A Crossroads, Not an Island: A Response to Hanoch Dagan
Zoë Hitzig* and E. Glen Weyl**
Obsessed, bewildered
By the shipwreck
Of the singular
We have chosen the meaning
Of being numerous.
—George Oppen, “Of Being Numerous” (1968)
Hanoch Dagan has written one of the most detailed and thoughtful reviews of a book one of us authored with Eric Posner, Radical Markets: Uprooting Capitalism and Democracy for a Just Society.[1]It offers so much grist for discussion that it is the only one of roughly a hundred reviews so far that seems to demand a detailed response. There is much to agree with and to be grateful for in his analysis. However, many of his strongest criticisms of the book are a result of miscommunication or misunderstanding, which we briefly clarify here. Most importantly, Dagan appears to be sympathetic to an atomized individualist view of human agency held by some other critics of the book, particularly Gaspard Koenig.[2]While the book itself was somewhat agnostic on agency, the two of us have since developed a perspective on this issue that differs from the critics’ views in important ways and that relates closely to the ideas developed in the book. We devote most of this response to briefly articulating this view.
Before turning to the issue of autonomy, we will clarify a few points in response to Dagan’s broader criticism of what he sees as the ultimate welfarist foundations of the book’s proposals. Dagan claims that the softer, open-minded autonomist view that he endorses causes problems for the ideas in Radical Markets.[3]We will show why this claim is mistaken. Most fundamentally, Dagan alleges that Radical Marketsassumes an ultimately welfarist moral view.[4]It does not.
Radical Marketsemploys welfarism much in the same way economists frequently employ clearly false assumptions—such as the absence of income effects, risk-neutrality, and the like—that provide a tractable basis for novel mathematical analysis. These assumptions, while wrong, are analytically useful. In fact, we are not aware of many novel policy proposals that have been formally derived or tightly inspired by other, nonwelfarist policy frameworks, as these are difficult to reason about formally.[5]Using other frameworks as a formal starting point thus tends to be conservative in the sense that they make it hard to derive novel policy ideas.
On the other hand, stringently optimizing given false assumptions can lead to very poor ideas that are overfit to a false view of the world. Therefore, none of the policies in Radical Marketsare optimal, even under the narrow assumptions considered; instead they are simple and approximately optimal under a robust set of assumptions. The book justifies them not just according to these models but also, as Dagan highlights, using a range of other informal arguments.[6]This balance can steer a course between the rock of conservatism and the hard place of fragile overoptimization.[7]In fact, Radical Marketsexplicitly rejects the speculative proposals that Dagan most heavily criticizes precisely because they are too complex and insufficiently robust. It seems that most of the concrete disagreements we have can be resolved by understanding the role welfarism plays in our thinking.
Let us consider one example. Dagan argues that while the COST[8]might be appropriate for certain kinds of administrative property, it would limit choice and autonomy if applied to, for example, primary residences.[9]While the book concedes that such an application should not be the first, we think Dagan’s analysis is largely misguided. Certainly, applying a COST to a permanent residence would directly limit the freedom of the citizen (call her A) on whom the COST is imposed. But it would significantly increase the freedom and autonomy of every citizen who might consider purchasing A’s home. A’s freedom falls by a bit, as she now must pay a premium if she wants absolute and unchallenged security. Meanwhile, every other citizen gains a potentially substantial freedom to access a good from which A could have arbitrarily excluded them, and every other citizen also gains income. Given the extremely unequal distribution of housing wealth, the income gained by every other citizen must on average make wealth more equally distributed.
Whether this latter freedom outweighs the freedom lost by A requires some metric on freedom, but clearly in extremisDagan would have to concede that such a limitation would be freedom enhancing overall. For example, in a feudal society, imposing a COST on the fiefs of lords that gives landless serfs a realistic possibility to purchase pieces of these fiefs must overall enhance freedom. Now whether a COST imposed on primary residences in the contemporary United States or in the State of Israel would overall enhance freedom might require subtler calculation. Dagan, however, does not offer a metric other than aggregate preference satisfaction by which to judge this. Furthermore, academic studies of the COST suggest that it would generate Pareto improvements or something quite close, and thus any metric consistent with preference would tend to endorse it.[10]
While the specifics are different, a structurally similar argument applies to Dagan’s critique of Quadratic Voting (QV)[11]: it is not clear, for example, why QV’s derivation from the welfarist argument about preference aggregation implies that QV will reduce the role of rational argumentation in the formation of public policy (and there are many reasons to believe the reverse). Nor is it clear why his argument for its flaws applies any less to his preferred application of the system to campaign finance.
