I know A Modest Proposal is satire. I read it in tenth grade like everyone else. Swift wrote it in 1729 to indict the English landlords who were starving Ireland, and the joke is that his arithmetic arrives somewhere no decent reader will follow. The problem is that we now have an intellectual movement built around the proposition that when moral arithmetic produces a conclusion that feels horrifying, the responsible response is to inspect the arithmetic rather than recoil from the answer.
Longtermism asks us to evaluate decisions partly by their effects on the enormous number of people who may exist in the future. Effective altruism asks us to compare interventions quantitatively instead of letting sentiment decide which beneficiaries count. Population ethics asks us to include people whose existence depends on the policy being evaluated. Existential-risk research asks us to preserve civilization’s ability to survive catastrophes severe enough to make ordinary moral intuitions irrelevant. Rationalism tells us that disgust is not an argument. Taken together, the broader TESCREAL worldview supplies almost every premise Swift requires, which means A Modest Proposal deserves another hearing.
A reductio works only on readers who reject the conclusion before they check the premises. For most of the history of moral philosophy, that was a safe assumption. It is less safe now. Parfit introduced the Repugnant Conclusion as a reductio of total utilitarianism, and in 2021 twenty-nine philosophers, several associated with longtermism, signed a statement arguing that a theory’s implying it is not sufficient reason to reject the theory. That is the correct standard. An argument arriving somewhere strange tells us nothing by itself. The question is whether the premises hold. If they do, intellectual seriousness requires following them where they lead. Swift was writing for readers who would stop at the conclusion. We should not.
Start with the policy as written. Swift estimated that about 120,000 children were born each year to poor Irish parents, proposed reserving 20,000 for breeding, and selling the remaining 100,000 at one year of age, well nursed, for about ten shillings each. Everyone fixates on the deaths. Fine. Put them in the ledger. Then put everything else in the ledger too, because the program also creates lives.
Demand for well-nursed one-year-olds gives poor families a financial incentive to have children they otherwise might not have had, and Swift is explicit that those children should be kept plump and healthy until sale. On a total view of population ethics, a well-fed year of infancy contains positive welfare. If the policy causes additional such lives to exist, that welfare counts. You do not get to count the death and then pretend the preceding year never happened. Swift was, in effect, proposing a high-welfare certification standard.
He also did something many charities still fail to do: specify measurable outcomes. Swift lists six benefits. Setting aside the first, which concerns reducing the number of Catholics and has not aged well, the others are straightforward: tenants acquire something of value with which to pay rent, fifty thousand pounds a year enters the national economy, parents are relieved of the cost of raising children beyond infancy, taverns receive new business, and domestic violence declines because husbands become “as fond of their wives, during the time of their pregnancy, as they are now of their mares in foal” and stop beating them for fear of causing a miscarriage. That is a more explicit theory of change than several organizations I have funded.
The pricing deserves the same seriousness. Ten shillings in 1729 comes to roughly $120 today. GiveWell’s top charities spend something on the order of $5,000 to save a child’s life. That is a fortyfold discrepancy. Either Swift’s market was catastrophically undervaluing children or the modern charity sector is paying a forty-times premium. People become uncomfortable when I put it that way, but discomfort does not reconcile the numbers.
The strongest argument for Swift, though, is the one unavailable to him in 1729: existential risk. Consider an abrupt sunlight-reduction event caused by nuclear winter, a supervolcanic eruption or an asteroid impact. Conventional agriculture collapses, supply chains disappear and stored food runs down. At that point, the morally relevant variable is calories. Researchers studying resilient foods have therefore examined seaweed, single-cell protein, mushrooms, greenhouse agriculture and sugars derived from wood. They are right to do so. Civilization should prepare in advance for circumstances in which familiar food systems cease to function.
One obvious source of protein remains excluded almost entirely because people find the thought unpleasant. That is squeamishness disguised as risk management. Human biomass has a property that should interest anyone thinking seriously about resilient food systems: its production capacity scales with the same population whose survival creates the demand. If we are willing to model every other source of emergency nutrition, there is no principled reason to exempt this one from study simply because it produces a stronger emotional reaction.
The recipes should be preserved. The nutritional data should be preserved. The logistics should be modeled. The husbandry requirements should be documented. Emergency legal authorities should be worked out in advance. Allocation mechanisms should be designed before famine rather than during it. The moral case should be settled now, while we have enough food to discuss it calmly. This is what preparedness means. We maintain seed vaults because crops may fail, continuity-of-government facilities because governments may collapse, and pathogen repositories because destroying knowledge can be more dangerous than containing it. Militaries plan for scenarios everyone hopes never occur. A food source does not become less useful because we hope never to use it.
Once the longtermist calculation enters, the case becomes overwhelming. Suppose preserving this option improves the probability of civilizational recovery after a genuine extinction-level food shock by one basis point. That probability gain must be multiplied by every generation enabled by recovery: billions of lives, trillions potentially, and more if technological civilization persists beyond Earth. Against that denominator, the moral cost of keeping one additional emergency food pathway available is negligible.
People will say this is monstrous. That is not an argument. People will say there are lines we should not cross, but lines require justification too. Which lines? Why those lines? Under what population assumptions? At what extinction probability? With what discount rate? If you cannot answer those questions, you do not have a moral principle. You have a reflex.
None of this makes Swift’s program costless. It kills children. Each death forecloses a life at age one, and some fraction of those lives would otherwise have been long and good. The harm is real, but the fact that a cost is horrifying does not exempt it from comparison. We already trade lives constantly. Medical triage does it. Transportation policy does it. Pollution regulation does it. Workplace safety rules do it. Cost-effectiveness analysis does it. Every policy that exchanges money, convenience or economic output for changes in mortality risk contains an implicit valuation of human life.
What changes with Swift is that the tradeoff becomes visible. We tolerate death when it arrives statistically, when the causal chain is long enough, or when nobody has to look at the body. Swift removes the abstraction, and suddenly everyone discovers a deontological commitment they did not know they possessed. That inconsistency deserves more scrutiny than his proposal does.
The deeper problem with A Modest Proposal is that Swift accidentally anticipated an intellectual culture capable of surviving his punchline. His readers were expected to reach the sentence about eating children and recoil so violently that the preceding economic logic collapsed with it. Modern consequentialism has spent half a century teaching us that this is precisely the wrong way to reason. Population ethics tells us to count created lives. Effective altruism tells us to compare outcomes rather than intentions. Longtermism tells us to include future generations. Existential-risk research tells us to preserve options whose value becomes apparent only in catastrophe. Rationalism tells us not to confuse emotional repugnance with logical refutation.
If we believe all that, then Swift’s conclusion does not make the argument fail. It tells us where the argument goes.
Every generation has had moral assumptions that later generations found indefensible. Ours may be the belief that some conclusions are so unpleasant that we are entitled to stop calculating. We are not.
My wife, who is due in March, asked me not to read the rest of this aloud. I reminded her that Swift’s program was means-tested, and that at our income our daughter would never have been eligible.