By Susumu Cato & Hun Chung
In the half-century since its publication, Robert Nozick’s Anarchy, State, and Utopia (1974) has become a pillar of contemporary political philosophy. Most scholars, however, focus their attention on the first two-thirds of the book—Nozick’s entitlement theory of justice and his famous derivation of the minimal state from a state of nature. The final section, Part III, titled “A Framework for Utopia,” is frequently treated as an inspired but somewhat underdeveloped afterthought.
In our paper, “A Formal Theory of Robert Nozick’s Framework for Utopia,” (published online in Noûs, open access), we argue that Part III is not a mere addendum but the essential logical culmination of Nozick’s project. To assess its theoretical success, we provide the very first formal model of Nozick’s “possible worlds” framework, transforming his informal intuitions into a precise mathematical formalization.
The Problem: Formalizing the Possible Worlds
Nozick’s vision of utopia is famously pluralistic. He does not propose a single vision of the “best” life; instead, he argues for a “meta-utopia”—a framework of many voluntary sub-communities where individuals can live according to their own diverse values. Nozick justifies this framework by imagining a “possible worlds” model: a thought experiment where every individual has the power to create a world of their own design by their imaginations, but where others are equally free to stay or leave for an alternate world that they can also create by their imaginations.
Nozick’s central claim is that any world that remains stable in this environment—where everyone is free to imagine and move to a better alternative—must be one where no one can be lured away by the promise of something better, and, hence, qualify as “utopia,” understood as “the best of all possible worlds.” Nozick’s “framework for utopia” is then identified as a collection of such stable worlds within this possible worlds model.
However, Nozick left the mechanics of this model informal. What are the precise conditions under which a world is “stable?” How does Nozick’s notion of stability compare with other notions of stability? Does a stable framework always exist? Would everybody (as Nozick conjectured) receive their marginal contributions from their worlds once a stable configuration is finally reached? Without formal answers, Nozick’s “framework for utopia” remains a brilliant metaphor rather than a robust theory. Our paper fills these gaps.
From Possible Worlds to Formal Model
We begin by modeling a set of possible individuals and defining a “world” as a voluntary association (a non-empty subset of individuals). A “framework” is a partition of all individuals into such worlds. Each individual has preferences over the worlds in which they might belong and possesses an imagination set specifying which worlds they can conceive as viable alternatives.
Nozick’s core idea is that a world is stable if none of its members can imagine and prefer another world that would persist under the same freedom of exit and entry. Nozick’s original definition of stability is circular: it defines stability in terms of the absence of preferable stable alternatives. However, as we argue in our previous work—”Stable Dystopia: A Critique of the Circular Definition of Stability in Nozick’s Model of Utopia” (2024, Analysis)—Nozick’s circular definition of stability has many inherent theoretical problems (that made even Nozick abandon it himself.) We therefore adopt a non-circular formulation of stability: a world is Nozick-stable if and only if none of its members can imagine a strictly preferable alternative world.
A Nozick-stable framework (Nozick’s “meta-utopia”) is then a partition of individuals into worlds, each of which is Nozick-stable. Nozick’s ultimate aim was to argue that not only does such a Nozick-stable framework exist, but such a framework is equivalent to the minimal state. This constitutes Nozick’s ultimate justification for the minimal state.
First Result: Stable Worlds Do Not Guarantee a Stable Framework
One might think that if every individual has at least one stable world available to them, then a stable framework should exist. We show that this inference is invalid. Even when stable worlds exist individually, they may overlap in ways that prevent them from forming a partition. The existence of stable components does not imply the existence of a stable global structure.
This already introduces pressure on Nozick’s project. His argument requires not merely the existence of some stable communities, but the existence of a complete and rights-respecting framework covering everyone.
Second Result: Nozick Stability Is Stronger than Core or Nash Stability
To understand how demanding Nozick’s stability notion is, we compare it with familiar solution concepts from cooperative game theory.
We show that Nozick stability is strictly stronger than both core stability and Nash stability. While core and Nash stability are logically independent of one another, Nozick stability implies both. As a consequence, many configurations that would count as stable under standard economic notions fail to qualify as Nozick-stable.
This means that even if coalition structures exist that are immune to blocking (core-stable) or to unilateral deviation (Nash-stable), they may still be destabilized by Nozick’s imaginative exit condition. The upshot is that Nozick’s concept makes stable frameworks significantly harder to achieve than standard models of voluntary association would suggest.
Third Result: Existence Requires Restrictive Conditions
Given the strength of Nozick stability, we identify a set of sufficient conditions under which a Nozick-stable framework does exist. However, these conditions turn out to be highly restrictive. Roughly speaking, they require strong alignment and independence of preferences that are unlikely to hold in pluralistic societies.
