Political Violence

1. Recorded leader deaths by decade and region

Central / South AsiaE. Europe / former USSRLatin America / CaribbeanMiddle East / N. AfricaNorth AmericaNortheast AsiaOceaniaSoutheast AsiaSub-Saharan AfricaWestern Europe

The decade-by-region chart preserves the page’s main historical comparison while making its counting boundaries explicit. The supplied series contains 136 recorded deaths from the 1900s through an incomplete 2020s category. The largest decade is the 1970s, with twenty entries, followed by the 1980s with sixteen. Colours distinguish the ten regional groups and remain consistent in the next chart, allowing the composition of each column to be followed across time.

These are counts from the existing compilation, whose underlying case list and final observation date were not supplied. They should not be described as a complete current census. A decade with fewer entries may reflect real historical differences, incomplete coverage or decisions about who qualifies as a national leader. Countries also differ in the number of people and years exposed to risk. The chart therefore cannot, by itself, demonstrate that protective security improved or that political institutions became safer. The incomplete final decade is particularly unsuitable for direct comparison with ten-year periods.

Source: existing page chart data, retrieved 20 October 2026. Underlying case list and final observation date not supplied; retained as an incomplete, undated compilation, not a complete current census. 136 entries. Decade bins begin in 1900; 2020s incomplete.

2. Regional shares of the decade compilation

Central / South Asia — 15E. Europe / former USSR — 15Latin America / Caribbean — 14Middle East / N. Africa — 35North America — 2Northeast Asia — 6Oceania — 1Southeast Asia — 6Sub-Saharan Africa — 25Western Europe — 17

The Middle East and North Africa accounts for 35 of the 136 entries in the decade series, or about 25.7%. Sub-Saharan Africa follows with 25, while Western Europe contributes seventeen. The pie uses the same regional colours as the preceding stacked chart and calculates every slice from that chart’s underlying rows. This keeps the two views internally consistent rather than carrying forward conflicting totals.

The earlier standalone pie contained 26 entries for Sub-Saharan Africa, one more than its decade series. The rebuilt pie follows the detailed series and identifies the discrepancy rather than assigning an unexplained extra death to a decade. Regional shares are descriptions of this compilation, not estimates of personal risk. They are affected by historical boundaries, the number of governments covered and the inclusion rules for coups, executions or other violent deaths. A large share cannot establish a single cause across very different events. Individual historical examples can give context, but they should not substitute for a documented case-by-case dataset.

Source: existing page chart data, retrieved 20 October 2026. Underlying case list and final observation date not supplied; retained as an incomplete, undated compilation, not a complete current census. Recalculated from chart 1: Sub-Saharan Africa 25, not the standalone pie’s 26; denominator 136 rather than 137.

3. Recorded failed and fatal attempts since 2000

Failed attemptsFatal outcomes

The supplied post-2000 comparison contains 49 failed attempts and seven fatal outcomes across nine regional groups. Eastern Europe and the former Soviet area has the largest failed-attempt count, with twelve; Sub-Saharan Africa follows with ten. Stacking the two categories makes the recorded total visible while keeping fatal and nonfatal outcomes distinct. Grey-blue represents failed attempts and muted red represents recorded assassinations.

The starting year is specified by the earlier page, but the final date and incident list are not. It is also unclear whether repeated attempts against one person are counted separately in every case. For those reasons, the chart should not be turned into a probability of surviving an attack or used to judge the effectiveness of a country’s security services. Alleged plots, intercepted preparations and attacks that actually reach a target can have very different reporting thresholds. The figures are retained as a labelled compilation, without claiming that they include all attempts or every recent incident worldwide.

Source: existing page chart data, retrieved 20 October 2026. Underlying case list and final observation date not supplied; retained as an incomplete, undated compilation, not a complete current census. Starting year 2000; final cutoff unspecified. 49 failed attempts and seven fatal outcomes. Counts are not survival probabilities.

