A typo here, a typo there, and soon it can begin to add up to a “story”, a point of view. All of us have made data errors in print — the editors allow a “correction” post-publication, and do so easily on the web pages. I have had occasion to correct my own data, and everyone should. We have been in the era of data-based policy for some time now, not just in India but around the world. There are several “fact-check” websites around the world. This article should be seen primarily in that light.
But there is more, much more, to the recent data release of the preliminary results of the National Family Health Survey (NFHS) for 2019-20. These preliminary results have been released for 2019-20 along with comparable data for the 2015-16 survey. Data are presented for 21 states (12 “big” and nine small), with a population coverage of approximately 700 million. Tabular data for 131 variables have been released for 2019-20, along with comparable data for 2015-16. The data show that there has been considerable improvement in many of these 131 indicators of social welfare; and that in very few (you can count such instances on the fingers of one hand) there has been stagnation rather than a decline.
But that is not the first impression one obtains from the writings of distinguished scholars like Jean Dreze (‘Give children weight’, IE, December 19) and Abhishek Anand, Vikas Dimble and Arvind Subramanian (‘New welfarism of India’s Right’, IE, December 22), as well as editorials like ‘Hunger Alarm’ (IE, December 15), and pronouncements on worsening teenage pregnancy.
Dreze starts his essay with the conclusion that “leaving aside two or three countries like Niger and Yemen, India has the highest proportion of underweight children in the world…”, and adds for effect that this proportion is much higher than in neighbouring countries like “Bangladesh (22 per cent) and Nepal (27 per cent)”. Subramanian and co-authors, write that “after showing a slow but steady decline, especially between 2005 and 2015, there has been a disappointing reversal thereafter”. The ‘Hunger Alarm’ editorial notes that “there have been several indicators of the slowing down of economic growth and employment distress, which are bound to have an effect on hunger and nutrition” (emphasis added). The Hindu’s (December 15) verdict on worsening in India was reflected in the title “Rise in teenage pregnancy, higher anaemia rate in women: NFHS-5 phase data”.
The collective interpretation one obtains from reading these and many other articles on the subject is that after years of progress on health, nutrition, and child welfare, things have really worsened in the last four years — 2015/16 to 2019/20. Some authors are explicit, others indulge in nudge-nudge, wink-wink — these four years were also the first four years of the changing political order in India — the beginning of Narendra Modi as Prime Minister.
I have often stated in my articles (starting from ‘India, the Incomparable’, The Economic Times, March 3, 1989) that Indians suffer from a special disease — we think we are unique. To help negate this falsehood, it is important to compare India with the rest of the developing world, about 180 countries and a population of 6.5 billion. Thus, data presented in the table tries to put interpretation, and data, into both an absolute (for example, what happened in India) and a worldwide perspective (what happened elsewhere in comparable economies). The World Bank provides these comparable data on an annual basis and the data are easily available to all who are interested. Trends in hunger, nutrition, inequality, and growth among others can be judiciously assessed in a comparative perspective.
Let us look at some of the typos, falsehoods, misinterpretation of the data. Dreze states that the Bangladesh underweight percentage in 2015 was 22 per cent; World Bank (WB) data states it to be 30.1 per cent — some 4.7 percentage point (ppt) lower than India, but not 12.8 ppt lower as implied by Dreze. WB figures for all developing countries with data on stunting (63 countries, population 2.5 billion): In 2000, the stunting average excluding India was 29.3 per cent; in 2010 23.8 per cent; and in 2015, 28.7 per cent. In other words, a lot of zig-zag, reversals, and no net improvement over 20 years. In contrast, the stunting average in India was 46 per cent in 2000, 10 ppt lower in 2015, and still lower in 2019.
Teenage pregnancies on the rise? Really? In 2005, teenage pregnancies in India were 19.9 per cent; in 2015, 8.7 per cent, and 1 ppt lower in 2019. Net improvement — about 12 ppt in 15 years. Contrast this with a constant 10 ppt higher level in the developing world over the same period. The decline in teenage pregnancies in India is a genuine success story. India had the 10th lowest level in 2015, with Nepal nine positions lower (16.7 per cent) and Bangladesh (30.8 per cent in 2014) with the ninth highest magnitude of teen pregnancies (women 15-19 years of age).
Wasting: Not one of the authors looked at the data on wasting (weight for height). A curious but revealing omission. The wasting percentage in India was 17.5 per cent in 2000, and 3.3 ppt higher in 2015, and did not deteriorate further in 2019. In contrast, in the same time-period, wastage in the developing world improved by 1.6 ppt.
None of the data presented above shows India’s performance to have worsened post-2015. In other words, the pessimistic conclusion of “many” (or some) of a deterioration is false. Further, according to a very large number of NFHS indicators, welfare improved “big-time” between 2015 and 2019. A summary version of these data is as follows (comparable data for other countries is not as yet available).
All data are for change (difference in percentages) as revealed by NFHS on an aggregate population-weighted basis, rather than on an elevator economics basis as presented in many data analyses on the subject, for example, for X, Y states showed an improvement and Z states showed a decline.
Child Mortality: Data on three indicators (neo-natal, infant and under-five) show an improvement between 3 and 4 ppt. Immunizations: This is one big story of improvement; for example, the hepatitis B vaccine percentage (for children 12-23 months) increased from 64.8 per cent in 2015 to 83.6 per cent in 2019. Breast-feeding and diet of children 6-23 months: Average gain of 4 ppt for four indicators.
Negative trend in 26 children indicators: Only two show a worsening. Children with fever taken to a health facility (71.3 per cent in 2019, as compared with 72.6 per cent in 2015); and children under five years who are severely wasted — increase to 8.6 per cent from 7.6 per cent in 2015.
Negative trend in seven young women indicators: Zero. Modern birth control methods, age at marriage, teenage pregnancies, and sexual violence all show improvement. The average improvement in these seven indicators — 3.9 percentage points.
Negative trend in 15 adult indicators: Zero. Average gain 4.1 percentage points. Some of the indicators show an increase of 9 percentage points and above: Households with electricity (9.2 ppt), improved sanitation facilities (17.3 ppt), clean fuel for cooking (18 ppt) and women having a bank account that they use (29.8 ppt increase to a level of 77.2 per cent in 2019).
Why such a large divergence between perception and reality? If there is improvement in close to a 100 indicators and stagnation (not worsening) in very, very, few (less than five) then why are there editorials, special shows on TV, and scholarly (but motivated) conjectures about the larger “reality” that maybe causing a non-existent worsening? The intelligentsia, both domestic and international, got the Modi election, and his re-election, horribly wrong. As I have argued earlier (The New Wealth of Nations and Citizen Raj) one consistent explanation for this mis-reading is that the old elite has not come to grips with the reality that there is a change in leadership on issues relating to politics, economics, culture, and society.
This article first appeared in the print edition on December 26, under the title “Ideology trumps analysis.” Bhalla is executive director IMF representing India, Sri Lanka, Bangladesh and Bhutan. Views expressed are those of the author and do not necessarily represent the views of the IMF, its Executive Board, or IMF management