Population census: Need better, more honest data

The fundamental error in the religious census could be the fudge it manages between people who may be civilisationally grouped as Hindu and those who are Hindu by religious affiliation and practice. The only sound way to list anyone under a particular religious head is by self-affirmation and self-selection. It cannot be a default condition, or done by adding people to the Hindu count in case he or she does not claim to be Muslim, Christian, Buddhist, Sikh or Jain.

The census surveyor is (as this blogpost suggests) allowed only six options for indicating religion: the above five and Hinduism. Moreover, while a scheduled tribe (ST) citizen can indicate any religion of his or her choice, a scheduled caste (SC) Indian can only belong to Hinduism if he or she claims to be SC. This is because reservations based on caste status can only be given to SCs who are Hindu. This itself suggests that Christian SCs may be undercounted, making the Hindu total bloat to that extent.

Given the significant proselytisation and conversion activities of Muslim and Christian religious organisations, the chances are a large chunk of the SC population has probably formally been designated as Hindu to avoid being made ineligible for reservations.

As for Muslims, given that a significant chunk of them are illegal immigrants from Bangladesh in border states such as Assam and Bengal, what is the chance that all of them have declared their true religious affiliations? I have heard anecdotal stories of Bangladeshi maids in Mumbai who wear the sindoor and claim to be Hindu Bengalis in order to avoid detection as illegal immigrants.

So, even though the Muslim population proportion has risen in almost every single Indian state between 2001 and 2011 (barring Chhattisgarh, Manipur and Daman & Diu) it is more than likely that even this data may undercount Muslims.

Also, given that there is no category called “atheist”, and that most atheists are likely to be ex-Hindus (from hard-core Dravidian partisans to Communists and Naxals), there is more than a passing chance that the Hindu totals have been overcounted here too. Since anyone who does not claim to belong to any of the six major religions has to declare his religion or persuasion separately, which has to be noted down on the survey sheet separately, one can visualise the possibility where some may just get listed as Hindu.

Can any self-respecting country call this reliable data?

The question then is this: is this mere incompetence, or something deliberate?

My own suspicion is the latter, for India’s “secular” state would like all of us to believe that all Hindu fears of being gradually outnumbered are baseless. So declaring the Hindu population to be an overwhelmingly large number – whether 80.45 percent in 2001 or 78.35 percent in 2011 – may be considered vital for vote bank politics. This is enough to rubbish Hindu fears of going down in numbers due to lower birth rates and illegal immigration, even while frightening Muslims into voting en bloc for the “secular” parties.

This is why the Marxist-Left-secular groups consistently contradict themselves while arguing with Hindu groups which talk of being gradually outnumbered. While this fear may be exaggerated, it has been borne out in Assam, and is increasingly being realised in Kerala and West Bengal.

First, they will say that Hindus are a very huge proportion of the population, and then criticise Hindu groups for claiming tribal populations are being part of Hindu totals or for trying to convert them through "ghar wapsi" or other mechanisms. But the same rules don't apply to proselytisers of other religions.

Second, secular critics will deny that there is large-scale immigration in Assam and Bengal, and then, when the data points in the other direction, they will say this is the historical trend and its over. Nilim Datta, writing in Kafila, came to this conclusion two years ago: “It would be pertinent to point out that…(the) high population growth rate in Assam has declined since 1971 and has remained lower than that of India, categorically refuting assumptions of continuing illegal immigration from Bangladesh.”

How then does he explain the huge jump in the Muslim proportion of Assam’s population by a stupendous 3.3 percent – from 30.9 percent in 2001 to 34.2 percent in 2011? Pure natural growth may not be able to explain these numbers. The stork didn’t get these extra people. Or did a significant number declare themselves as Hindu in 2001 and have now changed their minds?

Third, the critics will say Muslim birth rates are higher because they have been discriminated against and are poor. But this can’t be entirely true, even granting that Muslims are among the poorer sections in India. Both Pakistan and Bangladesh are poorer than India, but still have lower birth rates. According to Business Standard, Muslim decadal population growth is 24 percent in India, 20 percent in Pakistan and just over 14 percent in Bangladesh. The Hindu decadal growth in India was 14.5 percent, according to The Indian Express.

Our own state-wise data shows that the poverty argument for showing high birth rates for Muslims is not the whole truth. In Kerala, where Muslims are not poor by any Indian standard, the Muslim proportion has actually risen substantially in 2011 from 24.7 percent to 26.6 percent.

So, when overall Muslim decadal population growth exceeds the Hindu decadal growth by almost 10 percent, one has to consider cultural and political factors as important variables in Indian Muslim demographics. This needs investigation and effective counter-strategies for promoting family planning.

The purpose of this article is not to raise unwarranted fears among Hindus or to advocate anything drastic, but to point out that we cannot have such shoddy data collection which explains nothing.

The only sensible way to collect religious data is by having a questionnaire that asks every adult Indian the following:

#1: Which religion do you belong to? Apart from the known big religions, which can be just ticked, there ought to be several others mentioned, including atheism (or no religion).

#2: All non-adults should be excluded from religious totals, and listed separately as under-age children belonging to parents whose religion is clear. If a child is born to parents who profess two different religions, we need a separate category for it, for the child could end up adopting any of the two, or neither. The non-adult totals should be mentioned separately, with religion merely being indicative and not for real.

#3: Scheduled castes and OBCs should not be excluded from reservation benefits just because they don’t declare themselves Hindu. In fact, a lot of OBCs converts who now get reservations may also be tempted to misdeclare their religions for fear of losing quota benefits. Making quotas contingent on which religion one belongs to is the surest way of getting doubtful data.

Our religious census data sucks. We need better and more honest data collection.

The writer is editor-in-chief, digital and publishing, Network18 Group