The Food System’s Hidden Import Dependency

When people picture food-trade power, they picture exporters of grain. The 2024 data confirms that picture is real and concentrated: Brazil alone supplies 55.4% of world soybean exports ($44.4 billion of $80.2 billion), the United States 23.9% of maize, India 29.3% of rice, and Russia 16.1% of wheat. A short list of countries dominates each staple.

But that map describes the output of farming. Underneath it sits a second map that describes an input -- fertilizer — and it looks different. Almost every harvest in those exporting countries, and in the importing countries that buy from them, depends on nitrogen, phosphate, and potash that is itself traded across borders and concentrated in a handful of suppliers. World fertilizer exports were about $54.6 billion in 2024, and the five largest exporters accounted for 51.0% of that trade.

This article holds the two maps side by side. The first is the familiar one: who exports the food. The second is the hidden one: who supplies the fertilizer, and how concentrated each importer’s supply is. The argument is not that grain exporters do not matter — they plainly do — but that a country’s food-system exposure is not captured by its grain trade alone. A country can be a grain powerhouse and still sit downstream of a narrow fertilizer supply chain. Throughout, we treat trade value as a measure of commercial flow, never as a direct measure of food on the table.

Read group by group, the concentration of food exports is striking but uneven. Soybeans are the most concentrated of the four: Brazil (55.4%) and the United States (28.9%) together ship more than five-sixths of all traded soybeans, with Paraguay, Canada, and Argentina trailing far behind. A soybean buyer is, in practice, choosing between two suppliers.

Maize and wheat are more contested. In maize, the United States (23.9%) leads, but Brazil (19.5%), Argentina (16.3%), and Ukraine (13.1%) keep the top of the table close. Wheat is the most distributed of the group: Russia leads at 16.1%, followed by Canada (13.7%), the United States (11.5%), Australia (9.6%), Ukraine (8.1%), and France (7.5%) -- six countries before you reach a tenth of the market each. Rice sits in between, with India a clear front-runner at 29.3% ahead of Thailand (17.2%) and Vietnam (10.5%).

The practical lesson is that 'food exporter concentration’ is really four different stories. A disruption to soybeans would route through two countries; a disruption to wheat would have more alternatives. This is the map most food-security commentary already uses, and it is worth getting right — but it is only the surface. The exporters on this map are themselves dependent on something further upstream, which is where the next section turns. Note that these are export values in current dollars; they reflect both price and volume, and they describe commercial trade, not the food actually consumed in any country.

Now turn the map over. Measured by 2024 exports of the five fertilizer HS6 products in scope, the supply side is led by Russia ($9.9 billion, 18.2% of world fertilizer exports) and Canada ($8.2 billion, 15.1%), followed by Saudi Arabia (7.0%), Morocco (5.4%), China (5.2%), and Oman (4.9%). The top five suppliers together are 51.0% of all fertilizer trade — a concentration comparable to the food map, but pointing at a very different set of countries.

What matters for any given importer is not the global picture but its own supplier mix. Some buyers spread their fertilizer purchases across dozens of sources; others rely overwhelmingly on one. The chart ranks importers that bought at least $500 million of fertilizer in 2024 — a threshold that keeps tiny buyers, where a single shipment can read as 100% dependence, from dominating the leaderboard — and plots each importer’s purchase scale against its top-supplier share and supplier concentration (HHI).

Brazil is the single largest fertilizer importer in the world at about $7.5 billion, ahead of the United States ($6.9 billion) and India ($6.2 billion). But scale and concentration diverge. The United States buys slightly less than Brazil yet draws 50.9% of it from a single supplier, Canada (an HHI of 0.30); China draws 32.8% from Russia; India, by contrast, is notably diversified, with no supplier above 20.8% and a low HHI of 0.12. The same dollar figure can hide a thin supply chain or a broad one — which is exactly why we keep value and concentration on separate axes. Top-supplier share is a concentration proxy: it measures where purchases originate, not the route they travel or whether a buyer could switch.

The aggregate ranking flattens out how different these supply chains really are. Five case countries make the contrast concrete — and they span almost the full range of fertilizer-import situations.

Brazil is the world’s largest fertilizer importer ($7.5 billion) and a soybean superpower at the same time: its top supplier is Russia at 32.6%, with Canada second (18.6%), across 51 supplier countries (HHI 0.16). The exposure is real but diversified by count. India ($6.2 billion) is the most spread-out of the five: Saudi Arabia leads at just 20.8%, with Russia, Oman, and Morocco close behind and 44 suppliers in total — the lowest concentration in the group (HHI 0.12). Indonesia (~$1.1 billion) leans harder on two countries, Russia (33.2%) and Canada (32.3%), which together are nearly two-thirds of its imports (HHI 0.24).

