Is the undocumented migrant counted?

Can India’s official statistics detect large-scale undocumented migration? Undocumented immigrants can evade the law, but can they evade the Census? If millions of them live in India, they must be somewhere in the country’s official statistics. Union Minister of Parliamentary Affairs Kiren Rijiju told Parliament: “As per available data, around 20 million illegal Bangladeshi immigrants remain in India.” They must then encounter enumerators during the census. So millions of people must be meeting enumerators during India’s decennial censuses. What can an undocumented immigrant do in the face of such an encounter?

They can respond in three ways: Avoiding the census or avoiding it altogether, contributing to reporting error; is counted but reports India rather than Bangladesh as birthplace, which introduces measurement error; or accurately report the place of birth.

Each of these options leaves a different trace in the data that can be checked. This provides a simple framework for asking whether India’s official data is consistent with the alleged scale of undocumented migration. But a framework based on place of birth has a built-in limit: The Census records where a person was born, not whether they entered legally. The same “born in Bangladesh” category includes citizens, visa holders and the undocumented, with nothing to distinguish them from each other. What the data can speak to, then, is scale: whether a population of the claimed size exists.

Getting wet

Ducking hides completely from the census taker. But hiding isn’t the only way to escape. This is what a demographer would call coverage error. A more subtle version, for example, is to deflect. For example, tell the enumerator, “I don’t actually live here. I will be counted at my usual residence elsewhere,” and then tell them not to be counted at the other location either. Such a response, given without any intention of being counted elsewhere, is legally no different from refusing to respond – both are offenses under the Census Act – but in practice it is likely to go unnoticed.

The result is an omission; This is what the Census was created to detect through the Post-Enumeration Survey (PES), an independent reenumeration of block samples individually matched to the Census. This has been done after every Census since 1951 to estimate coverage error (the people the Census missed) and measurement error (the reasons that led to them being missed).

The 2011 PES found a net neglect rate of 23 per 1,000 people nationally, or 2.3%; this figure is unchanged from 2001 (up from 1.8% in 1991). As shown in the table below, the Eastern region, which borders Bangladesh and where the alleged “infiltration” is concentrated, is best covered.

Of the missing number of about 27.7 million, the Central and Northern regions combined account for 19 million, or about 66%.

The most overlooked people nationally are babies and young, active adults in their twenties; this is a pattern that has continued since at least 1991. However, this pattern is concentrated in the Northern and Central states. In the Eastern and Northeastern States, coverage is not only better overall but also flatter across age groups; There is no similar increase among young adults, as seen in the graph below. The people the census misses are disproportionately babies, not adult immigrants.

As the chart shows, coverage in the Eastern and Northeastern States is not only better overall, but also flatter across age groups; There is no similar increase among young adults

Even assuming, for the sake of argument, that everyone missed by the Census anywhere in the Eastern and Northeastern regions were undocumented Bangladeshi immigrants, this would remain well below the 20 million stated by the government.

What can therefore be said with confidence is that the census-overlooked population is small, stable over decades, and smallest along the border. This leaves little room for unenumerated populations on the scale required by political assertion.

Incorrect reporting

The second option – being enumerated but recording one’s place of birth as India rather than abroad – is the least costly and can neither be verified nor refuted at the point of enumeration. But there’s one point to this: The person who misreports is still the person counted in the Census. This is what might be called measurement error. They remain in the total population, census by religion, age and gender tables, but place of birth records are incorrect.

For example, if large numbers of undocumented Muslim immigrants were doing this, the effect would not be a hidden population but an inflated population (a Muslim population with a population growth rate that natural increase and differential fertility can account for) and concentration in border areas.

This inflation was examined in an earlier article on India Forum (https://www.theindiaforum.in/politics/has-demography-border-districts-eastern-india-changed) and no such signature was found. If misreporting hides a large influx of immigrants, this would show up as border-driven Muslim growth beyond what fertility explains. Instead, the counts showed growth was moderate, consistent with fertility, and geographically dispersed.

Correct Answers

The third option is to report their place of birth accurately, in the belief that the census is anonymous and will not put them at risk. Immigrant stock is a commonly used metric to track immigration levels. A person who is enumerated in India but born in another country is considered a migrant under this measure.

The composition of the migrant stock in India has seen significant declines in those born in Bangladesh and Pakistan, while large increases have been seen in those born in Nepal, as shown in the chart below. However, despite the decline in the overall immigrant stock, around half of India’s immigrant stock consistently consists of Bangladeshi-born individuals.

Chart shows share of people whose birthplace was not India in the last three Censuses

In the last three censuses, roughly 94% of India’s Bangladeshi-born residents lived in five border states (West Bengal, Assam, Tripura, Meghalaya and Mizoram); This share has remained virtually unchanged, although the national Bangladeshi-born population has fallen by nearly a third, from approximately 4.0 million to 2.7 million.

The table shows data from districts in the Eastern States bordering Bangladesh.

In terms of both share and absolute numbers in these regions, immigrants, the vast majority of whom come from Bangladesh, are aging and possibly disappearing. Younger age groups show insufficient numbers. This pattern is consistent with a specific history: a large portion of this population arrived during Partition in 1947 or the Bangladesh Liberation War in 1971. This generation is now old and its numbers are decreasing as its members die.

Has there been a fluctuation?

Thus, using official data in all its forms, the evidence suggests that there was at least no increase in the number of undocumented immigrants from Bangladesh between 1991 and 2011. The claim here is not that there are no undocumented immigrants from Bangladesh in India, the question here is about the scale of migration.

The census cannot tell us that there are no undocumented immigrants from Bangladesh. From the three different possibilities discussed in this article, it appears that their number is not 20 million.

The writer is a lecturer at the Economics Group, Faculty of Humanities and Social Sciences, Azim Premji University. The content and opinions expressed are those of the author and do not constitute endorsement by the university or reflect the views of the university.