A CSH overview on the COVID-19 pandemic

UPDATED ON NOVEMBER 27th 2020

The COVID-19 epidemic:

The current COVID epidemic will undoubtedly become a significant landmark in the history of the human kind. The reason for this does not lie so much in the health consequences of the epidemic itself, however impressive these may seem. COVID-19 is estimated indeed to have killed a little more than 1.4 million people worldwide since its beginning a year ago. Of course, this if far from being a final assessment as the epidemic is still spreading. Yet, for the moment, this dead toll is somewhat modest when compared to those of past epidemics, be they relatively recent, or more ancient. The Hong Kong flu epidemic of 1968-1969 for instance is estimated to have killed from 1 to 4 million people worldwide, for a world population of 3 billion (as compared to the 7 billion people who live on Earth nowadays. The 1957-1958 H2N2 pandemic is also estimated to have killed around 1.1 million people worldwide, for a world population of 2.5 billion. And these figures are obviously very modest when compared to the 20-50 millions of death caused by the Spanish Flu in 1918-1923 (over a population of 1.5 billions), or to the third of the European population that was apparently killed by the Black death of 1346-1353.  The health consequence of COVID-19 are also modest when compared to those of other major causes of death – many of them avoidable – that the human kind is currently facing. According to the World Health Organization, in 2017, 1.3 million people worldwide died from road accidents, 15.2 millions died from cardiovascular diseases, and 9.6 millions died from cancers.

The main reason why COVID-19 will become an episode of significant importance in the human history is the astonishing fear it has caused, and which is responsible for the adoption of policy measures of unprecedented stringency to prevent its spreading. While the actual impact of those measures on mitigating the health consequences of COVID-19 are far from clear, their dire economic and social consequences are amazing, and unprecedented in the last century. The following table, showing some estimations of the economic growth rate for 2020 in some of the largest economies, is amazingly clear from that perspective.

Country name Estimated GDP growth rate 2020 (%)
(source: The Economist, October 4th 2020)
Australia -4.5
South Africa -8.0
Brazil -5.2
Canada -5.8
China 1.7
France -10.2
Germany -5.9
India -8.5
Italy -10.4
Japan -6.4
Pakistan -2.8
Russia -5.7
South Korea -1.8
UK -9.5
US -5.3

If we except China, all countries of the table are therefore expected to experience severe reduction in their national income. India is expected to see its GDP drop by more than 8%, something that has never happened in the country since its independence. France, whose GDP is expected to fall by more 10% this year, will also witness the most severe recession of its history. Behind these (spectacularly) negative numbers lie the dire realities of 150 millions of people worldwide who have been plunged into the hell of extreme poverty (defined by an income of less than 1.9 US dollar per day in purchasing power parity) according to the most recent estimate of the World Bank. It is the first time in the last 20 years that the world is experiencing an increase in the rate of extreme poverty.  India, which concentrates one third of the world’s population of extremely poor people is obviously heavily concerned by this descent into hell.

COVID-19 is hurting us. The policies that we have put into place to prevent its adverse health consequences are hurting us even more. This is what makes the COVID-19 epidemic so exceptional.

 The next table provides numbers that describing the current state (as of October 2020)  of the COVID epidemic in East Asian countries as well as in a few others that are “salient” in terms of their population size, severity of the COVID, or their geographical localization. Countries in this table are ranked decreasingly in terms of the lethality rate of the COVID-19. As can be seen, countries are extremely diverse on this matter. In the world’s most severely affected country so far, Peru, more than a thousand people per million have died from COVID-19. This number is quite impressive, especially considering the severity of the lockdown implemented in Peru and the relatively young age of its population (half of which being below 28). Latin American countries are all on a par from that point of view. All of them have paid an extremely severe dead toll to COVID-19, despite the fact of having a rather young population. At the other extreme, one finds East Asian countries such as Bhutan, Cambodia, Laos, Sri Lanka, Taiwan, Thailand and Vietnam who have seen less than 1 person out of a million dying from COVID. If one excepts Hong Kong, Myanmar, Japan and the countries of the Indian subcontinent, the overall lethality of COVID in East Asia is everywhere below 10 deaths per million, which is extremely low (in average a hundred time lower than what is observed in Latin American countries). India is certainly the East Asian country that has paid the highest dead toll to COVID-19 (a bit less than 100 deaths per million) so far, even though this dead toll, is 8 to 10 times lower than what is find in the most affected European and Latin American countries.    

