Introducing the case studies

Mass atrocities – widespread and systematic violence against civilians – have proven easier to condemn than to stop. Yet these atrocities do eventually stop. Valuable lessons for halting future atrocities may be learned by studying past terminations.

This website provides case studies of endings (alpha listing of cases) assembled in this of “very massive atrocities,” defined as widespread and systematic killing of unarmed people (civilians or prisoners of war) within a single country resulting in at least 50,000 civilian fatalities in cases post-1945. This includes cases in which people were under the direct control of perpetrators, held in camps and prisons, and denied the means for sustaining life. We use a threshold of 5,000 civilian deaths in a given year, with onsets marked by the first year above this level, and endings marked as the final year at this threshold under a single perpetrator, when followed by two consecutive years below this level. If more than 5,000 civilians are killed by a separate perpetrator group within the subsequent two years, this is coded as a subsequent atrocity.

The universe of cases was established by drawing on diverse existing lists and datasets of atrocity and genocides plus additional research. Case studies for each atrocity episode in the dataset drew on expert qualitative analysis to outline the context for instability, describe the scale and pattern of atrocities, present the evidence base for fatality figures, and detail terminations. This process is inherently subjective and not all experts will agree on all conclusions. We consulted outside experts to help with these decisions and obtained in-depth reviews on difficult cases including the Africa cases, Iraq, Indonesia and Vietnam.

Unlike some datasets on mass atrocities, we include examples from anti-colonial conflicts, where the primary perpetrators are either state or non-state actors, and both war time and peacetime atrocities.

Throughout, we sought the most credible minimal number of fatalities.

We also offer a framework that organizes these endings into three types. First, atrocities end as planned when perpetrators achieve their aims, often by eliminating the threat they believed was posed by the targeted group. Second, atrocities can end when perpetrators were prevented from achieving their aims, notably through military defeat. The third ending type, labeled strategic shift, occurs when perpetrators retain power and capacity, and continue to pursue their initial goals, but through means that no longer rely upon mass killing of civilians. Such a shift may be a response to rising (military, political, or economic) costs of continuing the atrocities, failure to achieve the intended goals, or shifts in political power and preferences within the perpetrating organization.

Each case study is divided into five sections. These include: an introduction; atrocities, discussing the patterns of violence across the atrocities period; fatalities, a discussion of the research surrounding the numbers; endings, describing how the violence ended, in line with our criteria; and coding, explaining how and why we coded the ending. We further provide a list of works cited. (All variables in the dataset are detailed here).

We welcome your feedback, and ask that you include a real name and email address along with any comments or questions.

Posted in Introduction | 1 Comment

Dataset Variables

Variables in the dataset are:

id: A unique identifier for each atrocity, using the country in which it occurred concatenated with the start year.

crisis: A shorthand indicator of the crisis during which the atrocity occurred, such as a “civil war”, “war of independence”, “partition”, or other common name frequently used to refer to the context of the event by scholars familiar with the country’s history.

• cause.asplanned, cause.shift, cause.defeat. Three binary indicators (mutually exclusive) labeling the type of ending we assigned to each case. cause.asplanned equals one if the atrocity was determined to have ended as planned, and zero otherwise. cause.shift equals one if the atrocity was determined to end through strategic shift. cause.defeat equals one if the atrocity was determined to end through defeat of the primary perpetrator.

• secondary. This variable offers a secondary coding in the nine cases where we contemplate that other analysts might have come to a difference conclusion. It takes the value “P” for endings as planned, the value “S” for endings by strategic shift, and the value “D” for endings by defeat of the primary perpetrator.

• secondary.note. A text note indicating the logic by which the secondary coding could be chosen.

• atrocitybegin, atrocityend. The years in which the atrocity is coded to begin and to end.

