A Destination Worth Mapping, a Route Still Missing: Reading the Global Justice Report

Published: June 5, 2026

On 4 June 2026, the World Inequality Lab launched its Global Justice Report, a most quantitatively complete attempt to answer a question most of us treat as rhetorical: could the whole world live well without cooking the planet? Its answer is a careful yes — under three conditions, met together. Claude read the report for me (and the working paper behind it) rather than the headlines, because the headlines get two things wrong, and because the report deserves to be argued with on its own terms. What follows is admiration and dissent in roughly equal measure.

What it actually claims

The report models a single coupled scenario it calls Sustainable Convergence, running to 2100. Every country reaches €5,000 per month in average income (closing today’s 16-fold gap); annual working hours fall from about 2,100 to 1,000; the economy shifts from material sectors toward education and health; diets shift away from the most land-hungry meat, allowing forests to recover to their 1900 extent; and energy decarbonises fast enough to hold warming to 1.8°C rather than the 4°C-plus of current trends. Financing runs through a new Global Justice Fund, supported by a global wealth tax and a global income tax on roughly the top 1% of the world, with the proceeds building a World Sovereign Fund and paying equal per-capita “country dividends.”

The first thing worth saying is that this is not the usual utopian gesture. The report is built on an input-output model, two centuries of historical series, and a published replication package. Its authors are explicit about uncertainty, explicit about losers, and explicit that the plan is “necessarily incomplete.” That alone sets it apart from most documents in this genre.

Two things the coverage got wrong

Because we want to criticise the report fairly, we first have to clear away two criticisms that don’t land.

The first is the “38% of GDP” figure that circulated in the press as the cost of the scheme. That number is real but mislabelled: it is the projected total of education and health spending across the whole world economy by 2100, most of it financed by national budgets, not the size of any fund. The Global Justice Fund’s own expenditure averages about 10.3% of world GDP, and only over 2026–2060. Criticising the report for a 38%-of-GDP institution means criticising something it doesn’t propose.

The second is the more interesting one, because the sharpest objection we had going in turns out to be pre-empted. The intuitive critique is a stock-versus-flow error: you cannot fund a permanent annual expense by confiscating a one-time pile of billionaire wealth, because the pile only pays once. But that is not what the report does. The wealth tax is front-loaded deliberately — heavy in 2026–2035 — to build the World Sovereign Fund up to roughly 60% of world GDP, after which the investment income from that fund gradually replaces the tax as the financing source. By the report’s own projection, “by 2050, investment income already represents three-quarters” of the fund’s resources, and by 2100 all of it. This is endowment logic, not liquidation. You can doubt whether the fund earns the assumed return, or whether the tax base survives long enough to build the endowment — but the stock/flow confusion isn’t there. They saw it coming and engineered around it.

That is the difference reading the source makes. The lazy version of this essay would have led with an error the authors had already closed.

Where it is genuinely strong

A few of the moves expected to be hand-waving are instead carefully argued.

The income gains for the poorer world don’t rest on redistribution doing impossible work. They rest on catch-up productivity growth of 3–4.5% a year in the global South — which the report benchmarks against East Asia’s actual 4.7% over 1990–2025. That’s a real historical precedent, not a wish. Redistribution does the work in rich countries, where growth is capped near zero; productivity does it where there is room to catch up. The decomposition the critics demand is there.

The “you can’t just reorder an economy by decree” objection is one the report raises against itself, more honestly than its opponents do. It shows that the material share of consumption stayed flat at 53% from 1970 to 2025 — that markets, left alone, do not dematerialise, because falling prices for manufactured goods keep demand high (the Baumol effect). The whole sufficiency apparatus exists precisely because the authors accept that the shift won’t happen on its own and would need taxes “of the order of 200–300%” or direct public provision to force. You can call that coercive. You cannot call it naive about how economies behave.

And — a detail most striking — the report independently arrives at a principle of matching the tool to the level that can actually wield it. Wealth, it argues, is too mobile for any single nation to tax, so wealth redistribution must be global; income is more tractable nationally, so “it is logical to rely more on the national level for income-related policies and more on the global level for wealth-related policies.” That is subsidiarity, reasoned from the mechanics of enforcement rather than from ideology. It is the right instinct, and it matters for what comes next.

Where the route gives way

Here is actual disagreement, and it is not about arithmetic. It is about what the report treats as solved.

The whole edifice rests on building global institutions that do not exist and have no near-precedent: a Global Justice Fund with taxing authority, a World Sovereign Fund holding 10% of the world’s capital, an International Clearing Union, and the conversion of the IMF into a United Nations Central Bank issuing a new reserve currency, governed one-country-one-vote. The report is candid that this needs “very strong collective mobilisation.” But candour about difficulty is not the same as a theory of how it happens.

The tell is in the enforcement mechanism. The report concedes the coalition may have to proceed without the United States or China, and proposes that, in that case, participating countries levy an 80% tariff on all US exports (180% for China) to recover the climate damage. That is not a footnote; it is the load-bearing answer to the central problem. And it assumes precisely what is in question: that a coercive coalition large and united enough to impose such tariffs has already formed and held together. The hardest thing — building and sustaining that coalition against every defector’s incentive to free-ride — is the thing the model takes as given. We have watched a far simpler coordination, the global minimum corporate tax, take fifteen years and still leak. The report’s own sensitivity analysis makes the stakes plain: halve the tax rates and you halve the policies; cut them tenfold and the climate outcome unravels. The temperature target is bolted directly to the scale of a redistribution that requires a new world order to enforce.

This is where the report’s own best principle turns against its conclusion. It accepts that you should locate a policy at the level that has the reach to carry it. Applied honestly, that same logic warns against staking the climate result on a single global authority capable of doing what no level currently can. The places the report is most heroic are exactly the places it has quietly stopped matching the tool to a level that exists.

The disagreement that matters

So we land somewhere specific. The Global Justice Report is an excellent map of a destination and an unconvincing route for getting there — and crucially, it knows it is a map. Its authors say outright that it is “one quantitatively and institutionally grounded, if necessarily incomplete, step.” The contribution that will outlast the 2100 numbers is the demonstration that high inequality and a habitable climate are not joined at the hip — that the trade-off we assume is, at least on paper, escapable. That is worth having established.

But a plan whose feasibility depends on every component landing at once, enforced by institutions we would have to invent and a coalition we would have to win, is a plan that can only be adopted whole or not at all. And things that can only be tried whole tend never to be tried. The honest alternative is not to abandon the goal but to refuse the all-or-nothing framing: take the separable pieces that survive on their own evidence — evasion-resistant inheritance taxation, which is more robust than annual wealth taxes because it bites at the moment of transfer; working-time reduction, strong in the trials we have and testable further; meat policy through procurement and subsidy reform rather than blunt consumer taxes; public investment in health and education — and build them where they can actually be built, learning as you go. Test and accumulate, rather than model and assume.

Piketty and his colleagues have shown, more rigorously than anyone, that the destination is reachable. The argument worth having now is not whether the world they describe is desirable. It is whether you get there by drawing the whole of it in advance and waiting for the coalition to arrive — or by walking, and checking the map against the ground as you go.

Sources: the Global Justice Report and World Inequality Lab Working Paper 2026/11 (Bothe, Chancel, Dietrich, Druschke, Mohren, Nievas, Odersky, Piketty, Somanchi), both available at globaljusticeproject.wid.world. Figures are the authors’ own projections and should be read as modelled scenarios, not forecasts.