Monday, May 18, 2009

Analysis of Predictions vs Results

Well, I'm certainly pleased with the Indian election results, mostly because I wanted the Manmohan Singh government to return, but also because my pre-results predictions proved to be remarkably good (see post below for full predictions). Obviously I made some major howlers, but on the whole, I think I called this one correctly. Read on for the analysis and my thoughts on the whole thing.

What did I get right?
1. That Congress would be the largest party and UPA will return to power.
2. Mulayam and Lalu appear set to support the UPA government (albeit from a weaker position than I was expecting).
3. BJP allies fared badly (the exception is Nitish Kumar in Bihar - more on that below).
4. Mayawati was a non-factor. I predicted she would get 26 seats, and she got 21. I'm pretty happy with that, considering there was talk of her getting over 40 seats at one point.
5. The communists lost in both Bengal and Kerala.

On all the high level predictions, I was proved correct. Drill down into the numbers, and you see that the main mistake I made was to be a little too conservative when it came to picking a Congress win. Strangely enough, when you drill down into the states, you find that quite a few of the predictions were off. It seems all the errors even each other out!

Of the 12 big states, the ones I got correct (or close enough that I can legitimately claim that I called the trends correctly) are Gujarat, Kerala (bonus: Shashi Tharoor won), Orissa, UP and West Bengal (the last two with a little reservation - keep reading).

But I also failed to see a large Congress resurgence in Andhra (where my hung assembly prediction totally failed to happen), Madhya Pradesh, Maharashtra, and to some extent, UP, where I think the results match my predictions with this one caveat. There was also Rajasthan, where I predicted a small swing towards Congress, when in fact it ended up being a Congress landslide.

I also missed the trends in the two states that definitively went against the national grain - Karnataka and Bihar, where the NDA did surprisingly well. Oh, and of course, Deve Gowda won three seats in Karnataka, so we still have him and his absurd son to kick around.

And finally, there's Tamil Nadu, where I expected Jayalalitha to catch up to Karunanidhi, but instead he came out with twice as many seats as her.

On the issue of West Bengal, I got the trend correct (that the commies would falter), but failed to predict the beneficiary - Mamta. I'm still going to put this one in my correct column because Mamta will be supporting the Congress government, so the upshot is the same.

So, what are the lessons (or "take-aways" as I used to say in my previous, corporate incarnation)?
1. It's all regional. Every party that won or lost did so because of how they have governed in their states. For example, Nitish, YSR, DMK, Patnaik all benefited from their recent positive track record in power in their states. The Communists (both in Bengal and Kerala), and the BJP in Rajasthan were seeking votes in states where their local governments had fared particularly badly in recent years. The Thackeray family fight in Maharashtra benefited the Congress, just as the honeymoon period of the Yediyurappa government helped the BJP in Karnataka. Mayawati and Mulayam's mis-governance in UP over the last decade hurt both of them.

It's all regional.

2. To the extent that there is any "national" trend or lesson, it's not about substance but a more fuzzy question of "national vision". What I mean by this is that when voters think beyond their own constituency and state, they focus more on which party offers a more compelling vision for India (or if you prefer, definition of India) rather than specifics (for which they turn to recent local history). On this count, the BJP failed to provide a compelling vision of India to the voters, and this further aided the Congress resurgence.

3. The following is a pet theory of mine that I can't back up with any evidence, but I think it fits the story. The national elections of 2009 can be viewed as a re-playing out of last year's confidence motion in parliament on a national scale. If you remember, even then, parties horse-traded and voted according to their local concerns (ex. Mulayam was scared of Mayawati and looked to fend her off by moving closer to the Congress) and the larger national vision was lost in the process. But somehow, the parliament muddled through and reached the correct decision. In this election, something similar happened, with the added bonus that everyone who proved themselves to be cravenly opportunistic politicians during the nuclear deal and confidence vote saga lost big time. In fact, support for the confidence motion was possibly the best indicator for how well parties did in this election.

