the days ahead, we can get a better picture.

5. Corruption is likely to be by and large a non-issue, given that AIDMK and its leader Jayalalithaa has previously been embroiled in many a corruption scam.
Accordingly a campaign that focuses on the 2G scam would be akin to a pot calling the kettle black! Apart from providing some marginal dividends to AIDMK in some urban pockets of the state, it is highly unlikely to touch a chord with the electorate. Instead, the arrest of Raja, ex Telecom Minister at the centre could even help the DMK alliance wean away significant sections of Dalits who are largely one of the backbones of the AIDMK political coalition. In the case of a landslide win by the DMK, it is expected to deflate the high decibel opposition campaign in what is known as the 2G scam.

6. The AIDMK alliance needs at least a 5% swing by Election Day to ensure a dead heat, if not forge into the lead. This though not impossible, looks a very tall order, given that the dice of electric arithmetic is heavily loaded against them. However an imaginative campaign could close the gap it faces. The AIDMK alliance could perhaps take a leaf out of Chandra Babu Naidu, opposition leader in Andhra Pradesh who bolstered his personal stock by taking on the for-profit micro-finance industry in the state. Banning the latter or castrating them like Andhra Pradesh did through legislation could provide an issue that offer some promise in generating some vote share swing in their favour.The AIDMK's best chances lie in the DMDK's capacity to mop up anti-incumbency votes and transfer them in full to the alliance. But whether this happens at all remain the moot question.


The defining feature of all national coalition politics from 1999 had been Andhra Pradesh and Tamil Nadu being the key game changer states. During this period, no government could be formed at the centre without their key participation. Governed by regional parties, the state accordingly wields a clout, largely disproportionate to their actual numbers at the centre.

The politics of Tamil Nadu for a long time had been the story of the bitter rivalry between the two stalwarts of the Tamil film industry, friends turned foes, M. Karunanidhi and MG Ramachandran (MGR). After having been a faithful follower of the DMK and Karunanidhi for long, MGR who till then had kept himself out of electoral politics successfully managed to create a huge political force (AIDMK) with his entry into politics. It was then carried forward briefly by his late wife, and then by his friend, co-star and protégé, Jayalalithaa.

The politics till the last decade of the 20th century in the State was therefore was more about being pro-MGR (AIDMK) or pro-Karunanidhi (DMK). Caste initially did neither matter nor religion, both had support within all religions and caste and the supporters were equally passionate about their leaders. But as caste and to a lesser extent, religious parties began to be formed and assert themselves during the last decade, the two Dravidian parties, began to experience a progressive decline in their vote share so much they are now at sub-critical levels preventing winning elections on their own steam. To overcome these limitations, the two started contesting in alliances. So the last decade witnessed coalitional politics in the state characterized by a contest essentially led by two political poles- the DMK and AIDMK respectively.

Another peculiar characteristic of Tamil Nadu is its huge propensity to deliver highly one sided verdicts. One alliance normally gets more or less wiped out, while the other makes a sweep. In 2001, AIDMK clobbered together a formidable alliance that included the Tamil Manila Congress (TMC); INC; PMK, CPM, CPI, Forward Bloc; Indian National League (INL); Republican Party of India (RPI) and some independents. A gap of more than 10 per cent translated into a massive 197 of the 234 seats victory for the ADMK-led alliance with the ADMK alone winning 132 seats.

In the 2004 General Elections, an alliance led by the DMK clean swept the state 40-0 despite being decimated by the AIDMK in the Assembly polls four years earlier. The DMK alliance then included the re-united Congress (after the merger of Tamil Manila Congress with its parent body); the two Communist parties; the PMK and the MDMK. Because the decimation of the opposition was as comprehensive as it could be as reflected in the 40-0 result, 2004 is widely considered by political pundits as the benchmark for the ultimate formidability of the DMK alliance.

Since the two Dravidian parties became progressively weaker during the last decade in this state, issues do not matter as cold alliance arithmetic to all political parties. In the absence of ideology, allies bolting from one alliance in one election only to return back in the next are therefore a common phenomenon. True to form, by the time of the next General Election in 2009 came about, the DMK alliance accordingly lost a good chunk of their formidability with MDMK, PMK, CPI and CPM bolting from its stables. With most polls predicting an AIDMK win, the DMK alliance stunned pundits by running away with 70% of the seats. This they had to thank the polity of the state changing from a bi-polar contest to a tri-polar contest due to the emergence of a relatively new party, DMDK, led by another film star, Vijay Kant, who ran away with over 10% of the votes. The Index of Opposition Unity (IOU) being low gave the DMK alliance a strike rate of 70%.

