Space Resource Economies

A space resource economy would be a system in which material extracted off Earth is produced, priced, sold and consumed. The framing under test is that an off-world resource economy has a market structure.

It does have one, and naming it is this brief's job. The structure is monopsony: one class of buyer, buying for policy reasons rather than for use, at prices set to establish a legal precedent rather than to clear a market. The total realised revenue from off-world resources in the whole of human history is $25,001, and the buyer was NASA. That is a sharper and more damning fact than the claim this page previously carried — that no off-world resource has yet been sold — and it is the correct one.

This brief is the market structure and the economics: who buys, at what price, whether the transaction clears, and what the two competing break-even analyses actually assume. Space Law and Governance owns the treaty regime — the Outer Space Treaty and non-appropriation, the Artemis Accords and their contested reading, the national resource statutes, safety zones and the bloc split. Whether a mined kilogram can lawfully be sold is unsettled and is taken up there; this page does not restate it. Asteroid Mining owns how you get the material; Lunar Industry owns lunar extraction; this page owns whether it sells. The one legitimate crossing is the regolith purchase itself, because it is simultaneously a legal instrument and the only transaction in the market: reported here as revenue and price, and there as precedent.

Established On 3 December 2020 NASA selected four companies to collect lunar regolith and transfer ownership on the lunar surface, and the prices are the entire revenue history of the sector. Lunar Outpost, $1. ispace Tokyo, $5,000. ispace Europe (Luxembourg), $5,000. Masten Space Systems, $15,000. Total: $25,001. The terms are unusual and deliberate: the companies collect samples, send back images and location data, and — in NASA's words — “Upon receipt, NASA will pay the companies and take ownership,” with ownership transferring while the sample is still on the Moon.

Established The stated purpose was legal, not industrial. NASA's aim was to demonstrate that “space resources can be mined and owned without violating the Outer Space Treaty’s prohibition on claims of national sovereignty,” establishing “the policy principles of space resource collection and ownership.” ispace described its contract as “the world’s first commercial transaction of lunar resources.” Frontier The price was not set by supply and demand. It was set to make a legal point, and a dollar is not a market-clearing price for anything.

Frontier The second transaction is contracted and undelivered, and it has the same shape. Interlune announced on 7 May 2025 an agreement with the US Department of Energy Isotope Program for 3 litres of lunar helium-3, delivery deadline April 2029, at “approximately today’s commercial market price”; plus a commercial agreement with Maybell Quantum for “thousands of litres” annually from 2029 to 2035, for dilution refrigerators cooling quantum computers below 10 mK. Interlune's own price reference is about $3,000 per litre, against market reports around $2,500. Its funding is an $18 million seed plus $4.84 million from the Texas Space Commission, plus DOE, NASA TechFlights and NSF SBIR Phase I awards, and its Vermeer-built full-scale prototype excavator, unveiled the same day, is designed to process 100 tonnes of regolith per hour. Mission sequence: Crescent Moon, Prospect Moon, Harvest Moon. A company on its own prospects — interested throughout. Established But note the structure the deal reveals: the anchor customer is again a US federal agency, and the commercial customer is buying an isotope for cryogenics, not a fuel.

Established There is essentially one buyer, and the evidence comes from three directions. First, the regolith purchase: buyer NASA, purpose legal precedent. Second, CLPS, where NASA is the customer and the model is described by an independent tracker as competitively financing commercial companies to build a commercial ecosystem around lunar exploration — with a delivery record of Astrobotic's Peregrine (8 January 2024, failed on a propellant leak, reentered 18 January); Intuitive Machines IM-1 Odysseus (22 February 2024, success, landed tilted); IM-2 Athena (partial, tilted with limited power); and Firefly's Blue Ghost Mission 1 (launched 15 January 2025, soft landing in Mare Crisium on 2 March 2025, success). Third, and most telling: the one serious NASA cost study of lunar in-situ propellant assumes no commercial market at all.

Established That third point deserves its own sentence. Jones, Pensado, Clark, Grande, Ivanco, Judd, Klovstad and Reeves, in NASA's 2019 break-even analysis of lunar in-situ propellant production, examine a NASA-only customer base with a multi-year lunar surface presence and multiple crewed Mars missions. No commercial market is considered — not as an oversight, but because none exists to model.

Established And the anchor customer is expensive and politically exposed. NASA's Office of Inspector General reported in November 2021 that Artemis will cost $93 billion across FY2012–FY2025 at $4.1 billion per SLS/Orion launch, warning that “without capturing, accurately reporting and reducing the cost of future SLS/Orion missions, the Agency will face significant challenges to sustaining its Artemis program in its current configuration.” Frontier A monopsony whose own auditor calls it unsustainable is a fragile demand curve, and that is the honest structural observation.

