Geographical Variations in the Propagation of COVID-19

By Jean-Philippe Platteau

One puzzling question that arises in connection with the spread of the virus SARS-CoV-2 is why there are such large variations in its incidence (the infection rate) and its lethal consequences (death toll) across countries and across regions within countries. At this stage, at least, nobody is able to bring forward a general hypothesis that would satisfactorily answer such a thorny question even if attention is focused on a limited geographical area, such as Europe, Asia, North or South America, Africa or the Middle East. The difficulty comes from the fact that there are many confounding factors at play, and they involve variations in geographical situations, biological determinants, public health policies, economic circumstances, and social or cultural characteristics. In this short essay, attention will, therefore, be focused on Europe and my specific contribution, will more particularly, consist of highlighting the role of some neglected factors, socio-cultural factors in particular, without pretending that they play a dominant role, let alone an exclusive role.

A striking contrast offered by the map of COVID-19 in Europe is between the relatively low rates of infections (and hospitalizations) in countries like Germany, Austria, Scandinavia (with the exception of Sweden), and Eastern Europe, on the one hand, and the relatively high rates observed in countries like Italy, Spain, France, the United Kingdom, Belgium and the Netherlands, on the other. Equally striking are the intra-country variations found in some countries, as attested by the examples of Italy, France and Switzerland. As is well-known, Lombardy in northern Italy has been the hotspot of the country’s epidemic with the city of Bergamo at its centre. In the central and southern parts, including the big cities of Roma and Napoli, the penetration of the SARS-CoV-2 has been much better contained. In France, while the virus has hit the Paris region, Bourgogne and Eastern France with particular vigour, the western and southern parts of the country (and Brittany) have been largely spared. In Switzerland, the French-speaking part, Romandy, has epidemiological statistics close to France whereas its German-speaking, Alemannic part evinces strong similarity with Germany and Austria, and its Italian-speaking part, the Tessin, strong similarity with northern Italy.

Variations in public health policies
In addressing inter-country variations, a lot of attention has been paid to explanations that privilege public health facilities and policies, including the capacity of the government to plan and anticipate and its ability to act decisively at the right moment. In this regard, South Korea, Singapore and Taiwan have been presented as models to emulate because, following early warning signals, they were able to intervene early in the transmission chain of the virus. Concretely, this means that they tested people belonging to a hotspot of infection and isolated them as soon as they were detected positive. By thus breaking or slowing down the propagation process at a critical stage, they could avoid to lock down their people with all the adverse economic and social consequences that this implies. The availability of an adequate testing and contact-tracing capacity was a major factor of success, and there is no doubt that in building it, the governments of Southeast Asia had learned from their experience of the first epidemic of SARS (2003), now called SARS-1, which eventually did not propagate outside Asia. And it is not only a matter of good public governance but also of self-discipline on the part of the citizens themselves; the latter, too, had learned from the SARS-1 experience and quickly adjusted their behaviour when the threat of SARS-CoV-2 emerged.

Returning to Europe, an essential fact is that it did not go through the SARS-1 painful experience and could not, therefore, draw useful lessons from it for the future. Moreover, it did not take seriously the outburst of the SARS-CoV-2 in Hubei province (China) since the belief was that, like SARS-1, it would remain confined to Asia. Hence, the general state of preparation of European public health systems when the virus penetrated into Europe through Italy was poor. Worse, even the Italian plight did not cause a serious alert in many European countries because Italy has a reputation of ill-organized and poor public health that cannot be generalized to other (West-) European countries. A number of countries, however, reacted faster than others and had available a better testing capacity from the very start of the epidemic. Foremost among these were Germany and Austria which have been considered ex-post as the European best emulators of South Korea, Taiwan, and Singapore in terms of public health systems and policies.

