The Coronavirus Outlook: Worse than Expected?

In my Tuesday podcast, I expressed extreme skepticism about the smooth trend in the reported coronavirus (nCoV) numbers. Everyone on the ground--and I mean everyone--agrees that infections and deaths in Wuhan were vastly under-reported through most of January. Hardly any test kits were available and sick people were routinely just told to go home and stay isolated there. If reporting has since improved, we should have expected some kink in the line as all the newly uncovered cases were revealed. But we haven't seen anything like that. The line has been perversely smooth--a bit like China's GDP numbers. The signs of managed data are everywhere.

Well, earlier today in China, we finally got our kink in the line.

New total cases per day, after slowing to under 3,000, jumped by 15,407. New total deaths per day, after showing a linear trend of about 100, jumped by 253. [Data as of 1 pm EST, 2/13/20]

Why the jumps? Unclear. The NYT suggests it is due to a shift in the criteria used in Hubei province for diagnosing patients as positive for nCoV.

Until yesterday, officials were using tests that directly detect the virus's genetic signature. These tests are complex and time…