Tamil Nadu during the last decade had been the Waterloo for psephologists.  The frequent changes within an alliance are part of the problem. Poll forecasts often tend to be prone to carry errors of both omission and commission in the constituent membership of alliances as forecasters get confused which party belongs to which coalition. Additionally, as and when they switch alliances, it is difficult to get a sense of their intrinsic popular support, since it is only when political parties contest on their own that realistic estimates can be drawn. For example PMK claims an individual vote share of 7% but it is difficult to ascertain their veracity since the last time they contested on their own was way back in 1991.

Partly Tamil Nadu proving the pollster’s nightmare is also accounted by the changing character of state elections to a tri-polar contest, increasing the complexity to forecast seats. With the DMDK, the third force during the last election, now contesting within the AIDMK coalition umbrella, the state has reverted back to a bi-polar contest that makes poll forecasting very much more less hazardous as compared to their experience of the last decade!

So What’s Different this Election? 


Both alliances have expanded their constituent party numbers since 2009 - the number of parties within the AIDMK going up from 5 to 14 and the DMK from 3 to 7. 
Under the deal early this week, DMK will contest 121 seats, PMK 30, Viduthlalai Chiruthaigal Katchi  (VCK)10, Kongu Nadu Munnetra Kazhagam 7, IUML two and Moovender Munetra Kazhagam one, the last two seeking election on the DMK symbol.

In the rainbow alliance led by opposition AIADMK, DMDK floated by actor Vijaykant has been allotted 41 seats, Manitha Neya Munnetra Kazhagam three, Puthiya Tamizhagam two, Samathuva Makkal Katchi of actor Sarathkumar two and Natalum Makkal Katchi floated by actor Karthick, All India Moovendar Munnetra Kazhagam, AIFB, Kongu Makkal Ilaignar Peravai and RPI one each.
The alliance with the  MDMK, CPI and CPM are uncertain with negotiations hitting a road block.
 The 2009 General Election results could be used as the bench mark to get a sense of the formidability of the alliances in the state. The table above provides a snapshot of the last parliamentary elections but the one below illustrates better the difference between the central poles of the two alliances:

It is evident from the table that while the DMK as the leader of its alliance is strong enough to lift the vote share of the entire alliance by performing better than the other constituents, the AIADMK being totally discredited and on the decline has lost this ability. The AIDMK is in fact piggy backing on its allies rather than the other way around. The AIDMK was even not able to capitalise on anti-incumbency factors gifted by its rival DMK that controls the government in the state as these votes were gleefully mopped up by the DMDK led by film actor, Vijay Kant.

The starting line-up for the DMK is accordingly 42.68% as there is no desertion within its ranks since 2009. It augments its strength further by bringing into the alliance - the PMK with 6.75% vote share; KNMK who contested the parliamentary elections independently and recorded a vote share of 1.99% and two minor parties like the IUML and MMK that together has a vote share around 2%. Adding up to 12%, this when added to the DMK vote kitty of 2009 gives the alliance a whopping 52% of the votes, at least on paper. Under the first passed the post system, such a vote share makes the alliance almost impregnable.

The starting line-up for the AIDMK is 38.74%, which when adjusted for desertions within its ranks (PMK with vote share of 6.75%) would drop to 32%. To offset this decline, the AIDMK brings within its fold the DMDK that recorded a vote share of 10.1 in the last parliamentary elections and eight other smaller parties that together command another 2%. This gives the AIDMK alliance a vote share around 43% at least on paper.

The two alliances (DMK & AIDMK) accordingly control a whopping 95% of the votes, making the present contest basically a bi-polar one. In 2009, the DMDK in contrast acted as the third force as it positioned as an alternative to both the alliances. Accordingly, the DMDK ate into the vote share of all the major players. Though founded only in 2005 quickly touched the 10% mark in 4 years. The DMDK hopes its growth momentum continues to enable it to reach a vote share of 15% these elections. However, it is more likely to see a massive erosion in its vote bank.

The DMDK is widely seen as a Congress prop, designed to cut the votes of its rival, a role another film actor, Chiranjeevi played to perfection in Andhra Pradesh.  The entire growth of the party had been financed and with the blessings of the Congress. While ties with the Congress still continues even as the DMDK joined the rival AIDMK, Congress finance for its current election has dried up, throwing a spanner into the momentum growth of the party. Further, since the DMDK had positioned itself as an alternative for both - the DMK and the AIADMK - in the previous polls, its vote bank could see significant level erosion since it has aligned with the AIADMK.

