(Art direction and design; illustration
created with AI: Jean-Dominique Lavoix-Carli)
When exploring how geopolitics and critical mineral supply chains — here, Lithium — may interact and impact each other, the first thing to do is to find out where the deposits of a critical mineral are located, how much resource is estimated within those deposits and, finally, which part of those resources can be estimated as reserves. This is the fundamental basis for analysis.
Now, when seeking the most recent assessment of the amount of Lithium resources and reserves worldwide, if we follow the data provided by the European Commission RMIS – Raw Materials Information System, we discover that the country holding the largest reserve of Lithium in the world would be Bolivia, with 39 million tonnes (data for 2023 — latest data available).

This would represent 57% of world reserves according to RMIS, 1.9 times the entire world reserves of 2023 according to the U.S. Geological Survey (USGS), Mineral Commodity Summaries, 2024, “Lithium” (pp.110–111, data for 2023), considered internationally as the “Gold Standard” for such information, and 1.05 times the entire world reserves of 2025, according to USGS 2026.
We are thus faced with a strange mystery. How much lithium does Bolivia actually hold as reserves? Do those reserves truly reach 39 million tonnes, or should we rather trust the USGS? How and why does the EC-RMIS reach such a figure for a mineral that is critical for the EU, notably used in Lithium batteries necessary for Electric Vehicles and electrification? Should all actors involved in the Lithium supply chain rush to Bolivia? Or are the reserves given by the EC-RMIS simply a mistake?
This is what we shall investigate in this article. In the first part, we explain why it matters, in geopolitical terms, to know which actors produce which knowledge and what status that knowledge holds in the world. Along the way we present the USGS, the major actor in the production of knowledge on mineral resources and reserves worldwide (and thus on lithium), and the European Commission RMIS – Raw Materials Information System, alongside other European actors.
In the second part we focus on the mystery itself — the discrepancies between the official figures given by the European Commission and those of the USGS — and seek to explain where the difference could come from. Along the way we highlight elements that should be strengthened to ensure decision-makers have access to better knowledge on this critical mineral.
Geopolitics for Lithium 2040: Eight Scenarios for the Upstream Supply Chain
Art © Jean-Dominique Lavoix-Carli [2026] –
jdlavoixcarli.comMost reports on critical minerals acknowledge that geopolitics matters. This one focuses on what this actually means — through eight scenarios built on a causal model of 383 variables and 1,528 relationships, using structural analysis (Godet, revisited with network centrality measures). Read more…
Official sources, knowledge production and geopolitics
Why the source of knowledge matters
As political scientists, we do not transform ourselves into geologists or resource-and-reserve specialists. We rely on serious assessments provided by official sources and reputable experts.
Which reserves of Lithium — or of any other critical mineral or resource, for that matter — a country holds is key to any political and geopolitical analysis pertaining to the Lithium supply chain.
Which reserves our own country holds, compared with the reserves held by other countries, in the context of demand and international competition for the resource, is critical for a government, as it determines its strategy regarding Lithium and its exploitation, through both domestic and international actors.
Which reserves exist in the world, and what the status of our influence and diplomatic relations is with each country holding reserves — compared with the influence and relations of other countries also seeking Lithium — is key for any state needing Lithium.
These calculations are then complicated by the international situation, geology, technology, the capital necessary for mining and extraction, the specifics of the supply chain, and so on. Yet for government officials and strategists, building a first picture of who holds what is the foundation.
Because this vision will determine what governments do regarding Lithium, international relations analysts, business strategists, and firms involved in the supply chain — from mining to capital investment — must also understand it.
This means that we must understand who officially says what and who sets the norms for knowledge.
The Architecture of the Critical Minerals Competition: Mapping Geopolitical Risk
Art © Jean-Dominique Lavoix-Carli [2026] –
jdlavoixcarli.comThis strategic foresight report based on a 161-variable Causal Graph Model uses centrality measures to reveal the levers shaping the future of supply chain and how the system can structurally evolve. Read more…
The IEA, the USGS, and the making of normative knowledge
For Lithium, the first two official sources to consult are the USGS for reserves and production, and the International Energy Agency (IEA) for supply and demand. For another mineral such as uranium, for example, we would also need to take into account the “recognised global reference on uranium,” jointly prepared by the Nuclear Energy Agency (NEA) and the International Atomic Energy Agency (IAEA) — the “Red Book.”
