Indian Economy: Who’s Right? - Arvind Panagariya

Q &A Arvind Panagariya

In a series of papers over the past year, you have argued these statistics are a complete myth, a gross exaggeration at the very least. Why do you think so?
When two sets of indicators lead to diametrically opposite conclusions, you either have a reasonable explanation for it or must reject one set of indicators. When we compare Indian children to those from sub-Saharan Africa (SSA) in terms of life expectancy, infant mortality rate (IMR), under-five mortality rate and maternal mortality rate (MMR), they look significantly healthier than the latter. But the picture turns on its head when we compare them in terms of incidence of stunting (low height for age) and underweight (low weight for age). The contrast is nothing short of dramatic.

Compare India with Chad, which has half of India’s per capita income. Using 2009 data, Chad has life expectancy at birth of 48 compared with India’s 66, IMR of 124 per 1,000 live births relative to India’s 50, MMR of 1,200 per 1 lakh live births in relation to India’s 230 and under-five mortality rate of 209 per 1,000 live births in contrast to India’s 66. Every one of these indicators places the health of Indian children miles ahead of those from Chad. Yet, child malnutrition indicators say that the proportion of children stunted and underweight is higher in India than in Chad!

Even more shocking is the comparison between Senegal and Kerala. With life expectancy of 74 years, IMR of 12 and MMR of 95, Kerala is the crown jewel of India when it comes to health. In comparison, Senegal exhibits a life expectancy of 62 years, IMR of 51 and MMR of 410. Yet, we are told that Kerala has a higher proportion of stunted and under-weight children than Senegal. It cannot get more absurd than this.

Where are we going wrong?
We have been applying a uniform World Health Organisation (WHO)-specified height to decide whether or not a child of a given age and gender is stunted. And similarly, a uniform WHO-specified weight to decide whether or not the child is underweight, regardless of the child’s race, socio-cultural background, geographical location or time or vegetarian versus meat diet. Any failure to meet the WHO-specified standard is attributed to malnutrition and the child classified as malnourished.

But what if Indian children are on average genetically shorter and lighter than the population from which the WHO standards are derived? Then, even perfectly healthy Indian children would be classified as malnourished just because they fail to meet the height and weight standards derived from the WHO population that is taller and heavier on an average.

So well-nourished populations may not be similar in height and weight?
My reading of the evidence is — not by a long shot. Japanese men and women are about 12 cm shorter than their Dutch counterparts.

The differences are not limited to adults. A 2006 study of infants born to Indian mothers in the US during 1995 to 2000 finds higher incidence of low birth weight and small-for-gestational age, and yet lower infant mortality rates for most part than the children of white mothers.

A study of Moroccan children in the Netherlands show that the height gap between the latter and the Dutch children can be observed as early as two years of age. The gap eventually rises to as much as 9 cm.

When did you first begin to doubt the Indian statistics? Why?
I’m not an expert on health, let alone child nutrition, by any stretch of imagination. But soon after my 2008 bookIndia: The Emerging Giant, in which I reported vital health statistics with approval, I began to notice the exceptionally poor child nutrition statistics and felt they could not be reconciled with the former. But I seriously focussed on the issue only when I took upon myself to write the chapter on health in a jointly authored book on the performance of Indian states. That is when I noticed that Kerala showed worse child nutrition statistics than many SSA countries and that, in turn, led me to dig deeper into the methodology leading to these absurd comparisons.

Why has there been no effort on the part of the Indian government to establish a set of height/ weight-age standards specific to India?
I think the United Nations, WHO et al have bulldozed us into believing that a single standard is scientifically right. Remember that all this ultimately ties into the Millennium Development Goals (MDG), which has behind it a huge lobby of very powerful international organisations, academics, NGOs and journalists. Without this common standard, it would be very difficult to assess progress on the MDGs. This is perhaps one reason that everyone has played along without asking tough questions. The belief is so deep-seated, especially in Delhi, that they refuse to see the obvious.

Jean Dreze and Angus Deaton have also pointed out flaws in malnutrition statistics. In an article written in 2009, they conclude “there are unresolved puzzles about anthropometric indicators in India, such as the high prevalence of stunting among privileged children”. But unlike you, they seem to believe that with adequate nutrition and time, Indian children will be able to catch up with ‘western’ populations. What do make of their arguments?
To set the record straight, let me first note that writing with Prof Amartya Sen in another article for a newsmagazine in November 2011, Jean Dreze makes no reference whatsoever to these “unresolved puzzles”. Instead, the authors go on to report child malnutrition numbers the way everyone else does, without any qualifications.

But coming to your main question, we will certainly narrow the gap as the diet of Indian children improves (as it has indeed been doing over the past several decades). But narrowing is not the same as eliminating the gap. If the Japanese newborns and adults have not been able to eliminate the gap, and the same also holds true of the newborn of Indian mothers in the US, why should we believe that Indian children would eliminate the gap?

One last point while we are on this subject. The catch-up is supposed to take many generations (we don’t know how many, of course). This means that some children who are classified as stunted and underweight using the WHO standards can simply not exit their height and weight category, no matter how good their diet is, because they were born with low height and weight and to mothers who were malnourished themselves. If these children are being given a proper diet, what sense does it make to label them malnourished? Won’t we be tempted to beef up their diets even further, and risk making them obese?

In a recent paper, you wrote, “A myth similar to the one considered here has also plagued policy discussion on adult hunger in India. There are widespread claims that more than one-fifth of the population, or approximately 240 million Indians, suffer from chronic hunger. This too is a much-exaggerated claim, as discussed in my forthcoming book.” Tell us a little more about this.
The claims of widespread adult hunger are principally based on the decline in calorie consumption observed in the surveys conducted by the National Sample Survey Organisation (NSSO). But there are good reasons to be sceptical that this decline reflects increased hunger. The decline has occurred across all consumer classes, including the richest ones. This points to factors other than access that have led to reduced calorie consumption across all category of consumers, rich and poor.

One obvious such factor is reduced need for calorie consumption due to improved absorption of calories consumed as well as reduced physical work. The former (improved absorption) has occurred due to improved epidemiological environment and access to healthcare.

Finally, measured calorie consumption is probably understating actual calorie consumption because it does not properly record midday meals, which too have progressively expanded.

These observations are consistent with the answers people give in the NSSO surveys when asked whether they had enough to eat on all days of the year. In the 2004-05 survey, 97.4 percent of the respondents in rural and 99.4 percent in urban areas replied to this question in the affirmative.

Do you think that standard WHO indicators of adult nutrition (like the Body Mass Index or BMI) that we are currently using might be similarly flawed?
This was indeed the conclusion of a 2008 study by Maarten Nube. Nube studied South Asian populations living in South Africa, Fiji and the US, and compared them with populations of other ethnicities living in the same countries, concluding that “there exists among adults of South Asian descent an ethnically determined predisposition for low adult BMI”.

How significantly do you think child malnutrition and adult hunger statistics have been overstated?
Answering this question requires identification of alternative norms to identify stunting and low weight among children. So, strictly speaking, I cannot yet answer this question. But let me say this: While releasing the much-publicised Hunger and Malnutrition (HUNGaMA) Report in January 2011, the prime minister said, “The problem of malnutrition is a matter of national shame.” Once we do our malnutrition numbers correctly, we will find that India has no more to be ashamed of its malnutrition level or the progress made in combating it than of other vital statistics such a life expectancy, IMR and MMR.