Fourth in the set, same format: small chunks, a one-line summary at the top of each, an analogy in every section. This one is different from the first three in one important way — it's a contested idea, not a settled body of knowledge, so where people disagree you'll find both cases laid out rather than a verdict.
Nothing marked yet. Each piece has a “Mark as read” button at the bottom — the page remembers, even if you close it.
The short version: Not infinite stuff. The usual claim is narrower: that a society could produce enough of what people genuinely need that access to it no longer has to be rationed by price or by work.
"Post-scarcity" gets used loosely enough that arguments about it are often two people using two definitions. It's worth pinning down before anything else, because the strong version is easy to dismiss and the weak version is hard to dismiss, and people switch between them constantly.
| Version | The claim | Status |
|---|---|---|
| Strong / absolute | Anything anyone wants, in unlimited quantity, free | Physically impossible. Nobody serious defends it. |
| Moderate | Material necessities — food, shelter, energy, healthcare, information — produced so cheaply they can be provided universally without rationing by price | The actual debate lives here |
| Domain-specific | A particular category becomes effectively unlimited (information, computation, light) | Already happened, several times |
Air. It's genuinely scarce in an economic sense — it's finite, it can be polluted, oxygen tanks cost money in a hospital. And yet nobody meters your breathing, nobody works a shift to earn air, and no one has ever had a serious conversation about air distribution policy. It sits so far past the threshold of enough that we stopped treating it as an economic good at all.
Post-scarcity, in the version worth debating, means moving other things across that same line. Not eliminating limits — just pushing them far enough out that the limit stops governing daily life.
The short version: Economics defines itself as the study of allocating limited means among unlimited wants. Post-scarcity proposes removing the premise the entire field is built on.
The standard textbook definition of economics is something close to: the study of how people choose to use scarce resources that have alternative uses. Scarcity isn't one topic within economics. It's the assumption that makes the subject exist.
From that assumption everything else follows. If resources were unlimited you wouldn't need prices, which exist to ration and to signal where things are most wanted. You wouldn't need opportunity cost — the recognition that doing one thing means not doing another. Trade-offs, budgets, efficiency, competition: all of it is machinery for handling not-enough.
Imagine a field devoted entirely to the study of traffic — signals, congestion pricing, road design, rush hour psychology. Now someone proposes that teleportation is coming. They're not offering a new traffic theory. They're suggesting the conditions that make traffic a subject may not be permanent.
That's why post-scarcity arguments tend to generate more heat than an ordinary policy disagreement. Economists aren't usually objecting to a detail — they're often objecting to a claim that their foundational premise has an expiration date. And the standard reply is worth taking seriously: even with teleportation, somebody has to decide who goes first, and the destinations are still finite.
The short version: Some scarcity is physical, some is created by rules we chose, and some is mathematically permanent. Confusing the three is the most common error in this whole conversation.
Three reasons you might not get a seat at a concert. The hall physically holds 500 people (physical). The promoter held back 200 seats to drive up prices (artificial). Or everyone wants to sit in the front row, and there is exactly one front row no matter how big you build the hall (positional).
The first can be engineered around — build a bigger hall. The second can be changed by decision, since somebody chose it. The third is a mathematical fact. Build a hall for a billion people and there's still only one front row.
This three-way split explains why the same person can be right and wrong at once. Someone arguing "scarcity is manufactured" has a real point about patents and zoning, and is plainly wrong about lithium. Someone arguing "scarcity is eternal" has an unanswerable point about status goods, and is dodging the question of why insulin costs what it does.
The short version: Several things that were luxuries within living memory are now effectively unlimited. That's the strongest evidence the trend is real, and the best place to start taking the idea seriously.
The argument doesn't rest on predictions. It rests on cases that already happened:
Imagine explaining to someone in 1900 that their great-grandchildren would leave lights on in empty rooms without thinking about it, throw away food routinely, and carry a device holding more text than any library they'd ever heard of — and would describe themselves as ordinary, not rich.
That's the core of the optimistic case, and it's not a small one. Post-scarcity has already happened to several categories of good. The disagreement is about whether the remaining categories work the same way.
The short version: The things that got cheap were mostly manufactured goods. The things that got dramatically more expensive — housing, healthcare, education, care work — are the ones people actually feel squeezed by.