Much less publicized are the sound approaches of Eastern European (and Baltic) countries in which the propagation of the virus has been relatively well controlled, with few hospitalizations and deaths. In addition to having younger populations and lower population densities than the worst-affected countries (Italy, Spain, France, the United Kingdom, Belgium, and the Netherlands), Eastern Europe benefitted from effective governments which took tough measures rather early in order to counter the first wave of the epidemic. This is probably due to the fact that they took the events in Italy more seriously than the comparatively self-confident countries of Western Europe. Thus, for example, the Czech Republic and Slovakia quickly imposed the wearing of the face mask outdoors. Also, the lockdown was announced earlier than elsewhere. Whereas public meetings and events were still authorized in the United Kingdom and the Netherlands in the second and third weeks of March, a partial or complete lockdown was already in force in most Eastern European countries.

In Romania, the government responded to COVID-19 with a harsh lockdown, declaring a state of emergency even before the country’s first official death. This implied, among other things, that written declarations of purpose were required to leave home. Croatia required a government-issued pass to travel between towns whereas in Poland (among the first European countries to shut its borders), children under 13 were barred from leaving home without an adult, shoppers were compelled to wear disposable gloves and face masks were required in public. Serbia, Hungary and Bulgaria were also quite fast in enacting harsh lockdown policies, contrasting with the United Kingdom, the Netherlands, and above all Sweden, which never imposed a lockdown and preferred to opt for a policy of ‘freedom under responsibility’ (Logo Business AM, 2020, May 6; Economist, 2020, May 2: 16-7).

The particularly mild policy of Sweden allows for an interesting experiment since neighbouring Denmark, which shares many similarities with Sweden, followed the way of a rather harsh lockdown. The comparison is in favour of Denmark where infections and deaths caused by the virus are significantly lower than in Sweden – to this date, there are about 2,600 cases/1million people in Sweden compared to 1,800/1M in Denmark, while the death rate is 320/1M in the former and only 90/1M in the latter. If Sweden is compared to Norway, the comparison is still more disadvantageous for Sweden since the rate of infection in Norway is hardly 1,500/1M, and the death ratio is as low as 40/1M (statistics extracted from Worldometer as of May 7, 2020).

Genetic variations
Recently, microbiologists from the University of Ghent in Belgium have argued that a part of the differences in the intensity of the epidemic is attributable to genetic variations. More precisely, some population groups carry a gene (ACE1) that facilitates the fixation of the SARS-CoV-2 while other groups exhibit a higher frequency of the polymorphism D of the same gene, which makes them more resistant against this virus (Delanghe et al., 2020). Interestingly, the more one moves toward the eastern parts of Europe, the higher the incidence of this favourable variant of the ACE1 gene, and not only Eastern European countries but also Austria-Germany, Scandinavia, and southern Italy (where the Norman conquest left its biological imprint) are included in the zone where the polymorphism is found. Spain, Northern Italy, France, Belgium, the Netherlands, and the United Kingdom are not.

Although it is hard to say, at present, how far this genetic variation goes toward explaining the aforementioned inter- and intra-country differences in the propagation of the epidemic, it cannot be ignored and one wishes that the supporting research will soon be extended to other countries beyond Western and Eastern Europe, including Russia and other continents.

Cultural variations
In addition to genetic variations and variations in public health policies and capacities, cultural differences can also potentially account for the observed contrasts in the incidence of COVID-19. It has, thus, been noticed that contact habits and attitudes may differ significantly between some countries. Thus, the Japanese habit of keeping reasonable distances between people while interacting, strikingly contrasts with the Western European habit, especially in southern Europe, of kissing and hugging friends, relatives, and acquaintances. Moreover, in some countries like South Korea, China, and Japan again, people are accustomed to wearing face masks as a way to protect themselves against air pollution, an attitude which is an oddity in Europe. It is evident that these East Asian cultural habits are a big advantage under a virus attack when precisely these attitudes are conducive to effective protection against contamination.

