The citizenship amendments are usually argued about as if they were a single event, and the argument usually fastens on the last of them. Read in sequence, the amendments to the Citizenship Act, 1955 tell a longer and more precise story. Between 1986 and 2015 Parliament amended the Act five times in ways that matter for status, and a sixth amendment in 2019 closed the sequence. Each amendment can be dated exactly, each changed a small number of identified provisions, and the changes point in a direction that is visible only when the whole trail is laid out together. The direction is double. Within India, automatic entitlement narrowed step by step: citizenship by birth, which the original Act granted on the fact of birth alone, became conditional first on one parent being a citizen and then on a stricter two-part test, while a defined category of person, the illegal migrant, was created and barred from every route into citizenship. Outside India, the movement ran the other way: a documentary status for persons of Indian origin abroad was created in 2003, widened in 2005 and consolidated in 2015, so that the same decades that made the statute harder to enter from inside made it more accommodating to the diaspora outside.

This article is the trail itself. It states each amending Act by its number and year, gives the date from which each change operated, and identifies the provisions each amendment touched. The statute that the amendments altered is treated in full in the article on the Citizenship Act, 1955 explained, and the wider field, including the constitutional provisions and the proof of status, is mapped in the complete guide to citizenship law in India. The amendment of 2019 is stated here only at the mechanical level, as the end point of the sequence, because its merits are the subject of a separate article and are not reopened here. What this article adds to the series is the pattern: five amendments before 2019, read as one instrument of change, moving in two directions at once.

Three habits of error recur in writing on this trail, and the article corrects each where it arises. The first is dating: the end of unqualified citizenship by birth is often placed in 2004, when in fact the first condition was imposed for births from 1 July 1987, and 2004 brought the second and stricter condition. The second is attribution: the statutory basis for a register of citizens is often assigned to the politics of 2019, when in fact the provision was inserted in 2003 and rules under it were made in the same year. The third is retrospectivity: the amendments are often described as if they reclassified people wholesale, when in fact each was drafted to operate by date bands, so that a person’s position continues to be governed by the law in force at the date of the birth or other event relied on. That third feature is not a technicality. It is the reason the trail can be read at all, because the statute carries its own history inside Section 3 and Section 4, clause by clause, and the clauses are arranged by date.

The starting point: what the 1955 Act granted

The Citizenship Act, 1955 (Act 57 of 1955) was enacted under Article 11 of the Constitution, which left the acquisition and termination of citizenship to Parliament. Part II of the Constitution had identified who was a citizen at its commencement on 26 January 1950; the Act governed everyone after that date. It provided five routes to citizenship: by birth under Section 3, by descent under Section 4, by registration under Section 5, by naturalisation under Section 6, and by incorporation of territory under Section 7. Of these, the two that the amendment trail reshaped most deeply are birth and descent, because both operated automatically at the moment of birth and neither originally asked much about the parents.

What did the original Section 3 of the Citizenship Act, 1955 provide?

The original Section 3 provided that every person born in India on or after 26 January 1950 was a citizen by birth, whatever either parent’s nationality. Two exceptions stood in Section 3(2): a child whose parent held diplomatic immunity and was not a citizen, and a child of an enemy alien born in a place under enemy occupation.

The simplicity of that rule deserves emphasis, because everything in the trail is a departure from it. A child born in India to two foreign parents, neither of whom had any right to remain in India, was an Indian citizen by birth. The rule was a rule of the soil, and it matched the circumstances of a republic that had just absorbed the movements of Partition and had reason to prefer a generous and administrable test: the place of birth is a fact that a birth register can prove, while the status of two parents is a fact that may be contested, undocumented or unknown. Section 4, on descent, was less even-handed. A person born outside India was a citizen by descent if the father was a citizen of India at the time of the birth. The mother transmitted nothing under the original text, a reflection of the drafting conventions of the period rather than of any stated policy about the diaspora. Where the father was himself a citizen by descent only, conditions about registering the birth at an Indian consulate applied, so that descent did not propagate indefinitely through generations born abroad without some act of connection to India.

Registration and naturalisation, the two discretionary routes, were comparatively open in the original scheme and were tightened at points along the trail. Section 5 allowed registration of persons of Indian origin ordinarily resident in India, of women married to citizens, of minor children of citizens and of certain others. Section 6 allowed naturalisation on the qualifications in the Third Schedule, which measured residence in years. Neither route contained any category of person who was barred in advance. The statute of 1955 had no concept of an illegal migrant. It knew citizens, and it knew foreigners, and the Foreigners Act, 1946 governed the presence of the second class; but it did not sort foreigners into those who could in principle become citizens and those who could not. The creation of that sorting, thirty years later and then decisively in 2003, is one of the two great movements of the trail. The other, the building of the overseas category, began in the same amendment that completed the sorting, and the coincidence is the article’s central finding.

It is also worth fixing what the 1955 Act did not contain, because later debate often projects later machinery backwards. There was no provision for a register of citizens. There was no provision for national identity cards. There was no overseas citizenship, and no cardholder status of any kind for persons of Indian origin abroad; the Constitution itself, in Article 9, barred dual citizenship by providing that a person who voluntarily acquired the citizenship of a foreign state ceased to be an Indian citizen, so any diaspora status Parliament later built had to be something less than citizenship in the full sense. The Act was, in short, a statute about people present in India or born to Indian fathers abroad, administered through birth registers, consulates and the discretion of the Central Government. The amendments that follow changed each of those silences, one at a time.

The constitutional frame the amendments worked within

The amendments were all made to a statute that itself sits inside a constitutional settlement of a particular kind, and the settlement explains several features of the trail that otherwise look arbitrary. Articles 5 to 8 of the Constitution identified the citizens of the new republic at its commencement: persons domiciled in India who were born in India, or either of whose parents was born in India, or who had been ordinarily resident in India for not less than five years before commencement; persons who had migrated to India from Pakistan in the circumstances the articles describe; persons who had migrated to Pakistan and returned under a permit for resettlement; and persons of Indian origin residing outside India whose births, or whose parents’ or grandparents’ births, were in India as defined in the Government of India Act, 1935, provided they registered with an Indian diplomatic or consular representative. Article 9 then withdrew citizenship from any person who had voluntarily acquired the citizenship of a foreign state. Article 10 continued citizenship subject to parliamentary law, and Article 11 gave Parliament the power to make any provision with respect to the acquisition and termination of citizenship and all other matters relating to it. The Constitution, in other words, answered the question of 1950 and then handed the subject to ordinary legislation.

Three consequences of that design run through the trail. The first is that every change described in this article was made by ordinary Act of Parliament, by simple legislative process, without any amendment to the Constitution itself. The narrowing of 1986, the creation of the illegal migrant in 2003 and the widening of 2005 were all exercises of the Article 11 power, which is why they could be enacted quickly, in single sessions, and why their reversal would require no more than another Act. The second consequence is that the statute rather than the Constitution is where a person’s status is found for any date after 26 January 1950, and the statute’s date banded clauses are the operative text. The third consequence is the Article 9 bar on dual citizenship, which fixed in advance the outer limit of diaspora policy. When the Singhvi Committee recommended dual citizenship in 2002, it was recommending something the constitutional text forecloses in its literal form, because an Indian citizen who takes another citizenship ceases to be an Indian citizen by force of Article 9 read with Section 8 and Section 9 of the Act. The overseas citizen scheme was drafted to give the diaspora everything except the name and the political rights: a lifelong visa, residence without registration, parity in economic and educational fields, and a card. It is best understood not as a half measure adopted for want of generosity but as the maximum status available without touching Article 9, and the amendments of 2005 and 2015 are movements within that constitutional ceiling.

The constitutional frame also explains why the trail contains no amendment about loss of citizenship after 1955 of comparable importance. Sections 8, 9 and 10 of the original Act, on renunciation, termination on voluntary acquisition of another citizenship, and deprivation by order of the Central Government on grounds including fraud, disloyalty and seven years of continuous residence abroad, were left substantially as enacted. Parliament’s attention across five amendments was directed at entry into citizenship and at documentary status, not at exit. The statute that results is asymmetric in a way the table later in this article makes visible: it is much easier to trace how the door inwards narrowed than to find any corresponding change in the door outwards, and a person tracing a family’s status across the decades will find the amendments waiting at the moments of birth, descent and registration, and almost nowhere else.

The Assam prelude: the Accord and Section 6A

The first amendment in the trail was not one of the five general amendments at all, and its separate character matters. Through the years from 1979 to 1985, Assam was the site of a sustained agitation led by the All Assam Students’ Union and allied organisations, directed at the presence on the electoral rolls of persons who had entered the State from across the border with Bangladesh, and at the political and cultural consequences the movement attributed to that entry. The agitation ended in a settlement, the Assam Accord, signed on 15 August 1985 between the Government of India, the Government of Assam and the leaders of the movement. The Accord fixed its own date bands: persons who had come to Assam before 1 January 1966 were to be treated as citizens; persons who had come between 1 January 1966 and 25 March 1971 were to be detected as foreigners, registered, kept off the electoral rolls for ten years and then treated as citizens; persons who had come after 25 March 1971 were to be detected and removed.

Parliament gave that settlement statutory form within months. The Citizenship (Amendment) Act, 1985 (Act 65 of 1985), in force from 7 December 1985, inserted Section 6A into the Citizenship Act, headed as special provisions for persons covered by the Assam Accord. Section 6A enacted the Accord’s bands almost exactly. Persons of Indian origin, a term the section defines by birth in undivided India of the person, a parent or a grandparent, who came to Assam from the specified territory before 1 January 1966 and had been ordinarily resident in Assam since entry were deemed to be citizens from 1 January 1966. Persons of Indian origin who came on or after 1 January 1966 but before 25 March 1971, who had been resident in Assam since entry and who had been detected to be a foreigner under the Foreigners Act, 1946 and the Foreigners (Tribunals) Order, 1964, were to be registered, and were deemed citizens for all purposes on the expiry of ten years from the date of detection, with every right except inclusion in an electoral roll in the interval. Entry on or after 25 March 1971 attracted no such provision at all: such a person remained a foreigner, to be dealt with under the foreigners law. The constitutional validity of Section 6A was challenged decades later and was upheld by a Constitution Bench of the Supreme Court in October 2024 by a majority of four to one, a judgment whose reasoning belongs to the article on the principal Act rather than to this trail.

