212.69
excess deaths per 100,000 people, 2020–2021
Moderate trust, 68 of 100 Only WHO counts, so it cannot be checked against another source. The other 2 sources shown do not count; "Where the number comes from" says why.
Show the data as a table
| Year | WHO global excess mortality estimates 2020-2021 (May 2023 release) |
|---|---|
| 2020 | 110.01 |
| 2021 | 102.69 |
All countriesAll of Kyrgyzstan's figuresCompare
Kyrgyzstan among 237 countries
73 have a higher figure and 163 a lower one. Definitions differ between countries, and their years run from 2021 to 2021.
What this number counts
- Counts
- Excess deaths as each source publishes them: deaths in 2020 and 2021 minus the deaths that source expected without the pandemic, after any deaths the source removes for other events (WHO removes other mortality shocks, and each line says what it removes), as a two-year total per 100,000 people, not a yearly rate.
- Where
- Kyrgyzstan
- When
- 2020-2021: 1 January 2020 to 31 December 2021, one total for the two years (2020 and 2021 combined, not one year)
Where the number comes from (3 sources)
| Source | Figure and year | Definition | Quality | Counts for |
|---|---|---|---|---|
WHO global excess mortality estimates 2020-2021 (May 2023 release)World Health Organization (WHO) · Model-based estimateRetrieved 2026-11-02 · Data (zip file) | 212.692021 | WHO mean estimate for 2020-2021 in the country workbook: 14,348 excess deaths (95% interval 13,775 to 14,919), which is 204.2 to 221.16 per 100,000.WHO publishes counts, so the rate uses UN WPP 2024 population on 1 January 2021 (6,745,818), the midpoint of the period. WHO's methods document describes removing deaths from other mortality shocks (conflicts and natural disasters) from its excess figures. The chart points are the 2020 and 2021 rows, each per calendar year on the same population; the headline is their two-year total. Raw file: WHO_COVID_Excess_Deaths_EstimatesByCountry.xlsx inside 2023-05-19_covid-19_gem.zip, sheet "Deaths by year"; population from WPP2024_Demographic_Indicators_Medium.csv.gz. | 5 of 5 checks
| 100% |
World Mortality Dataset excess deaths, January 2020 to December 2021 (Karlinsky and Kobak)World Mortality Dataset (Karlinsky and Kobak) · Death and health registersRetrieved 2026-11-02 · Data (CSV file) | 222.72021 | Published "Excess per 100k": 14,532 excess deaths (standard deviation 1,133) from monthly registered deaths, compared with the authors' baseline, a linear trend with seasonal terms fitted to 2015-2019.Rate on the authors' own population base. Shown, not counted: WHO names this dataset among its repositories, so WHO's line and this one share deaths data and differ mainly in the baseline. Raw file: excess-mortality-jan2020-dec2021.csv; period checked against world_mortality.csv as of 9 March 2023. | 5 of 5 checks
| Not counted |
The Economist excess deaths modelThe Economist · Model-based estimateRetrieved 2026-11-02 · Data (CSV file) | 210.292021 | Central cumulative estimate per 100,000 in the row labelled 2021-12-27.The export script labels weekly rows by their first day, so this is a weekly approximation of the period, not an exact cut at 31 December 2021. Rate on the file's own population column (6,630,621). The file gives no range here because the figure rests on recorded excess deaths. Shown, not counted: The Economist says it takes most of its all-cause deaths from the World Mortality Dataset, which has a line here, so the two are not separate collections. Raw file: export_country_per_100k_cumulative.csv (last changed 18 June 2024), column cumulative_estimated_daily_excess_deaths_per_100k. | 4 of 5 checks
| Not counted |
- WHO100%
Shown but not counted: WMD (WHO names this dataset among its repositories, so WHO's line and this one share deaths data and differ mainly in the baseline); Economist (The Economist says it takes most of its all-cause deaths from the World Mortality Dataset, which has a line here, so the two are not separate collections).
