Death and journalism: India’s Covid deaths are undercounted, but twice the Partition toll?

Last 3 months have been gravest period of national tragedy, and a lot of Indians have died, mostly avoidably. But an undercount of total deaths at a mass scale is tougher.

Ambulances queue up outside outside a crematorium as chimneys bellow smoke | Photo: Praveen Jain | ThePrint

Tossing around this week’s National Interest in my head, I Googled ‘Death and Journalism’. Funnily, 19 of the first 20 entries were about the death “of” journalism. These included a couple, I must mention in supreme self-interest, that argued that it was wrong to say that journalism was dead. Thank you.

My search, however, was on how journalism covers, or relates to, death. Or more specifically, mass death. I found only this article on the Poynter website, which is a useful set of SOPs for a journalist reporting death. But not quite the opening gambit I was searching for this rather messy argument I was getting into this week. Inspired, of course, by that much-celebrated New York Times report estimating the real number of Covid deaths in India yet, especially in this catastrophic second wave.

It’s a marvellous piece of data journalism. Graphics are great, explanations pithy and short, sourcing, attribution, credits, brilliantly done. And this isn’t just about this one story. There’s been much discussion in the global media, the think-tank circuit and among social media influencers on what India’s real toll is.

That there is an undercounting of Covid infections and death numbers in India is certainly true. There’s been undercounting everywhere, but given our very weak municipal and public health systems, India’s data is fairly expected to be much poorer than, say, the US, which it now seems might have undercounted by about one-third.

The highest estimates of US deaths so far are nine lakh (against the official figure of six lakh), which Dr Anthony Fauci disputed the other day. He also raised questions about the reliability of the science of modelling. The WHO has also come with an estimation of global deaths closer to a little less than twice the officially stated numbers by all countries put together. But never mind. In America and across the world, we are still floating between 50 per cent and a 100 per cent undercounting.

The NYT report uses three scenarios to estimate India’s undercount as of 24 May. Against the government figure of 26.9 million cases and 307,000 deaths, it makes a conservative estimation, which is closer to WHO’s global calculation. It takes the positive cases to 404.2 million and doubles the deaths to six lakh. The denominator is 15 infections missed for each one recorded, and a fatality rate of 0.15 per cent. Deaths then double to twice the official figure.

The next scenario is “more likely”. It estimates 539 million infections, calculated at 20 cases missed for one tested positive. But, while the denominator goes from 15 to 20 for missed cases, for infection fatality rate (IFR), it doubles from 0.15 to 0.30 per cent. Deaths then become 1.6 million, or more than five times what India is officially stating.

And, finally, we come to the headline numbers. The “Worse” scenario. Mind you, there is no calculation yet for the worst. Or, you or I might not even have been alive to write or read this. This presumes 26 cases missed for each tested, and a fatality rate of 0.6 per cent. The total death number then comes to 4.2 million. Exactly 14 times what India has reported. Now, between the first scenario and the third, cases presumed missed have gone up from 15 to 26 times. But the fatality rate denominator has gone up four times, from 0.15 to 0.60. Why, it isn’t clear.

There are some mathematical/statistical reasons. There is also statistical extrapolation from the earlier Indian sero-surveys, the last of which covered the period up to 8 January. At least that one official data point from India is considered trustworthy. A straightforward doubling and quadrupling, in any case, helps the ease of doing calculation business.

We have no basis to question any of the figures yet. We are only asking for the basis of these figures. If that comes from the cricket team-strength of epidemiologists and mathematical modellers listed under the story, it would be instructive to check out what each one’s models predicted in the first wave in 2020. Or possibly, they entered the wrong year in their supercomputers.

The Government of India countered these calculations through NITI Aayog member and top neonatologist V.K. Paul. If anything, his estimate of the number of infections missed is close to the NYT’s “worse” case scenario. Read this wonderfully detailed story on his numbers claims by ThePrint’s Health Editor Abantika Ghosh. You might also wish to check out her book on the first wave in India, Billions Under Lockdown (Bloomsbury, 2021).

If you apply the Bharat Sarkar denominator for missed infections, then the estimated numbers (630 million) come tantalisingly close to the NYT’s “worse” case, 700 million. The difference is the basis on which deaths are calculated. Dr Paul says infection fatality rate is 0.05 per cent. The 0.6 that NYT uses, and that makes an immediate, unquestioned headline, is exactly 12 times higher. That is the difference between 3 and 42 lakh (4.2 million) deaths.