So to what level could its erosion be? We estimate this around 2% overall. This would bring down the vote share of the AIDMK alliance to near sub-40% levels, giving over a double digit lead to its rival DMK alliance. In the 2001 Assembly elections, a lead of 10% translated into a landslide win for the AIDMK alliance. It won a massive 197 of the 234 seats with the AIADMK alone winning 132. Such a double digit lead could see the DMK trying to replicate the 2001 results.

But it has to be kept in mind that while the real contribution of the DMDK to the AIADMK alliance vote kitty maybe around 8%, by piggy backing on the alliance and contesting 44 seats, the DMDK may still make headway to realise its dreams of recording 15% vote share or near about - the additional 7% is what the AIDMK alliance contributes to its vote kitty.

The DMDK got 8.38% votes in 2006 elections and bettered its performance by securing 10.08% votes in the parliamentary elections in 2009. The DMDK secured more than 20,000 votes in 27 assembly segments, more than 15,000 votes in 75 segments, more than 10,000 votes in 169 segments and more than 5,000 votes in 228. In the last Assembly election, the DMDK secured only one seat. But this time, by piggy backing on the AIDMK, it hopes to get a haul of 15 seats. This is the most likely outcome as DMDK supporters are likely to vote in constituencies where their party is contesting in large numbers and mass abstains from voting where the party is not contesting. In short, the full transferability of the DMDK vote to the AIDMK is greatly in doubt. The party itself will concentrate its campaign in the 27 seats it secured more than 20,000 votes. The result: while its own Alliance gets thrashed, the DMDK makes a quantum leap in evolving as a major contender for power in the future.

So Can AIDMK Bridge the Gap?

The AIDMK alliance needs to whip up a 5% swing to target a dead heat, if not forge into the lead in the next 30 days. Is this even possible? Affirmative. The DMK being in government is vulnerable to the anti-incumbency effect and thus the possibility always exists. But is this even probable? This remains a more difficult question. 


Since the formation of the DMDK who positions itself as an alternative to both the DMK and AIDMK, it was this party that was the primary beneficiary of mopping up the anti-incumbency vote in the state and as a result put a spanner in the wheel of AIDMK to make a comeback as it deprived the latter of the anti-incumbency vote.

This also seem to suggest that the AIDMK is too discredited for anti-incumbency votes to be attracted to it as an alternative magnet. The growth of the DMDK is more likely to be an indicator of the increasing public fatigue with both the Dravidian parties. Now with the DMDK within the AIDMK umbrella, the alliance looks to the former to boost it vote share by continuing to mop up the anti-incumbency vote which in turn boosts the vote share of the entire alliance. Simply put if the DMDK is to contribute 15% (than the 10% it demonstrated in 2009) to the AIDMK alliance vote share, then the Tamil Nadu Assembly elections immediately competitively livens up instead of being a drab one-sided affair.

As explained earlier, leave alone anti-incumbency it is expected to garner this elections, it  looks highly unlikely that the DMDK will totally transfer its core votes to the AIDMK alliance as a whole.  At best it should transfer only a fraction of its total vote. This character would possibly give the DMDK a higher strike rate than the AIDMK alliance as a whole under whose banner it currently contests under.

So if the AIDMK intends to generate a 5% swing in its favour, rather than look to the DMDK growth momentum to generate it, the AIDMK should identify an issue that cuts an emotional chord with the people. In the 2G Spectrum scam involving A Raja, DMK ex-Minister of Telecom Ministry at the centre, the AIDMK think it found such an issue. But the record of the AIDMK, particularly its party supermo, Jayalalithaa, is equally sullied and that’s the reason why Tamil Nadu voted her out.   

Trying to capitalize on the 2G scam would be akin to a pot calling the kettle black. The wily patriarch of the DMK realized very early an opportunity in a crisis by declaring that Raja was being targeted only being a Dalit. The net outcome could possibly be that the some vote gains in a few urban pockets are more or less offset by weaning away some of the Dalits who are among the major support base of the AIDMK. In short, what Jayalalithaa gains in the swings she may end up  losing in the roundabout by focussing on the 2G scam. In the days ahead, opinion polls would provide a greater insight whether corruption can be a winning issue.


1. This analysis is based solely on the party performance of the Lok Sabha elections in 2009 and not based on any opinion polls. Besides both alliances are yet to firm up agreements with their respective constituents and neither alliance have even begun candidate selection yet. Accordingly, the confidence levels of this study is extremely low but all the same there is nothing to suggest it it cannot be representative of broad trends blowing at the start line of the campaign.

