Go to the interactive web app for detailed projections.

Below the latest estimates for how well different jurisdictions are able to contain the spread of Covid 19 (updated as of 15 June 2020)

The dots represent best estimates of the model, and bars are uncertainty margins (95% confidence intervals) as estimated by the Bayesian model we employ. The models estimates the effectiveness of government policies in reducing the rates of new infections across a range of jurisdictions. These estimates are used to project the evolution of the epidemic over the next few weeks.

If g < 0% then new infections are declining, and the epidemic should eventually fade away if the current policy remains in place. In practice, in most geographies this will take too long. For instance, at the current rate of decline in Italy it would still take approximately a year to have local transmission of the disease fully recede.

For further background:

• Model code in R, posted on GitHub

• Pre-print of the academic paper, including methodology

(C) 2020 Phebo Wibbens, Wesley Koo and Anita McGahan.

All code and data are available through the GNU General Public License v3, but without any warranty, not even the implied warranty of merchantability or fitness for a particular purpose.