Frontier The most useful thing this page can set out is a live, specifiable disagreement between two economic analyses whose authors' interests run in opposite directions. The optimistic pole is Metzger's Economics of In-Space Industry and Competitiveness of Lunar-Derived Rocket Propellant. His framework uses a gear ratio on cost and a production mass ratio — kilograms of propellant per kilogram of capital landed. Baseline production mass ratio 167; tent sublimation 442–534, which he calls an order of magnitude better than the competitiveness threshold even against delivery to LEO; strip mining 3.7–43.4, near the threshold; gear ratio 6–15 depending on architecture. His conclusion is that lunar propellant gains an absolute cost advantage at geostationary transfer orbit by year 5–7 and possibly at LEO by year 15–23, and “will be commercially viable and should lower the cost of doing everything else in space” — explicitly contradicting sceptics who say falling launch costs eliminate lunar competitiveness.

Established Now the assumptions that result rests on, because they are the disagreement. Metzger assumes terrestrial launch falling from $2,000/kg to $30/kg over thirty years — a 67-fold reduction — and an optimistic demand scenario of 436,000 tonnes a year of LEO capacity, with moderate and pessimistic cases at 10% and 1% of that; finance at a discount rate falling from 21.7% to 12% by year 30. Frontier Against ESA's measured ~1,500–2,000 tonnes actually injected to LEO in the record year 2025, the optimistic demand case is about 200 times the entire world's current annual launch mass. Metzger is a long-standing advocate of lunar in-situ resource utilisation; interest runs with the finding.

Established The conservative pole is NASA's own 2019 analysis, on the same subject with the opposite temperament. Break-even there depends on in-situ system lifetime (1 to 5-plus years), lander inert mass fraction (0.20–0.30), campaign duration (5–15 years) and demand rate, with launch costs taken as SLS Block 1B at $1,000 million per launch and commercial heavy-lift at $200 million. Their finding: “the magnitude and duration of the lunar campaign, more so than the Mars campaign, drive the breakeven,” and critically, “without long lifetime ISRU systems, with greater than 5 years of autonomous operation before replacement, the demand in cis-lunar space for a Mars campaign favors propellant delivery from Earth.”

Established Setting them side by side is the point: the disagreement is not about physics. It is about three things — how fast launch prices fall, how long an autonomous mining plant survives on the Moon, and whether a demand curve exists above the government's. Frontier All three are unmeasured. That is the state of the market-structure question, stated honestly, and neither side is being unreasonable within its own assumptions.

Established The oldest statement of the condition is still the sharpest, and it belongs here as well as next door. Sonter in 1997: resource recovery is profitable “given an in-space market of some thousands of tonnes per year, in competition against Earth-launch costs of several hundred dollars per kilogram.” Established Neither condition holds in 2026: the in-space market for extracted material is zero tonnes, and launch is $1.4k–2.7k/kg.

Established The cost-to-orbit dependency the whole sector leans on is weaker than assumed, and a NASA analyst says so about a NASA talking point. Jones gives the history in consistent units: launch cost to LEO dropped from a 1950s high approaching $1,000k/kg; from 1970 to 2010 it was usually $10–20k/kg, with the Shuttle much higher at $62k/kg; Falcon 9 brought it to $2.7k/kg and Falcon Heavy halved that to $1.4k/kg. And the load-bearing sentence: “Launch cost is roughly 10 to 30% of total mission cost.” His supporting comparison has commercial satellites at about $120k/kg against commercial launch at about $12k/kg, so launch is only 10% of payload cost. Frontier A tenfold reduction in launch price therefore improves total mission economics by 9 to 27%, not by tenfold — which does not close a gap measured in orders of magnitude.

Frontier Is there a buyer above the government? This is the field's central open question and every other one is downstream of it. The regolith purchase, CLPS, the Interlune DOE contract and NASA's own modelling assumption all point the same way. The most-cited prospective private customer is orbital manufacturing — and Space-Based Manufacturing's own finding is that no named non-government volume customer appears anywhere in its evidence base. Two hypothetical markets do not make one real one.

Frontier Does lunar propellant beat Earth launch, and when? Metzger says year 5–7 at geostationary transfer orbit and possibly year 15–23 at LEO, on a 67-fold launch price fall and 436,000 t/yr of demand; NASA's 2019 analysis says Earth delivery wins without ISRU plants surviving more than five years autonomously. The unmeasured variables are the same three every time: launch price trajectory, plant lifetime, and demand above the government's.