The mechanics of the second band repay attention, because they show the settlement thinking in administrative detail. A person in the 1966 to 1971 band did not become a citizen on entry, on detection or on registration. Detection by a tribunal under the Foreigners (Tribunals) Order, 1964 started a clock: the person was to be registered by the registering authority, the person’s name was to be deleted from the electoral roll for the constituency, and for ten years from the date of detection the person held every right of a citizen except the right to be included in an electoral roll, becoming a citizen for all purposes, the franchise included, only when the ten years expired. The provision thus created a statutory waiting room, a decade of citizenship in substance withheld at the single point of voting, and it did so for a class defined by two dates and a tribunal’s finding. Nothing else in the Citizenship Act resembles it. The general amendments work by attaching consequences to births and applications; Section 6A worked by suspending a consequence for a defined population across a defined decade, and its presence in the Act from December 1985 is a standing reminder that the citizenship statute has been used, from the start of the trail, as an instrument for managing a migration settlement and not only as a code of status.

Section 6A matters to the trail for two reasons. The first is that it was the statute’s first experiment in sorting entrants by date and origin and attaching different citizenship consequences to each band, a technique the later general amendments used repeatedly. The second is that it was deliberately local. It applied to Assam alone, to entrants from the specified territory alone, and to a closed set of dates that lay in the past when it was enacted. The general amendments that followed took the underlying anxiety, the undocumented entrant whose presence the State could neither regularise nor remove, and addressed it through rules of general application. The connection between the Assam settlement and the amendment of 1986 is stated in the 1986 Act’s own Statement of Objects and Reasons, which recited the entry of large numbers of persons from Bangladesh, Sri Lanka and certain African countries and the Government’s decision to make the grant of citizenship more stringent. The local settlement and the general tightening were products of the same political moment, and the trail is easier to follow if they are read together.

The amendment of 1986: the first condition on birth

The Citizenship (Amendment) Act, 1986 (Act 51 of 1986) received the assent of the President on 28 November 1986. Its central change was to Section 3. For births before its operative date, the old rule stood untouched: every person born in India on or after 26 January 1950 and before 1 July 1987 was a citizen by birth, subject only to the two exceptions in Section 3(2). For births on or after 1 July 1987, the substituted Section 3(1) required that either of the person’s parents be a citizen of India at the time of the birth. The date of 1 July 1987, rather than the date of assent or of publication, is the hinge the section itself uses, and it is the date that determines status in any contested case. A child born in India on 30 June 1987 to two foreign parents was a citizen by birth. A child born to the same parents a day later was not, unless one parent was a citizen. Unqualified citizenship by soil, which had operated for thirty seven years, ended on a named day, and it ended in 1987 rather than in 2004, the date to which the change is most often misassigned.

The reasoning offered at the time was stated plainly in the Statement of Objects and Reasons and repeated in the parliamentary debate: automatic citizenship by birth was said to reward clandestine entry, because the child of an entrant who had no right to be in India acquired a status the parent could never acquire directly, and the presence of citizen children complicated every later attempt at detection and removal. The answer chosen was not to make the child’s status depend on the legality of the parent’s presence, a test that would have required the State to adjudicate the parent before it could register the child. The answer was simpler and blunter: to require a citizen parent. That choice had a consequence its authors accepted. The child of two long settled foreign residents, present in India lawfully for decades, was placed in the same position as the child of a clandestine entrant. Both needed a citizen parent, and neither had one. The 1986 amendment drew its line by parentage, not by the lawfulness of the parents’ stay, and it took the further narrowing of 2003 to introduce the lawfulness of presence into the birth rule at all.

The birth rule was the centre of the 1986 Act, but it was not the whole of it, and the rest of the Act moved in the same direction. The Statement of Objects and Reasons proposed, and the Act carried, a lengthening of the residence qualification for citizenship by registration: a person of Indian origin applying under Section 5 had previously needed to be ordinarily resident in India for six months before applying, and the period was raised to five years. The same statement proposed raising the residence period for naturalisation from five years to ten years for an applicant who was not a citizen of a Commonwealth country listed in the First Schedule. One change in the 1986 Act ran against the grain of tightening and is worth recording because it shows the Act was a revision and not merely a restriction: the registration category for spouses, which the original Act had confined to women married to citizens of India, was widened by substituting the word person, so that a man married to a citizen became eligible on the same terms as a woman. Gender entered the trail here for the first time, in the registration route; it returned six years later in the descent route, in the amendment of 1992.

Measured against what followed, the 1986 amendment was moderate. It kept citizenship by birth as an automatic route for the great majority of children born in India, since the great majority had at least one citizen parent. It created no barred category. It inserted no register and no identity card. What it did was to establish the principle on which everything later depended: that birth in India was no longer, by itself, enough. Once that principle was conceded, the questions for later Parliaments were only how much more than birth should be required, and what should happen to the children of those whose presence the State treated as unlawful. The first question was argued again in 2003. The second was created in 2003, because before that year the statute had no answer to it at all.

The 1986 Act also deserves a word about what it reveals of the politics of the trail, because it was the first citizenship amendment of general application since 1955 and it set the terms in which every later amendment was debated. Its Statement of Objects and Reasons named the countries from which entry was said to be clandestine and named the mischief as the acquisition of citizenship by the children of such entrants, and its passage was not seriously opposed on the principle of the birth condition itself. The principle commanded a breadth of assent in 1986 that surprises readers who meet citizenship politics only in its later and more polarised form: the proposition that birth in India, without any parental connection, should confer citizenship was defended by almost no one once the question was put, and the argument of the period concerned the residence periods and the marriage category rather than the central change. Two features of that consensus matter for what followed. The first is that the 1986 Parliament treated the birth rule as an instrument of migration control, a characterisation that has attached to it ever since, although the rule’s actual operation falls overwhelmingly on the children of lawful residents and of citizens’ spouses rather than on the children of clandestine entrants, who are a small fraction of the births it governs. The second is that the amendment left the determination of parentage undocumented and unmachined: it required a citizen parent and provided no register by which parentage and status would be recorded, so that the condition it imposed would be proved, in contested cases, by the ordinary civil evidence of birth records, passports and the testimony of record. The gap between the condition and the machinery was present from 1987. It is the gap Section 14A was later inserted to close, and a reader of the trail should notice how long the statute ran, sixteen years, with a conditional birth rule and no register to administer it.

The amendment of 1992: descent made gender neutral

The Citizenship (Amendment) Act, 1992 (Act 39 of 1992) made a single change of principle, and it moved in the opposite direction from the amendment before it. Section 4 governs citizenship by descent, for persons born outside India. Under the original text, and under the text as it stood until December 1992, descent ran through the father alone: a person born abroad was a citizen by descent if his father was a citizen of India at the time of the birth, and the citizenship of the mother was legally irrelevant. The 1992 Act substituted a new sub section for Section 4(1) with effect from 10 December 1992. A person born outside India on or after 26 January 1950 but before 10 December 1992 remained governed by the father rule. A person born outside India on or after 10 December 1992 was a citizen by descent if either of his parents was a citizen of India at the time of the birth. The provisos carried the older safeguards into the new rule: where the parent through whom descent was claimed was a citizen by descent only, the birth had to be registered at an Indian consulate within one year, or later with the permission of the Central Government, or the parent had to be in service under a Government in India at the time of the birth.

The change is small in compass and large in what it discloses about the trail. The 1992 amendment was not a security measure. It responded to an obvious inequality inside the statute, under which two children born in the same foreign city to two Indian parents in opposite configurations, an Indian father with a foreign mother and a foreign father with an Indian mother, received opposite answers. After 10 December 1992 they received the same answer. The amendment also had a diaspora dimension that connects it to the overseas story told later in this article: as emigration from India grew and as Indian women settled abroad in their own right, a father only rule excluded a widening class of children of Indian mothers, and the correction enlarged the class of persons with a statutory connection to India at exactly the period when policy towards the diaspora was becoming more active. The amendment is therefore the first point in the trail where the diaspora movement is visible, a decade before the overseas category itself was created. It widened entitlement, deliberately, and it did so by date band like every other amendment in the sequence, leaving the position of persons born before 10 December 1992 exactly as the old law had left it.

The 1992 amendment also repays a closer look at its limits, because they are characteristic of the trail’s method. It did not disturb the provisos that had always hedged descent about with conditions. Where the transmitting parent was a citizen by descent only, the claim of the next generation depended on an act of registration at a consulate within a year, or on the parent being in government service in India at the time of the birth, and those conditions survived the substitution of either parent for the father untouched in substance. Descent, in other words, was widened at the point of gender and left narrow at the point of generation: the statute was willing to let a mother transmit what a father could transmit, and equally unwilling to let citizenship by descent propagate indefinitely through a line that had ceased to register its connection with India. The further tightening of the registration condition in 2003, with its declaration against a foreign passport, was therefore not a reversal of the policy of 1992 but the strengthening of a condition that 1992 had preserved. Reading the two amendments together, as the trail requires, the consistent principle is connection: a parent of either sex may transmit citizenship to a child born abroad, and the family must perfect the transmission by registration while the child is an infant, or the claim lapses into the discretionary territory of permission from the Central Government.

Why the dates decide: the prospective operation of the amendments

Every amendment in this trail was drafted to operate prospectively, and the drafting technique is the same in each case. The substituted sections do not state a single rule and a commencement date. They state a series of rules, each attached to a band of dates, so that the present text of Section 3 reads as a compressed history of the trail itself. Clause (a) of Section 3(1) states the unqualified rule for births from 26 January 1950 to 30 June 1987. Clause (b) states the one parent rule for births from 1 July 1987 to 2 December 2004. Clause (c) states the stricter rule for births from 3 December 2004 onward. Section 4 does the same for descent, separating births before 10 December 1992 from births on or after that date. A reader who knows nothing else about the amendments can reconstruct their order and their operative dates from the two sections alone, which is why errors about the trail are so often errors that the statute itself refutes.

The legal consequence is that citizenship by birth or descent is determined by the law in force at the time of the birth, and a later amendment does not disturb a status that had already attached. This proposition is frequently misunderstood in both directions. It is misunderstood by those who assume that a tightening amendment exposes people born earlier to re examination under the new test: it does not, because the date bands preserve the earlier rule as the governing rule for earlier births, and the person born in India in 1980 to foreign parents holds citizenship under clause (a) by force of the same section that denies it to a person born in 2010 on identical parentage. It is also misunderstood by those who assume that a liberalising amendment repairs earlier exclusions: the 1992 amendment did not make a citizen of a person born abroad in 1985 to an Indian mother and a foreign father, because clause (a) of Section 4(1) continues to ask about the father for births before 10 December 1992. Prospectivity cuts both ways, and any honest account of the trail has to hold both edges in view. The amendments changed the law for the future, repeatedly and substantially. They did not rewrite the past, and the status of every person born before a given amendment remains an inquiry into the text as it stood on the day of the birth.

