Najmuddin A. Farooqi.
I acknowledge the expertise and oratory skills of the top layer of BJP leaders. They come out clean from even the most complex debates through their articulate manoeuvring, twisting and churning. At the end of the day, their fans, followers and devotees blindly fall in love with their performance, while their adversaries are left shocked and numb. After the game is over, the adversary is left wondering how a sure-shot winning battle was lost and how the ground was so smartly taken away. I do not know whether the party or the Sangh prepares them through professional training camps. At least, I am indeed impressed by their swift and smart handling of issues and their art of proving and justifying what may appear to be the utmost wrong and false narratives, leaving the spectators and audience almost hypnotised by the situation.The speech delivered by the Union Home Minister as the Narendra Mohan Memorial Lecture at a Dainik Jagran Conclave on October 10, 2025, is a glaring example of this sort. I do not know whether the topic, “Illegal Immigration, Demographic Change and Democracy,” was deliberately planned for the occasion, as it does not appear to match the character of a literary programme.The rest I leave to the reader to decide: whether the figures and data presented by the Home Minister were factual and verifiable from the records, or exaggerated or false.Three important points specifically need to be noticed:First, according to him, the Hindu population of Pakistan and Bangladesh has undergone a substantial reduction in its proportion of the total population over the period from 1951 to 2014. He probably forgot that East and West Pakistan were divided in 1971 and that the majority of the Hindu population resided in the East. The data on record and what was stated by the Home Minister may be cross-checked.Second, 2.5 to 3 crore innocent/ persecuted Hindu refugees migrated to India and have been living here for three to four generations without basic rights and amenities. He then asked: “इन ढाई से तीन करोड़ लोगों का गुनाह क्या था? उनसे पूछकर तो विभाजन नहीं किया गया था।”In English, the substance is: “What was the fault of these 2.5–3 crores people? The Partition was not carried out after asking them.”Third, the percentage of the Muslim population increased exorbitantly from the 1951 to the 2011 Census and, as ever, the culprit is said to be the infiltrators (GHUSPAITHIYE). If the increased percentage is converted into numbers, it becomes 3.55 crore. If the number of persecuted Hindu refugees and Muslim Ghuspaithiyas are clubbed together, the total comes to between 6.05 and 6.55 crore.The ruling party frequently alleges that the previous governments, particularly the Congress regime, never cared about the sentiments and plight of Hindu community, failed to provide them with their due rights and on the contrary, looked the other way when it came to taking action against the Muslim “ghuspaithiyas” (illegal immigrants), allegedly as part of a policy of Muslim appeasement aimed at retaining their vote bank.According to this narrative, the BJP claims to be the only political party that has consistently championed Hindu interests. It was, therefore, eventually the BJP-led government that addressed this long-pending issue and succeeded in getting the Citizenship (Amendment) Act (CAA) enacted. But its rules were framed on March 11, 2024. However the reason of this inordinate delay of over four years in framing of the rules has not been made public so far.The figures cited later in this article present a striking picture: despite nearly six years having elapsed since the enactment of the CAA, only a few thousand persons appear to have been granted citizenship under its provisions so far. This inevitably raises an important question: at the present pace, how many more years will it take to reach the last eligible persecuted persons who remain in the queue and to ensure that the promise of the law is translated into justice in its full letter and spirit ?
2.5 to 3 Crore Persecuted Hindu Refugees
The statement was made by HM describing a large historical population of refugees, whom he said had suffered the consequences of Partition and remained without citizenship for generations. But an important statistical question remains: What is the underlying Government dataset or methodology by which the figure of 2.5–3 crore was calculated ?
The published speech does not provide a decade-by-decade statistical table from 1951 to 2014 showing the number of persecuted Hindu refugees who migrated, their countries of origin, their community, their legal status and their subsequent movement into citizenship. That missing statistical bridge is central to the discussion.
The demographic background: Pakistan in 1951.
The demographic record provides a very important background. The 1951 Census of Pakistan covered what were then West Pakistan and East Pakistan the territories corresponding broadly to present-day Pakistan and Bangladesh. The combined population was approximately 75 million. The Hindu population was approximately 9.6–9.8 million, constituting roughly 13% of the combined population. But the geographical distribution was radically different.