There is yet another important sense in which cultural variations matter and they relate to the frequencies of contacts between people. For example, Italian society is strongly centred on the family with the consequence that relatives frequently pay visits to each other. In particular, children and grandchildren often visit their grandparents and not doing so (when feasible) is considered a grave breach of duty or social norm violation. In Scandinavia, by contrast, interpersonal contacts are not only more distant but also less frequent. It happens that so-called contact matrices have been estimated for a large number of countries, and they display such contact frequencies both within and between different age classes. A simple glance at these matrices shows, for instance, that the density of contacts is comparatively high for Italy and much lower for Germany. As for Belgium, it occupies an intermediate position.

In a recent paper, with Vincenzo Verardi, I have carried out an exercise that proceeds in two steps: first, using a standard epidemiological model (the SEIR model) calibrated on Belgian data, we simulate the effects of different lockdown exit strategies on the evolution of the epidemic, once its peak has passed; and, second, we repeat the same simulations after having replaced the Belgium-specific contact matrix by that of Germany and then, by that of Italy. The results are striking. If only Belgians could have the contact habits of the Germans, they would be able to (gradually) re-open their economy by resorting to milder measures than what they actually need now. And, conversely, if they had the contact habits of the Italians, they would need to use much harsher measures. In more concrete terms, with their own culture, the Belgians must implement rather ambitious testing (and contact-tracing) schemes and enforce rather strict social-distancing measures if they want to avoid a rebound of the epidemic while re-opening their economy. If they had German cultural habits, they could avoid a rebound by implementing only moderate measures of testing and social distancing. But if they had Italian cultural habits, they would have no choice other than implementing very severe public health policies and proceeding much more slowly in re-opening the economy.

Our discussion suggests that a country like Germany is probably cumulating all the advantages that work toward a successful lockdown exit: (1) it possesses a strong public health infrastructure and has chosen sound public health policies that prepared the ground for an effective battle against COVID-19; (2) its people probably evince genetic characteristics that make them less vulnerable to the virus; and (3) the social norms that guide individual behaviour, including the habits regarding meetings and visits, help slow down an epidemic. If in ordinary circumstances, the comparatively weak role of the family in Germany is not necessarily an advantage and the family-based model of Italy may seem preferable, the situation is modified in conditions of a raging epidemic when such a model is suddenly transformed from an asset into a liability.

Contact frequencies may also provide a (partial) explanation for the aforementioned variations in Switzerland where in terms of rates of infection and deaths, the French-speaking part is close to France, the German-speaking part is close to Germany and Austria, and the Italian-speaking part is close to northern Italy. It is true, on the other hand, that the important variations, such as those observed between northern and southern France, are not accounted for, attesting that there is no unique explanation for all the geographical differences observed.  However, there is a key lesson to draw from our work and from the foregoing discussion – there is no one-size-fits-all solution that could be uniformly applied to all countries and even to all regions inside a given country. It is perhaps not coincidental that the European Union has been unable or unwilling to suggest, let alone prescribe, a common lockdown exit strategy for all its members, leaving them free to make their own decisions in the matter. The diversity of peoples and cultures inside Europe is too large to allow for a general solution to the complex problems raised by the present pandemic. The same conclusion also applies to large federal political entities, India, Russia, and the United States, for example.

AUTHOR
Jean-Philippe Platteau, Emeritus Professor of Economics, University of Namur, Belgium

REFERENCES

Delanghe, J.R., M.M. Speeckaert, and M.L. De Buyzere, 2020. “The host’s angiotensin-converting enzyme polymorphism may explain epidemiological findings in COVID-19 infections”, Clinica Chimica Acta, June, 505: 192-193.
Economist, 2020. Infectious Doubt, 2-8 May, pp. 16-17.
Logo Business, 2020. Pourquoi les pays de l’Europe de l’Est sont-ils beaucoup moins impactés par le coronavirus?, 6 May.
Platteau, J.P., and V. Verardi, 2020. How to Exit Covid-19 Lockdowns: Social Structure Matters, Centre for Research in Economic Development (CRED), University of Namur, Mimeo.

Disclaimer: The views and opinions expressed in this article are those of the author/s and do not necessarily reflect the official policy or position of Azim Premji University or Foundation.