There is one great exception to prospectivity in this territory, and it proves the rule by the care with which it was confined. Section 6A, inserted in 1985, operated retrospectively by design: it deemed persons who had entered Assam before 1 January 1966 to be citizens from 1 January 1966, a date two decades before the section was enacted, and it fixed the status of the 1966 to 1971 entrants by reference to a detection and a ten year period that the section itself constructed. Retrospectivity of that kind was possible because Section 6A dealt with a closed class, identified by entries that had already happened, under a political settlement whose whole purpose was to fix the past. The general amendments did the opposite. Each addressed an open future: births that had not yet occurred, applications not yet made, a register not yet compiled. The contrast is the cleanest evidence that the prospective operation of the 1986, 1992 and 2003 amendments was a deliberate legislative choice and not a drafting habit. Parliament knew how to write a citizenship provision that reached backwards, had done so in the same statute two years before the first of the general amendments, and chose, for the general law, not to.

Worked examples fix the point better than abstraction. A person born in Delhi on 15 May 1987, both parents being citizens of another country lawfully resident in India, is a citizen by birth under the original rule. A person born in Delhi on 15 May 1990 on the same facts is a citizen by birth only if one parent was a citizen, and on these facts is not. A person born in Delhi on 15 May 2010, one parent a citizen and the other a foreigner lawfully resident, is a citizen by birth under clause (c), because one parent is a citizen and the other is not an illegal migrant. A person born in Delhi on the same date, one parent a citizen and the other an illegal migrant, is not a citizen by birth, because neither limb of clause (c) is satisfied. A person born in Toronto on 15 May 1990 to an Indian mother and a Canadian father is not a citizen by descent, because the father rule governed 1990. A person born in Toronto on 15 May 1995 on the same facts is a citizen by descent, because either parent sufficed by then, subject to the registration proviso if the mother was a citizen by descent only. Six births, three cities’ worth of identical human facts, and the answers are supplied entirely by the calendar. That is what it means to say the amendments operate prospectively, and it is why the FAQ literature on this subject, which often asks what the rule is without asking when the birth was, misleads the very readers who most need the date first.

Between the amendments: detection and the foreigners machinery

The years between 1992 and 2003 saw no general amendment to the Citizenship Act, but the machinery for deciding who was a foreigner, on which any citizenship regime ultimately depends, was contested throughout the period and the contest forms part of the setting of 2003. In Assam, the Illegal Migrants (Determination by Tribunals) Act, 1983 had created a special tribunal system for determining whether a person in Assam was an illegal migrant, defined by entry into India on or after 25 March 1971, and it applied to Assam alone. The ordinary law elsewhere, and the law to which Assam would return, was the Foreigners Act, 1946, under which the Central Government’s power to make orders restricting or removing foreigners is wide, tribunals were constituted under the Foreigners (Tribunals) Order, 1964, and Section 9 of the 1946 Act places the burden of proving that a person is not a foreigner on that person. The Assam experiment was attacked from both directions for twenty years: by those who said its procedures made detection nearly impossible, because a complaint required support from persons living within a defined distance and the burden arrangements favoured the person proceeded against, and by those who said any special and harsher regime for one state was discriminatory. The constitutional challenge succeeded after the period covered by this section, when the Supreme Court in Sarbananda Sonowal v. Union of India, (2005) 5 SCC 665, struck the 1983 Act down as unconstitutional, describing large scale illegal migration in terms of external aggression for the purposes of Article 355 and holding that the Act had made the identification of illegal migrants next to impossible. The judgment postdates the 2003 amendment, but the debate it concluded predates it, and the amendment’s insistence on defining the illegal migrant inside the Citizenship Act itself, rather than leaving the concept to the foreigners statutes, is best read against this background of a detection regime widely believed, by the government of the day, to have failed.

The point matters because it distinguishes two questions that public discussion habitually merges. Whether a person is a foreigner present in India without authority is a question for the foreigners law, with its tribunals, its burden of proof and its orders of removal. Whether a foreigner may become a citizen, and whether a child born in India is a citizen at birth, are questions for the Citizenship Act. Before 2003 the two statutes barely touched: the Citizenship Act contained no definition drawn from the law of entry, and a person proceeded against as a foreigner could in principle still pursue registration or naturalisation on the ordinary criteria. The 2003 amendment built the bridge between them by importing the fact of unlawful entry and overstay into the citizenship statute as a defined status with citizenship consequences. That is a legislative choice of a recognisable kind, and identifying it correctly is a precondition of arguing about it intelligently. Criticism of the burden of proof in foreigners proceedings is criticism of the 1946 Act and the 1964 Order. Criticism of the barred category is criticism of the 2003 amendment. The trail keeps the two apart, and the articles that own the foreigners statute argue the first on its own ground.

The setting of 2003: a security review after a conflict

The amendment of 2003 cannot be understood from its text alone, because its two most distinctive features, the register of citizens and the national identity card, arrived in the Citizenship Act from a security review rather than from citizenship policy. After the Kargil conflict of 1999, the Government constituted the Kargil Review Committee under the chairmanship of K. Subrahmanyam to examine the events leading to the conflict and the state of the national security system. The Committee reported at the end of 1999, and the Government then constituted a Group of Ministers, chaired by the Home Minister, L. K. Advani, on 17 April 2000, to consider the Committee’s recommendations and formulate proposals for implementation. The Group’s report, titled Reforming the National Security System, was submitted to the Prime Minister on 26 February 2001 and approved by the Government in May 2001.

What did the Group of Ministers recommend on identity cards in 2001?

The Group of Ministers recommended compulsory registration of citizens and non citizens living in India, a national register of citizens prepared from that registration, a multi purpose national identity card for every citizen and differently designed cards for non citizens, beginning in the border districts and extending across the country in stages, at the cost of the Central Government.

The passage of the Group’s report dealing with illegal immigration, at paragraph 5.110, is the direct ancestor of Section 14A. It recorded that illegal migration had assumed serious proportions and proposed registration and documentation as the remedy: a State that could not say who on its territory was a citizen and who was not could neither detect unlawful presence nor protect lawful status, and a register with a card was presented as the administrative foundation on which every other measure would rest. The Kargil Review Committee had itself urged identity cards for border villagers in vulnerable areas as a priority measure. What is significant for the trail is where the recommendation was implemented. The recommendation was not left on paper even before the statute was amended. The Government approved the Group’s report in May 2001, and a pilot project for the issue of multi purpose national identity cards was taken up in a limited number of sub districts in selected states and union territories, chosen to include border areas of the kind the Kargil Review Committee had singled out, with cards issued to residents above the prescribed age in the project areas after enumeration and verification. The pilot was small against the scale of the recommendation, and it was not converted into a general issue of cards to citizens, but it established the administrative template, enumeration, verification, card, that Section 14A and the 2003 Rules would shortly afterwards write into citizenship law for the whole country. The register and the card could have been placed in a census statute, a foreigners statute or a new registration statute. They were placed, two years later, in the Citizenship Act, as a new Section 14A, by the same amendment that defined the illegal migrant and tightened the birth rule. The placement was itself a policy statement: the register was to be a register of citizens, compiled under citizenship law, by a Registrar General of Citizen Registration, and the card was to be a national identity card issued on compulsory registration of every citizen. The 2003 amendment therefore joined three things that had previously been separate: the test for acquiring citizenship, the definition of the person barred from acquiring it, and the machinery for recording who had it.

The amendment of 2003: the pivot of the trail

The Citizenship (Amendment) Act, 2003 is cited as Act 6 of 2004, because although Parliament passed it in December 2003 it received the assent of the President on 7 January 2004 and bears the number of the year of its publication. It came into force on 3 December 2004. The gap between its short title, its Act number and its commencement date is the single most common source of citation error in this territory, and it is also a useful discipline: the amendment is the 2003 amendment by title, Act 6 of 2004 by number, and operative for births and applications from 3 December 2004. It was introduced in Parliament by L. K. Advani, by then Deputy Prime Minister as well as Home Minister, the same minister who had chaired the Group of Ministers whose report supplied its register provisions. The Bill, numbered Bill 39 of 2003, was introduced in the Rajya Sabha on 7 May 2003. It was passed by the Rajya Sabha on 18 December 2003 and by the Lok Sabha on 22 December 2003, in the closing days of the winter session, and the interval between its passage and its commencement was nearly a year, because the provisions it inserted, the register machinery and the overseas scheme among them, required rules, schedules and administrative preparation before they could operate. That interval is itself instructive for readers of the trail: an amendment’s date of passage, its date of assent and its date of commencement are three different facts with three different legal significances, and this amendment separates all three more widely than any other in the sequence, which is why its citation has generated more error than the rest of the trail combined. It made four changes of the first importance, and several of consequence, and each is taken in turn.

The first change narrowed the birth route for the second time. Section 3 was substituted wholesale, and its new clause (c) provided that a person born in India on or after the commencement of the amendment is a citizen by birth only where both parents are citizens of India, or where one parent is a citizen of India and the other is not an illegal migrant at the time of the birth. The architecture of the test deserves attention. The 1986 test had asked one question: is either parent a citizen? The 2003 test asks that question and a second one: is the non citizen parent an illegal migrant? A citizen parent plus a foreign parent lawfully present satisfies the test. Two citizen parents satisfy it. A citizen parent plus an illegal migrant parent fails it, and two foreign parents fail it however lawful their presence. The lawfulness of a parent’s presence thus entered the birth rule for the first time, seventeen years after parentage itself had entered it.

What is the position of a child born in India after 2004 if one parent is an illegal migrant?

A child born in India on or after 3 December 2004, one of whose parents is a citizen and the other an illegal migrant, is not a citizen by birth. Section 3(1)(c) is satisfied only where both parents are citizens, or one parent is a citizen and the other is not an illegal migrant; that child meets neither limb.

The second change created the category on which the first change depends. A new clause in Section 2(1), clause (b), defined an illegal migrant as a foreigner who has entered India without a valid passport or other travel documents and such other document or authority as may be prescribed, or who has entered with valid documents but remains in India beyond the permitted period of time. The definition turns on documentation and on time: entry without papers, or presence beyond permission. It does not turn on the reason for entry, the country of origin, or the length of residence after entry, and a person who had lived in India for decades on an expired permission fell within it on the same terms as a person who had crossed the border the previous week. Detection of such a person, and the orders that follow detection, remained matters for the foreigners law: it is under the Foreigners Act, 1946 explained that a person is determined to be a foreigner and subjected to restriction or removal, and the Supreme Court’s decision in Sarbananda Sonowal v. Union of India, (2005) 5 SCC 665, which struck down the separate tribunal statute for Assam and restored the 1946 Act as the governing instrument there, belongs to that side of the line. The 2003 amendment’s contribution was not detection machinery. It was to attach citizenship consequences to the defined status, and to do so at every entry point of the Act.

