2014P_ / Codex / Radical Markets

Radical Markets. Mechanism design, and the parts that survive the stress test.

Weyl and Posner attempt to generalise Henry George's analytic move into a programme of twenty-first-century mechanisms. The Codex includes the book critically: Quadratic Funding earns its place; COST and immigration-sponsorship do not.

Codex · Western Canon · Read critically · ≈13 min read · Posner & Weyl, Radical Markets · 2018
TL;DR

The apparatus is worth learning. The proposals survive the stress test unevenly. Glen Weyl and Eric Posner's Radical Markets (2018) is the most sustained twenty-first-century attempt to generalise Henry George's analytic move into a programme of mechanism-design proposals. The ancestry is explicit. The execution is mixed. Quadratic Funding is in production now at Gitcoin and elsewhere and is a genuine contribution to commons-funding engineering. Quadratic Voting is elegant and operationally useful in bounded settings. COST (Common Ownership Self-assessed Tax) would turn every possession into a continuous auction and structurally destroy the rooted relationship to land, home, and inherited place that Pañca Ṛṇa exists to honour. The immigration-sponsorship and data-as-labour proposals are commodification moves that the book's own egalitarian commitments cannot adequately answer. The Sāmatvārtha view: keep the mechanism-design apparatus, accept what survives the Pañca-Ṛṇa stress test, refuse the commodification, and route the wider programme through commons-first instruments rather than through ever-finer market mechanisms. The book belongs in the Codex because it forces the line to be drawn explicitly.

The ancestry — Henry George at the mechanism-design lab

The intellectual lineage is the first thing to know. Weyl is an economist who has moved over the past decade from neoclassical microeconomics through mechanism design into the territory he and Audrey Tang now call Plurality. Posner is a law professor at the University of Chicago. Their starting point in Radical Markets is explicitly Georgist: the recognition that a great deal of the dysfunction of contemporary capitalism is the dysfunction of monopoly rent — returns captured by titleholders on assets (land, network position, regulatory privilege, data) whose value was created by parties other than the titleholder. The Georgist remedy, the Land Value Tax, addresses the land case directly. Weyl and Posner ask: can the same analytic move generalise?

The answer, in their hands, is a programme of five mechanism-design proposals. Each takes a contemporary economic relation that the standard market-and-state apparatus handles badly, and proposes a redesigned mechanism intended to dissipate the rent and route the value more equitably. The proposals are intellectually serious. They are also, individually, of very different merit, and the book's effect of bundling them together as a single programme has done some damage to the reception of the parts that genuinely work.

What survives the stress test — Quadratic Funding

Quadratic Funding (QF), formalised in Buterin, Hitzig and Weyl's 2018 paper "Liberal Radicalism", is a matching-fund mechanism for public goods. The mechanic, simplified: the match a project receives scales with the square of the sum of square roots of individual contributions. The practical consequence is that a project supported by many small contributors receives disproportionately more match than a project supported by a few large contributors, even when total contribution amounts are similar.

The intuition is breadth-of-support detection. A flat matching mechanism cannot distinguish ten thousand people each giving five dollars from ten people each giving five thousand dollars; QF can, and credits the former pattern as a stronger signal that the project is a genuine public good rather than a private interest dressed as one. The formal property the mechanism satisfies — optimal funding under specific assumptions about preference distribution — is genuine; the live deployments have produced meaningful funding flow to commons projects (Gitcoin Grants since 2019 and related Ethereum public-goods experiments; Optimism's Retroactive Public Goods Funding is adjacent but not itself QF).

The known weaknesses are equally real. Sybil attacks: a single party splitting into many fake-identity contributors can exploit the breadth-of-support signal. Collusion: coordinated groups can amplify their own funding by coordinating small contributions across many wallets. Capture by well-organised constituencies: subcommunities that are coordinated enough to flood QF rounds can extract disproportionate match against more diffuse public-good recipients. Each of these is being addressed in production (proof-of-personhood mechanisms, collusion-resistant variants, scoring functions that down-weight suspicious patterns), but the mechanism is not the policy-design free lunch its enthusiasts sometimes claim.

