The particular horror evoked by genocide exists for three main reasons. Genocide is murder at a mass scale, it is usually perpetrated against the defenseless and, although tensions may build up over decades or even centuries, it can flare up very rapidly.

In the Rwandan case, 800 000 people were killed in just 100 days[1]. A related phenomenon is politicide. In genocides victimized groups are defined in communal terms, whereas in politicides they are defined in terms of their political opposition to the regime and dominant groups[2]. These are ideal types: in some cases, it may be that both motives for murder are present.

The Appendix reports 49 genocides and politicides in 30 countries between 1955 and 2014.

Because genocide is so terrible, efforts have been devoted to assess the risk of its occurrence. Three influential studies will be reviewed in this brief. But first we need a guide to the phenomenon itself.

Stanton’s ten stages of genocide

Stanton’s ten stages of genocide[3] are well known in the field, and they form the basis for the work of the Genocide Watch project. Stanton states that stages may occur simultaneously, and that later stages require the presence of the earlier stages. The stages are:

  1. Classification, particularly where the distinction between ‘us’ and ‘them’ is insufficiently countered by tolerance, understanding and a common national identity.
  2. Symbolization, the assignation of names or other symbols to classification.
  3. Discrimination, where a dominant group uses law, custom and political power to deny the rights of other groups.
  4. Dehumanization, where one group denies the humanity of another.
  5. Organization, usually by the state, often using militias to provide deniability of state responsibility.
  6. Polarization, with extremists driving groups apart, hate groups broadcasting polarizing propaganda, and, possibly, laws forbidding intermarriage or social interaction.
  7. Preparation, the planning of a ‘final solution’ and the building of armies, the purchase of weapons and the training of militias.
  8. Persecution, with identification and separating out of victims, often with property expropriation and sometimes placement in ghettoes, camps, or confinement to a particular region. Massacres begin.
  9. Extermination which quickly becomes the mass killing legally called ‘genocide’.
  10. Denial, which always follows a genocide, and often is in an indicator of further genocide to come.

A risk factor is associated with the stages.

  1. Classification: Deeply divided societies.
  2. Discrimination: State-led targeting of groups.
  3. Dehumanization: An exclusionary ideology
  4. Organization: Autocratic government
  5. Polarization: An ethnically divided elite
  6. Preparation: An economically closed system and resistance to influences outside country borders.
  7. Persecution: Massive violations of human rights
  8. Denial: Impunity after genocides

Stanton makes the point that his model was designed to place the risk factor in Barbara Harff’s analysis of country risks of genocide and politicide into a processual structure. Harff distinguished between ideological genocide, associated with outcomes of elite succession struggles, and retributive genocides which are forged during and in the immediate aftermath of civil wars. Accordingly, she identified six risk factors:

  1. Political upheaval, an abrupt change in system caused by state or regime formation through violent conflict, redrawing of state boundaries or defeat in an international war.
  2. Prior genocides, with elites and security forces habituated to mass killing. Nearly half of the countries listed in the Appendix had more than one genocidal/politicidal episode with 60 years.
  3. Autocratic rule and exclusionary ideologies.
  4. Ethnic and religious cleavages
  5. Low economic development. Twelve out of the 30 countries listed In the Appendix were in sub-Saharan Africa, the least developed region in the world
  6. International isolation. Leaders of isolated states are more likely to calculate that they can eliminate unwanted groups without international repercussions.

The Sentinel Project

The Sentinal Project has considered a number of models and distilled from them thirty factors divided into four categories[4].

The categories are:

1. Political

Institutional: Quality of democracy, constraints on state security agencies, isolation from the international community, level of military expenditure, frequent changes in political leadership.

Regime and ideology: Orientation towards force and coercion to seize and maintain power, commitment to a harmful ideology (exclusionary, antagonistic, legitimation of victimization of outgroup, revolutionary, dominance by one group), charismatic leadership that generates mass followership, degree of freedom of speech, installation of a newly-created regime, ruling group deems outgroup to be dangerous, severe government discrimination or active repression against communal groups, ethnic minority rule.