• conflictbegin, conflictend. Dates of the associated multi-sided armed conflict (if present), as coded by the Uppsala Conflict Data Program (UCDP/PRIO Armed Conflict Dataset version 19.1, Pettersson, H¨ogbladh & Oberg, 2019). ¨ 2

• primaryperp. A name for the primary perpetrator in the atrocity.

fatalities. Estimated number of civilian fatalities.

• normalize. Whether (0/1) the perpetrator remained in power and normalized relations with the target group(s) thereafter (if the regime was the primary perpetrator).

• leaderchange. If same regime remains intact at the ending, but a key leader is shifted out of power, whether through elite coup, natural death or democratic change.

• defeatdom, defeatint. Whether (0/1) the primary perpetrator was militarily defeated at the hands of domestic (defeatdom) and/or international (defeatint) forces.

• shiftdom, shiftint. Wheter (0/1) a strategic shift was substantially influenced by domestic (shiftom) and/or by international (shiftint) actors.

withdrawalint. Whether (0/1) a withdrawal of forces occurred during the atrocities and that was associated with a reduction in atrocities.

popularviolence. Whether (0/1) the violence involved popular participation or was conducted only by members of specialized organized groups;

• initiatornotworst. Whether (0/1) atrocities occurred at a higher level under a perpetrator who was not the considered to have been the first.

• nsa.primary, nsa.secondary. Whether (0/1) the primary perpetrator was a non-state
actor (NSA); whether the secondary perpetrator (if there was one) was a non-state actor.

subsequent. Whether (0/1) a more than 5,000 civilians killed by a new perpetrator, or
the same perpetrator targeting a new and distinct group, within two years of ending the
primary atrocity event.

Posted in Introduction | Leave a comment

Cambodia under the Khmer Rouge

Introduction | Atrocities | Fatalities | Ending | Coding | Works Cited | Notes


The US bombing of Cambodia came to a halt in August of 1973 when the US Congress legislated its conclusion, following the signing of a peace agreement between the US and North Vietnamese. The Khmer Rouge and Lon Nol armies continued to fight for two more years until 1975 when Phnom Penh fell to the Khmer Rouge.

Atrocities, 1975 – 1979

Once the Khmer Rouge (KR) took over the country, they began to implement radical plans to restructure society. Among the influences and populations targeted for elimination were: Buddhism; “foreign” influences (including targeting the urban, educated class); entire minority groups (the Chinese, Vietnamese, Muslim Cham, and Thai[i]); as well as those Khmer who were deemed unfit as determined by an ever-widening criteria. The KR goal was to create an agrarian communist state by controlling family relations, restructuring agriculture, re-organizing the political, legal, and military institutions of the state, and collectivizing economic production and consumption—in short, to establish complete control over society.

The Khmer population was divided into “old citizens” – those who lived in KR zones before their 1975 victory and “new citizens” those who fell under their control thereafter[ii] As Alex Hinton writes, “in the new revolutionary society, each person had to be reworked, like hot iron, in the flames of the revolution.”[iii] However, despite the regime’s emphasis on targeting foreign elements in society, by the end of the conflict the majority of killing had been perpetrated by Khmer citizens against other Khmer citizens.[iv] The lack of popular base for the KR and upheaval wrought by their policies required extensive use of lethal force to maintain control.[v] Kiernan notes that the “most horrific slaughter was perpetrated in the last six months of the regime” in areas bordering Vietnam.[vi]


Our research suggests a range of 650,000 – 1.4 million violent deaths between 1970-1979.

Many attempts have been made to count or estimate the scale of deaths under the KR.[vii] While the KR officials claim that only around 20,000 civilians were killed, the true estimate likely falls somewhere between 1-3 million total deaths, with upper range estimates of those directly killed by the regime approaching 1 million.