Frontier How long does an autonomous plant last on the lunar surface? This is the single quantity that flips the NASA result, and there is no data point anywhere: the longest-running resource-production experiment off Earth is MOXIE, which ran sixteen times over two and a half years on Mars. An answer would be worth more to this brief than any economic model.

Frontier Is helium-3 an isotope business or an energy business? These are separated by five to six orders of magnitude in mass flow and the distinction is almost never made. The isotope market is real, tiny and has a government buyer; the fusion market requires a reactor type that does not exist for a fuel cycle harder than the one that also does not exist. Conflating them is how the concept survives.

Frontier What kind of market is a market with fewer than ten viable sites? Elvis, Krolikowski and Milligan find that the Moon “presents finite and scarce areas with rare topography or concentrations of resources of special value … Typically, there are fewer than ten key sites of each type, each site spanning a few kilometres across,” and that diverse actors pursuing incompatible ends there “could soon crowd and interfere with each other, leaving almost all actors worse off.” Speculative The economics that implies is not a commodity market but site-specific rivalry with first-mover exclusion — closer to spectrum or port-berth allocation than to mining. The authors frame it as crowding and interference rather than explicit monopoly, and this page reports it that way.

Speculative Would legal certainty create a market? It is commonly asserted and no operator in the retrieved record names it as the reason they failed. Planetary Resources' policy director named the customer base; AstroForge names demand when declining the water and depot market; NASA's own study models a government-only buyer because there is no other. The legal question is real and belongs to Space Law and Governance; promoting it to a cause here is not supported by the evidence.

Frontier And a revealed preference that cuts against the field's consensus product. Water and propellant are named as the defensible near-term product by Sonter, Ross, Sercel, Crawford and Metzger alike. AstroForge, the only firm currently flying hardware, explicitly declines that market for want of demand. An operator refusing the product the literature recommends is a market datum, not an engineering one.

Handwave The position at the edge of the space, stated because leaving it out would be worse. The popular claim is that space contains effectively infinite resources and therefore ends scarcity. Established Access, energy and cost bind off Earth exactly as they do on it, and the measured record — $25,001 of revenue and 122 grams of oxygen produced on another world — is what abundance looks like when it has to be delivered.

Established The bottleneck is demand, and it is named as such by the people who ran out of money. Peter Marquez of Planetary Resources: there is “no customer base for asteroid mining in the next 12 to 15 years” — said in 2019, so the window closes in 2031–2034. No property regime, launch price or drill design creates a buyer.

Established Autonomous plant lifetime is the technical bottleneck that decides the economics. NASA's analysis turns on whether an in-situ system runs more than five years without replacement. Nothing has ever produced a resource off Earth for more than a few hours of cumulative operation: MOXIE totalled 122 grams of oxygen across sixteen runs.

Established Launch price is not the bottleneck and treating it as one is the sector's standard error. Launch is 10 to 30% of mission cost. The entire fall from $62k/kg to $1.4k/kg — a factor of forty-four — did not produce the resource industry it was supposed to unlock.

Frontier Concentration of viable sites is a bottleneck of an unusual kind. Fewer than ten key sites of each type, each a few kilometres across, means capacity is bounded by geography rather than by capital, and that early actors can foreclose later ones by occupying rather than by outcompeting.

Established Capital formation has already failed twice at the firm level. About $50 million into Planetary Resources including a $21.1 M Series A and a €25 M Luxembourg package, wound down with assets to ConsenSys on 31 October 2018; $3.5 million into Deep Space Industries before acquisition and redirection. Frontier Here that record is market-structure evidence rather than engineering evidence: roughly $53 million of venture capital subsequently chased a market whose own policy director dated at twelve to fifteen years away.

Established And a bottleneck in the evidence itself, which this page has to declare. The OECD's space-economy figures were blocked at every attempt during research, so no total for the size of the space economy is asserted anywhere on this page; the industry-participant lunar propellant business case could not be fetched either. A sector whose headline market size cannot be cited from a fetched source is one where the honest move is to give transactions instead of totals.

Established The dependency on Space-Based Manufacturing is the most-cited prospective customer relationship, and that customer refuses the role. Its own rewrite finds its market may be structurally small precisely because orbital manufacturing works so well, that the sector is still substantially on public seed capital with more than twenty-four projects and about half SBIR-derived, that NASA's programme exists to move products that will ultimately be sold to customers on Earth, and that no named non-government customer buying at volume appears anywhere in its source base. This page owns the supply side of that feedstock question and has to report that its principal named customer has no volume market either.