Country Number of cases Number of deathsPopulationMedian age Number of tests per million peopleNumber of deaths per million people
Belgium558 77915 61811 539 32841,4485054,71353,5
Peru949 67035 59532 510 45328,034853,01094,9
Spain1 606 90543 13146 736 77642,7346657,9922,8
Italy1 431 79550 45360 550 07545,5336722,6833,2
Argentina1 374 63137 12244 780 67731,745490,8829,0
UK1 527 49555 23067 530 17240,5516981,8817,9
Brazil6 071 401169 197211 049 52732,630426,2801,7
US12 777 371263 687329 064 91738,1539931,7801,3
Mexico1 049 358101 926127 575 52928,317721,7798,9
France2 144 66049 23265 129 72841,4394591,5755,9
Colombia1 254 97935 47950 339 44330,093507,2704,8
Sweden208 2956 40610 036 37941,2224746,2638,3
Iran866 82145 25582 913 90630,370293,5545,8
Poland876 33313 77437 887 76840,7151255,6363,5
South Africa769 75920 96858 558 27027,190353,9358,1
Canada337 55511 52137 411 04742,2290867,4308,0
Russia2 114 50236 540145 872 25639,6500092,3250,5
Germany946 64814 58383 517 04547,1316784,0174,6
Saoudi Arabia355 4895 79634 268 52827,5271257,6169,1
Turkey453 53512 51183 429 61530,9206708,6150,0
India9 177 840134 2541 366 417 75428,197029,498,3
Maldives12 75846530 95328,2435019,786,6
Philippines420 6148 173108 116 61523,547848,875,6
Egypt113 3816 560100 388 07323,9N.A65,3
Indonesia502 11016 002270 625 56830,213230,359,1
Nepal222 2881 33728 608 71024,157746,246,7
Bangladesh449 7606 416163 046 16126,716455,239,4
Australia27 84390725 203 19838,7385236,436,0
Pakistan379 8837 744216 565 31823,824089,535,8
Myanmar80 5051 76554 045 42028,25879,632,7
Japan132 3581 981126 860 30147,329195,815,6
Hong Kong5 7021087 436 15444,4537223,014,5
Malaysia56 65933731 949 77728,579129,110,5
South Korea31 35351051 225 30841,855942,710,0
New Zealand2 039254 783 06337,9258030,35,2
Singapore58 165285 804 33734,6728399,64,8
Sri Lanka20 9679421 323 73332,835061,34,4
China86 46946341 433 783 68637,4N.A3,2
Thailand3 9226069 037 51337,717609,60,9
Vietnam1 3163596 462 10630,5N.A0,4
Taiwan618723 773 87640,74523,40,3
Bhutan3860763 09227,6N.A0,0
Cambodia306016 486 54225,3N.A0,0
Laos3907 169 45523,0N.A0,0

A common explanation put forth to explain the diversity of the countries in terms of their COVID death toll is their age structure. COVID typically hit very severely old people, and it was recently recalled by the French president Emmanuel Macron that 90% of the French people who die from COVID are above 65. Hence, one could expect countries with a young population to pay a lower dead toll to COVID than countries with an older age structure. Yet the Figure 1 below shows that this impression is not really corroborated by the existing data. For one thing, Latin America countries have a very young population structure and yet they pay an impressive dead toll to COVID. At the other extreme, countries like Taiwan and Japan have a rather aged population structure and do not suffer much from COVID mortality.

Figure 1: Number of COVID related deaths/million against countries’ median age

Another common explanation invoked to explain the amazingly different performances of the countries in avoiding COVID-19 mortality is their differing testing capacities. Yet, here again, there does not seem to be a significant correlation between a country’s testing capacity – measured by the number of tests performed per million people – and its performance in avoiding COVID-19 test. As shown on Figure 2 below, while countries who pay the largest dead toll to COVID tend to be those who test the least, the negative relationship is rather weak. One reason for this may lie in the fact that the COVID epidemic and the testing policy are simultaneously determined. On the one hand, an wide testing policy may reduce the death toll from COVID by easing the tracking of people, and preventing therefore the spreading of the disease. On the other hand, when COVID is very active and people get infected and develop symptoms, they want to be tested. This induces a positive relationship between the intensity of COVID and the number of tests done. The balance between the two effects is what is captured on Figure 2.

