The Problem of Evil and its Coasian Solution: a dialogue on the economic theory of theodicy
The year is 2035. Breakthroughs in genomics, artificial intelligence and synthetic biology have led to a scientific renaissance, a second enlightenment, as the brains of the great intellectual and academic figures of the past (at least those with surviving DNA fragments) are regrown and reprogrammed with the contents of their original corpus mentally intact.
Meanwhile, at the massively online University of Chicago, a controversy is brewing. The brain of the late, great economist George Stigler has been made a Professor Emeritus at – of all things – the Divinity School. His appointment causes especial uproar among a coalition led by the brain of the late biologist Stephen Jay Gould, under the slogan that economic science and religion belong to “non-overlapping magisteria.”
Fresh from the hereafter, Stigler’s first post-posthumous seminar bears the title “The Problem of Evil and its Coasian Solution,” which, according to Stigler, will launch a field of research known as “Theology and Economics.” Unbeknownst to Stigler, Gould is present in the audience taking notes, ready to pounce when the lecture concludes.
STIGLER: In summary, heretofore theologians have developed incomplete theodicies due to their failure to grasp the Coase Theorem. The evil god allows, as a type of harm, is always and everywhere reciprocal. We should not waste our time moralizing about victims and perpetrators of evil – the best we can do is seek the least costly way of reducing said evil: that mutually beneficial outcome which leads to minimized social cost. This means the existence of evil is not only consistent with a benevolent and all powerful God, but further demonstrates his providence through Coasian means. Any questions?
GOULD: Yes. I have several questions to pose to you. You seem to be saying that evil does not exist, that it is only apparent. This is a most absurd and reactionary view, it seems to me, given mass technological unemployment and the enduring plight of the Global South.
STIGLER: Ah, Professor Gould! So glad to see your mind was salvaged. You went much too early. But yes: to say that “evil is only apparent,” as Hume pointed out, is indeed contrary to human experience. That is not my view. My view is simply that in some sense there is an optimal amount of evil in the world, which, like pollution, is unlikely to be zero. If there is an excess of evil in the world it is due to an incompleteness of property rights. The crux, then, is not evil per se, but the quantity thereof.
GOULD: So you’re saying the ten billion humans living in subsistence is optimal? Do you not err in much the same was as the naïve adaptationist who treats natural selection as ruthlessly molding an organism for optimal utility? “Our world is not an optimal place, fine tuned by omnipotent forces of [economic] selection … History matters.” The evolution of societies, just like organisms, often “reflects inherited patterns more than current environmental demands.”
STIGLER: That’s right. But it is precisely this ubiquitous influence of history and transaction cost (stretching back to the Garden of Eden) that forces us to live in a “second best world” – that is, one separated from God. Every durable evil is efficient for the context or else it would not persist over time. Otherwise, the rational agents involved would have practiced their God given free will (i.e. negotiated) and improved their situation through a Coasian bargain. It is nonsensical to declare, in the abstract, that a state of affairs is evil. It must be put in context, measured against the opportunity cost of the alternative state of affairs. In other words, evil compared to what? My position is similar to the “greater good” theodicy developed by Richard Swinburne many decades ago. In a finite world, God allows for the best of possible logical worlds. That implies efficiency in the pareto sense, constrained, as it may logically be, by path dependency. I therefore am with Hume, too, in rejecting the view that God will rectify evil in the future. If an evil can be efficiently rectified it will already have been – there are, after all, no $1000 bills left lying around on the sidewalk.
GOULD: Ah, the economist’s version of the “myth of natural harmony—all is for the best in the best of all possible worlds. Is that not the foolish vision of Dr. Pangloss, so vividly satirized by Voltaire in Candide—the world is not necessarily good, but it is the best we could possibly have.” The quasi-optimum you describe is at best merely local, not global. The world is too random and complex for your stylized analysis. Do you defend the evil of African American slavery on the basis that it persisted for over 80 years? Mere efficiency could not be less relevant to the problem of evil. The concept is an amoral technocratic farce: pure ideology.
STIGLER: On the contrary, Stephen. Violations of the pareto principal create outcomes in which one person is made worse off without anyone receiving any benefit. To quote the philosopher Joseph Heath, “one way of formulating the Pareto principle would therefore be to say that it recommends the elimination of gratuitous suffering.” Indeed, from my readings on the problem of evil it seems to be more typically framed as the problem of gratuitous suffering. Efficiency, or the seeming lack thereof, is thus the issue at the heart of any theodicy. The end of American slavery did not come without costs. The Britons, for example, got rid of slavery without having a destructive civil war by buying the slaves up and setting them free. In economics, we call that approach compensating variation.