Established Deep Space Infrastructure is the delivery half of any resource transaction. Depots, transfer stages and cislunar logistics are how a produced kilogram reaches a buyer. This brief owns the price and the buyer; that one owns the pipe, and neither is useful without the other.

Established Lunar Industry owns extraction; this page owns whether it sells. The permanently shadowed inventory numbers belong there; the dollars per kilogram and the customer belong here. Where the two touch is the excavator rate, and it is worth stating: Interlune's prototype is designed for 100 tonnes an hour, which is a production figure that only means something once there is a price.

Established Asteroid Mining supplies the firm-level failure record, and it is read differently here. There it is evidence about engineering and capital formation; here it is evidence about market structure. The same paragraph is deliberately not printed twice.

Established Space Law and Governance answers “may you sell it?”; this brief answers “to whom, at what price, and does the transaction clear?” Frontier And one correction runs back across that seam: legal certainty should not be claimed as the binding constraint on the industry, because the constraint the operators name is absence of a customer, and no property regime creates one.

Frontier A measurement dependency this page cannot satisfy from any fetched source: the size of the space economy. The standard totals are published behind access controls that refused every attempt during research. Everything quantitative on this page is therefore a transaction, a contract or a cost estimate rather than a market total, and that is a deliberate consequence of what could be verified.

Established The experiment that has actually been run is a purchase, and it worked as designed. Four contracts, $25,001, ownership transferring on the lunar surface. It was constructed to test a legal proposition rather than an economic one, and as a legal experiment it produced a result: nobody has successfully challenged it.

Established The only resource-production experiment ever conducted off Earth is MOXIE, and its scale is the most useful number in this brief. It produced 122 grams of oxygen total across sixteen runs between February 2021 and its final run on 7 August 2023, at a peak rate of 12 grams per hour and 98% purity or better — twice its target hourly rate. Frontier Put it beside OSIRIS-REx's 121.6 grams returned from Bennu. The entire history of resource production on another world and the entire history of asteroid material delivered to Earth are both about 122 grams.

Frontier The decisive economic experiment is a lifetime test, and it is a surface mission rather than a study. An in-situ plant operating autonomously for more than five years is what NASA's own break-even analysis turns on. Nothing has run for more than a few hours cumulatively, and the difference between four years and six years of plant life reverses the published conclusion.

Frontier Interlune's Harvest Moon would be the first end-to-end extraction and return of a saleable off-world commodity. Preceded by Crescent Moon, a hyperspectral camera, and Prospect Moon, a lander. The DOE contract that would be satisfied is 3 litres of helium-3 by April 2029 — about 0.4 grams — which is a delivery test rather than a market test, and the company says the price is approximately today's commercial rate.

Speculative The experiment nobody has run is a standing purchase commitment at a stated price. An offer to buy water or propellant delivered in orbit at a published rate would reveal immediately whether the demand curve exists above the government's. The nearest real-world instance is the $25,001, and it was a legal instrument wearing a commercial costume.

Established And the delivery record that constrains every schedule here. Of four CLPS landings attempted: one total loss on a propellant leak, one success landing tilted, one partial with limited power, one clean soft landing. A supply chain with a fifty per cent full-success rate at the landing step is not yet a supply chain, and every resource contract on the books depends on it.

Established The engineering requirement that actually prices this market is autonomous endurance, not throughput. NASA's break-even turns on more than five years of autonomous operation before replacement; Metzger's turns on production mass ratios of hundreds of kilograms of propellant per kilogram of capital landed. Both are statements about how long hardware survives on a surface nobody has operated on for a season.

Frontier Throughput requirements, where they exist, are enormous and specific. Interlune's excavator is designed for 100 tonnes of regolith per hour. Established Derived from Crawford's figures, supplying 10% of projected 2040 world electricity from lunar helium-3 requires processing 500 km² of the highest-concentration regolith to 3 m depth every year, which at about 1,500 kg/m³ is roughly 2.25 billion tonnes a year — about 2.6 million excavator-years annually at that machine's rate. Both inputs are published; the multiplication is this brief's.

Established The isotope requirement, by contrast, is small enough to be plausible, and the contrast is the lesson. Helium-3 gas is about 0.1346 g/L at standard conditions, so 1 kg is about 7,430 litres, worth about $22 million at Interlune's $3,000/L reference. At 20 ppb, 1 kg of helium-3 needs about 50,000 tonnes of regolith — roughly 500 hours of a 100 t/hr excavator. The DOE's 3-litre contract is about 0.4 grams, requiring about 20 tonnes of regolith. These derivations are this brief's, from Crawford's concentration figures and Interlune's published rate and price.

