Modeling rare-earth and energy materials supply chains under theoretical China-outer-Mongolia political reunification scenarios

Wenxi Fang; Xi Liu

Modeling rare-earth and energy materials supply chains under theoretical China-outer-Mongolia political reunification scenarios

T0 review · 6 major / 3 minor · reviewed 2026-08-01 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A simulated China–Outer Mongolia union would cut rare-earth prices by about 15% and add $25–35 billion to Chinese welfare over a decade, provided the political event itself is costless.

desk verdict Interesting counterfactual, but the quantitative results are not computable from the model as written — the welfare and security numbers rest on undefined variables and the reported outputs are pre-announced. read the letter →

arxiv 2607.18019 v4 pith:BVCLHFAU submitted 2026-07-20 physics.comp-ph

Xi Liu , Wenxi Fang This is my paper · ORCID

classification physics.comp-ph MSC 91B7691A6591B55

keywords rareearthelementscriticalmineralsChina–MongoliaintegrationsupplychainmodelingStackelbergequilibriumexporttaxresourcesecurityinfrastructureinvestment

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tests what would happen to global rare-earth and energy-mineral markets if China and Outer Mongolia were politically reunified. It builds a ten-year dynamic supply-chain model in which Outer Mongolia supplies raw minerals, Baotou processes them, and the rest of the world buys the finished goods, with China acting as a Stackelberg leader that sets export taxes. Simulating a 'deep integration' scenario—full supply linkage, faster infrastructure investment, and an optimized export tax—the model predicts Outer Mongolian rare-earth output rising to 438×10³ tonnes, a 14.8% long-run price reduction, and $25–35 billion in cumulative Chinese welfare gains, while Rest-of-World consumers lose $8–15 billion. The paper also finds that a two-year railway delay would erase roughly 30% of those gains and cause persistent price inflation. The results matter because they quantify the resource-security stakes behind a politically charged hypothetical, but they rest on treating the reunification itself as a smooth, costless parameter change.

What carries the argument

The central object is a dynamic partial-equilibrium Stackelberg supply-chain model (Eqs. 1–13) with three geographic nodes: Outer Mongolian extraction, Baotou processing, and Rest-of-World residual demand. Its moving parts are a supply curve bounded by infrastructure capital, lagged capital accumulation with depreciation and investment efficiency, a Leontief processing function, profit-driven investment, market clearing, and an export-tax instrument that China optimizes as the Stackelberg leader. The deep-integration scenario is implemented as parameter shifts: Outer Mongolian supply share θ_i → 1, investment efficiency φ_i up 50%, depreciation δ_i down 20%, and the export tax made endogenou

What would settle it

A concrete check: add a one-time integration cost to the model—say a 5% GDP shock, a 10% tariff on Chinese exports, or a shift in RoW demand elasticity—and re-solve Eqs. (9)–(13). If the 14.8% price decline and $25–35 billion welfare gain do not survive a modest cost shock, the claim that full integration is welfare-positive fails. Alternatively, track actual Outer Mongolian rare-earth output and global prices through 2036 after the new railway opens; a sustained price path far above the model's deep-integration trajectory would contradict its predicted 438×10³-tonne output response.

Watch

Extended reading notes

Core claim

The paper's central discovery is a quantified mechanism: integrating Outer Mongolian mineral supply with China's Baotou processing hub would shift the global rare-earth market onto a lower, more stable price path. Under deep integration, the model's supply curve, infrastructure accumulation, and optimal export tax interact to raise Outer Mongolian output to 438×10³ tonnes within ten years, cut long-run equilibrium prices by 14.8% relative to baseline, lift China's supply-security index from 0.72 to 0.91, and generate $25–35 billion in discounted welfare gains for China, with $8–15 billion of surplus shifted away from Rest-of-World consumers. The authors present these as four core empirical f

Load-bearing premise

The load-bearing premise is that a China–Outer Mongolia political reunification can be represented by smooth, costless parameter shifts—full supply integration, 50% higher investment efficiency, 20% lower depreciation, and an endogenous export tax—with no sanctions, retaliation, capital flight, or change in world demand; the authors concede in the conclusion that the model abstracts from geopolitical retaliation, so if such costs are real, the 14.8% price reduction and $25–35