Demographic analysis has been undertaken by many parties including the US Central Intelligence Agency[viii] and a variety of historians.[ix] By estimating population numbers before and after the genocide and taking into account expected fertility rates, migration rates, and normal expected mortality rates, these demographic analyses attempt to calculate the excess mortality rate. One of the more thorough demographic studies, conducted by Patrick Heuveline, also attempts to separate out violent civilian deaths from a general increase in mortality caused by famine, disease, working conditions, or other indirect causes. He does so by grouping deaths within different age and sex brackets and analyzing treatment of these age and sex groups by the Khmer Rouge and violent regimes in general. His conclusion is that an average of 2.52 million people (range of 1.17-3.42 million) died as a result of regime actions between 1970 and 1979, with an average estimate of 1.4 million (range of 1.09-2.16 million) directly violent deaths.[x]

Researchers have also attempted to estimate deaths attributable to the regime by interviewing segments of the Cambodian population and extrapolating this information to the wider population. Through this method, Cambodia scholar Ben Kiernan originally estimated that 1.5 million deaths resulted from the regime’s policies (his figure was later scaled up to 1,671,000)[xi]. This figure amounts to a loss of 21% of the population[xii]–approximately 15% of the rural population and 25% of the urban population[xiii]–but includes both people who were violently killed and those that died of starvation, overwork, and disease.[xiv] As Ben Kiernan notes in his book, Milton Osborne valued the percentage of overall deaths that occurred as a result of execution to be 31 percent.[xv] If this percentage is applied to Kiernan’s most recent estimate of overall deaths, then the number of violent deaths by execution is 539,032.

A much larger estimate of overall deaths was put forward by the People’s Republic of Kampuchea (PRK), the Vietnamese backed government that took control of Cambodia after Vietnam defeated the Khmer Rouge. In the early 1980s, the PRK attempted to quantify deaths by interviewing Cambodians at the household or village level throughout the country. They estimated that 3,314,768 people lost their lives under the KR.[xvi]
However, this figure has been criticized as an over-estimate resulting from the PRK’s methods, which did not take into account the fact that families may have been geographically dispersed during the period and also involved adding reported deaths to estimates from exhumed mass graves.

The Documentation Center of Cambodia (DC-Cam) with Craig Etcheson undertook a mass grave survey to estimate deaths by the KR. Although this method of estimation is complicated by the difficulty of determining cause and time of death, in 1999, DC-Cam estimated that their ongoing investigations had uncovered 20,492 mass graves containing the remains of roughly 1.1 million victims of KR government execution. Although larger than some other estimates, Etcheson argues that these numbers are plausible, given the nature of the mass grave and DC-Cam’s methods, which are more likely to produce an under-count of bodies rather than an over-estimate.[xvii]


The KR initiated several attacks against Vietnam, despite its comparative weaknesses. Several skirmishes occurred along the Cambodian-Vietnamese border beginning in 1977, escalating to large-scale battles by early 1978.[xviii] Purges of KR leadership also caused defectors to flee to Vietnam and request military assistance. In December of 1978, Vietnam broadcast that a Kampuchean Front for National Liberation had been established in “liberated Cambodian territory”. This announcement was followed by a second invasion on Christmas day of 1978.[xix] In January of 1979, Vietnamese forces took control of the capital; Pol Pot and the remaining KR leadership fled to the Southwest of the country.[xx] The Vietnamese capture of Phnom Penh and establishment of the People’s Republic of Kampuchea marked the end of mass atrocities in Cambodia.

International support for the KR continued long thereafter, both in the refugee camps in Thailand and in international diplomatic arenas. The KR, which separated itself from Vietnamese and Soviet influence, received backing from China, the U.S. and most western countries even after the end of the Cold War. The KR, for instance, held Cambodia’s seat at the UN until 1993, long after it had been defeated by Vietnam.

Armed groups composed of the remnants of the KR together with the FUNCINPEC and KPNLF forged a coalition in 1982 to continue to challenge the Vietnamese-backed state. The 1991 Paris Peace Accords, co-chaired by the Ministers of Foreign Affairs of France and Indonesia, four Cambodian parties, and the Secretary General of the United Nations, in addition to many national representatives, culminated in a final agreement signed by 19 governments. The agreement further established the United Nations Transitional Authority in Cambodia to implement the agreements, an ambitious mission with military, civil administration, human rights, policing, repatriation, rehabilitation, and electoral components. Elections were held in May 1993, and a new constitution was promulgated on 24 September 1993.


