The research, published in BMC Environmental Science, was led by Yaw Asamoah, Kow Ansah-Mensah, Adams Osman and Raphael Ane Atanga of the University of Education, Winneba. The team chose the Bawku area deliberately. It is the driest part of Ghana, home to predominantly rain-fed subsistence farming, and simultaneously one of the country’s most notorious ethnopolitical conflict hotspots. If the theory of resource war, which holds that competition for scarce natural resources inevitably breeds violence, were to hold anywhere in Ghana, this would be the place. The study spans five districts: Bawku Municipal, Bawku West, Binduri, Garu and Tempane, lying between latitudes 10 degrees 30 minutes and 11 degrees 11 minutes north.

To build their evidence base, the researchers combined hard climate data with the lived experience of residents. Rainfall and temperature records were drawn from three gauge stations at Manga, Binduri and Garu-Tempane, covering a forty-year span from 1976 to 2015 supplied by the Ghana Meteorological Agency. The team aggregated daily values into monthly and annual figures, fitted a twelve-month Standardised Precipitation Index to characterise drought, and ran simple linear regressions of rainfall and temperature against time. On the qualitative side, they conducted sixteen focus group discussions across eight purposively selected communities, involving 172 participants, plus fifteen in-depth interviews with chiefs, opinion leaders, assembly members, security analysts and officials from water and environmental agencies, bringing the total to 187 participants. Transcripts were coded inductively using NVivo 11.

The statistical results contain a genuine surprise. Rainfall showed a positive but statistically non-significant trend, increasing by 2.72 millimetres per year (p = 0.163, R-squared = 0.051), meaning barely five per cent of the variability in rainfall could be explained by time. Temperature, by contrast, rose significantly by 0.023 degrees Celsius annually (p = 0.003, R-squared = 0.213), a warming trend that translates into roughly 0.23 degrees per decade. Annual rainfall totals mostly ranged between 800 and 1000 millimetres, peaking at 1339.4 millimetres in 2007 and bottoming out at 669.8 millimetres in 2008. The climate signal, in short, is one of warming with erratic, highly variable rainfall rather than a simple long-term decline.

Here lies the study’s most striking finding: residents’ perceptions contradict the gauge data. Farmers interviewed across the communities insisted that rainfall had declined, that the rains now arrive late in June rather than April, and stop by August, leaving crops damaged and planting calendars upended. Elderly participants recalled childhood downpours that no longer come. This gap between measured trends and perceived decline is not a trivial discrepancy. As the authors note, if people believe water is vanishing faster than the instruments suggest, their water management decisions, and their grievances, will be shaped by perception rather than measurement. The researchers argue this makes seasonal forecasting and drought warning systems urgently needed, so that adaptation can be grounded in evidence while still acknowledging lived experience.

The seasonal rhythm of scarcity emerged with brutal clarity from the interviews. During the roughly five-month wet season, residents reported no water problems at all: rivers, dams, wells and rainwater harvesting provide enough for domestic use and farming, and barrels and buckets stand full. But two or three months after the rains cease, streams vanish entirely. Residents dig temporary pits in dry riverbeds to scrape out irrigation water. Cattle walk long distances in search of fodder and water. In communities such as Yaanatinga, Aniisi, Yirigungu and Apotdabogo II, dams, hand-dug wells and streams were reported to be insufficient or completely dry, forcing some residents to sit idle through the dry season, unable to farm or even reliably cook and clean. Field observations confirmed the picture: dried-up streams, marshy remnants, empty water pans at boreholes, and bare, cracked earth.

