di Alberto Ciancio, University of Lausanne and University of Pensylvania

Sub-Saharan Africa (SSA) has been one of the last regions to be hit by the COVID-19 pandemic. Similar to developed countries, the populations of this region experienced a surge in cases and deaths prompting a drastic policy response to restrict social activities. However, there seem to be also important differences in the way the pandemic is unfolding in these countries. With the exception of the relatively rich South Africa, the number of cases is lower than in Europe or America. But the stark difference with the developed world it is in the number of deaths.

No country in SSA has anywhere near the number of deaths by population experienced in Europe or America. In Malawi, there are currently 9 deaths per 1 million people (https://www.worldometers.info/coronavirus/) while Europe average is around 300 deaths per million people and United States 650 deaths per million. The low mortality of COVID in Africa could be at least partially the consequence of limited testing capacity. However, the first studies that compare mortality with previous years are showing no particular signs of extra-mortality due to COVID (Uyoga et al, 2020). It is then possible that individuals in Africa are less vulnerable to the COVID epidemic perhaps because of SSA particular demographic structure. At the same time, the economic consequences of social distancing policies and the drastic decline of economic activity in the entire world can have particularly strong effects on poverty and mortality in SSA, a region which includes the vast majority of the individuals living in poverty around the world. The first reports are showing dramatic scenarios of 100 millions people pushed to extreme poverty because of COVID (World Bank, 2020).

Figure 1: Science post on the pandemic in Africa


There is then a need to bring evidence on how COVID-19 is affecting the continent through reliable micro-level data with surveys that possibly have a longitudinal component and can quantify the effect of the pandemic comparing the world pre and post pandemic. However, conducting a survey in times of COVID-19 in Africa is challenging. No in-contact survey can be ethically approved and internet surveys are not an option in rural areas of SSA.

The Malawi Longitudinal Study of Family and Health (MLSFH) presented a unique opportunity to answer these questions. The MLSFH is a long-standing survey following the same individuals in rural Malawi from 1998. The survey has rich information on health (initially focused on HIV/AIDS but more recently also on non-communicable diseases), family dynamics, economic and social activity. As part of the MLSFH team, lead by professor Hans-Peter Kohler from the University of Pennsylvania, we were preparing to go in the field to collect new data on the long-term consequences of the AIDS epidemic but COVID-19 travel restrictions did not allow in-person fieldwork. We planned instead a cell phone-based data collection and focus the questions on the new epidemic. Importantly and differently from many COVID-19 surveys in developing countries we have twenty years of data about the individuals we surveyed. The target sample was all the MLSFH respondents surveyed during 2017-2019 with recorded phone numbers as well as village chiefs surveyed in 2019. The total target sample was 3,172 individuals of which we managed to interview 2,256 between June 2nd and August 17th 2020.

Figure 2: Subjective probability of becoming infected in the next 6 months with and without social distancing. Source: author's elaboration from MLSFH Covid surveyImmagine2

The boxplot-like graph displays the mean (dot) and median (center line) of the subjective infection probability, as well as the 10th (lower whisker), 25th (bottom of box), 75th (top of box), and 90th (upper whisker) percentiles of the distribution.

Some clear patterns emerge from the survey. First, COVID-19 prevalence seems to be low. In the absence of testing, we can only look at reported symptoms: we see very low rates of fever and coughing (less than 1% of respondents have both symptoms) suggesting low prevalence of Covid-19 in rural Malawi. Second, individuals in rural Malawi are very knowledgeable about COVID-19 symptoms and transmission methods. Government radio campaigns have probably been effective in spreading correct information even though we find evidence of misinformation: 27% of respondents reported receiving fake news while only a small group believed traditional practices (3%) or using herbs could prevent COVID infection. Respondents also have plausible risk perceptions. For example, they understand that the probability of getting infected with social distancing is much lower than without distancing. Third, respondent report very good levels of social distancing and hand washing although only 60% of them use face masks. Finally, respondents are very worried about the economic consequences of the pandemic. The great majority is already reducing non-food consumption and almost half of them believe they will face a shortage of food in the next six months.

More research is needed to understand why the African COVID experience is different. One question to explore is whether exposure to previous epidemics can explain a more resilient response to COVID. It is also important to understand what should be an appropriate policy response tailored to the African context, taking into account the limited state capacity of these countries and their fragility to economic shocks. The benefits of lockdowns in these contexts may be outweighed by its costs not just in terms of economic outcomes but also in terms of health and mortality.


Uyoga et al. (2020). Seroprevalence of anti-SARS-CoV-2 IgG antibodies in Kenyan blood donors. medRxiv.

World Bank. (2020). Updated estimates of the impact of COVID-19 on global poverty: The effect of new data. Retrieved from https://blogs.worldbank.org/opendata/updated-estimates-impact-covid-19-global-poverty-effect-new-data

logo fblogo fb


Our activities


Recent publications on emerging economies:


Scientific articles and book chapters



Slides and videos


Data and documentation