Illegal Cross-Border Migration and India’s Security

Few questions in India’s neighbourhood policy are argued with more heat and less evidence than cross-border migration. The subject sits where demography, citizenship law, federal politics and diplomacy meet, and each field pulls the argument a different way. What makes it a question of regional cooperation is a single mechanism: India’s attempts to count, document, detain and expel undocumented residents convert an internal administrative problem into a bilateral grievance.

The vocabulary the political debate collapses

The whole topic turns on distinctions that public argument routinely erases. A note that does not fix them first will reproduce the confusion it is meant to analyse.

• A migrant moves voluntarily, on a calculation about work, land, income or family, and may or may not do so lawfully.
  • Irregular or undocumented migration describes the manner of entry or stay, not the motive — a person may enter lawfully and overstay, or enter clandestinely for entirely economic reasons.
  • The term “illegal migrant” carries a precise Indian statutory meaning under Section 2(1)(b) of the Citizenship Act, 1955: a foreigner who entered without valid travel documents, or who entered with them and stayed beyond the permitted period.
• A refugee is displaced, not self-moved, and the displacement is what defines the category in law.
  • The 1951 Refugee Convention defines a refugee by a well-founded fear of persecution on five enumerated grounds — race, religion, nationality, membership of a particular social group, political opinion — and by being outside the country of nationality.
  • The 1967 Protocol removed the Convention’s original European and pre-1951 limits, universalising the definition without widening its grounds.
  • The definition is therefore narrow by design: it does not cover people fleeing generalised violence, war, famine or environmental collapse.
• An asylum-seeker is a person whose refugee claim has been made but not yet determined — a procedural status, not a lesser one.
  • The Convention grants a right to seek asylum, not a right to receive it; states declined to accept an unconditional obligation to admit.
• A stateless person is one whom no state recognises as a national under the operation of its law.
  • The Rohingya are the paradigm South Asian case, stripped of Burmese nationality by the 1974 Emergency Immigration Act and definitively excluded by the 1982 Citizenship Law.
  • Statelessness is what makes deportation practically impossible: no state is obliged to receive a person it does not acknowledge as its own.
• An internally displaced person (IDP) is forcibly displaced but has not crossed an international frontier, and therefore remains within the jurisdiction of the government that may itself be the persecutor.
  • The UN Guiding Principles on Internal Displacement place primary responsibility on the national government, and there is no agency with a protection mandate for IDPs comparable to UNHCR’s for refugees.
  • Francis Deng‘s formulation, on which the Guiding Principles rest, deliberately keeps the list of causes open — armed conflict, generalised violence, human rights violations, natural and man-made disasters.

Why the distinction between forced and voluntary movement does the real work

• Displacement is forced migration; migration proper is a conscious decision taken under constraint but not under coercion. The two produce different legal claims and different political responses.
• The distinction is analytically clean and empirically blurred. Mixed flows are the norm: the same boat, bus or riverine crossing carries persecuted minorities, economic migrants and trafficked persons together.
• States exploit the blurring in both directions — receiving states reclassify refugees as economic migrants to avoid obligation, and claimants do the reverse.
• Migration is a process, not an event. Displacement recurs across a lifetime and across generations, which is why cut-off dates in citizenship law are administratively convenient and sociologically arbitrary.

Category	Defining criterion	Governing instrument	India’s position
Migrant (irregular)	Voluntary movement; unlawful entry or stay	No treaty; the 2018 Global Compact is non-binding	Governed as a foreigner; liable to detention and expulsion
Refugee	Well-founded fear of persecution on five grounds	1951 Convention and 1967 Protocol	Not a party; no domestic refugee law; treated as executive discretion
Asylum-seeker	Claim made, not yet determined	Convention procedure; UNHCR mandate	No statutory status-determination procedure; UNHCR determines status for some groups
Stateless person	Recognised as a national by no state	1954 and 1961 statelessness conventions	Not a party to either; no statelessness determination procedure
IDP	Forced, but within national borders	UN Guiding Principles — non-binding	No national resettlement and rehabilitation law of general application
Climate-displaced	Movement driven by environmental change	No category exists in law	No recognition; the gap is unfilled

Push, pull and the chains that make migration self-sustaining

• Push factors operate in the sending society — land scarcity, riverbank erosion, cyclone and flood loss, unemployment, persecution of minorities, and in Myanmar’s case civil war.
• Pull factors operate in the receiving society — wage differentials, demand for informal labour in construction, brick kilns, domestic work, fisheries and waste handling, and the near-absence of employer sanction.
• Chain migration is what converts a wage differential into a durable flow: the first arrivals lower the cost and risk for those who follow by supplying housing, contacts, work and documents.
  • This is why enforcement at the frontier alone is weak. The network, not the border, is the operative variable once a corridor is established.
  • It also explains spatial concentration — clustering produces visible local demographic change out of nationally modest numbers, and visibility is what generates politics.
• Myron Weiner‘s “politics of scarcity” names the mechanism that turns arrival into conflict: competition for education, employment and public resources sharpens dividing lines that were previously latent.

India’s refugee policy without refugee law

India has hosted very large refugee populations for seven decades and has never legislated for them. That is a deliberate position, not an oversight, and everything distinctive about Indian practice follows from it.

The statutes that actually govern

• India is not a party to the 1951 Refugee Convention or its 1967 Protocol, and has no domestic refugee or asylum statute. Neither is any other South Asian state a party.
• Until 2025 the field was governed by three colonial-era statutes plus one modern one, all of which regulate foreigners and know nothing of refugees:
  • The Passport (Entry into India) Act, 1920 — power to require documents for entry.
  • The Registration of Foreigners Act, 1939 — registration and reporting.
  • The Foreigners Act, 1946 — the operative instrument: near-plenary executive power to restrict movement, order departure and detain, with the burden of proving nationality placed on the person under Section 9.
  • The Immigration (Carriers’ Liability) Act, 2000.
• The Immigration and Foreigners Act, 2025 repealed and replaced all four and came into force on 1 September 2025.
• The Foreigners Order, 1948 and the Foreigners (Tribunals) Order, 1964 supply the machinery — refusal of entry, restriction of movement, and the quasi-judicial tribunals that determine nationality.
• Constitutionally, the subject is central: Union List Entry 17 covers citizenship, naturalisation and aliens, and Entry 19 covers admission into, and expulsion from, India.

