For whom the bell tolls? Covid 19 death and demography in India
Anirban Mukherjee and Nilanjana Roy
The Covid 19 outbreak, the deadliest since the Spanish flu pandemic of 1919, is creating unprecedented havoc all across the world. After almost a year into the pandemic, the number of Covid 19 infected people worldwide has reached a staggering 4 crore of which 11 lakh patients expired. As of 19th October, there are approximately 1 crore active patients all over the world. Given the impact it created, it is no wonder the Covid 19 now acquires the center stage in vaccine research. However, the pandemic has also stimulated the minds of social scientists who besides looking at the impact of Covid on our society, also examine how different socio-economic parameters facilitate the spread of the disease. The research on the contagion pattern shows, that despite our belief that death is a great leveller, death in Covid 19, as well as infection and hospitalization rates vary a lot across ethnicities, age, and income groups. Data published by the Centers for Disease Control and Prevention (CDC) show that in the United States compared to the White, non-Hispanic population, the Covid 19 infection, hospitalization, and death rates are much higher for minority groups such as Hispanics and African-Americans. The case of the United States is not an isolated incident; it reflects the importance of demography which is corroborated by studies done in the context of several countries which mostly look at other important factors such as age and sex. We follow suit by looking at the age and sex distribution of Covid infected people in India
The data we use, however, is only a sample of all the Covid 19 patients in India up to the date 6th September, 2020. By 6th September, the number of total patients in India was around 42 lakhs while our dataset which we obtained from a crowdsourced website (www.covid19india.org) however has information about 1 lakh patients. While we cannot rule out the possibility of reporting bias in our sample, this is the only patient-level data publicly available in India which allows us to look at the difference in infection rates across sex and age groups.
Figure 1 shows a pattern that we expect in general. The incidence of Corona is much higher among the people falling in the 16–60 category with the maximum infection happening to the age group 31–45; the group which is more likely to go in crowded work or market place and get infected. However, there is an important gender dimension in this data; for most of the age groups, the number of male patients is twice the number of female patients.
Next, we examine the difference in case fatality rates (CFR) between males and females for different age groups. The CFR is obtained by representing the total number of death as a percentage of total cases. From figure 2, it becomes clear that the CFR is way higher for the older age group than the younger ones which is the exact opposite of the infection rate scenario. For people below 30, the CFR is below 1% for both males and females. For the age group of 31–45, the rate is a little below 3% for both males and females (2.95% for males and 2.65% for females). The CFR makes a significant jump as we move to older age groups. For people between 46 and 60, the rate jumps to 11% for males and 10.3% for females. For people aged more than 60, are much higher than the younger groups; the CFR for male in this age group it is between 27% and 41% while for female it stays between 21% and 32%. More interestingly, the difference between male and female CFR is much larger for the old age population.
Together, these observations mean that people aged between 16 and 60 have a much higher chance of contamination than the old age population. But once infected, the aged population face a higher risk of death than their younger counterpart. Moreover, within the elderly group, male members are more risk-prone than female ones. However, there is one caveat in our data. The overall CF rate across all age groups in our data is 7% for males and 6% for females. But this is a sample. For the population, on 6th September, there were around 42 lakhs Corona infected patients in the country of which around 71000 died. This makes the overall case fatality rate 1.7% which is much lower than the CFR that we calculate using our sample. In other words, in our sample death is over-represented which can be the result of reporting biases that are not very uncommon in a crowdsourced database. Nonetheless, the pattern involving the relative positions of different age groups and gender in terms of infection and CF rate is likely to hold even we could get individual data for all 42 lakhs patients.
How do we explain these observations? Given the paucity of data, we can only provide some speculative explanations. The group of people who go out more is more likely to get infected. The young, work-age population, therefore, are facing a higher risk of infection. Given the low women’s workforce participation rate in India, following the same logic, women in India are less likely to get infected. Both these speculations are confirmed by the same data. The death rate however, is higher for men and this is much higher for old age men than their women counterparts. One possible explanation can be the greater prevalence of the smoking habit among men in India. However, that cannot fully explain why CFR is so much higher for old men than old women. But without more detailed, patient-level data this cannot be ascertained. Besides these speculative explanations, our article tells us an important lesson — the proverbial funeral bell does not toll for everyone with equal likelihood. There is considerable variation across gender and age groups in terms of Covid 19 infection and death; some groups of people are more risk-prone than the other. We can combat the Covid 19 situation much better if our policies internalize this lesson.