How the trust score was worked out
- Quality40 / 40
How well the sources are made
- Agreement15 / 30
Unknown with one source, so set to half
- Independence6 / 20
How many separate ways the number was measured
- Freshness7 / 10
How recent the newest data is
Who reports it, and since when
- Single source: nothing to check it against
figure no figure here, inside its range no later figure recorded change
Why the numbers differ
- Each source subtracts a different expected number. WHO fits a yearly linear trend and a seasonal curve, usually to 2015-2019 where a country has monthly history and usually to 2000-2019 where it has only yearly totals. (source, retrieved 2026-11-02)
- The World Mortality Dataset fits a linear trend across years and a separate term for each week or month, to 2015-2019. (source, retrieved 2026-11-02)
- The Economist's own baselines fit a linear trend for year and a term for each week or month, on data up to February 2020. (source, retrieved 2026-11-02)
- Many countries have no all-cause death data for the period. In the WHO workbook 88 of 194 Member States are modelled for all 24 months and 95 rest on reported deaths for all 24; The Economist models all or part of the period for 127 of 236 places. The World Mortality Dataset models no deaths and so has 114 countries here. (source, retrieved 2026-11-02)
- Sources treat incomplete registers and other events differently. WHO adjusts reported deaths for completeness, and its methods document describes removing deaths from other mortality shocks (conflicts and natural disasters), so its figures are not unrestricted all-cause excess either. (source, retrieved 2026-11-02)
- The World Mortality Dataset uses registered deaths and its published excess is not always deaths minus expected: Armenia 15,882 against 19,882; Azerbaijan 32,796 against 36,796; Belgium 18,902 against 20,428; Germany 87,403 against 91,660; France 78,463 against 80,008; Luxembourg 12 against 43; Netherlands 27,983 against 28,646. The authors subtract 4,000 deaths in each of Armenia and Azerbaijan for the 2020 war and exclude weeks 32-34 of 2020 in the five European countries. Those lines get half weight where they count. Its paper says coverage varies by country and gives Peru as an example of an estimate that may be too low. (source, retrieved 2026-11-02)
- Excess deaths and reported COVID-19 deaths measure different things. Excess deaths count deaths from every cause, apart from any deaths a source removes for other events. WHO says some COVID-19 deaths may not have been certified as such, that countries certify differently, and it treats excess deaths as direct effects plus knock-on effects, minus deaths averted. (source, retrieved 2026-11-02)
- The population bases differ. WHO lines and reported deaths here use UN WPP 2024 on 1 January 2021; the World Mortality Dataset uses its own base (mostly UN WPP 2022 for 2020, with national figures for Russia and Ukraine); The Economist uses the population column in its file. (source, retrieved 2026-11-02)
- The cut-off dates are close but not identical. WHO runs by calendar month to December 2021; weekly World Mortality Dataset countries run to week 52 of 2021; The Economist's export labels weekly rows by their first day, so its row labelled 2021-12-27 is a weekly approximation of the period and gets half weight where it counts. (source, retrieved 2026-11-02)
- These are not fully separate collections. The Economist says it takes most of its all-cause deaths from the World Mortality Dataset, so wherever that dataset has a line the Economist line is shown but not counted (113 places). It counts only where the dataset has no data. (source, retrieved 2026-11-02)
- WHO also names the World Mortality Dataset among its repositories, so where both give a figure the two share deaths data and differ mainly in the baseline. There WHO's line counts and the dataset's is shown for comparison. (source, retrieved 2026-11-02)
Common mistakes
- Reading a table figure as a yearly rate. Each table figure is a total for two years, so it can't be set beside a one-year death rate. Only the WHO chart points are per calendar year.
- Calling excess deaths "COVID deaths". They are all deaths above an expected number, whatever the cause. WHO describes both direct and indirect effects and deaths averted. (source, retrieved 2026-11-02)
- Adding reported COVID-19 deaths to excess deaths. Reported deaths are part of all deaths, so the two overlap.
- Treating a negative figure as a measured fall. It means fewer deaths than expected, and WHO says it can also be an artefact of the model's assumptions. (source, retrieved 2026-11-02)
- Quoting a central estimate without its range. The Economist says the 95% range should be reported when ranges are not very narrow, and each line here gives the range its source publishes. (source, retrieved 2026-11-02)
Download, cite and report an error
From the table How many more people died than expected in 2020 and 2021 in each country? Updated 2026-11-02.
Download every value in this table (CSV)
Cite as: RealStats, “Excess deaths per 100,000 people, 2020 and 2021 combined in Kyrgyzstan”, updated 2026-11-02, with the original sources listed on the page.