Net assessment for the Sāmatvārtha frame: QF is a useful and largely Pañca-Ṛṇa-compatible mechanism. It funds the knowledge-commons and software-commons (Ṛṣi Ṛṇa discharge), it routes capital toward genuinely public goods rather than rent-extractive private goods, and the known failure modes are engineering problems rather than philosophical ones. The Codex endorses QF as a serious instrument for the commons-funding toolbox alongside the Techno-Memetic Commons licence and the federated-unicorn architecture.

What survives partially — Quadratic Voting

Quadratic Voting (QV) is the collective-decision counterpart of QF. Voters are given a budget of vote-credits which they can allocate across issues; spending n credits on an issue produces √n votes on that issue. The non-linear scaling allows expression of intensity of preference (a voter who cares deeply about one issue can spend most of their credits on it, sacrificing influence elsewhere) while discouraging extreme concentration (each additional vote costs disproportionately more credits).

QV is most useful in bounded settings: corporate governance, DAO decisions, internal team prioritisation, civic-technology pilots, citizen assembly structures. Colorado Democratic legislators used QV-like mechanisms in 2019 for prioritising budget items. Several DAO communities use QV in production for protocol decisions. In these bounded settings the mechanism is a genuine improvement over one-person-one-vote when the decisions to be made are multi-issue and the intensities of preference are heterogeneous.

The structural critique applies at scale rather than at mechanism. QV makes preference intensity legible and influential; preference intensity is correlated with wealth, attention, free time, and political identity in ways that flat voting partially counteracts. A society that adopted QV as its main electoral mechanism would, in most realistic political-economy scenarios, give greater relative voice to organised, well-resourced minorities at the expense of broad-but-shallow majorities. Whether that is desirable depends on which questions one is trying to resolve; for some kinds of decisions it is, for others (basic rights, distributional fundamentals) it almost certainly is not.

Net assessment: QV is a useful instrument inside its proper scope (bounded decisions, multi-issue prioritisation, internal governance). It is not a replacement for democratic majoritarianism in the high-stakes constitutional register, and it should not be marketed as one. The Codex carries it in the mechanism-design toolbox with the scoping note attached.

What does not survive — COST

COST (Common Ownership Self-assessed Tax, sometimes called the Harberger tax after the economist Arnold Harberger who first proposed it) is the proposal that fails the Pañca-Ṛṇa stress test most clearly. Every property owner publicly posts a self-assessed value of their property. The assessment is the tax base — owners are taxed at a fixed percentage of their declared value. The assessment is also the standing-offer price: any willing bidder can buy the property at the declared price, and the owner must sell.

The mechanism is analytically elegant. It cleanly solves the property-assessment problem (owners reveal true value because over-declaring increases tax burden while under-declaring increases risk of forced sale). It dissipates monopoly rent on illiquid assets in a single stroke. It would in principle improve allocative efficiency by routing assets to their highest-value users continuously. As a piece of theoretical mechanism design, it is a beautiful result.

The Sāmatvārtha critique is structural and follows immediately. Under COST, every possession is permanently for sale. The household lives in continuous defensive auction against any willing bidder. The home one's grandparents built. The ancestral farm. The workshop one has tooled to one's particular craft over decades. The village land that the community has tended for centuries. The bookcase one's father made. All of it sits under standing offer, at all times, to anyone with the capital to bid above the owner's tax-minimising assessment.

This is not the dissipation of monopoly rent that Georgism targets. It is the dissipation of rooted relationship to land, home, inheritance, and place — precisely the substrate that Bhūmāta and Pitra Ṛṇa exist to honour, the substrate whose erosion Scott documents in his case studies of villagisation, and the substrate whose preservation is a load-bearing component of any coherent civilisational accounting framework.

