Friday, February 28, 2020

The Human Mind, Risk & Statistics

Humans did not evolve to think of risk in terms of statistics. Because of that, most people are bad at assessing risk. Slow moving risks and unremarkable events due to well-known risks are hard to assess. This chart shows relative risks of death from various causes.




Terrorism
Fear of death and injury after a terrorist attack on US soil are grossly overestimated. People estimated there was about a 30% chance of personal involvement in a terrorist attack in the next 12 months.  Emotional reactions of either fear or anger alter risk perception. This is a great example of how emotional responses impairs our ability to think rationally. Lack of sleep makes the irrationality problem worse. Emotion-driven irrationality influences policy and that can make policy more irrational.





According to one analysis, the annual risk of injury from terrorist attack in the US is about 1 in 678,000 and the risk of death is about 1 in 3.8 million.





Flu virus vs terrorists vs coronavirus
Seasonal flu that tends generally starts spreading in the fall and peaks during the winter months. Flu infections can become life-threatening from complications such as pneumonia. The Centers for Disease Control and Prevention (CDC) estimated there were 61,100 flu deaths for the 2017-2018 flu season. The rate of death from flu virus infection is about 0.1%, meaning that about 1 infected person in 1,000 will die. To be rational, people would fear flu far more than terrorists, but most don't. Terrorist attacks are unusual, spectacular and heavily reported in the media. That elicits fear and/or anger and increases or decreases the appearance risk. Citing statistics doesn't seem to change that much.

By comparison, coronavirus appears to be more lethal with an associated rate of death of about 0.4%. That makes it about four times more lethal than influenza.

Infections and spread of flu is limited by annual vaccinations and antivirals that make an infection less severe. At present, there is no drug or vaccine available to treat or prevent Coronavirus infections. Most Coronavirus infections are fairly mild and resolve on their own. The unknown risk about Coronavirus is how many people will be infected. At present, quarantines are used to limit spread of the virus. If the quarantines work reasonably well, flu will pose a greater risk of death than Coronavirus. If the quarantines do not work and the virus spreads freely, it is possible that Coronavirus will turn out to pose a higher death risk.

How the Coronavirus outbreak will turn out cannot be predicted with certainty, but quarantines in the US are likely to work well. Time will tell if that prediction is accurate or not. We will probably have a fairly good feel for the potential risk within the next 3-4 months.

No comments:

Post a Comment