1951
| Area | Total population | Hindu population | Approx. Hindu share |
|---|---|---|---|
| West Pakistan | 33.7 million | about 0.53 million | 1.58% |
| East Pakistan | 42.2 million | about 9.3 million | about 22% |
| Pakistan as then constituted | about 75.9 million | about 9.8 million | about 13% |
The underlying 1951 Pakistan census material is preserved in the Census of India archive, including its tables on population by religion. This distinction between East and West Pakistan is extremely important. The Hindu population of West Pakistan was already very small after partition whereas East Pakistan retained a substantial Hindu population roughly one in every five inhabitants in 1951.
What happened to the Hindu population in East Pakistan / Bangladesh?
The demographic change in the eastern part of Pakistan was substantial. Historical census analysis shows that the Hindu share declined markedly over subsequent decades.
One important study of the census record notes that the number of Hindus in the territory that became Bangladesh fell sharply around Partition from approximately 11.8 million in 1941 to about 9.2 million in 1951 before subsequently recovering in absolute numbers, even while their percentage of the total population continued to decline.
This distinction is crucial: A decline in percentage does not necessarily mean an equivalent decline in absolute population. For example, a minority population can increase in absolute numbers while declining as a percentage because the total population is growing faster. That is precisely why both absolute population and percentage share need to be examined.
Bangladesh: 2022 Census
The latest Bangladesh census provides a useful contemporary comparison.
The 2022 Population and Housing Census recorded:
- total population: approximately 165.16 million enumerated, with an adjusted population of approximately 169.83 million;
- Hindu population: approximately 13.14 million;
- Hindu share: approximately 7.96%;
Thus, compared with approximately 22% Hindus in East Pakistan in 1951, the Hindu share in Bangladesh in 2022 was approximately 8%.
That represents a very substantial change in demographic share. But the demographic fact itself does not automatically establish where every person who disappeared from the Hindu share went. Possible components include:
- migration to India;
- migration to other countries;
- mortality;
- differences in fertility;
- changes in census enumeration;
- and other demographic factors.
The census data establish the change in population composition. They do not, by themselves, provide a complete migration ledger.
Pakistan: 2023 Census
The latest Pakistan census provides similarly important evidence. Pakistan’s 2023 Census recorded a total population of 240,458,089. The census tables record: Hindu : 5.22 million which roughly 2.17% of Pakistan’s total population.
Therefore, the figure sometimes quoted as approximately 5.2 million Hindus, or about 2.15%, is broadly close, but the precise percentage depends upon how the census categories are treated. The demographic table tells an important story but not the whole story. The available figures can therefore be presented approximately as follows:
| Period | Territory | Total population | Hindu population | Hindu share |
|---|---|---|---|---|
| 1951 | Pakistan, including East & West | ~75.9 million | ~9.8 million | ~13% |
| 1951 | West Pakistan | ~33.7 million | ~0.53 million | 1.58% |
| 1951 | East Pakistan | ~42.2 million | ~9.3 million | ~22% |
| 2022 | Bangladesh | ~169.8 million adjusted | ~13.14 million | ~7.96% |
| 2023 | Pakistan | ~240.46 million | ~5.22 million* | ~2.17%* |
These figures demonstrate a large long-term decline in Hindu share of population mainly in Bangladesh. They do not, however, by themselves establish that every person corresponding to that decline migrated to India. That distinction must be maintained.
Migration to India: a documented historical phenomenon
There is nevertheless substantial historical evidence that migration from East Pakistan to India continued for years after Partition.
For example, a 1952 letter from Jawaharlal Nehru to Chief Ministers recorded official estimates of movements between East Pakistan and India for 1950–1952. The figures showed very large flows in both directions and specifically recorded an excess movement of Hindus from East Pakistan to India during the period. The document also expressly noted that these figures did not include West Pakistan migration and did not include migration before 1950.
But a demographic decline cannot automatically be converted into an Indian refugee total
This is perhaps the most important statistical caution in the entire discussion.