The attachment of consequences was done in three places. Section 5, on registration, was recast so that the Central Government may register as a citizen a person not being an illegal migrant who falls within the listed categories; the words operate at the threshold of the section, so an illegal migrant cannot be registered even if the category itself, such as marriage to a citizen or Indian origin with seven years of ordinary residence, is made out. Section 6, on naturalisation, was amended in parallel, so that a certificate of naturalisation cannot be granted to an illegal migrant. Section 3, as described above, made the status decisive for the next generation as well. Before 2003, unlawful presence was a matter between the foreigner and the foreigners law, with expulsion as its sanction and no necessary consequence for citizenship applications, which were assessed on their own criteria. After 3 December 2004, unlawful presence as defined became a permanent disqualification from the two discretionary routes and a disqualifying fact in the automatic route. No other amendment in the trail changed the legal meaning of a fact so completely.

The drafting of clause (c) contains a choice that analysis of the amendment often passes over, and it is worth isolating because it determines who the clause actually catches. Parliament could have required that the non citizen parent be lawfully resident in India. It did not. It required that the non citizen parent not be an illegal migrant, a negative test set against the Section 2(1)(b) definition. The difference between the two formulations is the whole population of foreigners whose position is irregular in some respect short of the definition, or regular under a status the definition does not reach: the foreign parent who entered lawfully and holds a subsisting permission satisfies both formulations, and the foreign parent who entered without documents or overstayed fails both. Between those poles lie cases the negative test was designed, or can be taken, to spare: the parent whose papers are in process, the parent holding a status under a protection or exemption regime, the parent whose documentation is disputed but who has not been determined to have entered without authority or to have overstayed a permitted period. A positive lawfulness test would have put the burden on the family to establish the parent’s regular status in every case, including the uncontentious ones. The negative test puts the disqualification only where the defined status is established. That is a narrower exclusion than the rhetoric of the amendment suggested, and it matters in practice, because Section 3 operates at the registration of a birth, where the question asked is not whether a parent can prove lawful residence but whether anything on the record shows the parent to be an illegal migrant. The clause is severe in its central case, the child of a citizen and a person who entered without documents, and deliberately less severe at its edges. Both the severity and the limit are in the drafting, and accounts of the amendment that report only one of them misdescribe the instrument.

The third change touched descent, and it is the least noticed of the four. The 2003 Act added a further proviso to Section 4 under which a person born outside India on or after its commencement is a citizen by descent only if the birth is registered at an Indian consulate, in the prescribed form and manner, within one year of the birth or of the commencement of the amendment, whichever is later, or afterwards with the permission of the Central Government; and no such birth may be registered unless the parents declare that the minor does not hold the passport of another country. The direction of this change runs against the liberalising amendment of 1992, and the two together describe the trail in miniature. In 1992 Parliament widened who could transmit citizenship abroad, from the father to either parent. In 2003 Parliament tightened how the transmission had to be perfected, by requiring a timely act of registration and a declaration against a foreign passport, so that descent became easier to qualify for and easier to lose by inaction. A diaspora that was being courted with one hand, through the overseas category created in the same Act, was being required with the other hand to register its children promptly or forfeit their claim.

The fourth change was Section 14A, the register and the card. The new section provided that the Central Government may compulsorily register every citizen of India and issue a national identity card to each citizen; that it may maintain a National Register of Indian Citizens and establish a National Registration Authority for that purpose; that the Registrar General, India, appointed under the Registration of Births and Deaths Act, 1969 (Act 18 of 1969), would act as the National Registration Authority and function as the Registrar General of Citizen Registration; and that the procedure for compulsory registration would be as prescribed. Rules followed within the year: the Citizenship (Registration of Citizens and Issue of National Identity Cards) Rules, 2003, which set out how a register would be built, through a house to house enumeration for a National Population Register, verification, the preparation of local, sub district, district and state registers, and a claims and objections process in which a person whose citizenship was in doubt could be marked for further inquiry. The distinction the Rules drew, between a population register recording usual residents and a register of citizens recording citizens, is examined below, because the machinery was in the event operated on the population side of that line. Section 14A is also the point at which identity documentation enters citizenship law, and it should not be confused with the later and quite separate identity programme built on biometric enrolment under the Aadhaar Act, 2016: the Section 14A card was a card of citizenship, to be issued after a determination of status, where the Aadhaar number is issued on residence and expressly is not proof of citizenship. Readers comparing the amendment texts with the rules made under them will find the two instruments repay side by side reading, and the office file viewer for working through the amendment Acts and the 2003 Rules in document form is the practical tool for that comparison.

Two lesser changes of the 2003 Act complete the record. The Act omitted the provisions of the principal Act that had recognised Commonwealth citizenship, a status of diminishing content that the 1955 scheme had inherited from the constitutional arrangements of 1950, and removed the First Schedule list of Commonwealth countries that went with it. And the Act lengthened the qualifying residence for registration of a person of Indian origin to seven years of ordinary residence in India, completing a progression that the 1986 Act had begun when it raised the same figure from six months to five years. Neither change attracted debate at the time. Both belong in a complete trail, because the first closed a category that had existed since 1955 and the second shows the tightening of the registration route proceeding in step with the tightening of the birth route.

The register in practice: a population register, not a citizen register

Section 14A has stood on the statute book since 3 December 2004. The use made of its machinery is a matter of record, and the record needs to be stated structurally, because the gap between the provision and its operation is itself one of the findings of the trail. The 2003 Rules contemplate a sequence: enumeration of usual residents into a National Population Register, verification of the entries, separation of citizens from non citizens through a process that includes marking doubtful cases and hearing claims and objections, and the issue of national identity cards to those determined to be citizens. The enumeration stage was carried out on a national scale, alongside the house listing operations of the Census of 2011, and the population register was prepared and later updated. The later stages of the sequence, the determination exercise that would convert a register of residents into a register of citizens for the country as a whole, and the general issue of identity cards under Section 14A, were not carried out on a national scale. One state level exercise in citizenship determination did take place, in Assam, where a register of citizens was updated under the special provisions of the same 2003 Rules that govern Assam, under the supervision of the Supreme Court, with a final list published on 31 August 2019; that exercise rested on the Section 6A framework and the legacy of the Accord rather than on a general national determination, and its conduct and consequences are part of the Assam story rather than of the general trail.

The population register that was built deserves a precise description, since it is the largest administrative artefact of the 2003 amendment. A usual resident, for the purposes of the National Population Register, is a person who has resided at a place for the six months preceding the enumeration or intends to stay at that place for the six months following it, a definition borrowed from census practice and deliberately indifferent to nationality and to the lawfulness of presence. The particulars collected in the house to house enumeration of 2010, conducted with the house listing operations of the Census of 2011, were demographic: name, relation to the head of household, sex, date of birth, marital status, education, occupation and the names of parents and spouse, with biometric particulars gathered in a parallel exercise. Nothing in that enumeration asked the enumerator to decide, or even to record an opinion, whether the resident was a citizen. The register that resulted is therefore a register of the resident population in the fullest sense, citizens, lawful foreigners and unlawful entrants recorded on the same footing, and its value to the State is the value of a population database, not of a citizenship determination. The updating of the register undertaken in 2015 proceeded on the same basis. When public argument speaks of a national register prepared under the 2003 framework as though a citizenship sorting of the country had already occurred, it confuses the first stage of Rule 4 with the whole of it. The sorting stages, verification against citizenship, doubtful marking, hearings and the preparation of the register of citizens from the population register, remain, at the national level, unperformed, and the identity cards of Section 14A remain, at the national level, unissued. The amendment built the doorway to a register of citizens. What has been carried through the doorway is a register of residents.

The position, stated without ornament, is therefore this. Since 2003 the Citizenship Act has contained a complete statutory basis for compulsory registration of citizens, a national register of citizens and national identity cards. Since 2003 the machinery actually operated under that basis, nationally, has been a population register, which records who usually resides at an address and makes no determination of citizenship at all. Arguments that treat a register of citizens as an innovation of 2019 are wrong about the date by sixteen years. Arguments that treat the register of citizens as an accomplished national fact are wrong about the operation: what exists nationally is the population register, and the conversion of the one into the other would require the determination stages of the 2003 Rules to be run, with their verification, doubtful marking and objection hearings. The provision and the practice are both part of the record, and the trail falsifies any account that reports only one of them.

The detail of the 2003 Rules rewards a closer statement, because the Rules are where the difference between the two registers is enacted. Rule 3 of the Citizenship (Registration of Citizens and Issue of National Identity Cards) Rules, 2003 provides for the National Register of Indian Citizens, to consist of registers at the local, sub district or taluk, district, state and national levels, and to be prepared under the direction of the Registrar General of Citizen Registration. Rule 4 prescribes the procedure: an enumeration of the population, conducted through house to house visits by a local registrar, producing a National Population Register of usual residents; verification of the particulars collected; the preparation of a draft local register; the marking of entries whose citizenship is doubtful, with notice to the persons concerned; a period for claims and objections; hearings by the sub district registrar; an appeal; and only after that process, the separation of citizens from others and the issue of identity cards. The architecture is instructive. Citizenship determination under the Rules is not a single act of inspection but a cascade of administrative stages, each with its own officer and its own remedy, and the doubtful marking is the hinge on which the whole cascade turns, because it is the stage at which a resident recorded in the population register is singled out for inquiry into status. Everything that has been argued about registers of citizens since 2003, about burden, about documentation, about the consequences of error, is an argument about how that hinge stage would operate in practice. Nationally, it has not operated at all: the enumeration and population register stages were run, the doubtful marking and determination stages for the country as a whole were not, and the distance between the two halves of Rule 4 is the precise measure of the distance between the statute’s promise and its performance.

The 2003 amendment argued both ways

The 2003 amendment is the pivot of the trail, and the stance of this series requires that the case for it and the case against it each be stated at full strength before any verdict is given. The case for the amendment, as made by the Government that carried it and by the security review that prompted its register provisions, rests on the condition of the eastern border and on the logic of documentation. Unregulated migration across a long and porous border had, on this account, produced a large undocumented population whose size no official figure could state, precisely because the State had no register from which to count. The Kargil Review Committee and the Group of Ministers had identified the absence of compulsory registration and of identity documentation as a vulnerability in the security system, and the Group’s report said so in terms. A definition of illegal migrant was, on this account, a precondition of any lawful administration: without it, the Act could not distinguish the person entitled to apply for citizenship from the person whose presence was itself unlawful, and every route into citizenship stood open in principle to a person who had entered in defiance of the law governing entry. The tightening of the birth rule followed from the same premise. A rule that granted citizenship to the child of two persons with no right to be in India allowed the fact of an unlawful entry to generate, one generation later, a status that no authority had ever granted. On the strongest version of this case, the 2003 amendment did not withdraw anything from anyone lawfully present: the foreign parent lawfully resident in India continued to satisfy the birth test with a citizen spouse, and the lawful resident continued to qualify for registration and naturalisation on the stated conditions. What changed was the position of unlawful presence, which the Act had previously ignored.