2. Economic

Economic status of the regime, sudden and severe economic hardship, long term difficult life conditions, socio-economic deprivation combined with group based inequality

3. Sociocultural

Existence of distinctive groups separated by social divisions, cultural devaluation of the outgroup, prior persecution of the outgroup, legacy of intergroup hatred and grievance, population growth and youth bulge, ethnic nationalism, outgroup viewed as an obstacle to economic progress.

4. Conflict and upheaval

Political upheaval, conflicts of status, power and rights, history of conflict, history of genocide, ongoing insurgency of civil war.

The United Nations Framework of Analysis of Atrocity Crimes[5]

The United Nations divides atrocity crimes into three categories: genocide, crimes against humanity, and war crimes. It identified eight common risk factors and two specific risk factors per category. Indicators for each risk factor are specified.

The ten relevant to genocide are:

1. Armed conflict or other instability

Armed conflict, security crisis, humanitarian crisis or emergency, political instability caused by abrupt or irregular regime change, growing nationalist, armed or radical opposition movements, autocratic rule, severe political repression, economic instability caused by dispute about resources, a severe economic crisis, or acute poverty, mass unemployment or deep inequalities, mass protest against state authorities or policies, and tensions based on identity issues

2. Record of serious violations of international human rights and humanitarian law.

Past or current serious violations of human rights and humanitarian law that have not been prevented, punished or adequately addressed, creating a risk of further violations.

3. Weakness of state structures

National legal framework that does not offer ample and effective protection, lack of resources for judiciary, law enforcement and human rights institutions, lack of an independent and impartial judiciary, lack of effective civilian control of security forces, high levels of corruption, poor governance, absent or inadequate mechanisms of oversight and accountability, lack of access to recourse by victims, inadequate training on human rights to military forces, lack of resources for reform or institution building, lack of resources for protecting the population.

4. Motives or incentives.

Political motives, economic interests, strategic or military interests, ethnic cleansing, real or perceived threats against interests of perpetrators, real or perceived membership or support for armed opposition groups, supremacist ideologies, politicization of past grievances, social trauma caused by past incidents of violence not adequately addressed.

5. Capacity to commit atrocity crimes.

Availability of personnel, arms, ammunition and financial resources, capacity to transport and deploy personnel and other resources, capacity to recruit large numbers of supporters, strong culture of obedience to authority and group conformity, links with armed forces, support by business and the wealthy, support by external actors.

6. Absence of mitigating factors.

Absence of factors which could contribute to preventing or lessening the impact of serious acts of violence

7. Enabling circumstances or preparatory action

State of emergency, interference with vital state institutions, strengthening and use of the security apparatus, acquisition of large quantities of arms and ammunition, creation of militias and paramilitary groups, strict control of communication channels, restrictions on NGOs, international organizations and media, serious acts of violence against women and children, imposition of life threatening living conditions or deportations, destruction of plundering of essential goods and services, marking people on the basis of identity, increase politicization of identity, increased inflammatory rhetoric, propaganda campaigns and hate speech.

8. Triggering factors

Sudden deployment of armed forces, spillovers of conflicts in neighbouring countries, measures perceived as threatening state sovereignty, abrupt changes in the political power of groups, attacks against life, physical integrity, liberty or security of leaders, destabilization of elections, exploitation of resources which have a serious impact on the livelihoods of civilian populations, inflammatory commemoration events.

9. Intergroup tensions or patterns of discrimination

Past or present conduct that creates stress in the relationship among groups or with the state.

10. Signs of intent to destroy a group as a whole or in part

Official document, political manifestos, media records or other documentation containing incitement, targeted physical elimination, widespread and systematic discriminatory practices, threats to women and children, actions leading to humiliation, fear and terror, public expressions of euphoria at having control over a group, attacks against homes, farms, businesses or other livelihoods, and against cultural and religious symbols.