The defenders of COST argue that the tax rate could be set low enough that displacement is rare. But low rates undermine the rent-dissipation argument, and any rate high enough to dissipate rent is high enough to make rooted possession structurally precarious. The mechanism is internally coherent and externally toxic to substrates the model does not include in its objective function.

The crucial point: the diagnosis COST attempts to address is correct. Monopoly rent on illiquid assets is structural injustice. The standard Land Value Tax — calibrated annually by professional assessors, no standing-sale obligation — captures the same monopoly rent without the dispossession dynamics. Georgism's original instrument survives the Pañca-Ṛṇa stress test; the COST extension does not. The book bundles them together; the Codex disentangles them.

Take the rent on the land. Leave the rootedness of the home.
The two are not the same target.

What does not survive — Data as Labour, Migration as Service

The data-as-labour proposal argues that users whose data trains AI systems should be paid for their contribution — a wage relation rather than the current free-input relation. The diagnostic motivation is correct (uncompensated value-creation by users captured by platform titleholders) and overlaps cleanly with Mazzucato and Zuboff. The proposed instrument is the problem: marketising data contributions converts the relational substrate of everyday life — the conversations, the photos, the purchase patterns, the location traces — into a transactional surface, with structural pressure to expand the surface (more sharing produces more income) and to make the surface auditable (you cannot pay for what you cannot measure). The cure is the same disease.

The Sāmatvārtha alternative is structural rather than transactional. The Techno-Memetic Commons licence refuses enclosure of substrate contributed by users; the federated-unicorn architecture distributes returns to contributors via ownership rather than via micro-payment; data protection regimes (GDPR-style rights, ownership-of-self models) constrain what can be extracted in the first place. Data is not labour because it is not an alienable product of effort; it is a relational trace of being-in-the-world, and treating it as labour is the marginalist sleight-of-hand Mazzucato warns against in a different register.

The VIP immigration-sponsorship proposal — allowing ordinary citizens to financially sponsor migrants in exchange for a share of the migrant's future earnings — is the case where the book's own egalitarian commitments are weakest. Citizenship becomes a sponsorable asset; the migrant enters the receiving polity as a financial obligation to a private citizen-sponsor rather than as a person under public protection; the relationship between sponsor and sponsored carries structural pressure toward asymmetric dependency. The motivation (expand legal migration channels, route the gains-from-migration equitably) is legitimate. The instrument converts citizenship and personhood into commodities in ways the book's own values cannot answer, and the proposal has been roundly criticised on these grounds from across the political spectrum.

Where the programme is going — Plurality, after the book

The post-2018 trajectory of the work is itself important to the assessment. Weyl has moved noticeably from the mechanism-design-as-cure framing of Radical Markets toward the Plurality framework he now develops with Audrey Tang. Plurality reframes the project around protocols for collaboration across difference, digital-democracy infrastructure, and the recognition that purely market-mechanism solutions to substrate problems are insufficient. The community around Plurality overlaps substantially with the commons-funding and digital-public-goods movements; the language of rent-dissipation through market design is much less load-bearing than it was in Radical Markets.

This is the relevant context for reading the 2018 book in 2026. The authors themselves have moved past some of its proposals. The mechanism-design contributions (QF, QV in bounded settings) are now production instruments in the civic-technology and commons-funding stack. The substrate-commodifying proposals (COST, data-as-labour in transactional form, migration-sponsorship) have not aged well and are not load-bearing in the post-book Plurality programme. The Sāmatvārtha read aligns straightforwardly with the post-book trajectory: take the engineering, leave the commodification.