The case against the amendment, as made by its critics in Parliament and in the scholarship that followed, accepts none of that framing. The historian Anupama Roy, in her study Citizenship in India, described the amendment as a hinge point in Indian citizenship, the moment from which two tendencies that had previously been latent, a documentary and exclusionary citizenship at home and an expansive citizenship for the diaspora abroad, became the organising features of the law; the description has been widely taken up because it fits the text, which did both things in a single Act. On the strongest version of the critical case, the vice of the amendment lies in the combination it created. A class of persons was defined by want of documents and barred from every route to citizenship, including the route of birth for their children. In a country where documentary proof of an earlier generation’s status is unevenly held, where birth registration itself was historically incomplete, and where the poor hold the fewest papers, a regime that sorts by documents will not confine its effects to persons who crossed a border. It will catch the undocumented citizen in the same net as the undocumented entrant, and the doubtful marking procedure of the 2003 Rules places the burden of escaping the net on the person marked. The critics’ sharpest point concerns the destination of the exercise. A person determined not to be a citizen of India does not thereby become a citizen of any other state, and a neighbouring state that does not accept the person as its own will not receive the person back. Exclusion from citizenship, on this account, produces not removal but statelessness inside the territory: a population without status, without the franchise, and without a lawful future for its children, who inherit the disqualification at birth under clause (c). That consequence, the critics say, was not an accident of the drafting but the predictable operation of a definition built on papers in a society built substantially without them.

The verdict this article reaches is narrower than either case, because the trail itself supplies it. On the character of the amendment there is no serious room for doubt: 2003 was the most consequential amendment in the sequence, the point at which the Act acquired its barred category, its register and its diaspora status together, and Roy’s description of it as a hinge is borne out by the table that appears later in this article, in which three of the five rows change their content at the 2003 column. On the operation of the amendment, the record since 2004 supports a distinction both camps neglect. The exclusionary machinery has operated mainly as definition and deterrent rather than as mass determination: the register nationally compiled was a population register, and the one completed citizen determination, in Assam, proceeded under a state specific framework older than the amendment. The diaspora machinery of the same Act, by contrast, was operated at once, widened within two years and consolidated within twelve, and it is the side of the 2003 Act that has touched the largest number of lives. A reader who wants to know what the 2003 amendment did should therefore look less at what it threatened than at what was built under it, and what was built, on both sides of the line, was documentary: a State that proposed to know its citizens by register, and a diaspora invited to hold its connection by card.

The discretionary routes across the trail: registration and naturalisation

The birth and descent rules dominate discussion of the amendments because they operate automatically, but the trail also worked steadily on the two discretionary routes, and a complete account has to follow them. Registration under Section 5 is the route of connection: it is available to a person of Indian origin ordinarily resident in India, to a person of Indian origin resident outside undivided India, to a person married to a citizen of India, to minor children of citizens, to a person whose parents are registered as citizens under the section, to a person who or whose parent was earlier a citizen of independent India and who has been residing in India for one year before applying, and, after the overseas scheme was built, to a person registered as an overseas cardholder for five years who has been residing in India for the prescribed period before applying. The categories have been stable in outline since 1955. What the trail changed, repeatedly, was the price of admission in years. The Statement of Objects and Reasons of the 1986 Act recited that a person of Indian origin had needed six months of ordinary residence before applying, and raised the requirement to five years. The recast of Section 5 in 2003 set the figure for the same category at seven years, where it stands. A qualification that began, in 1955, as a formality of settlement ended, across two amendments, as a residence requirement of real substance, and the progression six months, five years, seven years is the registration route’s equivalent of the birth route’s progression from no condition to the two part test. The same 1986 Act that began the progression also widened the marriage category from women to persons, and the 2003 recast of Section 5 is the text that placed the illegal migrant bar at the head of the section, so that the section’s opening words exclude the barred class before any category is reached. The 2015 Act then adjusted the section’s language for the merged cardholder scheme, including the substitution of a twelve months ordinary residence formulation in the spouse category and the alignment of the overseas cardholder category with the single scheme.

Naturalisation under Section 6 followed a parallel course through the Third Schedule. The 1986 Act raised the residence qualification for an applicant who was not a citizen of a Commonwealth country from five years to ten, on the reasoning stated in its objects. The 2003 Act’s omission of the Commonwealth citizenship provisions and of the First Schedule removed the distinction between Commonwealth and other applicants from the Act altogether, and the Third Schedule in its operative form states a single set of qualifications for every applicant: that the applicant is not a citizen of a country whose law prevents its citizens, or Indian citizens by naturalisation, from the relevant reciprocity; that the applicant, if a citizen of another country, has renounced that citizenship in the manner required; that the applicant has resided in India or been in government service in India throughout the twelve months before applying, and during the fourteen years preceding those twelve months has resided or served for periods amounting in aggregate to not less than eleven years; that the applicant is of good character; that the applicant has an adequate knowledge of a language specified in the Eighth Schedule to the Constitution; and that the applicant intends, if granted the certificate, to reside in India or to enter or continue in service under the Government or an international organisation of which India is a member or a society established in India. The eleven year aggregate, against which the 2019 amendment’s reduction to five years for its defined class is measured, is the product of this settled schedule rather than of any single amendment, and care is needed in attributing it: the trail’s contribution to naturalisation was the raising of periods in 1986, the removal of the Commonwealth distinction in 2003, the bar on the illegal migrant, and the single class exception of 2019. Between those points the route stood still, which is itself a datum: Parliament’s repeated attention fell on birth, descent, registration and status, and the naturalisation door, once its residence price had been set high, was left alone for three decades.

A word should also be said about the Commonwealth dimension, because its disappearance is the trail’s quietest change and the one that best marks how far the statute travelled from its origins. The Act of 1955 was drafted in a period when Commonwealth citizenship was a living legal idea: the Constitution itself had been framed for a republic that remained within the Commonwealth, and the Act as enacted provided that every citizen of India would also have the status of a Commonwealth citizen, recognised a corresponding status for citizens of the countries listed in the First Schedule, and gave Commonwealth citizens advantages in the discretionary routes, including the shorter naturalisation residence that the 1986 Act raised from five years to ten for applicants outside that charmed circle. The 2003 Act omitted the Commonwealth citizenship provisions and the First Schedule together. By 2003 the status conferred little in Indian law and less abroad, and its removal attracted no debate, but the omission completes the trail’s logic in a way the headline changes do not. The Act of 1955 had organised citizenship partly by a supranational affiliation inherited from the empire. The Act as amended organises it by parentage, by documentation and by a diaspora status of India’s own design. The Commonwealth citizen, the person of Indian origin and the overseas cardholder are three successive answers to the question of what connection short of citizenship the law should recognise, and only the third is a connection India invented for itself.

Determination contested: the courts, the rolls and the burden of proof

A trail of amendments about status generates litigation at its edges, and the case law that matters here is not about the amendments’ validity, which was never seriously in issue for the general amendments, but about determination: how a person’s citizenship is to be decided when it is disputed, and what process is owed before an adverse conclusion is recorded. The recurring setting is the electoral roll, because inclusion in a roll is the most visible everyday consequence of citizenship and because roll revision exercises have repeatedly been the occasion on which large numbers of residents were asked to prove status. In Lal Babu Hussein v. Electoral Registration Officer, decided in 1995, the Supreme Court considered a revision of electoral rolls in which names had been deleted on the footing that the persons named were not citizens. The Court held that an electoral registration officer could not treat a person as a non citizen on suspicion or on the report of an enumerator: a person whose name stood on the roll was entitled to an inquiry, with notice and an opportunity to produce material, before deletion, and citizenship, once enjoyed, was not to be displaced by an administrative assumption. The decision belongs to the law of elections rather than to the Citizenship Act, but its principle runs through the whole territory of this article, because it fixes the point at which the documentary turn of the amendments meets the law of procedure: a statute may define status by documents and parentage, and a register may record the result, but the determination of an individual’s status remains an act that must be done fairly, person by person.

The foreigners side of determination was settled, for Assam and by extension for the country, by the Sonowal judgment already noticed, which struck down the special tribunal Act of 1983 and left the Foreigners Act, 1946 as the governing instrument, with its rule in Section 9 that the burden of proving that a person is not a foreigner lies on the person. The juxtaposition of the two lines of authority is the honest summary of the legal position the trail produced. Under the citizenship and elections law, status once held is protected by procedural fairness and by the prospective drafting of the amendments themselves. Under the foreigners law, a person proceeded against as a foreigner carries the burden of proving the negative. Which side of that line a disputed case falls on is determined by its history: whether the person claims a status the statute conferred at birth, or is met by a proceeding that treats the person as an entrant to be detected. The amendments of the trail did not create that duality, which is older than all of them, but they sharpened its consequences, because after 2003 the outcome of a foreigners determination governed not only removal but the citizenship of the person’s children and the availability of every route back. A reader who understands why the same individual can be, in one proceeding, the holder of a status the State must disprove fairly and, in another, the bearer of the burden of proving it, understands the practical law of Indian citizenship better than a reader who knows every date in the trail by heart.

The diaspora side: from the PIO card to overseas citizenship

The overseas story begins before the statute moved, with an administrative scheme and a committee. The Person of Indian Origin card scheme was formulated in 1999 and revised in 2002. It was not a creature of the Citizenship Act at all: it was an executive scheme under which a person of Indian origin holding a foreign passport could obtain a card giving visa free travel to India for fifteen years and exemption from the registration requirements that otherwise applied to foreigners for stays up to one hundred and eighty days. It conferred no status under citizenship law and was not registered under the Act. Its significance for the trail is that it established the pattern of the card as the instrument of diaspora policy, and its limitations, a fixed term, a fee, and benefits that fell short of what the diaspora organisations were asking for, set the agenda for what followed.

