Policy lessons from data gaps in Covid death toll and Maha Kumbh mumbers

There is statistical data that suggests the impact of the pandemic is still being felt, and that it had a far higher death toll than is acknowledged. In financial year (FY) 2024-25, the Indian Railways (IR) carried 7.15 billion passengers. (Suburban commuters account for around 55 per cent of all rail passengers.)

 

While this traffic was 5 per cent higher than the 6.9 billion carried in FY24, it was well below the 8 billion mark that IR had carried in every financial year from 2015 to 2020 with a peak of 8.44 billion in FY19. Traffic crashed to just 1.25 billion in FY21. It hasn’t fully recovered yet.

 

Recently released mortality data from the pandemic period also indicates that the official death toll from Covid-19 was massively undercounted. If you know the size of a population, and you have a sense of age cohorts, birth, and mortality trends, you can project how many people are likely to die in the next 12 months. If the number of actual deaths greatly exceeds that estimate, there must be some cause.

 

The total number of deaths in 2018 and 2019 was 14.5 million, according to civil registration data. But the number of deaths in the pandemic years of 2020 and 2021 was 18.3 million. India, therefore, recorded roughly 3.74 million excess deaths in 2020 and 2021 compared to the two pre-Covid years of 2018 and 2019.

 

It’s reasonable to assume that a very large proportion of the excess deaths were due to Covid. But the official Covid toll was 533,000 — roughly one-seventh of the actual excess mortality. Undercounting has consequences. If health policy and resource allocation are based on 533,000 deaths, fewer resources are allocated.

 

The newly released data implies that the health care system coped very badly. The likely under-allocation of resources due to undercounting suggests it may also cope poorly in the next major health crisis.

 

The late release of 2021 data (on May 7) also has consequences. Insurers don’t know how many people died, and from what causes. Actuarial calculations for life and health care may, therefore, be wrong by orders of magnitude.

 

Insurance float — money collected as premium but not yet paid out in claims — is foundational to investment. Insurers invest the float after estimating how many claims there may be, and provisioning accordingly.

 

Lower premiums, charged due to undercounting, combined with a surge in unexpected claims, could lead to financial stress. India remains under-insured, though the pandemic did trigger greater interest in insurance. The next time there’s a pandemic, there may also be a financial crisis if insurers come under stress.

 

Coming back to passenger data, better civil aviation infrastructure may have led to some people switching to air. Domestic air traffic has grown quickly as schemes such as Udaan have connected smaller cities and as the IR has hiked First Class and AC fares. In FY24, around 160 million Indians took domestic flights and around 166 million did so in FY25.

 

But the difference between traffic volumes on IR and in the air is vast. It’s clear that the vast majority of travellers either can’t afford to fly or don’t need to. All this traffic data leads to the diagnosis of a case of likely overcounting in a more recent instance. The latest iteration of the Maha Kumbh, celebrated every 12 years, took place between January 13 and February 26.

 

The authorities made elaborate arrangements for pilgrims. Roughly ₹6,400 crore was spent on the bandobast and the state of Uttar Pradesh reckons it generated ₹2 trillion in  economic activity. This should be a nice, little gross domestic product (GDP) bump and a great return on investment. It was estimated over 400 million pilgrims travelled to take a dip.

 

Did they really? If we assume that a large proportion of the pilgrims took trains or flights, the railway and civil aviation data should have seen a tsunami in those 45 days. The data suggests far fewer people actually went to the mela. There is no traffic bump.

 

In turn that implies revenue generation may be a lot less. We may get an inkling of that when GDP data and UP’s GDP data are released. There may also have been over-invoicing of costs by large magnitudes by invoicing for ghost pilgrims, and thus, wastage of public funds in arrangements. Overcounting of pilgrims has consequences too, even if those are less serious than the consequences of undercounting pandemic victims.