Paul calculates backwards: This is my death figure (3 lakh, and rising at about 4,000 a day) and here are my total infection numbers, so divide one by the other. He doesn’t tell us from where he gets his numbers which, back of the envelope, would show 23 infections missed for each tested positive. Particularly as there has been no sero-survey since 8 January. So, it would be reasonable to argue that there is no hard science to these Government of India claims. But I can’t find any science in the other calculation that takes the IFR from x to 12x either.

Death is a serious thing. More so for those still alive. Because each death leaves family, friends, colleagues, unfinished tasks behind. It is traumatic. That’s the reason we journalists also need to be that much more sensitive talking about death, or counting the dead.

The last three months have been the worst period of national tragedy, incompetent governance and national embarrassment for India since Partition. Pictures, of bodies piled up waiting for last rites, shortage of firewood and burial spaces, people gasping to death on the streets without oxygen, hundreds of bodies floating in the rivers, thousands strewn on the banks, will be the nightmare permanently imprinted on the minds of this generation, just as the massacres and mass-rapes of Partition were on our parents’. The world press covered this in great, graphic and accurate detail, as did much of the Indian media. Unless, of course, you were watching the wrong channels at prime time.

You also saw all of it, in the writings, pictures and videos of ThePrint reporters and camera-persons who, if I may bask in the halo of their brilliant work, did more than any other news organisation to bring you the brutal truth from across the country. And continue to do so. It was the brave and brilliant Dainik Bhaskar which unleashed hundreds of reporters to follow and count the bodies along the Ganga. Do their reports suggest a massive undercount? Yes. Are the official figures a huge understatement? Yes. How much? Does it look like millions of deaths were missed out? These armies of sharp reporters on the ground may not have the fancy degrees that modellers do, but they can see and they can count what’s visible, and not what’s imagined.

We will know the numbers by next year for sure. There is definitely an undercount of Covid deaths. But an undercount of total deaths at a mass scale? I toss some data for 2020 here, thanks to our reporter Aneesha Bedi. Mumbai and Delhi have nearly 100 per cent registration of births and deaths. In Mumbai, total deaths in 2020 were 1,11,942 compared to 91,223 in 2019. The city reported 11,116 Covid deaths. You can then presume that the remaining additional deaths, around 10,000 were also from Covid, and missed.

Delhi is more intriguing. Its total deaths in 2020 under all four municipalities — 1,42,693 — are actually considerably less than 2019’s 1,49,998. This despite 10,557 Covid deaths reported. I am told it is because deaths from ‘usual’ factors in Delhi, such as road accidents, murders, drink driving etc. dropped to almost zero because of the lockdowns. There are also two more stories. One by ThePrint’s Tenzin Zompa and the other by Anand Mohan J. in The Indian Express, who meticulously compiled data from Delhi’s biggest crematoriums and graveyards. These show an undercount by almost 40 per cent, bad enough, but not 4,000 per cent.

The fact is that the total death figure last year couldn’t be more than twice the recorded number, if at all. In which case, check out the epidemiologists who had “modelled” 2.5 million (25 lakh) Indians dead by September 2020. Then reflect on the wisdom of painting the new ‘worse’ scenarios based on their projections for 2021.

That’s the reason we say, death is a serious business for those left alive. A lot of Indians have died, mostly avoidably, in the past two months. But, if just about 2 lakh had died last year, have 40 lakh died in these seven weeks? That is twice as many as the number of people massacred during Partition in almost two years.

I am conscious of the hazards of exposing my delicate neck and my backside in making this argument. Amitava Kumar, a famous author, asked me in some social media posts if I do not have the taste of ashes in my mouth over what happened this April, given that I had written saying no such thing was happening in April 2020. I sure have had the taste of ashes in my mouth looking at so many compatriots die, often denied the dignity of a decent funeral, oxygen, hospital bed, medications. I have the taste of ashes in my mouth over how incompetently my government has handled this second wave. I have the taste of ashes in my mouth over what’s gone wrong in April 2021. Not over what I got absolutely right in April 2020. You can re-read it here.