Established Energy inputs are the part of the helium-3 case that is usually omitted. Heating regolith from about −20 °C to about 700 °C consumes about 5% of the fusion energy obtained, assuming no heat recycling; Crawford puts end-to-end efficiency at “could exceed 50%, and may be much less.” Advocates' claims of 60–70% direct conversion obscure extraction, collection, purification and transport.

Established The positional advantage that makes any of this worth attempting is quantified and it is real. Earth's gravity well is about 6,400 km of equivalent depth against the Moon's 290 km — a 22-fold energy advantage for delivering material to Earth orbit from the lunar surface rather than from the ground. That is the entire physical basis of the sector, and it is sound.

Frontier And the products where the economics could work are named and narrow. Propellant, with oxygen dominating the LOX/LH₂ mass budget; geostationary solar power satellites; and titanium and aluminium for orbital construction, for which lunar lithophile concentrations exceed asteroid abundances. Crawford's own conclusion is that it is difficult to identify any single lunar resource sufficiently valuable to drive an extraction industry by itself.

Established Space Law and Governance owns the treaty regime and this page uses the pointer rather than restating it. The 1967 Outer Space Treaty and non-appropriation, the Rescue, Liability and Registration instruments, the dead-letter 1979 Moon Agreement, the Artemis Accords and their contested reading that extraction is not national appropriation, the national resource statutes, safety zones, the Accords-versus-ILRS bloc split and the equity question between spacefaring and non-spacefaring states all belong there. Whether a mined kilogram can lawfully be sold is unsettled and is taken up in that brief.

Established The one legitimate crossing is the December 2020 regolith purchase, because it is two things at once. This page reports it as revenue and price; that page reports it as precedent. Neither should carry the other's framing, and the number that belongs here is $25,001.

Established Asteroid Mining is the supply side for the non-lunar route, and it supplies this page with the strongest single piece of demand evidence in the corpus: an operator declining the water and depot market for want of buyers, and a policy director dating the customer base at twelve to fifteen years out. The extraction technique is theirs; the revealed preference is ours.

Established Lunar Industry and Moon-Based Manufacturing hold the production half. Extraction chemistry, ice prospecting, dust and power belong to the first; forming and fabricating objects to the second. What this brief adds to both is the observation that a produced kilogram with no buyer is inventory, not revenue.

Frontier Commercial Fusion is the adjacency that the helium-3 story depends on entirely, and it does not hold. Deuterium–tritium fusion is not yet demonstrated for power generation, so deuterium–helium-3 is speculative twice over, and Crawford notes the reduced-neutron advantage is “often exaggerated.” The isotope market for cryogenics and neutron detection is entirely independent of that and should never be argued from it.

Frontier Deep Space Infrastructure is where a market, if one appeared, would physically clear. Depots make propellant a tradable commodity by giving it a place to sit; without them a delivery is a rendezvous. Metzger's competitiveness result assumes that infrastructure exists; NASA's conservative result assumes campaign-driven delivery instead, and the difference in assumed logistics is a large part of the difference in conclusion.

Established The institutional design of this sector is unusually explicit: a government created the market in order to create a legal fact. The four regolith contracts transfer ownership on the lunar surface for a total of $25,001, and NASA said in advance that the purpose was to establish the policy principles of space resource collection and ownership. That is industrial policy conducted through procurement, and it worked as policy.

Established CLPS is the same instrument at larger scale. An agency competitively financing commercial companies to build a commercial ecosystem, with a landing record of one loss, one success, one partial and one clean soft landing. The programme's stated theory is that a government customer bootstraps a private one; the evidence that a private one has appeared is, so far, absent.

Frontier Public capital is doing most of the work on the private side too. Interlune's disclosed funding is an $18 million seed plus $4.84 million from the Texas Space Commission plus DOE, NASA TechFlights and NSF SBIR Phase I awards; its first contract is with a federal isotope programme. A company on its own prospects is an interested source, and the structure of its balance sheet is a fact about the market.

Established The anchor institution is under audit pressure from its own inspector general. $93 billion through FY2025, $4.1 billion per launch, and a warning that the agency faces significant challenges to sustaining Artemis in its current configuration. The agency's own auditor against the agency's programme is interest running against the finding, which is why it carries weight here.

