Prices for the 12 minerals where the US is 100% import-reliant moved 11% on average in 2025 — far less than the 26% swing across all tracked critical minerals, whose biggest jumps (bismuth +270%, antimony +144%, germanium +106%) came from export-restricted commodities the US does NOT fully depend on imports for

Summary

The commodities the U.S. is most import-dependent on — the roughly dozen nonfuel minerals USGS rates at 100% net import reliance (Arsenic, Fluorspar, Gallium, Graphite, Indium, Manganese, Niobium, Scandium, Tantalum, Titanium sponge, Yttrium, plus Cesium and Rubidium, which USGS excludes from price tracking) actually moved less, on average, than the broader universe of critical-mineral prices in the most recent year on record (2024 to 2025). Among the 12 of these with a tracked price, the mean absolute 1-year price change was 10.8% (median 7.5%), ranging from Manganese's -19% to Gallium's +32%. Across all 55 critical minerals USGS tracks prices for, the mean absolute change was 26.1% (median 11%) — pulled up by a handful of extreme spikes (Bismuth +270%, Antimony +144%, Germanium +106%, Rhenium +91%) that USGS explicitly attributes to newly imposed export restrictions, not to how import-dependent the U.S. happens to be on those specific commodities.

In other words: within this one-year window, being 100%-import-reliant did not correlate with bigger price swings. The largest swings hit commodities in the 50-90% import-reliance band that became geopolitical targets for export controls in 2025.

Step 1 — Defining "most import-dependent"

USGS's Mineral Commodity Summaries 2026 (ver. 1.3, May 2026) is the primary federal source for net import reliance (NIR), defined as imports as a share of apparent consumption. In 2025 the U.S. was 100% net import reliant for 13 of the 60 commodities on the Final 2025 List of Critical Minerals (an additional 20, including the 14 lanthanide rare earths, were 50-99% reliant). Per USGS Table 7 (ordered by descending NIR), the 100%-tier / data-unavailable-but-effectively-total-reliance group is: Arsenic, Cesium, Fluorspar, Gallium, Graphite, Indium, Manganese, Niobium, Rubidium, Scandium, Tantalum, Titanium sponge, Yttrium, and the heavy rare earths as a block.

Step 2 — How much their prices moved

USGS Figure 10A reports the exact 1-year (2024→2025) percent price change for each tracked critical mineral. Restricting to the 100%-NIR tier (12 of 14 have tracked prices — Cesium and Rubidium are excluded by USGS itself "because there was not enough information available regarding prices"):

Computed in Python from these 12 values: mean absolute change 10.8%, median absolute change 7.5%, mean signed change +4.1%, population standard deviation 13.9 points.

Step 3 — Comparison against the broader critical-minerals price universe

USGS Figure 10A covers 55 critical-mineral prices in total (spanning the full NIR spectrum, not just the 100% tier). Across all 55: mean absolute change 26.1%, median absolute change 11%, mean signed change +20.3%, standard deviation 45.9 points — roughly 2-3x more volatile than the fully import-reliant group, both in typical size of move and in dispersion. The extreme values driving that gap: Bismuth +270%, Antimony +144%, Germanium +106%, Rhenium +91%, Tellurium +60%, Tungsten concentrate +51%, Ruthenium +53%, Ytterbium oxide +50%. USGS's own narrative ties the three largest (antimony, bismuth, germanium, all >100%) directly to export restrictions — several of these are commodities where China (and in a few cases Congo-Kinshasa, Gabon, Malaysia) imposed new controls in 2025, not commodities on the 100%-NIR list itself.

USGS also reports that 5-year (2021-2025) compound annual growth rates (Figure 10B) were positive for many critical minerals but declining for cobalt, dysprosium oxide, gadolinium oxide, holmium oxide, lanthanum oxide, palladium, rhodium, silicon metal, and vanadium — none of which sit in the 100%-NIR tier. The underlying bar-chart figure did not carry machine-readable numeric labels in the extracted PDF text (unlike Figure 10A, which included an explicit data table), so the 5-year CAGR is described qualitatively here rather than quantified per-commodity; a reader wanting exact 5-year CAGR values should consult Figure 10B in the source PDF directly.

Caveats

• n=12 for the 100%-tier price comparison is small, and one year is a short, noisy window — 2025 specifically saw a wave of new export-control actions (China, DR Congo, Gabon, Malaysia) that could make any single year unrepresentative.
• Two of the fourteen 100%-NIR commodities (Cesium, Rubidium) have no USGS-tracked price series at all, so "most import-dependent" and "has a trackable price" are not perfectly overlapping sets.
• Import reliance is calculated as a share of apparent consumption and can hit 100% partly because the U.S. has essentially zero domestic production (not necessarily because global supply itself is tight) — a different vulnerability than the export-restriction-driven spikes seen in commodities like antimony and bismuth.
• This askamerica connector's own schema (econ.trade_imports, econ.usitc_tariffs) does not carry a "net import reliance" measure — that concept exists only in USGS's own tables, so it was sourced directly from the USGS MCS 2026 PDF rather than computed from a connector table.