When a Word Becomes a Sentence: The Political Economy Of Witchcraft Accusations

Lariba built her house herself. Brick by brick, room by room, across years of labor in the fields of northern Ghana. She raised a five-room house from the red laterite earth. The house stood as proof that a woman, alone, could accumulate wealth. When the jealousy started, it arrived in the language of the supernatural. Neighbors whispered. Then the whispers became a verdict. Two days of beatings with sticks, and Lariba was cast from everything she had made and sent, with nothing, to the Gambaga witch camp, where she remains. Her house was taken. The word "witch" had done its work.

Across four camps in Ghana's Northern and North East Regions—Gambaga, Gnani, Kpatinga, Kukuo—more than 500 people currently live in a condition that philosopher Achille Mbembe would recognize as a death-world: alive but stripped of home, property, political status, and the right to return. Most are between 50 and 90 years old. They arrived via a mechanism that, on its surface, looks like belief but, in practice, functions like law. A misfortune occurs. A soothsayer is consulted. A chicken is slaughtered and cast to the ground; if it lands on its back, the woman has witchcraft. The judgment is without appeal. The sentence is permanent. Across 18 months of field research ending in January 2025, Amnesty International interviewed 93 people in these camps, 82 of them women, and documented a reality that should force a confrontation with the assumptions embedded in the word "traditional." There is nothing pre-modern about what is happening. This is capitalism's oldest move, wearing a different mask.

To understand why women are expelled from their communities for crimes they did not commit, you have to stop looking at the accusation and start looking at what the accusation takes. In case after case, the Amnesty researchers found the same architecture beneath the spiritual language. Fushiena was accused by her brother after his son died, despite having carried the boy to the hospital herself. Yamina's co-spouse accused her after their children's academic success provoked jealousy. One resident of Gambaga camp was beaten with a bicycle chain by her accuser's brother; when she reported the assault to the police, they declined to act. Awa, widowed, was accused by her husband's brothers within months of his death. Salamatu states it without ambiguity: her father-in-law wanted the cows, the land, and the money her husband had left, and when she refused, the community convened to find a reason to expel her. The reason was witchcraft. The effect was expropriation. Adisa Mahama, a resident of Kukuo camp, summarized the logic with unsparing precision:

"They always have plans of putting allegations against you, especially if you are hardworking and are still strong and doing well as a woman."

More than 70 percent of women in the surveyed camps were accused and banished after their husbands died. 

The feminist political philosopher Silvia Federici spent decades documenting how this mechanism works in history. Caliban and the Witch, published in 2004, argued that the European witch trials of the 15th through 17th centuries were integral to the transition from feudalism to capitalism—a systematic terror campaign designed to break women's autonomy, enclose communal lands, and enforce a new sexual division of labor in which women's reproductive work became invisible and uncompensated. What makes Federici's argument genuinely disturbing is its epilogue: primitive accumulation, she insisted, is an ongoing process, endlessly repeated wherever capital meets resistance and wherever land or labor can be stripped from those too vulnerable to defend themselves. 

In her 2010 essay on witch-hunting and enclosures in Africa, she follows this logic into the present. The concentrated witch-hunts in Ghana, Tanzania, the DRC, Kenya, and Malawi, where over 40,000 people in Tanzania alone were murdered on suspicion of witchcraft between 1960 and 2000 and reports indicate that more than 700 women were killed in Tanzania in a single year (2014), emerged, she argues, from exactly the same structural conditions that produced their European predecessors: the breakdown of communal land tenure, the imposition of privatized property regimes, and the intensified competition for resources among communities stripped of their reproductive safety nets by structural adjustment. The World Bank's prescriptions for African agriculture, such as land titling, commercialization, and the dismantling of communal and clan-based tenure systems, produced the dispossession that witch-hunts then accelerate. What Federici observes is a feedback loop: structural adjustment destabilizes communities; destabilized communities generate accusations; accusations strip women of whatever land and property they held. The stripped assets concentrate further into male-controlled patrilineal inheritance systems.

This is not a story about Africa's failure to modernize. This distinction matters because the standard colonial narrative runs in the opposite direction: colonialists arrived at communities that practiced lethal witch-hunting, and their legal codes eventually suppressed the practice. The historical record does not support this. Prior to British administration, communities across what is now northern Ghana held beliefs about witchcraft and malevolent spiritual power, but the response to those beliefs was rarely lethal. A community member suspected of witchcraft might be required to confess, pay restitution, undergo ritual purification, or accept social sanctions. Expulsion to permanent exile camps was not the customary response. The anthropologist Alexandra Crampton, whose fieldwork in northern Ghana produced some of the most granular documentation of the practice, finds no evidence of pre-colonial witch camps functioning as they do now.

