Japan’s Rice Crisis Swings from Shortage to Surplus

After empty shelves and emergency releases, Japan now carries far more rice into a new harvest while household demand is weakening. Consumers want relief; farmers fear that a sharp fall will erase margins just as fertilizer, fuel and labor remain expensive. The reversal exposes a system tuned too tightly to a shrinking market.

The reversal in four numbers

At the end of May 2026, Japan’s private-sector rice inventory stood at 2.23 million tonnes of brown rice, the agriculture ministry reported on June 30. That was 740,000 tonnes above the same month a year earlier. Shippers had collected more rice but sold substantially less: cumulative sales of the 2025 crop were 1.322 million tonnes, down 272,000 tonnes year on year.

The government’s March supply-and-demand guidelines put the 2025 staple-rice crop at 7.47 million tonnes, against demand of 6.91 million to 7.04 million tonnes. Together with opening stocks and 230,000 tonnes of government reserve releases, that pointed to private stocks of 2.21 million to 2.34 million tonnes at the end of June 2026—up sharply from 1.55 million tonnes a year before.

Retail prices are finally responding. Agriculture ministry supermarket data showed an average ¥3,458 for five kilograms in the week of June 29 to July 5, down ¥96 in one week and 4% from a year earlier. The average fell below ¥3,500 for the first time in 79 weeks. Yet this was still a costly bag for households accustomed to pre-crisis prices near or below the low ¥2,000s.

How a shortage becomes a surplus so quickly

Rice is produced once a year but eaten every day. Farmers decide acreage and inputs months before harvest. Consumers can switch to bread, noodles or cheaper imported and blended rice in a week. Wholesalers, retailers and government can release stocks at different speeds. The result is a classic delayed-feedback system: high prices signal farmers to plant more only after the shortage has already changed behavior.

What caused the 2024–25 shortage?

There was no single cause. The 2023 heat wave damaged grain quality and reduced the share that met ordinary grading standards. Private stocks entered summer 2024 at 1.56 million tonnes, 20% below a year earlier and the lowest since 1999. At the same time, rice demand rose for the first time in roughly a decade as its relative price looked attractive and inbound tourism recovered.

Then psychology accelerated the physical tightness. A major-earthquake advisory in August 2024 and typhoon warnings led households to buy extra rice. Supermarkets introduced limits. Empty shelves became information: people who saw scarcity bought earlier, making local scarcity worse even if the national balance sheet still showed rice.

Distribution mattered too. Japan’s government initially argued that total supply was adequate and hesitated to release strategic reserves. When it auctioned reserve rice in early 2025, grain moved slowly through established collectors and wholesalers. By late April, only a small share had reached stores. Later direct contracts with major retailers accelerated delivery and offered five-kilogram bags near ¥2,000, but the emergency intervention then became additional supply that had to be digested.

A national balance can say “enough rice” while a neighborhood shelf is empty. Inventory location, ownership, quality, packaging and willingness to sell are part of supply—not footnotes to it.

Why retail prices and farmer prices do not move together

“Rice price” can mean at least four things: the advance payment or procurement price offered to farmers, the negotiated price between collectors and wholesalers, the wholesale price to retailers and restaurants, and the shelf price paid by consumers. Each includes different timing, quality, transport, milling, packaging, storage and risk.

A retailer may still be selling rice bought under an expensive contract even as the expected harvest price falls. A farmer may see a lower offer before shoppers see cheaper bags. Branded single-origin rice can remain expensive while government-stock or blended rice pulls down the average. This lag explains why consumers can still complain about high prices at the same moment producers fear a collapse.

The May ministry survey illustrated the split. Wholesalers’ selling volume was only 88% of a year earlier. Prices to retailers were roughly flat at 101% of the prior year, while prices to prepared-food and restaurant customers were 119.6%. Different channels were clearing old contracts at different speeds.

Why farmers fear a “fair” consumer price

Rice farmers do not receive the supermarket price. From the shelf price come milling loss, bags, transport, storage, finance, wholesaler and retailer operations, and consumption tax. A decline of ¥1,000 per five-kilogram bag cannot be read as ¥200 less per farm kilogram.

Meanwhile, farm costs have risen. Fertilizer, fuel, machinery, drying, land improvement and hired labor are expensive; the yen has increased imported input costs. Many Japanese growers are small or part-time, and field fragmentation raises machinery and travel costs. Large operators may be efficient but often rent many scattered plots and face high fixed investment.

A rapid price fall is therefore asymmetric. Consumers regain purchasing power, but farmers may be locked into seed, fertilizer, land and equipment decisions made when the market was high. If the farm-gate price falls below full cost, older producers may exit and tenants may decline marginal paddies. That can shrink future capacity and revive shortage risk.

The 2026 crop: one target, two production numbers

The March basic guidelines set a policy production outlook of 7.11 million tonnes of brown rice for 2026, matching the upper end of projected 2026/27 demand. But farmers’ January planting intentions—1.361 million hectares of table rice plus 14,000 hectares of reserve rice—could yield 7.19 million to 7.32 million tonnes of staple rice after planned reserve procurement is considered, depending on yield and classification.

That difference is the heart of the surplus fear. Opening private stocks were projected at 2.21 million to 2.34 million tonnes. Add a harvest potentially above the policy assumption, then subtract demand of 6.96 million to 7.11 million tonnes, and June 2027 stocks could reach 2.29 million to 2.71 million tonnes under the planting-intention scenario. The range is wide because weather, milling yield, imports, reserve operations and demand are uncertain.

The government plans to buy 210,000 tonnes of 2026-crop rice for the reserve through advance contracts. That can remove some volume from the table-rice market. But officials must also decide how and when to replace as much as 590,000 tonnes sold from the reserve during the crisis. Buying back too quickly could support farm prices while rebuilding security stocks; it could also conceal oversupply and transfer carrying cost to taxpayers.

