March 09, 2020
Lockdown measures are not working
If the Chinese Covid-19 data are even remotely correct, they are telling us that the severe lockdown measures in Wuhan have worked. Italy is at approximately the same stage in the spread of the virus where the Hubei region was in mid-January. But Italy’s lockdown measures are not of the same quantity and quality as China's. As newspapers in Germany and the UK reported, there was no problem getting in and out of Milan. FAZ notes that the critical Brenner motorway, the biggest connection between northern and southern Europe, was unperturbed yesterday. There were no controls at the border or anywhere on the motorway itself. Journalists also tried to get from Bologna, the capital of Emilia Romagna, to Rome without any disturbance. We realise that some of the measures will only take effect today, but the early signs are that this is not a lockdown in the Wuhan sense.
We said before that it was a deadly mistake by the EU not to close the borders several weeks ago because the infrastructure is there. A border closure would have prevented or slowed down the spread from Lombardia and Veneto to Austria, Germany, Switzerland and France. Severe lockdown measures will still be effective even now, but the longer they are delayed the more intrusive they will eventually become.
We estimate that France and Germany, each with over 1000 reported cases now, are approximately one to two weeks behind Italy. In Germany at least, we see even less readiness to act than in Italy based on a false belief that the government has the situation under control. This is aggravated by a high degree of statistical illiteracy among journalists and politicians.
Public discussions in Germany are still informed by the illusion that Covid-19 is a slightly worse case of the flu. This is wrong. Covid-19 is much more contagious than the flu, and the death rate is higher. The WHO now puts the mortality rate at 3.5%. In a worst-case scenario, where 70% of the population catches the virus, this would translate to about 2m deaths in Germany alone.
We note a widespread failure in public discussion to think about catastrophic risks. The stochastic components of economic methods assume known probability distributions. This gives people the illusion that they are in control of situations when they are not.
One of best articles we read over the weekend is from Matthew Syed in the Sunday Times. He illustrates to failure to grasp large risks in the following way. The chances of dying in a road accident are statistically higher than the chances of being killed by a terrorist. But a viewpoint that focus on averages ignores extremes at its peril. The chances of your community getting annihilated by road accidents is near zero. Your community is at a significantly higher risk from a terrorist attack. For the same reason, Covid-19 should not be compared to previous diseases or other risks. If you make comparisons based on averages, you may end up underestimating the risk.