Frontier The institutional question the concentration result raises is allocation, not ownership. Fewer than ten key sites of each type, each a few kilometres across, with crowding and interference leaving almost all actors worse off, is a coordination problem that markets do not solve by themselves. The instruments belong to Space Law and Governance; the observation that the relevant model is closer to spectrum allocation than to mineral rights belongs here.

Established And a declaration about what this page cannot say. The standard institutional source for the size of the space economy was blocked at every attempt during research, and the industry-participant lunar propellant business case could not be obtained. No space-economy total appears anywhere on this page, and nothing from that business case is cited. A brief that quoted either from memory would be doing the thing it criticises.

Frontier The distributional question is the live one, and the concentration result sharpens it. If there are fewer than ten key sites of each type, each a few kilometres across, then early access is close to exclusive access, and the actors able to reach them first are a small set of states and firms. The equity argument between spacefaring and non-spacefaring states belongs to Space Law and Governance; what belongs here is that the market structure makes it acute rather than theoretical.

Established Crawford's environmental argument is the strongest ethical case for the sector and he flags its complexity himself. If ecosystem services were properly priced into terrestrial extraction, off-world sources become comparatively attractive, and “obtaining raw materials from uninhabited ones may come to be seen as ethically preferable.” Speculative The comparison is currently untestable because there is no delivered off-world cost to compare against.

Established The same author supplies the counterweight in one sentence: “one would not go into space for a source of metals for use on Earth.” Terrestrial ores and infrastructure are abundant; the value of off-world material is positional. Which means the environmental argument, if it ever applies, applies to a substitution that is not currently on offer.

Frontier There is a public-communication ethics problem in this subject with a measurable cost. Helium-3 as an energy resource has attracted policy attention and capital for four decades on a concentration figure — 4 to 20 parts per billion — that is almost never stated alongside the claim. Established Restoring the concentration figure changes the argument completely, and doing so is the single most useful thing a page like this can do.

Frontier Interest marking is unusually load-bearing here and this page applies it consistently. Metzger is an in-situ resource advocate and his result is favourable; Interlune is a company on its own prospects; NASA's inspector general is an auditor against its own agency; Crawford is a lunar-science advocate whose review deflates the headline lunar resource. The pattern is that the deflating results come from people with reason to want the opposite, and they are weighted accordingly.

Speculative And a quieter question about public money. A sector sustained by government procurement, SBIR awards and state space commissions is being funded on the expectation of a commercial market that its own most careful analyses do not model. Whether that is sensible industrial policy or a subsidy in search of a customer is a judgement this page does not make; it notes that the record so far is $25,001 of revenue.

Established The civilisational claim is that off-world resources remove the mass constraint on everything humans do in space, and the measured base for it is 122 grams. MOXIE's total oxygen production on Mars, and OSIRIS-REx's returned Bennu sample, are the same order of magnitude. Every projection in this subject starts from there.

Frontier The positional argument is the durable one and it is physically sound. A 22-fold energy advantage in delivering material to Earth orbit from the lunar surface rather than from the ground is a permanent feature of the solar system. What is missing is not the physics but somebody standing in orbit wanting the material.

Speculative If a market ever forms, the concentration result says what shape it takes. Fewer than ten viable sites per type turns a resource economy into a rivalrous allocation problem from the first day — not after depletion, as on Earth, but immediately. Space scarcity would arrive before space abundance.

Frontier The helium-3 story is the clearest case in this corpus of a civilisational claim surviving on a missing number. Four to twenty parts per billion; 2.25 billion tonnes of regolith a year for a tenth of world electricity; best deposits lasting about 2,000 years at that rate and about 200 years if all demand were met. Established And Crawford's closing comparison: covering the same lunar area with solar panels yields as much electrical energy in seven years as all the helium-3 beneath it, and at equal 20% conversion efficiency solar out-produces it in 1.4 years — continuously and renewably, against a finite deposit.

Handwave “Space is infinite resources, so scarcity ends” is the standing civilisational claim and it fails the same way every other claim here does. Frontier The binding constraints off Earth are the ones that bind on it — access, energy and cost — and the record of the sector so far is one government buyer, four sellers who have delivered nothing, and a price chosen by a lawyer.

