This chart illustrates what the global mortality distribution of COVID-19 fatalities could look like if 40% of global COVID-19 deaths were unreported under different scenarios of underreporting for high-income and developing countries. The scenarios are anchored to the WHO’s finding that for all-cause mortality roughly 40% of global deaths are unregistered (see here for a comprehensive assessment of data quality on a country basis). Assuming that a similar order of magnitude is likely for COVID-19, the 40% estimate is applied to the analysis of mortality during the current pandemic. 

The current distribution in mortality across high-income and developing countries is shown in the blue bar on the left of each income group. The second bar illustrates that the distribution will not be affected if the same degree of underreporting were to apply to both income groups (with the average of the two leading to a 40% global average). The third bar illustrates the mortality distribution if the degree of underreporting in developing countries is double the degree observed in high-income countries (while yielding again the 40% global average). The fourth bar makes the (unrealistic) assumption that there is no underreporting in high-income countries, only in developing countries, subject again to the condition that we obtain a 40% global average.

The latter scenario would raise the developing country share in the global death toll to levels that are similar to the counterfactual demography scenario that isolates the effect of demography (population size and age structure) keeping all other factors constant. The estimates in this chart therefore suggest that data quality likely plays an important role in explaining the difference between the current distribution of mortality and the counterfactual one indicated by demography.


Pandem-ic uses the World Bank income classification as a major building block in the analysis of the impact of the pandemic.


The income classification groups countries in four buckets by per capita income levels: high-income countries (HICs), upper-middle-income countries (UMICs), lower-middle-income countries (LMICs) and low-income countries (LICs). We use the current FY2023 classification, which determines the thresholds of the buckets as follows:  


  • LICs are defined as those with a GNI per capita, calculated using the World Bank Atlas method, of $1,085 or less in 2021;
  • LMICs are those with a GNI per capita between $1,086 and $4,255; 
  • UMICS are those with a GNI per capita between $4,256 and $13,205;
  • HICs are those with a GNI per capita greater than $13,205.


A good part of this site also analyzes the pandemic by region (where we use the World Bank regional classification and the UN geo-scheme of subregions). In both cases (i.e. across income groups and regions), the universe of countries is based on the World Bank income classification. More on that in the next note.

The universe of countries on this website is determined as follows.


  • We start with the FY2023 World Bank income classification, which comprises of World Bank member countries as well as other economies with populations over 30,000 people (see World Development Indicators database). This means that we exclude the Holy See, Cook Islands and Niue since they are are not included in the WB income classification.  
  • We then narrow down this list to 196 countries/economies by retaining only the member states of the UN, one non-member state with observer status (the State of Palestine listed as West Bank and Gaza in the income classification) and two members/observers in UN Specialized Agencies  (Kosovo and Chinese Taipei listed as Taiwan, China in the income classification). This leads to the exclusion of a further 22 territories and dependencies (21 HIC and 1 UMIC) from the World Bank income classification. 


Note that the vaccination data is pulled from Our World in Data, which utilizes a slightly different universe of locations. In sticking with the above 196 countries and economies, we have made the following adjustments relative to the OWID universe.

  • Given their UN membership status, we extract the following UN members from US totals and list them separately:
    • Federated States of Micronesia;
    • Marshall Islands;
    • Republic of Palau.
  • Conversely, given that they are not identified as separate members of the UN or UN specialized agencies, we do not separately mention the following entities but instead include their data into the totals of the country they are a territory or dependency of:
    • Hong Kong SAR and Macao SAR are added to China totals;
    • Faroe Islands and Greenland are added to Denmark totals;
    • Aruba and Curaçao are added to the totals of The Netherlands;
    • Data for Anguilla, Bermuda, Cayman Islands, Falkland Islands, Gibraltar, Guernsey, Isle of Man, Jersey, Montserrat, Saint Helena and Turks and Caicos are add to the totals of the United Kingdom.


For each of the above adjustments to the vaccination data, we make adjustments to the demographic data that vaccine information is related to (including population size, age structure and priority group size).


Finally, note that no adjustments are required to the totals for France as its overseas territories and dependencies are already included.

Excess mortality can be defined as the gap between the total number of deaths that occur for any reason and the amount that would be expected under normal circumstances.  Given the massive undercounting of the mortality toll both directly and indirectly attributed to COVID-19, excess mortality provides a useful way to get a glimpse of the true mortality toll.


Unfortunately, however, data on excess mortality are not universally available. Only 84 countries release some sort of data (national or subnational; regular or one-off) on excess deaths.  This is where the excess deaths model of The Economist comes in as it tries to fill the gaps on the basis of a well-calibrated model that takes advantage of various types of data that CAN be observed.


At its core, the model relies on a machine-learning algorithm (a gradient booster) that learns from official excess-mortality data and over 100 other statistical indicators. Where data on excess deaths is available, they are used.  Where such data are not available, the model fills the gaps in the form of single-point estimates.


Given the vast degree of uncertainty surrounding any point estimate, the model then uses a bootstrapping method to calculate standard errors. This amounts to using subsets of the full dataset (in terms of country-week pairs) and training different gradient-boosting models on each of these data subsets. The central estimate is derived then from the trained model on the full set of data, whereas the middle 95 of the predications generated by the 100 other models produce the 95% confidence intervals. 


Further details on the methodology can be found here and the full model can be viewed here.