Atrocities ended through an international defeat of the primary perpetrators when the Vietnamese invasion overthrew the Khmer Rouge.

Works Cited

Banister, Judith and Paige Johnson. 1993.”After the Nightmare: The Population of Cambodia,” in Genocide and Democracy in Cambodia: The Khmer Rouge, the United Nations and the International Community, ed. Ben Kiernan. New Haven, CT: Yale University Southeast Asia Studies.

Chandler, David. 2008. A History of Cambodia, 4thed. Boulder, CO: Westview Press, 2008.

Etcheson, Craig. 1999. “‘The Number’: Quantifying Crimes Against Humanity in
Cambodia.” Mass Graves Study, Documentation Center of Cambodia.

Gottesman, Evan. 2003. After The Khmer Rouge: Inside the Politics of
Nation Building.
New Haven: Yale University Press.

Heuveline, Patrick. 1998. “’Between one ad three million’: Towards the demographic reconstruction of a decade of Cambodian history (1970 – 1979).” Population Studies 52: 49–65.

Hinton, Alexandar Laban. 2009. “Truth, Representation and the Politics of Memory after Genocide” in People of Virtue: Reconfiguring Religion, Power and Moral Order in Cambodia Today. ed. Alexandra Kent & David Chandler. Copenhagen: NIAS Press.

Hinton, Alexandar Laban. 2005. Why did they Kill?: Cambodia in the Shadow of Genocide. Berkeley and Los Angeles: University of California Press.

Kiernan, Ben. 1985. How Pol Pot Came to Power : A History of Communism in Kampuchea, 1930 – 1975. London: Verso.

Kiernan, Ben. 2008. The Pol Pot Regime: Race, Power, and Genocide in Cambodia under the Khmer Rouge, 1975-79, 2nd ed. New Haven, CT: Yale University Press.

Kiernan,Ben. 2009. “The Cambodian Genocide” in Century of Genocide, ed. Samuel Totten and William Parsons. Third Edition. New York: Routledge, 340– 373.

“Report of the Research Committee on Pol Pot’s Genocidal Regime.” 1983. Phnom Penh, Cambodia, July 25.

Sharp, Bruce. “Counting Hell” Available at: May 26, 2015.

Sliwinski, Marek. 1995. Le Génocide Khmer Rouge: un analyse démographique. Paris: Editions L’Harmattan.

U.S. Central Intelligence Agency. 1980. “Kampuchea: A Demographic Catastrophe.” Washington, DC, January 17.

Vickery, Michael. 1984. Cambodia 1975 – 1982. Boston, MA: South End Press.

[i] Kiernan 2009, 346 – 347.

[ii] Kiernan 2009, 349.

[iii] Hinton 2009, 63. 

[iv] Hinton 2005, 15.

[v] Kiernan 2009, 346.

[vi] Kiernan 2009, 349.

[vii] Sharp 2015.

[viii] U.S. Central Intelligence Agency, 1980.

[ix] Sliwinski 1995; Banister and Johnson, 1993; Vickery 1984.

[x] Heuveline 1998.

[xi] Kiernan 2008, 458.

[xii] Kiernan 1996, 351.

[xiii] Kiernan 2009, 349.

[xiv] Kiernan 1996, 351.

[xv] Kiernan 1996, 456.

[xvi] “Report of the Research Committee on Pol Pot’s Genocidal Regime” 1983.

[xvii] Etcheson 1999.

[xviii] Etcheson 1999, 191.

[xix] Gottesman 2003, 10 – 11.

[xx] Chandler 2008.

Posted in Asia, Defeat | Tagged , , | Leave a comment