GOULD: That’s all fine and good, but this focus on efficiency completely ignores the issue of massive class inequality around the globe. Forget simply about class – God fails so much as to provide equality of opportunity – and for whole nations! Your vision of God strikes me as brutally and implausibly meritocratic. Perhaps even racist.
STIGLER: Hold on now. I’m with Swinburne again on this, in that we both recognize that such disparities are not evil, per se. To quote the man himself, “[I]f [God] gives to some ten good things, and to others twenty good things, no one is wronged; nor has he failed to be perfectly good. He has been generous, and, more so, he has made it possible for us to be generous.” Swinburne may not have realized it, but that is pure Pareto principal. And, if I may proffer another Coasian formalism, the notion that in some sense poverty creates the conditions for compassion strikes me as eminently describable in terms of positive externalities.
GOULD: Some show of compassion. So do you follow the Irenaean line that suffering – even if not strictly gratuitous in your view – exists to aide in spiritual development?
STIGLER: More or less. I prefer to cite the models by my colleague, the brain of Gary Becker. Indeed, he has a forthcoming paper in the Journal of Theology and Economics titled “Spiritual Capital: A Theoretical and Empirical Analysis.” Preliminary results indicate spiritual capital formation adds two percent annually to the growth of gross domestic spirit.
GOULD: And what spiritual development comes out of the ichneumon parasites? That genus of wasp which plants its eggs in a caterpillar’s body, but not before permanently paralyzing the host so that it can live to be devoured alive, bit by bit, when the eggs hatch? Is the total, senseless deprivation of the animal kingdom not the death knell of your economic theology?
STIGLER: I think not. Perhaps if my theology gave moral status to a caterpillar, but I have my doubts that something so small should feel much pain at all. More to the point, suffering demands a marginal analysis. How is the life of a caterpillar likely to resolve if it manages to avoid contact with an inchneumon? Death by fire ants, bird beak, or being baked in the sun? Everything biological ages and decays; and, until recently, death was a human inevitability. Parasites exist in humans too, and have evolved symbiotically to help regulate our immune systems. The providence of God is evident in the balance of marginal costs caused by the earth’s carnivora against the marginal benefits of a stable but dynamic ecology, our natural resource endowment, on which to build human civilization. As they say, there are two kinds of theologians: those who understand price theory… and everyone else.
GOULD: And of other natural evils, like earthquakes or tsunamis?
STIGLER: God is the kind of parent that lets his children learn from their mistakes. In other words, God recognizes that intervention in these areas tends to lead to moral hazard, so God in his infinite wisdom gave man skin in the game. The Coasian corrective to your specific examples is to make sure the people who live on fault lines or near the coast pay a premium on their homeowners insurance to internalize the cost of their risky choices.
GOULD: Disgusting – you are simply substituting one evil for another: insurance. Perhaps God’s laissez faire attitude to suffering is due to the fact that he’s been swindled by a ‘free market’ hack.
STIGLER: You’ve reminded me of a remark I made in my Nobel Prize banquet: A Swedish physicist cannot discuss his work with fifty people unless he goes abroad. A Swedish economic theologian can get opinions and instructions in his native language from thousands upon thousands of his fellow citizens. It seems we will have to leave matters at that. I am giving a lecture at my prayer tank on finite horizon models of Christian eschatology that I’m quite late for.
What “God” Means to a Materialist
Explaining just what is meant by the word “god” has proven very problematic for generations of philosophers and theologians. Given the abstract and invisible nature of god, the standard approach has been rationalistic – applying “pure reason” to divine from logic and a few plausible premises what, by deduction, god must be. Since Kant, this approach has clearly failed, and still no consensus exists on god’s essential properties. Where agreement abounds, the logical dilemmas of qualities such omnipotence, omniscience and omnipresence have been inadequately resolved. What follows is my own proposal for what is in practice meant by “god,” which will in turn provide an explanation for why the standard attempts have been so problematic.
Beforehand, I need to contextualize this essay by confessing my own basic convictions with respect to language and meaning more generally. Unless justification or elaboration is required on a specific point, these position statements will have to stand on their own in the interest of space and focus.