Figure 2: COVID death toll by countries’ testing capacity

But all in all, the main causal explanation for the cross-country differences in COVID mortality remains far from clear.

Are lockdown policies Effective? a cross-comparison of France and India

A widely discussed issue pertaining to the management of COVID-19 is the impact of lockdown policies implemented in various countries to limit the spread of the virus. France for instance is currently under a second lockdown (that has been extended to December 15th by the French president). India is still in the process of eliminating the last restrictions of the very severe lockdown it has implemented on March 22 2020. How successful have been these lockdowns in reducing COVID mortality? Answering a question like this is not easy, because we don’t know the counterfactual situation that would have been observed in absence of those lockdowns. However, the historical juxtaposition of the timing at which the lockdown policies have been implemented and the daily number of COVID death provides some useful insights on those matters. Figure 3 below show the daily number of COVID deaths observed in India since March 1st 2020, along with the various steps of the lockdown policies implemented by the Modi government, starting from the first stringent lockdown initiated on March 22nd to the current phase 5 and 6 of the delockdown. 

Figure3: number of daily Covid deaths in India during the different lockdown phases

This figure is suggestive of a failure of the Indian lockdown episodes to affect the lethality of the COVID. Any policy restricting the contact between individuals and the circulation of the virus at time t will see its effect on COVID lethality some 20 to 25 days later. Indeed, it takes between 5 to 7 days to be contaminated by the virus and it takes again 10 days after the beginning of the symptoms for the disease to evolve into the dangerous and potentially lethal phase. Hence, this graph suggests that the first phase of the lockdown has slowed down a bit the daily lethality of the COVID (one can see a little drop in the number of death roughly in the first two weeks of April). However, the progressive alleviation of the lockdown initiated on April 14th and after seem to have generated a continuous upsurge of the daily death that has peaked in mid-September without any clear connection to the severity of the lockdown (which has regularly decreased over the period). Hence, except perhaps at the very beginning of the epidemic, lockdown policies in India seem to have had no visible effect on the dynamic of the daily number of daily deaths. To the contrary, it looks as if the reduction in the severity of the lockdown initiated on August 1st with the Unlock 3.0 has contributed to decrease the rate of growth of those daily deaths, thus leading to the mid-September peak. Again, one must be careful in over-interpreting this juxtaposition, because we do not know what would have been the dynamic of daily Covid deaths in India had the lockdown policies be different. 

This figure is suggestive of a failure of the Indian lockdown episodes to affect the lethality of the COVID. Any policy restricting the contact between individuals and the circulation of the virus at time t will see its effect on COVID lethality some 20 to 25 days later. Indeed, it takes between 5 to 7 days to be contaminated by the virus and it takes again 10 days after the beginning of the symptoms for the disease to evolve into the dangerous and potentially lethal phase. Hence, this graph suggests that the first phase of the lockdown has slowed down a bit the daily lethality of the COVID (one can see a little drop in the number of death roughly in the first two weeks of April). However, the progressive alleviation of the lockdown initiated on April 14th and after seem to have generated a continuous upsurge of the daily death that has peaked in mid-September without any clear connection to the severity of the lockdown (which has regularly decreased over the period). Hence, except perhaps at the very beginning of the epidemic, lockdown policies in India seem to have had no visible effect on the dynamic of the daily number of daily deaths. To the contrary, it looks as if the reduction in the severity of the lockdown initiated on August 1st with the Unlock 3.0 has contributed to decrease the rate of growth of those daily deaths, thus leading to the mid-September peak. Again, one must be careful in over-interpreting this juxtaposition, because we do not know what would have been the dynamic of daily Covid deaths in India had the lockdown policies be different. 

Figure4: number of daily Covid deaths in France during the different lockdown phases

Nicolas Gravel, Director of the CSH.

Regarding the CSH scientific production linked to the topic:

You may also like...