”Telescopic altruism” is when progressives are supposed to care about distant strangers at the expense of those close to them. Scott Alexander recently argued against the concept (without quoting anyone specific making the claim). He countered that concern for distant and proximate others is correlated rather than opposed: the people who object to Israel’s actions in Gaza also support school lunches, the people who protest factory farming would also protest if a billion of their friends were caged.
When much of the developed world’s population was subjected to inhumane isolation during COVID, the protests came largely from the moderate right, not from the progressives Scott is defending. Serious proposals that might have actually helped, such as variolation, challenge trials, and mass deployment of far-UVC sterilization, were largely ignored, while medical remedies and mitigation measures were politicized in bad faith on all sides. What the correlated altruism population mostly did was follow orders and enforce compliance on their neighbors.
Local care pays for itself: your neighbor helps you raise your barn, you help them with theirs. Concern that flows from identification with an altruistic collective rather than from relations of shared production or exchange has to be paid for by something else.
Warm applesauce and cold ICE
I have neighbors with toddlers. We finally met them because my three-year-old asked why we send him to preschool a few days a week. I offered three reasons:
With two toddlers we need some help. It’s too much work for mama and me to do well all the time.
It’s good to get information from different people than just your parents.
It’s helpful to make friends outside your family, especially if you ever want to have children of your own.
All three were enthymemes, so I explained their shared hidden premise: we don’t have friends or family close enough to meet these needs adequately, and while we might want to befriend our neighbors to help with this, we haven’t managed to yet.
So one night, when we were bringing home a pizza, he told me that he wanted to go over to a neighbor’s house for dinner. I think he was also trying to apply some messages about neighbors from children’s television he’d recently watched. I explained why this wasn’t appropriate if we weren’t invited, and also I was tired and wanted to stay home. A modern-day Abraham, I bargained him down to bringing presents to two of our neighbors. One got a chocolate covered Oreo; the other household, with the toddler, got a toy car and a note. They texted their thanks, and I began to try to figure out how to befriend them further.
They told me that their child doesn’t do well with gluten. I invited them to come over and make fresh applesauce with my toddler. I chose applesauce specifically because it was something their child could eat. They responded to the invitation not by accepting or declining, but by texting me a flyer for a Stop ICE rally.
I don’t know whether they personally know someone affected by ICE’s recent activity, because they don’t really talk much with their neighbors. Which is itself the point. Perhaps they couldn’t tell me how they know ICE is a problem for anyone they’re in a position to help, because they don’t relate to the problem that way. They know it’s a problem the way one “knows” crime is declining: through convergence of indicators produced by an abstraction layer, not through contact with the phenomenon. Or the way one “knows” crime is increasing: through media that present themselves as informing you about the world, but function in practice as a way to calibrate your anxiety to the perceived norm.
[1]
They’ve created structural distance between themselves and the people next door by adopting identities that put them closer to an unaccountable system of political action than to their literal neighbors.
[2]
Unlike friends I’ve made online, these neighbors were not selected for being unusual or for being very online. They’re just the people who happened to move in on the corner. They’re responding to the same pressures that shape nearly everyone’s engagement with the world in a modern economy.
But progressives support school lunches! If progressive concern for distant others isn’t about sacrificing those close to them, then a fortiori we should expect that their own children, over whom they have much more direct influence, eat enough for lunch. Do they?
My two toddlers are both around the 99th percentile for height and weight, even though my extended family aren’t particularly large people. So I can be expected to say no, other people’s children are not getting enough lunch, progressives included. The same class that supports school lunch programs produces pediatricians who tell me to withhold food from my healthy child. My partner grew up around children from much wealthier and classier families who would come to her house to eat, because at her house they could access fresh fruit freely, unlike at home. One family I know doesn’t seem to salt their toddler’s food or feed him much meat, and complained to me that he undereats to the point where it impacts his sleep, but visibly blanked out when I suggested they try a nutritionally dense ice cream such as Van Leeuwen French. Another family has repeatedly expressed surprise, but not much curiosity, when their preteen ate the adult food I prepared (e.g. pasta in meat sauce) instead of insisting on his usual buttered noodles.