Suppose the Hindu population share in a country declines substantially. That does not mean that the entire numerical difference represents people who migrated to India. A proper demographic accounting would have to distinguish:
Initial population: births − deaths; immigration − emigration=subsequent population.Consequently to convert the demographic decline in Pakistan and Bangladesh into a number of refugees who came specifically to India would require migration records. The census figures are therefore strong evidence of demographic change, but they are not by themselves a census of Indian-bound refugees.
Two to two-and-a-half crore Hindus settled abroad
The decline in population of Indian Hindus, since 1951, estimated two to two-and-a-half crore Indian Hindus have migrated from India and permanently settled abroad. As permanent emigrants, such persons would no longer part of India’s resident population recorded in subsequent Censuses. However, this figure should be treated as estimated.
Where does the CAA enter the picture ?
This demographic history provides the broader context, but the CAA has a much narrower legal universe. The Citizenship (Amendment) Act, 2019 introduced Section 6B into the Citizenship Act, creating a special route for specified persons belonging to the Hindu, Sikh, Buddhist, Jain, Parsi or Christian communities from Afghanistan, Bangladesh and Pakistan who entered India on or before 31 December 2014 and satisfy the statutory conditions. Interestingly implementing Citizenship (Amendment) Rules, 2024 were notified after more than four years on 11 March 2024. The first citizenship certificates were issued on 15 May 2024.
Thus, the CAA’s administrative operation began with certificates only after the 2024 Rules. The CAA therefore cannot be used as a statistical measure of the entire historical Hindu migration from Pakistan and Bangladesh. It addresses a specific statutory category.
Long-Term Visas came before the CAA
The framework was regulated by RBI under Manmohan Singh government in 2010. Official records show that LTVs were being issued to Pakistani and Bangladeshi nationals since then.
For example:
| Year | Pakistani nationals | Bangladeshi nationals |
|---|---|---|
| 2013 | 3,085 | 109 |
| 2014 | 2,779 | 126 |
| 2015, up to 29 July | 1,934 | 11 |
These figures come from Ministry of Home Affairs parliamentary records. (Nehru Archive)
From 1 January 2015 to 31 December 2019, the Government reported 26,786 LTVs in specified categories, including 25,782 persons belonging to minority communities from Bangladesh, Afghanistan and Pakistan.
Citizenship before CAA
Citizenship was also being granted under the Citizenship Act, 1955 before the CAA.
For example, an official parliamentary answer recorded 1,246 Pakistani nationals granted Indian citizenship between 2010 and 20 July 2014.
The figures were:
The Government also cautioned that yearly figures could include applications made in earlier years. This again demonstrates why simple addition of annual figures can produce misleading totals.
The first CAA certificates
The first 14 citizenship certificates under the 2024 Rules were issued on 15 May 2024 after more than four years of the caa act came into existence. Subsequently the number of certificates issued so far:
- 188 certificates in Gujarat in August 2024;
- 185 displaced persons receiving certificates in Gujarat in July 2025;
- 35 certificates in Odisha in December 2025;
- and other state-level distributions.
West Bengal reported to have received 2,30,605 applications August 14, 2026 and 23,956 certificates issued.
These figures demonstrate the difference between applications and approvals. They also demonstrate why the number of certificates cannot be treated as a measure of the entire historical refugee population.
Enormous numerical gap
We can now see the central issue more clearly.On one side stands the Home Minister’s statement referring to: 2.5–3 crore Hindus refugees living in India since decades without proper rights. On another side are administrative numbers involving:
- LTVs in the thousands and tens of thousands;
- pre-CAA citizenship grants;
- CAA applications;
- and CAA citizenship certificates numbering in the tens of thousands in publicly reported state-level data.
These are obviously not the same statistical universe. Nor should they simply be added together. But it cannot yet be answered by simply adding published figures. Then CAA has its own statutory eligibility conditions. Not every Hindu, Sikh, Buddhist, Jain, Parsi or Christian migrant from Pakistan or Bangladesh automatically qualifies. Therefore, a verified unique-person total requires a consolidated database. The CAA numbers citizenship numbers are not remotely equivalent to 2.5 to 3 crore historical figure.
