The Geopolitics of Water — Scarcity Manufactured by Bad Design

The hydrological sciences have known for decades what policymakers have been slow to accept: rainfall is not the primary driver of water availability in most of the world's agricultural regions. Land management is. This distinction has profound implications for how water scarcity should be understood, and for who bears responsibility for it.

The Rainfall-Absorption Gap

Annual precipitation volumes across most inhabited regions have not changed dramatically enough to explain the acceleration of water stress documented over the past half century. What has changed is the proportion of that precipitation that enters aquifers and sustains river flows versus the proportion that runs off the surface and out to sea within hours of falling.

In a healthy forested watershed, between 30 and 50 percent of precipitation infiltrates into groundwater. In a degraded agricultural watershed with compacted soils, that figure can fall to 5–10 percent. The difference is not the rain. It is what happens to the rain when it hits the ground. This infiltration loss is a direct consequence of how land above the watershed is managed: tillage that breaks soil structure, removal of vegetative cover, compaction from heavy machinery, replacement of perennial root systems with annual crops that leave soil bare for portions of the year.

The Loess Plateau restoration in China provides a large-scale natural experiment. Beginning in the late 1990s, the Chinese government undertook one of the largest watershed restoration projects in history across the Yellow River's heavily degraded upper basin — approximately 35,000 square kilometers of terracing, tree planting, and grazing restriction. Within fifteen years, satellite data showed dramatic increases in vegetation cover. Hydrological studies documented increased dry-season flows in previously intermittent streams, rising groundwater levels, and substantial reductions in sediment load in the Yellow River. The rainfall did not change. The landscape's capacity to hold water did.

Transboundary River Politics

Of the 263 internationally shared river basins, fewer than 60 have functional treaty frameworks, and fewer still have enforcement mechanisms with teeth. This means that the majority of cross-border water sharing operates on informal norms, historical precedent, and power differentials — which is another way of saying that upstream countries do what their domestic political economy demands, and downstream countries negotiate from weakness.

The Mekong provides a case study in what happens when upstream development proceeds without downstream agreement. China has built eleven major dams on the upper Mekong (the Lancang) and plans additional development. Studies by the Stimson Center using satellite data showed that in 2019, a severe drought hit the lower Mekong countries — but upstream at the Chinese dams, water levels were unusually high, suggesting that dam operations were holding water rather than releasing it during a period when downstream communities and agriculture needed it most. China disputed the methodology. The downstream countries — Thailand, Laos, Cambodia, Vietnam — had no enforcement mechanism and limited diplomatic leverage.

The Mekong River Commission, established in 1995, requires "prior notification" for major water infrastructure projects but not prior consent — a legal distinction that renders the mechanism advisory rather than constraining. China and Myanmar are not signatories. The political architecture for managing a river basin whose fish stocks feed 60 million people is fundamentally inadequate to the challenge.

The Indus Waters Treaty between India and Pakistan, brokered by the World Bank in 1960, is frequently cited as a successful transboundary water agreement — it survived two major wars between the signatories. But the treaty was designed for mid-twentieth century conditions: fixed infrastructure, stable climate basins, and population levels substantially lower than current ones. Indian development of run-of-river hydroelectric projects in the Indus basin has periodically brought treaty compliance disputes to international arbitration. Climate change is altering the seasonal distribution of glacial melt that feeds the Indus system. The treaty framework, while formally durable, is under increasing strain from conditions it was not designed to manage.

Aquifer Economics and the Race to the Bottom

Groundwater presents a distinct governance problem from surface water. Aquifers are invisible, their boundaries are uncertain, recharge rates are difficult to measure, and extraction is largely private. The economic structure of groundwater use in most countries creates a tragedy of the commons: every individual user has an incentive to pump as much as possible before others do the same, because the water they don't pump, their neighbors will. The result is systematic over-extraction.

The Ogallala Aquifer, which underlies the Great Plains of the United States and makes that region one of the most productive agricultural zones on Earth, is being drawn down at rates that dramatically exceed natural recharge. In some areas, the water table has dropped tens of meters in decades. The USGS has estimated that portions of the aquifer could see 70 percent depletion within fifty years under current extraction rates. This is not a secret. The agricultural states dependent on the Ogallala have known for decades. Governance responses have been inadequate because any state or district that moves to restrict pumping unilaterally simply loses competitive advantage while neighbors continue extracting.

The same pattern plays out in the Central Valley of California, the Gangetic Plain of India, the North China Plain, the Middle East and North Africa region. Each is drawing down a non-renewable or slowly-renewing resource at rates that are unsustainable on any reasonable planning horizon. The political economy that sustains the extraction is short-termist by design: individual farmers, water districts, and agricultural corporations are optimizing for near-term profitability under institutional frameworks that do not price water at its replacement cost.

The Virtual Water Trade and Hidden Dependencies

The concept of "virtual water" — the water embedded in traded agricultural goods — reveals a geopolitical dependency layer that conventional trade statistics obscure. When Saudi Arabia imports wheat, it is importing the water that grew that wheat. For a water-scarce country, this can be rational: domestic water is worth more for drinking and high-value crops than for low-value grain production. But it also means that food security becomes dependent on the water security of export partners.

Saudi Arabia pursued water independence in grain through the 1980s and 1990s, mining its non-renewable fossil water aquifers to grow wheat in the desert. By the 2010s, the aquifers were severely depleted and the wheat program was abandoned. The country reverted to import dependency — but had permanently destroyed a portion of its hydrological heritage in the interim.

The virtual water trade also creates leverage: countries that are major food exporters are effectively major water exporters. If water constraints force reductions in US, Australian, or Indian agricultural exports, the food-insecure countries currently depending on those exports face supply disruptions that their own depleted agricultural systems cannot compensate for. Water stress in exporting countries translates into food insecurity in importing countries, mediated through commodity markets.

Design Solutions at Scale

The path toward reduced water geopolitical conflict runs through landscape restoration and distributed water management rather than centralized engineering and treaty frameworks — though both remain necessary.

Keyline design, developed by Australian farmer and designer P.A. Yeomans in the 1950s, uses subtle contouring of agricultural landscapes to direct water movement across slopes, maximizing infiltration and minimizing surface runoff. Applied at watershed scale, keyline principles can substantially increase the water retained in a landscape. The technique requires minimal infrastructure — primarily shaped earthworks — and produces compounding benefits over time as soil organic matter builds and soil structure improves.

Farmer Managed Aquifer Recharge (FMAR) schemes, implemented across India, Australia, and the western United States, enable individual farmers to manage check dams, infiltration basins, and managed flooding to direct surface runoff into aquifers rather than allowing it to leave the watershed. These are bottom-up interventions at household or community scale that aggregate into watershed-level hydrological improvement.

Urban water management through green infrastructure — permeable surfaces, bioswales, urban trees, green roofs — dramatically reduces the proportion of urban precipitation that enters storm drains versus infiltrates to recharge urban aquifers or sustain urban greenery. Cities that have invested in green infrastructure at scale, including Philadelphia's Green City Clean Waters program and Singapore's comprehensive ABC Waters program, have documented substantial reductions in combined sewer overflows and improved urban water resilience.

The fundamental reframing required is this: water scarcity is primarily a symptom of landscape dysfunction, not a natural resource limitation. Countries that restore their landscapes restore their water. Countries that continue to degrade their landscapes while building dams are treating symptoms while accelerating the underlying cause. The geopolitics that follow from manufactured scarcity are also, therefore, manufactured — and preventable.