Numbered in the millions

It has been reported for some time that the Union government has sought to block the World Health Organization (WHO) from publishing a report about excess mortality during the Covid-19 pandemic. It is now clear what the authorities in India were worried about. According to news reports, India reported by far the highest excess deaths during the pandemic — almost 4.7 million such deaths. Indonesia and the Russian Federation both had just over a million, and the United States just under a million.

In terms of population size, this is not completely out of line. India would rank in the low 30s among the countries for which calculation has been done. This would be a relatively unsurprising result. India’s relatively fragile healthcare system would tend to increase the number of deaths per capita; its younger population — since Covid-19 is particularly virulent in those above 65 — would tend to reduce its deaths per capita. Finally, India’s unfortunate position as the first location where the virulent Delta variant appeared during a period of relatively lax restrictions is likely to push deaths per capita up, again. Given these multiple pressures, a rank of 33rd in the world in terms of excess mortality per capita would not be a shock. It certainly does not call for India to be the only nation objecting to a report from a multilateral agency.

What, precisely, is “excess mortality”, and why is it being used instead of registered Covid-19 deaths — which, in India, are 10 times lower? Many, if not most, countries have incomplete or inefficient death registration systems that tend to miss out on recording the cause of death particularly when they are overburdened. However, given extensive historical data about mortality from all causes, well-understood models exist to predict the average “expected” deaths in any country in a particular year. Researchers at the WHO extrapolate actual deaths in the pandemic years from the data available, and compare that to the model’s predicted deaths in a non-pandemic year. The excess mortality, if any, is probably a result of year-specific circumstances — in this case, the pandemic (or associated lockdowns).

There is nothing particularly problematic about this exercise. The Union health ministry has nevertheless objected to the WHO’s use of models in the first place, saying that “authentic” data is available through its “robust” death registration system. The WHO’s researchers could have questioned the quality of this data — but, indeed, they merely needed to point out that the Indian government, although it possesses monthly mortality data, reportedly did not make these numbers available to the WHO’s research team.

It is perfectly fair to dispute the exact methodology used by the WHO’s research team. Yet the fact remains that there have been multiple different models applied by various different teams of researchers to the problem of India’s pandemic death toll. The WHO’s estimate is not the lowest of these estimates, but neither is it the highest. (That came from two University of Chicago economists estimating excess mortality using the Consumer Pyramids Household Survey, which pegged the number of extra deaths at 6.3 million, with 4.3 million of those coming during the Delta wave. The economists also found, intuitively, that the number of excess deaths peaked at the same time as the two waves, and that their estimate of excess deaths in a district correlated strongly with the registered Covid-19 death in that district, but were much higher. Counter-intuitively, they found that excess deaths were higher in families with higher household income.)

What matters surely is that all the results that have emerged from these different teams and methodologies are an order of magnitude greater than the officially registered deaths. This is not surprising in the least given the huge gaps in India’s death registration system that existed and were documented well before the pandemic.

Why is the government closing its eyes to objective evaluations of the toll that the pandemic took on Indians? It is now more than two years after the national lockdown began to be lifted, a year since the devastating Delta variant-driven second wave, and months after the Uttar Pradesh election results. It should now be absolutely clear that neither the Union nor the state governments are likely to pay a political price for any mishandling of the pandemic. So there is no real political capital at stake here. The government can easily engage with this data in a less confrontational manner. It may reveal structural flaws in the Indian state which should be addressed, not papered over. The government is under no obligation to accept any of these numbers — but an attitude of denial is disrespectful to the dead, who almost certainly number in the millions.