Any honest account has to sit with this. Over recent decades, televisions, clothing, toys, and computers collapsed in real price. Over the same period, housing, college, medical care, and childcare rose faster than wages in most wealthy countries. People who feel poorer despite owning more devices are not confused — they're accurately reporting where their money goes.
Two explanations do most of the work, and they point in opposite directions politically:
Some work can't be made more productive. A string quartet takes four musicians and forty minutes today exactly as it did in 1800 — there has been no productivity gain in performing Beethoven. But those musicians' wages have to keep pace with the wider economy or they'd leave for other jobs. So the cost of the quartet rises relentlessly, because everything else got more productive. The same logic applies to teaching, nursing, childcare, haircuts, and therapy: labor-intensive services with no automation path, whose relative price must rise as manufacturing gets cheaper. On this account, rising service costs are a mechanical side effect of success, not a failure.
The other account says these sectors are expensive because of rules and market power: zoning and land restrictions that prevent building where people want to live, credential requirements that limit supply, administrative bloat, patent protection, and the fact that land itself can't be manufactured. On this view the costs are chosen, and different choices would produce different prices.
Your grocery bill drops every year while your rent doubles. You could reasonably describe yourself as living in an age of abundance or an age of scarcity, and you'd be accurately describing different aisles of the same store. Which aisle you shop in most determines which story feels true — and that, more than any disagreement about facts, explains why people talk past each other on this subject.
The short version: Some goods cost a fortune to make once and nearly nothing to copy. Those goods behave so differently from ordinary goods that normal economic intuitions break down on them.
Two terms carry this section. Marginal cost is what it costs to produce one more unit. Non-rival means my using it doesn't reduce your ability to use it.
An apple is rival and has real marginal cost: I eat it, you can't, and growing another one takes land and labor. A piece of software is non-rival with a marginal cost of essentially zero: my copy doesn't diminish yours, and the millionth copy costs the same as the second — nothing.
A recipe versus a meal. The meal is rival and costly — one person eats it, and another one requires more groceries and another hour of cooking. The recipe is non-rival and free to copy — a thousand people can hold it at once, and sharing it costs the originator nothing.
Humanity has been accumulating recipes for ten thousand years, and the recipes never get used up. The optimist's core argument is that an ever-larger share of what we value is turning out to be recipe rather than meal.
This is not hypothetical. Wikipedia replaced an industry that sold physical encyclopedias for hundreds of dollars, and did it with volunteers. Free and open-source software runs most of the internet's infrastructure and is given away. In these domains, something recognizable as post-scarcity already operates at scale.
The counterargument is that the zero-cost part is the least of it. A recipe doesn't feed anyone. Digital goods still require physical data centers, electricity, cooling, rare metals, and the labor of people who maintain all of it — costs that are real but invisible to the person downloading a file. And the hardest problems in human life, from housing someone to caring for them when they're sick, remain stubbornly in the meal category.
The short version: When something can be copied for free, nobody can profit from selling it — so we invented legal barriers to make copying illegal. That's intellectual property, and it's an intentional scarcity machine.
Here's the problem that forces the issue. A new drug might cost two billion dollars to develop and three cents to manufacture. If anyone could copy it the day it launched, the price would fall to roughly three cents, the developer would never recover the two billion, and nobody would fund the next one.
Society's answer was patents and copyright: a temporary, legally enforced monopoly. We deliberately make an abundant thing scarce so that someone has a reason to create it in the first place. The US Constitution is explicit that the purpose is "to promote the progress of science and useful arts" — the monopoly is a means, not a right.
A toll booth on a bridge that costs nothing to cross. Physically, unlimited people could walk over it today at zero cost, and every person turned away by the toll is pure waste. But without the toll, nobody would have financed the bridge. The toll isn't there because crossing is expensive. It's there because building was.
Every serious argument about drug prices, copyright terms, streaming rights, and open-source software is an argument about where to set that toll and how long to keep collecting it. Notice that both sides are usually right about something: the toll genuinely does fund bridges, and it genuinely does turn away people who could have crossed for free.
This matters for post-scarcity because it demonstrates something uncomfortable: technical abundance does not automatically produce experienced abundance. We have had the technical capacity to give every person on Earth a copy of nearly every book ever written for about two decades. We don't, and the obstacle is not technological.