The agenda was written by the High Level Committee on the Indian Diaspora, constituted in September 2000 under the chairmanship of L. M. Singhvi. The Committee reported in January 2002 and recommended that the Citizenship Act be amended to provide for the grant of dual citizenship to persons of Indian origin belonging to certain specified countries. Article 9 of the Constitution stood in the way of dual citizenship in the literal sense, and what the 2003 Act enacted was the closest available substitute, deliberately named to stop short of citizenship itself. The Act inserted Sections 7A to 7D into the Citizenship Act and the definition of an overseas citizen of India in Section 2(1). A person of Indian origin who was a citizen of a specified country, or who had been a citizen of India immediately before acquiring that citizenship, could be registered as an overseas citizen of India under Section 7A. The specified countries were those listed in a new Fourth Schedule, sixteen countries at the outset, chosen because their own laws permitted their citizens to hold another citizenship or status of the kind proposed. Registration brought a package of rights under Sections 7B and 7C: parity with non resident Indians in economic, financial and educational fields, a lifelong visa facility in practice, and exemption from the foreigners registration regime, balanced by express exclusions. An overseas citizen was not entitled to the rights of a citizen under Article 16 of the Constitution, equality of opportunity in public employment, and the scheme carried no vote, no eligibility for constitutional office and no right to acquire agricultural or plantation property on the same footing as a citizen. The status was a documentary connection, substantial in daily life and deliberately incomplete in constitutional law.

The sixteen countries of the Fourth Schedule were Australia, Canada, Cyprus, Finland, France, Greece, Ireland, Israel, Italy, the Netherlands, New Zealand, Portugal, Sweden, Switzerland, the United Kingdom and the United States of America. The list is a map of the older Indian emigration as the Government of 2003 understood it: the Anglophone countries of settlement, the states of western Europe with established Indian communities, and Israel. It omitted the Gulf states, where the largest number of Indian workers lived, because those states do not offer their residents a citizenship to hold alongside an Indian status; it omitted East and South East Asia and the Caribbean and Pacific countries of the older indenture diaspora, whose connection to India was as old as any on the list; and it omitted South Africa. The criterion was never antiquity of connection. It was the law of the other country: whether that law permitted its citizens to hold a second citizenship or a status in the nature of one, since a scheme built to stop short of dual citizenship still required the other state to tolerate the connection. The narrowness of the list, and the accidents of emigration history it froze into a schedule, explain why the list could not survive. A diaspora policy that reached a person of Indian origin in Toronto and London but not a person of Indian origin in Durban, Georgetown or Kuala Lumpur was indefensible on its own stated premise of connection, and the correction of 2005, which replaced the list with descent from the constitutional starting point, was driven by that indefensibility rather than by any change in the security climate.

The specified country device lasted two years. The Citizenship (Amendment) Act, 2005 (Act 32 of 2005), in force from 28 June 2005, recast the eligibility for overseas citizenship by substituting Section 7A. Under the substituted provision, overseas registration was opened to a person who was a citizen of India on 26 January 1950 or at any time thereafter, or who was eligible to become a citizen of India on that date, or who belonged to a territory that became part of India after 15 August 1947, together with the children, grandchildren and great grandchildren of such a person, and to certain spouses, in each case provided the person was a citizen of a country other than Pakistan or Bangladesh or a country notified by the Central Government for the purpose. The Fourth Schedule and its list of sixteen countries were omitted, because a list was no longer needed: eligibility ran by descent from the constitutional starting point rather than by the nationality the person happened to hold. The movement of 2005 is easy to miss because it changed no headline category, but in the geography of the diaspora it was the largest widening of the three. The 2003 scheme had reached persons of Indian origin in sixteen named states. The 2005 scheme reached a person of Indian descent in any state in the world, the Gulf, East Africa, the Caribbean and South East Asia included, subject to the Pakistan and Bangladesh exclusion, which has remained a constant of the overseas provisions from their first enactment and reflects the particular history of Partition that the whole subject carries.

For a decade after 2005, two diaspora documents coexisted: the PIO card, administrative in origin, with a fifteen year term, and the overseas citizen registration, statutory in origin, lifelong in effect and carrying the fuller package of rights. The coexistence confused applicants, produced anomalies in fees and validity, and was ended by the Citizenship (Amendment) Act, 2015 (Act 1 of 2015). That Act had begun as an ordinance promulgated on 6 January 2015, and the Act itself, which received assent on 10 March 2015, was deemed to have come into force on 6 January 2015. By notification of 9 January 2015 the PIO card scheme ceased, and every existing PIO cardholder was deemed to be an overseas citizen cardholder from that date. The Act then rewrote the statutory vocabulary: Section 2(1)(ee) was substituted to define an Overseas Citizen of India Cardholder, and Sections 7A to 7D were recast in the language of the cardholder, so that the Act speaks of one category, registered under Section 7A, holding the rights in Section 7B and liable to cancellation under Section 7D on grounds that include fraud, disaffection towards the Constitution, unlawful trade or communication with an enemy in wartime, a sentence of imprisonment of two years or more within five years of registration, and the interests of the sovereignty, security and public order of India. The 2015 Act also relaxed conditions on the citizenship side of the ledger in a detail that completes the picture of its purpose: the registration route in Section 5 was adjusted for overseas cardholders and spouses, aligning the waiting and residence conditions with the merged scheme, so that the cardholder who later seeks full citizenship by registration faces a single coherent set of provisions rather than the residue of two schemes.

The merger had a political prelude that belongs in the record. The proposal to merge the two schemes had been under discussion for years on the ground of administrative confusion, and it was announced publicly by the Prime Minister, Narendra Modi, on 28 September 2014, during an address to the Indian diaspora in New York, as part of a package of measures directed at that audience. The ordinance followed in January 2015, timed to the Pravasi Bharatiya Divas convention of that year, and the gazette notification of 9 January 2015 converted every PIO cardholder into a deemed overseas cardholder on that date, without application, fee or fresh scrutiny. The manner of the change illustrates the character of the diaspora side of the trail. Where the domestic side proceeded by conditions, definitions and bars, each requiring the individual to prove a position within them, the diaspora side proceeded by announcement, ordinance and deeming provision, conferring the better status on an entire class at a stroke. The two techniques are visible in the statute side by side, and the contrast between them is as much a part of the pattern this article records as the substantive direction of the changes themselves.

What rights does overseas citizenship not confer on a cardholder?

Overseas citizenship confers no vote and no right to stand for Parliament or a state legislature, no eligibility for constitutional offices, President, Vice President or judge of the Supreme Court or a High Court, and no right under Article 16 to equality of opportunity in public employment. It is a status with civic and economic facilities, not citizenship.

The content of the cardholder package should be stated plainly, because argument about overseas citizenship often proceeds without it. A registered cardholder holds a multiple entry, multi purpose lifelong visa for visiting India and is exempt from registration with the Foreigners Regional Registration Officer for any length of stay. The cardholder is placed on a par with non resident Indians in financial, economic and educational fields, other than in the acquisition of agricultural and plantation properties, and enjoys parity with resident Indians in domestic air fares and in entry fees for national parks and wildlife sanctuaries, among the facilities notified under Section 7B. Against that package stand the exclusions: no franchise, no membership of Parliament or of a state legislature, no constitutional office, no claim under Article 16 to equality of opportunity in public employment, and no appointment to public services other than as the Central Government may specify. The status is therefore best described as a right of abode with economic parity, held by a person who remains, in law, a citizen of another state. Its attraction to the diaspora was never the name. It was the removal of the recurring frictions of the foreigner’s position in India, the visa renewals, the registration visits, the uncertainty of re entry, exchanged for a single registration that does not expire. Measured by uptake and by the politics of its announcement, the exchange was one the diaspora valued highly, and the amendments of 2005 and 2015 are best read as the State successively improving the terms of that exchange while conceding nothing on the constitutional exclusions.

The overseas sequence, taken with the birth sequence, is the evidence for the article’s central claim. The amendments that narrowed entitlement inside India and the amendments that widened the diaspora category are not two separate stories that happen to share a statute. They are, at the decisive point, the same instrument: the Act of 2003 that defined the illegal migrant and barred that person from registration and naturalisation is the Act that created overseas citizenship, and the Parliaments that maintained the birth conditions of 1986 and 2003 raised no obstacle to the widenings of 2005 and 2015. The statute was becoming more restrictive towards persons present in India without documents and more accommodating towards persons of Indian descent living abroad, and those two movements were carried, repeatedly, in the same amending Acts and by the same governments. A description of the trail as restriction alone misses half of what the statute was doing. A description of it as diaspora liberalisation alone misses the other half. The table that follows is the trail arranged so that both movements can be seen at once.

Five amendments, one direction: the trail as a table

Position after each amendment	Amendment of 1986	Amendment of 1992	Amendment of 2003	Amendment of 2005	Amendment of 2015
Birth route	One citizen parent required for births from 1 July 1987	Unchanged	Both parents citizens, or one citizen and the other not an illegal migrant, for births from 3 December 2004	Unchanged	Unchanged
Descent route	Unchanged	Either parent transmits for births from 10 December 1992	Consular registration within one year and a declaration against a foreign passport required for births from 3 December 2004	Unchanged	Unchanged
Illegal migrant	No such category	No such category	Defined in Section 2(1)(b) and barred from registration and naturalisation	Unchanged	Unchanged
Register and identity cards	No provision	No provision	Section 14A inserted: compulsory registration, National Register of Indian Citizens, national identity cards; Rules made in 2003	Unchanged	Unchanged
Diaspora category	No category	No category	Overseas citizenship created for persons of Indian origin in sixteen specified countries	Opened to persons of Indian origin in all countries except Pakistan and Bangladesh and notified countries	PIO card merged into a single Overseas Citizen of India Cardholder category

Read the table by rows rather than by columns and the argument of this article makes itself. The birth row moves in one direction only, from no condition before 1986 to a citizen parent after 1987 to the two part test after 2004, and it never moves back. The diaspora row moves in the opposite direction only, from nothing before 2003 to a sixteen country scheme, to a near universal scheme, to a consolidated card, and it never moves back either. The middle rows explain how the two movements are connected. The illegal migrant row and the register row both change at the 2003 column and nowhere else, because the documentary turn of that year supplied the concepts the other rows needed: the birth test of clause (c) cannot be administered without the definition in Section 2(1)(b), and the claim that a register of citizens was thinkable in 2019 rests on a provision inserted in 2003. The descent row is the subtlest of the five, widening in 1992 and tightening in its conditions in 2003, and it is the row that best deserves the series description of a sequence described in one direction moving in two. No single column of the table tells the story of the statute. The story is the pattern of the rows.