Editorial extensions

If this is right

If deep integration occurs as modeled, global rare-earth prices fall roughly 15% and stay lower, benefiting Chinese processors and downstream users while tightening the terms of trade against Rest-of-World manufacturers.
China's cumulative welfare rises by $25–35 billion over ten years (2015 prices), with about 60% of the gain attributed to reduced import-price volatility.
A two-year delay in cross-border railway construction cuts that welfare gain by about 30% and substitutes a 14.6% sustained price inflation for the price decline.
Rest-of-World consumers lose an estimated $8–15 billion in surplus as China tightens processed rare-earth exports and prices.
China's supply-security index for lithium and rare earths rises from 0.72 to 0.91, implying less exposure to external supply disruption.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A natural extension is to price the political event itself: if reunification carries sanctions, capital flight, or a demand-regime change, the model's price and welfare numbers should be read as upper bounds, since none of those channels appears in Eqs. (1)–(13).
The optimal export-tax formula (Eq. 18) makes a testable prediction for current Chinese policy: tax rates on unprocessed rare-earth exports should move with the rest-of-world demand elasticity, a relationship that could be checked against actual export-tax and quota data.
The same three-node structure could be reapplied to other cross-border mineral pairs (e.g., lithium corridors or copper) to see whether the 14.8% price effect is specific to rare earths or a generic property of supply integration under Stackelberg leadership.
The model's sharp sensitivity to infrastructure lag—30% of welfare gains lost to a two-year delay—suggests the infrastructure construction-time parameter is doing heavy lifting; targeted historical validation of rail-project delays would sharpen confidence.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

What's genuinely new is the application: no one has modeled a China–Outer Mongolia reunification supply chain for critical minerals, and the qualitative structure—extraction constrained by infrastructure, Leontief processing, Stackelberg export-tax leadership—is a sensible way to frame the question. The directional predictions are uncontroversial: deep integration expands supply and lowers price; delayed infrastructure does the opposite. The model itself is textbook dynamic partial equilibrium, so there's nothing novel in the mechanics.

The soft spots are load-bearing. The welfare objective in Eq. 10 sums processing revenue at prices P^Y that never appear in the market-clearing condition, the inverse demand curve, or the parameter table. That means the $25–35 billion headline is not a computable output of the model. Similarly, the supply-security index uses D^China, which is not defined anywhere. The 'derived' optimal export tax in Eq. 18 is asserted without a first-order condition, is dimensionally inconsistent, and the accompanying sentence says the tax rises with RoW demand elasticity while the formula has 1/η. The Monte Carlo design contradicts itself: 1,000 MVN draws in Eq. 14, 200 draws from triangular/normal/uniform in Section 4.1 and Figure 5, and ε_REE = 0.6 in the table versus μ=0.8 in Eq. 16. Several cited sources do not appear in the reference list. The paper claims four minerals but reports parameters only for rare earths. And the price decline is attributed to 'scale economies in Baotou's processing cluster,' but the model's Leontief technology is constant-returns; in the equations the price fall is just supply expansion against a fixed residual demand curve.

The pre-announcement in Section 4.1 is the clearest sign that the results are not independent findings. The 'expected outputs' listed there—10–15% price decline, $25–35B welfare gain, 30% reduction with delay—are exactly the numbers reported in Section 5. That's fitting, not prediction.

The paper does acknowledge that it abstracts away geopolitical retaliation, which is honest, but that concession doesn't address the internal issues. This is not a paper I would send to referees. The counterfactual is worth explaining to a colleague, but the quantitative claims should not be cited. If the authors re-specified the model with all variables defined, made the welfare computation explicit, and cleaned up the MC and references, it might become a solid applied paper. As it stands, I'd desk reject.

Referee Report

6 major / 3 minor

Summary. The paper constructs a dynamic partial-equilibrium Stackelberg supply-chain model for 2026–2036, linking Outer Mongolian mineral extraction, Baotou processing, and Rest-of-World demand, and compares three scenarios: baseline trade, deep Sino-Mongolian integration, and delayed infrastructure. Deep integration is implemented as parameter shifts (θ→1, φ+50%, δ−20%, endogenous export tax), with Monte Carlo sampling for uncertainty. The headline claims are that deep integration raises Outer Mongolian rare-earth output to 438×10^3 t, lowers long-run prices by about 14.8%, generates $25–35 billion in cumulative Chinese welfare gains, improves China's supply-security index from 0.72 to 0.91, and imposes $8–15 billion in RoW welfare losses. The paper concludes with policy recommendations on infrastructure, stockpiles, and export-tax design.