These horizons track contracts and landings, because those are the only things in this sector with dates attached:

• 10 yr: Frontier Interlune's DOE contract falls due in April 2029 — 3 litres of helium-3, about 0.4 grams — preceded by Crescent Moon, Prospect Moon and Harvest Moon; its Maybell Quantum agreement runs 2029–2035 for thousands of litres a year. Established Delivery on any of the four 2020 regolith contracts would be the first physical performance of a resource sale in history. Frontier Marquez's twelve-to-fifteen-year window from 2019 expires in 2031–2034 and nothing in the record since contradicts him. Speculative No commercial volume customer for propellant is expected to appear, on the evidence of every source on this page.
• 25 yr: Frontier This is the horizon on which Metzger's result would resolve: lunar propellant competitive at GTO by year 5–7 and possibly at LEO by year 15–23, conditional on launch falling to $30/kg and demand reaching 436,000 t/yr — about 200 times the world's 2025 launch mass to LEO. Frontier NASA's 2019 analysis reaches the opposite conclusion on plausible plant lifetimes, and the disagreement is settled by data nobody has: how long an autonomous plant lasts. Speculative A helium-3 isotope business at thousands of litres a year is the plausible commercial outcome at this range; a helium-3 energy business is not.
• 50 yr: Speculative If in-space demand ever reaches Sonter's thousands of tonnes a year, this is the earliest horizon on which it could, and it depends on crewed and infrastructure programmes rather than on anything resource-specific. Handwave Any figure for the size of an off-world resource economy at this range is an assertion — and this page declines to give one even for the present-day space economy, because the standard source could not be obtained.
• 100 / 250+ yr: Handwave Beyond forecasting. Speculative The defensible structural statement is that the physics is permanently favourable — a 22-fold energy advantage for lunar material delivered to Earth orbit — and that everything else in this brief is a statement about demand, which is a fact about what else humanity chooses to do in space rather than about resources at all.

• Depends on Five edges, and they divide into the two halves of a transaction plus the rule-set it happens under. Space-Based Manufacturing is the most-cited prospective customer and refuses the role: no named non-government customer buying at volume appears anywhere in its evidence base. Deep Space Infrastructure is the pipe — depots and transfer stages are what make a produced kilogram a tradable one. Lunar Industry and Asteroid Mining are the two supply routes, and both hand this brief the same finding from opposite directions: their operators name absence of a customer, not technique or law, as what stopped them. Space Law and Governance answers may you sell it; this brief answers to whom, at what price, and does it clear. The one place they overlap is the December 2020 regolith purchase, which is a legal instrument there and $25,001 of revenue here.
• Requires (not on this map) Three constraints that are not briefs on this map, and the first is the whole subject. The market constraint is a buyer: every transaction in the sector's history has a government at one end, NASA's own break-even study models a NASA-only customer base because none other exists to model, and the most-cited prospective private customer reports no volume buyer in its own evidence. The industrial constraint is the quantity that decides the published disagreement — NASA finds Earth-delivered propellant wins without in-situ plants surviving more than five years autonomously, and nothing has ever produced a resource off Earth for more than a few hours cumulatively. The institutional constraint follows from concentration: with fewer than ten key sites of each type, each a few kilometres across, crowding and interference could leave almost all actors worse off, and that is an allocation problem closer to spectrum than to mineral rights.
• Enables No enabling edge is claimed and none would be honest. A functioning resource economy would change the mass budget of everything done in space — but total realised revenue is $25,001, the only production experiment off Earth made 122 grams of oxygen, and the sector's most careful cost study models a NASA-only customer base because there is no other one to model. An edge from that to anything downstream would record an intention.
• Adjacent Commercial Fusion is adjacent only through the helium-3 claim, and the adjacency does not hold: deuterium–tritium power generation is not demonstrated, so deuterium–helium-3 is speculative twice over, and the isotope market for cryogenics and neutron detection is independent of it. Moon-Based Manufacturing is the downstream use for lunar feedstock. Mars Colonization is where MOXIE's 122 grams were produced, and where the only measured demand for in-situ resources — a crew's own consumables — would actually sit.

Established “No off-world resource has ever been sold.” This page said that, and the precise version is worse. One has been sold, four times, on 3 December 2020 — $1 to Lunar Outpost, $5,000 each to ispace Tokyo and ispace Europe, $15,000 to Masten, a total of $25,001 — with ownership transferring while the sample was still on the lunar surface. Frontier The buyer was a government, the purpose was to establish the policy principles of space resource collection and ownership, and the price was chosen to make a legal point. “Never sold” understates the problem; “sold for twenty-five thousand dollars to prove a legal proposition” states it.

Handwave “Lunar helium-3 will power the world.” The concentration is about 4 ppb by mass, at most about 10 ppb measured, possibly 20 ppb in the most favourable maria — and that figure is almost never given alongside the claim. Established Supplying 10% of projected 2040 world electricity requires processing 500 km² of the best regolith to 3 m depth every year, which is about 2.25 billion tonnes annually — roughly 2.6 million excavator-years at Interlune's 100 t/hr prototype rate. The best deposits last about 2,000 years at that rate; meeting all demand exhausts accessible reserves in about 200. Frontier And Crawford's killer comparison: covering the same lunar area with solar panels yields as much electrical energy in seven years as all the helium-3 beneath it, and at equal 20% conversion efficiency solar out-produces it in 1.4 years — continuously, against a finite deposit.