The amendment of 2019: the end point, stated mechanically

The sequence closed, for the purposes of this trail, with the Citizenship (Amendment) Act, 2019 (Act 47 of 2019). It received the assent of the President on 12 December 2019 and came into force on 10 January 2020. Its operative mechanics are three. It inserted a proviso into the definition of illegal migrant in Section 2(1)(b), under which a person belonging to the Hindu, Sikh, Buddhist, Jain, Parsi or Christian community in Afghanistan, Bangladesh or Pakistan, who entered India on or before 31 December 2014 and who has been exempted by the Central Government under the Passport (Entry into India) Act, 1920 or the Foreigners Act, 1946, is not to be treated as an illegal migrant for the purposes of the Act. It inserted Section 6B, empowering the Central Government or a specified authority to grant a certificate of registration or of naturalisation to a person covered by that proviso, subject to conditions, and providing that such a person would be deemed a citizen from the date of entry into India. And it amended the Third Schedule so that the aggregate residence qualification for naturalisation, eleven years in the period preceding the twelve months before application, stands reduced to five years for a person covered by the proviso. The same Act added a ground of cancellation to Section 7D, violation of the provisions of the Citizenship Act or of other laws notified by the Central Government, with a safeguard that no cancellation order be passed without giving the cardholder a reasonable opportunity of being heard. Each of these provisions engages questions of equality, religion and the Assam settlement that are argued in full in the article on the Citizenship Amendment Act 2019 debate, and this article takes no part in that argument. Its place here is narrower: it is the first amendment in the trail to work by carving an exception out of the illegal migrant definition of 2003, which confirms, from the direction of its own drafting, that the definition and the bar built in 2003 are the load bearing provisions of the modern Act.

What residence period does naturalisation require under the Third Schedule?

Naturalisation under Section 6 requires the qualifications in the Third Schedule, among them residence in India throughout the twelve months before application and, in the fourteen years preceding those twelve months, residence amounting in aggregate to not less than eleven years. For a person covered by the proviso inserted in 2019, the aggregate period stands reduced to five years.

A note on citing the amendments

Because the trail is so often misdated, a short note on citation is a practical necessity rather than a bibliographical luxury. An Indian Act is cited by its number in the year of its publication in the Gazette, which is not necessarily the year in its short title. The Citizenship (Amendment) Act, 2003 is Act 6 of 2004. The Citizenship (Amendment) Act, 2015 is Act 1 of 2015, and its short title year, publication year and deemed commencement, 6 January 2015, happen to coincide, which is why it causes little trouble. The Citizenship (Amendment) Act, 1986 is Act 51 of 1986, the Act of 1992 is Act 39 of 1992, the Act of 2005 is Act 32 of 2005, the Assam Act of 1985 is Act 65 of 1985, and the Act of 2019 is Act 47 of 2019. A citation that gives the short title and the Act number together, Citizenship (Amendment) Act, 2003 (Act 6 of 2004), is unambiguous in a way that neither element is alone. The second discipline concerns commencement. The operative date of a change in this trail is the date from which the changed rule governs the event in question, and the sections themselves state it in their date bands: 1 July 1987 for the first birth condition, 10 December 1992 for descent through either parent, 3 December 2004 for the second birth condition, the illegal migrant provisions, the descent registration requirement and Section 14A, 28 June 2005 for the widened overseas eligibility, 6 January 2015 for the cardholder merger, and 10 January 2020 for the provisions of 2019. The text of the principal Act as maintained on India Code carries the history in its footnotes, which record each substitution and insertion with the amending Act and the date from which it took effect, and a reader who checks a proposition about this trail against those footnotes will find the footnotes and the sections agree, because both are printing the same sequence of amendments. Most published error in this field would be prevented by the single habit of reading the footnote before repeating the date.

Working a status question along the trail

The practical value of the trail is that it converts most real status questions into an exercise in dating, and the order of inquiry matters more than any single provision. The first question is always the event relied on and its date: a birth in India, a birth outside India, a marriage, a period of residence, an application. The second question is which band of the relevant section covers that date, and the bands are printed in the sections themselves. For a birth in India, the three bands of Section 3(1) decide the matter, and the inquiry then moves to the parents’ status on that date and to no other date: the citizenship of a parent is tested at the time of the birth, not at the time of the application, the hearing or the argument, and a parent who became a citizen later does not retrospectively satisfy clause (b) or clause (c) for an earlier birth. For a birth outside India, the bands of Section 4(1) decide, with the registration provisos to be checked for the descent only cases and, for births from 3 December 2004, the consular registration and the declaration against a foreign passport to be verified as facts. For an application for registration or naturalisation, the inquiry is different in kind, because those routes are discretionary and are governed by the law in force when the application is decided, subject to the single hard rule the trail inserted: an applicant who is an illegal migrant as defined is barred at the threshold, whatever the application’s other merits.

Three errors recur in practice and the trail refutes each. The first is applying the birth rule of the wrong band because the amendment is known by the year of its short title rather than by its operative date: the 2003 amendment governs births from 3 December 2004, and a birth in November 2004 is a clause (b) birth, tested by the one citizen parent rule, even though the amending Act had been passed and published by then. The second error is the mirror image: treating an amendment’s passage as irrelevant until commencement for every purpose, when the descent provisions inserted in 2003 themselves use the commencement of the amendment as their trigger date, so that the operative date and not the assent date is uniformly the right reference point. The third error is assuming that documentary absence settles status. It does not. Status under Sections 3 and 4 attaches by operation of law on the facts of birth and parentage; documents are evidence of those facts, and a register entry, a passport or an identity card records a determination rather than creating the status it records. That distinction, elementary in the text, is the one most often lost in public argument, and it is the reason the register provisions of 2003, which are about recording and determining, should not be confused with the acquisition provisions, which are about the status itself. This article states the law of the trail for information; a person whose own status is in issue needs advice on the facts of the particular case, because the bands make the outcome turn on dates and parentage that only the record of that family can supply.

The same discipline serves the researcher. A proposition about the amendments should be testable in one of three places: the date banded text of Sections 3 and 4 for birth and descent, the substituted or inserted provision itself for everything else, and the footnotes to the India Code text for the instrument and date that made the change. A proposition that cannot be located in one of those places, however often it is repeated in commentary, should be treated as unsettled until it is located. The trail was built by six instruments across thirty four years, and its whole content is recoverable from the statute book without recourse to memory or to summary. That recoverability is the final respect in which the amendments reward the reader who approaches them as a table rather than as an argument.

What the trail establishes, and what it does not

The reader who has followed the sequence can date every change in the modern law of citizenship by birth and descent, and can state the direction of travel without assistance. Automatic entitlement narrowed in 1987 and again in 2004. A barred category was created in 2004, on definitions settled in the Act of that year. A statutory basis for a register of citizens and national identity cards has existed since 2004, operated nationally as a population register rather than as a citizen register. The diaspora category was created in 2004, universalised in 2005 and consolidated in 2015, in the same period and partly in the same instruments as the narrowing. Those propositions are the payload of the trail, and each of them is checkable against the text of the Act as it stands, because the Act carries the trail inside its own date banded clauses.

What the trail does not establish is equally worth stating. It does not establish that the narrowing amendments achieved the purpose their authors stated for them, since the size of the undocumented population was unknown when they were enacted and remains a matter of estimate rather than of count. It does not establish that the overseas widening and the domestic narrowing were coordinated as a single design in the mind of any one government, since the two movements have different institutional origins, the security review on one side and the diaspora committee on the other, and their meeting in the Act of 2003 may owe as much to the legislative calendar of a single session as to any unified intention. And it does not establish the merits of the amendment of 2019, which this article has deliberately confined to its mechanics. What the trail establishes is a pattern in the text, visible to any reader who lays the amendments side by side: over five amendments across three decades, the statute asked more and more of the person born inside India to parents without status, and offered more and more to the person born outside India to parents with an Indian past. The pattern is the finding. The arguments about whether it was justified belong, in the main, to the articles that own the individual provisions, and they are the stronger for being conducted on a trail that has first been dated correctly.

One further use of the trail deserves record, because it is the use to which the sequence is most often put in examination halls and in newsrooms alike. Almost every contested proposition in the citizenship debate reduces, on inspection, to a proposition about one of the five amendments: that India ended citizenship by birth alone at a particular date, that a register of citizens entered the law at a particular date, that a class of persons was barred from citizenship by a particular instrument, that the diaspora was given a status by a particular instrument. Each of those propositions has a right answer, and the right answer is in the table in this article and in the date bands of Sections 3 and 4. The trail’s discipline, date first and instrument second and evaluation last, will not settle the arguments. It will at least ensure that the arguments are about the law that exists.

Frequently Asked Questions

Q: How did the 1986 amendment narrow citizenship by birth?

Before the amendment, every person born in India on or after 26 January 1950 was a citizen by birth, whatever the parents’ nationality, subject to the exceptions for children of diplomats and enemy aliens. The Citizenship (Amendment) Act, 1986 (Act 51 of 1986) substituted Section 3 so that a person born in India on or after 1 July 1987 is a citizen by birth only if either parent was a citizen of India at the time of the birth. Births before that date kept the old rule. The soil alone stopped being sufficient, and parentage became a condition of the birthright for the first time.

Q: What did the 2003 citizenship amendment introduce?

The Citizenship (Amendment) Act, 2003, cited as Act 6 of 2004 and in force from 3 December 2004, introduced four things of the first importance. It defined the illegal migrant in Section 2(1)(b) and barred such a person from citizenship by registration and by naturalisation. It tightened citizenship by birth, requiring both parents to be citizens or one parent to be a citizen and the other not an illegal migrant. It inserted Section 14A, providing for compulsory registration of citizens, a National Register of Indian Citizens and national identity cards. And it created overseas citizenship, in Sections 7A to 7D, for persons of Indian origin in sixteen specified countries.

Q: When did India stop granting citizenship by birth alone?

India stopped granting citizenship by birth alone for births on or after 1 July 1987. Until that date, birth in India by itself conferred citizenship under the original Section 3 of the Citizenship Act, 1955, regardless of the parents’ nationality. The Citizenship (Amendment) Act, 1986 (Act 51 of 1986) added the requirement that either parent be a citizen of India at the time of the birth, and its substituted section used 1 July 1987 as the dividing date. The change is often misdated to 2004; that year brought a second and stricter condition, but the first condition, and the end of citizenship by soil alone, belong to 1987.

Q: What is the statutory basis for a register of citizens?

The statutory basis is Section 14A of the Citizenship Act, 1955, inserted by the Citizenship (Amendment) Act, 2003 with effect from 3 December 2004. Section 14A empowers the Central Government to compulsorily register every citizen of India and issue national identity cards, to maintain a National Register of Indian Citizens, and to establish a National Registration Authority, with the Registrar General, India acting as Registrar General of Citizen Registration. The Citizenship (Registration of Citizens and Issue of National Identity Cards) Rules, 2003 prescribe the procedure. The basis dates from 2003, not from 2019, although the machinery has been operated nationally as a population register rather than as a completed register of citizens.

Q: How did the overseas citizenship scheme evolve?