Mental States and Translation
First, I am an eliminative materialist in the pragmatic and radically empirical traditions of Hume, W.V. Quine or C.S. Peirce. This is the view that most of our everyday terms for mental states are useful fictions; that numinous entities such as “self” or “thought” can in principal be eliminated with a more fundamental physical explanation.
This position is not nihilistic about the language of mental states. On the contrary, it is both acceptable and unavoidable to carry on truly meaningful conversation using the conceptual animism embedded in our ordinary language. Communication is possible because our usual interlocutors, other humans, have a common evolutionary inheritance and thus a common set of ontological commitments that place them in the same language game.
In particular, human brains share folk models of psychology and physics which help imbue the world with meaning and structure for interfacing in our environment. We have stubborn instincts about causality, being, object, purpose and identity. These instincts are so fundamental to functioning in our daily lives that even professional philosophers can be fooled into thinking that the elements of our mental asset library are not merely an interface but are literally fundamental to reality. To paraphrase Quine: Metaphysical thinking arises from the sensation of computation modeling the world.
Imagine if a Martian Scientist were to land on earth and begin to make observations. How should we explain ourselves? How could we explain even a simple to understand (for us) statement like: “As payback, Johnny stole Susie’s lunch money.” We could point to examples, but what is our plan B if our ostensive explanations are lost on them?
Other humans understand that “payback” refers to the retributive quality of Johnny’s intentions, but what is intention anyway; and how do we make the attribution without reading Johnny’s mind? Human attempts to elucidate these irreducible concepts usually just beg the question: “Well, if you intend something that means you meant to do it!” Indeed, it would probably be easier for the Martian and the human to converse about quantum mechanics than a simple intentional sentence.
Suppose the human-Martian discourse on intention continues. As the human circles back for the fifth time to contrast “intentional” with “indifferent,” the Martian interrupts. “When two atoms come into proximity and begin to share an electron, forming a covalent bond, does one say that the atoms intended to form the bond? Do atoms ever intend to find another atom with which to bond but fail? Are there atoms which form bonds on accident?”
“No of course not, that’s absurd,” replies the human. “How can atoms have intentions?”
“That is our question to you. You are composed of atoms after all.”
The communicative difficulty that inevitably frustrated the human is due to his reliance on his intuitive tool kit of stand-on-there-own concepts, for which the Martian has no innate ability to process. A solution is to shift from the misnomers of folk psychology to a more fundamental level of analysis – one hopefully within the grasp of the Martians. Applied to folk psychological concepts such as intention, this summarizes the practical program within eliminative materialism. Move from the ordinary language of mental states to the neuroscience of mental states and the chances of Martian communication go up.
Thus one way to proceed would be to offer high resolution scans of a human brain – some during the act of intention and others while at rest. One study that tried something like this concluded that intention is just the sensation of early movement, and that awareness of intention reliably occurred after the execution of executive motor functions.
Whatever the case may be, we can be sure the Martians would discover that our ordinary language for certain types of mental states does an awful job at describing what is really taking place. By analyzing the brain, a sufficiently intelligent Martian would be able to arrive at a definitive translation of “intention” defined in computational terms. If the Martian is Turing-equivalent, it will in time discover the physical pathways of the sophisticated pattern recognition algorithms that search for cues of vitality in other beings; that scan eyes for sincerity; and that discover conspiracies couched in random events.
Perhaps the Martian discovers a subtly within his analysis, delineating mental states correlated with intentional action and mental states correlated with the perception and attribution of intention in other beings. Without knowing it, the Martian has discovered the phenomenon of the intentional stance – the ability to represent the knowledge, beliefs and intentions of another. Indeed, with sufficient resources the Martian will be able to understand any of our ontological commitments on the basis of the mental state’s physical correlates. In general, since the abstract mental state A is understood as supervening on the physical correlates X, it is a definitive definition. The Martian will be able to search for any instances of X-like physical systems and by induction conclude that it embodies the mental state A.
God and the Intentional Stance
With the above establish I can now turn to what I think is meant by God.
God has been defined as an invisible universal being of omnipotence, omniscience, benevolence, and design. God is a higher power capable of consolation, intervention, forgiveness, and punishment. More than betraying a general anthropocentricism, these attributes are connected by a common thread: they all illustrate the phenomenon of a human mind taking an intentional stance to the universe – that is, of attributing wants, desires, and motives to the natural environment.
