[This is the first of a three-part guest post series by Rudraksh Lakra, examining delimitation in Assam and the SIR in West Bengal.]
Introduction
In the 2026 Assam Assembly elections, the BJP won 82 seats, while its allies, the Bodo Peoples Front and the Asom Gana Parishad, secured ten seats each, taking the alliance tally to 102 out of 126 Assembly seats. In the last two Assembly elections, the BJP had peaked at around 60 seats. The 2026 election marked an increase of 22 seats. In West Bengal, the BJP secured 207 out of 294 seats (two-thirds majority) despite never having governed the state previously and winning only 77 seats in the preceding assembly election.
Conventional explanations account for part of what happened. In Assam, the BJP successfully linked Hindutva politics with Assamese identity through the slogan “jati, maati, bheti” (ethnicity, land, and hearth) alongside welfare politics and organisational consolidation. In West Bengal, anti-incumbency sentiment, opposition fragmentation, and Hindu consolidation all played an important role. Yet the sheer scale of these victories generated intense controversy. Opposition parties raised allegations that the elections had been manipulated, while the impartiality and institutional independence of the Election Commission of India came under serious scrutiny. This controversy was shaped by two major administrative exercises that preceded the elections. In Assam, the 2023 delimitation exercise redrew constituency boundaries across the state, restructuring the electoral map in ways that significantly altered representational distribution. In West Bengal, the Special Intensive Revision (“SIR”) of electoral rolls led to more than 90 lakhs gross deletions being carried out.
This series of essays examines these developments through the lenses of constitutional law, competitive authoritarianism, and digital structural authoritarianism. The first part focuses on delimitation in Assam, while the subsequent piece will explore the SIR in West Bengal. This essay analyses the constitutional concerns surrounding Assam’s delimitation exercise, the allegations of gerrymandering that followed, and the central role of digital technologies in reshaping electoral representation. The second piece examines how West Bengal’s SIR of electoral rolls produced large-scale exclusion through algorithmic classification, unstable software systems, and procedurally burdensome verification mechanisms. The third piece will place Assam’s delimitation and the Bengal SIR in conversation with one another to develop five broader lessons about the changing relationship between technology, electoral administration, and the conditions necessary for genuinely competitive and fair democratic contestation.
Background to Delimitation
Delimitation plays a key role in representative democracy. At its core, it is the process through which electoral constituency boundaries are redrawn to reflect population changes. This exercise is undertaken because populations shift over time. Cities expand, migration patterns change, districts grow unevenly, and demographic balances evolve. Without periodic readjustment, some constituencies would end up containing far more people than others, resulting in unequal voting power. The constitutional idea underlying delimitation is therefore simple yet fundamental: one person, one vote, linked to political equality. In India, this principle is enshrined in Articles 81 and 82 of the Constitution.
The power of the map lies precisely in its invisibility. Most voters see elections as contests fought through speeches, campaigns, manifestos, and political mobilisation. However, the design of constituency boundaries can predetermine who is likely to win. A constituency map can be shaped to amplify certain communities while diluting others (gerrymandering). It is precisely for this reason that political actors have sought to abuse the process. Gerrymandering can weaken political opponents, fragment communities, consolidate favourable voter blocs, and structurally tilt elections.
Modern gerrymandering is deeply data-driven. Political actors now rely upon advanced technologies, demographic databases, GIS mapping systems, predictive analytics, and algorithmic modelling to engineer electoral outcomes with extraordinary precision. In countries such as the United States, sophisticated software can simulate thousands of electoral scenarios to determine which configuration of boundaries maximises partisan advantage. The Voting Rights Act, 1965, which was designed to prevent racial discrimination in voting practices and electoral procedures, has been progressively diluted by the United States Supreme Court. In the recent ruling in Louisiana v. Callais (2026), Justice Kagan, in dissent, described Section 2 of the Voting Rights Act as having been rendered “effectively a dead letter.”
However, at the same time, technology has also created tools for public transparency. In the United States, applications such as Dave’s Redistricting App allow citizens, journalists, researchers, and advocacy groups to independently analyse constituency boundaries and propose alternative maps. These platforms democratise access to electoral geography and make redistricting processes more transparent. Independent analysts can examine demographic balance, partisan fairness, and potential bias. Public scrutiny acts as a potential safeguard against covert manipulation.