Frontier “But Interlune has helium-3 customers, so the fusion case must be working.” These are two different claims separated by five to six orders of magnitude in mass flow, and this brief keeps them apart. Established The isotope market is real, tiny and government-anchored: helium-3 is a byproduct of tritium decay separated at the Savannah River Site, thousands of litres are made available annually, demand peaked at about 70,000 litres in 2008 and projected federal demand is now under 6,000 litres a year after recycling and alternative technologies; a White House inter-agency group was created to allocate supply during the 2008 shortage. Frontier Derived from published figures: 1 kg of helium-3 is about 7,430 litres, worth ~$22 million at $3,000/L, and needs about 50,000 tonnes of regolith at 20 ppb — roughly 500 hours of a 100 t/hr excavator. The DOE's 3-litre contract is about 0.4 grams, or about 20 tonnes of regolith. As an isotope play the mass flows are plausible. As a fusion-fuel play they are not, and conflating them is how the concept survives.

Speculative “Helium-3 fusion is nearly here.” Deuterium–tritium fusion is not yet demonstrated for power generation, so deuterium–helium-3 is speculative twice over — a reactor type that does not exist, for a fuel cycle harder than the one that also does not exist. Established Crawford notes the reduced-neutron advantage is often exaggerated, and that advocates' 60–70% direct-conversion claims obscure extraction, collection, purification and transport costs.

Established “Cheap launch is the unlock.” Launch is 10 to 30% of total mission cost, so a tenfold price fall improves total economics by 9 to 27%. Frontier The fall from $62k/kg to $1.4k/kg has already happened — a factor of forty-four — and the resource industry it was supposed to unlock did not appear. This is a NASA analyst deflating a favourite NASA talking point, which is interest running against the finding.

Frontier “The economics are settled and lunar propellant wins.” Metzger's result assumes launch falling to $30/kg and a demand scenario of 436,000 tonnes a year to LEO — roughly 200 times the world's entire 2025 launch mass. Established NASA's own 2019 analysis reaches the opposite conclusion on plausible plant lifetimes: without in-situ systems surviving more than five years autonomously, Earth-delivered propellant wins. The disagreement is about three unmeasured quantities — launch trajectory, plant lifetime, demand above government — and this page declares the tie rather than averaging it.

Speculative “Property-rights uncertainty is why the first-generation firms failed.” Their own people said otherwise. Planetary Resources' policy director named the absence of a customer base for twelve to fifteen years; AstroForge declines the water and depot market on demand grounds; NASA's cost study models a government-only buyer because none other exists. Established The legal question is real and belongs to Space Law and Governance. No property regime creates a customer, and it should not be promoted to a cause here.

Frontier “The space economy is worth $X hundred billion, so resources have a market.” Two problems. The first is a category error: satellite services and launch are not off-world resource markets. Established The second is evidential and this page declares it: the standard source for space-economy totals was blocked at every attempt during research, so no such figure appears anywhere on this page. The verifiable numbers here are transactions: $25,001 realised, 3 litres contracted for 2029.

Handwave “Space is infinite resources, so scarcity ends.” Frontier Concentration says the opposite arrives first: fewer than ten key sites of each type, each a few kilometres across, with actors pursuing incompatible ends liable to crowd and interfere and leave almost all of them worse off. The relevant model is spectrum or port-berth allocation, and scarcity would show up before abundance did.

Established “We are already producing resources off Earth at meaningful scale.” The entire history is 122 grams of oxygen from MOXIE across sixteen runs, at a peak of 12 g/hr and 98% purity, ending 7 August 2023. Frontier That is about the same mass as OSIRIS-REx's 121.6 g of Bennu. Both totals are the whole record of their kind, and both fit in a hand.

Frontier And the framing itself. “An off-world resource economy has a market structure” implies buyers, sellers and a clearing price. Established What exists is one buyer, four sellers who have each delivered nothing, and a price chosen by a lawyer. Frontier The structure is monopsony, the anchor customer's own auditor calls its programme unsustainable at $93 billion and $4.1 billion a launch, and the most-cited prospective private customer reports no volume buyer of its own. That is a market structure. It is simply not the one the phrase is usually used to suggest.