Hello and welcome to Episode 19 of The Moving Curve. I’m Rukmini, a data journalist who lives in Chennai. Every night on this mini-cast, I examine one question around the novel coronavirus epidemic in India. Tonight I’m considering this question: should we trust the mathematical models that show that India’s lockdown prevented over 800,000 new cases?
Here’s that chart provided by the government on what it says is the impact of the lockdown
Gautam Menon, the professor of physics and biology at Ashoka University whom I spoke to in Episode 17 on testing, had a terrific piece in The Wire asking if we should believe what the models say: https://science.thewire.in/the-sciences/covid-19-pandemic-infectious-disease-transmission-sir-seir-icmr-indiasim-agent-based-modelling/
If those are things the health ministry would like to collaborate with epdemiologists and modellers on, that would be a great direction to move in. Instead, the government’s belated embrace of models is just plain shoddy. That graph is a simple numerical exercise with none of the demographic and epidemiological assumptions that went into the others. Of course a lockdown is going to produce fewer cases in a linear or exponential progression than without a lockdown — even the models predicted that. For me this is the real doomsday prediction — that we ignore evidence when it seems to sound like things might go badly, and then we borrow the barest of its bones to make government measures sound successful. This isn’t to say that they aren’t — it’s just to say that their statistical underpinnings are weak.
Thank you for listening. This episode was edited by Anand Krishnamoorthi. Tomorrow — a new question.
