Surjit S Bhalla: What's going on ... Noncensical

CALLING THE BLUFF

"There's something happening here. What it is ain't exactly clear". So sang Buffalo Springfield about guns and war, circa 1966. They could have been talking about the Indian census, and the RSS, and Tarlochan Singh of the National Commission for Minorities, circa 2004.

How can it be that the census, whose life and job it is to produce accurate numbers, could have sensationally published the growth rate of the population of various religious groups in India without controlling for the fact that the previous census (1991) did not include Jammu and Kashmir?

It took non-specialist journalists to point out this simple error. For Muslims, this error is huge""without J&K, the Muslim population growth rate is 2.6 per cent per annum 1991""2001, compared to a growth rate of 1.8 per cent for Hindus; with J&K in 2001 (and not in 1991) the Muslim population growth rate is 3.1 per cent per annum.

The latter is good for sensationalism, but not of much use otherwise.

Though he should know better, the 3.1 per cent per annum growth rate was enough for Mr Tarlochan Singh to go into cataclysms of his own. Speaking with authority that comes naturally to our non-expert bureaucrats, he states that the census has raised "some startling facts about Muslims' growth"; and that "Muslims can't go on producing children and then not send them to schools even where these are within their easy reach".

The last time there was a bizarre hare-brained scheme, and concern, about population growth was when Sanjay Gandhi of the Congress instituted forced sterilisations amongst the poor.

Thirty years later, our new "experts" are making the same mistake. Fertility is a function of education (primarily of the woman) and income, i.e. the less poor you are, and more educated you are, the less children you have, and less the population growth rate of your community.

There is another factor in operation here. Population growth is also a function of decreases in infant mortality rates, and the decreases are possibly largest among the poor.

Why? Because the rich always had medical care, their babies were always born healthy (though now they like to kill 'em real young, especially and only if the unborn child is a girl).

So with development, population growth is biased upwards for the poor communities, in the very short run (less than 10 years). Soon the declining fertility effects take over, and population growth decline.

This is what Indira and Sanjay Gandhi did not realise in 1975, and Tarlochan Singh, and BJP wannabe Neanderthals are not recognising today.

Which is why this is what Ram Madhav, RSS spokesman, had to say: "If the (Muslim) leadership encourages you to produce more, it results in higher fertility rates."

VHP Working-President Chinubhai Patel chimed in: "The situation is alarming going by the 36 per cent growth in the number of Muslims in the country. The community is conspiring to convert Hindu rajya into a Muslim country" (The Indian Express, September 8).

This ideological and prejudiced nonsense aside, population growth should not be of concern for policy makers""or even ideologues.

What should be of concern is education, and health care, and poverty, especially of women. Let people make their choices, and with development, as sure as the sun shines in the desert, population growth will decline.

This has happened around the world, across religions, across space, and across time. The fertility rate (average number of children per woman) was 6.1, 4.1, and 2.9 in all-Muslim Bangladesh for the three years, 1980, 1990, and 2002.

The corresponding numbers for all-Muslim (fundamentalist?) Iran: 6.7, 4.7, and 2. The numbers for Hindu-dominated India: 5.0, 3.8, and 2.9. The decline in Pakistan is less, but nevertheless there: 7, 5.8, and 4.5.

The National Family Health Survey (NFHS) for India for 1998-99 shows that once the education of the mother, and standard of living are controlled for, the same red blood and the same fertility pattern are observed.

The distorted rear view mirror effect
Census data, both on literacy and population growth, reflect the "wisdom" of a rear-view mirror. Both these statistics suffer from a severe "overhang" problem.

Literacy, for example, is affected by the presence of old people who never had a chance to go to school in the thirties, forties, fifties, and sixties.

Surely, precious little is going to be learnt from looking at the literacy figures of the entire population in 2001.

Analogously, population growth statistics in the nineties are affected by the characteristics of women born in the fifties and sixties. Equally useless.

The difference time, development, and education make to fertility patterns is well brought out by a simple analysis of NFHS data. Young less educated Muslim women (less than 30 years and less than 7 years of schooling) had near identical fertility pattern to that of corresponding Hindu women; to be precise, about 0.13 child higher in rural areas, and no difference in urban areas. For young women, religion is immaterial to prediction about fertility.

It is among older women (> 30 years) that differences are large, sometimes as much as 0.8 children (rural and uneducated). It is the young women who will determine the future pace of population growth""and that is going down, and is the same, for all communities.

It is useful to examine the pattern of population growth revealed by an alternative to the census source""namely, the National Sample Survey (NSS) of India.

Data for the three years""1983, 1987-88, and 1999-00""were taken and the total population matched with that of the census. (Unfortunately, for much the same reason as the census, NSS data for 1993-1994 are also plagued by the unavailability of reliable data for J&K).

The census and NSS data show parallel trends, though the magnitude of decline in population growth of Muslims post-1987, -1.2 percentage points, is much larger for the NSS than for the census post- 1991, -0.3 percentage point.

The real story of the census may not be the politically induced "analysis" of population growth among the Muslims but the sharp decline in growth rate among the scheduled castes/tribes (SC/ST)""from a growth rate of 2.7 per cent 1981""91 to only 2 per cent per annum, 1991""01. Unless the SC/STs have sharply escalated their income levels, it is unclear as to what might have caused this oversized drop. NSS data do not reveal a similar tendency""here the decline is only 0.1 percentage point.

One explanation is that there is no more "caste deflation", i.e. households and individuals can benefit from reservations for OBCs, etc. so there is no need to classify oneself as an SC/ST.

Support for this view is also obtained from the NFHS""the percentage of SC/ST "currently pregnant" women in 1998-99 was around 6.3, compared to a Hindu average of 5.5 and a Muslim average of 6.9, i.e. such differences do not imply an equality of SC/ST and Hindu population growth rates.

We are all economic animals first, and religious identities second, and political entities third.

Perhaps this reality has struck our fundamentalists, of all shades of colour, khaki, or red, or shades of pink""which is why they so often resort to demagoguery, rather than facts.

That is the only logic they know. Have pity on them, for in this globalised, information age, the old tactics just don't work anymore. The lie is so easily seen, the bluff so easily called.

ssbhalla@oxusresearch.com