Significance. If the quantitative results were valid, the paper would quantify an important and underexplored supply-chain counterfactual and could inform policy debate. The qualitative direction of the price effects follows from the model's construction: more integrated supply and better infrastructure lower equilibrium prices, while delayed infrastructure raises them. The Stackelberg export-tax formulation is also a relevant policy tool. However, the central quantitative claims are not supported by the model as written. The welfare objective in Eq. (10) depends on processed-product prices P^Y that are never defined, calibrated, or solved; the supply-security index depends on an undefined D^China; the deep-integration parameter in Eq. (5) is reversed relative to its verbal definition; and §4.1 pre-announces the headline numbers before any results are derived. These are internal inconsistencies, not mere modeling simplifications. The manuscript does present its model structure transparently, but no reproducible code or tabulated results are provided, and the figures are not included in the submitted text.

major comments (6)

[§4, Eq. (5)] Equation (5) is internally inconsistent with its own definition and with the deep-integration scenario. The text defines θ_i as 'the share sourced from Outer Mongolia,' but the equation places θ_i on Q^IM (Inner Mongolia): Q^P = θ_i Q^IM + (1−θ_i) Q^M. Thus θ_i→1, which §4 says gives 'full integration of Outer Mongolian supply,' actually makes Baotou's raw-material mix 100% Inner Mongolian and zero Outer Mongolian. This reversed sign flips the main integration mechanism, and every quantitative result in §5 depends on it.
[§4, Eq. (10)] The welfare objective in Eq. (10) includes ∑_j P^Y_{j,t} Y^P_{j,t}, the revenue from processed products. P^Y_{j,t} is never defined, calibrated, or solved: it does not appear in the market-clearing condition (Eq. 9), in the inverse demand relation (Eq. 12), or in Table 1. Consequently W is not a computable function of the model's state variables, and the headline cumulative welfare gain of $25–35 billion (abstract; §5; §8) cannot be evaluated. The same issue affects the reported RoW consumer-surplus loss, whose components are not expressed in model variables.
[§4, Supply security index] The supply-security index defined as SSI_t = Σ_i ω_i (Q^M + Q^IM − D^China) / Σ_i ω_i D^China uses D^China_{i,t}, which is not defined anywhere in the model. It does not appear in Eqs. (1)–(13), is not in Table 1, and no calibration or data source is given. The abstract's claim that 'supply security index improves from 0.72 to 0.91' is therefore uncomputable as written. A precise definition of D^China and its mapping to model variables is required.
[§4.1] The numerical-algorithm section pre-announces the findings: 'expected outputs including a 10=15% price decline under deep integration, a 25-35 billion cumulative welfare gain for China, and a 30% reduction...' before any simulation results are reported in §5. Reporting these pre-specified numbers as computational findings makes the results non-independent. In addition, the Monte Carlo protocol is inconsistent: the surrounding text states 1,000 draws; the algorithm first says 'Repeat for 10^3 Monte Carlo draws' but later says '200 iterations'; and Figures 5–6 are described as '200 Monte Carlo sensitivity draws.' The uncertainty quantification is not reproducible.
[§8] The deep-integration counterfactual is imposed as smooth parameter perturbations (θ→1, φ+50%, δ−20%, τ^X free) with no cost, duration, or transition dynamics. The paper acknowledges in §8 that it 'abstracts away geopolitical retaliation risks' and ignores strategic stockpile responses, but this is not a peripheral limitation: the model contains no channel through which any negative consequence of the political event could enter. The headline magnitudes (14.8% price fall, $25–35 billion welfare gain, 0.72→0.91 SSI) are therefore conditional on an assumption the model itself does not test. The conclusion that integration is a 'cost-effective, high-impact strategy' (§8) is not supported within the model's scope.
[§5, Eq. (18)] The derived optimal export-tax formula is τ^{X*}_{i,t} = (1/η_i) · (D^RoW/(Q^M−Y^P)) · (1−∂C^P/∂Q^M). The accompanying text states 'the tax increases with RoW demand elasticity,' but for η_i>0 the formula decreases with η_i. Since the paper presents this formula as a tractable policy tool and calibrates η_i as a positive number, the sign error should be corrected. The standard inverse-elasticity property of a monopolistic markup would also imply a negative relationship.

minor comments (3)

[Table 1] Table 1 lists calibration values only for rare earths (and a coal reserve figure); no baseline values are shown for lithium or copper, nor for the full parameter vector (K_0, θ^0, ω_i, ΔS_i,t, A_i,0, etc.). Please provide a complete calibration table or a reference to an appendix.
[Figures 1–9] The manuscript describes nine figures, but the submitted text contains only figure captions/descriptions and no actual plots. If figures are to be part of the paper, the images must be included in the source so results can be inspected.
[General presentation] There are numerous typos and inconsistencies, including '10=15% price decline' in §4.1, 'UnderDeepIntegration' in §5, and the title's 'political reunification' framing versus the model's 'deep integration' parameter shifts. A careful proofread and terminology alignment are needed.

