As a simple rule of thumb, it takes three weeks from the moment of infection to reported death, for most people who die from Corona. Which means that, we have to wait three weeks before our changed behaviour shows up in the number of reported deaths. Read more here: http://www.gapm.io/corona3
This is the third video in our series listed here: http://www.gapminder.org/corona
Here you can download:
—Slideshow: https://www.gapm.io/corona3_download_slides
—Video: https://www.gapm.io/corona3_download_video

— DATA SOURCES
Numbers are based on (but not identical to) the mean durations from CDC — COVID-19 Pandemic Planning Scenarios, Table 2: https://www.cdc.gov/coronavirus/2019-ncov/hcp/planning-scenarios.html#table-2
In this film we present our estimate of a common duration, which is several days shorter than the average duration of 26 days. Our purpose is to have a useful rule of thumb for what delay to expect between a behavioural change, and a change in the weekly deaths. We can’t use the longer averages for that purpose, because they are influenced (made longer) by a minority of cases that have much longer durations than the normal, for all three durations: days before symptoms, days of illness before death or and days before reporting. Long durations are quite common, and there are huge variations, but those later deaths won’t be the first change to indicate the effect of a behavioural change. Instead we think the within three weeks, if the curve dips, it's because of an absent of those that would have been reported after the normal duration which captures most deaths.
See data documentation here: http://www.gapm.io/corona3