And, now I'm sure you're asking whether I have any advice for our wonderful Parties. Why, of course! Here goes:

To the BJP - focus on regional government, try a little less cynical opportunism (see: nuclear deal) and tone down the Hindutva nonsense.

To Mayawati, Mulayam, Commies, Shiv Sena, Deve Gowda etc. - Please leave politics.

To Congress and UPA - don't screw it up!

Now, where's my TV contract?

Monday, May 11, 2009

2009 Indian Parliamentary Elections Predictions

Right, so the elections are coming to a close and we find out on May 16th what the results are. I've decided (in a significant bout of folly) to predict the unpredictable and to call the Indian election results before they're announced. I've read the tea leaves, consulted the stars, ruffled through wikipedia and a few websites of Indian news organizations, and have come up with my very own set of predictions. 543 seats being contested, in 28 States and 7 Territories, with over 30 political parties, all kinds of crazy characters, and very high stakes. This is a little long, so I'll post the predictions up top and discuss the thinking behind this and state-by-state analysis further down. Feel free to pull this list out on the 16th and ridicule me for getting it completely wrong!

FINAL TALLY
UPA - 257
NDA - 156
3rd Front - 108
Other - 22

PARTY - NUMBER OF SEATS (ALLIANCE AFFILIATION)
1. Congress - 179 (UPA)
2. BJP - 128 (NDA)
3. Commies - 36 (3rd Front)
4. Mulayam - 27 (UPA)
5. Mayawati - 26 (3rd Front)
6. Lalu - 19 (UPA)
7. Chandrababu Naidu - 15 (3rd Front)
8. Nitish Kumar - 13 (NDA)
9. Karunanidhi - 13 (UPA)
10. Jayalalitha - 11 (3rd Front)
11. Naveen Patnaik - 11 (3rd Front)
12. Sharad Pawar - 9 (UPA)
13. Shiv Sena - 8 (NDA)
14. Chiranjeevi - 6 (Other)
15. Raj Thackeray - 6 (Other)
16. Akalis - 5 (NDA)
17. Ram Vilas Paswan - 4 (UPA)
18. Omar Abdullah - 4 (Other)
19. Telengana - 3 (3rd Front)
20. Ramdoss - 3 (3rd Front)
21. Vaiko - 3 (3rd Front)
22. Mamta Banerjee - 2 (UPA)
23. Shibu Soren - 2 (UPA)
24. Ajit Singh - 2 (NDA)
25. Independents - 2 (UPA)
26. Independents - 6 (Other)


Phew! You'll notice I've used the names of politicians to denote the parties they lead. This is largely because too many acronyms will get too confusing, and also because most of these parties are glorified personality cults of their leaders anyway. The tallies above are based on a state-by-state analysis, which made sense because the national election is really a collection of separate regional elections. I'm including my state-by-state analysis below as well.

Well, the high level analysis is that we will see a return of the UPA government at the center. There are a few assumptions I'm basing this on. The first is that Mulayam will support the UPA. If he doesn't, then the equations will change all over again, but I have a feeling he'll stick with it. The second is that while the BJP will not be hurt very much, it's allies will virtually disappear. Third is that the Congress will improve its tally, continuing the BJP's stint in the wilderness of opposition. Fourth is that the Mayawati phenomenon will not materialize as expected. She will gain a few seats, but it won't be a dramatic gain, and her appeal will not translate outside UP. And fifth, the Communists will lose ground in Kerala and Bengal, but the 3rd Front will cross a hundred seats because of the gains of newly minted allies like Mayawati, Jayalalitha, Naveen Patnaik and Chandrababu Naidu.

One other major consideration is the confidence vote which took place in the Lok Sabha last year. The vote was very close, and so it will be again if the 3rd Front and NDA gang up again to avoid a UPA government. I doubt that will happen, because I think the UPA will be able to entice some of the Other parties (like Omar Abdullah's) as well as some constituent members of the 3rd Front and NDA to create the government.