What non-citizens can and cannot claim

• Articles 14 and 21 apply to “any person”, not merely to citizens, and therefore protect a foreigner’s equality before the law and life and personal liberty.
  • NHRC v State of Arunachal Pradesh established the point decisively for the Chakma refugees from East Pakistan, holding that the state was bound to protect their life and liberty against forced eviction.
  • The Gujarat High Court, in the case of Iraqi nationals, read non-refoulement into Article 21 where the person’s presence was not prejudicial to national security — a reading no Supreme Court bench has adopted as binding.
• Article 19(1)(e) — the right to reside and settle in any part of India — is available only to citizens, and this asymmetry does most of the work in the Rohingya litigation.
• Non-refoulement — the prohibition on returning a person to a territory where life or freedom is threatened — is widely regarded as customary international law binding regardless of treaty ratification.
  • India’s consistent official position is that it is not bound by Article 33 of a convention it has not joined, and that expulsion of persons without valid documents follows domestic law.
  • India nevertheless sits on UNHCR’s Executive Committee and permits UNHCR to conduct status determination for some nationalities in Delhi — an acknowledgement in practice of what is denied in principle.

The consequences of the void: ad hoc, group-specific, reversible

• Because no statute defines who qualifies, the executive decides group by group — and what an executive grants, an executive may withdraw.
• Treatment has therefore been strikingly unequal across groups, with the differences tracking foreign-policy calculation and domestic political cost rather than any test of persecution.
• The absence of a status-determination procedure means an individual cannot establish a claim; only membership of a favoured group confers protection.
• Reversibility is the sharpest cost. A Tibetan or Sri Lankan Tamil enjoys documented long-term residence by administrative decision, not by right, and the order that grants it can be withdrawn.

Group	Period	Legal basis of stay	Documents	Outcome
Tibetans	1959 onwards	Executive; Tibetan Rehabilitation Policy, 2014	Registration Certificates, identity certificates for travel	Settlements, schools, an exile administration; most generous treatment
Sri Lankan Tamils	1983, 1989, 1995, 2006 waves	Executive; state-run camps	Camp registration; state relief	~58,000 in camps in Tamil Nadu; large organised repatriations in 1987–89 and 1992–95
Chakmas and Hajongs	1964–69	Settled in Arunachal by the Union	Contested	Supreme Court directions on citizenship still not fully implemented
Afghans	1980s; 1992; 2021	UNHCR determination; long-term visas	UNHCR cards; long-term visas	Sikh and Hindu Afghans naturalised; others in protracted limbo
Rohingya	2012, 2017 onwards	None — classed as illegal migrants	UNHCR cards not recognised by government	Detention, deportation orders, litigation
Bangladeshis	Continuous	None — classed as illegal migrants	None	Tribunals, detention, push-backs

The three durable solutions, and which of them India uses

• International practice recognises three ways a refugee situation ends, in a settled order of preference, and India’s record maps onto them unevenly.
  • Voluntary repatriation — return to the country of origin by free choice — requires the causes of flight to have been removed, a credible amnesty, and resettlement support on return.
  • Local integration in the country of asylum, which the 1951 Convention urges through naturalisation.
  • Resettlement in a third country, the least preferred and most expensive, used for compelling humanitarian reasons.
• India has used the first extensively and the second selectively, and has never been a resettlement country.
  • For Rohingya and Bangladeshis, none of the three is available: return is blocked by non-recognition, integration is politically refused, and resettlement is not on offer.

India’s refugee record is generous in volume and arbitrary in principle: what determines protection is not the strength of a claim but the identity of the claimant’s group.

The Supreme Court and the Rohingya orders

• In Mohammad Salimullah v Union of India, refugees detained in Jammu sought to restrain deportation. The Court declined to bar it, holding that removal could proceed if the prescribed procedure were followed.
  • The reasoning turned on Article 19(1)(e) being a citizens-only right, while accepting that Articles 14 and 21 extend to all persons.
  • The Court acknowledged the argument from non-refoulement but did not hold India bound by it.
• Subsequent orders have followed the same line. In May 2025 a bench declined interim relief, finding no prima facie basis for allegations that UNHCR card-holders had been deported at sea.
• In July 2025 the Court agreed to examine the underlying question — whether Rohingya are entitled to be declared refugees and what rights follow — which remains undecided.
• The doctrinal position is therefore stable and uncomfortable: constitutional protection of life, liberty and equality for non-citizens coexists with an unfettered executive power of expulsion.

The corridors, each with its own logic

Treating “cross-border migration” as one phenomenon is the commonest analytical error in this field. India’s frontiers with Bangladesh, Nepal and Myanmar generate movements that differ in legality, drivers, scale and the instruments available to manage them — and the maritime and Himalayan arrivals differ again.

Bangladesh: the long border and the deep history

• The 4,096 km India–Bangladesh boundary is the longest India shares with any neighbour, running through West Bengal (2,217 km), Tripura (856 km), Meghalaya (443 km), Mizoram (318 km) and Assam (263 km).
• Its geography defeats conventional border management.
  • Long riverine stretches and shifting char islands make a fixed line meaningless; the Brahmaputra–Padma system moves the frontier physically.
  • Settlement and cultivation run up to the zero line, and the 150-yard construction restriction under the 1975 border guidelines leaves gaps that cannot be fenced.
  • The enclaves — 111 Indian enclaves inside Bangladesh and 51 Bangladeshi enclaves inside India — were a standing anomaly until the Land Boundary Agreement of 2015, given effect by the 100th Constitutional Amendment.
• The historical layer is decisive and is what distinguishes this corridor from every other.
  • Partition in 1947 divided a single Bengali linguistic and economic space, and movement across it continued as it always had.
  • The 1971 war produced roughly ten million refugees into India, the largest single displacement in the subcontinent since Partition, most of whom returned.
  • Successive waves followed communal violence and dispossession, so the flow has always mixed persecuted minorities with economic migrants — precisely the mixture the legal categories cannot handle.
• Economic pull was the classical explanation, and it has weakened. Bangladesh’s per capita income converged with India’s over the 2010s, which removes much of the differential that the standard account assumes.
  • The pull that persists is sectoral and informal — demand for cheap labour in specific Indian industries, which no income convergence dissolves.
  • The post-2024 political turmoil in Bangladesh has reintroduced a displacement driver that the economic account cannot capture.
• Fencing is incomplete after four decades. About 3,232 km of the border had been fenced by early 2025, leaving roughly 864 km unfenced, of which some 175 km is officially classed as non-feasible.
  • The obstacles the Union government itself cites are marshland and landslide-prone terrain, land acquisition delays, short working seasons, and objections from Bangladesh’s border guards.

