Understanding Tamil Nadu’s Vaccination Drive
In my previous post, I had looked at the state-wise trends in vaccination rates in India. One finding was that Tamil Nadu was falling behind in terms of vaccination rate when compared to other similar populous states. It is indeed a matter of great surprise to see a state like Tamil Nadu fall behind in the vaccination drive.
I hence took a closer look at the data, by amalgamating vaccination data with district-level population information from the Census of India 2001 and 2011 and economic well-being information from the 66th and 72nd rounds of the National Sample Survey. The main aim is to explore some commonly provided explanations for Tamil Nadu’s performance, from a data perspective.
Chennai — The State’s Capital City
The starting point of this analysis is with Chennai, the state’s capital city. To our surprise, Chennai is one of the best performing cities in India, with a full vaccination rate (i.e. proportion of total population that have received both shots) of over 12%.
The same applies to partial vaccination as well (i.e. the proportion of total population that has only received one dose). The partial vaccination rate stands at an admirable 28.7%.
If the city is performing so well, why is the state falling behind? Well, it is obviously because the other districts are not performing as well as Chennai. A common response one gets is that it’s all a consequence of the relative economic position of the population across the districts. Can we establish this correlation?
The Rest of Tamil Nadu — Does Relative Economic Position Correlate with Vaccination Rates?
A proxy that I use for the measuring the economic status of the population is the mean monthly per-capita expenditure (MPCE) of a district. This is computed using the National Sample Survey Organization’s 66th round (conducted between 2009 and 2010) and 72nd round (conducted between 2014 and 2015).¹
Information on expenditure is only captured on a quinquennial basis, and hence these two are the most recent surveys we have. I extrapolate MPCE to 2020 by using the Compounded Annual Growth Rate (CAGR) of MPCE between the two rounds, for each district, after adjusting for inflation.²
The results for the full vaccination rate are as follows,
Clearly, Chennai is leading among all the districts, surprisingly a fair majority of the districts (all but for the Nilgiris to be precise) have a full vaccination rate of around 4% or less (even among other economically better-off districts such as Kancheepuram, Coimbatore, and Madurai).
One hypothesis could be that the people in the other districts do not see the need for vaccination given that COVID-19 infections were less prevalent in these districts when compared to Chennai.
We can test this by looking at the proportion of population that were infected (i.e. total number of confirmed cases divided by the total population) across the various districts. Figure 4, presents the results. It is puzzling to see why a district like Kancheepuram, which has a substantially higher infection rate, still has poor vaccination rates.
If not for economic position or relative prevalence of the disease, what could be the reasons behind the poor vaccination rates in the other districts?
Some questions that arise are if there is widespread vaccine hesitancy or if there are logistical challenges in getting the vaccines to the people? Irrespective, Tamil Nadu’s poor vaccination rate certainly warrants a closer look as the state seeks to ramp up its vaccination efforts. There is an urgent need to identify potential areas of weaknesses and address them.
1: I use the MPCE URP for the 66th round NSS, and the “Household usual monthly consumer expenditure” in the 72nd round. I consider the 66th round to be in 2010 and 72nd round to be in 2015 for inflation purposes and use the national Consumer Price Index (CPI)for India to deflate the data. The CPI data is from World Bank.
2: Tiruppur was formed from the Coimbatore and Erode districts in 2009. Therefore, it is not captured as a separate district in the 66th round of the NSS. Hence, I take the simple average of the CAGR of Coimbatore and Erode to extrapolate the numbers for Tiruppur from 2015 to 2020. I use the pre-2019 districts, as much of the underlying data is from this period. Hence, Ranipet and Tirupathur districts fall under Vellore. Tenkasi district falls under Tirunelveli. Chengalpattu district falls under Kancheepuram. Kallakurichi district falls under Villupuram.