Climate change, natural resources, and conflict: Navigating a complex nexus

Reprinted by permission from VoxDev

Extreme weather and the green transition are reshaping global conflict. We summarise the evidence on this trend and explore how policy can respond.

Editor's note: This article is part of series covering CEPR's Reducing Conflict and Improving Performance in the Economy (ReCIPE) programme. Oliver Vanden Eynde and Juan Vargas are the ReCIPE Theme Leaders on Climate Change, Natural Resources, and Conflict.

From parched rangelands to flooded villages, climate shocks are becoming increasingly frequent and severe. To mitigate their impact, governments worldwide have committed to a global shift to low‑carbon energy. However, while a large body of research shows how extreme weather events can increase the risk of conflict, an equally important strand of research documents the extent to which resource booms contribute to this. This is problematic as the green transition entails boosting the demand for key minerals, with the risk of further contributing to instability and violence.

In this article, we distil what we know, what we don't, and where policy can make a difference, drawing on decades of evidence across economics and other social and environmental sciences, which we have synthesised in a recent review (Vanden Eynde and Vargas 2025).

The climate-conflict nexus

Weather shocks - droughts, heatwaves, floods - systematically raise the likelihood or intensity of violence, especially where livelihoods depend on rain‑fed agriculture. Early work used rainfall as an instrument for income to show that bad years brought more conflict at the country level (Miguel et al. 2004). Later research drilled down to the grid‑cell level and improved measurement of droughts and heat, confirming the pattern across places and types of violence (Burke et al. 2015, Harari and La Ferrara 2018). The mechanisms are multiple: falling farm earnings can lower the opportunity cost of joining armed groups; scarcity may sharpen competition over land and water; and heat itself can heighten aggression, including in urban settings (Jacob et al. 2007, Bollfrass and Shaver 2015).

Two channels deserve emphasis:

1. Opportunity costs: when crops fail or wages fall, recruitment into armed groups and social mobilisation becomes easier (Dube and Vargas 2013, Vanden Eynde 2018).
2. Rapacity: when weather shocks increase the incentives for raiding, groups may fight to capture access to less‑affected pastureland or upstream water sources (McGuirk and Nunn 2025, Decet and Marcucci 2024). Migration and displacement can compound tensions, particularly where inflows meet pre‑existing ethnic or political cleavages (Ghimire et al. 2015).

But causality can also run the other way. Conflict often degrades the environment: illegal mining, coca or poppy expansion, and sabotage of pipelines destroy forests and pollute rivers, perpetuating vicious cycles in which environmental deterioration and violence sustain one another (Fergusson et al. 2013).

Natural resources: The green transition's double-edged sword

If weather shocks reshape opportunities locally, resource booms do so across sectors. Oil and mineral price increases have repeatedly intensified violence in producing zones, especially where extraction is capital‑intensive or lootable (Dube and Vargas 2013, Berman et al. 2017, Blair et al. 2021). By contrast, higher prices for labour-intensive agricultural commodities often reduce conflict, consistent with higher opportunity costs of fighting.

The green transition will exacerbate these dynamics. Demand for 'transition minerals' (lithium, cobalt, rare earths) is rising quickly, threatening to amplify rapacity incentives in some places, even as fossil‑fuel revenues shrink elsewhere (Hund et al. 2020). The forms of extraction will potentially shape the forms of violence observed: while bulky coltan invites protection rackets at mine sites, easily concealed gold pushes armed groups to tax downstream spending in settlements (Sanchez de la Sierra 2020). Pollution from mining these minerals - especially water contamination - can also depress farm yields far beyond the pit, creating the very livelihood losses that fuel conflict risk, though pollution remains largely understudied as a trigger of violence.

Importantly, risk factors overlap. Using publicly available data, our review documents that drought‑prone regions in Africa often sit atop mineral deposits. This implies that climate and resource risks will likely exacerbate one another. However, quantifying those complementarities is a priority for future research.

Climate policy priorities: Insure, adapt, and govern

Based on the rigorous evidence that has established causal links or robust predictive capacity, we identify a few key policy priorities.

1. Protect against shocks without feeding rapacity. Index‑based insurance and social protection can break the link from droughts to recruitment (Fetzer 2020, Gehring and Schaudt 2023). But their design matters: Insurance that stabilises incomes in bad years may unintentionally embolden predation in good years, implying the need for careful contract design and credible monitoring mechanisms (Jensen et al. 2025). For the case of concomitant shocks, rapid humanitarian assistance is essential - but targeting and timing are crucial. The evidence is mixed on whether food aid tempers or aggravates violence, underscoring the need for early‑warning systems and evaluation of delivery models (Nunn and Qian 2014, Mary and Mishra 2020).
2. Invest in adaptation and connectivity. Irrigation, drought‑resistant crops, and transport links can mitigate local weather shocks and reduce famine risk (Burgess and Donaldson 2010, Gatti et al. 2021). Yet, roads and markets are double‑edged: they can also help armed groups tax trade or move contraband (Müller‑Crepon et al. 2020). Infrastructure choices should therefore be paired with governance upgrades and community oversight.
3. Regulate extraction and share benefits credibly. Transparency and certification can reduce financing for armed groups in some contexts (Berman et al. 2017), though others - such as early Dodd‑Frank 3TG rules in the Great Lakes - appear to have raised violence in targeted areas by shifting incentives (Parker and Vadheim 2017, Stoop et al. 2018). The Kimberley Process shows stronger conflict‑reduction effects for alluvial diamonds, the type easiest to loot (Heffernan 2016, Binzel et al. 2023). This implies that regulation is context-specific and should consider structural factors, such as whether exploitation is industrial or artisanal, border proximity, and state capacity. But complementary measures matter too: local revenue‑sharing, information campaigns that set realistic expectations, and decentralised water and forest management can amplify the positive effects of well-designed regulation and reduce the scope for political capture (Armand et al. 2020, Krakowski and Mazur 2024).

Two qualifications are in order. First, none of this is cheap. But much of it - early‑warning plus targeted aid, insurance with guardrails, certification that actually fits mineral realities- costs less than protracted conflict, and can yield additional benefits in terms of economic growth.

Second, the design of effective policies requires additional scientific evidence. We argue that three gaps stand out:

1. Prediction: climate change and the clean‑energy build‑out feed back into the structure and volatility of climate shocks. Models must move beyond extrapolating from past experiences to integrate behavioural, political, and ecological mechanisms, including biodiversity loss.
2. Complementarities: we need integrated datasets and designs that test whether and how climate and resource risks multiply each other.
3. Pollution and displacement: both likely depress livelihoods and inflame tensions around mines, but rigorous evidence on conflict effects is thin.

Researchers and practitioners must turn this new evidence into anticipatory action before today's climate shocks become tomorrow's wars.