The organization Our World in Data is running an article on the severity of government lockdown policies used to deal with the Covid-19 disease, “How Are Governments Responding to the Coronavirus Pandemic?”. It is shows rankings of the severity (“stringency”) on nine measures of lockdown policies throughout the world; it also addresses testing and tracking. A click-able global map showing overall lockdown severity ratings for individual countries can be found here.
The measures Our World in Data uses to calibrate lockdowns are the following: the severity or comprehensiveness of school closures, work closures, cancelling public events, restrictions on public gathering, staying at home, restricting public transport, and limiting internal & international travel. Click on any country and see the 0-100 overall score. Some of the scores may or may not conform to your impression of the government’s policy based on what you are reading; I took their ratings as given for the purposes of my analysis, below.
Unfortunately, the article did not also asses the success of these policies. I e-mailed Our World in Data asking them to do so; I have received no response.
Thus, I looked at Our World in Data’s analysis on confirmed Covid-19 deaths per million in 17 selected countries.
The table below shows what I found.
(I characterize those countries that score from 80-100 as “most severe;” those between 60-80 as “less severe,” and those less than 60 as “least severe.” The data both for severity and death are mostly as of May 8.)
So what do we get?
All of the most severe lockdown countries shown have almost all of the worst death rates on the table; these are all Western European countries. Add to the high death rate countries the US (228 deaths per million) and Sweden (301/million) from the lower severity groupings.
Except for the US, the less severe lockdown countries all have a fraction of the deaths per million of the most severe lockdown countries.
The least severe lockdown countries, which surprisingly include China, all have tiny death rates, except Sweden, which – as mentioned – ranks with the highest death rate counties.
There is much this quick analysis does not measure: including the competence of the implementation of the various policies and the promptness of the policies given a nation’s first Covid-19 positive test case or death, or even before either.
Also, given the possibility that not all of the countries listed here are past their peak death rates (and given the predictions by some of second waves of Covid-19), there may be more to tell some weeks from now (or months from now if the second wave predictions prove out). Prolongation of the first wave, and the second wave problem, may be particularly serious for countries that do not implement (or continue) effective and comprehensive testing, tracing and individual quarantining procedures as they relax various lockdown policies.
Despite some hype that cherry picks the data, the US appears to be past its peak death rate. Nonetheless, the US may be especially vulnerable to prolonged or additional waves of death as states undo their lockdown policies while not simultaneously requiring widespread testing, follow-up tracing and individual quarantining of people carrying or exposed to the virus. The US was very late to introduce testing, let alone reliable testing, and even today tests in the US public have shown poor reliability and/or are not even approved by the FDA. That they are allowed to be marketed is stunning. It will exacerbate the potential problem in the US.
However, people anxious to point fingers for the demonstrated incompetence in the US should not point just at the chief executive or selected governors. Evolving journalism on the US handling of the Coronavirus shows that the CDC, HHS and FDA richly deserve some of the blame — however, seriously diagnostic research on the role of NIAID and it head remain very hard to find. It would have been helpful if Our World in Data had attempted to measure competence as well as strictness – the likely controversy notwithstanding. Given the blatant demonstrations of incompetence on a key issue like testing, it is hard to imagine the US scoring as competent as other developed nations. Nonetheless, it remains anathema in the mainstream media to criticize anyone beyond Trump (and to a lesser extent Pence) for fear that to criticize the CDC, FDA and HHS – let alone Dr. Fauci – is tantamount to supporting Trump politically — or even as a normally functioning human being. No one benefits from this mentality, except Trump for the most part: he avoids criticism of the entire structure of his reaction to the Coronavirus.
Looking at the individual Our World in Data ratings of severity, it would be extremely interesting to assess what, if any, individual measures generally seem to coincide with high and low death rates. For example, the data for the countries listed above show spectacular success in five countries: China, Japan, South Korea, Taiwan and Australia. Three of them (China, South Korea and Japan) have more available public testing – compared to most of the western European countries, which generally have had both high death rates and less available testing. (The frequency of testing is not among the nine measures of severity discussed heretofore, but it is addressed in a map and assessment in the Our World in Data article.)
The data on high death rate countries show some coincidence with more severe restrictions on perhaps three measures: workplace closures, strict stay at home policy and more severe public transport closures. While it is counterintuitive that a policy like workplace closures would cause additional deaths, it is thoroughly worth further inquiry to determine if strict stay-at-home policies did, rather are.
Promptness, the imposition of each of the policies upon the discovery of the first Covid case or death, if not before, is another area that clearly needs further inquiry. Especially telling could be whether promptness in one area, such as testing and tracking, trumps the need for any of the lockdown policies, especially the ones that result in economic collapse.
These and my other cursory observations are not an attribution of cause and effect, but they clearly are reason to conduct more study – with finer grain data and statistical and other appropriate expertise.
You may complain that mine is an ex post facto analysis selecting only certain countries. Select your own from the map at the bottom of the article here.
Don’t like Our World in Data’s overall scoring of the nine measures in their penultimate map? Select your own individual ones in the body of the article.
Don’t like the measure of success/non-success I used: Our World in Data’s Covid-19 confirmed deaths per million? Pick your own; Our World in Data shows a long list of them, but be careful: some of the commonly used ones, especially positive test cases, will get you a result that is a function of the prevalence of tests, not of the disease. But even that, in a test, showed about the same results as above.
Don’t trust the China data? Take it out; doing so changes nothing.
This is not a statistical analysis; it is a brief study of cases. Please feel free to do your own analysis on the question of severity of anti-Covid policy and measures of success, using data of your own selection. Please show your results and your sources.
The key is to further probe what in the less and least severe lockdowns are enabling the vastly reduced death rates. In that respect the South Korea and Taiwan models appear very promising. In addition, the severe lockdown policies need further investigation to determine what may be killing people needlessly and should never be repeated again.
The most, and even some less, severe lockdown countries have crashed their own economies. But some have not. For example, the unemployment rate in the US is now passing 14 percent while that in Germany is just 5.9 percent, as a result of Germany’s far less onerous and carefully disease-monitored workplace policy. Germany has only 35 percent of the death rate of the US – and just 17 percent of the average of the seven other West European countries shown.
Very clearly, the severe lockdown policies are not the only option available to try to protect people. The available data need further investigation to find the best of both worlds: low death rates and lesser economic penalty. To be avoided is the worst of both worlds, which appears to be the case for the US and much of Western Europe.
The mainstream knee-jerk reaction is to howl that the lockdowns saved us from greater death and not imposing them means even more death. That is clearly not the case in some countries. There is a real need to establish exactly what made that possible. No one should settle for the worst of both worlds when both lower death rates and lesser economic penalty are empirically demonstrated.
When what engenders low death rates and lesser economic impact are established, there will be many politicians and their advisers (epidemiological and otherwise) who have fostered the high death rate, economy-trashing policies that will need to be called to account for the human and economic disasters they have helped to create.
The sooner the better.