The physician and the lens
Consider a physician whose body is visibly rotting. You look at their patient charts and the numbers seem fine. But at some point the body becomes evidence that the numbers are misleading; that whatever process is generating those outcomes isn’t tracking health because it cares about health the way patients do. Because if it were, the physician’s own body would implement that understanding. A physician suffering from an injury or terminal cancer through no fault of their own might still serve patients well. But we want heuristics like “is this physician healthy” precisely because we can’t fully verify the track record directly. If the legible metrics were adequate, we wouldn’t need other controls.
We only know about things in the world through our bodies interacting with them. (This is a crucial proposition in Spinoza’s Ethics.) A poorly ordered body is like a badly ground lens. The looker might try to compensate in a principled way for the distortions the lens introduces, but if the looker is disordered, their adjustments are likely to be distorted as well. We rely on those close to us to help us become aware of and interpret our world, and if we are dissociated from our relationships with them, we have a bad lens and a bad error-correction system.
An organized person who knows how to care about themselves and their environment is doing one sort of cognitive-emotional operation when becoming aware, abstractly and indirectly, of people they know about only through institutional mediation. They have some idea of the instrumentation by which they know of such people. And their beliefs about what is good for others can be checked against their own functional needs, rather than drifting helplessly with legible approval metrics, which can be checked for consistency but not soundness. The only calibration available to a human being is the life they are actually living.
When someone’s concern with others takes place in a story that includes their own self and problems, I can credit that concern fully. I know a visceral massage practitioner, Valentin, who’s worked out his own methods and tools. Part of his interest is in sports medicine, and he’s a genuine amateur athlete who works extensively on his own body. He helped my partner, who had longstanding gut issues, unkink her abdominal muscles, which probably made the difference between a prolonged and painful labor, and arriving at the hospital fully dilated. His interest in helping others is visibly continuous with his interest in his own physical functioning, and his recommendations can be checked against his own condition.
I trust concern moderately when it comes from demonstrated abstract competence applied to a domain the person finds intrinsically interesting, like a mathematician who helps others by doing good math, or a programmer or engineer who wants to design something excellent with integrity. But I trust it very little when the primary motive is altruism directed at people the altruist has no particular reason to understand.
This is not a complaint about “virtue signaling.” Nor is it a proposed alternative on the same level, an inverted, evil version of an inauthentic virtue one might want to signal. This is a serious account of virtue, in the sense of functional integrity. It’s not about how to be a good little boy or girl and get on Santa’s nice list, or how to be naughty and receive combustible hydrocarbons gratis; it’s about the psychic capacity to appropriately employ means to ends. The difference is between virtue so defined and compliance with the norms of a concerned-seeming class.
The wrong instruments
Scott offers evidence for “correlated altruism”: people who care about distant others also tend, at the population level, to show indicators of caring about proximate others (lower divorce rates in blue states, lower child abuse rates, support for school lunch programs). But every one of these is a population-level aggregate largely explained by (or subsumed in) political affiliation. The difference in divorce rates, as a commenter called “bean” points out, reflects different patterns of marriage and cohabitation more than different levels of devotion. In Oklahoma, a young couple who’ve been together three years and it isn’t working get divorced. In California, the equivalent couple were never married. The child abuse data almost certainly reflects reporting standards and agency effectiveness rather than actual rates of abuse. Bean notes that adjacent, culturally similar states show wildly different rates, with a distribution implying extreme below-average outliers that are simply not plausible as real data.
These are exactly the kind of convergence that looks robust until you check whether the instruments share a systematic distortion. Are progressives kinder, or are our metrics for kindness progressive?
How distance-altruism pays for itself
The examples above are drawn from progressive culture because that’s what I live in and can observe directly. But the dynamic is a general feature of modernity. It affects anyone whose engagement with the world is primarily mediated by institutions rather than direct relations, which in modern economies is nearly everyone.
Vitalik Buterin built Ethereum, a platform for decentralized contracts that don’t require trusting intermediaries. It worked. Then the speculators came.
[3]
Defending it would require an interest not only in cryptographic protocols, but in adversarial social dynamics. Buterin understood his work as a public good rather than as self-defense, so the defense didn’t get done.