2. The next few weeks will see a spate of opinion polls hitting the market which will provide a better sense in changes in vote share since 2009. This blog would re-produce these as well as those in Assam and West Bengal who are also going to polls. In the first week of April, we will publish a more detailed seat forecast and another as a commentary on exit poll results.  This post is to be treated as more as a backgrounder to the election to the state

About the Author: Rajan Alexander is a management consultant in the non-profit sector who dabbles in psephology as a hobby just to cook a snook at professional opinion and exit polls. When the latter during the Lok Sabha 2005 was predicting a Jayalalithaa-BJP win in Tamil Nadu, Rajan predicted their complete rout, a 40-0 win for the DMK-Congress alliance, which came to pass. In the 2009 Lok Sabha, Rajan predicted 27-30 seats for the DMK-Congress alliance, who secured 28 seats.  


  1. Faulty analysis. Difficult to trust someone who uses 'AIDMK' for AIADMK,for it is the second 'A' that differentiates it from DMK. Fails to understand that vote share of any party is embedded in that of alliance partners, especially in the Lok Sabha poll. For example, deducting 6.75% from PMK (percentage of votes polled by PMK in 2009 as part of AIADMK front. ) is absurd, as this figure includes AIADMK votes. Same logic for other parties holds. The analysis fails to take into account the failure of arithmetic in 2006 assembly polls. After winning 40-nil in 2004, the DMK alliance ought to have translated this into at least 200 seats, but DMK did not even get majority of its own. AIADMK alliance won 69, implying that in an assembly poll, the party still has big clout (In 2009, AIADMK front did not project a PM candidate but still got 12. In terms of pure arithmetic, it should have won much fewer). Another factor author fails to take into account is the narrow margins of victory in 2009 Lok sabha polls, and only a proper segment-wise analysis would be relevant to current situation. As DMDK polled in excess of margin of victory in 135 constituencies,AIADMK alliance needs only a minor swing -- as little as 5,000 extra votes in its favour - in each constituency to sweep this election. Analysis fails to see the all-round dissatisfaction with DMK regime unrelated to 2G scam. Given the price and power situation, never has an election taken place in unhappier circumstances, and an unhappy electorate never votes back the same party. Also not taken into account is the fact that in every election, neutral votes decide the outcome, as difference in committed votes is only five to 10 per cent between two main fronts, whereas neutral voters exceed 25%. Parties are elected to power only because uncommitted voters choose them above their rivals. In the present polls, an overwhelming majority of neutral voters are likely to vote against DMK,which will get only the support of its members, sympathisers and those of alliance parties. My analysis will be -- AIADMK front will get 30% through its vote bank, 10 % through DMDK and Left vote bank and another 10-12 per cent through neutral voters, taking it well beyond the 50% mark. DMK may get the total vote banks of DMK, Cong,PMK and VCK to touch about 40-44%. Analysis fails to take note of schism between DMK and Congress. Congress, which i believe doesn't have an individual vote bank of more than 6-8% is unlikely to transfer its support base to DMK. That is, most traditional Congress supporters are likely to vote only for Congress candidates this time, but against DMK, if a DMK candidate is fielded in their constituency. In fact, some Congress voters may vote against the alliance on issues of corruption, price rise, while some DMK supporters are likely to vote against Congress candidates because of arm-twisting tactics and the DMK's need to contain the aspirations of the Congress.

  2. Thank you so much for a detail reaction. So let me respond:

    1. "Deducting 6.75% from PMK (percentage of votes polled by PMK in 2009 as part of AIADMK front. ) is absurd, as this figure includes AIADMK votes"

    That is of course true. But lets take Chennai South for example which AIDMK (sorry AIADMK) won, how much of it belongs to AIADMK alone and how much to PMK and other alliance partners? Shouldn't AIADMK's vote share be accordingly adjusted if we follow your logic?
    I think in my post I did highlight that total votes polled is not a true indication of instrinsic support in a coalition case. Secondly, contesting in an alliance also means that the seat alloted to partners like PMK is also restricted. In 2009, the PMK contested in 4 constituencies while their sphere of influence extends in several more. Then there is a problem that partners do not get the seat where they are the strongest. Take Kanyakumari or Pondicherry for example where Congress was the strongest by the DMK and PMK grabbed both. So I agree much with you but going by votes poll does give a broad indication since it often more or less reflect their total vote share taken into consideration during seat negotiations.

    2. "After winning 40-nil in 2004, the DMK alliance ought to have translated this into at least 200 seats, but DMK did not even get majority of its own."