While the virtues of the proposals do not rest on welfarist foundations, we acknowledge that Radical Marketslargely skates over the role of welfarism in its proposals. The complex interplay between welfarist reasoning, economic theory, and informal argumentation about the fairness of institutions indeed demands deeper methodological exposition that was beyond the scope of Radical Markets.[12]
However, precisely because we are not committed to welfarism, we do not wish to dwell on Radical Markets’s purported commitment thereto. There is, in our view, a deeper and more interesting issue related to Dagan’s premise: namely, that some idea of aggregated individual autonomy is the ultimate good. This view is reasonably widely held among liberal political thinkers, such as Joseph Raz.[13]We struggle to understand this view, and, as noted above, some forms of it that focus on preference satisfaction (or “properly informed” preference satisfaction) do not differ much from the form of welfarism most commonly applied by mechanism designers. But there is a version of this view that we think is mistaken. This view has been vocalized in conversations surrounding Radical Marketsfor some months now and thus we wish to offer a clear response to it here. We do not claim Dagan endorses this view, though it strikes us that he is somewhat sympathetic to it.
The view to which Dagan seems to be sympathetic is broadly construed as a “perfectionist liberal” one. On this view, the ultimate good lies in maximizing “self-authorship” and “self-ownership,” meaning independence of individuals from social influences and entanglements beyond what individuals would choose or wish for themselves even on reflection.[14]Gaspard Koenig is an interlocutor who has repeatedly expressed this view; he has held out the ideal of maximizing the capacity of individuals to engage in deeply isolated acts.[15]His favored examples include individuals who choose to spend extended periods in isolation—surfing, living isolated in the woods, or choosing the moment of one’s own death through suicide—as the height of self-actualization. Many who express this view admit that no individual can truly be an island unto herself, but still hold out the vision of such island-like independence as the ideal realization and liberation of an individual that a liberal state should enable.
This “island” view is mistaken about the nature of agency humans aspire to and what makes for a self-actualized person. The vision of selfhood we believe Radical Marketsshould be judged against is not the “island” vision but rather one in which individuality is defined by the communities to which individuals belong. The primary source of individuality is not independencebut a unique pattern of interdependence. Most activities and characteristics that make a person an “individual” are meaningful not in isolation or independence but only in relation to the cultural communities that give these identities substance and value. The fundamental importance of community to identity has long been emphasized by conservative, communitarian, and various sorts of radical critics of liberal societies.[16]Yet some liberals, such as Alexis de Tocqueville and Hannah Arendt, have emphasized the centrality of community organization to the preservation of liberty.[17]
In fact, the people we know who epitomize the sort of self-actualized individuals that seem impossible outside of a liberal society are not isolated surfers or renegade forest dwellers, archetypes that are far too familiar and scripted to strike most citizens as “original.” Let us offer two examples of people who, in contrast, strike us as unique.
One is the gay nephew of the former ultraorthodox chief rabbi of the State of Israel. He is deeply steeped in nearly every Jewish theological tradition and is ordained as a conservative rabbi, but runs a “God-optional” shul and was thrown out of the conservative rabbinate for marrying a Jew to a non-Jew. He knows and incorporates into his sermons more social theory and sociology than most rabbis, yet his shul is driven primarily by the contributions of a wide range of artists who draw on ancient Jewish traditions from all over the world.
A second is an Egyptian living in D.C. She has extensive experience working for the Saudi Arabian government and her public-facing images show her wearing a head scarf. She owns half a dozen of the best restaurants in the city and is a serial technology entrepreneur. She is deeply involved in global sustainable development efforts led by the main international institutions. And she teaches and studies applied ethics at a divinity school and is the mother of three children, all as a young woman and leading light of society.
