Three Definitions, One Choice: Defining Gender Apartheid in the Draft Crimes Against Humanity Treaty

[Lisa Davis is a professor of law at CUNY Law School, and the ICC special adviser on Gender and Other Discriminatory Crimes.

Kirby Anwar is a Visiting Associate Professor at Human Rights and Gender Justice Clinic, CUNY School of Law, senior legal advisor at MADRE and member of the Gender Persecution in Afghanistan Accountability Working Group.]

As governments move to negotiate a long-awaited treaty on Crimes Against Humanity (CAH), a consequential problem sits at the core of the treaty’s criminal provisions: the draft retains a definition of apartheid that belongs to a bygone era—one that risks excluding many who most need protection, including Palestinians, Afghan women and girls, and LGBTQI+ persons.

It currently defines the crime of apartheid as “inhumane acts … committed in the context of an institutionalized regime of systematic oppression and domination by one racial group over any other racial group or groups and committed with the intention of maintaining that regime.” The definition is from the International Criminal Court’s Rome Statute, which in the late 1990’s modestly refined that in the 1973 Apartheid Convention. 

At first glance, this definition may seem fine. It was based in part, however, on false assumptions: that “racial groups” are biologically fixed and identifiable through immutable characteristics. Both the travaux prépartoires for criminal law treaties and international criminal law jurisprudence around the concept of protected “groups” make plain this flaw. 

International legal instruments such as the Genocide Convention, the Apartheid Convention, and the Rome Statute all carry mid-century baggage. They reflect language that treats race (and racial groups) as something one is born into, not something socially constructed. Courts applying this language have sometimes reinforced these harmful assumptions, which now risk being embedded into the CAH treaty itself.

This will have profound consequences. By relying on outdated concepts, the treaty risks rendering invisible some victims who have been oppressed through racist, patriarchal, political, colonial, or ideological structures. Advocates and experts have long recognized the conditions faced by Palestinians, for example, as a form of apartheid. Yet Palestinians and others enduring systems of oppression and domination, have suffered legal invisibility for far too long. A proposed definition of gender apartheid built on the same antiquated assumptions risks similar exclusions, particularly for LGBTQI+ persons (to learn more about this risk of exclusion see our webinar).

Decades of research findings—and developments in human rights law—reject biological or hereditary bases for categories like race. The UN Committee on the Elimination of All Forms of Racial Discrimination has repeatedly affirmed that race is socially constructed, and produced through systems of power, hierarchy and exclusion. Yet international criminal law has been slow to catch up. The CAH treaty presents a rare and consequential opportunity to expand protections at precisely the moment when systems of oppression and domination are growing more complex and entrenched.

Compounding the risks of exclusion, the draft treaty lacks an interpretive provision comparable to Article 21 of the Rome Statute, which instructs the International Criminal Court (ICC) to read its text in harmony with international human rights law. Absent a similar provision, courts may limit accountability in the face of the principle of legality, which requires them to strictly construe the language of criminal provisions. Without explicit guidance to incorporate contemporary human rights jurisprudence, courts default to narrow interpretations of a treaty’s plain text. As a result, they base decisions on outdated assumptions embedded in travaux préparatoires and reinforced through jurisprudence—leaving many victims outside the scope of protection.

As negotiations intensify, several proposals have emerged to integrate gender apartheid into the treaty. These proposals are grounded in the lived experiences of Afghan women, girls, and LGBTQI+ persons targeted under the Taliban regime. One proposed definition would ensure inclusive recognition of all those victims, while others could entrench the very exclusions their advocates seek to correct. They would do so by limiting gender to a rigid binary that erases LGBTQI+ victims, or by structuring apartheid in ways that obscure how patriarchal and racist oppression operate together. 

Intersectionality is not an academic embellishment—it is historical fact. Apartheid in South Africa and Namibia was enforced through mutually reinforcing racist and patriarchal oppression and domination; yet Black women and LGBTQI+ Southern Africans received little formal recognition as victims on that basis. Repeating that erasure now would be a consequential mistake.

The differences between the proposals are not semantic; they will shape the future of accountability. The definition adopted into the treaty will determine who the law recognizes, whether intersectional oppression can be prosecuted, and whether apartheid remains a coherent crime in international criminal law or fractures into disconnected parts.

Civil society has influenced the draft treaty before, and it can do so again. It’s time for the treaty to change once again to ensure the recognition of all victims of apartheid, including Palestinians, women, girls, LGBTQI+ persons and those targeted on intersecting grounds.

Three Proposed Definitions for Codifying Gender Apartheid

Below, we examine three proposed definitions to codify gender apartheid in the draft treaty and explain why only one aligns with the doctrinal, political, and moral imperatives of our time. We also describe why the others risk reproducing outdated biological frameworks, fragmenting the concept of apartheid, and leaving vulnerable populations outside the scope of legal protection.