4. US political assassinations by century and office

PresidentsMembers of CongressFederal judgesGovernorsState legislatorsState judgesOther state officesMayors

The American compilation contains 72 entries across eight categories of public office. Its nineteenth-century group is largest, with 37, followed by the twentieth century with 27. State legislators account for a substantial part of the nineteenth-century column. Showing the office categories separately makes clear that this topic extends beyond the four presidential assassinations most familiar to many readers.

The century labels require care. The twenty-first century is incomplete, while the eighteenth-century entry reflects the scope of the supplied list rather than a full century of comparable United States institutions. The underlying definitions and final update date are unavailable. Counts are therefore retained as recorded entries, with no claim that the shorter recent column demonstrates a falling annual risk. The number of officeholders, their terms of service and the reporting of local cases all affect interpretation. Different offices also have different responsibilities and exposure. This chart describes the composition of a historical list; it does not provide a controlled comparison of how dangerous each office was.

Source: existing page chart data, retrieved 20 October 2026. Underlying case list and final observation date not supplied; retained as an incomplete, undated compilation, not a complete current census. 72 entries. Century lengths and coverage differ; twenty-first-century bin incomplete.

5. Selected locations: domestic and foreign leaders

Domestic leadersForeign leaders

The selected country comparison distinguishes leaders representing the country in which they were killed from foreign leaders killed there. Egypt has six entries in the supplied ranking, split into five domestic and one foreign. Afghanistan and Iran each have five domestic entries. France’s total of three includes two foreign leaders, illustrating why the location of an assassination and the nationality or political office of its target must be recorded separately.

This distinction is easy to lose in a simple country league table. An attack during a visit can add to the host country’s count without involving its own head of state or government. The chart keeps the two categories visible and preserves only the seven countries in the earlier selection. It is not a complete ranking of all countries, and its undated coverage should not be presented as current. The reference table later on the page provides separate columns for the represented country and the location country, allowing readers to see that distinction in named historical cases.

Source: existing page chart data, retrieved 20 October 2026. Underlying case list and final observation date not supplied; retained as an incomplete, undated compilation, not a complete current census. Seven selected locations; domestic/foreign follows the source classification.

6. Total recorded deaths across decade bins

Adding the regional segments produces a simpler view of the decade series. The total rises to twenty entries in the 1970s and then falls to sixteen in the 1980s and eleven in the 1990s. The line emphasises changes between successive bins, while the earlier stacked chart remains available for readers who want to see which regional groups contribute to each point. Every value comes from the same 136-entry compilation.

The line joins decade totals, not annual observations. It should not be read as showing a smooth trajectory between the plotted periods. The final category is an incomplete decade, and the source’s update date is unknown. Low recent totals consequently cannot establish that assassination has nearly disappeared or that modern leaders face a particular level of risk. Changes in the number of countries and officeholders would also matter in any rate calculation. The purpose of this additional view is to make the supplied historical pattern easier to read while retaining the limitations that a visually clean trend line can otherwise conceal.

Source: existing page chart data, retrieved 20 October 2026. Underlying case list and final observation date not supplied; retained as an incomplete, undated compilation, not a complete current census. Same 136 entries as chart 1; points are decade totals, not annual rates. Final decade incomplete.

7. Regional composition across three periods

Central / South AsiaE. Europe / former USSRLatin America / CaribbeanMiddle East / N. AfricaNorth AmericaNortheast AsiaOceaniaSoutheast AsiaSub-Saharan AfricaWestern Europe

Grouping the decade series into three broad periods shows how the regional composition changes within this compilation. The first period covers 1900–1949, the second 1950–1999 and the third the available entries from 2000 onwards. Each bar totals 100%, so its segments describe shares within that period rather than equal numbers of recorded deaths. Europe contributes much more strongly to the early period, while other regions account for a larger part of the later twentieth-century collection.

The periods contain 54, 75 and seven entries respectively. That small final denominator makes each recent entry especially influential: one record represents about 14.3 percentage points. The periods also differ in duration and completeness, so equal-width bars should never be taken to mean equal exposure to risk. Regional groups retain the earlier page’s boundaries, including its combined Eastern Europe and former Soviet category. This chart is a composition comparison within supplied records, not evidence that violence moved neatly from one region to another or that a single historical explanation applies everywhere.