The neighbours — and where the diagnosis thickens

  • Henry George, Georgism: the analytic ancestor. The land-rent specific case from which the Weyl/Posner programme generalises. Georgism's original instrument (LVT) survives the Pañca-Ṛṇa stress test cleanly; some of the Radical Markets generalisations do not.
  • Mariana Mazzucato, The Value of Everything: the parallel modern generalisation of the value-creation-vs-extraction distinction. Mazzucato preserves the public-investment frame Radical Markets partially abandons in favour of pure mechanism design. The two programmes overlap diagnostically and diverge in instrument philosophy.
  • Audrey Tang & Glen Weyl, Plurality: the post-book trajectory that tempers the market-mechanism enthusiasm with explicit commons and collaborative-protocol infrastructure. The project Sāmatvārtha is closer to than to Radical Markets itself.
  • Elinor Ostrom, Commons Governance: the empirical and theoretical refutation of the premise that markets-or-states is the only available choice for resource governance. Ostrom's design principles for actual commons institutions are the structural alternative Radical Markets under-engages with and the Codex centres.
  • Vitalik Buterin: co-author of the foundational QF paper and a useful steady voice on which mechanism-design contributions are operationally sound. The Ethereum-research stack has been the most sustained live laboratory for the proposals in the book.
  • Michael Sandel, What Money Can't Buy (2012): the explicit philosophical case against commodification of relational and civic substrate. The book Radical Markets reads as if the authors had not engaged with. Sandel's framework is the cleanest external articulation of the Pañca-Ṛṇa stress test the Codex applies.

The Indic counter-frame — when is a market the right instrument?

The Indic tradition is not anti-market. The Arthaśāstra is one of the world's most sophisticated treatises on political economy; the śreṇī guilds, the nagara commercial regulations, the long-distance trade networks of the subcontinent operated complex market structures for millennia. What the tradition is allergic to is the move that treats every relation as eligible for marketisation on the grounds that the marketised form is more efficient than the un-marketised. The categorical distinction — what is properly transactional, what is properly relational, what is properly sacred — is structural to the tradition.

Pañca Ṛṇa is the Indic-Codex articulation of this distinction in operational form. Bhūta Ṛṇa (debt to the elements) is discharged by stewardship and restoration, not by purchase. Manuṣya Ṛṇa (debt to fellow humans) is discharged by genuine reciprocity, not by transactional settlement. Pitra Ṛṇa (debt to lineage) is discharged by honouring inheritance and continuing the line, not by auctioning the ancestral house. Ṛṣi Ṛṇa (debt to the knowledge-commons) is discharged by adding to the commons, not by paying per use. Dev Ṛṇa (debt to governance) is discharged by maintaining the institutions one inherits, not by purchasing exemptions from them.

Each Ṛṇa specifies a substrate that is improperly marketised. Each Radical Markets proposal can be stress-tested against the relevant Ṛṇa. Quadratic Funding passes — it routes capital toward discharging Ṛṣi Ṛṇa (commons funding) rather than against it. Quadratic Voting passes inside its scope (bounded decisions in Dev-Ṛṇa-honouring institutions), fails outside it. COST fails — it imposes continuous auction on the rooted-place substrate Pitra and Bhūmāta Ṛṇas honour. Data-as-labour fails — it converts relational substrate into a transactional surface in ways Manuṣya Ṛṇa cannot accept. Migration-sponsorship fails — it converts personhood and citizenship into commodifiable assets in ways every Ṛṇa refuses.

The Pañca-Ṛṇa stress test is therefore not a generic anti-market move. It is a categorically precise instrument for distinguishing which mechanism-design contributions are compatible with civilisational substrate-maintenance and which are corrosive to it. Radical Markets is intellectually generative precisely because it forces the stress test to be applied proposal by proposal — and the uneven result is the most informative reading the book can produce.

