# Lithium Triangle Chile, Argentina and Bolivia, and the high-altitude salt flats between them. The case in [[Conflict Resources Atlas]] with **no armed group at all** — where the contested questions are consent, water and who owns the state's share. > [!warning] "The Lithium Triangle holds over half the world's lithium" is a resources claim > It is not a reserves or production claim, and the three are very different things. The Triangle holds roughly **43% of global resources**, about **37% of reserves**, and produced roughly **27% of world output in 2025**. Bolivia has **no USGS reserves entry at all** — only resources. That distinction is the entire Bolivian story and is constantly misreported. | | 2025 production (t Li) | Reserves | Resources | |---|---:|---:|---:| | Chile | 56,000 | 9.2 Mt | 13 Mt | | Argentina | 23,000 | 4.4 Mt | 28 Mt | | Bolivia | ~460 (≈2,462 t LCE) | **none listed** | 23 Mt | | *World* | *290,000* | *~37 Mt* | *~150 Mt* | ## Three states, three completely different models **Chile — state-centric, and it survived a change of government.** Boric's April 2023 National Lithium Strategy kept lithium non-concessionable state property, with private access only through special operating contracts. The **Codelco–SQM joint venture was formally constituted on 27 December 2025** as NovaAndino Litio, running to 2060, with Codelco holding the controlling economic interest. Rio Tinto took 51% of Salares Altoandinos alongside the state minerals company in July 2025. José Antonio Kast took office in March 2026 and the state-centric model **survived** — though most of it rests on presidential decree rather than legislation, and Chile has dismantled a lithium policy on a change of president before. **Argentina — the structural opposite.** Since the 1994 constitutional reform, **provinces own the subsoil**, so Salta, Catamarca and Jujuy each run their own regime and each has resisted congressional attempts to nationalise lithium. Milei's RIGI regime offers 30-year fiscal, customs and currency stability for projects over $200 million. By May 2026, **13 mining projects were approved under RIGI with over $42 billion committed**. Production grew **66% in a single year**, from 13,800 to 23,000 tonnes — the fastest growth in the world. **Bolivia — the largest single-basin deposit on earth, supplying under 1% of the world.** The state company YLB holds a statutory monopoly and acts simultaneously as developer, operator and de facto regulator, with no independent lithium regulator. **Bolivia produced about 2,462 tonnes LCE in 2025 against a 15,000 t/yr nameplate plant — roughly 16% utilisation.** Uyuni's problem is chemistry as well as politics: a high magnesium-to-lithium ratio and high rainfall make solar evaporation uneconomic, which is why Bolivia bet on direct lithium extraction. Its two big contracts — with Rosatom's Uranium One and with a CATL-led Chinese consortium, each around $1 billion — have **never been ratified by the legislature**, as the constitution requires, and were **suspended by a court in Potosí** following Indigenous complaints of absent consultation and missing environmental assessments. Neither has reached production. Rodrigo Paz took office in November 2025, ending twenty years of MAS rule, and is reviewing the contracts as "opaque" — but state-only extraction is constitutionally entrenched, so genuinely opening Uyuni needs a referendum he has not called. ## The water argument — where both sides overstate > [!warning] Handle this carefully; it is the most misrepresented claim in the transition-minerals literature > **What is agreed:** extracted brine is hypersaline, not potable freshwater, and is about 70% water by mass. The Atacama basin has been in prolonged drought, and very large **copper** operations there use far more *fresh* water than lithium does. > > **The "impact is small" evidence:** a 2025 hydrological study found average groundwater drawdown outside the salar nucleus of **13 cm**, with the net evaporation effect under 4% of basin-wide evaporation. Moran et al. (2022) found lithium extraction is not the cause of observed water-level change — meteorological drivers dominate, about a third of salar water is direct rainfall, mean groundwater age exceeds 65 years, lithium firms use under 10% of allocated freshwater, and regulators had **over-allocated** water on a flawed basin balance. *That study was funded by BMW and BASF. Disclose it.* > > **The "impact is real" evidence:** a 2022 PNAS study found a negative correlation between lithium extraction and Andean and James's flamingo abundance in the Salar de Atacama over about thirty years. The lagoon question is unresolved even in the reassuring literature. > > **The defensible statement:** peer-reviewed hydrology finds lithium brine extraction is **not the dominant driver** of basin-wide water loss — drought, copper and historical over-allocation matter more — while **localised drawdown in the salar nucleus and ecological effects on lagoon systems remain genuinely unresolved**, and much of the reassuring literature is industry-funded. Both "lithium is drying up the Atacama" and "lithium uses no water" are wrong. ## Consent — the variable that actually distinguishes the three All three states ratified ILO Convention 169. The FIDH finding across all three is that, except in exceptional cases, lithium operations have not been the subject of consultation with Indigenous peoples in an appropriate form and in good faith. - **Argentina** produced the strongest judicial win in the Triangle: in March 2024 Catamarca's Supreme Court **ordered the province to halt new permits** pending a cumulative environmental assessment, over the alleged disappearance of the Trapiche River. At Salinas Grandes, the Mesa de las 33 Comunidades drafted its own consultation protocol in 2015 and litigates on it. - **Chile** ran the first full-dress *consulta indígena* for a lithium deal in 2024–25, producing a package worth up to **$150 million a year to 28 communities through 2060**, against roughly $10–15 million previously. **And the four communities closest to the operations — Socaire, Peine, Toconao, Camar — suspended participation and physically disrupted the first session**, citing a methodology designed in under two months. - **Bolivia** is where consent was most clearly never sought — and where central-government control of the area has kept independent researchers out, so there is no independent field data on impacts at all. That contrast is the useful one. **Indonesia's Hongana Manyawa and Bolivia's Uyuni are cases where consent was never sought. Chile's Atacama is the case where a formal process ran and still split the affected communities.** Procedure is not the same as consent, and the vault should not treat a completed consultation as a settled question. ## See also [[Conflict Resources Atlas]] · [[Indonesian Nickel Boom]] · [[Critical Minerals]] · [[Artisanal and Small-Scale Mining]] · [[Conflict Minerals Hub]]