Thus, while stability is not logically impossible, it depends on structural assumptions that undermine the realism of the model. The more diversity of preferences and imaginative possibilities we allow—precisely what Nozick celebrates—the harder stability becomes to secure.
Fourth Result: Marginal Contribution and Exploitation
Perhaps most strikingly, we show that even when a Nozick-stable framework does exist, it does not guarantee (as Nozick claimed) that individuals receive their marginal contributions within their stable associations. Contrary to Nozick’s conjecture (based on analogy with competitive markets), stable worlds may systematically assign individuals less than their marginal product.
This opens the door to a troubling possibility: stable frameworks may institutionalize and perpetuate exploitation. Stability alone does not ensure fairness, reciprocity, or distributive justice. A configuration can be stable under voluntary exit and still leave some individuals structurally disadvantaged.
This finding is especially significant because Nozick’s identification of the minimal state with utopia depends not merely on voluntary stability, but on the claim that such stability embodies a morally attractive outcome that inspires.
Philosophical Implications
Nozick’s ambition in Part III of Anarchy, State, and Utopia was to complete his libertarian defense of the minimal state by showing that it is not merely the only morally legitimate state, but also an inspiring utopian ideal. Our analysis casts doubt on whether that argument succeeds.
If Nozick-stable frameworks frequently fail to exist, then this implies that it would be difficult to arrive at its equivalent—the minimal state—through purely spontaneous and voluntary associations. If they exist only under highly restrictive assumptions, then the success of the minimal state depends on implausible structural conditions. And if they can exist while stabilizing exploitative arrangements, then stability does not deliver the utopian property Nozick hoped for.
The broader methodological lesson is that formal modeling can illuminate the internal dynamics of normative political theories. By reconstructing Nozick’s possible worlds framework with precision, we are able to test its coherence, feasibility, and distributive implications. In doing so, we aim not merely to criticize Nozick, but to clarify the conditions under which a voluntaristic theory of political association could genuinely yield a stable and attractive social order.
Whether the minimal state remains “an inspiring vision” depends, we suggest, on questions that only become visible once the framework is formally articulated.
Conclusion
For 50 years, critics have argued that Nozick’s Part III was a detour. Our paper proves that Part III is the engine of the entire book. By defining utopia as a stable equilibrium of voluntary associations, Nozick provides the normative justification for the minimal state that Parts I and II only partially achieve.
In this sense “A Formal Theory of Robert Nozick’s Framework for Utopia” is more than just an interpretative commentary; it offers a systematic reconstruction of one of the central pillars of Nozick’s project. By developing the first fully rigorous possible-worlds model of Nozick’s framework, we identify with precision the structural and theoretical difficulties confronting his vision of meta-utopia as a robust and sophisticated response to what may be the central question of political philosophy: “how can we all live together while living differently?”
We hope this summary encourages researchers in political philosophy and PPE to read our paper and revisit Part III of Anarchy, State, and Utopia—not as a relic of the 1970s, but as a living, breathing model for the governance of complex, pluralistic societies. As an open access paper, our paper can be freely accessed and downloaded at: https://doi.org/10.1111/nous.70036
References
CATO, S. & H. CHUNG. (2026). “A Formal Theory of Robert Nozick’s Framework for Utopia” Published Online in Nos (Open Access): https://doi.org/10.1111/nous.70036
CATO, S. & H. CHUNG. (2024). “Stable Dystopia: A Critique of the Circular Definition of Stability in Nozick’s Model of Utopia.” Analysis 84 (3), 465-475, Jul 2024 (https://doi.org/10.1093/analys/anad091)
Interesting to read. I want to comment on this: "This opens the door to a troubling possibility: stable frameworks may institutionalize and perpetuate exploitation. Stability alone does not ensure fairness, reciprocity, or distributive justice. A configuration can be stable under voluntary exit and still leave some individuals structurally disadvantaged."
This leads me to think a definition of utopia is in question. If the world's are voluntarily created and left, they still must be possible (I can't imagine a world where no one hunts or farms but we all magically have food). The opt in opt out "utopia" doesn't seem to promise our desired outcomes, but our desired structure. What the structure leads to is our just desert, yet if we don't like the consequences, and believe a new structural system would yield better ones, we are free to opt out and create a world with a new structure and try it with others.
All this to say the existence of exploitation or unfairness doesn't seem to reflect that the "utopia" has failed, only that an individuals preferred (at the time) structure created unfairness and exploitation