Three weeks after the health ministry claimed to have achieved in 2020, the Registrar General of India has indirectly confirmed the figure. This, then, is the first time that India claims to have registered every death – in a pandemic-hit year, no less.
In a bulletin on Wednesday, the registrar general estimated India’s 2020 mortality rate at six per 1,000 people, the same as the previous year. For a projected population of just over 135 crore, that means 81.2 lakh deaths from all causes. The estimate is based on the findings of a baseline survey called the Sample Registration System, or SRS.
Newslaundry that the health ministry had declared 99.99 percent death registration for 2020 without explaining how the number was arrived at, raising questions about its accuracy.
The ministry was disputing the World Health Organisation’s estimate of Covid deaths in India – 8.3 lakh in 2020 and 47.4 lakh over the course of the pandemic so far, nine times more than the official count of 5.25 lakh – claiming it was out of agreement with the “extremely robust” Civil Registration System, or CRS, which is used by states to record births and deaths. But the CRS report for 2020, in a departure from practice, was silent on the death registration rate as the SRS report had not been released yet.
Now that the SRS data is out, will it settle the dispute over India’s Covid deaths? Not quite.
Here’s what it all means.
SRS and CRS surveys
The decadal census is the gold standard for capturing demographic data, covering as it does every household in the country. In intervening years, annual surveys such as the SRS and the CRS help the government chart its course of action, not least when it comes to providing public services. The SRS survey covers around 83 lakh people across 9,000 sampling households, collecting data on death, birth, and child mortality rates. Though the survey is done annually, the report is generally published with a lag of two years.
PC Mohanan, chairperson of the Kerala Statistical Commission, calls the SRS an “efficient, scientific” system as death and birth records are validated by enumerators in person. The CRS, in contrast, relies on people visiting a registration unit – like panchayat or ward office – to get deaths and births registered. The system also collects data on births and deaths reported by hospitals. The data is collected by states and union territories and sent to the registrar general.
To assess the severity of the pandemic in India – marked by two years of lockdowns, deaths, migrations, and losses of livelihood – academics and journalists have relied on CRS and SRS reports. The latest CRS report, for 2020, counted 81.15 lakh deaths, 4.75 lakh more than the previous year.
There is, however, that states have Covid deaths – and the WHO’s recent estimate confirmed as much.
So, how many people in India have actually died of Covid?
We may never know. The CRS reports don’t ascribe cause of death, while the SRS reports are an estimation. Of all the registered deaths, only 20 percent are medically certified, according to another government survey. Moreover, in a year when the pandemic made it harder for people to visit registration units and delayed the SRS survey. K Narayana Unni, former deputy registrar general of India, has that many families may not have registered all deaths under the circumstances. Another limitation, he told Newslaundry, was related to the registration processes of the CRS and the SRS. “CRS takes into account the place of occurrence, SRS the place of residence,” he explained. “For example, if a man from Delhi dies in another state, CRS will register the event in that state while SRS will record it in Delhi. So CRS and SRS are not straightaway comparable.”
In addition, some officials involved with registration exercise were drafted for Covid duty which could have hampered data collection as well. All these factors may have caused under-reporting of estimated and registered deaths.
It’s a thumb rule that the sooner a death or birth is registered, the more efficient the system is. In India, however, around 7.6 lakh deaths which occurred in 2019 were registered in 2020. It remains to be seen how many of the deaths in 2020 appear in the registration data for 2021.
As Mohanan pointed out, all projections have errors and limitations. For example, there was a divergence of 2-3 crore between modelled population projections and the censuses in 2001 and 2011. “So, every crore addition or subtraction makes a huge difference in estimated deaths by SRS,” he added.
The recent National Family Health Survey shows that India registered only 71 percent deaths in 2019-21. Sayeed Unisa, who heads the Department of Biostatistics and Epidemiology at the International Institute for Population Sciences, Mumbai, had told Newslaundry that the realistic death registration rate could be 80-85 percent. This means India may have missed 14 lakh deaths from all causes in 2020.