Tuesday, April 14, 2020

Does Testing for Coronaviris Reduce Death?

The Nobel Laureate Paul Romer explained his plan to restart the U.S. economy. He argues that conducting 30 million tests a day and allowing those who test negative to resume working would restart the economy sooner. Certainly, this is a smart idea that I have not heard anyone in this entirely infected world think about. Although they have been thinking about it for sure, so far, no government has provided any plan on how to restart the economy. Trump in fact said that he might stop testing altogether! 

Germans, on the other hand, have been testing people more efficiently. Say that there are five people in your bubble. They do a combined test for all five at once. If the result is negative then all five people are safe. If positive, they redo the test to each one in the bubble. So Germany could easily follow Romer's plan to restart the economy.

The population is very small in NZ so we have a very good chance that we can restart the economy in a week if the government follows these strategies. 

Ten days ago I began working on examining the effect of testing for Coronavirus on death. Ten days ago I found two data sets. The first one is published by the  EU (here) and the other is by Oxford University (here). The EU data set is large and reports data for deaths and infections by more than 200 countries and territories. Oxford University data set is smaller and reports data on testing for Coronavirus. The data are daily. Some countries report fewer data than other, different dates, and many have missing data. I combined both data sets to arrive at a balanced panel of 8 countries only that have continuous data from March 1 to March 31 for all three variables, tests, death, and infections.

I found that, on average, a one percent increase in daily testing for the Coronavirus reduces death by about 4 a day. When I allowed the coefficients to vary across countries, I found that the U.S. and Italy could reduce death by about 13 and 68 respectively. At a lower significance level, the Belgium and the U.K. could reduce death by about 2 and 32 respectively. Japan could reduce death by 25. So there is reasonable evidence in this small panel that testing for the virus reduces death. You can read the paper on Massey University Website in a couple of days (here). 

No comments:

Post a Comment