    That too is of course true. Lok Sabha and Assembly elections are different. But 2009 being more recent can give a better indication of current vote share than 2006 and we can get around this by examining assembly level vote polled within a constituency.

    3. "DMDK polled in excess of margin of victory in 135 constituencies,AIADMK alliance needs only a minor swing -- as little as 5,000 extra votes in its favour - in each constituency to sweep this election."

    Not if you deduct PMK vote share from those of then AIADMK Alliance and add both PMK & KNMK (who contested independently) into the DMK tally.

    The rest of your observations is just an opinion and opinions can differ and not not reflect ground realities.

  3. @ Anonymous

    My first reaction is that if 2006 (Assembly) was different than 2005 (Lok Sabha) the main factor was DMDK, though it won only one seat. It ate into the votes of both the DMK and AIAMK. These are the vital statistics -

    2006 13th Assembly 2,764,223 1
    2009 15th Lok Sabha 3,073,479 0

    Its cumulative growth in 3 years was around 11% of its 2006 base. Assuming the same rate of growth, the DMDK maximum vote contribution to the AIADMK could be assumed as 10.5% - little change from 2009.

    The article rightly observes the DMDK's vote share comprises of protest votes - anti-Dravidian. As long as DMDK contest independently, it grows but its growth is stymied once it aligns to either of the Dravidian party.

    I am typical supporter of DMDK. I will certainly vote if a DMDK candidate is in my constituency. If there are none, I rather stay at home or note some independent.

  4. By chance I came across this great post. Its a good backgrounder for dummies.

    One of the comments suggested that the Congress supporters may not vote for the Congress and hence DMK votes may come down, enabling Anna DMK to win.

    But if the lead of DMK alliance is over 10% as the post suggest; then it is possible that the Congress may not want such a landslide win as a stronger DMK does not suit the Congress. Only if DMK falls short bu 25 or more seats can Congressman can hope to become Ministers.

    But if the statistics tell its tale what it says is that even if a section of the Congress supporters do not vote DMK candidates, it has enough vote share to sail through but this does not hold vice versa.

    Just sharing the thought of a novice

  5. Excellent post. @Karuna

    You say that choice of a DMDK supporter is confined to abstaining or voting independents. I disagree.

    Let's agree that the DMDK alliance is one of tactics, the long term strategy to oust both Dravidian party with the intermediate strategy to ally with the Congress. If this be the case than this elections DMDK supporters like Congress voters should vote tactically - vote each other when their parties are not standing. The consequence would be that DMK & AIADMK take a hit and their vote and seat share grows. This performance should be the platform for a DMDK-Congress alliance next election.

  6. Whatever you arm chairs analyze, Amma will win. Is this blog in the payroll of the DMK

  7. Karuna's stats interested me. Why did Chiranjeevi merge with the Congress - his handful of MLAs were threatening to defect to the Cobgress and TDP and he knew that next time he had to start from scratch and vote share peaked.

    If Vijay Kant vote share growth is only 11% in three years then, it looks like he is peaked or peaking, explaining his alliance with Jaya.

    The media exaggerates DMDK's defection

  8. One of the comment has stated that why DMK could not win 2006 assembly elections landslide when they swept all 40 seats in 2004 loksabha elections. I think it is due to the fact that MDMK was not part of DMK alliance in 2006 and the important thing is DMDK has formed a new party and taken away 8% and thereby reducing the DMK combines vote share and reducing it to lesser seats in 2006 compare to 2004.

  9. one more think I forgot to tell that it is time that DMDK should have stood seperately or with left and MDMK(even there is chance now) and in that event even if DMK come back to power the AIADMK will be nowhere int he picture, may be with 10 or 15 seats. In that case AIADMK will be disintegrated before the next election and that will be time for DMDK to allign with congress and wrest the state in next 2014 election. I think DMDK should stand seperately now.

  10. @MPD

    I agree. Even now with allies mad at Amma for arbitrarily deciding seats, we see AIADMK allies looking up to Vijaykant. If all allies pull out and assuming AIADMK ties up with BJP & Subramaniam Swamy; even then Vijaykant led Third Front will be just trailing AIDMK vote share.

    People are looking for choices outside the two Dravidian parties and with a 20% vote share (DMDMK + other allies of AIADMK); they can seek to attract 10% from (AIDMK) and 10% from (DMK) - as the momentum will be with them - the two Dravidian poles on the decline. They can either give a tough fight to DMK or even pull off a stunning victory. Even if this does not materialize,DMDK should play for long term - marginalize Amma and for Vijay Kant to emerge as the really challenger to DMK


    for latest election news


Note: Only a member of this blog may post a comment.