The critics of artificial scarcity see it as proof that our institutions are the binding constraint, and that a great deal of present-day deprivation is a choice dressed up as a necessity.
The defenders point out that the abundance being distributed had to be created by someone, usually at great cost and risk, and that removing the incentive is a fast way to end up with much less to distribute. The pharmaceutical case is the sharpest version: cheap copies of drugs that already exist, or expensive funding for drugs that don't yet.
The short version: With enough cheap energy, most other material shortages become solvable problems. This is why energy cost is the single number the optimistic case rests on.
Almost every physical scarcity is, underneath, an energy problem. Fresh water is abundant if you can afford to desalinate seawater — and desalination is mostly an electricity bill. Metals are rarely gone; they're dispersed, and recovering them from seawater, low-grade ore, or landfills is again mostly energy. Fertilizer is made by pulling nitrogen from the air, which takes energy. Food can be grown indoors under lights. Carbon can be pulled back out of the atmosphere, expensively, using energy.
So a plausible route to broad material abundance runs through one variable: drive the cost of energy toward zero and a long list of separate shortages collapse into a single solved problem.
A universal currency that converts into any other resource. Water, metals, food, materials — energy buys all of them at some exchange rate. The rates vary and some are terrible, but the conversion always exists, because matter can be moved, separated, and recombined if you're willing to spend enough energy doing it.
Which means the question "will we run out of resources?" quietly transforms into "how cheap can energy get, and how fast?"
The trend line here is what optimists point to: solar photovoltaic costs have fallen by roughly 90% over the past decade and a half, following a learning curve where each doubling of installed capacity cuts the price by about 20%. Batteries followed a similar path. This is the empirical core of the modern abundance argument.
Energy isn't the only bottleneck. Conversion rates matter enormously. Extracting metals from seawater is theoretically possible and absurdly expensive in practice. Land, time, ecosystems, and human attention aren't purchasable with kilowatts.
Jevons paradox. Historically, when a resource gets cheaper we consume dramatically more of it rather than banking the savings. Cheaper energy has reliably meant more energy used, not the same amount used at lower cost. Whether abundance satisfies demand or simply expands it is an open empirical question with a discouraging track record.
Infrastructure is slow. Cheap generation is not the same as a built-out grid, storage, and transmission. Those take decades, permits, minerals, and political will.
The short version: If machines do the work, goods can be abundant without human labor. Whether that means liberation or mass unemployment is the oldest live argument in the field, and both sides have real evidence.
Post-scarcity requires that things get made without a corresponding amount of human toil, which makes automation central. The historical record is genuinely reassuring and genuinely incomplete.
Around 1900, roughly 40% of American workers were employed in agriculture. Today it's under 2%. That is a near-total automation of the largest job category that had existed throughout human history — and unemployment today is not 38%. The displaced labor moved into work that hadn't been invented yet. The economics term for assuming otherwise is the lump of labor fallacy: treating the amount of work in an economy as a fixed quantity to be divided up, when in practice new wants generate new work indefinitely. On this reading, automation raises output per person, which is the definition of getting richer, and the transition — however brutal for individuals — resolves.
Every previous wave automated muscle, and humans moved up into cognitive work. A wave that automates cognition removes the destination. "This time is different" has been wrong repeatedly, which is why it's a weak argument — but it is not a logically impossible one, and the mechanism being proposed is different in kind. There's also solid evidence that the transitions matter enormously even when the endpoint is fine: the displaced farmworker doesn't become a software engineer, and the adjustment can take a generation and wreck particular regions permanently. The polarization of work into high-wage and low-wage ends, with the middle hollowing out, is a documented feature of recent decades rather than a forecast.
Horses. Mechanization made human farm labor vastly more productive, and humans moved on to better work. It did not do that for horses — the US horse population collapsed once tractors arrived, because horses had exactly one economic capability to offer and it was replaced.
The whole argument turns on whether humans are more like the farmers or more like the horses, and that in turn depends on whether there is some capability humans have that machines will keep failing to replicate. That's an open question about the future of technology, not a question economics can settle on its own — which is precisely why confident answers in either direction should be treated with suspicion.