Everyone has experienced the pain of stubbing their toe on a piece of furniture with the concurrent feeling of anger directed at the furniture piece responsible. But why lash out against a hardwood chair? It is not as if the chair intended to inflict harm. Yet this simple example demonstrates the powerful human ability to construct a representation of an inanimate object’s motives, to see intention in the non-mental. Indeed, the typical human is more fluent at perceiving light flickers in terms of demons than wiring.
So when I consider what is meant by god, I ignore the specific attributes. God as a universal intentionality is prior to these product differentiations. Take the god of the deists. 17th century deists were enlightened enough to recognized the indifference of the universe, that is, the absence of an active intentional agency. Yet they could not help project intentionality onto the prime mover, conflating a naturalistic higher power with a weakly human consciousness. As an aside, this was in part because the self-evident moral ontology of Natural Law felt so real as to demand an intrinsic reality.
This picture of god can be successfully elucidated to other humans with the use of more of these sorts of thought experiments. However, if we sought verification in a more definitive, translatable way, then we could measure the mental states of someone “perceiving” god and someone perceiving more prosaic demonstrations of the intentional stance, and compare their computational structures. In both cases, I bet one will find algorithmic systems that function to detect reliable cues of other beings which embody representational goal oriented software, too, each with their uniquely assigned reaction function. Perhaps the main difference between a secular adult and a spiritual mystic comes down to a set of neurological parameters controlling the sensitivity of these detection systems.
The example of the malicious hardwood chair demonstrates that in some sense our propositional attitudes towards the world can be more or less appropriate and more or less useful, according to the proposition’s “direction of fit.” As a benchmark, seeing intention in the chair is an error that will lead nowhere, while seeing intention in another awake and active animal is highly useful and predictive. This is because in the case of another human, our mental representations (or attributions) of their intentions A supervene nicely onto the complex arrangements of their brain matter X, while failing to map onto Y, a homogenous arrangement of wood fiber.
The same analysis can be done for god. If notions of god arise from people taking an intentional stance towards our natural environments – that is, actively constructing a representation of, for example, what nature “wants” – then the next question to ask is whether this projection can find a direction of fit. While the abstraction of intentionality A can be easily projected onto any number of things, from voodoo dolls to lightning bolts, this says more about the functional sensitivity of our brain than voodoo dolls or lightning bolts. Neuroscience has established that intention is a specific kind of physical state X in the brain. If some object Y lacks the properties of X then it is unwise to project A on to it. In this way, god is a type of attribution error or category mistake and thus can be rejected as a useless, meaningless, and erroneous application of the intentional stance. The lack of a direction of fit implies meaninglessness because, though humans find no problem talking meaningfully as if nature has intention, the Martian will be locked out from communicating without an X for empirical reference.
It should not be surprising that so many cultures across space and time have independently invented gods. Human mental faculties were not selected for on the basis of Ultimate Truth. During our evolution, “accuracy” always served an end. It is extremely useful to be able to tell the difference between an angry stranger and a happy one; to model the knowledge inside the heads of our friends and enemies; to judge two organism’s interests as potentially divergent; to form alliances; and to distinguish theft from “accidental borrowing”. Having an internal knowledge of precisely what intentional mental states are on a physical level is not evolutionarily important compared to the ability to actively infer and project intentionality in social interactions or in the wild.
It is so important that, in error management terms, our intentionality detection system would rather take on many false positives than to accept a genuine and potentially lethal false negative. Monotheists are simply those who choose to worship one false positive in particular.
To the extent that foreign investment occurs, it raises the amount of capital in the country, driving wages up and profits down. The effect is exactly analogous to the effect of free migration. If people move from labor-rich countries to labor-poor ones, they drive wages down and rents and profits up in the countries they go to, while having the opposite effect in the countries they come from. If capital moves from capital-rich countries to capital-poor ones, it drives profits down and wages up in the countries it goes to and has the opposite effect in the countries it comes from.
The people who attack “economic imperialism” generally regard themselves as champions of the poor and oppressed. To the extent that they succeed in preventing foreign investment in poor countries, they are benefiting the capitalists of those countries by holding up profits and injuring the workers by holding down wages. It would be interesting to know how much of the clamor against foreign investment in such countries is due to Marxist ideologues who do not understand this and how much is financed by local capitalists who do. David Friedman, Price Theory (via aboveauthority)
Why Religion Has Declined in the US
At Overcomingbias.com Robin Hanson speculates that greater wealth explains the decline in American religiosity, but this is unlikely. The United States has a seen a surge in secularism since 1990, jumping from 8% “non-religious” then to 20% now according to the latest polls by Pew and Gallup. Has median wealth really changed so dramatically since 1990?