It was colonialism that introduced an escalation. The British colonial administration restructured northern Ghana's chieftaincy system, concentrating power in lineage heads and traditional authorities while simultaneously closing off women's access to colonial courts. The restructuring had a specific logic: indirect rule required identifiable, stable, male-controlled hierarchies. Customary institutions that had operated more fluidly, with women occupying recognized roles as mediators, healers, and land custodians within certain ethnic communities, were formalized into patriarchal bureaucracies. At the same moment, colonial law withdrew from precisely the communities where women's customary protections had been most substantial. The vacuum this created was filled by the oracle, by the traditional priest, and the slaughtered chicken, whose verdict, unchecked by either customary communal accountability or formal legal recourse, became permanent. The camp, in other words, is a colonial institution dressed in pre-colonial language. The British made it possible. Structural adjustment made it profitable. The modern Ghanaian state, in its ongoing abstention, makes it necessary.

Dzodzi Tsikata, Distinguished Research Professor at SOAS and one of Ghana's foremost scholars of gender and land, has spent three decades mapping the fragility at the center of Ghanaian women's economic existence. Women in northern Ghana, she documents, access land almost exclusively through men, fathers before marriage, husbands after, sons in widowhood. The arrangement is not merely inequitable; it is structurally precarious in a way that becomes catastrophic at exactly the moment when men disappear. A woman married from her natal community loses her claims there. If her husband dies and his family turns hostile, she has no customary fallback. The land goes. The house goes. The cattle go. If she is old enough, past the reproductive and productive phases in which patriarchal economies find women useful, no one argues on her behalf.

This is the architecture that witchcraft accusations exploit. The accusation is the trigger; the target was always the property.

In this light, the demographic profile of camp residents reads as a portrait of the system's internal logic. Women who are widowed. Women who are childless, or whose children are absent. Women who are elderly, past the reproductive utility that structures their social value. And, most revealingly, women who have accumulated too visibly: those who are hardworking, successful, and land-holding. The accusation's threshold is triggered by both extremes. Too poor, and you are an inconvenient dependent. Too independent, and you are a transgression. Either way, the word comes for you. Witchcraft accusations, the scholar Mensah Adinkrah argues in his comprehensive study Witchcraft, Witches, and Violence in Ghana, are traceable to

"deeply held misogynistic attitudes and gynophobic beliefs, which are the effects of patriarchal arrangements and ideology embedded in the society."

The witch is, in the end, any woman who exceeds the boundaries the patriarchal order draws around female existence. The supernatural language gives social permission for what is, at base, a property crime.

On July 23, 2020, a 90-year-old woman named Akua Denteh was beaten to death in the village of Kafaba, in Ghana's Savannah Region, while dozens of people watched. A fetish priestess named Sherina Mohammed had been brought in to cleanse the community. She declared Denteh, who is a devout Muslim, mother of seven, grandmother of 2, a witch, responsible for irregular rainfall and the burning of a political shed. Denteh was forced to drink a concoction. Then Mohammed and a 25-year-old accomplice named Latifah Bumaye slapped, kicked, and caned the 90-year-old for approximately an hour. A witness filmed it. The video circulated across Ghana and then across the world. Denteh died the next day. The video did not disappear. It lodged in the national conscience like a splinter, and it changed what was politically possible.

Ghana's Parliament passed the Criminal Offences (Amendment) Bill—legislation that would criminalize witchcraft accusations, impose penalties of up to five years' imprisonment, require accusers to compensate victims, and provide reintegration services—unanimously by voice vote on July 28, 2023. Mohammed and Bumaye were convicted of manslaughter and sentenced to 12 years each. Three women died from witchcraft-related violence in the months while the bill awaited presidential assent. President Nana Akufo-Addo declined to sign it, citing "constitutional concerns" and "potential financial implications." The bill lapsed with the dissolution of the 8th Parliament.

The structure of this sequence demands attention. The state witnessed the public murder of a 90-year-old woman on video. It passed a law in response. The president vetoed the law on procedural grounds. In the gap between those two facts, the video and the veto, lies the full weight of what Ghanaian feminist scholars mean when they say the camps are not an aberration but a manifestation.

The bill has not yet been enacted. As of March 2026, it remains in Parliament.

The camps remain full. Women are still arriving.

Before examining what the bill's passage would mean, it is worth examining what previous camp closures have meant, because Ghana has closed camps before, and the results complicate the promise that legislation alone makes whole what dispossession has broken.