Elon Musk built SpaceX to put rockets in orbit and Tesla to make electric cars. Both still function, because he still wants to put rockets in orbit, and he still wants to make electric cars (as a substrate for self-driving car software). He bought Twitter to secure a communications channel, but didn’t have or develop an adequate theory of what broke the tool Jack Dorsey built (and then the next tool Jack Dorsey built to replace it), so Twitter decayed again into a gracefully censored platform. There were new censors with new prejudices, but the wrong kind of speech was still shadowbanned. Musk’s DOGE wasn’t trying to divert government funds to support the state capacities he specifically needed. He wasn’t cutting down specific obstacles in his way. He was trying to be a good citizen, to reduce “waste” in the abstract. Most government waste is disputed by people whose salary or identity compels them to dispute it, and DOGE built no instrument to distinguish genuine objections from interested ones.
A monocrop doesn’t turn into parasites and pests on its own; they show up and eat it. Creating a big new public good is similar. If you still use the thing, defending it is just part of using it. If you don’t, it’s thankless extra work to keep spraying a field for pests when you don’t depend on the harvest.
And in memetic space, unlike a physical field, the pests are imitative. They present as more of the crop. An undefended altruistic project doesn’t visibly decay to anyone who isn’t trying to use it. It fills up with people who perform altruism, because that’s what the niche rewards. The field looks green and productive, until you try to harvest the wheat and discover it’s tares.
[4]
What stabilizes is not the original project but an altruism-performing class that sustains itself by purchasing participants’ willingness to overlook things “for the greater good”. The “greater good” is the currency in which silence about the infestation is bought.
This is why the track record of the institutionally-mediated altruism class compares poorly to communities like the early Puritans and Quakers, who organized around reciprocal direct accountability. Your Puritan neighbor who might reproach your ungodly conduct was also the neighbor you traded with, whose own conduct you could scrutinize, who depended on your good opinion for their standing in the community. You depended on each other’s cooperation for your own survival, and on each other’s children as potential mates for your own. The judgment stayed calibrated against shared reality rather than against institutional imperatives, because the person judging you had to live with the consequences of being right or wrong.
Robinson Crusoe and the cannibals: the epistemological case for liberal restraint
This constrains positive institutionally-mediated altruism much more than negative duties. Negative duties (don’t harm, don’t intervene where you lack standing) work at any distance, because they require only recognizing the limits of your own knowledge. I exercise this kind of restraint constantly with my own children, whom I know far better than I know any foreigner. Much of their development depends on my judging when not to intervene, when to let them struggle with a banana or a stuck zipper rather than solving the problem for them. But to owe others help, to have a positive duty to improve their conditions, we first need to understand them and their conditions well enough to know what would help them. This is classical liberalism arrived at not through rights theory but through epistemology, through asking what it is possible to know well enough to act on.
In Daniel Defoe’s novel Robinson Crusoe, the castaway finds himself alone on an island where groups of cannibals periodically arrive to kill and eat captives. His immediate impulse is to attack them. But he reasons it through: no one has appointed him judge over these people; they aren’t threatening him or his interests in any way that demands a response; he has no reasonable hope of actually rescuing the victim against a group that outnumbers him. Attacking would mean satisfying his moral feelings at the cost of a pointless mass murder. Crusoe’s restraint comes from recognizing what he doesn’t know and what standing he doesn’t have, and this recognition is available to anyone, at any distance, without local grounding.
Crusoe thereby avoided not only material danger to himself and others, but a whole shadow realm of perversion. When you fight someone, you awaken and attract two kinds of attention in yourself and others. One rationally understands the fight as relevant to some other interest they mean to protect or pursue. The other simply identifies its interests as winning (or losing) this kind of fight.
Who works at a grade school, prison, or psychiatric ward? Some no doubt mean to help those under their care. Many are attracted by pay and working conditions favorable enough to compensate them for spending their time and effort on those in their custody. And for others, their carceral duties to wield power over others are assessed not as a cost, but as a benefit. When some of our faculties are persistently thwarted, they learn helplessness, and we learn to spare ourselves the effort of employing them. And when others meet with success, we are more inclined to return to those wells. This is why well functioning custodial institutions are vigilant about abuse of power. It is no accident; it is an attractor.