Nepal: an open border is not illegal migration

• The 1,751 km India–Nepal border is open by treaty, and getting this right is essential to the argument.
  • The Treaty of Peace and Friendship, 1950 grants nationals of each country, on a reciprocal basis, national treatment in residence, property, participation in trade and commerce, and movement.
  • Nepali movement into India is therefore lawful, and describing it as illegal migration is a category error.
• The security problems on this frontier are real but different in kind — they are problems of unregulated rather than unlawful movement.
  • Human trafficking, especially of women and girls, for which the open border and the absence of documentation checks are the enabling conditions.
  • Fake Indian currency notes and gold routed through Nepal, exploiting the absence of customs scrutiny on persons.
  • Third-country infiltration — nationals of other states, including operatives handled from Pakistan, using Nepal as an entry corridor into India, which is the genuine counter-terrorism concern.
• Management is by policing rather than restriction — the Sashastra Seema Bal and Nepal’s Armed Police Force, with joint patrolling and district coordination meetings.
• Proposals to regulate the border are politically impossible on both sides — in Nepal because national treatment is the treaty’s principal benefit, in India because millions of Nepali citizens work in India and Indian citizens hold property and business in Nepal.
  • Nepal’s demand for a review of the 1950 Treaty and the unreceived report of the Eminent Persons Group keep the framework contested without changing the practice.

Myanmar: the Free Movement Regime and its scrapping

• The 1,643 km border runs through Arunachal Pradesh, Nagaland, Manipur and Mizoram, and it cuts through communities — Naga, Kuki-Chin-Zo, Mizo — that predate it.
• The Free Movement Regime permitted residents on either side to cross up to 16 km without a visa and stay for a limited period, formalising movement that the boundary had never actually interrupted.
• On 8 February 2024 the Union Home Minister announced that the FMR would be scrapped and that the entire border would be fenced, on the reasoning that free movement facilitated insurgent movement, narcotics and unregulated migration.
  • A permit-based system has replaced it — border passes for residents of villages within a limited belt, for short stays, issued and recorded rather than assumed.
  • Fencing has proceeded very slowly, constrained by terrain, land ownership and local resistance, and is a fraction of the 1,643 km after two years.
• Opposition from Mizoram and Nagaland is on kinship grounds, and is bipartisan within those states.
  • The Mizoram Assembly resolved against both the fencing and the ending of the FMR, on the argument that the boundary divides a single Zo people.
  • Naga organisations, including the United Naga Council, have demanded restoration of free movement on the same reasoning, that the line severs ancestral territory.
  • This is a federal disagreement about the definition of the nation’s edge, not merely about a policing method.
• The post-coup inflow since February 2021 has been substantial and locally managed.
  • Tens of thousands displaced from Chin State have sheltered in Mizoram, with the state government providing relief and refusing central directions to collect biometric data on the ground that they are kin, not foreigners.
  • Smaller numbers have entered Manipur, where the state government asserted large-scale illegal immigration as a driver of the 2023 conflict while civil society organisations disputed both the scale and the causal claim. The scale is contested.

Sri Lanka, Tibet and Afghanistan: the regimes that worked, and the selectivity charge

• Sri Lankan Tamils are the one long-running Indian refugee regime that reached a workable outcome, and it is instructive that no statute was involved.
  • Arrivals came in distinct waves — after the 1983 pogrom, after the IPKF withdrawal and the resumption of war, and again after 2006 — totalling well over two lakh persons.
  • Roughly 58,000 remain in state-run camps in Tamil Nadu, with a comparable number living outside them, sustained by state relief that has been repeatedly enhanced.
  • Two large organised repatriations returned tens of thousands in 1987–89 and 1992–95, and Colombo has since opened a framework for voluntary return subject to security clearance.
  • The regime worked because the host state’s population was co-ethnic and politically invested, and because the flow was maritime, bounded and enumerable — none of which is true of the eastern land border.
• Tibetans received India’s most generous treatment, and the most institutionally elaborate.
  • Settlements, schools, a Tibetan Rehabilitation Policy and Registration Certificates created a de facto status without any legal category, and courts have recognised citizenship by birth for those born in India in the relevant window.
• Afghans were treated in two tiers, and the tiering is the point.
  • Sikh and Hindu Afghans have been naturalised; others depend on UNHCR status determination and long-term visas, and their position tightened sharply after 2021.
• The selectivity charge is the critical scholarship’s central claim, and the comparison is what supports it.
  • Tibetans, Sri Lankan Tamils and Afghan Sikhs received durable protection; Rohingya and Chakmas did not, and Bangladeshis were never in the frame at all.
  • The variable that best predicts the outcome is not the severity of persecution but the group’s religion, ethnic proximity to a domestic constituency, and the foreign-policy value of the state it fled.
  • Defenders answer that discretion is what allowed the generosity — a statute with defined criteria would have admitted fewer people, not more, and would have bound India to obligations it could not fund.

Corridor	Legal status of movement	Principal driver	Dominant security concern	Instrument
Bangladesh	Unlawful without documents	Economic and episodic displacement	Demographic change; trafficking; infiltration under cover	Fencing, tribunals, push-backs
Nepal	Lawful under the 1950 Treaty	Labour market integration	Trafficking, counterfeit currency, third-country transit	Policing, not restriction
Myanmar	Formerly permitted; now permit-based	Civil war and kinship	Insurgent movement, narcotics, refugee inflow	FMR scrapped 2024; fencing
Sri Lanka	Camp-based, executive	War, then economic collapse	Minimal; maritime and fisheries friction	Camps and organised repatriation

The Rohingya question and India’s stand

The Rohingya case is where every element of the topic converges — statelessness, a mass expulsion by a neighbouring state, a hosting neighbour under strain, an Indian security framing, and litigation testing constitutional protection for non-citizens.