As far as critical mineral reserves are concerned, the IEA uses USGS figures (see for example, IEA, Global Critical Minerals Outlook 2025 p.211), as well as for its regional outlooks, alongside the USGS, S&P Global (2025), and the European Commission (2025), RMIS – Raw materials’ profiles (IEA, Ibid. p.259). We do not know, however, which source corresponds to which piece of information.
The IEA, as an international organisation, represents a consensual body of knowledge shared by the international order of states and thus holds normative power. This normative power is in turn granted to the sources it uses, since that usage amounts to a kind of international validation.
As a result, USGS knowledge becomes the standard and the norm, accepted and recognised even by international agencies. Meanwhile, the USGS is also part of the US administration. Thus, irrespective of the quality and validity of its work, when the USGS shares knowledge with the world, this is knowledge held by the US — still, for now, the most powerful state in the world.
Knowledge, sovereignty, and national interest
- The Strange Mystery of Bolivia’s Lithium Reserves
- Beyond Geopolitics: Understanding the Critical Pressures Reshaping Our World.
- WATCH OUT – The Iran War, Saudi Arabia, and the Houthis
If we remain at the international and state levels that concern us primarily here in terms of geopolitics, the actors producing knowledge should ground it in their own capabilities: their experts, their geological services, reputable companies of their own nationality and, ideally, without overseas ties. If a country wants to be sovereign and independent, it cannot depend solely and exclusively on knowledge produced by other countries or by entities from other countries. The absence of ties to foreign territory, capital, or interest more broadly is essential to ensure that the knowledge produced — the fundamental building block for policy — is not tainted, even involuntarily, by others’ national interest. That knowledge must remain independent of outside influence.
When one uses USGS data, then, despite its reputation, we must keep its origin in mind. Of course, a power such as the US must produce data that are as accurate as possible, lest it lose influence in international knowledge production. Nonetheless, it is always useful to bear in mind where knowledge originates.
For example, the USGS defines itself as follows:
Created by an act of Congress in 1879, the USGS provides science for a changing world, which reflects and responds to society’s continuously evolving needs. As the science arm of the Department of the Interior, the USGS brings an array of earth, water, biological, and mapping data and expertise to bear in support of decision-making on environmental, resource, and public safety issues.
Among other things, it produces the knowledge upon which the entire US policy on mineral resources is built — from the list of critical minerals to all related domestic and foreign American action, from trade and diplomacy to possibly more aggressive moves such as Trump’s vision on Greenland, and perhaps one day invasion to secure resources.1e.g. Hélène Lavoix, “The American Threat against Greenland: When Should It Have Triggered a Watch?“, the Red Team Analysis Society, 21 January 2026. If the USGS were wrong one day, the entire strategy and policy of the US would be skewed and, ultimately, the US would fail to secure its national interest.
Thus, even though other states should, for reasons of state and national interest, always bear in mind the origin of figures shared by the USGS, the national and international obligations placed on the USGS mean that its process for producing knowledge may itself be considered a standard, or at least a useful example.
For instance, in the USGS Mineral Commodity Summaries appendix, where all terms are defined and sources for the data are given — for example in “Part B—Sources of Reserves Data” — we find :
“National information on reserves for most mineral commodities found in this report, including those for the United States, is derived from a variety of sources. The ideal source of such information would be comprehensive evaluations that apply the same criteria to deposits in different geographic areas and report the results by country. In the absence of such evaluations, national reserves estimates compiled by countries for selected mineral commodities are a primary source of national reserves information. Lacking national assessment information by governments, sources such as academic articles, company reports, presentations by company representatives, and trade journal articles, or a combination of these, serve as the basis for national information on reserves reported in the mineral commodity sections of this publication…”
Mineral Commodity Summaries 2024, appendix ,”Part B—Sources of Reserves Data”, p. 210.