Stocks are insurance—but also a price signal

Japan’s strategic reserve target is about one million tonnes at the end of June. The system is designed to cover a once-in-ten-years poor harvest or two ordinary shortfalls, and rice is generally held for about five years before being rotated.

Private stocks serve another function. Millers and wholesalers need working inventory between annual harvests, across regions and quality grades. Too little creates fragile shelves; too much ties up capital, adds storage costs and signals that future purchases should slow. A healthy stock-to-use ratio is not zero.

With annual demand near seven million tonnes, 2.2 million tonnes equals roughly 3.8 months of consumption before adjusting for seasonal and channel differences. That is not intrinsically absurd after a crisis. The risk comes from the direction and composition: stocks rose quickly, sales slowed and another harvest was approaching.

A century of rice politics

From gentan to “demand-led production”

The famous acreage-reduction policy, gentan, is often described as having ended in 2018. The statement is both true and incomplete. The national government no longer assigns compulsory table-rice reduction quotas. Yet policy still publishes demand-aligned production figures and subsidizes alternative uses of paddy land, including feed rice, rice flour, wheat, soybeans and processing crops.

The reason is structural. Consumption fell from 118 kilograms per person in 1962 to 51.5 kilograms in fiscal 2021. Total table-rice demand declined by roughly 80,000 to 100,000 tonnes annually in recent years. If every productivity gain became more table rice, chronic surplus would depress prices and leave government with costly stocks.

But a tightly balanced system saves carrying cost by sacrificing shock absorption. When heat reduces quality, visitors add demand and households stock up, there is little slack. The answer is not simply “produce more forever” or “return to reduction.” It is to separate national resilience capacity from the annual price target.

Imports: open enough for commitments, closed enough for protection

Japan imports about 770,000 tonnes of rice annually under minimum-access commitments, much of it for processing, feed or food service rather than ordinary branded table-rice shelves. A high out-of-quota tariff—commonly stated as ¥341 per kilogram—protects domestic table rice. A simultaneous-buy-and-sell system allocates some imported rice to specified users.

During the shortage, supermarkets and restaurants experimented more visibly with U.S. Calrose and other foreign rice, and private out-of-quota imports rose despite the tariff. Imports can cap extreme prices and diversify climate risk. But an emergency opening that is closed after farmers have planted creates another delayed signal. Stable, transparent import rules are easier to plan around than crisis exceptions.

Exports and rice flour are often proposed as surplus outlets. They can grow, but they are not a magic sink. Export markets require consistent variety, price, branding and logistics. Rice flour competes with wheat and needs processors and product demand. Turning a temporary surplus into sustainable demand takes years.

The quality problem hidden inside tonnage

Rice balance sheets are usually expressed in brown-rice tonnes. Consumers buy milled rice and care about taste, variety, origin and crop year. Heat can reduce the share of first-grade rice through chalky kernels while leaving physical tonnage less affected. Milling yield varies. Old reserve rice and new premium rice are not perfect substitutes.

This explains an apparent paradox from 2023–24: official production numbers did not look catastrophic, but usable preferred rice in familiar channels was tight. The opposite can occur in 2026: ample total inventory does not mean every variety in every prefecture is excessive. Good policy needs grade-, location- and channel-level data, not one national total.

Climate change makes “just right” harder

Warmer temperatures can extend growing seasons in some regions but damage grain filling and quality elsewhere. Heat-tolerant varieties, changed planting dates, water management, forecasting and crop insurance reduce risk. They do not eliminate it. Water availability, extreme rain, typhoons and pests also matter.

A system that responds to one hot year by maximizing acreage may overshoot in a normal year. A system that assumes trend decline every year may undershoot when quality falls. Climate adaptation requires a buffer and flexible uses—not only a single national tonnage forecast.

Who bears the risk?

A better stabilization design

Publish one balance sheet, with uncertainty

Combine production, grade, private inventory, reserve releases, imports, sales and household demand in timely ranges. The public should see where rice is, not only how much exists nationally.

Use rules for reserve release and replenishment

Trigger bands based on stocks, prices and distribution stress would reduce political delay in a shortage and opportunistic buying in a surplus. Replenishment should be gradual and announced.

Support farm income without forcing consumer prices up

Targeted income support, insurance and payments for maintaining paddies can preserve capacity more transparently than scarcity pricing. Price support makes every consumer—including low-income households—pay through food bills.

Build flexible demand before the glut

Feed, rice flour, sake, prepared foods, exports and emergency aid require contracts, standards and processing capacity in advance. A disposal plan invented after harvest will offer poor prices.

Reward consolidation and climate adaptation

Land banks, field consolidation, heat-tolerant varieties, efficient dryers and shared machinery can lower cost. Keeping every current business model unchanged is not the same as keeping rice production resilient.

The lesson: resilience needs slack, not permanent scarcity

Japan’s rice market was managed for decades around predictable demand decline. That made chronic surplus the policy enemy. The 2024–25 crisis showed the cost of running too close to the edge; the 2026 reversal shows the cost of responding after prices have already changed planting and consumption.

The consumer and farmer do not have to be enemies. Consumers need affordable rice; efficient farmers need revenue that covers long-run cost; the country needs enough land, skill and stocks for a bad harvest. Those goals conflict only when policy uses one shelf price to accomplish all three.

The next test is not whether five-kilogram bags become cheaper—they already are. It is whether Japan can let prices normalize without pushing viable growers out, rebuild the strategic reserve without disguising excess, and preserve a transparent buffer for the next heat wave. A successful rice policy should make neither shortage nor surplus a surprise.