• The Rohingya are a Muslim population of Rakhine State, described by the United Nations as among the most persecuted minorities in the world, and rendered stateless by Myanmar’s citizenship law.
  • Myanmar’s official position is that they are Bengali migrants from the colonial period; their own claim is to indigeneity in Arakan. The dispute over origin is itself the instrument of exclusion.
• The 2017 expulsion from northern Rakhine, following ARSA attacks and the military’s clearance operations, drove roughly 700,000 people into Bangladesh in weeks.
  • The Kofi Annan-led Advisory Commission on Rakhine State had reported days earlier, urging review of the 1982 citizenship law; its recommendations were overtaken by events.
  • The Gambia’s case against Myanmar at the International Court of Justice, alleging breach of the Genocide Convention, has proceeded past the jurisdiction stage. India is not a party to it.
• Bangladesh now hosts well over a million Rohingya, most in the Cox’s Bazar camps, of which Kutupalong is the largest refugee settlement anywhere.
  • Roughly 35,000 have been relocated to Bhasan Char, a silt island in the Bay of Bengal — presented by Dhaka as decongestion, criticised as isolation with restricted freedom of movement.
  • Renewed fighting in Rakhine since 2024, as the Arakan Army took control of the border districts, has driven a further large inflow into Bangladesh.
  • Two organised repatriation attempts collapsed because no refugee volunteered to return to a state that would not restore citizenship, and the conditions have since worsened.
• India hosts roughly 40,000 Rohingya, of whom about 22,500 are registered with UNHCR, spread across Jammu, Hyderabad, Delhi, Haryana and the North-East.
• India’s stand has been consistent across three dimensions, and evaluating it requires taking each separately.
  • Domestically, they are illegal migrants, not refugees. A Union Home Ministry advisory directed states to identify and deport them; UNHCR cards are not recognised as conferring status.
  • Diplomatically, India framed the 2017 events primarily as a security and development problem in Rakhine rather than as persecution — it condemned the ARSA attacks, and abstained from or dissented on international texts naming violence in Rakhine.
  • Humanitarianly, India ran Operation Insaniyat, delivering relief to Bangladesh for the camps, and signed a Rakhine State Development Programme with Myanmar, building prefabricated housing intended to enable return.
• The case for India’s position: Myanmar’s cooperation is indispensable for North-East security and for the Kaladan and trilateral highway projects; a formal refugee designation would be irreversible and would set a precedent for a much larger flow; and India’s own security agencies allege recruitment attempts among the displaced.
• The case against it: the security claim has been asserted in litigation rather than demonstrated; the numbers are trivially small against India’s population; the position sits badly with India’s simultaneous claim to be a first responder in the region; and it exposes India to the charge that protection tracks religion.
  • The sharpest evidence for that charge is comparative — the CAA extends protection to persecuted minorities from Myanmar’s neighbours but not from Myanmar, and to no Muslim group anywhere.

The security argument, at its strongest

The argument that large-scale undocumented migration threatens Indian security deserves to be stated in its most serious form before it is tested. Stated weakly, it is easy to dismiss; stated properly, parts of it survive.

Demographic change in border districts and its electoral consequences

• The claim is that sustained inflow into a contiguous belt alters the composition of border districts, and that composition translates into political power through the franchise.
• The measurable change in Assam is real: the Muslim share of the population rose from roughly a quarter at the 1951 census to about 34 per cent in 2011, with the highest concentrations in the districts adjoining Bangladesh.
• In West Bengal, the districts of Murshidabad, Malda, Uttar Dinajpur and North 24 Parganas show the same pattern of concentration along the frontier.
• The electoral mechanism asserted is enrolment: that undocumented residents obtain ration cards, then voter identity, then inclusion in the rolls, and thereafter constitute a bloc that political parties have an incentive to protect.
• Myron Weiner gave this argument its most rigorous form long before it became a slogan, describing the anti-immigrant movement in Assam as a case of natives responding to a perceived loss of political control rather than merely of jobs.

Pressure on land, employment and services

• Land alienation is the most concrete grievance — the transfer of tribal and smallholder land to newcomers, whether by sale, encroachment or benami transaction, in states where land is the basis of community identity.
• Competition for unskilled employment in a labour surplus economy depresses wages at the bottom, which is where the political effect is felt even when the aggregate effect is small.
• Public services are rationed, so any addition to the claimant pool is experienced as subtraction by existing claimants.
• Encroachment on forest land, grazing reserves and river islands produces recurrent eviction drives, which are among the most politically volatile events in Assam.

Infiltration under cover of migration

• The argument is not that migrants are terrorists but that a high-volume, unmonitored flow supplies cover — that hostile agencies insert operatives into a movement too large to screen.
• The documentary consequence is the serious one: once obtained, an Indian identity document is indistinguishable from any other, and the fraudulent acquisition of citizenship documents is a genuine counter-intelligence problem.
• Sanctuary and transit have been the historically demonstrated versions of this concern — North-East insurgent groups operated from camps across the eastern border until Dhaka’s cooperation from 2009 dismantled them.

Trafficking, smuggling and counterfeit currency

• Human trafficking is the best-evidenced cross-border criminality in South Asia, and unlike the migration debate it rests on prosecutions and rescues rather than estimates.
  • Women and children are trafficked from Bangladesh and Nepal into Indian cities; the SAARC Convention on trafficking in women and children exists but is weakly implemented.
• Cattle smuggling is the largest single driver of border violence and of the border killings that poison the bilateral relationship.
• Narcotics move along the eastern corridors, with the Golden Triangle‘s methamphetamine and heroin entering through Myanmar.
• Counterfeit currency enters principally through Nepal and Bangladesh, and is a monetary as well as a security problem.

The North-East’s fear, and the case where it came true

• The region’s specific anxiety is not about crime or terrorism but about becoming a minority in one’s own homeland — a claim about cultural and political survival, not about law and order.
• Tripura is the case where it actually happened, and it is the reason the fear cannot be dismissed as imagined.
  • The tribal share of Tripura’s population fell from over half at the 1941 census to under a third by 2011, as roughly six lakh Bengali refugees arrived between 1947 and 1951 and around a million more during the 1971 war, most of whom stayed.
  • The Tripura Land Revenue and Land Reforms Act, 1960 prohibited transfer of tribal land, and benami transfers continued regardless.
  • The political consequences — tribal parties, an autonomous district council, and two decades of insurgency — followed the demographic inversion rather than preceding it.
• Tripura therefore functions in North-Eastern political argument as a demonstration case: proof that the outcome feared elsewhere is possible, achieved by displacement rather than by conquest.

Tripura is why the North-East’s demographic fear cannot be answered as prejudice: it is the one place in India where an indigenous majority became a minority within a generation.

The consequences borne by the displaced themselves

An account that treats migrants only as an object of security is incomplete, and the second-order effects are what turn a movement of people into a durable social problem for the receiving region.