We now have a picture of the normative production of knowledge on the resources and reserves of critical minerals, and an ideal vision of what this could, or should, look like for a single state.
The EU’s fragmented knowledge production on critical minerals
As far as the European Union is concerned, a list of Critical Raw Materials and one of Strategic Materials were set by the 2024 Critical Raw Materials Act. A review of those lists is due by 24 May 2027 and every three years thereafter. The European Commission RMIS – Raw Materials Information System, led and maintained by the Joint Research Centre (JRC), is in charge of those lists, and notably produces raw material profiles, country profiles, and factsheets.
Alongside this, we find another initiative, SCRREEN 3, coordinated by the French Geological Survey (BRGM), a EU consortium comprising “16 partners and 8 affiliated entities from 12 countries” financed through the European Union’s Horizon Europe Research and Innovation Programme under Grant Agreement N° 101138060, which supports the review and validation of Critical Raw Materials data. It is “running from January 2024 to December 2027, with a budget of €2.99 million”.2Call for Experts for the 2nd SCRREEN3 Validation Workshop.
As an indication – although a direct comparison is impossible, since we would need to consider the share of the JRC budget devoted to minerals plus, possibly, the budget of every relevant agency in each member state – US President Trump requested $892.7 million for the USGS for fiscal year 2027, while the enacted budget for the agency in 2026 was $1.420 billion.3“The U.S. Geological Survey (USGS): FY2027 Budget Request“, EveryCRSReport.com, 16 April 2026.
SCRREEN participates – or should participate – in the Raw Materials Information System (RMIS), as explained at a 2024 JRC / European Health and Digital Executive Agency (HaDEA) Technical Workshop “Channeling knowledge from European projects into the Raw Materials Information System (RMIS)” on 9th – 10th October 2024.
Indeed, in the top tier of the RMIS factsheet webpage for Lithium, we find a link giving access to the SCRREEN criticality factsheets of Lithium.
The very existence of this link suggests that two distinct factsheets on Lithium exist for the European Union. How, or whether, the data from one feeds into the other is not made transparent.
The EC-RMIS could be considered a compiler of data, but not a producer of knowledge on a par with the USGS. In future, part of that role may fall to the Geological Service for Europe (GSEU), or a similar structure — yet these remain “only” projects, “funded by the European Union,” and thus carry no official character.
One may wonder whether the multiplication of projects, programmes, networks, and consortia under a “European Union” umbrella is conducive to the best possible use of resources, or to an optimal strategy.
Let us now turn to our initial mystery — the official European Commission statement on Bolivia’s Lithium reserves, and on world Lithium reserves more generally — compared with USGS knowledge.
Explaining the Gap Between EC-RMIS and USGS Figures
Discrepancies and outdated data
For lithium, according to USGS 2026 public data, we have the following figures, as displayed on the charts below:

More specifically, Bolivia’s resources reach 23 million tonnes, and there are no reserves. We may assume that, when the USGS collected its data, Bolivia was not yet exploiting any lithium mine, and thus had no reserves.
Now, if we look at the EC-RMIS Lithium profile, the first chart shows the top five holders of Lithium reserves in the world as Bolivia 50%, Australia 12%, China 10%, Argentina 8%, and other 20%. EU reserves reach 1.2%. No year is given for this data. If we download “About the data – Dashboard” and look at the chapter for Lithium, we read: “Reference Year: 2022… Latest RMIS Dashboard update: November 2024.” Let us compare those figures with the USGS 2026 figures for 2025:



The figures are radically different, notably for Bolivia. The RMIS “reserves” correspond neither to USGS reserves nor to USGS resources. The USGS notes that resources were considerably revised worldwide in 2025, so some of the discrepancy may stem from that revision, and from earlier ones. Meanwhile, the first data presented by the RMIS date from 2022 — hardly comparable to the far more recent USGS 2026 Critical Summary.
This alone shows that the three-year revision cycle set out in the EU Critical Raw Materials Act is far too long. It implies, as the charts show, that EU policymakers and other decision-makers relying on EU data are working from outdated figures — which could carry serious consequences when dealing with critical and strategic materials. And if the discrepancy is already known, the RMIS data discarded and the USGS is treated as the real standard, then EU taxpayer resources are being spent producing data that goes unused.
