Gujarat Covid-19 Mortality Rate: Why Rahul Gandhi Is Wrong

Looking at epidemic data in isolation is nearly the worst thing one can do. This is since most of the numbers reported, especially in the early stages, are patchy, scattered, and non-representative of the ground situation.

The only worse thing is to use the statistics of a grave health crisis to try and drum up political mileage. Today, Rahul Gandhi of the Congress has managed quite a feat, by making both mistakes in one go.

In a catty tweet, the political dynast compared Gujarat’s high mortality rates with those of a few other states to conclude that he had exposed a flawed ‘Gujarat model’. This was the usual dog-whistle style of politics, which capitalises on calamities to put down an opponent. Unfortunately, as usual, Gandhi’s proclamations are not borne out by facts.

Here’s why he is wrong:

Gujarat mortality rates are high because a large number of cases went unreported in the old, walled city area of Ahmedabad. These cases went unreported because many parts of containment zones there became no-go areas, thanks to large, violent mobs that agitated furiously, and indulged in stone pelting, to senselessly protest against the lockdown.

This prevented health officials from doing their duty, and the resultant lower number of reported cases has artificially skewed mortality statistics in the state.

The data does not lie:

As the mortality statistics curve above shows, these rates were already abnormally high from March through mid-April, on account of the massive Tablighi Jamaat cluster which developed in Ahmedabad’s old city area during that period.

The rates dipped only when testing went up in other parts of the state, before rising again to stabilise, as deaths (and testing) in the old city area eventually rose further. This is what caused the apparent, misleading, numerical distortion.

We see the same pattern in a second plot of cumulative deaths versus time, where again, the trends were initially quite steep, before the reasons listed above came into play.

A third point is brought out by the mortality cross plot above: if cases are being under-reported in Gujarat, leading to artificially higher mortality rates, then deaths are being under-reported in Kerala (resulting in artificially lower mortality rates).

The point is that even if deaths can never be hidden, there appears to be a high degree of subjectivity and variance between states, on how they report fatalities. The issue of Covid/non-Covid, and co-morbidities or not, remains to be standardised or reconciled from state to state across India. This is a factor which analysts and politicians need to take into account before blindly doing arithmetic and arriving at misleading values.

Indeed, even without factoring in an error margin to Gujarat data because of case under-reporting, we see from the above chart that, in reality, the state is not too distant (look at the blue, dashed line) from both Madhya Pradesh (MP) and West Bengal.

Now, if one were to add the unreported deaths from MP and Bengal, we see that both states too, would converge onto the same blue, dashed line upon which Gujarat presently sits.

This subjectivity, in the reporting of deaths, and its attendant impact on mortality rates, became most evident when both Maharashtra and Delhi ‘revised’ their figures on 16 June. Delhi added 437 deaths in a single day, taking its mortality rates up from 3.3 per cent to 4.1 per cent.

Far more alarmingly, Gandhi’s coalition government in Maharashtra added 1,409 deaths the same day. Whatever the wishy-washy, untenable, justifications his Aghadi might submit to a stunned public, the net result is that the mortality rates in Maharashtra surged overnight from 3.7 per cent to almost 5 per cent.

These unconscionable modes of tallying, more than anything else, highlight the high sensitivity of epidemic statistics to human caprice, and serve as a stern reminder against mindless number crunching.

With specific reference to Gujarat, then, the issue therefore is not of deaths being under-reported, or of Wuhan virus deaths being classified as otherwise, but very simply, of far fewer cases being identified and reported than actually existed in the hotspots.

In relative terms, more deaths than cases were registered for long enough to inflate mortality rates irreversibly. This is because now, if ever state mortality rates are to come down to 3-4 per cent, Gujarat would need to report an additional 20,000 death-free cases – and that is a near statistical impossibility (especially since the state looks set to flatten its case curve soon).

What is most ironic is that in addition to the above available public domain data, to disprove Gandhi, the rebuttal to his accusation actually lay in the very BBC report he quoted for his political broadside.

If only he had bothered to read it before deploying it as a political weapon. Quoting an economics professor, it states unambiguously that testing was low in the old, walled city area of Ahmedabad (where the bulk of the cases are).

This is the truth: a bulk of the cases in that part of Ahmedabad went unreported because officials were prevented from doing their duty by locals. Of course, the BBC being the BBC, this lack of testing is spun in their report as laxity on the government’s part, but numerous news reports tell a very different story: that roving mobs of violent stone-pelters repeatedly clashed with the police in these containment zones, preventing the collection of samples for testing in fuller measure.

So, Gandhi will not condemn the reason why adequate samples could not be collected in the old, walled city of Ahmedabad (the agitations and stone pelting, which is what led to the misleadingly high mortality rates in Gujarat), but he will use such resultant, flawed data to demonstrate the incompetence of his political opponents to his prized vote banks. Where is the logic or morality in that?

Perhaps it is time politicians of a certain hue stopped misusing epidemic data to attack their opponents, or politicking on the miseries of the people – if they don’t want to embarrass themselves, that is. If they do, then the world of electoral irrelevance is their oyster.

All data from Covid19india.org and MoHFW.