First Proposal: A Separate Crime of Gender Apartheid

The first proposal would create a standalone crime of “gender apartheid,” mirroring the Rome Statute’s apartheid provision but replacing “racial groups” with “gender groups.” This new offense would sit alongside the existing proposed definition of apartheid. The two provisions would read:

“The crime of gender apartheid means inhumane acts … committed in the context of an institutionalized regime of systematic oppression and domination by one gender group over any other gender group or groups and committed with the intention of maintaining that regime.”

“The crime of apartheid means inhumane acts … committed in the context of an institutionalized regime of systematic oppression and domination by one racial group over any other racial group or groups and committed with the intention of maintaining that regime.”

At first glance, this seems like progress. Elevating gender apartheid to the same conceptual level as the crime of apartheid appears to give long-ignored harms—especially those experienced by women, girls, and LGBTQI+ people—the same gravity and visibility as others under international criminal law. Yet this approach contains two fundamental flaws that risk replicating exclusions that have plagued the international criminal law definition of apartheid for decades.

A. Re-inscribing Outdated and Exclusionary Notions of “Group” Identity

By adopting the existing apartheid definition structure, this proposed gender apartheid definition imports the conceptual weaknesses that have hampered recognition of racial group membership in international criminal law jurisprudence. If added, treaty drafters risk “gender groups” being interpreted as fixed, binary categories of “male” and “female”, much like the mid-20th-century assumption that “racial groups” were biological or hereditary. 

Gender, however, is not biologically fixed. It is socially produced, enforced, and contested through cultural norms, state power, and institutional practices. Transgender, non-binary, and other gender-nonconforming and intersex persons are targeted precisely because they do not fit within the rigid categories that perpetrators impose. A legal framework that only recognizes domination between binary gender groups is structurally incapable of capturing domination against those who fall outside them.

This weakness stems from defects in the understanding of protected “groups” under international criminal law. Shaped by theories that underpinned negotiations around the Genocide Convention, international criminal law has long required some, often undisclosed, evidence of biological or hereditary traits to establish membership in, for example, a “racial group.” While there is no finalized apartheid jurisprudence yet, as our law review article explains, the travaux préparatoires for the Apartheid Convention and Rome Statute indicate drafters took the same approach to racial groups under apartheid.

Thus, this proposed gender apartheid definition risks importing the conceptual errors that have historically narrowed the reach of apartheid—and that continue to leave whole populations unprotected.

B. Creating a Separate Crime Fractures the Legal and Symbolic Unity of Apartheid

Apartheid is not always expressed as two different systems—racist or gendered—operating independently. It can, and has, operated as a unified political project grounded in domination, enforced through mutually reinforcing hierarchies. For example, South African and Nambian apartheid regimes were sustained through simultaneous racist and patriarchal oppression. Black women experienced racist segregation, forced removals, land dispossession, and political and economic exclusion that were inseparable from and compounded by gendered subordination via housing, education and work restrictions, reproductive control, state-sanctioned sexual violence, and other pervasive gender violence carried out with impunity. Black women and LGBTQI+ persons endured yet another layer of gender policing and punishment of sexuality and gender expression. Yet the South African Truth and Reconciliation Commission largely overlooked the intersectional harms they endured. 

Under a separate proposed definition, however, these overlapping harms may not be charged as one system of apartheid. Instead, prosecutors would have to artificially disentangle racist from gendered domination and oppression, even when they are fused in practice. 

Prosecutors may have to charge two offenses—(racial) apartheid and gender apartheid—to capture intersectional harms, doubling the work and creating a patchwork of overlapping charges rather than a unified account of victims’ lived experiences. Two separate crimes also create two separate bodies of jurisprudence, potentially with different thresholds for what qualifies as “systematic oppression,” “domination,” or “intent.” Such bifurcation would not only generate doctrinal inconsistencies across cases, it would also risk producing conflicting evidentiary standards. This would weaken the coherence and enforceability of apartheid as a unified international crime.

Worse still, victims that suffer simultaneous racist and gendered oppression may be forced into one box or the other. This would not merely be inefficient; it would represent an epistemic failure.

Second Proposal: Expanding the Existing Apartheid Definition

Another proposal attempts to avoid these problems by retaining apartheid as a single, unified crime, while expanding the definition to include gender oppression and domination. Under this approach, the revised definition reads:

“The crime of apartheid means inhumane acts committed in the context of an institutionalized regime of systematic oppression and domination by one racial group over any other racial group or groups, or by one gender group over another gender group or groups, and committed with the intention of maintaining that regime.”

This formulation improves upon the first proposal in important ways. It maintains the conceptual unity of apartheid: a single crime, encompassing multiple modalities of domination. This may avoid creating parallel and potentially divergent bodies of jurisprudence and preserve the symbolic coherence of apartheid as a structural system of oppression. Yet two major problems remain.

First, the definition preserves the same exclusionary “group” language. By inserting “gender group” into the existing formula, this proposal repeats the conceptual flaws of the current apartheid definition—namely, the outdated notion that race and gender are fixed, biologically identifiable groups. 