For its part SCRREEN 3 is carrying out its own revision of its Critical Raw Materials factsheets, with the latest round taking place over the summer 2026.4Call for Experts for the 2nd SCRREEN3 Validation Workshop. However, a look at the draft version of the Lithium factsheet (SCRREEN/2025) shows that, for reserves and resources, it uses USGS data published in 2025 – i.e. 2024 figures. It states: “Global measured and indicated lithium resources reached approximately 115 million tonnes, according to the latest USGS data.”5Ibid. The newer USGS 2026 report instead reads: “Owing to continuing exploration, measured and indicated lithium resources have increased substantially worldwide and total about 150 million tons.”
Thus, despite the effort involved, the newly synthesised knowledge is already outdated before it is even published. As with the EC-RMIS dashboard, this points to a serious problem for national interest, strategy, and policymaking.
Bolivia’s Reserves Under Closer Scrutiny: The 2023 Data
Below the first charts on the RMIS dashboard, Lithium reserves are broken down and presented as bubbles. Bolivia’s reserves are shown crushing all others, as shown in the first screenshot.
Here, a date is finally given: 2023. Downloading the specific information sheet for this chart, looking at the item “Lithium”, 2023 is indeed the reference year.
We should therefore compare this 2023 data with the USGS 2023 data, published in the 2024 report:
For 2023, the discrepancies are even larger than those observed above. Comparing the figures country by country, we obtain the following result:



No obvious pattern emerges — there is no systematic over- or under-estimation that might point toward an explanation.
For its part, SCRREEN3, across its various Lithium factsheets, uses USGS data for resources and reserves. The forthcoming edition, relying on USGS 2025, like the 2023 factsheet relying on USGS 2022 (thus 2021 data), shows Bolivia with 23 million tonnes of resources and no reserves (tables 9 and 10).
Let us try to unpack the EC-RMIS data.
What does the EC-RMIS mean by “reserve”?
What resources and reserves actually are is well codified in the mining world, even though different systems exist. Specialists working across those systems are aware of the differences, specify which system they use, and make an effort to keep their data consistent across countries.
Internationally, the system evolved out of the interactions between three frameworks: the American USGS Circular 831, the Australasian Code for Reporting of Exploration Results, Mineral Resources and Ore Reserves (the JORC Code), and the CIM Definition Standards for Mineral Resources & Mineral Reserves (interested readers may find detail on the historical contribution of each here).
The resulting common CRIRSCO standard has now been adopted by many countries’ mining industry representatives, which is not the same as state’s entities. That standard is defined as below:

What must be noted is that moving from resources to reserves requires a number of factors to be satisfied, demanding evaluation and expertise by a “competent person.” These modifying factors include: mining and metallurgy (do we have the technology to separate lithium from magnesium?), infrastructure (are there roads, water, and electricity to run the plant?), law and environment (does the company hold the necessary environmental permits?), and society and government (does the local community or the sovereign state allow extraction?).
As far as the USGS is concerned, definitions of resources and reserves are still governed by USGS Circular 831, which differs from the CRIRSCO standard:

Resources are categorised as “Measured and Indicated” and grouped together under the term “Demonstrated.” Reserves are explicitly defined as the portion of those “Demonstrated” resources that is economically extractable at the time of determination.
Unfortunately, “economically extractable” (USGS) does not map exactly onto the “modifying factors” of the CRIRSCO standard.
As for the EU-RMIS, the two available information sheets tell us that “the mineral reserves estimate provided by (S&P, 2024) are mostly based on CRIRSCO reporting standards (last update May 2024).” The word “mostly” is worrying, since it suggests no real effort at standardisation was made, in contrast to the USGS’s approach. This may be a first source of discrepancy.
In any case, the EC-RMIS deals in reserves, not resources. Given the different logic of the two systems — CRIRSCO and USGS — some discrepancy between figures is to be expected. But since the USGS remains the gold-standard reference for resources and reserves, we still need to compare against it.
