WHO Estimates of India’s Covid Deaths Are Highly Suspect

On May 5th, the World Health Organisation (WHO) issued a new report estimating global excess deaths at 14.9m for two years of the pandemic 2020-21 as the true COVID-19 mortality toll, nearly triple the official toll of 5.44m. “Excess mortality” is the difference between the number of deaths that would be expected in any time period based on data from earlier years and the number of deaths that have occurred. For countries with robust data surveillance, reporting and recording systems, this poses no real difficulty. Unfortunately, these conditions are not met in many countries. Therefore their excess mortality can only be estimated and the accuracy is a function of the reliability of the methodology and modelling used in the exercise. Given the overwhelming evidence about the flaws and deficiencies of Covid-related modelling over the last two years, and the damage caused by governments trusting modelling projections over real-world data, this should immediately throw up a forest of red flags about the WHO report.

A second reason to be sceptical is the less than stellar role of the WHO in its well-known Covid-related deference to China, the abandonment of its own summary of the state of the art science on managing pandemics from October 2019, its willingness to manipulate definitions of ‘herd immunity’ in relation to vaccines and natural immunity in order to fit with the experimental pharmaceutical and non-pharmaceutical interventions (NPIs) that came to dominate Covid policy around the world, and its self-interest in expanding its budget, authority and role in steering global health policies and management by means of a new international treaty.

A third ground for scepticism is they ascribe the total death count to the direct effects of Covid “due to the disease” and indirect effects “due to the pandemic’s impact on health systems and society”. The first part is questionable because it fails to distinguish between deaths with and from Covid. The second is disingenuous because the indirect toll of the NPIs (lockdowns, masks, induced fear, lost schooling, lost jobs, cancelled screenings and operations, aborted immunisation programs, disruptions to global food production and distribution, etc.) and vaccine-related adverse events will prove to be significantly higher than the indirect effects of the disease per se. Any study that fails to disaggregate deaths caused by the disease and by policy interventions to mitigate it lacks credibility.

Like many others including Will Jones on this site, I was especially struck by the new figures for India. The report pushes India up to the very top of the Covid mortality toll with 4.74m deaths, nearly 10 times more than the count of 481,486 (as of December 31st 2021), almost one-third of the world total. Sorry, but that is simply not credible.

India’s geographic diversity, population size and economic conditions make data collection especially challenging. In public lectures in Australia and Canada, to drive home the point about the scale, I usually comment that the entire Australian population is a rounding error in 1.3bn-strong India. It suffers from persistent and widespread mass poverty – India is a country of a few mega-billionaires amidst the world’s biggest pool of poor, illiterate and sick people bar none. It might be nuclear-armed, but state capacity when it comes to administration and public and social services is easily the worst of all major economies. The public sector scores high on petty corruption but low on efficiency. The public health service is risible and high quality healthcare is neither accessible nor affordable for ordinary Indians. The best doctors work in the public sector, in medium to large clinics and hospitals in metropolitan centres and as individual practitioners in most towns and villages. Consequently, health statistics are not all that reliable. But this is a general pathology, not one unique to COVID-19.

From everything I know about India, the WHO estimate does not align with overall death data, historical trends and Covid death compensation claims on the Indian Government from states. Indian experts believe that official statistics capture over 90% of all deaths. But this also means that about 10% of deaths would have been missed in previous years, yet the WHO’s ‘excess deaths’ count uses the official numbers as the baseline against which to estimate the impact of Covid. In a related vein, why would under-reporting be limited to Covid-related deaths and not, say, to suicides with its heavy social stigma and traffic accidents where the operators of overloaded buses and vans would try to drastically reduce actual numbers in order to hide the illegal loads (Figure 2)? The WHO estimates are flawed also in relying on 2019 deaths instead of using a five year average 2015-19 to wrinkle out anomalies in any given year.

Estimates of India’s total annual death rate range from 738 per 100,000 people by the World Bank to 1,030 per 100,000 people by World Life Expectancy. The total annual death toll therefore would be somewhere in the 10-13 million range: a very wide range. The WHO estimate of the death rate for 2021 is within the higher range from World Life Expectancy. Simply put, the WHO estimate of all-cause deaths is within any realistic estimate of the margin of error in India’s unique circumstances of scale and state capacity.

The caveats to official data notwithstanding, the WHO estimate would mean almost one-quarter additional deaths than normal. In fact it’s worse. Looking at the detailed tables, the 4.74m excess deaths is calculated from a combined excess death rate for 2020–21 of 171 per 100,00 people. This is disaggregated into 60 and 280 per 100,000 people for 2020 and 2021, respectively. That would imply a 38% jump in all cause deaths in 2021. Despite all the horror scenes we saw on TV of corpses lying in the streets and washed ashore on riverbanks, that’s just not possible. Perhaps the clue to the error lies in the title of the actual document: “Global excess deaths associated with COVID-19 (modelled estimates)” (emphasis added).

Some Daily Sceptic readers had fun with this aspect of the WHO announcement. My favourite exchange was this:

India’s own estimates of excess deaths for 2020 compared to 2019 is 480,000, of which Covid-related deaths were just under 150,000. So over 300,000 excess deaths were due to non-Covid causes, which in itself is far more believable because of the impact of the lockdown measures on exacerbating most of the conditions underlying India’s leading causes of deaths. By contrast, in 2021 the Covid-related death toll was much higher at 332,492.

Much as I have been critical in the past of official dismissals of international reports on India including weakening democratic practices, in this instance the Government is right to reject the WHO methodology of mathematical modelling based on data on 17 Indian states collected from websites and media reports: “This reflects a statistically unsound and scientifically questionable methodology of data collection for making excess mortality projections in the case of India.” As well as defective data collection methodology, the report is marred also by three critically flawed assumptions: that uncounted excess deaths occurred only in 2020-21 and not before; they occurred only for COVID-19 and not other diseases; and Covid-related deaths were due solely to the disease and not caused by policy interventions to control and eradicate it.

Ramesh Thakur is Emeritus Professor at the Australian National University’s Crawford School of Public Policy and a former UN Assistant Secretary-General.