• Undocumented status is the source of exploitability. A person who cannot approach the police, a court or a labour inspector works at whatever wage is offered and lives wherever tenure is least secure.
• Women and children bear the sharpest costs — interrupted schooling, loss of the language and cultural transmission that schooling carries, and exposure to trafficking, sexual violence and forced labour.
• Common property resources absorb the shock — grazing land, forest, riverine char, groundwater — so the environmental cost falls on the host locality and generates the resentment that drives the politics.
• Urban concentration strains housing, water, sanitation and health infrastructure, and pushes arrivals into peripheral settlements without services, which is where clearance drives then take place.
• Occupational deskilling is a recurring pattern: fishing, weaving and cultivating households end up in construction and waste handling.
• These effects are identical whether the person is a refugee or an economic migrant, which is the practical case for a status determination procedure that at least sorts the claims.

Testing the argument: what the evidence actually supports

The security case is not uniform in quality. Some of it rests on documented events, some on plausible inference, and some on assertion that has survived four decades without acquiring an evidentiary basis.

The absence of a reliable count is itself the central finding

“Motivatedly exaggerated.”

Samir Guha Roy, on the official estimates of illegal migration

• No Indian government has ever published a methodology for estimating the undocumented population, and the official numbers have varied by an order of magnitude.
  • A Home Minister told Parliament in the late 1990s that there were about ten million illegal migrants nationally, with the largest concentrations in West Bengal and Assam.
  • A Minister of State put the figure at twelve million in 2004, with five million in Assam, and effectively withdrew it days later by conceding that no accurate figures existed.
  • A Minister of State told Parliament in 2016 that there were about twenty million Bangladeshi migrants — again without a survey.
  • The Home Ministry has repeatedly replied to parliamentary questions that no authenticated data exists, because clandestine entry is by definition unrecorded.
• The independent evidence points to a far smaller order of magnitude.
  • Census birthplace data recorded roughly three million Bangladesh-born residents, and the figure has declined across successive censuses as the Partition and 1971 generations age.
  • International migrant-stock estimates put the Bangladesh-origin population in India at around three million — a fraction of one per cent of India’s population.
  • Samir Guha Roy of the Indian Statistical Institute concluded that a substantial share of those counted as foreigners were in fact Indian citizens from adjacent states, misread as aliens because they were Bengali-speaking and poor.
• Two large administrative exercises have now tested the political claims against enumeration, and both fell far short of them.
  • The Assam NRC processed 3.3 crore applicants and excluded 19,06,657 — under six per cent — and the excluded population was split roughly between Bengali Hindus, Bengali Muslims and others, which fits neither of the competing political narratives.
  • The Election Commission’s special intensive revision of the West Bengal rolls found around eleven lakh untraceable entries out of a very large electorate, against political claims that a crore or more illegal voters would be removed — and untraceability is not proof of foreign nationality.
• The honest conclusion is that the flow is real, historically substantial, and very much smaller than the political rhetoric asserts — and that the state has never seriously tried to measure it, because a credible number would constrain the politics in both directions.

The single most important fact about illegal migration into India is that nobody knows how many there are, and that four decades of asserted figures have never once been accompanied by a stated method.

Claim by claim

Claim	Mechanism asserted	What the evidence supports
Demographic change in border districts	Sustained inflow alters composition	Change is real and measurable; causation is mixed — differential fertility, internal migration and district reorganisation each contribute
Electoral capture	Enrolment converts residents into a voting bloc	Documented individual cases; the aggregate claim is unsupported by the NRC and roll-revision findings
Land and employment pressure	Competition at the bottom of the market	Strongest where land is the identity — Tripura, Assam’s char belt; weak as a national claim
Infiltration under cover	Volume defeats screening	Plausible and partly demonstrated for documents and sanctuary; rarely evidenced as a route for operatives
Trafficking and smuggling	Unpoliced frontier	Best-evidenced of all, resting on prosecutions rather than estimates
Indigenous displacement	Newcomers become the majority	Established in Tripura; asserted but not established elsewhere

How a demographic fact became a security question

• The Copenhagen School‘s theory of securitisation — developed by Barry Buzan, Ole Wæver and Jaap de Wilde — is the precise tool for this question, because it asks not whether a threat is real but how an issue becomes a security issue.
  • Securitisation is a speech act: a securitising actor declares an existential threat to a referent object, and if a relevant audience accepts the claim, the issue moves out of ordinary politics into emergency politics.
  • The pay-off of the move is exceptional measures — detention without ordinary process, reversed burdens of proof, tribunals outside the regular judiciary — that would be unavailable in normal political argument.
• Wæver’s category of societal security fits the North-Eastern case exactly: the referent object is not the state’s survival but the survival of a collective identity, a “we”.
  • This is why the Assam movement’s demand was never primarily about crime or terrorism. It was about who would remain politically and culturally dominant in Assam.
  • It also explains why the same states that fear Bangladeshi migration welcome Chin refugees: kinship determines whether an arrival is read as reinforcement or dilution.
• Sujata Ramachandran applies the framework to India directly and unsparingly. She argues that “the illegal Bangladeshi” is a constructed political category that fuses two different groups — undocumented border-crossers and Indian Muslims of Bengali origin — and that the fusion is the point.
  • She notes that the wildly divergent estimates function as propaganda rather than demography, and that Operation Pushback of 1992 in Delhi established detention-and-deportation as routine policy while collapsing when Bangladesh refused to receive the deportees.
  • Her charge is that securitisation licenses exceptional treatment of a domestic minority under cover of a foreign-policy problem.
• Desecuritisation — Wæver’s own normative preference for returning an issue to ordinary politics — is the analytical name for what the regularisation argument proposes.

The scholarship, and where it disagrees

• Myron Weiner is the foundational name and the most useful, because he neither dismisses the security concern nor accepts it uncritically.
  • His “sons of the soil” framework explains conflict between natives and migrants as competition for middle-class positions and political control in a rapidly expanding but still scarce opportunity structure, not simply for wages.
  • His work on migration and security identifies the circumstances in which a receiving state perceives migrants as threatening — as a political risk, as a cultural threat, as a social and economic burden, as an irritant in relations with the sending state, and as an instrument the sending state may use.
  • The crucial move is his insistence that a threat perception is a political fact in its own right, worth explaining rather than merely debunking.
• Sanjib Baruah supplies the Indian corrective. He argues that the “illegal immigrant” in the North-East is produced by an administrative and documentary regime as much as by any border crossing, and that the region is governed through exception.
  • His sharpest point on citizenship is that the Assam movement’s logic and the CAA’s logic are incompatible — the first is linguistic and nativist and religion-neutral, the second is religious and pan-Indian — which is why Assam’s own agitationists opposed the CAA that was supposed to answer their grievance.
• Niraja Gopal Jayal situates the whole controversy in the long drift of Indian citizenship from jus soli towards jus sanguinis — birth-based to descent-based — through the 1986 and 2003 amendments, with the CAA adding a religious criterion to a descent-based structure.
• Partha Chatterjee‘s distinction between civil society and political society explains the everyday reality: undocumented residents are governed as populations to be managed — through ration cards, slum rehabilitation and electoral patronage — rather than as rights-bearing citizens, which is how they acquire documents in the first place.
• B.G. Verghese argued that the region’s problem was development and governance rather than demography alone, and advocated regulated work permits in place of a deportation policy that could not be executed.
• Sanjoy Hazarika reached the same practical conclusion from a different direction — that the border cannot be sealed, that deportation is a fiction without a receiving state, and that documented seasonal work is the only policy that engages the actual flow.

