Natural immunization: are economies on a faster recovery path than predicted?

While the development of highly effective COVID-19 vaccines has been extremely quick and certainly is a success story, their global distribution is and remains highly uneven. High-income countries originally absorbed large parts of global supply, except for Chinese and Russian vaccines. Upper- and lower-middle-income countries only gradually gained access, while low-income countries basically remain at the side-lines, see chart below.

Economists widely believe that the uneven progress on vaccination campaigns has far reaching implications. In July, the International Monetary Fund (IMF) wrote in its “World Economic Outlook” update: “Vaccine access has emerged as the principal fault line along which the global recovery splits into two blocs: those that can look forward to a further normalization of activity later this year (almost all advanced economies) and those that will still face resurgent infections and rising Covid death tolls.” Whenever economists write about the near-term outlook of any country, they mostly put forward the pandemic as the key risk that makes the outlook subject to a higher than usual uncertainty, particularly in emerging markets and developing economies (EMDEs) where vaccination campaigns have not gone very far yet, and believe that recoveries will likely be subject to setbacks as occasional flare-ups in infections take a toll on economic activity.

However, the IMF’s fault lines and (most) economists’ risk perceptions miss out on one important factor: natural immunization. Over one and a half years into the pandemic, many people have contracted Covid-19 and gained some natural immunity. While immunity is likely fading over time and might not offer full protection against new variants, the same applies to vaccinations, potentially even more so. A recent study in Israel for instance found that "natural immunity confers longer lasting and stronger protection against infection, symptomatic disease and hospitalization caused by the Delta variant [...] compared to [...] two-dose vaccine-induced immunity."[1]

With regards to the effective number of infections (contrary to the reported number of infections), there is high uncertainty, particularly in countries with lesser institutional capacities, which tend to be lower income countries. Reported figures are not getting anywhere close to effective caseloads, as documented by The Economist. Still, research has been undertaken to estimate the effective number of infections, combining available data (on cases, deaths and testing) with epidemiological knowledge and assumptions. The Economist has produced “extremely rough” estimates of cumulative infection rates, based on “excess deaths” divided by context-adjusted infection-fatality risks. [2] A model developed by Imperial College London (ICL) has a similar approach, though based on confirmed deaths. [3] Given Covid-related deaths are subject to underreporting, this poses a very serious short-coming with regards to many EMDEs and leads to a significant underestimation of effective caseloads. As The Economist notes: “In middle-income and especially poor countries […] excess mortality almost always exceeds the official death toll, often by large margins.” [4] In what follows we present estimates by The Economist, which look plausible, particularly also for EMDEs on which this article focuses.

With regards to the most pressing question of this article, The Economist’s message is clear: in many EMDEs, large parts of populations were likely exposed to the virus in the past and have hence gained some (natural) immunity. Rather than point estimates, findings are presented in ranges, and those ranges widen as income levels decrease, meaning there is higher uncertainty about effective cumulative infections in poorer countries, probably due to more acute data limitations. Looking at mid-points of range estimates, the median of cumulative infections is at 64% of the population in EMDEs, compared to 14% in Advanced Economies (AE), see chart below. Differentiating EMDEs by income level does not provide much further insight, mid-points being at broadly similar, elevated levels. The regional picture though is interesting: Developing Asia is the region with the lowest cumulative infections (mid-point at 33%). Cumulative infections and hence natural immunity are quite high in Eastern Europe, the Caucasus and Central Asia (EECCA), the Middle East and Northern Africa (MENA) as well as Latin America (LatAm), where even the lower bound of range estimates points to roughly half of the populations having contracted Covid-19 in the past. For Sub-Saharan Africa (SSA), the mid-point suggests a relatively elevated level of natural immunity too (62%), but the range is extremely wide (0% - 120%). As the virus has spread around the globe, cumulative infections seem likely a function of institutional capacities and the possibility to follow social distancing and basic hygiene measures, which typically are not strongholds of poorer countries. As such, it is doubtful that SSA fared markedly better than other regions, and the assumption of rather elevated natural immunity should also hold there.

Natural immunization hence appears too big of a factor to be ignored, and a factor that could imply that ongoing recoveries in many EMDEs are less prone to setbacks than widely thought. Pandemic-related vulnerabilities are mainly concentrated in countries with low vaccination rates and – importantly – low natural immunization. Today, only few countries enter that category, mostly (South) East Asian ones including Thailand and Vietnam, where the spread of the Delta variant in recent months proved harder to control than earlier outbreaks, see chart below. In these countries, the threat of surging infections remains imminent until vaccination campaigns are well advanced. That said, in response to recent struggles, affected countries have ramped up efforts to vaccinate large parts of their populations in due course. In contrast, there is a large sample of countries, where vaccination campaigns have not yet advanced very far, but where natural immunization is supposedly quite elevated. In this category, there are in particular countries in the Caucasus (Armenia, Georgia), Central America (Guatemala, Honduras, Nicaragua) and many countries in Africa, not least Côte d’Ivoire, Kenya and South Africa.

The pandemic certainly remains an important threat for the time being, new variants, possibly waning immunity and vaccine hesitancy being among the key concerns. That said, it might well be that many analysts overestimate the likeliness of further setbacks along the road, particularly in countries were the virus has run its course, by ignoring natural immunization.

[1] See Gazit et al. (2021): Comparing SARS-CoV-2 natural immunity to vaccine-induced immunity: reinfections versus breakthrough infections.

[2] The Economist’s model looks at “excess deaths”, i.e., the gap between how many people “normally” die in a given region and year, and how many have effectively died since the pandemic erupted. From there, the share of people who have been infected is calculated, by dividing a country’s total excess deaths by a context-adjusted infection-fatality risk. Further details can be found here.

[3] ICL estimatesthe number of infections based on confirmed deaths and age-specific infection fatality rates. The authors explicitly say that they do not make any adjustment for under-reporting of deaths. Further details can be found here.

[4] The Economist: “How we estimate the true death toll of the pandemic”, 13 May 2021