None of these characteristics are even explicable when removed from the deep traditions and large communities associated with them; they would have no meaning to an isolated individual. Yet the individuals we described could hardly exist or thrive in a society dominated rigidly by those traditions. They are only possible within an open, liberal society. What makes them unique is their ability to draw on and combine a wide range of rich cultures in innovative ways to produce new and beautiful combinations, an intercultural version of what Claude Lévi-Strauss has called bricolage.[18]In this view, a society that facilitates self-actualization is not one that allows every citizen maximal independence, but instead one that offers an ideal balance of rich cultural traditions that are raw material for bricolage and the flexibility to build new identities and cultures from these. It is one that helps individuals become not isolated islands but vibrant social crossroads.
Our view of the individual as a crossroads resonates with many feminist, communitarian, conservative, and radical criticisms of liberalism—and yet we want to be clear that we are tapping into a disagreement within liberalism rather than calling out a reason to reject it full stop.[19]We do not suggest that individuals are justtheir connections to present and historical communities, defined in terms of geography, kinship, culture, and so on. Instead, the individual, significantly shaped by these communities, plays a key role in her own dynamic self-actualization through her connections to groups, traditions, and ideas. There is room to value autonomy on this view, and in fact periods of isolation may be a component of self-actualization (indeed, periods of isolation have critical social meaning in many cultures, e.g. monastic cultures). But autonomy ultimately constitutes an emergent property of a complex system of individuals rather than a static attribute of an isolated individual.
While such a view might appear abstract and philosophical, it suggests several social institutions. Most directly, this view of human identity suggests that systems that aim to formalize and verify identities should draw on rather than subvert (as centralized identity systems do) or ignore (as so-called “self-sovereign” identity schemes do) social connections. In particular, nearly all properties that individuals use such systems to prove (e.g., educational qualifications or date of birth) are already known to or shared with others (e.g., professors or parents). Preexisting networks of social cooperation and trust will typically include such “intermediate verifiers,” allowing verification of identity properties through the social relationships in which they are embedded rather than through abstract or externally imposed platforms. One of us is leveraging these facts to design new identity protocols with Nicole Immorlica and Matthew Jackson.
It also suggests that the primary aim of economic institutions should be to facilitate the formation of novel social communities and public-good-providing polities while allowing older ones to gradually decay to the extent they no longer serve citizens, a primary basis of our recent work with Vitalik Buterin.[20]It suggests that property should not belong absolutely and permanently to individuals. Rather, property should belong to a variety of communities in ways that mirror their rich and overlapping patterns of connection. Its uses and value ought to flow among these communities.
On Dagan’s reading, Radical Markets are welfarist means to autonomy-enhancing ends. This arrangement, to Dagan, threatens to undermine noble ends through a stubborn attachment to misguided means. We have argued here that Radical Markets are not built on welfarist values, nor are they in pursuit of a wholly individualistic notion of autonomy. On our view, Radical Markets are both means and ends to fostering dynamic collectives and individual freedom to move about them, securing a greater degree of equality, a greater diversity, and a greater depth of collective organization than today’s standard capitalist institutions could bring about.
As a book, Radical Marketsis far from fully instantiating these ideas and even contrasts with them in some ways. For example, the primarily individualistic focus of the chapters on migration sponsorship and Data as Labor is in tension with the above ideas. The division of COST revenues described in the book is relatively arbitrary and conforms poorly to actual community structures; QV assumes a fixed rather than a flexible polity. Overall, though, Radical Markets ideas point toward social institutions that come much closer to honoring these values than do standard capitalist institutions. The COST embodies the idea of partial collective ownership that decays to the extent that it is hoarded away from the communities to which it belongs. QV, precisely by allowing an expression of strength of preference in collective decisionmaking (of which Dagan is skeptical), allows variable and fluid identification with a range of collective communities and issues. The Data as Labor vision—of collective organization to protect individual economic rights in data and to help balance power in the digital economy—resonates with de Tocqueville’s and Arendt’s emphasis on intermediate institutions as a safeguard of liberty.
To the significant extent to which Radical Marketsfalls short of its potential, it is because the institutions it describes do not fully confront the collective-crossroads nature of identity, not because they fail to safeguard island-like autonomy.
[1]. Hanoch Dagan, Why Markets? Welfare, Autonomy, and the Just Society, 117 Mich. L. Rev.1289 (2019) (book review).
