Data show India’s coronavirus lockdown may not be working

Prime minister Narendra Modi’s government, which today (April 14) extended the lockdown till May 3, has argued that all this was necessary to “flatten the curve” of infections. However, the evidence it offers to back this claim is entirely unconvincing.

An external affairs ministry official, for instance, claimed on April 9 that in the absence of a lockdown, there would have been 820,000 cases by April 15, according to an internal assessment. When asked about it, the Indian Council of Medical Research (ICMR) initially said no such study existed and later clarified that the claim was made based on a “statistical extrapolation” by the ministry of health as depicted in chart 1.

The problem with this claim is that we are not told what, if any, model was used to arrive at these numbers. No such model has been made public or even talked about by government officials. As it happens, the only model that ICMR has shared publicly is not about the lockdown, but an analysis of the border screening measures India introduced relatively earlier than other countries.

So what is the right way to know if the lockdown helped?

In the absence of an official reference model in India, all we have to go on is aggregate data on total confirmed cases nationwide, which is updated every day. The chart below plots the natural logarithm of total cases on the vertical axis with the time index on the horizontal.