Source: existing page chart data, retrieved 20 October 2026. Underlying case list and final observation date not supplied; retained as an incomplete, undated compilation, not a complete current census. Period denominators: 54, 75 and seven. Periods are unequal in duration and completeness.

8. Offices represented in the US compilation

Presidents — 4Members of Congress — 11Federal judges — 3Governors — 4State legislators — 33State judges — 5Other state offices — 2Mayors — 10

Pooling the American entries by office gives a different perspective from the century chart. State legislators contribute 33 of the 72 recorded entries, the largest category. Members of Congress contribute eleven and mayors ten, while presidents account for four. The doughnut makes the breadth of the list visible, with colours identifying offices rather than political parties or the motives of attackers.

A large count for an office does not mean that an individual officeholder faced the greatest risk. There are many more state legislators and mayors than presidents, and the number serving has changed across history. Calculating comparable risks would require the population of officeholders and their time in office, as well as a complete and consistently defined incident list. None of those denominators is supplied here. The chart instead answers a narrower question: how are the entries in the existing American compilation divided among its eight categories? The presidential portrait below connects one familiar case with that much broader institutional picture.

Source: existing page chart data, retrieved 20 October 2026. Underlying case list and final observation date not supplied; retained as an incomplete, undated compilation, not a complete current census. Same 72 entries as chart 4, pooled across centuries; no officeholder or person-year denominator.

9. Geographical spread and decade totals

The scatter plot compares two features of each decade in the regional series. The horizontal axis counts how many of the ten regions have at least one entry; the vertical axis gives the decade’s total recorded deaths. The 1980s spans seven regions, the widest spread in this compilation, while the 1970s has the largest death total across six regions. These are related but distinct descriptions of the data.

A high total can be concentrated in relatively few regions, and a wider geographical spread does not necessarily produce the highest total. The plot helps separate those ideas without introducing a causal trend line. Points represent decade bins, not individual attacks or countries. Some decades share the same coordinates, so they can overlap visually; the supporting data file retains every point and its label. The incomplete 2020s remains a partial observation. As elsewhere, the chart measures the supplied compilation and cannot establish the true worldwide incidence of political violence or compare population-adjusted risks between regions.

Source: existing page chart data, retrieved 20 October 2026. Underlying case list and final observation date not supplied; retained as an incomplete, undated compilation, not a complete current census. One point per decade; exact overlaps possible. Spread counts nonzero regions among the ten source groups.

10. Offices held in the selected historical reference table

The final chart summarises the offices held by the people in the selected historical reference table below. It groups prime ministers and comparable chief ministers as heads of government, kings and emirs as monarchs, and presidents or equivalent national leaders as presidents and heads of state. Every person contributes once, even when more than one leader died in the same event. The categories describe office types rather than political systems or methods of attack.

This is a separate, explicitly selected collection covering cases from 1900 to 2021. It is not the missing case list behind the earlier charts, and its totals should not be expected to reproduce them. The selection provides named examples with year, office, represented country and location, making the history easier to explore. It excludes failed attempts and does not attempt to catalogue every politically motivated death. The count of one office type depends on which examples were selected, so the bars should not be interpreted as showing which form of government is inherently more vulnerable to assassination.

Source: selected entries from historical leader-death compilation. 61 named cases, 1900–2021. Office grouping is editorial; this table is separate from the earlier aggregate data.

11. Selected assassinated leaders, 1900–2021

61 named examples, arranged by year. Scroll across and down for office, represented country and location. This is a selected historical reference, not a complete list, a list of perpetrators, or the case dataset behind charts 1–9. It includes coups and other assassinations; no claim is made that every case has the same legal classification. Years avoid conflating attack dates with later dates of death. Regional groups here are broad editorial groupings based on location, distinct from the older charts.