The last point on a general national level is that of the issues. The major issues affecting India today, the threat of terrorism and the global economic crisis were rarely discussed on the campaign trail. The US-India Nuclear Deal and the subsequent confidence motion that the government barely survived in parliament are also not discussed nearly as much as they should be. Instead, this campaign was dominated by complaints of personal slights, random inflammation of religious hatred and its twin, over-reaction by the state apparatus, the ridiculous sight of pampered film stars in designer clothes urging dhoti and sari clad voters to blindly follow their recommendations and the prospect (as a very close friend put it to me) of "an 81 year old and a 76 year old vying to lead a country where the median age is 25." The media, as expected, has been weak, shameful, shrill, excessive and unwilling and unable to mediate a grand national debate. More money has been spent on this election then ever before, and I'm not sure if the issues could have been more ill-served than they were. One thing is for sure: nobody, not the politicians, not the parties, not the media, nor (heck, I'll say it) the electorate have lived up to the grandness of the occasion.

Below is my state-by-state analysis. It's in order of size of state, with a little blurb of explanation where I felt it was required. I'll stand by the predictions in the first 12 states on this list (ie. every state with 20 seats or more). These top 12 states account for 440 of the 543 total seats being contested. After that, what was an exercise that was mostly guess work became an exercise in complete guess work, so keep that in mind. Righto!

1. UTTAR PRADESH (80 Seats)
Mulayam - 26
Mayawati - 24
Congress - 14
BJP - 12
Ajit Singh - 2
Independent - 2

So here I'm predicting gains for Mayawati and Congress, and losses for Mulayam. I think that makes basic sense, so Mulayam goes from currently holding 35 seats to 26 seats, and Mayawati goes from 19 to 24. Mulayam still comes out on top, and with an alliance with Congress can fend off Mayawati and the BJP.

2. MAHARASHTRA (48 Seats)
BJP - 11
Congress - 10
Sharad Pawar - 9
Shiv Sena - 8
Raj Thackeray - 6
Mayawati - 2
Mulayam - 1
Independent - 1

This was quite possibly the hardest one to pick. So I'll start with the Independent. I'd like Meera Sanyal to win, so I've put her down as the winner in Bombay South (with the expectation that she will support the UPA). Congress, BJP, Shiv Sena and Sharad Pawar have all basically been evenly footed the last several years in the state, but with the split within the Thackeray family, and the heavy investment Mayawati is making in the state, it's all completely up in the air. My notes for Maharashtra have the most number of crossed out tallies than any other state.

3. ANDHRA PRADESH (42 Seats)
Congress - 16
Chandrababu Naidu - 15
Chiranjeevi - 6
Telengana - 3
Commies - 2

Another very difficult state to call. They're concurrently holding elections for the state legislative assembly, so that means that this vote is a referendum on the YSR led Congress government. Usually AP swings in a big way one way or the other. This time however, with the addition of Chiranjeevi in the fray I think Naidu won't be able to capitalize entirely on the expected swing, leaving Congress at the top, but not by much. Expect a hung assembly as well, where Chiranjeevi will play a much bigger spoiler role.

4. WEST BENGAL (42 Seats)
Commies - 20
Congress - 20
Mamta- 2

Major loss of face for the Communists is predicted here. They've pissed off secularists and internationalists by allying with the NDA to try to topple the government over the nuclear deal, and they've pissed off local leftist agitator types by supporting industry in Singur and Nandigram.

5. BIHAR (40 Seats)
Lalu - 19
Nitish Kumar - 13
BJP - 4
Ram Vilas Paswan - 4
Congress - 0

Gains for Nitish Kumar in the wake of the successful start to his Chief Ministership, but Lalu maintains enough seats to stay relevant.

