The chance of dying from coronavirus is hard to know because iffy numbers go into the calculation.
I believe that the number of people who die is kept track of pretty closely in the US and many countries. Therefore the number of deaths should be known. Are all people who die of infections tested for the virus? If so, number of deaths from coronavirus (so far) is well known. If not, the fatality count is an underestimate.
I don’t believe that the number of people who have gotten sick from the virus is known. Like other strains of flu, coronavirus causes symptoms ranging from “mild to severe”. Therefore many people who got sick and merely stayed home from work would not have been counted in total number sick. Number of people exposed, including people with no symptoms, is even less well known.
Data from an Accidental Experiment
An accidental experiment gives more authoritative numbers. Passengers on the cruise ship Diamond Princess were exposed to the virus. According to latest news reports, all 3711 passengers and crew were tested, 705 tested positive, and so far 7 have died (6 by Feb 28, followed by 1 more. So:
deaths/exposure = 1/530
deaths/infection = 1/101
More people from the ship could die, raising the death rate. The passengers had been stuck on the ship since before February 3, but I don’t know how long the disease can incubate or worsen.
Another point to consider is the demographics of the disease. Worldwide, most deaths have been people who were 65 or older, though at least 1 48-year-old and 1 36-year-old have also died. This could be an important point for how well the economy deals with the disease. The lower risk to people of prime working age could make people more willing to risk getting the disease in order to stay on the job. Of course these people should also take into account whether they regularly come into contact with more vulnerable people who they might infect.
The demographics of the ship also affect our calculation of risk. The Japanese National Institute of Infectious Diseases lists statistics from the ship. About 58% of people on the ship were 60 or older, compared to roughly 25% in the US . If all the deaths were among age 60+ people on the ship, that comes to a death rate of 1/359 for that age group (and zero for younger people).
Extrapolating to the US
Health authorities throughout the world are working hard to catch instances of the disease and stopping its spread. Due to the nature of the disease, stopping its spread seems impossible. There is value in slowing its spread in order to give society more time to prepare for it, but I am guessing (based on zero professional knowledge) that a large fraction of the population will eventually be exposed to the disease. If the ship’s death rate holds for the US as a whole, that implies a total of about half a million deaths. If we apply the ship’s death rate for ages 60+ to ages 60+ in the US, that’s about 200,000 deaths.
So worst case scenarios extrapolated from the cruise ship statistics show low chances of any one person dying from the disease, but with a big total impact of up to hundreds of thousands killed. For comparison, the 2009 H1N1 influenza outbreak, which was considered very serious, caused around 300,000 hospitalizations and about 12,000 deaths in the US. On the other hand, the influenza epidemic at the end of World War I killed almost 700,000, mostly young adults, out of a total US population of about 100 million.
If efforts to stop the spread don’t work, we could face many more deaths than recent flu pandemics, but many fewer deaths than the worst the United States has seen. It is less challenging to protect part of the population than all of it, and we may be able to shield more vulnerable parts of the population from exposure. And of course if a vaccine can be developed quickly, the danger would decrease drastically.
Last Update: 6 Feb 2020