At Macro Level, Voters Match Population Projections. But There Are Vast Disparities Across States

On the directions of Election Commission of India (ECI), a special intensive revision (SIR) exercise is being carried out in some states in line with Article 326 of the constitution. 

The ECI has the  motto that no eligible voter is to be left out, and no ineligible person is to be included in the electoral rolls. However, the ongoing SIR continues to attract criticism regarding the procedures adopted. Before every parliamentary or assembly election, voter lists are revised and updated. Yet, the current SIR resulted in the deletion of 47 lakh voters in Bihar. We do not know how many of them would have voted in parliament elections in 2024. 

In Uttar Pradesh, 2.89 crore voters included in the current voter lists are at risk of deletion.  Such a large scale of possible deletions raises the issue of whether the earlier electoral rolls used for parliamentary elections (2014, 2019, and 2024) and several assembly elections were distorted to that extent. 

In this article, we attempt to understand whether large-scale deletions from the pre-existing (i.e. before SIR 2025-26) voter lists are mainly due to duplication, deaths, migration etc. or if there is a probability of large-scale omission of eligible voters as a result of  the current SIR.  We do this by analysing Elector Recorded Percentage (ERP) – taken as the number of electors recorded in the electoral roll as percentage of the 18+ population estimates.

For this analysis, the estimated state-wise population of eligible voters i.e. 18+, for the year 2021 and 2026 has been taken from Table T-17 of Report of the Technical Group on Population Projections for 2011–2036 (published by Ministry of Health and Family Welfare). The estimates for the requisite interim time-points have been interpolated by applying State-wise compounded annual population growth rates. It is germane that all the vital factors like mortality, fertility, migration etc. are duly incorporated in these projections by the Registrar General and Census Commissioner. The number of electors in the general elections 2024 has been taken from the ECI portal. 

A caveat here is necessary. While the number of electors is an actual count, the 18+ populations are estimates from an expert committee headed by the Registrar General. Of course, the accuracy of population projections can be questioned. However, we feel that as the cohort of first-time voters becoming eligible for registration in 2024 were born five years before the last census conducted in 2011, 18+ population projections are likely to be quite reliable. These projections are only likely to be affected due to significant changes in the mortality rates during the interim period. There are no known factors supporting any such eventuality, if we go by the number of deaths in COVID-19 as per the Union government’s data. Moreover, regarding net inter-state migration, it is pertinent that the same has been assumed to remain constant as observed during 2001-11. Deviation therefrom, particularly such as observed during COVID-19, would obviously impact statewise ERPs (not that of all-India ones) to some extent, albeit in our view to a very small. 

Coverage in various states

The table below gives ERPs in respect of 21 major states for the year 2024. ERP of 100% indicates complete coverage, having no duplication or omission of voters. It would also mean almost correct up-dation of voter lists for deaths/migration etc. 

As it is not feasible to keep the rolls perfectly updated, ERP in practice is expected to be a little more or less than 100%, depending upon the difference between the number of omissions due to non-registration of eligible voters and the extent of duplication as well as non-updation due to mortality etc.

 Table: Number of electors and estimated 18+ population (in 2000) in 2024

*State-wise data on ECI portal excluded Surat constituency since a candidate there was elected unopposed. Requisite number of electors added from the all-India figures in the December 26, 2024 press release.

Findings

At the all-India level, ERP is 99.5% indicating that the extent of missing voters may largely be equal to duplicate voters or unrecorded deaths. 

At the state level, in respect of 16 major states, 2024 ERPs between 95-105% are likely to be within the realm of possibility and a small departure from 100% can be attributed to a little inaccuracy in projected population due to non-constant inter-State migration, duplicate voters or non-deletion of deaths etc. 

We find that ERPs for Telangana (115.8) and Karnataka (108.1) are remarkably high. The issue of great concern however is ERPs for Haryana (92.9), Punjab (91.5) and J&K (87.7), indicating that even in the 2024 rolls, a large number of eligible voters may not have been included. Therefore, in these states, an intensive revision of electoral rolls would be justified and it should result in substantial net inclusion of voters.

Before SIR 2025-26, the number of total electors in Bihar was 7.89 crore (on June 24, 2025), which was updated to 7.24 crore (on August 1, 2025) in the draft roll, and finally became 7.42 crore (on September 30, 2025). In the process, its ERP was reduced from 97.7% in 2024 electoral rolls to 90.3% after SIR. This fall in the coverage of electors vis-à-vis the eligible population could be an indication that almost 10% of the eligible population may perhaps be out of coverage as a result of SIR. In the absence of digital lists after SIR, it is difficult to make a detailed assessment of reasons for each constituency.

For other states, we depend on media reports. Taking a cut-off of January 1, 2026, in Uttar Pradesh and Gujarat, ERPs are seen to be merely 77.9% and 81.7% respectively against 99.1% and 96.7% in the 2024 electoral rolls. It implies that around 22% and 18% of the eligible population may fall out of coverage in these states. Similarly, significant declines in ERPs are observed in many other states: Tamil Nadu from 105.3% to 90.5%, Chhattisgarh from 99.8% to 86.4%, Kerala from 101.8% to 91.8%, West Bengal from 101.2% to 92.1%, Madhya Pradesh from 97.6% to 88.5% and Rajasthan from 98.2% to 89.3%.

In our view such significant decreases are not likely to be due only to inaccurate population projections of RGI. The most worrisome issue is not only the fact that ERP is decreasing in all the states, but also that the ERPs are much lower than 100% in all these states. 

Fundamental reason for considering these ratios a cause of concern is that if such is the omission at states or Union territories overall, at disaggregated levels – for instance, constituency-wise – coverages are likely to be much more skewed. Release of granular data in digitised form will enable a deeper examination of missing voters and its reasons.

Conclusion 

In all the states where SIR is either complete or draft electoral rolls released, there are clear indications that a considerable number of eligible voters are left out of the electoral rolls. Likely disenfranchisement of such a large percent of electors belies the ECI’s mandate and it may be a big blow to Indian democracy. A major reason seems to be treating the 2002-04 SIR as sacrosanct. Issue of notices to several well-known persons in the country shows that treating 2002-03 voter lists as sacrosanct is a grave error of judgement by the EC. The solution lies in complete digitisation of the process and the release of the same in public domain.

Sanjay Kumar retired as Additional Director General of the Ministry of Statistics & Programme Implementation.

N. K. Sharma retired as Director General from the Ministry of Statistics & Programme Implementation. 

Siraj Hussain is former Union Agriculture Secretary.

The Wire is now on WhatsApp. Follow our channel for sharp analysis and opinions on the latest developments.