Thinking through UP election results with numbers: Rahul Verma

Guest post by RAHUL VERMA

Here’s a closer analysis of Uttar Pradesh 2012 election results

In an article the Times of India says the Samajwadi Party’s victory in Uttar Pradesh seems to be an even more impressive sweep than the BSP’s 2007 showing, but it’s actually a less comprehensive domination. According to the same article, the SP did not do well in western UP and Bundelkhand. They do not provide any reason for this.

My analysis of the election results data shows that average number of candidates per assembly constituency and average number of candidates per one lakh electorate in west UP and Bundelkhand, was slightly lower than other regions of the state. In west UP and Bundelkhand, the average number of candidates per one lakh electorate was approximately 8.5 and average number of candidates per assembly constituency was approximately 16. Whereas in other parts of the state average number of candidates per one lakh electors was approximately 9.5 and average number of candidates per assembly constituency was 17. This means that in other regions of the state votes were more divided and thus the SP got an edge in terms of winning seats. In the first-past-the-post (FPTP) system with multi-cornered contests, even such small gaps lead to big swings in terms of seats a party can win.

2. In another article, the Times of India said, “Mulayam gets the ‘M’ equation right”. They argue that there are around 140 constituencies in UP where Muslims constitute more than 30% of the population. While in 73 assembly seats the community is between 20-29% of the population, it is over 30% in 67 constituencies. Results of these 140 constituencies show that the SP won 72 seats, more than half in contention. I wondered how did the ‘M’ equation worked here because–

Total Seats in UP = 403
SP won = 225* (Counting Raja Bhaiya’s seat as well)
Winning Percent = 55.8 %

Muslim Dominated Seat in UP according to ToI =  140
SP won = 72
Winning Percent = 51.4 %.

So when the SP’s winning percentage on an average is more than in Muslim dominated seats…. what’s the need for analyzing ‘M’ equation separately?

3. Christophe Jaffrelot in an article in Outlook writes, “…the pundits who kept telling us that she was bound to lose because she was only paying attention to her core constituency, the Dalits, will have to think of some other grouse. The party has lost 27 percentage points among the Jatavs, according to a CSDS post-poll survey (which excludes the last phase of voting). The Jatavs and other SCs (down 15 percentage points) deserted the party, while the Brahmins have remained with it.” In my opinion Prof. Jaffrelot’s argument of shift in Dalit vote (as gauged by the post-poll survey) stands correct only when the gap between the Samajwadi Party and the BSP is 10% – as was predicted by the CSDS post-poll survey. The poll estimated 34% vote share for the SP and 24% for the BSP. The actual difference has come down to 3.2% – as SP has got 29.1%, and the BSP 25.9% respectively. Mathematically, even if there was a shift in Dalit vote, it would be to a much lesser extent than was previously estimated.

4. It’s being written in many newspapers that the SP has gained across communities and the BSP has lost in all. Do we need to be surprised by that? Winners most often gain among all sections, and losers see secular decline across communities.

5. Whose victory is bigger – the SP’s in 2012 or the BSP’s in 2007?

It’s just role reversal this time. The Samajwadi Party is in the same position in 2012 as the BSP was in 2007. Despite winning lesser number of seats this time, the number of seats in which the BSP remained second in 2012 is marginally greater than number of seats in which SP was first runner up in 2007.

6. Did the SP win more seats in 2012 with a close margin than the BSP won in 2007?

One can only draw three conclusions:

  1. The SP in 2012 lost more seats in close fights than the BSP lost in 2007.
  2. Proportionately SP won more seats with bigger margins than the BSP in 2007.
  3. Most of BSP seats were won in 2007 between 5% and 10% vote gap, whereas the SP won more seats in 2012 between 2% and 5% vote gap.

(Rahul Verma is a PhD student in the Political Science department at the University of California, Berkeley. He thanks Aaditya Dar ( for scrapping the raw data from Election Commission’s website. Conversation with Shivam Vij and Aaditya Dar helped him in this analysis.)

