A persistent 80% reduction in rare earth supplies could cut GDP by 1.8% in Japan, 1.5% in the United States and 1.2% in Germany when manufacturers have little time to find alternatives, according to a new IMF working paper.

The results come from a stress scenario rather than a forecast. The authors model what could happen during a severe, prolonged shortage lasting less than a year, when companies cannot easily replace rare earths with other materials.

Small inputs can create large disruptions

Rare earth elements are used in relatively small quantities, but they perform specialized functions in electric motors, permanent magnets, semiconductors, wind turbines, consumer electronics and defense equipment.

The global markets for rare earth oxides and permanent magnets were valued at about $6 billion and $25 billion, respectively, in 2024. Their economic importance is much larger because shortages can interrupt production of high-value goods.

China’s control of several stages of the supply chain adds to the risk.

The paper estimates that China accounts for about 90% of permanent magnet production. Its share of heavy rare earth processing is even higher, including approximately 97% of oxide separation and 95% of metal refining.

Building alternative processing capacity is difficult in the short term. Separation and refining facilities can require large investments, lengthy approvals and specialized technical knowledge.

Supply networks amplify the losses

The researchers incorporated rare earths and permanent magnets into a detailed U.S. production map covering 405 industries. They then matched the results with OECD input-output data for the United States, Japan, Germany, France, the United Kingdom and India.

A preliminary measure identified 34 U.S. sectors that rely on rare earth inputs. Those sectors generated about $155 billion in value added in 2017, equivalent to 0.8% of U.S. GDP.

Germany appeared more directly exposed under that measure, with rare-earth-dependent activities accounting for about 2.5% of GDP. Yet the model produced a larger short-term GDP loss for the United States.

The difference comes from how industries are connected.

U.S. motor vehicles, electrical equipment, computers and electronics buy from and supply many other parts of the economy. A shortage affecting those industries can therefore spread beyond the companies that purchase rare earth materials directly.

Under the stress scenario, output falls across 19 U.S. sectors. Automotive production records one of the largest declines, dropping 9%, compared with a 5.8% decrease in Germany.

Japan experiences the largest overall GDP loss because it combines sizeable rare-earth-intensive industries with strong links between manufacturers and their suppliers.

Why it matters

The risk became more visible in April 2025, when China introduced export-licensing requirements covering seven rare earth metals and several related products, including permanent magnets.

Chinese permanent-magnet exports fell roughly 70% from a year earlier at the low point in May. Shipments subsequently recovered, making the disruption temporary rather than the prolonged shortage examined in the model.

The episode nevertheless showed how quickly licensing delays can reach automotive, electronics and other manufacturers.

The paper also notes that the number of export measures affecting critical minerals increased fivefold over the previous 15 years. Greater use of trade restrictions could make supply disruptions more frequent even when minerals remain available underground.

Longer-term adjustment changes the result

The economic losses largely disappear when the model gives companies several years to redesign products, find alternative materials or change suppliers.

Under assumptions representing adjustment periods beyond five years, the estimated GDP loss averages just 0.006% across the six countries.

The ability of rare-earth-intensive manufacturers to replace these materials with other inputs explains 97% of the difference between the paper’s short- and long-term results.

That finding suggests the greatest danger comes from an abrupt disruption rather than a gradual change in supply.

The estimates are not forecasts

The scenario assumes a persistent 80% reduction in all rare earth inputs available to each economy, including oxides, metals, compounds and magnets.

The researchers also assume that the same industries use rare earths at similar rates across the six countries. Differences in the results therefore reflect industry size, production methods and domestic supply networks rather than country-specific engineering data.

The model relies partly on 2017 input-output tables and does not cover every economy with substantial manufacturing exposure. The authors identify South Korea, Mexico, Canada and Southeast Asian countries as possible additions to future analysis.

The findings should therefore be read as an estimate of vulnerability under an extreme disruption, not as a prediction of an approaching recession.