When the researchers asked directly whether water scarcity had caused conflict, they received a revealing double answer. Residents emphatically denied any water conflict, yet immediately admitted that quarrels, fights and tension over water access, use and management were daily occurrences. Women and children scuffled over borehole queues, sometimes dragging parental enmities home with them. Vegetable farmers quarrelled at irrigation gardens when someone’s water was used by another. Livestock owners allowed cattle to roam into crops, triggering repeated disputes that police in Bawku Municipal were routinely called to settle. The residents’ narrow definition of conflict, shaped by memories of the deadly, decades-old Bawku ethnic conflict involving arms, deaths and displacement, led them to dismiss these frictions as mere normality. A security expert at the Kofi Annan International Peacekeeping and Training Centre interviewed for the study offered the corrective: every violent conflict begins at the latent stage, and unmanaged quarrels and tensions are precisely where escalation starts.

The sharpest case study is the 2014 Kuka-Lansa confrontation, which nearly escalated into full-scale violence. Kuka, upstream, controls a dam that should serve downstream Lansa. When water levels fell dangerously low, Lansa demanded the spillway be opened to relieve their drought-stricken farmlands. Kuka refused, arguing that crops still needed irrigation within their own community and that opening the dam would leave them with almost nothing. Lansa residents, in the words of one opinion leader, grew angry because they no longer received water while Kuka alone used it for farming. The dispute ended in a fight, though community leaders from both sides eventually met and defused it. The episode mirrors upstream-downstream conflicts documented in Iran’s Doroodzan Dam network and Tanzania’s Rufiji Basin, but with a crucial difference: in Bawku, local leadership and customary mediation stepped in before the situation metastasised.

That mediating capacity is exactly what the resource war theory, as classically formulated, fails to capture. The theory, associated with Michael Klare’s work on petroleum, river water and timber, identifies three drivers of resource conflict: rising demand, shortages and contests over ownership. The Bawku evidence supports the first two stages, demand and shortage clearly produce competition and friction. But the leap to war is not automatic. The authors point to governance quality, adaptive capacity, local mediation skills and societal cohesion as moderating factors that determine whether scarcity hardens into violence. Comparable cases in Kenya, Mali and Nigeria, where pastoralist and farmer-herder conflicts have turned lethal amid drought, institutional collapse and small-arms proliferation, illustrate what happens when those moderators fail. In Bawku, customary water-sharing arrangements, the involvement of chiefs and informal negotiation have so far kept tensions chronic rather than catastrophic, even in an ethnically polarised environment where any spark could ignite wider violence.

The study’s implications reach well beyond northern Ghana. It suggests that the scarcity-conflict relationship is non-linear and context-dependent, and the authors propose what they call a resource war moderation framework: scarcity produces tension, but escalation into violence occurs only when local institutions, social cohesion and governance are ineffective. Their recommendations are correspondingly practical. They call on the Ghanaian government to invest in household wells, community boreholes, dams and pipe-borne water targeted at the most water-stressed northern communities, with equity in distribution to avoid sowing future grievances. Crucially, they urge that traditional authorities and district officials be formally integrated into participatory water management, embedding climate sensitivity and local responsiveness into governance rather than relying on infrastructure alone. The Bawku case, in the end, is a warning and a lesson at once: water scarcity will not by itself start a war, but in a fractured society, every quarrel at a borehole is a conflict in waiting, and the institutions that absorb those quarrels are the region’s most precious, and most fragile, resource.

Subject of Research: The relationship between water scarcity driven by climate variability and latent or violent conflict in the Bawku area of Ghana's savannah ecological zone

Article Title: Does water scarcity drive conflict? The Bawku area case in the Savannah ecological zone, Ghana

Article References: Asamoah, Y., Ansah-Mensah, K., Osman, A., & Atanga, R. A. (2025). Does water scarcity drive conflict? The Bawku area case in the Savannah ecological zone, Ghana. BMC Environmental Science, 2(1), Article 20. https://doi.org/10.1186/s44329-025-00034-6

Image Credits: AI Generated

DOI: 10.1186/s44329-025-00034-6

Keywords: water scarcity, conflict, climate change, Ghana, Bawku, savannah zone, resource war theory, dry season, boreholes, latent conflict, water governance, rainfall variability