No, but what did change was the expansion of the US welfare system. The literature in the economics of religion consistently finds a significant and strong social-insurance aspect to religious attendance. See for example, one or two. This fits the model proposed by Iannaconne of religious orgs as emerging to provide private social insurance, and using religious ritual, stigma, sacrifice etc. as costly signals of commitment to prevent free-rides.
Thus wealth does not lead to religious decline — social security does. The correlation with wealth exists because people’s income elasticity for social welfare is greater than unity. That is, overtime wealthier nations demand more public goods. The expansion of anti-poverty measures and income subsidies in the US after WW2 and through to the great society granted households greater independence from local mutual aid systems, churches included. These effects magnify across generations, as show by the life cycle data found here on page 22. Boomers who were still nominally religious were less likely to force their children to attend church; Gen X’ers all the more, and so on. Thus the demographics of the secular boom beginning in 1990 are biased towards the under-30 crowd. Beginning in the late 80’s there was a massive increase in subsidized medicaid across the US, as well, and now we must include Obamacare; so expect the religious decline to deepen.
As the state’s safety net moves in as a lower cost substitute to local safety nets, people don’t suddenly become Secular Humanists. People in general remain mystical, but their mysticism doesn’t need to be organized and tested around a particular focal point. Hanson, focusing on wealth, forecasts a resurgence of religiosity in the event that wealth suddenly declines (say, due to the singularity). Yet if wealth declines and social security remains high, I see no reason for there to be a major resurgence in organized religion. Religion will makes a come-back in scenarios where collective action and enforcement is weak. Most “whole brain emulation” or strong-AI scenarios seem to me to introduce extremely low transaction costs and thus high degrees of secularized collective action and coordination. People may still have a spiritual impulse, but there will be little pressure to form coalitions around a particular set of dogmas.
The secular boom and the decline in US religiosity post-1990 is the subject of my current thesis project, so I will eventually be publishing new empirical research on the topic. But the existing literature and my own preliminary research is strongly supportive of the model sketched above.
Telescopic Morality vs Myopic Morality
Over at The Umlaut Adam Gurri has written an interesting piece arguing that much too often moral discourse privileges the far over the near, the Big Cause over the kindness and virtuosity one can work to generate in his or her own community.To make his point he references the Dicken’s character Mrs Jellyby, the “telescopic philanthropist” who pursues distant projects at the expense of her duty to her own family. Ultimately Gurri suggests that if we all concentrated on the duties within our own groups, the decentralized web of local obligations (including property rights and market exchange) will be more resilient to disasters (like the Hatian earthquake) than a fragile network of self-righteous individuals or groups with discrete, far-flung projects.
The advice to concentrate on the problems that you can fix is always a good one, however I think Gurri’s call to turn inward is a non-sequitur. Some level of institutional, cultural and individual inwardness is required to keep civilization from collapsing, however it’s still valuable to consider the welfare of other people and nations. Indeed, the real “siren song” is for an overly myopic morality, epitomized by nationalism and xenophobia, Randian self-interest, or the rivalry between Springfield and Shelbyvile. Clearly some kind of balance has to be found between privileging the local and the principal of analytical egalitarianism.
Gurri gives Haiti as an example and it is a good one. Ideally Haiti would have functioning institutions and a healthy amount of social capital so it could be resilient to disasters like a massive earthquake, but in the meanwhile they still need our help. He compares Haiti to Japan, which recently suffered from a similarly powerful quake but rebounded. But how did Japan get to its current pre-quake state? Japan’s growth miracle after the reckoning of WW2 was a combination of a strong state/culture *and* the extension of international trade and aid.
At a more basic level I am reminded of the argument from Ayn Rand, who denied that supporting your family was a form of self-sacrifice. True altruism is to neglect your kin in favour of your neighbour’s interests, she argued. Nonetheless, we need not neglect our own in order to help the other. The way Rand saw things was in a kind of pre-marginal revolution dichotomy whereas an economist would say one should equalize the marginal rates of substitution between helping your own and the other. At some level, showering your own family with wealth and affection has diminishing returns and would be better spent on more ‘telescopic’ projects — which of course includes both philanthropy and entrepreneurship. Imagine if Elon Musk or BIll Gates took Gurri’s advice and dedicated themselves to their local community garden?