Bangladesh (~$1.5 billion) is more concentrated still: its top three suppliers — Morocco (38.5%), China (29.0%), and Saudi Arabia (27.6%) -- together make up about 95% of its fertilizer imports, an HHI of 0.31 across just 11 sources. Egypt is the outlier that proves trade value must be read carefully: it imports only about $0.11 billion of fertilizer, and 91.0% of that sliver comes from Jordan — but Egypt is itself the world’s seventh-largest fertilizer exporter at about $1.8 billion, a gas-based producer that sells far more than it buys. Egypt’s tiny, concentrated import line is not a sign of fragility; it is a sign that Egypt sits on the supply side of this map. Reading its 91% top-supplier share as 'dependence’ without that context would be exactly the kind of mistake this article is built to avoid.

Why does fertilizer concentration get attention now? The timeline shows it. World fertilizer trade value rose from about $38.6 billion in 2020 to $93.8 billion in 2022 — roughly 2.4 times higher in two years — before easing to $61.4 billion in 2023 and $54.6 billion in 2024. Wheat moved in parallel, with world wheat export value peaking at $78.9 billion in 2022, up from $54.3 billion in 2020. Food and its key input spiked together.

Two cautions are essential here, and the data itself enforces them. First, this is value, and a value spike of this size is dominated by price, not by more tonnage delivered. The 2022 jump reflects a price shock in fertilizer and grain markets, documented externally by the FAO, the World Bank, and others; it does not mean the world shipped 2.4 times more fertilizer. Second, BACI is annual, so it cannot show the within-year timing — the planting seasons missed, the export bans imposed and lifted — that made 2022 a crisis rather than a line on a chart. We mark 2022 as a documented disruption but keep the causes (the war in Ukraine, export restrictions, energy-driven production costs) as external context, not as something derived from trade data.

One structural shift the data does show clearly: Belarus's share of world fertilizer exports fell from 7.0% in 2018 to 2.1% in 2024, even as Russia's share held near or above its earlier level (18.2% in 2024). A major potash supplier receding from recorded trade is the kind of change that tightens an already concentrated market — and it is visible in the measured shares, independent of the price story.

Put the two maps together and a different way of reading agricultural risk emerges. The grain map tells you who can sell food; the fertilizer map tells you who depends on a narrow set of suppliers to grow it. The two do not line up. Brazil dominates soybean exports yet is the world’s largest fertilizer importer, drawing a third of that fertilizer from a single source. India is a top rice exporter and is, simultaneously, the most diversified fertilizer buyer of our five cases. A country’s strength on one map says little about its exposure on the other.

That is the core of the claim: fertilizer import dependence reveals a layer of food-system vulnerability that the grain-export map cannot show. It is a structural factor, and only one of several. Whether concentration on the input side translates into a harvest problem depends on things outside this dataset — domestic fertilizer production, stockpiles, subsidies, the substitutability of suppliers, and the weather. We are careful not to overclaim: a high top-supplier share is a measured concentration, not a forecast of scarcity, and certainly not a ranking of which populations would suffer.

What the trade data does well is locate the structure -- where the chokepoints in the input supply chain sit, and which importers have the thinnest cushion of alternative suppliers. That structural lens is most useful precisely when read alongside the external evidence — agronomic, agricultural-economics, and humanitarian data — that can speak to consequences. The trade record draws the map; it does not, and should not, be made to narrate the crisis on its own.

Every value in this article is computed from CEPII BACI HS92 bilateral goods trade by a committed, re-runnable query; no figure was reasoned out. The measured layer has three parts. Food-commodity exporter totals use exporter-country sums for wheat (HS6 100110, 100190), maize (100510, 100590), rice (100610, 100620, 100630, 100640), and soybeans (120100). Fertilizer dependency uses importer-by-source-country sums across five fertilizer codes (310210, 310230, 310280, 310420, 310530): for each importer we compute total fertilizer imports, the largest single supplier and its share, and the concentration of the supplier mix as a fractional Herfindahl-Hirschman Index in the range 0 to 1. The timeline uses annual world fertilizer trade value, and Russia's and Belarus's shares of it, from 2018 to 2024.

The proxy layer is the interpretive step from top-supplier share to 'dependency’ or ‘exposure’. Source-country concentration is measured; it is not the same as a route, and it does not tell us whether a buyer could substitute another supplier, draw down stocks, or rely on domestic production. We therefore call it a concentration proxy and apply a $500 million minimum-import threshold to the ranking so that trivial buyers cannot top it on a single shipment.

The external context layer is everything about consequences. The 2022 value spike is real in the data, but its causes — the war in Ukraine, export restrictions, and energy-driven production costs — and its human effects on food prices and availability come from the FAO, the World Bank, IFPRI, and similar sources, and are kept strictly separate from the BACI numbers. Standard caveats apply throughout: trade value is not food availability; annual data cannot capture within-year disruptions; BACI numeric country codes can differ from ISO codes (Russia is 643, the United States 842); the product history is HS92-compatible through 2024; and we deliberately do not rank hunger or suffering -- the figures describe trade concentration, full stop.