There is much to think about in these schemata, but there is also considerable overlap between them. The onset of political or economic crisis increases the risk of genocide. So do clear divisions between groups and high inequality between them, creating mistrust and a tendency to escalate tensions. Autocratic rule with adverse implications for recourse, the closing down of communication channels and restriction of human rights facilitate genocide. So does inflammatory, and sometimes coded, rhetoric. Accumulation of arms and ammunition and the formation of militias indicate a high risk of imminent genocide, as do deportations and restrictions on where people may live.

How should we think about the risk of genocide in South Africa?

There has been much loose talk about the possibility of genocide in South Africa. A more measured response, in line with international understanding, suggests the following observations:

  1. We have a legacy of socio-cultural division, a high level of inequality, and inadequate protection of human rights before 1994. Many thought that these factors would tear South Africa apart, but the advent of a negotiated political transition and the development of a new constitution demonstrated the society could manage them. The risk to this outlook is created by members of this or that group taking the view that the constitution sold out their interests and is therefore not worthy of support. To the extent that this view gains traction, the risk increases.
  2. South Africa is not currently in an economic crisis, but the probability of one is rising. Real per capita GDP has fallen since 2014 and the prospect over the medium term is at best one of very slow growth. Unemployment, long high, has been increasing as has the public debt to GDP ratio.
  3. South Africa is not currently a failed state, though the Fund for Peace’s fragile state index in 2019 place it in the ‘elevates warning’ category. Key parts of the public service and state owned enterprises are dysfunctional, and it has become increasingly clear that there has been a criminal conspiracy (or, more precisely, criminal conspiracies) at the heart of the state, not yet completely eradicated.
  4. The legal status of inflammatory speech is not clear. Aspects of the Promotion of Equality and Prevention of Unfair Discrimination Act (PEPUDA) are likely to be unconstitutional and a case is heading for the Constitutional Court to clarify the matter. A Prevention and Combating Hate Crimes and Hare Speech Bill has been published but not enacted.
  5. It is decidedly not helpful to advance two lines of argument appearing in a recent SA Human Rights Commission Report[6]. The first introduced the concept of differing vulnerability between groups, and associated this vulnerability with socio-economic status, regarding Indians as less vulnerable than Africans and Whites as a socio-economically powerful group. This view is historically and sociologically uninformed and pernicious in its effect. To take an example: by the early 1930s, German Jews were well represented in both business and the professions and, to that extent they were a ‘powerful socio-economic group’. Did that protect them from isolation, demonization and ultimate annihilation? It did not. The implication of the SAHRC’s approach is that some people can, at the very least, steer closer to the wind than others. The second line of argument is to distinguish between literal and figurative forms of speech. But what use is that? To return to the German example, German Jews were isolated by both literal and figurative speech and actions. The difference was unimportant. What mattered in both cases was the desire to isolate and stigmatize.

Is South Africa at imminent risk of genocide? No, even though Genocide Watch has placed us at Level 6 on the Stanton scale. Has the risk risen in recent years? Unfortunately yes, and the only way to come to terms with it is to focus more clearly on what is happening now and confront the challenges we face .

Charles Simkins
Head of Research

Appendix - Genocides and politicides, 1955-2014

Sub-Saharan Africa






Central African Republic

Democratic Republic of Congo


Equatorial Guinea

























Sri Lanka







Middle East and North Africa

















El Salvador











Source: Benjamin Goldsmith and Charles Butcher

Genocide Forecasting: Past Accuracy and New Forecasts

Atrocity Forecasting Project Working Paper, 2017

[1]BBC News, 4 April 2019

[2]Barbara Harff, No lessons learned from the Holocaust? Assessing risks of genocide and political mass murder since 1995, American Political Science Review, 97(1), 2003

[3] His document can be accessed at

[4]The Sentinel Project, Risk Factors at

[5]United Nations, Framework of analysis of atrocity crimes: a tool for prevention, 2014

[6] SA Human Rights Commission, Findings of the South African Human Rights Commission regarding certain statements made by Mr Julius Malema and another member of the Economic Freedom Fighters, March 2019