BTW: Gurri refers to Robin Hanson on the subject of far versus near, but as far as I know Hanson is a pretty telescopic individual (he is a futurist after all). To the extent that Hanson opposes “far” mode moralizing, it’s because he views the average person as quite bad at it. When people moralize about the future, according to Hanson, they are more likely to be making statements about the present. In particular, people are overly idealistic. It often takes exceptional people to be effective while being telescopic, but it can be done.
The law, biology, standard norms, all obligate a man or woman to be the safety net to his or her child, but that does not preclude chasing the highest marginal impact in other domains.
The best argument for a balance between virtue and duty, and more “far” concerns comes from Simon Blackburn’s account of quasi-realism, which points out different normative concepts supervene at different levels of conceptual abstraction. In other words, the Vulgar Moralist is only partly right about the dimensions of our moral sphere. Most of our moral language is emotive and deontological, and supervene nicely on our appraisal of individuals and character. Utilitarian calculus could be called ‘far’ because it supervenes on properties that are much more idealistic, rational, static, abstract, namely institutions and policies — talking about “virtuous nations” or the “duty of the government” is an category “error” in the J.L. Mackie sense, as the concepts have no direction of fit. The inverse error is found in the lowly husband caught applying “the greatest good for the greatest number” to justify his polyamorous extra-marital affair.
This goes a long way to explain Hanson’s reservations. Hanson would probably accuse Mrs Jellyby of selecting her philanthropic occupation on the basis of its high signal value. The cynical motive need not be conscious. A better description is that Mrs Jellyby has erroneously applied a near modality to a far problem (this one reason why I reject Peter Singer’s Good Samaritan intuition pump in The Life You Can Save, and other personal calls to stop genocide. While such arguments appear to be utilitarian they actually work by stoking instincts of a deontic flavour).
Quasi-realism’s logical conclusion is thus a pragmatic unification of the near and far types of morality as different kinds of applied ethics with different kinds of specialized roles. They are only in opposition when people bungle the moral gestalt.
I go so far as to define moral progress as the slow unwinding of a lot of these moral gestalt errors, such as the recognition of nations as complex systems rather than unified moral - and thus culpable - agents; or that tax design ought to be based on measures of efficiency rather than a priori theories of what’s “dutiful”.
The list could stretch a mile. But in every case progress wasn’t made by privileging the telescopic over the myopic, or vice versa. Rather the key was to make the conceptual demarcation within moral language as clear as possible, on the basis of which natural properties moral concepts are able to fit.
So - TL:DR - I am not really disagreeing with anything Gurri wrote. This essay was more about showing how Gurri’s point could be made more comprehensive within a broader framework of how moral language functions.
If Utilitarianism is true should individuals practice it?
Paradoxically, the answer is not obviously “Yes”.
Joshua is so sure of this that he treats it as a logical imperative. If it turns out not to bring world peace, you are passing a utilitarian style judgement on the system wide outcome. Therefore, whatever was done in the first place was not in fact utilitarian.
Unfortunately for Joshua, this reasoning invokes the fallacy of composition. He is considering the idea that good things happen if individuals become utilitarian agents. Yet the qualities of an individual’s actions and higher levels of social organization are not guaranteed to bear any resemblance.
Fundamentally, he is defending the wrong side of the utilitarian equation. Utilitarianism is about achieving better and worse states of the affairs. Start with naming the global level utilitarian worlds that you want and then consider which types of agent beliefs or behaviours are conducive to them. More simply, instead of rewiring or repressing human nature, build the institutions that will realize those goals by harnessing our natures.
In fact, many uncomfortable conclusions are entailed by utilitarianism. The kind of mind we need may be quite irrational and biased in its own way, in much the same way competitive markets prove opportunistic natures and self-love can be good for the social utility function. And even then, any governance system would only be approximately utilitarian if it’s not possible to satisfy unrestricted domains.
Different levels of social organization require different conceptual modes based on what type of real property our ethics is supervening to. At the policy level the real things are laws, regulations, institutions. A constitution can quite readily be rational, cosmopolitan and consistent (internally and overtime). it makes no sense for agencies to work at cross purposes, or beggar thy neighbour laws or tariffs. Apply coasian methods when making legal decisions or designing healthcare systems. Policy is cognitive, already in manual mode.
The level of the individual, however, is irreducibly non-cognitive. The real stuff are computations, attitudes, habits, and so we must adjust the metaethics required to avoid making a category error (in the same way it would be a category error to derive monetary policy from virtue ethics). Utilitarian criteria don’t disappear, however the role of virtue ethics play a much greater role, as do complex emotions like envy or love.