Hence the conclusions cannot responsibly be drawn from these figures alone. That every Hindu who disappeared from the percentage share of Pakistan or Bangladesh migrated to India.
Missing Statistical Bridge
The central unanswered question can therefore be expressed quite precisely. If 2.5–3 crore is the Government’s historical estimate of Hindu refugees referred to by the Home Minister, the Government could make the matter much clearer by publishing a consolidated statistical statement covering:
| Required information | Period |
|---|---|
| Persons migrating from Pakistan to India | 1951–2014 |
| Persons migrating from East Pakistan/Bangladesh to India | 1951–2014 |
| Community/religion | Year-wise |
| LTV granted | Year-wise |
| Ordinary citizenship granted | Year-wise |
| CAA applications | 2024 onwards |
| CAA certificates issued | 2024 onwards |
| Persons who obtained citizenship through more than one administrative stage | Deduplicated |
| Persons remaining without citizenship | Current |
| Persons who migrated onward to third countries | Where documented |
| Persons who returned to country of origin | Where documented |
Only such a consolidated dataset would allow the 2.5–3 crore estimate to be reconciled with actual administrative records.
A broader demographic question
There is also a second, separate question.
The decline in the Hindu share of Pakistan and Bangladesh is a documented demographic phenomenon. But the reasons behind that decline are more complex than simply saying that every person represented by the decline migrated to India.
A complete demographic analysis would have to examine: migration + mortality + fertility + age structure + census methodology + possible changes in religious identification + migration to countries other than India.
The available census figures establish the outcome. They do not by themselves allocate every component of that change. This is why demographic data and refugee data should not be treated as interchangeable.
The significance of the 2.5–3 crore statement
The Home Minister’s statement nevertheless raises an important empirical question.
If the Government considers the historical refugee population to have reached 2.5–3 crore, then the public should be able to know:
How was this number calculated ? Was it based on:
- historical census differences?
- Government migration registers?
- refugee registration?
- estimates of population movement?
- records from border states?
- LTV databases?
- citizenship applications?
- demographic modelling?
- or a combination of these sources?
At present, the published speech does not provide that methodology. The statement itself establish the figure as an independently verified census total. It remains a major historical estimate stated by the Home Minister whose underlying statistical basis deserves disclosure and examination.
From political narrative to verifiable data
The CAA debate can therefore be separated into three distinct questions.
How many Hindus and other minorities left Pakistan and Bangladesh for India over decades since 1951 and how many settled elsewhere ?
What explains the change in the Hindu population share in Pakistan and Bangladesh between 1951 and the latest available census ?
Conclusion: numbers require reconciliation not assumption
The available evidence establishes a substantial and historically documented demographic transformation in Pakistan and Bangladesh, accompanied by significant migration to India over successive decades.
It also establishes that the Home Minister, in October 2025, referred to 2.5–3 crore people in the context of refugees who, according to his account, had suffered for generations after Partition.
At the same time, the available administrative record of LTVs, ordinary citizenship and CAA certificates is vastly smaller and belongs to different legal and statistical categories.
The demographic figures from Pakistan and Bangladesh cannot by themselves establish how many people migrated to India.
The assertion that approximately 2–2.5 crore Hindus migrated from India after 1951 and settled permanently abroad and consequently are no longer part of India’s subsequent Census population, also needs to be taken into consideration as one of the factors that may have contributed to changes in India’s demographic composition.
The most useful next step is to ask the government for the underlying data.
What is the Government’s complete statistical basis for the 2.5–3 crore figure ?
How many people from Pakistan and Bangladesh migrated to India in each decade from 1951 to 2014 ?
How many received LTVs ?
How many obtained citizenship under the ordinary law ?
How many have received citizenship under Section 6B of the CAA ?
How many applications received across India under CAA ? how many pending, how many rejected ? and why ?
How many unique individuals are represented after eliminating duplication ?
And, finally: What is the present legal status of the remaining population referred to in the 2.5–3 crore estimate ?
Only when these questions are answered through a transparent, consolidated and non-duplicative dataset can the historical magnitude of migration and the actual reach of the CAA be properly understood.