Circularity Check

2 steps flagged · score 6.0 of 10

Headline price, welfare, and delay findings are pre-specified as 'expected outputs' in §4.1 and then restated as results in §5; the welfare formula also depends on an undefined processed-product price P^Y, so the core claims reduce to their inputs.

fitted input called prediction [Section 4.1 (Numerical algorithm) 'expected outputs' paragraph; Section 5 (Computational results)]
"The model is calibrated using empirical data from USGS reserves estimates, Baotou processing capacities, and bilateral trade statistics, with expected outputs including a 10=15% price decline under deep integration, a 25-35 billion cumulative welfare gain for China, and a 30% reduction in welfare gains if infrastructure construction is delayed by two years. ... UnderDeepIntegration,rare-earthoxidepricesdeclineby10-15%relativetobaselineduetoscaleeconomiesinBaotou’sprocessingcluster.China’scumulativewelfaregainover10yearsisestimatedatUSD25-35billion(2015prices),with60%derivingfromreducedimportpr"

The 'results' in Section 5 and the abstract are numerically identical to the 'expected outputs' announced in Section 4.1 before any simulation output is shown. A calibrated model should produce outputs; here the algorithm's expected-output paragraph fixes the headline values (price decline, welfare gain, 30% delay penalty), and Section 5 restates them as findings. No mapping from Eqs. (1)-(13) to these numbers is exhibited, and the welfare number is not computable as defined (see next step). The claimed predictions are therefore the pre-specified inputs renamed as results.
other [Eq. (10) in Section 4 (China’s marketing power optimization problem); used for welfare results in Section 5 and §4.1]
"max_{Y^P_{j,t}, τ^X_{i,t}} W = Σ_t 1/(1+r)^t [ Σ_i P_{i,t}Q^M_{i,t} + Σ_j P^Y_{j,t}Y^P_{j,t} − Σ_i c^{ext}_i Q^M_{i,t} − Σ_j C^P_{j,t} + Σ_i τ^X_{i,t} P_{i,t}(Q^M_{i,t}−Y^P_{i,t}) ]"

The welfare objective W in Eq. (10) is the stated source of the $25-35bn cumulative Chinese welfare gain, but it depends on P^Y_{j,t}, the price of each processed product. That price is never defined, calibrated, or solved for: it does not appear in market clearing (Eq. 9), inverse demand (Eq. 12), the optimal export-tax condition (Eq. 18), Table 1, or any calibration narrative. Consequently ΔW = Σ(W_deep − W_base)/(1+r)^t is not a well-defined function of the model's state variables. The reported 'estimated' welfare gain is instead the pre-announced expected output, so the headline welfare result is not derived from the model.

full rationale

The central circularity is internal rather than citation-based. Section 4.1's algorithm description lists as 'expected outputs' exactly the numbers that Section 5 and the abstract present as simulation findings: a 10-15% price decline, a $25-35bn welfare gain, and a 30% welfare reduction from a two-year infrastructure delay. Section 5 restates those same values as results. The welfare objective adds a further problem: Eq. (10) requires processed-product prices P^Y_{j,t} that are absent from every other equation and from the calibration table, so the $25-35bn figure cannot be computed from the stated model. While the scenario mechanics (supply curves, capital accumulation, Leontief processing, export-tax formula) are internally coherent, they do not supply the headline quantities; those quantities are imposed as expected outputs and then reported as findings. Auxiliary outputs such as the supply-security index and the elasticity sensitivity have some independent content, so this is partial rather than total circularity. There is no load-bearing self-citation chain. Score 6: one or more headline 'predictions' reduce by construction to pre-specified expected outputs, with an underdetermined welfare formula.

Assumptions & free parameters 8 free parameters · 5 assumptions · 0 invented entities

The ledger is dominated by unstated calibration freedom: only the REE row of Table 1 exists, four of its five cited sources are missing from the reference list, parameters entering the Monte Carlo (γ, λ, ρ, A_0) have no reported base values, and variables entering the objective and the headline indices (P^Y, D^China) are undefined. The scenario comparisons are perturbations of this under-specified structure, so the reported numbers owe more to unshown inputs than to model structure.

free parameters (8)

