The brouhaha over data on pandemic mortality
PremiumIndia has decried the statistical model used by the WHO’s estimate of excess deaths from the pandemic. This nitpicking, however, doesn’t obscure the need of a realistic record for history
It’s a novel mix: a virus whose effects vary vastly, a pandemic that took a toll in various ways, and an age of ‘post-truth’ riven by alternate facts. History’s need to record how many lives the world lost was never going to be easy to fulfil, but the count is proving even harder than expected. An estimate of excess mortality made by the World Health Organization (WHO) was held back on objections raised by India over its statistical model. As reported by The New York Times, the WHO pegged extra deaths above the pre-covid norm at about 15 million globally by the end of 2021. This figure, which includes indirect losses from a healthcare crunch caused by a covid crowding-out, is around 9 million above the combined tally reported by governments. India, with an official covid death toll of roughly 520,000, was estimated to account for at least 4 million deaths. While this broadly conforms with independent studies by other agencies, it has evoked a sharp response from New Delhi, which has sought to poke holes in the methodology that was used for it. As the controversy draws global attention, the validity of our stance is likely to face scrutiny around the globe.
The data in contention is not a head count, but a calculation drawn from national records in conjunction with local surveys and other variables fed into formulae designed to also account for covid-related deaths that were missed. In the sphere of statistics, this is a standard way to plug information gaps. Based on field data taken from random samples, probabilities can be worked out that offer a fairly accurate big picture. Of course, these representations of reality make no claim to absolute truth, offering us confidence ratios instead. Lack of clarity on this aspect of the WHO survey is among the reasons India has decried it as flawed. While the government’s basic objection that this country is too large for a one-size-fits-all formula seems rather facile, its actual critique goes deeper. Data taken from 18 Indian states remains not just “unverified", in its view, it’s too geographically narrow a sample for all-India use. The Centre also objected to the use of 2019 WHO health estimates rather than Indian datasets and of mortality data by age and gender generalized from other countries. It has also sought to point out other alleged flaws, such as the WHO model’s binary measure of income instead of a gradation and a temperature variation input that could plausibly distort results. A big complaint is that advanced countries were assessed on a different yardstick altogether. On its part, the WHO has shown no inclination so far to retract the claim that its approach was robust, tested and therefore appropriate.
Charges of a bias held by global institutions against developing countries in general and India in particular have been flung around in recent years. Some of these allegations do hold weight and deserve probes, but for Indian dissent on covid-led mortality to make headway, we must openly accept that an undercount problem exists. For the Centre’s litany not to read like a nit-picker’s list slapped together in indignation at WHO findings, we must not only acknowledge that more people may have died than we’ve counted, but also produce our own estimate with tools kept open to the world for testing. Given the anecdotal evidence and poor quality of official mortality statistics, an insistence that state data reflects all we need to know about how we fared against covid will just not work. If truth is our aim, let’s show it.