“The conundrum of social distancing — Situation Analysis of COVID-19 on Mar.29 by acesuzuki” | by Eisuke Suzuki

8 min readMar 31, 2020

“The conundrum of social distancing — Situation Analysis of COVID-19 on Mar.29 by acesuzuki”

This is my third article on COVID-19. If you have not, I’d recommend you to read the ones in the last week (the second article) and the previous (the first article). In this week’s version, I updated my analysis and added some thoughts on the conundrum with the difference of degree social distancing can work.

I’ve kept saying that we should treat “# of deaths per capita” as the key index of each country’s performances, and the timing of introducing general “social distancing” policy based on the index is extremely important to determine the fate of the country.

The followings are the graph and the table shown in the last article. The former shows the number of cumulative deaths per capita (100 Mil) and the latter the number on the day when major political reaction including any social distancing resumed.

Here are what I mentioned in the last article.

Italy: It has well surpassed Hubei now…

Spain: The curve looks as sharp as Italy. This may be because Italy had introduced draconian policies into Lombardy days before they did it nationally, while Spain did not have that kind of local measures.

Iran: It has been a bit worse pace than Hubei as expected.

France: Following the same type of path as Iran as expected.

UK: They could have saved many more lives if they decided on general social distancing when the likes of Germany, France, USA, and Spain reacted. The index was at 12 on Mar.13. They postponed the decision instead, and finally made a move on Mar.16 when the index reached at 51.

USA & Germany: Although the rapid increase of “nominal” # of confirmed cases is terrifying many, both of these countries will be kept relatively unscathed.

Adding the data of last week, the graph now looks like this.

-Countries with the key indicator in their 100s-

Italy: The death toll keeps increasing sharply. This will probably be eased next week as we see a sign of “flattened curve” as I discuss later.

Spain: The curve actually looks a bit worse than Italy. It looks like to become the worst case among the countries in my analysis.

-Countries with the key indicator higher than teens but lower than 100-

Iran: Keeping a bit worse pace than Hubei as expected.

France: It has surpassed Iran now. A bit worse pace than I expected.

UK: Following the same type of path as Iran and France.

-Countries with the key indicator lower than teens-

USA & Germany: In my first article, I expected they would be the same level as S.Korea, but that seems to have been a bit too optimistic.

The following is a closer look of the same graph.

Both of them have quickly surpassed S.Korea now. However, we can say that these countries are still relatively unscathed.

The following graph shows the “curve” of “Weekly # of Confirmed Cases by Countries/Region”. As I have always mentioned, focus on the “shape” of the curves, and do not compare the numbers of each countries here.

Here is what I said in the last article- “I’m expecting the decrease of the number of cases in Iran for the next week. USA and European countries in this analysis will still see the rapid increase for the next week. I expect it will be two weeks later to see the start of the effect of social distancing.”

First of all, I was right with USA and European countries, although I did not expect this much of increase with USA. This is most likely due to the massive increase of testing, and I don’t think the curve fairly reflects the reality.

As shown in the case of USA, the caveat of this analysis is a shape of the curve can become misleading when a country changes its testing policy drastically on the way.

Secondly, I was wrong with Iran.

My hypothesis is that it would take 2 to 3 weeks to see the effect of social distancing in the epidemic curve as it takes about 1 week for incubation and 1–2 weeks to get tested and to see the result.

It didn’t happen to Iran, however.

So, social distancing isn’t working in Iran? It worked in China and in S.Korea, didn’t it? To dig into this conundrum, I checked the time span between the date nationwide social distancing policy announced and the peak of epidemic curve.

The following graphs are showing a daily epidemic curve of each countries/region. The red arrow shows the date nationwide social distancing policy announced. The blue shows the timing of the peak of the epidemic curve. “Day 1” means the first day when the country sees no consecutive zero case.

The first one is Hubei. It took 12 days for Hubei to see its social distancing effect, and the curve quickly went down after that.

The second one is China (w/o Hubei). It took only 9 days for them.

The next up is S.Korea. Only 6 days. Are you kidding???

Well, in the case of S.Korea, the fact that they have tested almost all the suspicious cases with very well known “drive through” system probably made them possible to erase the time gap between having symptoms and getting the test result.

These are beautiful examples of social distancing effect. But now the conundrum.

Here is Iran. When I wrote the last article, they seemed to have hit the peak, but we are seeing a rapid increase of the curve again. So, even after 4 weeks of social distancing, Iran has not seen its effect yet.

There is another country which makes me puzzled. It is Japan.

I said Japan seemed to have succeeded in flattening the curve in my last article. It didn’t at all, however.

There are several hypotheses to explain this conundrum.

1) The measures were not strong enough.
This is the case for Japan. As I explained in my last article, the measures in Japan were very relaxed ones. This may apply to Iran as well. The measures of Iran look stronger than Japan, but weaker than countries like China or Italy. That being said, this hypothesis does not explain everything, as the measures in S.Korea were not that tight either.

2) People are not so compliant to the social distancing policy.
This is the case for Japan. Many Japanese had kept going out to work, to enjoy social gatherings, or to work out in fitness gyms. The following graph shows the change of number of people visited Shinjuku, which is well known crowded area of office/entertainment/shopping, in weekdays. We see a slight drop starting from March, but “less than 10% drop” doesn’t seem to be enough.

3) The timing of introduction of social distancing policy was too late.
This is consistent with my central hypothesis. The later you introduce social distancing policy, the harder you realize the effect. Iran fits into this case.

4) The earlier reported number of cases have been downplayed by governments.
There is a “rumor” that Japanese government tried to downplay the situation of COVID-19 to seek for the possibility to hold the Tokyo Olympic 2020 as scheduled. Nobody knows, but the data of the number of deaths would reveal if that is the case. So far, it doesn’t look like so.

5) The aggressiveness on testing or testing technique changed on the way.
Japan hasn’t changed its testing policy or technique. I’m not sure about Iran.

Combining the likely hypotheses and the policies of each countries, the situation looks like the following table.

As the general social distancing policies were resumed between Mar.9 and 13 for Italy, USA, France, Germany, and Spain, this week will be the time for us to see the effect if any.

My prediction is that we will clearly see it in Germany at least, and hopefully in USA. In case of USA, the measures look to be stronger than Japan, but weaker than European countries, and I’m not so sure about how much people are compliant to the measures there.

I want to add that there is a sign of hope in the data of Italy. They may already have hit the peak. If we see the clear decrease of the curve in the next analysis, it will be very encouraging.

1. This may sound wild to many readers, but there is a very interesting hypothesis that taking BCG vaccination may strengthen immune to the novel coronavirus. The hypothesis is well explained by my former colleague in the article “If I were North American/West European/Australian, I will take BCG vaccination now against the novel coronavirus pandemic.”

2. The above hypothesis may explain why we saw worse number of cumulative deaths per capita for France comparing with Iran, which were contradictory to my prediction. Iran currently has universal BCG vaccination program while France does not.

3. Japan seems to have succeeded in slowing the start of epidemic by introducing its rather weak social distancing policy at very early stage, but the policy is probably not strong enough to prevent the rapid increase of the epidemic curve forever. The government seems to try “gradual tightening” strategy, but I would recommend “lockdown first and gradual loosening” strategy at this stage as we can’t avoid the risk of “over medical capacity” with the former.

[Source of the data in the analysis]

“Coronavirus disease (COVID-2019) situation reports” (World Health Organization) for the number of cases and deaths

Various news websites for the date major political reaction resumed

Website of Tokyo Metropolitan (https://stopcovid19.metro.tokyo.lg.jp/) for the change of number of people visited Shinjuku

Written by Eisuke Suzuki