District-level data on census phase deletions in Odisha’s Special Intensive Review (SIR) exercise, released by the Election Commission of India on July 6, shows that the state is an outlier compared to other major states. Of the 17 completed census states that have released area-level numbers, all of the largest states show higher deletion rates in more urban areas. However, deletions in Odisha do not show any clear pattern at the district level, especially at the urbanization level of the district.
SIR is a two-stage process. The first phase of the exercise involves ECI distributing census forms and publishing the names of those who returned the filled forms in the draft list. The second phase makes corrections to this draft list after further verification, leading to the publication of the final list. Odisha’s draft list published on July 5 shows that the state’s electoral roll has shrunk by 2.01 million or 6% compared to the pre-SIR list to 31.3 million, the third lowest percentage of deletion after Lakshadweep (2.5%) and Mizoram (5.3%).
Read also: SIR launches in Delhi: 168 lakh forms distributed, 7.6 lakh forms digitized on first day
Highest deletion rate
A summary of the districts deleted in Odisha’s draft list released on Monday shows that the highest deletions were recorded in Ganjam, Cuttack, Mayurbhanj, Jajpur and Balangir, where the electoral list shrunk by 207,624, 155,166, 111,267, 105,014 and 99,258. The lowest deletions were recorded in Deogarh, Boudh, Nopada, Subarnapur and Jharsuguda, where the list shrank by only 11,004, 13,944, 16,829, 19,224 and 28,542. The highest percentage of deletions was recorded in Malkangiri, Balangir, Cuttack, Nayagarh and Ganjam, where the number of rolls reduced by 10.2%, 7.4%, 7.1%, 7% and 6.9%. These areas were at very different levels of urbanization in the 2011 census, with 91.9%, 88%, 72%, 91.7%, and 78.2% of the population rural in the census. Likewise, the districts with the lowest percentage of deletion are Nuapada, Subarnapur, Boudh, Deogarh and Khurda, where the number of voters has shrunk by 3.3%, 4%, 4.1%, 4.4% and 4.5%. The rural share in these regions was 94.4%, 91.8%, 95.4%, 92.8%, and 51.8%, respectively, in the 2011 census.
Clearly, there is no clear relationship between district-level deletions in Odisha and its urban or rural nature. This is in contrast to the trend seen in other states where most urban areas generally had the highest deletion rates, and a possible reason for this may be that rural-urban migrants choose to keep their names on their rural household lists.
Also, deletions at the draft stage have nothing to do with socially vulnerable groups. For example, the percentage of deletion at the district level is not related to the share of Scheduled Tribes (ST) population in the total district population, the share of Scheduled Castes (SC), or their combined share. The ST and SC groups account for 23% and 17% of the population of Odisha statewide.
Read also: The many controversies around SIR
At the AC level, the highest deletions in Odisha in absolute terms were recorded in Malkangiri, Sanakhimundi and Choudwar, where 27,653, 24,622 and 23,828 names were deleted from the lists. Bhubaneswar North, Udala and Sonepore recorded the lowest number of names removed, with only 4,077, 5,314 and 6,188 names removed.
The highest and lowest percentage deletions were at the same ACs as in absolute value. Malkangiri, Sanakhemundi and Choudwar recorded deletions of 11%, 10.5% and 10%. Only 1.4%, 2.4% and 2.6% names were removed in Bhubaneswar North, Sonepore and Udala.
To be sure, the IEC did not publish a summary of voter counts at the Electoral Commission or district level immediately before the start of the SIR exercise. Therefore, Hizb ut-Tahrir used the list as of January 1, 2025 (the latest list of which a summary is available on the website of the Odisha Chief Electoral Officer) to check the district deletions and the percentage of AC in the state. This list included 34.07 million voters statewide, 2% more than the 33.4 million voters on the lists immediately before the SIR exercise. While a 2% change is unlikely to lead to a trend in district-level numbers, it could change the ranks of some districts.