6. TAMIL NADU (39 Seats)
Karunanidhi - 13
Jayalalitha - 11
Congress - 8
Vaiko - 3
Ramdoss - 2
Commies - 2

Jayalalitha makes some gains, but mostly at the expense of the Congress.

7. MADHYA PRADESH (29 Seats)
BJP - 26
Congress - 3

This traditional BJP bastion stays that way.

8. KARNATAKA (28 Seats)
BJP - 10
Congress - 18
Deve Gowda - 0

You may have noticed that Deve Gowda is missing from my total tally up top. That's because I have him and his party getting the sound thrashing they completely deserve. BJP will also lose seats, and Congress will make a remarkable come back in this state. You heard it here first folks!

9. GUJARAT (26 Seats)
BJP - 16
Congress - 10

The trend of the greater BJP-fication of Gujarat continues, much to my displeasure.

10. RAJASTHAN (25 Seats)
BJP - 16
Congress - 9

Congress won recent state assembly elections here, following a few years of caste based violence, and a poor governing record on the part of the BJP's Vasundhara Raje Scindia. BJP is still strong, but the Congress will pick up a few seats here.

11. ORISSA (21 Seats)
Naveen Patnaik - 11
Congress - 6
BJP - 4

Recent communal violence, as well as being ditched as an ally by Patnaik means that the BJP will lose ground here. Patnaik is an odd fish, because he just left the NDA, and has given temporary allegiance to the 3rd Front. If, as I predict, his position in the Lok Sabha doesn't change (he keeps the 11 seat tally he currently holds) then his party will also be a good candidate for the UPA to try and woo.

12. KERALA (20 Seats)
Congress - 11
Commies - 9

Another thumping setback for the Commies, and Shashi Tharoor will join the Indian Parliament as a Congress MP.

That's the last of the confident predictions (or as confident as they can possibly be). Below are the rest of the states and territories (the final 103 seats in the Lok Sabha).

13. ASSAM (14 seats)
Congress - 11
BJP - 2
Independent - 1

14. JHARKHAND (14 Seats)
Congress - 7
BJP - 4
Shibu Soren - 2
Commies - 1

15. PUNJAB (13 Seats)
Congress - 6
Akalis - 5
BJP - 2

16. CHHATTISGARH (11 Seats)
BJP - 9
Congress - 2

17. HARYANA (10 Seats)
Congress - 9
BJP - 1

18. JAMMU & KASHMIR (6 Seats)
Omar Abdullah - 4
Congress - 1
BJP - 1

19. UTTARKHAND (5 Seats)
BJP - 3
Congress - 2

20. HIMACHAL PRADESH (4 Seats)
Congress - 3
BJP - 1

21. ARUNACHAL PRADESH (2 Seats)
BJP - 2

22. GOA (2 Seats)
BJP - 1
Congress - 1

23. MANIPUR (2 Seats)
BJP - 1
Congress - 1

24. MEGHALAYA (2 Seats)
Congress - 2

25. TRIPURA (2 Seats)
Commies - 2

26. MIZORAM (1 Seat)
Mizo National Front (counted as Independent) - 1

27. NAGALAND (1 Seat)
Naga People Front (counted as Independent) - 1

28. SIKKIM (1 Seat)
Sikkim Democratic Front (counted as Independent) - 1

29. DELHI (7 Seats)
Congress - 5
BJP - 2

30. ANDAMAN AND NICOBAR (1 Seat)
Congress - 1

31. CHANDIGARH (1 Seat)
Congress - 1

32. DADRA AND NAGAR HAVELI (1 Seat)
Bharatiya Navshakti Party (counted as Independent supporting UPA) - 1

33. DAMAN AND DIU (1 Seat)
Congress - 1

34. LAKSHADWEEP (1 Seat)
Congress - 1

35. PONDICHERRY (1 Seat)
Ramdoss - 1

If you've made it this far, then hats off to you! I think the numbers speak for themselves. We'll find out soon enough. Cheerio!