In 2014, one of the six original camps, Bonyase, was closed through joint civil society and government pressure. Between 2019 and 2020, a second, Nabuli, followed. ActionAid Ghana and the Ministry of Gender documented both closures and classified most former residents as successfully reintegrated. What this meant in practice for many women was relocation to a different community, rather than return to the land, household, and social position they had occupied before the accusation. The property transferred during their expulsion remained in the hands of those who had seized it. The camp closed. The crime persisted in the landscape.

The word "reintegration" carries an implicit premise: that there is a stable social position to return to. For a widow whose land passed to her in-laws during her absence, whose house was sold or occupied, whose community organized around the oracle's verdict for years, reintegration means negotiating re-entry into the exact social arrangement that produced the accusation in the first place. The Criminal Offences (Amendment) Bill, as drafted, mandates financial compensation from accusers and criminalizes future accusations. It does not require the return of transferred property. It does not alter the customary inheritance regimes that left the widow without a legal claim to her husband's land before the accusation was made. It criminalizes the mechanism; it does not dismantle the architecture.

Calling this a limitation of the bill is not arguing against the bill. Women are dying without it, and the criminalization of accusations would sever the confidence with which communities deploy the oracle as a property instrument. But it is to insist on clarity about what signing would accomplish and what it would leave intact. Sister Monica Yahaya, a Catholic religious working with camp residents for decades, has named the structural loop without abstraction:

"Relatives cannot allow widows to inherit husband's possessions. They will definitely look for a reason to accuse them and then send them away from their homes in order to take properties."

The reason is already prepared before any misfortune occurs. The accusation is the implementation of a decision already made in a meeting the widow was not invited to attend.

A Songtaba representative, speaking to Amnesty's researchers, captured the anticipatory logic that the camp, in all its visibility, obscures:

"With the threat of witchcraft accusation, women are afraid of growing old in their communities. They are even afraid of contributing to the community or providing their opinion. They know that they can be accused of challenging the norms." 

What the bill cannot address, what no single law in isolation can address, is the anticipatory architecture of self-suppression that operates across the north among women who have calculated, correctly, that accumulation of wealth or lack thereof invites accusation.

State services at each camp are effectively nonexistent. The Ghanaian constitution prohibits the denial of personal liberty and mandates equal protection; international law, including CEDAW and the UN's 2022 guidelines on witchcraft accusations, classifies the practice as a form of gender-based violence. None of this translates into material conditions at the camps. No government program finances healthcare, food, or shelter. No legal structure makes expulsion illegal. Police interviewed by Amnesty described themselves as unable to intervene in "cultural matters." The logic is circular and precise: the state's abstention from the camps is also the state's endorsement of them.

There is a counter-argument available, and it must be taken seriously. Some scholars of witchcraft in postcolonial Africa warn against reducing the supernatural to a simple economic alibi, insisting that belief in witchcraft carries cultural weight that demands respect on its own terms. This is correct as far as it goes. The Gallup survey finding that 77 percent of Ghanaians believe in witchcraft, against a sub-Saharan average of 55 percent, represents a cosmology in which spiritual and material causation are intertwined.

This month, corporations around the world will publish statements on women's empowerment. They will include statistics, wage-gap charts, and photographs of women in hard hats, laboratory coats or corner offices. This is what International Women's Month has largely become inside the institutions that most loudly observe it: a measure of how far women have penetrated the upper tiers of the existing order. The question it asks is whether women can access power. The question it rarely asks is what happens to women whom that order has already deemed to have no power worth accessing, those whose productive utility to capital is already exhausted.

The global anti-feminist backlash, the organized, increasingly well-funded movement to reassert patriarchal norms against feminist gains, primarily manifests as the insistence that women's independence is dangerous, that women who exceed their assigned position are illegible to the social order and must be made to disappear.

It is worth pausing on what it means, theoretically, to arrive at a witch camp. To understand it fully requires something like the framework Mbembe develops in his work on necropolitics and what he calls death-worlds

"new and unique forms of social existence in which vast populations are subjected to living conditions that confer upon them the status of the living dead."

The camps produce, with terrible exactitude, what Mbembe, drawing on Orlando Patterson's analysis of slavery, identifies as the triple loss of subjugation: loss of home, loss of rights over one's body, and loss of political status. Women in Gambaga, Gnani, Kpatinga, and Kukuo have forfeited all three. They are alive. They are not free. The community that expelled them retains the authority to determine whether and when they return. An authority almost never exercised. Once named, the name adheres.

At Gambaga, one resident has lived in the camp for 45 years. Bachalibanoya Anaberi arrived after her husband died and his co-wife's children blamed her for the family's misfortune. She was around 40 when she arrived. She is 85 now. The accusation was never proven, never revisited, never withdrawn. The word that expelled her has never expired.