Victims whose subordination is rooted not in biology but in racist beliefs embedded in political, legal, and ideological structures—victims such as Palestinians—may again fall outside the treaty’s scope. So too may transgender, non-binary, and intersex persons, whose persecution often stems from the policing of gender boundaries rather than domination by one “gender group” over another (see discussion above).

This formulation therefore risks protecting only those whose oppression neatly maps onto binary categories—reproducing the exclusions gender apartheid codification is meant to address.

Secondly, the definition also forecloses recognition of apartheid based on intersecting forms of discrimination. Although the single-crime model is conceptually stronger, its disjunctive structure—“racial” apartheid or “gender” apartheid—forces prosecutors to categorize systematic domination into one box or the other. This simplifies and distorts how apartheid can operate.

There are real-world apartheid systems that have relied on overlapping racial and gender hierarchies. As discussed above, under the apartheid regimes in Southern Africa, Black women and LGBTQI+ persons experienced racist oppression and domination alongside and compounded by gendered oppression and domination. These are not separate harms that can be neatly divided into racial or gender apartheid—they are mutually reinforcing manifestations of a single architecture of domination.

A definition that pushes prosecutors to disentangle these inseparable harms undermines the ability of the law to reflect victims’ lived experiences. It fragments identity, fractures accountability, and erases the complex ways in which systems of power operate. It also erases the historical record, painting a partial picture of discrimination during atrocities. Historical erasure invisibilizes victims and can perpetuate cycles of violence. 

Third Proposed Definition: Aligning Apartheid with Persecution-Based Group Protections

A third formulation offers a stronger approach:

“The crime of apartheid means [conduct] committed in the context of an institutionalized regime of systematic discrimination, oppression and domination by one group over another group or groups, based on race or gender, and committed with the intention of maintaining that regime.”

Or:

“…on race or gender grounds…”

This version draws on the proposal by the UN Working Group on Discrimination Against Women and Girls, which calls for the CAH treaty’s apartheid definition to “incorporate the human rights component of discrimination.” This definition is the most coherent, inclusive, and doctrinally grounded of the three. It aligns apartheid with the structure of the crime of persecution which can be committed “against any identifiable group,” on enumerated grounds. 

Referring simply to “groups,” rather than “racial groups” or “gender groups,” avoids resurrecting mid-century notions of race and gender as biological categories. Courts would not need to search for hereditary or physical features to determine who qualifies as a victim. Adding the words “based on” (or alternatively, “on the grounds of”) reinforces a perpetrator-based view of group identity, aligning with how persecution crimes are interpreted by international criminal tribunals and how discrimination is understood under international human rights law. It recognizes that both race and gender are socially constructed categories instead of rooted in fixed or biological criteria. This shift solves the exclusion problem and harmonizes apartheid doctrine with contemporary human rights jurisprudence.

As with persecution, this formulation also captures intersectional discrimination. Because this definition is grounded in the perpetrator’s intentions and imposed structures it captures situations where racial and gender oppression operate together, just as intersecting grounds for discrimination are recognized under persecution jurisprudence. This reflects apartheid as it has actually existed. It would capture, for example, the experiences of Black South African and Namibian women and LBTQI+ persons who lived under racist and gendered subordination. 

Under this definition, these intertwined harms can be charged as a single crime of apartheid. Currently, this is the only proposal capable of capturing the full historical and contemporary complexity of apartheid’s architecture. It offers treaty drafters a path toward an inclusive and contemporary understanding of apartheid that better reflects lived realities.

Why an Inclusive Definition Matters 

International criminal law’s core definitions rarely change. Once codified, they become difficult to revise. That is why the draft CAH treaty negotiations are an important moment for advocacy.

Civil society has successfully intervened in the draft treaty before. In 2018, a global campaign led by Lisa Davis and Jessica Stern working with MADRE, OutRight International, the Human Rights and Gender Justice (HRGJ) Clinic at CUNY Law School, and others—challenged the outdated definition of “gender” under gender persecution in the draft treaty. As with all forms of persecution, victim recognition under gender persecution can be determined by perpetrators’ intent, instead of hereditary characteristics, but the definition drew in part on outdated notions of gender as a binary. In 2019, the International Law Commission responded by removing the definition entirely, affirming that gender is socially constructed (see how we won here). It is a key example of how civil society can modernize international criminal law architecture.

We now face a similar inflection point. The apartheid provision in the draft treaty imports outdated concepts that will shape jurisprudence for decades. What is at stake is whether the law will recognize the full spectrum of systematic domination and oppression as it exists—or whether it will remain anchored to discredited assumptions of the past.

For Palestinians, Afghan women and girls, LGBTQI+ persons, and countless others experiencing layered forms of subordination, this is not an academic debate. It defines the line between visibility and erasure.

States now face a defining choice: preserve an outdated definition that fragments and excludes, or adopt a single, coherent, intersectional, modern definition of apartheid. The CAH treaty may be a once-in-a-generation opportunity to modernize the understanding of crimes against humanity under international criminal law. Whether states seize this moment will hinge on civil society’s efforts to demand justice and recognition for all victims.