What to do with this

Three operating heuristics for builders, founders, DAO designers, and civic-technology practitioners in 2026:

  1. Use Quadratic Funding for commons-funding rounds. QF (with serious Sybil-resistance via proof-of-personhood and collusion-aware scoring) is the best available mechanism for breadth-of-support-sensitive matching against public goods. It belongs in the commons-funding toolbox next to LVT-style recapture, federated-equity distribution, and TMC-licensed substrate. Gitcoin is a live QF reference; Optimism RPGF is a live adjacent reference for retroactive public-goods allocation.
  2. Use Quadratic Voting in bounded multi-issue decisions inside institutions that already enjoy democratic legitimacy. Internal governance, DAO protocol decisions, prioritisation pilots inside civic technology — yes. Replacement for one-person-one-vote at the constitutional level — no. The instrument is real; the scoping discipline matters more than its advocates sometimes admit.
  3. Run the Pañca-Ṛṇa stress test on every market mechanism before adopting it. For each proposal, ask: which substrate would this mechanism convert into a transactional surface? Which Ṛṇa would be defaulted on by structural design? If the answer is "rooted place", "kin and lineage", "relational substrate of everyday life", "personhood", "citizenship", or "the commons-funded base of someone else's livelihood", the mechanism has failed the test regardless of how elegant its theoretical properties are. Some Radical Markets proposals pass. Several do not. The discipline of running the test is the instrument the Codex contributes.

Quick answers

Why include a book the Codex partially rejects?
Because the book is intellectually generative and operationally consequential. The Quadratic Funding mechanism has moved tens of millions of dollars to genuine commons projects through Gitcoin and related rounds; adjacent public-goods allocation systems such as Optimism RPGF show the same problem-space maturing. Any commons-funding practitioner needs to understand the apparatus. The proposals that fail the Pañca-Ṛṇa stress test are themselves instructive — they show where market-mechanism extension hits substrate that cannot be marketised without damage, and they force the line to be drawn explicitly rather than carried as an unexamined assumption. A Codex that includes only books it endorses without reservation is a Codex too credulous to be useful. The discipline of critical inclusion is part of the Codex's working method.
Is COST actually being implemented anywhere?
In its full form, no. Variants are being explored — some NFT marketplaces use Harberger-tax-like mechanisms for digital assets where the dispossession dynamics matter less; some experimental jurisdictions have studied COST-adjacent proposals at the margins. The Sāmatvārtha objection applies most sharply to COST as a general framework for private property. For specific narrow domains (idle commercial land speculatively held, certain digital-asset classes, public-airspace allocation) the mechanism's dispossession dynamics are less morally weighty and the rent-dissipation case correspondingly stronger. The categorical critique stands; specific narrow applications may be defensible on their own terms.
Has the field updated since the book?
Substantially. The Plurality programme (Weyl + Audrey Tang + collaborators, 2023–) has moved the centre of gravity from market-mechanism-as-cure toward digital-democracy and protocol-for-collaboration framings that overlap more cleanly with commons-governance literature. The QF literature has matured around Sybil-resistance and collusion-resistance mechanisms (Gitcoin Passport, MACI). The COST literature has narrowed mostly to digital-asset and idle-commercial-property contexts. Reading Radical Markets in 2026 is best done alongside the post-2020 Plurality material, which represents the authors' own updated positions in many cases.
Where else should I read?
Posner and Weyl's Radical Markets (2018) is the canonical text. The Buterin–Hitzig–Weyl paper "A Flexible Design for Funding Public Goods" (2018, the formal QF paper) is the mechanism's foundational treatment. Audrey Tang and Glen Weyl's Plurality (Plurality Institute, 2024) is the updated programmatic statement. Michael Sandel's What Money Can't Buy (2012) is the philosophical case for limits on marketisation that Radical Markets does not engage. Inside this Codex, Georgism for the foundational diagnosis, The Value of Everything for Mazzucato's parallel generalisation, Commons Governance for Ostrom's structural alternative, and Pañca Ṛṇa for the stress test.

Designing commons-funding mechanisms?

If you're working on quadratic funding rounds, civic technology, or any other mechanism-design problem where the question of what should and should not be marketised is load-bearing — write in. The Pañca-Ṛṇa stress test is the working instrument the studio uses.