α_REE (supply curve scale) = 0.75 tonnes/USD^0.5
'Author's estimate' (Table 1); sets the level of Outer Mongolian REE output, hence the output and price trajectories.
ε_i (supply price elasticity) = ε_REE = 0.6 (Table 1); MC: μ=0.8, σ=0.15; ε for Li/Cu/coal unreported
Dominant sensitivity driver per §8; central value differs between Table 1 and Eq 16.
η_i (RoW demand elasticity) = η_REE = 1.4 (Table 1); MC: μ=1.5, σ=0.25; others unreported
Sets price response (Eq 12) and the optimal tax (Eq 18); cited to Christmann (2024), absent from references.
φ_i (investment efficiency) = 0.05 tonnes/USD, 'Calibrated'
The +50% integration boost directly generates the higher-output/lower-price result; no data source given.
γ_i, λ_i, ρ_j, A_{i,0} (MC-sampled parameters) = unreported
Enter the MC distribution (Eq 15) with no stated central values; λ_i also enters investment Eq (3).
c_ext_i (extraction cost) = 2,500 USD/tonne for REE only
Cited to 'Baotou Daily (2025)', not in reference list; Li/Cu/coal costs unreported.
δ_i (infrastructure depreciation) = 0.08 yr^-1
Cited to 'World Bank (2024)', absent from references; the −20% integration reduction sharpens the capital-accumulation advantage.
K_{i,0}, κ_i, θ^0_i, ΔS_{i,t}, ω_i, P^Y_{j,t} = unreported/undefined
Initial capital, capacity caps, baseline trade shares, stockpile paths, SSI weights, and processed-good prices are never given; P^Y and D^China enter headline welfare/SSI numbers but have no equation defining them.

assumptions (5)

domain assumption Rest-of-World demand is a passive constant-elasticity residual demand curve (Eq 8) with η_i > 1, and the RoW does not respond strategically to China's export taxes.
Invoked in Eqs. (8)-(12): the entire price-formation mechanism reduces to a single inverse-demand curve per mineral; no competing exporters, no entrants, no strategic stockpilers are modeled.
domain assumption Market clearing (Eq 9) with no Chinese final demand for processed products; all output not absorbed by Chinese processing flows to RoW.
Eq. (9) omits any D^China_final term for processed goods; the welfare maximization (Eq 10) consequently rewards processing revenue regardless of Chinese domestic consumption, shaping the welfare numbers.
ad hoc to paper Political reunification is equivalent to smooth parameter perturbations: θ_i→1, φ_i+50%, δ_i−20%, τ^X free, with zero transition cost and unchanged technologies.
Declared in §4 'deep integration' scenario; if false (sanctions, capital flight, demand-regime change), every headline number changes by an unknown amount. Conceded only partially in §8.
domain assumption Calibration values in Table 1 (Gholami 2025, Christmann 2024, World Bank 2024, Baotou Daily 2025) are accepted as valid external estimates.
These sources do not appear in the reference list, so the numbers cannot be checked; ε_REE, η_REE, δ_i, c_ext_REE all rest on them.
standard math Functional forms are taken as structural: power-law supply in price and capital (Eq 1), partial-adjustment capital with construction lag (Eq 2), profit-proportional investment (Eq 3), Leontief processing (Eq 4), geometric energy-intensity decline (Eq 7).
Standard modeling choices; the paper neither derives nor tests them, but they are not on trial as mathematical claims — they are assumptions of the exercise.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Modeling rare-earth and energy materials supply chains under theoretical China-outer-Mongolia political reunification scenarios." pith.science (2026). https://pith.science/paper/BVCLHFAU

@misc{pith2026260718019,
  author       = {Pith},
  title        = {Pith review of: Modeling rare-earth and energy materials supply chains under theoretical China-outer-Mongolia political reunification scenarios},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BVCLHFAU}},
  note         = {Machine review of arXiv:2607.18019}
}

read the original abstract

Critical rare earth elements, lithium, copper, and coal underpin global clean energy transitions and advanced manufacturing, yet China faces persistent supply volatility and resource security risks amid fragmented cross-border mineral trade with Outer Mongolia. This paper constructs a dynamic partial equilibrium Stackelberg supply chain model spanning ten years, integrating three geographic nodes: Outer Mongolia's mineral extraction sector, Baotou's rare earth processing hub in Inner Mongolia, and residual demand from the Rest of the World (RoW). The model endogenizes core mechanisms including mineral supply curves constrained by infrastructure stock, lagged capital accumulation, Leontief processing production functions, profit-driven investment, and optimal export tax policy maximizing China's discounted social welfare. Three comparative scenarios are calibrated and simulated: a baseline status-quo trade framework, deep Sino-Mongolian resource integration, and a delayed cross-border infrastructure counterfactual.