At first glance, the CRIRSCO modifying factors should be stricter than the USGS’s “economic viability” test. We would therefore expect all RMIS reserve figures to be smaller than USGS ones — which is not the case, as shown below (we have also kept the USGS resource figures for reference, in that case RMIS figures should always be smaller, thus appear in negative in the chart below):

The most striking incoherence concerns Bolivia’s reserves. They reach 39 million tonnes according to the EC-RMIS, yet do not exist at all for the USGS (production has not yet started, or may be about to, with a Chinese and a Russian firm involved).6Muflih Hidayat, “Bolivia Lithium Development: Massive Reserves Meet Production Challenges”, Discovery Alert (Mining News), 24 October 2025. Comparing RMIS’s Bolivian reserves with the USGS’s Bolivian resources (23 million tonnes), the RMIS figure exceeds the USGS one by 15 million tonnes — not a small margin.
While the discrepancy is most striking for Bolivia, very significant gaps also appear across most other figures. Since none of them are explained, it becomes effectively impossible to rely on the EC-RMIS for Lithium reserves at all.
It seems likely that the RMIS did not standardise the data it collected across countries, nor check the plausibility of its results against USGS findings.
Whose Data Is It? The Sources Behind EC-RMIS
The RMIS relies entirely on S&P 2024 for its data — that is, S&P Global’s Global Mining Intelligence platform. (We contacted S&P Global to ask whether they could help explain the differences, but had not received a reply as of publication.)
This is a surprising choice. First, how can a quasi-state entity rely entirely on a market-based one? The logics and interests involved are not the same. Second, how can an entity representing sovereign, independent states rely on an American private firm, given the various tensions between the EU and the US — tensions that predate the election of President Trump.
The USGS, by contrast, does not rely exclusively on data from a commercial firm, however reputable, even an American one.
To hand the production of state knowledge over to a foreign firm is to give up sovereignty and independence.
The EU could instead have been expected to compile and standardise world data on critical minerals sourced from its own member states — assuming such data exists at all.
Conclusion
It is impossible to rely on the 2024 EC-RMIS data for Lithium reserves, even though these are the Commission’s official statistics.
The difficulty of making sense of the gap with the USGS figures unfortunately discredits the entire factsheet as potentially unreliable — even where other figures in it may well be correct (we did not check them). No sensible decision-maker or researcher will spend the time needed to compare and verify every figure in the factsheet, unless specifically tasked to do so.
Worse, the problem identified here for Lithium reserves could also cast doubt on the entire RMIS effort for other Critical Minerals, should other major unexplained inconsistencies come to light.
In terms of influence, this reflects poorly on the European Commission and on the European Union as a whole.
Furthermore, where Lithium is concerned, if decision-makers ground strategies, policies, and actions in RMIS data, the consequences could be serious.
Other EU-funded initiatives and projects rely, for world data, on the USGS. While this may be the only realistic option, it is also a risky one, should the US one day withdraw or alter its USGS publications.
Policy recommendations
The European Union should ensure that the official data it provides is correct, and should allocate sufficient resources to this task on a permanent basis.
All European institutions and member states should, as rapidly as possible, take every measure necessary to ensure the official knowledge produced is valid.
Given how essential they are, data on critical minerals should be revised yearly, not every three years.
Efforts should be made to regain Europe’s independence in knowledge production, as this is key to effective action.
The European Parliament (EP) – and concerned Committees – should make sure the 2024 EU Critical Raw Materials Act is updated accordingly.
Notes
- 1e.g. Hélène Lavoix, “The American Threat against Greenland: When Should It Have Triggered a Watch?“, the Red Team Analysis Society, 21 January 2026. ↩︎
- 2
- 3“The U.S. Geological Survey (USGS): FY2027 Budget Request“, EveryCRSReport.com, 16 April 2026. ↩︎
- 4
- 5Ibid. ↩︎
- 6Muflih Hidayat, “Bolivia Lithium Development: Massive Reserves Meet Production Challenges”, Discovery Alert (Mining News), 24 October 2025. ↩︎