Enlightened self-interest seeks out fewer fights than altruistic coordination at scale, because the coordination has to purchase compliance through loyalty tests, and loyalty tests are defined by, or exist to define, enemies. The benefit of avoiding fights is not only from avoiding the direct harms fights cause, but in remaining the sort of person with interests beyond the fight.
What would change my mind
The most reliable indicator of whether a community’s way of life is functional is whether it reproduces its capacities. Fertility is very hard to game, and damage to an organism’s capacity for self-maintenance shows up in reproductive fitness within a few generations, even if the earlier generations otherwise appear happy and healthy — the psychic equivalent of Pottenger’s cats. And if a community isn’t reproducing at replacement but still persists, it is either extracting resources from a productive population elsewhere, being sustained as a tool by something that finds it useful, or disappearing.
A good counterexample to this heuristic would be a community organized primarily around concern for institutionally-distant others that also reproduces above replacement, maintains longer-than-usual healthspans, and sustains itself without harming others or working on projects its members believe will destroy the world. I don’t know of any such community. The most obvious candidate, the Effective Altruism and Rationalist communities, fails the last criterion: EA served as an intake funnel for AI capabilities research that its own members believed would endanger humanity, and continues to do so. Whether or not they were right about the danger, the community’s own stated beliefs condemn its track record.
The communities I know of that do pass this test, the Satmar and the Amish, are organized around exactly the kind of reciprocal direct accountability I’ve been describing, and they reproduce well above replacement. The Satmar maintain their own rabbinical courts (batei din) that adjudicate civil and commercial disputes within the community, with enforcement through social consequences: a ruling against you is backed not by state power but by the fact that everyone in your life will know about it. The Amish practice mutual aid through the congregation, with elders who know the parties personally mediating conflicts. In both cases, the person judging your conduct is embedded in the same web of obligations you are, which keeps the judgment grounded in shared reality rather than abstract principle.
On healthspan, the Amish had dramatically longer lives than other Americans a century ago (over 70 years when the US average was 47), and while overall lifespan has since converged as modern medicine closed the gap, the Amish maintain notably better late-life health: lower rates of cancer, cardiovascular disease, diabetes, and obesity. Amish men over 40 have significantly lower mortality from cancer and cardiovascular disease than the surrounding population. The general population caught up on raw longevity through medical intervention, but the Amish advantage in health quality persists.
Israel is an interesting anomaly: a modern, technologically integrated society reproducing slightly above replacement. But the most persuasive explanation I’ve seen, NonZionism’s “trickle-down natalism,” attributes Israel’s fertility to the cultural influence of the Haredim, a community with precisely the direct-accountability structure this thesis predicts would be necessary. I can conceive of other functional arrangements, in which a relatively celibate governing elite supports the fertility of the population it recruits from. Before Gutenberg and Luther, the Roman Catholic Church enjoyed considerable success.
Full integration into the modern global economy may itself require passing the kind of loyalty tests that corrode the relations of direct accountability on which genuine concern depends. If so, the best achievable arrangement may be something like Israel’s uneasy compromise: a society that preserves a directly-accountable core while participating in the global system selectively, accepting the tension rather than resolving it.
During the COVID-19 pandemic my father called me up one day and said I should be extra careful because on the news they said a COVID-related number went up in his state. I asked what number, what was the numerator, what was the denominator, what was being measured. He didn’t know and didn’t seem bothered by this. So the number wasn’t being used as part of a structured quantitative model, but as a social prestige claim, part of a process by which he calibrated to what he perceived as a socially conforming level of anxiety. Anecdotes likewise contain local information, but people reading or watching the news or social media might use them not to draw specific structured local inferences, but to, again, calibrate their level of anxiety to the perceived norm. ↩︎
They did share their sled in the blizzard, and months later we finally managed to visit them in their home. They’re not monsters, just crazy like everyone else. ↩︎
Anatomy of a Bubble. For a distinct but related perspective, see Geeks, MOPs, and sociopaths in subculture evolution. ↩︎
Matthew 13:24-30, KJV.. Though on the other hand, rye seems to have originally been a weed infesting wheat and barley fields, that was accidentally bred into a crop. By removing all the obvious weeds and replanting whatever of its seeds made it into the seed corn, farmers selected for similarity to crop grains (see also Sun et al. 2022). Oats might have developed the same way. ↩︎
Discuss Read More