From Kafila archives:

See also:

4 thoughts on “Thinking through UP election results with numbers: Rahul Verma”

  1. Extremely interesting, but I think the geographical spread argument made by the first TOI story you linked to (full disclosure: I work for TOI and the team that did that story) is an important parameter in deciding how complete a win is.

    We weren’t able to explain why the SP has fallen behind in western UP and Bundelkhand, but it has certainly happened. I think one of the defining features of Mayawati’s 2007 victory was the sheer dominance everywhere in the state (30% or thereabouts vote share and the #1 position in each phase) and the SP has simply not been able to replicate that. To state the obvious, in a state of such stupendous size, geographical dominance (or the lack of it in this case) says a lot about how emphatic the victory was.


  2. Rahul Verma ignores one basic fact in his analysis. The constituencies where a particular community dominates, most political parties select their candidates from that community. Therefore, even a badly losing candidate garners a few percentages of votes from that community. But the constituencies where this community is significant, but not dominant, and candidates belong to different communities, then the preference of this community becomes vital for the victory. If Rahul Verma’s analysis is applied to BSP, it will be found that this party did not win more than 50% of the SC reserved seats. Based on that analysis, an erroneous conclusion that the majority of SC votes did not go to Mayawati will be reached. I am not sure if pertinent data is available, but the overall percentage of Muslim votes in UP, and the fraction that voted for SP, can only prove or disprove if the TOI analysis is wrong. Similarly, the % of SC votes to Mayawati’s party. In spite of clear majority of BSP in 2007, and massive victory of SP in 2012, the vote percentages clearly indicate that the majority of population did not vote for them. For years to come, the political power in UP will swing between SP and BSP, if the Congress Party is not able to encroach in the Muslim votes and BJP in lower OBC votes. In latter scenario, no single party will win a majority of seats.


  3. firstly a brilliant and very insightful article…….so kudos to the writer………..that said i slightly differ on 2 aspects of the writer’s argument…….which may be a result of not analyzing the complete data as closely as the writer seems to have………..
    firstly, his first point about the western UP and Bundelkhand as different from the rest of UP, which Rukmini also seems to agree, maybe the reason was more in terms of question of how development should be perceived..which is a central political economy question…which was in a way more important than the rest of the state……so although land qustion failed to draw as much attention as other issues prior to election, as it was a year back…….in a way land question and the question of how the agricultural economy can move in the path of industrialization played a role…… the reliance power plant at dadri.including the complete mismanagement and failure of SP in handling the reliance power plant during its tenure and the success of the GNIDA at least on paper under BSP to my mind came to haunt the SP to some extent……one can go on explaining how SP’s failure at dadri was much more than mayawati’s at bulandshahr…the question of development is equally debatable in bundelkhand, the less said the better….i did not understand how a one percent gap in terms of average number of candidates per electors and assembly constituencies between western UP/Bundelkand and rest of UP can create as much difference in dividing votes as the writer argues….maybe a closer look of the writer’s argument in terms of region wise analysis of the same…….i.e analyzing western UP and comparing it with say eastern UP, central, south , north etc rather than the rest of the UP as a single unit might show that such slight variations might also exist elsewhere also……i am assuming this since UP is such a big state with such diverse issues playing in the same election in different parts simultaneousl, so the trend of a uniform figure may be difficult to discern.which might change the argument completely………so may be BSP succeeded where political economy was important, while SP succeeded where it was politics……..that said, i may be wrong since i did not see the data which he is analyzing…………
    secondly, his 5th argument analyzing whose victory is bigger……SP and BSP in the last 2 polls…….his contention that may be BSP’s since the number of constituencies BSP remained second is marginally better than that of SP during BSP’s win….may be refuted by his table showing 224 v/s 206……..and moreover victory can be better analyzed by the wins rather than comparing first and second runners’ up performance……..
    both said…..since i have not seen the data set which he has analyzed……..i guess he is in a much better position to defend his arguments than i am……….


We look forward to your comments. Comments are subject to moderation as per our comments policy. They may take some time to appear.

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s