The short version: Producing enough and everyone getting enough are separate achievements, and history shows the first does not deliver the second. This is where the real argument is.
The world currently grows enough food to feed everyone in it. Hundreds of millions of people are nonetheless undernourished. This is not a production failure.
The economist Amartya Sen made the sharpest version of this point studying famines — including the Bengal famine of 1943, which killed millions in a year when food availability had not collapsed. What collapsed were people's entitlements: their ability to command food through wages, land, or trade. Sen's conclusion was that famines are overwhelmingly failures of distribution and political accountability rather than of harvest, and that democracies with a free press essentially don't have them.
A banquet hall with more food than the guests could possibly eat, and a locked door with a man selling tickets. The abundance is entirely real. So is the hunger outside. Nothing about the quantity of food resolves the situation, because quantity was never what was rationing it.
This is why "we could produce enough for everyone" is a much weaker statement than it sounds. We already can, for several things, and have for decades.
The point generalizes into the central question of the whole subject: who owns the machines? If automated production is owned narrowly, abundance could coexist with deep deprivation — a world of enormous output where most people have no wage to claim any of it and no capital either. Nothing in the technology decides this. It's decided by ownership rules, taxation, antitrust, labor power, and politics.
One: Markets have distributed previous abundances remarkably well without anyone planning it. Light, food, and clothing got cheap and then got broadly available, through ordinary price competition, and the poorest people today have goods no king could buy in 1800.
Two: That distribution ran largely through wages — people got a share of the abundance by selling their labor to produce it. If labor's role shrinks, the main channel through which ordinary people have historically claimed a share narrows with it, and no equivalent channel currently exists.
The short version: If wages stop being the way people claim a share of output, something has to replace them. Here are the main candidates and the strongest objection to each.
These are policy proposals with active constituencies and active critics. They're listed with their best case and their sharpest objection rather than ranked.
| Mechanism | The case for | The case against |
|---|---|---|
| Universal basic income Cash to everyone, unconditional | Simple, no bureaucracy or means-testing, no work disincentive cliff, respects people's judgment about their own needs | Expensive at meaningful levels; may be inflationary where supply is constrained (housing); could justify dismantling services that were doing more; trials so far have been small and temporary |
| Universal basic services Free healthcare, transit, housing, education | Targets the sectors where costs actually bite; avoids handing cash into a market that may just absorb it as price | Requires state capacity to deliver well; less flexible; historically uneven results |
| Job guarantee Government as employer of last resort | Preserves the social role of work and the dignity many people attach to it; produces public goods | Doesn't fit a world where labor genuinely isn't needed; risks make-work; administratively heavy |
| Public ownership / sovereign wealth funds Everyone holds a stake in productive capital | Directly addresses "who owns the machines"; Alaska and Norway are real working examples | Politically fraught; exposes citizens to market risk; governance capture is a live danger |
| Shorter work week Spread remaining work thinner | Distributes both income and leisure; trials report improved wellbeing with maintained output | Only works where output isn't proportional to hours; doesn't help those already outside work |
A town where the mill was the only employer and the mill just installed machines that do everything. Something has to replace the paycheck, and the townspeople's options are roughly: send everyone a check (UBI), make the necessities free (basic services), invent new jobs (job guarantee), give everyone a share of the mill itself (public ownership), or have everyone work one day a week (shorter hours).
Each answer implies a genuinely different town — different in what people do all day, what they owe each other, and who holds power. These are not technical fixes with a correct answer waiting to be calculated. They're choices about what kind of place to live in, and reasonable people land in very different spots.
The short version: Keynes predicted in 1930 that his grandchildren would work fifteen-hour weeks. He got the productivity right and the behavior wrong, and understanding why is the most interesting puzzle in the subject.
In 1930, John Maynard Keynes wrote a short essay predicting that within a century the economic problem would be solved, and the main challenge facing humanity would be how to fill its leisure. He expected roughly a fifteen-hour work week.
His productivity forecast was close to right — output per hour in wealthy countries multiplied several times over as he anticipated. His forecast about how people would use it was badly wrong. Hours fell far less than predicted, and the surplus went overwhelmingly into consuming more rather than working less.