[2]. Gaspard Koenig, Quand les libéraux contestent la propriété [When Liberals Challenge Property], Les Echos(June 19, 2018, 4:00 PM), https://www.lesechos.fr/19/06/2018/lesechos.fr/0301845012112_quand-les-liberaux-contestent-la-propriete.htm [https://perma.cc/4DYM-FE5S]. Most of these conversations have not been written down; the most detailed hashing out of these views was in a live debate on “Reinventing Liberalism” on October 24, 2018. GenerationLibre,Conference – Reinventing Liberalism, YouTube(Oct. 25, 2018), https://www.youtube.com/watch?v=v5lwPgsKRj8.
[5]. While philosophers of science tend to disagree on the extent to which ceteris paribusclaims in economics can legitimately guide policy, many hold that such practice is largely defensible. For two prominent articulations of this defense, see Nancy Cartwright, Nature’s Capacities and Their Measurement161–64 (1989), and Daniel M. Hausman,The Inexact and Separate Science of Economics123–51 (1992).
[7]. For more on this point, see Vitalik Buterin & E. Glen Weyl, Central Planning as Overfitting, RadicalxChange(Nov. 26, 2018), https://radicalxchange.org/blog/posts/2018-11-26-4m9b8b/ (on file with the Michigan Law Review).
[8]. “COST” stands for “common ownership self-assessed tax.” It is an ownership sharing system in which a piece of property is held on the condition that a higher-value user does not seek to take the property for themselves. Under a COST system, a possessor of property loses the right to exclude. Similarly, the public at large gains the right to use what would otherwise be considered private property. Eric A. Posner & E. Glen Weyl, Radical Markets61–62 (2018).
[10]. E. Glen Weyl & Anthony Lee Zhang, Depreciating Licenses, SSRN(Jan. 24, 2018), https://ssrn.com/abstract=2744810 [https:perma.cc/W6TM-YHWN].
[11]. “Quadratic Voting” is a voting system different from the classic one-person, one-vote model. It allows election outcomes to be based on the intensity of individual voters’ preferences toward the political outcomes available to them. See Posner & Weyl, supranote 8, at 82.
[12]. An elaboration of the methodological lacunae of Radical Marketsis presented in Zoë Hitzig, Lily Hu & SaloméViljoen, The Technological Politics of Mechanism Design, 86 U. Chi. L. Rev. Online(forthcoming 2019).
[14]. This is similar to the definition of “liberalism” used by Amartya Sen, The Impossibility of a Paretian Liberal, 78 J. Pol. Econ.152, 153 (1970). Prominent articulations of perfectionist liberal views are offered in Charles Larmore, The Morals of Modernity(1996); Charles E. Larmore, Patterns of Moral Complexity(1987); and Joseph Raz, Autonomy, Toleration, and the Harm Principle, inJustifying Toleration155 (Susan Mendus ed., 1988). For discussions of how perfectionist liberals vary from other variants of liberalism, see, for example, Martha C. Nussbaum, Perfectionist Liberalism and Political Liberalism, 39 Phil. & Pub. Aff.3 (2011).
[16]. Particularly influential criticisms along these lines include Alison M. Jaggar,Feminist Politics and Human Nature(1983); Catharine A. MacKinnon, Feminism Unmodified(1987); and Michael J. Sandel, Liberalism and the Limits of Justice(2d ed. 1998). A more recent iteration of this criticism is in Patrick J. Deneen, Why Liberalism Failed(2018). See alsoWill Kymlicka, Liberalism, Community and Culture2–3 (1989) (arguing that the individualism for which liberals advocate “accords with” the individual’s communal relationships).
[17]. See generallyHannah Arendt, The Origins of Totalitarianism(1951); Alexis de Tocqueville, Democracy in America (Harvey C. Mansfield & Delba Winthrop eds. & trans., Univ. of Chi. Press 2000) (1835).
*PhD Candidate in Economics, Harvard University.
**Founder and Chairman, RadicalxChange Foundation; Principal Researcher, Microsoft Research; Visiting Research Scholar and Lecturer, Julis-Rabinowitz Center for Public Policy and Finance, Woodrow Wilson School of Public and International Affairs, Princeton University. We are grateful to many interlocutors of the RadicalxChange community for contributing to these ideas, particularly Marion Fourcade, and especially to Mona Hamdy and Charlie Thompson for detailed comments.