Impact of delimitation in Assam
The Assam delimitation exercise was officially justified as an attempt to protect indigenous Assamese and tribal representation. However, the process altered the state’s electoral map in a manner that disproportionately weakened Muslim political representation. Before delimitation, 35 out of Assam’s 126 Assembly constituencies had Muslim-majority electorates. Following the redrawing of boundaries, that number fell to 20. Around 30 Muslim legislators were elected to the Assembly. After delimitation, that number dropped to roughly 22.
Three classic techniques of gerrymandering can be identified within the Assam exercise: cracking, packing, and stacking.
- Cracking refers to splitting a concentrated voter community across multiple constituencies so that it cannot form a majority in any of them. In Nowboicha, Muslim voters were divided across four neighboring constituencies, diluting their collective influence across seats where they would remain a minority in each.
- Packing involves concentrating a community into a smaller number of constituencies so that their votes become electorally wasted through overwhelming margins. In Barpeta district, the Muslim-majority constituencies of Baghbar and Jania were merged into a single constituency called Mandia with approximately 2.81 lakh voters. Instead of influencing multiple seats, Muslim voters became concentrated into fewer constituencies with excessively high victory margins.
- Stacking involves adding favourable demographic blocs into constituencies to alter their political balance. In Barpeta, panchayat areas with substantial Hindu populations were added to the constituency while Muslim-dominated areas were shifted elsewhere. The result was that the Muslim share of the Barpeta Lok Sabha electorate fell from around 60 percent to approximately 35 percent after delimitation.
One analyst suggested that after delimitation in a hypothetical scenario where the opposition alliance secured 43 percent of the popular vote while the BJP-led North-East Democratic Alliance secured only 42 percent, the alliance could still win around 87 out of 126 seats, amounting to nearly 70 percent of the Assembly. This is because opposition votes became concentrated in a limited number of minorities dominated constituencies, producing overwhelming victory margins that translated into no additional seats.
The 2026 results appeared to validate this model. The NDA lost seats primarily in constituencies where minority voters formed overwhelming majorities. In Central and Lower Assam, where the opposition had won 32 seats in 2021, the total number of seats in the region was reduced from 49 to 44 after delimitation. Reporting suggests that the delimitation exercise directly benefited the BJP-led alliance in at least 19 seats, including five newly created seats in tribal areas, two formerly Muslim-majority constituencies reserved for Scheduled Caste and Scheduled Tribe candidates, and twelve constituencies that underwent major demographic restructuring.
This demonstrated how delimitation can structurally shape electoral outcomes without altering overall vote share. The map itself became politically determinative. The deeper concern is that these redrawn constituencies will continue to shape results across multiple election cycles to come, making it very difficult for the opposition to win. In doing so, the exercise risks undermining the constitutional principle of one person, one vote by reducing the effective value of certain votes through unequal constituency design. It also raises serious concerns regarding the integrity of free and fair elections, because electoral competition itself becomes structurally skewed before the voting process even begins.
Process of delimitation in Assam
Apart from its eventual electoral impact, the process through which delimitation was conducted in Assam itself raised serious concerns. The delimitation exercise in Assam formally began in 2023 under the supervision of the Election Commission of India, following years of discourse surrounding demographic change, migration, and the protection of “indigenous” Assamese communities. Understanding what the process produced requires a framework that looks beyond the intentions of individual actors. My concept of digital structural authoritarianism is useful here. This lens shifts analytical attention away from isolated decisions and toward the operation of institutional and technological systems. Once established, these systems shape incentives, administrative behaviour, and political outcomes through their routine functioning. The Assam delimitation exercise reflects this dynamic at every stage of its implementation.
The draft delimitation notification was released on 20 June 2023, and objections were required to be filed by 11 July 2023. Political parties, civil society organisations, researchers, and ordinary citizens were therefore given barely three weeks to examine one of the most consequential electoral restructuring exercises in Assam’s recent history. The compressed timeline shaped the process itself. Meaningful scrutiny requires time: time to analyse demographic data, examine constituency boundaries, conduct field verification, and organise responses. By the time such scrutiny could begin to take shape, the exercise had already acquired practical finality.
Several opposition parties challenged the delimitation before the Supreme Court of India. The bench comprised Chief Justice D. Y. Chandrachud, Justice J. B. Pardiwala, and Justice Manoj Misra. The petitioners argued that the exercise was arbitrary and discriminatory, and that it violated the constitutional principle of equal representation because constituencies had not been structured around roughly equal populations. They challenged the Election Commission’s methodology, arguing that districts were assigned different average constituency sizes, resulting in deviations of up to 33 per cent between the populations of the largest and smallest constituencies. Population density, the petitioners argued, has no constitutional basis as a criterion for delimitation. However, the Supreme Court declined to stay the exercise and did not hear the matter on the merits in time for the challenge to have any meaningful impact before the elections. This effectively allowed the process to become irreversible, reducing the legal challenge to a largely symbolic exercise. The case has remained pending since 2023. This is judicial inaction as a tactic through which the Executive Court retains the status quo.
