Half the World's Wheat Importers Buy From Exporters Whose Harvests Fail Together

Most of the wheat that crosses an international border in a given year leaves one of about a dozen countries. Most of the nations that buy it order from several of those exporters at once, which looks like ordinary prudence: spread the contracts around, and a poor harvest in one place is covered by a decent one somewhere else. There is a condition attached to that arrangement, and it is rarely stated out loud. Diversifying suppliers only helps as long as their bad years stay out of step with each other.

How much stress a synchronized bad year puts on the countries at the far end of that chain is the subject of a study published Aug. 3 in Environmental Research Letters. Elisabeth Vogel, of the University of Melbourne, UNSW Sydney and the Helmholtz Centre for Environmental Research in Germany, worked with five colleagues in Australia and Germany to follow the damage from the field into the shipping manifests and on to the importers. They joined national agricultural statistics from the U.N. Food and Agriculture Organization to bilateral trade records and gridded climate observations covering 1992 through 2022, using mapped growing calendars so that a hot spell counted only if it landed on wheat while wheat was in the ground.

The production half of the answer is the blunt half. Across the top wheat exporters, variation in the area exposed to heatwaves during the growing season accounts for 56% of the interannual variability in how much wheat those countries produce. One climate variable tracks more than half the swing. The same measure explains about a third of the variability in their total exports.

That number needs its boundaries stated. It describes heat-exposed area against production among the leading exporters over three decades, not the share of global crop-yield variability attributable to climate in general, which other work has put closer to a third and measures differently. The two figures are not in conflict, and they are not interchangeable either.

The exposure itself is growing, and growing fastest where the world can least afford it. The extent of wheat-growing land experiencing anomalously hot growing seasons rose 7.9% per decade globally between 1992 and 2022. Among the top producers it rose 6.7% per decade. Among the top exporters, the countries whose surpluses everyone else is buying, it rose 9.3%.

Drought shows a stranger pattern, and it is the one worth slowing down for. Measured by rainfall, drought in wheat regions has no significant trend across the three decades. Measured in the soil, it climbs steadily: agricultural drought, defined by soil moisture, rose 4.2% per decade globally, 4.8% among top producers and 8.5% among top exporters. Rain has not deserted these fields. Heat is pulling water out of them faster than it arrives, and the authors read the gap between the two indicators as a sign that rising temperature, rather than failing rainfall, is what is drying the ground.

Irrigation changes almost everything about that picture. In predominantly rainfed exporters, heat exposure and production move together tightly. In the three major irrigated producers, China, India and Pakistan, the relationship is not statistically significant at all. A canal is an effective piece of climate adaptation, at least until the water behind it becomes the constraint.

Then the wheat leaves the country, and the study starts asking a question that is harder to answer. Working importer by importer, the team checked whether each buyer's actual trade partners are the kind of countries whose harvests suffer under widespread heat. For 81 importing nations, 48% of all of them, the answer is yes. More than half of importing countries have partners whose production is significantly hurt by drought.

What does not follow is a straight line from a scorched field to an empty shelf. Production losses do not translate linearly into export reductions, the authors write, because exports are shaped by economic factors and policy decisions: prices move, governments release stocks or restrict shipments, buyers switch suppliers or switch crops. Effects on imports and on domestic supply, the paper says, are "more varied" than effects on production. The transmission through trade is real, and it is not proportional. That qualification is the study's own, and it belongs beside the finding rather than in a footnote.

To rank who is most exposed when the transmission does happen, the researchers built a composite risk index out of three ingredients: a country's socio-economic vulnerability, how much of its wheat consumption it imports, and how exposed its trade partners are to climate extremes. The highest values cluster in sub-Saharan Africa, in the Middle East and North Africa, and in East Asia and the Pacific. These are countries that buy a large share of what they eat from a narrow set of suppliers, and several of them buy heavily from the same few.

The production-side premise here is not new on its own. A separate analysis published in Geophysical Research Letters in April compared observations from 1982 to 2019 against eight CMIP6 climate models and projected that compound drought and heatwave events will grow more frequent and longer in regions including the Indo-Gangetic Plain and the central United States, with episodes stressing three or more breadbaskets at once becoming more common later this century. That work stops at the farm gate. Vogel and her colleagues add the leg that follows: who is buying from whom when it happens.

The limits are the ordinary ones for a study built on national annual statistics. The relationships are correlational, they cannot resolve what happens inside a single country's borders or in a single season, and the risk index ranks exposure rather than predicting a shortage. What the work produces is a map of where the wheat trade is wired so that one hot summer is felt in several places at once. For an agriculture ministry deciding where to sign next year's contracts, the useful part is the list of suppliers that tend to fail in the same summer.