A family gets a dishwasher, a washing machine, and a microwave — hours of drudgery eliminated every week. They could bank the time. Instead the standards rise: clothes get washed after one wearing, meals get more elaborate, the house is expected to be cleaner. The labor-saving devices arrived and the labor didn't leave, because "enough" moved.
That's the pattern Keynes missed, and it's the strongest practical argument against post-scarcity. It suggests wants aren't a fixed target that abundance eventually overtakes. They may be a moving one that keeps pace with whatever we produce.
There's a reasonably robust finding in psychology that involuntary unemployment is associated with worse wellbeing than the income loss alone accounts for, while voluntary retirement generally is not. The difference appears to lie in structure, social contact, identity, and a sense of contributing — not in the work itself.
That cuts both ways, which is why it's worth sitting with. It suggests "freedom from work" is not automatically experienced as freedom, and that a society removing paid work would need to be deliberate about what replaces those functions. It also suggests that what people need is not a job specifically, since retirees get those things elsewhere — which is an argument that the functions are separable from employment if a society sets out to separate them.
The short version: Four objections that the strongest critics make, none of which is easily dismissed. Anyone who finds the idea appealing should be able to state these before arguing for it.
A weather forecast from someone who has predicted rain every day for fifty years. Eventually they'll be right, and that won't make them a forecaster. The honest position is that a good argument and a poor track record can coexist, and you have to evaluate the specific mechanism rather than the vibe.
The steelman of the optimistic case is that the mechanism this time is specific and measurable — energy learning curves, automation capability, information goods — rather than a general faith in progress. The steelman of the skeptical case is that every previous forecaster thought that too.
The short version: This idea has no fixed political home. It has been argued for by Marxists, libertarian technologists, liberal policy writers, and science fiction novelists, often in mutually incompatible terms.
Knowing the lineage helps, because the same phrase carries very different cargo depending on who's using it.
Four people looking at the same rising tide. One says it proves the boats will all float. One says it proves whoever owns the harbor will own everything. One says look at the beach erosion. One says we've been promised a tide before and the water never came. They are not disagreeing about the water level — they're disagreeing about what follows from it, and mostly for reasons that predate the tide.
The short version: Turning a vague hope into a checklist of testable conditions is the most useful thing you can do with this idea. Here are the conditions, and how you'd check each one.
Rather than asking "will we reach post-scarcity," which can't be answered, ask what would have to hold — and then watch those specific indicators.
| Condition | What you'd watch |
|---|---|
| Energy becomes very cheap and very clean | Cost per kilowatt-hour including storage; actual deployment rates, not just prices; grid buildout |
| Physical labor is genuinely automatable at low cost | Robotics in unstructured environments — construction, elder care, agriculture — not just factories |
| Material throughput decouples from output | Whether resource use and emissions fall while economies grow, in absolute terms, not per dollar |
| The gains get distributed | Share of income going to labor vs capital; wealth concentration; whether any non-wage distribution channel is actually built |
| The expensive sectors get cheaper | Real cost of housing, healthcare, education relative to median income — the Baumol problem from section 5 |
| People find it livable | Wellbeing, social connection, and purpose among populations already outside paid work |
Instead of asking whether your team will win the championship, list the six things that would have to happen for them to win — a healthy quarterback, an improved defense, a favorable schedule — and then watch those. You lose the pleasure of a confident prediction and gain the ability to update honestly when a specific thing does or doesn't happen. It also protects you from the most common failure here, which is letting a general mood about technology stand in for an assessment of any particular claim.
Notice that these conditions aren't all technological, and that the last three are arguably harder than the first three. That's the shape of the whole subject in one table.
The short version: Five habits that separate a useful conversation about post-scarcity from an unfalsifiable one.
Treat it as a live hypothesis rather than a destination or a delusion. Hold the strongest version of both cases in mind at once, and let specific evidence move you — a cost curve that flattens, a distribution mechanism that gets built or doesn't, a category of good that crosses the line into abundance the way light did.
The reason to learn this subject isn't to arrive at a prediction. It's that the questions underneath it — what's genuinely scarce, what's scarce by our own arrangement, who gets what and why — are worth being able to ask clearly, and they'd be worth asking even if no technology ever changed again.
Deliberately mixed, because reading only one side of this makes you worse at it: