The New Cold War
America, China, and the Echoes of History
Wildfires in the western United States and hurricanes on the East Coast captured media attention this summer and fall. But throughout 2018, weather events also had devastating humanitarian consequences in developing countries, from immense floods in the Indian state of Kerala to an intense drought in Afghanistan that affected millions.
Over the past decade, academics and policymakers have vigorously debated the question of whether climate change poses a security threat, with particular emphasis on whether it causes internal conflict. Connections are complex, leaving policymakers to talk about climate change vaguely as a “threat multiplier” when combined with other forces. But saying that climate change is a threat multiplier isn’t all that helpful unless we know something about the characteristics that make countries more likely to experience instability.
In fact, several risk factors make some countries more vulnerable than others to the consequences of climate change. Three stand out in particular: a high level of dependence on agriculture, a recent history of conflict, and discriminatory political institutions. Research suggests that in countries that display some or all of these risk factors, climate extremes are especially likely to lead to disastrous outcomes, including violence, food crises, and the large-scale displacement of populations.
We have used these factors to identify the countries that are most at risk from climate-related instability and humanitarian crises in the coming years. In doing so, we hope to provide an early warning to policymakers about where climate impacts are likely to prove most destabilizing in the short term, and where efforts to minimize their effects are most needed.
The map below displays the countries we consider to be at major risk of climate-driven instability. A sizeable number of countries—20, in total—combine the first two risk factors, agricultural dependence and a recent history of conflict. Countries facing these two risk factors are shown in orange. A further nine countries (Angola, the Democratic Republic of Congo, Myanmar, Pakistan, Rwanda, South Sudan, Sudan, Uganda, and Yemen) have the first two factors, as well as high levels of ethnic political exclusion. These countries are displayed in red.
Why are these specific risk factors so important? High agricultural dependence means that a particularly large proportion of citizens’ livelihoods are bound up in farming. Since climate shocks such as droughts or floods tend to disrupt farming, they often translate, in countries that depend on agriculture, into widespread reductions in income and high levels of food insecurity. Reduced farming income makes alternative sources of income such as rebel activity more attractive. And governments in these countries tend to extract a considerable portion of their revenue from agriculture, mostly from taxes on exports. Using data from the International Labor Organization, we defined a country as having a “high” level of agricultural dependence if at least 40 percent of its population worked in agriculture—a threshold that 66 countries met. In some countries, such as Malawi and Somalia, over 80 percent of the workforce is employed in agriculture.
The second risk factor, recent violence, can also combine with climate shocks to produce crises: more than 40 percent of countries experiencing a food crisis in 2017 carried this double burden. More important, recent conflict is probably the best predictor that a country will have more conflict in the future, since these countries are likely to still have leaders and groups that can access weapons and mobilize people to fight. We used data from Uppsala Conflict Data Program to identify the more than 40 countries that had recorded conflicts (defined as a conflict to which the state was a party) in the five years prior to 2018.
The third factor, discriminatory political institutions—those that marginalize certain social groups or exclude them from power—increases the risk of instability and humanitarian emergencies by making governments less responsive to segments of their population. For instance, in the event of a food shortage, marginalized groups in these countries are less likely than others to receive government aid, increasing the likelihood of a humanitarian crisis and creating grievances against the state.
The political scientist Judith M. Bretthauer defines “high” political exclusion as occurring in countries where at least 20 percent of the population is formally excluded from political power. Using her criteria and information from the Ethnic Power Relations (EPR) dataset, we calculated that 35 countries currently have highly discriminatory institutions. Nine of those countries, indicated on the map, also had high agricultural dependence and a recent history of conflict. (At least two more countries, Cameroon and Ethiopia, had high levels of discrimination not captured by the EPR dataset.)
Countries facing the three aforementioned risk factors will struggle to deal with climate shocks. And one of the most serious of these shocks—especially in countries that depend on agriculture—is likely to be drought and water shortage. By looking at which of our at-risk countries are either currently facing water deficits or are projected to face them in the near future, we can identify not only the countries that are most vulnerable to the effects of climate change in the future but those that are already threatened by serious climate shocks.
Although water deficits are hard to measure globally, one approach is to model water balances by combining rainfall, temperature, geography, evaporation rates, and other factors to estimate water anomalies, where amounts of surface water will be higher and lower than normal. The consulting firm ISciences has developed a Water Security Indicator Model (WSIM), which uses both historical data and forecasts to produce a composite index of water surpluses and deficits. At our request, ISciences aggregated its subnational data and ranked all the countries in the world in terms of both the share of their population and the proportion of their territory that experienced water deficits between October 2017 and September 2018, as well as a similar list including projected water deficits until June 2019. The firm identified countries suffering a ten-year deficit, or a drought rare enough to occur on average once every ten years, and those with a more extreme 30-year deficit, which occurs on average once every 30 years.
We examined the 25 percent of countries with the worst water deficits according to ISciences’ data and cross-referenced them with our list of 20 at-risk countries. On the map below, at-risk countries without water deficits are displayed in light blue. Those that appeared in the top 25 percent of ISciences’ historical or projected lists for ten-year droughts are shown in medium blue, while those that appeared in the top 25 percent for 30-year droughts appear in dark blue.
Among the countries with both high agricultural dependence and a recent history of conflict, a few stood out in ISciences’ historical data for the twelve-month period from October 2017 to September 2018. In Afghanistan, 64 percent of the population had been exposed to a thirty-year drought; in Yemen, the figure was 43 percent, and in Somalia, 34 percent. Looking only at a share of the population, however, could mask the problems in less densely populated rural areas. Eighty percent of Afghanistan’s territory and 38 percent of the Democratic Republic of the Congo’s territory had undergone a ten-year drought, while 36 percent of Sudan’s territory suffered a 30-year drought. India was not among the top 25 percent of states in terms of exposed share of population or territory, but about 95 million Indians, or 7.2 percent of the population, had weathered a 30-year drought over the past year.
ISciences projects that Somalia, Sudan, and Yemen will continue to face water deficits through June 2019. But an additional six countries not included in the top 25 percent for ten-year or 30-year deficits in the 2017–2018 list—Chad, Cameroon, the Central African Republic, Ethiopia, Mali, and Niger—are projected to rise into the top quartile by June 2019.
Promising efforts are under way to further increase our ability to forecast political violence and food crises. The World Bank recently rolled out a pilot of its Famine Action Mechanism, which tries to anticipate famine locations using machine learning. And the Uppsala University’s Violence Early-Warning System (ViEWS) Project has been developing subnational conflict forecasts for Africa. Before such systems reach their full potential, however, our list can help detect which countries are likely to become hot spots of climate instability and humanitarian crises.
What should policymakers do with this information? In countries enduring large-scale violence, such as Yemen or South Sudan, the key priority clearly must be dampening armed conflict. Countries where violence has ceased, however, can benefit from building their capacities to detect and respond to crises, and from improving job opportunities. Here the kinds of interventions that food policy experts have developed to prevent famines—including food and income support and crop insurance—can be of help. When a major drought hit Ethiopia in 2015, its government was able to use similar policies to ward off famine, although even this success did not fully insulate it from protests. And given the particular susceptibility of agriculture to changes in the climate, governments and international institutions need to urgently consider how agricultural practices can be adapted to meet these changes and prevent large declines in crop yields.
The risk factors we identify are, naturally, not the only conditions that can contribute to instability and humanitarian crises. But unabated climate change is likely to amplify the challenges of these high-risk countries in decades to come. They will see more extreme consequences, and their already fragile governments will become even more hard pressed to manage violence and feed their populations. Understanding where instability is most likely to occur is an important first step to reducing risk.