The Indian policy response, assessed

India’s response has accumulated as a sequence of instruments, each answering the failure of the last, and none of them closing the question.

The Assam Movement and the Accord of 1985

• The Assam Movement of 1979–85, led by the All Assam Students’ Union and the All Assam Gana Sangram Parishad, demanded detection, deletion from electoral rolls, and deportation of foreigners.
• The Assam Accord, signed on 15 August 1985, settled the terms on which the agitation ended and remains the template for every later instrument.
  • Clause 5 fixed the cut-off: those who entered before 1 January 1966 would be regularised; those between 1 January 1966 and 24 March 1971 would be detected, disenfranchised for ten years and then regularised; those entering on or after 25 March 1971 would be detected and expelled.
  • The date is not arbitrary — 24 March 1971 is the eve of Pakistan’s crackdown in Dhaka, so the Accord’s logic is that those displaced by the liberation war and after are Bangladesh’s responsibility.
  • Clause 6 promised constitutional, legislative and administrative safeguards for the cultural, social and linguistic identity of the Assamese people, and was left undefined for thirty-five years.
• The Justice Biplab Sarma Committee finally proposed a definition of “Assamese people” anchored on 1951, along with reservations in legislature, local bodies, jobs and land.
  • The Assam government accepted the majority of the recommendations in September 2024, deferring those requiring constitutional amendment.
  • The tension between a 1951 definition of the protected community and a 1971 cut-off for citizenship has never been resolved, and it is the reason Clause 6 remains politically live.

Section 6A and the Constitution Bench of October 2024

This is the most consequential legal development in the field and the point on which the entire Assam settlement now rests.

• Section 6A of the Citizenship Act was inserted in 1985 to give the Accord statutory effect: it confers deemed citizenship on those of Indian origin who entered Assam before 1 January 1966, and a delayed citizenship with ten years’ disenfranchisement on those entering between then and 24 March 1971.
• Its validity was challenged on the ground that it created a special citizenship regime for one state, offending Article 14, and that it diluted the cultural rights of Assamese people under Article 29(1) and the Union’s duty under Article 355.
• On 17 October 2024 a five-judge Constitution Bench upheld Section 6A by 4:1.
  • Justice Surya Kant wrote the principal opinion, with the Chief Justice concurring separately, and Justices Sundresh and Misra joining the majority.
  • The majority held that Assam’s situation was distinct in “magnitude and impact”, which supplied the rational nexus for singling the state out, and that the 1971 date bore a rational connection to the war that produced the influx.
  • On Article 29(1), it held that mere demographic change does not establish a violation of cultural rights, and pointed to existing linguistic safeguards; a contrary holding would make every internal migration a constitutional question.
  • On Article 355, it declined to read the Union’s duty to protect states against external aggression as a justiciable right of citizens, warning that this would expand emergency powers.
  • It criticised the Union’s failure to detect and deport post-1971 entrants, locating the problem in implementation rather than in the provision, and directed continuing judicial supervision of the tribunals through a dedicated bench.
• Justice Pardiwala dissented, and his reasoning is the more interesting of the two.
  • He accepted Parliament’s competence to legislate for Assam’s extraordinary circumstances but invoked temporal unreasonableness — a provision valid when enacted may become unconstitutional as the rationale that justified it erodes.
  • His specific objections were that Section 6A has no sunset clause and no voluntary registration mechanism, so the burden of identification falls on the state while any individual may claim its benefit.
  • He would have declared it unconstitutional prospectively, preserving the citizenship already acquired under it while closing it for the future.

Question	Majority	Justice Pardiwala
Legislative competence	Present; Parliament may act on citizenship	Not contested
Article 14 — singling out Assam	Valid; Assam’s problem unique in magnitude	Valid at enactment
Effect of the passage of time	Immaterial to validity	Temporal unreasonableness — the justification has eroded
Design of the provision	Workable; the failure is in implementation	“Faulty” mechanism — no time limit, no registration route
Relief	Uphold, with judicial monitoring of tribunals	Prospective unconstitutionality, existing citizenship preserved

The NRC and its unfinished status

• The National Register of Citizens for Assam, first prepared in 1951, was updated under Supreme Court supervision following petitions including that of the Assam Sanmilita Mahasangha.
• The final list published on 31 August 2019 covered 3,30,27,661 applicants and excluded 19,06,657 persons.
• It has still not been formally accepted. The Registrar General has not notified it; the state government itself sought re-verification of a share of the entries; and the original petitioner called the outcome flawed.
• The outcome satisfied nobody, which is itself the most telling fact about it.
  • Those who expected millions of foreigners found a number an order of magnitude lower.
  • Those who feared a communal purge found the excluded distributed across communities, including large numbers of Bengali Hindus and Assamese-speaking Hindus.
  • A Comptroller and Auditor General audit found irregularities and a cost escalation from about ₹288 crore to over ₹1,600 crore.
• The exercise’s most durable effect was neither detection nor deportation but the creation of a documentary regime in which lifelong residents must prove ancestry from records that flood-prone Assam is least likely to have preserved — a burden that fell hardest on women married early, without independent land or school records.