Overseas citizenship began with the Citizenship (Amendment) Act, 2003, which inserted Sections 7A to 7D and allowed persons of Indian origin who were citizens of sixteen specified countries listed in the Fourth Schedule to register as overseas citizens of India. The Citizenship (Amendment) Act, 2005 (Act 32 of 2005) omitted the Fourth Schedule and opened registration to persons of Indian origin in all countries except Pakistan and Bangladesh and any notified country, basing eligibility on citizenship at the commencement of the Constitution or eligibility for it. The Citizenship (Amendment) Act, 2015 (Act 1 of 2015) then merged the separate person of Indian origin card scheme into the overseas scheme, creating the single category of Overseas Citizen of India Cardholder.

Q: What was the difference between a person of Indian origin card and an OCI card?

The person of Indian origin card was an executive scheme, formulated in 1999 and revised in 2002, outside the Citizenship Act. It gave a foreign passport holder of Indian origin visa free travel to India for fifteen years and exemption from foreigners registration for stays up to one hundred and eighty days, and it conferred no statutory status. The OCI card rested on Sections 7A to 7D of the Citizenship Act, inserted in 2003. Registration was lifelong, carried parity with non resident Indians in economic, financial and educational fields, and exempted the holder from registration for any length of stay. The OCI was the fuller and the statutory document, which is why the two were merged into it in 2015.

Q: Which amendment made citizenship by descent gender neutral?

The Citizenship (Amendment) Act, 1992 (Act 39 of 1992) made citizenship by descent gender neutral. Before it, Section 4 transmitted citizenship to a person born outside India through the father alone: the person was a citizen by descent if the father was a citizen of India at the time of the birth. With effect from 10 December 1992, the substituted Section 4(1) provides that a person born outside India is a citizen by descent if either parent was a citizen of India at the time of the birth. The change was prospective, so a person born abroad before that date to an Indian mother and a foreign father did not gain citizenship from the amendment.

Q: What pattern runs across the citizenship amendments?

The pattern is a double movement carried in the same instruments. Inside India, automatic entitlement narrowed in stages: citizenship by birth gained a one parent condition in 1987 and a stricter two part test in 2004, a defined category of illegal migrant was created and barred from registration and naturalisation, and a statutory basis for registering citizens and issuing identity cards was inserted in 2003. Outside India, the movement ran the other way: overseas citizenship was created in 2003 for sixteen countries, widened to nearly all countries in 2005, and consolidated into a single cardholder category in 2015. The amendments restricted entry by presence and expanded connection by descent at the same time.

Q: What else did the 1986 amendment tighten besides citizenship by birth?

The Citizenship (Amendment) Act, 1986 (Act 51 of 1986) tightened the discretionary routes as well as the birth rule. Its Statement of Objects and Reasons proposed, and the Act carried, an increase in the residence qualification for citizenship by registration for a person of Indian origin, from six months of ordinary residence in India to five years. It also raised the residence period for naturalisation from five years to ten years for an applicant who was not a citizen of a Commonwealth country listed in the First Schedule. One change moved the other way: the registration category for spouses was widened by substituting the word person for the word women, so that a man married to a citizen of India became eligible on the same terms as a woman.

Q: Which amendment inserted Section 6A into the Citizenship Act, and why?

Section 6A was inserted by the Citizenship (Amendment) Act, 1985 (Act 65 of 1985), in force from 7 December 1985, to give statutory effect to the Assam Accord of 15 August 1985, which settled the agitation in Assam over migration from Bangladesh. The section contains special provisions for persons of Indian origin who came to Assam from the territory of Bangladesh. Those who came before 1 January 1966 and remained ordinarily resident in Assam were deemed citizens from 1 January 1966. Those who came between 1 January 1966 and 25 March 1971, and were detected as foreigners, were to be registered and became citizens for all purposes after ten years from detection, without the vote in the interval. Entry after 25 March 1971 received no such provision.

Q: When did the Citizenship (Amendment) Act, 2003 come into force, and under what Act number was it enacted?

The amendment was passed by Parliament in December 2003, received the assent of the President on 7 January 2004, and is therefore cited as Act 6 of 2004, although its short title is the Citizenship (Amendment) Act, 2003. It came into force on 3 December 2004. The three dates explain most citation errors in this field: 2003 is the year of the short title and of passage, 2004 is the year of the Act number, and 3 December 2004 is the operative date from which its rules govern births, applications and the other matters it changed, including the tightened birth test in Section 3(1)(c) and the commencement of Section 14A.

Q: Can an illegal migrant acquire Indian citizenship by registration or naturalisation?

No. The Citizenship (Amendment) Act, 2003 placed the bar at the threshold of both discretionary routes. Section 5 opens by empowering the Central Government to register as a citizen a person not being an illegal migrant who falls within the listed categories, so a person who is an illegal migrant within the definition in Section 2(1)(b) cannot be registered even if a category, such as marriage to a citizen or Indian origin with the required residence, is otherwise made out. Section 6 was amended in parallel, so a certificate of naturalisation cannot be granted to an illegal migrant either. The only exception to the bar is the class carved out by the proviso that the Citizenship (Amendment) Act, 2019 added to the definition itself.

Q: Which committee reviewed national security after the Kargil conflict?

The Kargil Review Committee, constituted after the Kargil conflict of 1999 under the chairmanship of K. Subrahmanyam, reviewed the events of the conflict and the national security system and reported at the end of 1999. The Government then constituted a Group of Ministers, chaired by the Home Minister, L. K. Advani, on 17 April 2000, to consider the Committee’s recommendations and formulate proposals for implementation. The Group’s report, Reforming the National Security System, was submitted on 26 February 2001. Its passage on illegal immigration recommended compulsory registration of citizens and non citizens, a national register of citizens and a multi purpose national identity card, and that recommendation became Section 14A of the Citizenship Act in the amendment of 2003.

Q: What did the Citizenship (Amendment) Act, 2005 change about overseas citizenship?

The Citizenship (Amendment) Act, 2005 (Act 32 of 2005), in force from 28 June 2005, widened overseas citizenship from a listed scheme to a general one. It substituted Section 7A and omitted the Fourth Schedule, which had confined registration to persons of Indian origin holding citizenship of sixteen specified countries. Under the substituted provision, a person who was a citizen of India on 26 January 1950 or at any time thereafter, or who was eligible to become a citizen on that date, or who belonged to a territory that became part of India after 15 August 1947, together with children, grandchildren, great grandchildren and certain spouses, could register, provided the person was a citizen of a country other than Pakistan or Bangladesh or a country notified for the purpose.

Q: What happened to the person of Indian origin card scheme in 2015?

The scheme was merged into the overseas citizenship scheme and abolished as a separate document. The Citizenship (Amendment) Ordinance was promulgated on 6 January 2015, and by notification of 9 January 2015 the person of Indian origin card scheme ceased and every existing cardholder was deemed to be an Overseas Citizen of India Cardholder from that date, without a fresh application. The Citizenship (Amendment) Act, 2015 (Act 1 of 2015), which received assent on 10 March 2015 and was deemed to have come into force on 6 January 2015, then recast Sections 7A to 7D and the definition in Section 2(1)(ee) in the language of the cardholder, leaving one diaspora category where two had coexisted since 2003.

Q: Do citizenship amendments apply to births that took place before the amendment?

No. Each amendment in the trail operates prospectively, through date bands written into the sections themselves. Citizenship by birth is governed by the clause of Section 3(1) that covers the date of birth: the unqualified rule for births before 1 July 1987, the one parent rule for births from that date to 2 December 2004, and the stricter test for births from 3 December 2004. Citizenship by descent is governed in the same way by Section 4(1), with 10 December 1992 as the dividing date for transmission through either parent. A later amendment therefore neither removes a status that had already attached at an earlier birth nor repairs an exclusion suffered under the law in force at that birth. The one great exception is Section 6A for Assam, which deemed a closed class to be citizens from 1966.

Q: How did the 2003 amendment change citizenship by descent for children born abroad?

The Citizenship (Amendment) Act, 2003 added a further proviso to Section 4 under which a person born outside India on or after its commencement, 3 December 2004, is a citizen by descent only if the birth is registered at an Indian consulate, in the prescribed form and manner, within one year of the birth or of the commencement of the amendment, whichever is later, or afterwards with the permission of the Central Government. It also provided that no such birth shall be registered unless the parents declare that the minor does not hold the passport of another country. The amendment did not change who could transmit citizenship, which the 1992 amendment had settled as either parent. It tightened how the transmission had to be perfected, making timely consular registration a condition of the status.

Q: How does a population register differ from a register of citizens?

A population register records usual residents: persons who have lived at a place for the preceding six months or intend to stay for the following six months, whatever their nationality and whatever the lawfulness of their presence. It makes no determination of status. A register of citizens records persons determined to be citizens of India, and reaching it from a population register requires the further stages prescribed by Rule 4 of the Citizenship (Registration of Citizens and Issue of National Identity Cards) Rules, 2003: verification, marking of entries whose citizenship is doubtful, notice, claims and objections, hearings and appeals. The National Population Register was prepared nationally alongside the Census of 2011. A national register of citizens, which would require those determination stages to be run for the whole country, has not been prepared, the completed citizen register exercise having been confined to Assam under its special provisions.

Q: Why is overseas citizenship not the same as dual citizenship?

Article 9 of the Constitution provides that a person who voluntarily acquires the citizenship of a foreign state ceases to be a citizen of India, and Sections 8 and 9 of the Citizenship Act carry that rule into the statute, so Indian law does not permit a person to hold Indian citizenship together with another citizenship. Overseas citizenship was designed to stop short of that bar. Despite its name, it is a registered status for a citizen of another country, conferring a lifelong visa, freedom from foreigners registration and parity with non resident Indians in economic, financial and educational fields. It confers no vote, no membership of a legislature, no constitutional office and no right under Article 16 to equality of opportunity in public employment. The High Level Committee on the Indian Diaspora had recommended dual citizenship in 2002; the scheme enacted in 2003 delivered the facilities without the citizenship.

Q: Which five amendments before 2019 make up the main citizenship amendment trail?

The five general amendments are those of 1986, 1992, 2003, 2005 and 2015. The Citizenship (Amendment) Act, 1986 (Act 51 of 1986) imposed the first condition on citizenship by birth. The Citizenship (Amendment) Act, 1992 (Act 39 of 1992) made citizenship by descent available through either parent. The Citizenship (Amendment) Act, 2003, cited as Act 6 of 2004, defined the illegal migrant, tightened the birth rule again, inserted Section 14A on registration and identity cards, and created overseas citizenship. The Citizenship (Amendment) Act, 2005 (Act 32 of 2005) widened overseas citizenship beyond the sixteen specified countries. The Citizenship (Amendment) Act, 2015 (Act 1 of 2015) merged the person of Indian origin card into the overseas scheme. A sixth instrument, the Act of 1985 that inserted Section 6A for Assam, stands before the sequence as its local prelude.