The opacity of the process compounded every other concern. Researcher Srinivas Kodali has noted that the Election Commission did not release the machine-readable electoral maps. Constituency boundaries were published only as static PDF documents, without latitude or longitude markers, making independent computational analysis effectively impossible without first reconstructing the boundaries. No public record exists of what algorithm the Commission used, what data inputs drove its decisions, or how competing criteria, population equality, administrative coherence, and geographic contiguity were weighted against one another. The methodology was conducted entirely within the Commission’s internal process, shielded from democratic scrutiny.
Within this opacity forms part of the operation of digital structural authoritarianism itself. Institutional control over data and the formats through which information is released shape the ability of outside actors to examine and contest political outcomes. An institution that possesses structured and analysable electoral data, yet releases it only in forms that inhibit analysis, has made a consequential governance choice. By the time the demographic consequences embedded within constituency boundaries become visible to independent analysts, the constituencies may already have been used to conduct elections.
The character of the boundaries themselves made that opacity more troubling. As the shapes that emerged are not what neutral administrative and representational criteria produce. Several constituencies cut across rivers, fragment villages, and combine geographically disconnected areas into single seats. The Gauripur constituency cuts across roads and waterways in ways that make it physically impossible for a representative to traverse all parts of the seat they are supposed to represent. Mangaldoi is not one place but two geographically separate areas with no continuous border between them. Boundaries like these are not the residue of applying demographic and cartographic logic to a population map. They are what you get when the boundary is an output, when the political result has been decided first, and the line drawn to produce it.
Kodali has also drawn attention to the connection between the delimitation exercise and the National Register of Citizens. The NRC process, completed in 2019, required every resident of Assam to submit extensive personal and family documentation: electoral rolls, land records, birth certificates, bank records, passports, educational certificates, and family linkage papers tracing ancestry to pre-1971 records. In aggregate, these submissions produced a highly granular demographic database of Assam’s population, mapping community identity, family history, and geographic location. Combined with voter registration data, electoral history, and GIS mapping software, this dataset creates the conditions for extremely precise political profiling.
The NRC-delimitation relationship illustrates the cumulative logic of digital structural authoritarianism. Administrative exercises that appear separately justified can, in aggregate, generate interconnected systems of demographic and political management whose effects extend beyond the stated purpose of any individual exercise. Each process generates additional layers of data, which then shape the possibilities available to subsequent exercises. Over time, the infrastructure itself reshapes the conditions under which electoral representation and democratic participation operate.
Conclusion
A trajectory similar to Assam emerged in Jammu and Kashmir following the 2020 delimitation exercise (see here and here). Six additional Assembly seats were allocated to the Jammu region, five of them Hindu-majority constituencies, while the Muslim-majority Kashmir Valley received only one additional seat. In the 2024 Assembly elections, the BJP secured five of these six newly created Jammu seats. Jammu accounts for roughly 44 percent of the population of Jammu and Kashmir, yet received approximately 48 percent of the Assembly segments. Muslims constitute nearly 68 percent of the territory’s population. The redistribution of representation had a clear political effect.
These developments generated substantial concern regarding the now-defeated Delimitation Bill, 2026, which proposed redrawing Lok Sabha and State Assembly constituencies using 2011 Census data. If exercises similar to those conducted in Assam and Jammu and Kashmir were expanded nationally, India could witness a nationwide form of technologically driven gerrymandering embedded within ordinary constitutional procedure.
The Assam experience demonstrates how digital structural authoritarianism operates within electoral systems. The Election Commission functions formally, courts hear constitutional challenges, elections take place on schedule, and results are announced through established democratic procedures. Yet many of the most consequential decisions occur earlier in the process, during demographic data collection, GIS mapping, constituency design, and the structuring of electoral boundaries through methods that remain difficult to publicly examine. Once established, this infrastructure continues shaping electoral competition through its routine operation across multiple electoral cycles.
The debate over delimitation is therefore about far more than constituency boundaries. It concerns the constitutional character of representation, the equal value of political participation, and the integrity of democratic competition. At stake is the question of whether electoral systems remain genuinely open arenas of democratic contestation or evolve into technologically managed political structures where constituency design itself increasingly shapes the horizon of electoral possibility.
