Bengal Election Results

Forecasts vs Actuals

May 08, 2026

Background

A comparison of my projections for the just concluded elections in West Bengal and the actual results is presented below.

As the above table indicates, my forecasts were incorrect both directionally and in scale. A regional comparison of my forecasts with actual results is presented below.

*- Polling for one seat has been countermanded to May 21. I have removed ‘Others’ from above table who I projected would win 3 (2 in Malda and 1 in Presidency), while 6 have won (5 in Malda and 1 in Presidency)

It is clear from the above that I have been both inaccurate in magnitude and direction in the Presidency and Bardhaman regions. I had mentioned in my earlier writings on West Bengal demographics that both these regions are predominantly Hindu and have pockets of Muslim concentration. A closer examination of my projections in Burdwan, Presidency and Medinipur regions indicates that it is in Hindu dominated seats where my estimates have gone wrong. My estimates for Malda have been inaccurate for other reasons discussed subsequently. Let us start by looking at vote shares.

Vote Shares and Swings

The BJP has experienced a vote share gain of about 8% as shown in table below:

This represents a staggering rise since the BJP has been unable to make any inroads in vote share since 2019 in West Bengal. I did not foresee this, and expected vote share to settle at somewhere around 45-42 in favor of the TMC (half the actual swing that has actually happened). The reasons for the huge swing in favor of the BJP can only be ascertained from post-poll data, that is awaited.

The aggregate data also hides interesting regional variations, as we will see later.

My estimates are off in 87 seats due to this extra swing in favor of the BJP. Details of these are presented in Annexure 1.

Splintering of the Muslim Vote

In the Muslim dominated areas of Maldah and Murshidabad, there has been a splintering of the Muslim Vote in favor of the CPIM, INC and AJUP (Humayun Kabir’s party). I expected the Muslim vote to strongly consolidate in favor of the TMC as I expected that Muslims would perceive the TMC as best placed to defeat the BJP. My forecasts were wrong in 11 seats because of this splintering with the BJP, INC and CPIM gaining 8,2 and 1 seat as a result. (details of these seats are in Annexure 2).

Decline of the CPIM, INC in non-Muslim Areas

The INC, CPIM vote share has dropped overall by 0.3% but this hides regional variation. While INC/CPIM vote share has increased in the Muslim dominated areas of Maldah and Murshidabad, it has shrunk in other areas. I did not foresee this, and expected that their strong pockets of support would still persist. This led to my estimates being incorrect in 5 seats (details in Annexure 3).

Summary

I had projected a TMC win in West Bengal with 184 seats out of 294. However, the TMC was reduced to 80 seats and the BJP won an overwhelming majority of 207 seats.

Most of the variance between my projections and results were in the Hindu dominated seats of Bardhaman, Presidency and Medinipur regions. My estimates were wrong in 87 seats overall (primarily in these 3 regions) because the BJP experienced a swing in vote share of 8% in its favor (greater than the 4% I anticipated).

A split in the Muslim vote in the Malda/Murshidabad regions led to my estimates being inaccurate in 11 seats.

A drop in the votes of the Left Front and INC in their pockets of strength in Hindu dominated areas led to a 5 seat variance between my estimates and actual results.