Foreigners Tribunals, detention and deportation

• Foreigners Tribunals are quasi-judicial bodies constituted under the 1964 Order, unique to Assam in practice until the power was extended to all states in 2019.
  • Over a lakh persons have been declared foreigners since 1985, a substantial proportion in ex parte proceedings where the person never appeared, often because notice never reached them.
  • The reversed burden of proof under Section 9 of the Foreigners Act means the individual must establish citizenship, not the state establish alienage — the structural feature that generates most wrongful declarations.
• Detention followed: dedicated centres in Assam, and from 2023 the large transit camp at Matia in Goalpara, the country’s biggest.
  • The Supreme Court’s orders permitting release on bond after prolonged detention conceded the central problem — that indefinite detention without a receiving state is punishment without a terminus.
• Deportation is the point at which the whole apparatus fails, and this is the analytical crux.
  • Removal requires the receiving state to acknowledge the person as its national. Bangladesh does not, and Myanmar will not for the Rohingya.
  • India has therefore turned to push-backs — informal transfer across the border without a readmission process — with over two thousand persons removed this way from May 2025, alongside Assam’s revival of the Immigrants (Expulsion from Assam) Act, 1950 to bypass the tribunals.
  • The Gauhati High Court intervened on due-process grounds, and documented cases of misidentification of Indian citizens followed.
  • Bangladesh formally protested, calling the practice unacceptable and insisting that returns proceed only through verification of nationality — which is precisely the process India cannot obtain.

The Citizenship (Amendment) Act, 2019

• The CAA provides an accelerated route to citizenship for Hindus, Sikhs, Buddhists, Jains, Parsis and Christians from Afghanistan, Bangladesh and Pakistan who entered India on or before 31 December 2014, reducing the naturalisation residency requirement from twelve years to six.
  • It exempts the Sixth Schedule areas of Assam, Meghalaya, Mizoram and Tripura, and the states under the Inner Line Permit regime — a concession to exactly the demographic anxiety described above.
  • The rules were notified on 11 March 2024, and the first certificates were issued that May; the numbers granted have been small relative to the political salience of the measure.
• The case for it, put at its strongest, rests on four claims.
  • Partition created a specific historical obligation to the religious minorities left in the successor states.
  • Those minorities have demonstrably declined as a share of population in Pakistan and Bangladesh.
  • Unlike other displaced groups, they have nowhere else to seek protection, since the states they flee are the states of their nationality.
  • A statute that names its beneficiaries is more honest than an executive discretion that quietly favours the same groups without saying so.
• The case against it, also at its strongest: it is the first Indian statute to make religion a criterion for citizenship, which offends the secular premise of Articles 14 and 15.
  • The classification is under-inclusive by its own logic — it omits persecuted Muslim sects such as Ahmadis and Hazaras, omits Sri Lankan Tamils and Rohingya, and omits Tibetans, though each faces persecution on the Act’s own reasoning.
  • The pairing with a nationwide NRC was the source of the real alarm: a register that can exclude, combined with an Act that readmits only non-Muslims, produces a differential risk of statelessness by religion.
  • Challenges under Article 14 are pending before the Supreme Court, which declined to stay implementation.
• The change the Act makes is to the basis of eligibility, not merely its scope. Earlier Indian practice sheltered groups by executive choice without writing religion into law; the CAA writes it in, and that is what the constitutional objection is about.

Fencing, border management and the 2025 statute

• Fencing is the most heavily funded response and the least conclusive: it channels movement to identifiable points and does nothing about overstay and document fraud.
• Complementary measures — floodlighting, border out-posts, riverine patrolling, integrated border management systems and check posts — have improved interdiction without producing an enumeration.
• The Immigration and Foreigners Act, 2025 consolidates and hardens the legal regime.
  • It raises penalties sharply — up to five years and ₹5 lakh for entry without valid documents, and two to seven years and ₹1–10 lakh for forged travel documents.
  • It imposes reporting duties on universities, hospitals and hotels, extends carrier liability to land and water transport, gives statutory backing to the Bureau of Immigration, and permits arrest without warrant.
  • It makes the government’s decision on the entry of a foreigner final, and it contains no definition of a refugee and no asylum procedure — so a consolidation that could have created a status determination mechanism deliberately declined to.
  • The criticism that matters constitutionally is the combination of restriction powers, warrantless arrest and finality of determination without a defined route of periodic judicial review, which raises Article 21 and Article 22 questions.

Instrument	Underlying logic	Record
Assam Accord, 1985	Political settlement by cut-off date	Ended the agitation; Clause 6 unimplemented for 35 years
Section 6A	Statutory effect for the Accord	Upheld 4:1 in 2024; enforcement criticised by the Court itself
NRC, 2019	Enumeration of citizens	Excluded 19 lakh; never formally accepted; satisfied nobody
Foreigners Tribunals	Individual adjudication of nationality	Over a lakh declared foreigners; heavy reliance on ex parte orders
CAA, 2019	Protection by religious category	Rules notified March 2024; small numbers; constitutional challenge pending
Fencing	Physical interdiction	3,232 km of 4,096 km on the eastern border; 175 km non-feasible
Deportation and push-backs	Removal	Blocked by non-acceptance; substituted by informal push-backs and protest from Dhaka
Immigration and Foreigners Act, 2025	Consolidation and deterrence	In force 1 September 2025; still no refugee category

Why migration is an impediment to regional cooperation

This is where a domestic controversy becomes a problem of foreign policy, and the mechanism is worth stating precisely rather than assuming.

“If any Bangladeshi national is found illegally residing in India, they should be returned through the proper channels.”

Md Jahangir Alam Chowdhury, Bangladesh’s Home Adviser

• The core mechanism: enforcement requires the cooperation of the sending state, and seeking that cooperation converts an internal policy into a bilateral demand the neighbour cannot concede without political cost.
  • India cannot deport without Bangladesh acknowledging nationality. Bangladesh cannot acknowledge it without conceding a claim its politics has always denied.
  • The result is a structurally unresolvable bargain, which is why the issue recurs at every summit and is settled at none.
• Dhaka’s position has been consistent across every government — Awami League, BNP, caretaker and interim: there is no illegal migration of Bangladeshi nationals to India, and any person to be returned must first be verified as a citizen through proper channels.
  • The issue was last substantively raised at leader level in the early 1990s, and has since been kept off the formal bilateral agenda by mutual convenience.
  • In 2025 the interim government responded to the push-backs by protesting formally and warning of consequences — the first time in years that the question has been an open irritant rather than a suppressed one.
• The NRC and the CAA were read in Bangladesh as directed at it, whatever New Delhi intended.
  • India’s assurance that both were internal matters was undercut by the public reasoning offered for them, which named persecution in Bangladesh as a justification.
  • Bangladeshi ministers cancelled visits to India at the height of the controversy, and Dhaka rejected the characterisation of its treatment of minorities.
  • The episode illustrates a general point: domestic legislation that names a neighbour’s citizens is foreign policy, whether or not it is framed as such.
• No bilateral or regional framework for labour mobility exists, and its absence is the true policy gap.
  • India has concluded migration and mobility partnership agreements with European and Gulf partners and labour arrangements with several, but has no such instrument with Bangladesh, its largest land neighbour by border length.
  • SAARC has no refugee or migration instrument at all. Its only relevant convention concerns trafficking, and adherence to sovereignty and non-interference has kept forced migration off the regional agenda entirely.
  • A model national refugee law drafted in the 1990s under an eminent persons process chaired by Justice P.N. Bhagwati was offered to South Asian states and adopted by none.
  • Private members’ bills in Parliament proposing an Indian asylum law have never been taken up.
• The regional cost is cumulative. Migration blocks progress on the connectivity and trade agenda it is nominally unrelated to, because it supplies the domestic constituency in India’s border states that opposes transit, land ports and visa liberalisation.
• S. Chatterjee draws attention to the federal layer that most accounts of the bilateral relationship omit: domestic politics on both sides sets the limits within which diplomacy can operate.
  • Alternation between the Awami League and the BNP has historically produced discontinuity in Dhaka’s posture toward India.
  • On the Indian side, a state government’s veto — West Bengal on the Teesta, and border states on transit and land ports — is a standing constraint that Delhi can neither ignore nor overrule.
• The measures that would actually engage the criminality, as distinct from the migration, are institutional rather than physical.
  • A standing joint task force on trafficking and narcotics, which requires only police cooperation and no political concession from either capital.
  • Smart border management — sensors, surveillance and biometric records at crossings — in place of continuous fencing on terrain that cannot take it.
  • Regularised border haats and integrated check posts, which convert informal crossing into recorded movement.

The regularisation argument

• The constructive position in the scholarship is that prohibition without deportation is not a policy but a pretence, and that the alternative is to convert an unmanaged flow into a managed one.
• Veena Sikri, who served as India’s High Commissioner in Dhaka, has put it most directly:

“If there is a demand for workers from Bangladesh to India, the government can issue work permits. They can regularise their legal standing.”

Veena Sikri

• Her fuller argument has three parts, and each is a distinct claim.
  • Migration between the two countries cannot be stopped, because geography, kinship and labour demand all run against the border; the achievable aim is a work-visa regime that documents and time-limits it.
  • The push and pull factors must be addressed at source, which means development cooperation in the sending districts rather than fencing alone.
  • Climate change will aggravate the problem, because Bangladesh’s exposure as a low-lying delta guarantees a rising displacement pressure that no enforcement regime can absorb — so only a long-term arrangement can work.
• The case for regularisation turns on four gains, of which the first is the decisive one.
  • It makes the population countable, which is the precondition for every other policy in this field.
  • It removes the criminality that makes migrants exploitable, and therefore cheap to the employers who create the pull.
  • It gives Dhaka something to accept rather than something to deny, which is the only way past the bilateral impasse.
  • It separates the genuine security task of screening from the impossible task of mass expulsion.
• The case against it is not merely political squeamishness, and each objection is serious.
  • Any regularisation creates an expectation of future regularisation, and therefore a fresh pull factor.
  • It is unsellable in the very border states whose consent the Union would need.
  • It presumes an administrative capacity to enforce time limits that the record of visa overstay does not support.
  • It would be read in Assam and Tripura as ratifying the demographic change the movements were fought over.
• The honest assessment is that regularisation is the only instrument that engages the actual flow, and that its political feasibility is close to nil in the current climate — which is a statement about Indian politics rather than about the argument’s merits.

India’s migration policy has been built on a deportation power it cannot exercise; every instrument since 1985 has been an attempt to work around that single missing capability.

Climate displacement: the category that does not exist

• Bangladesh is among the most climate-exposed populous countries on earth, and the exposure is structural rather than episodic.
  • Most of its territory is deltaic floodplain barely above sea level; a metre of sea-level rise is commonly estimated to inundate close to a fifth of its land area.
  • Salinity intrusion in the south-western districts is already destroying rice cultivation and drinking water, converting agricultural livelihoods into migration decisions.
  • Cyclones and riverbank erosion displace hundreds of thousands annually, and Dhaka absorbs a very large annual inflow of rural migrants.
• The World Bank’s Groundswell projections put internal climate migrants across the developing world in the hundreds of millions by 2050, with South Asia among the largest shares — and internal displacement of that magnitude does not remain internal indefinitely where a porous land border adjoins a larger economy.
• There is no legal category for a climate-displaced person, and this is the decisive gap.
  • The 1951 Convention’s five grounds do not include environmental cause; a person fleeing salinity is not a refugee in law however involuntary the movement.
  • The OAU Convention and the Cartagena Declaration broadened the definition regionally to include generalised violence and disaster; Asia has no equivalent instrument, only the non-binding Bangkok Principles.
  • The UN Human Rights Committee, in the case brought by Ioane Teitiota against New Zealand, accepted in principle that climate effects could engage non-refoulement obligations while finding the threshold unmet — the furthest international law has moved.
• The policy implication for India is uncomfortable: the flow it currently treats as an enforcement problem will acquire a humanitarian character it has no legal vocabulary to recognise, and it will arrive on the border with the longest unfenced stretches.
• This is the strongest argument for a domestic refugee statute — not on humanitarian grounds alone, but because a state that must distinguish between claims needs a procedure for doing so, and India has never built one.

Conclusion

Illegal cross-border migration is best understood as three problems wearing one name. There is a real and historically substantial flow across the eastern border, driven by geography, kinship and labour demand, whose size no Indian government has ever credibly measured. There is a genuine set of security concerns — trafficking, smuggling, document fraud, sanctuary — that are best evidenced precisely where the political argument pays them least attention. And there is a securitised political construction that fuses undocumented migrants with a domestic minority and licenses exceptional treatment of both.

The policy failure is specific rather than general. India built an elaborate machinery of cut-off dates, registers, tribunals and detention on the assumption of a deportation power it does not possess, because removal requires a receiving state and none will acknowledge the persons concerned.

• Every instrument since, up to the push-backs of 2025 and the consolidating statute of the same year, has routed around that missing capability rather than replaced the assumption behind it.
• The alternatives that engage the actual flow are all available and none is politically live — a documented work-visa regime, a domestic asylum law with a status-determination procedure, and a bilateral labour mobility framework.

For the regional question the conclusion is sharper still. Migration impedes South Asian cooperation not because migrants threaten the region but because the enforcement response converts a domestic problem into a bilateral accusation, and because the region has built no institution capable of holding the conversation. Until a mobility framework exists, each Indian attempt at a domestic solution will read in Dhaka as an accusation, and each denial from Dhaka will read in India as bad faith.