The Argument

Foundations

The case for Impactism, in six steps. Each step is made in plain words first, with the studies underneath for anyone who wants to check. Nothing here asks you to take our word for it.

One promise before we start: where the research is contested, we say so on the page. A movement that needs you to skip the footnotes doesn't deserve you.

Every society keeps score.

Wanting to matter in the eyes of other people is not a modern disease. It is one of the oldest things about us. Psychologists who reviewed the evidence across cultures, ages, and personalities concluded that the desire for status — for being valued by the people around you — is a fundamental human motive. Not an American thing. Not a capitalist thing. A human thing.

Most plans for a better world miss the next part: abundance does not switch it off. The economist Fred Hirsch showed that when material things become plentiful, competition doesn't end — it moves to the things that stay scarce because their whole point is standing out. A bigger house. A better seat. He called these positional goods: goods that work like positions in a race, where one person moving up means someone else moves down.

You don't get a vote on whether your society has a yardstick. You only get a vote on what it measures.

Go deeper — the evidence for step 1

The status review: 2015, Psychological Bulletin. Anderson, Hildreth and Howland examined the empirical literature against the criteria for a "fundamental motive" and concluded the desire for status meets them — it appears "across individuals who differed in culture, gender, age, and personality." Well-being, self-esteem, and even physical health track the status others accord us.

Positional goods: Fred Hirsch, Social Limits to Growth (Harvard University Press, 1976). His insight is why "just make everyone rich" never ends the race: the scarcity of a front-row seat is social, not physical, so no factory can relieve it. Thorstein Veblen described the same engine in 1899 — "conspicuous consumption," spending whose purpose is display — before most of modern consumer culture existed to prove him right.

Anderson, Hildreth & Howland (2015), "Is the Desire for Status a Fundamental Human Motive?", Psychological Bulletin 141(3) · Hirsch (1976), Social Limits to Growth · Veblen (1899), The Theory of the Leisure Class
Right now the score is money — a brilliant tool, and a biased judge.

Let's be fair to money first, because the case against it is usually made dishonestly. Money works. At bottom it is a kind of shared memory — a portable record that you did something for someone, which a stranger on the other side of the world will honor. Economists have a formal version of this: anything money does, a complete record of who-helped-whom could do. Markets running on that ledger have accompanied the largest decline in extreme poverty in recorded history. We are not here to burn the ledger.

We are here to point at what it leaves out. Money records what was sold. It has almost no memory of what was given. The teacher who turns a classroom around, the open-source engineer whose free code runs half the internet, the daughter caring for a father with dementia — the ledger writes them down small, or not at all. That is not a moral failure of any person. It is a recording failure of the unit.

And the unit pays out less than people think. Read the famous money-and-happiness research without the meme, and it says this: from $30,000 to $60,000 of income is one step up in felt life. To feel that same step again, you need $120,000. Then $240,000. Each doubling buys roughly the same increment — researchers call this rising with the logarithm of income — so the returns never quite reach zero, but they shrink relentlessly. Meanwhile the things that keep predicting a life people are glad to be living — purpose, autonomy, being needed, belonging — sit mostly outside what the ledger counts.

Go deeper — the money-and-happiness research, reported honestly

What we are NOT claiming: that money stops mattering at $75,000. That famous "plateau" (Kahneman & Deaton, 2010) applied only to one coarse measure of daily mood, and it failed to replicate. Killingsworth (2021) collected 1.7 million real-time reports from 33,391 working US adults and found no plateau. In 2023 the two sides ran a joint adversarial re-analysis and settled it: happiness rises with log income for most people across the whole measured range; it flattens above roughly $100,000 of household income only for the unhappiest ~20%. All of this is correlational, US-based, and we cite it as such.

What survives every side of that fight: the logarithm. Each doubling of income buys about the same increment of well-being — which is also the strongest moral fact in the literature, because it means a dollar does enormous work in a poor life and almost none in a rich one. Money causally helps, too: Swedish lottery winners show durable gains in life satisfaction up to two decades on. We say so plainly. And one honesty about our own side: money's diminishing curve is measured to a standard nothing else in this debate matches. For purpose and connection there is no comparable dose-response curve — no one knows the mathematical shape of "more meaning." What the evidence shows is that their associations with well-being, health, and survival are large and don't fade at high income. That is the claim we make: not a measured curve, but a measured gap in what money's curve can reach.

The meaning side: across 136,265 people in ten prospective studies, a felt sense of purpose predicted roughly 17% lower mortality after adjusting for confounders (observational, like all of this). Losing work damages life satisfaction far beyond the lost income — the classic fixed-effects panel result. Self-determination theory, the best-validated needs framework in psychology, finds well-being runs through autonomy, competence, and relatedness. And the economists Anne Case and Angus Deaton documented rising deaths from suicide, overdose, and alcoholic liver disease among midlife Americans without a four-year degree — deaths they attribute, by name, to the unraveling of work, community, and meaning rather than income alone (their interpretation has critics, and we flag that too). Émile Durkheim noticed the deeper pattern in 1897: suicide rose in economic booms as well as busts. He called it anomie — desire outrunning the norms that give it meaning. Prosperity without purpose was already a documented hazard in the nineteenth century.

The "money is memory" point is real economics, not poetry: Kocherlakota (1998) showed formally that money's role can be replicated by a record of past transactions. David Graeber's Debt makes the historical case that ledgers of obligation precede coinage. If money is at bottom a societal memory, then asking what the memory forgets is not utopian — it's bookkeeping.

Killingsworth, Kahneman & Mellers (2023), PNAS 120(10) · Killingsworth (2021), PNAS 118(4) · Kahneman & Deaton (2010), PNAS 107(38) · Stevenson & Wolfers (2013), AER 103(3) · Lindqvist, Östling & Cesarini (2020), Review of Economic Studies 87(6) · Cohen, Bavishi & Rozanski (2016), Psychosomatic Medicine 78(2) · Winkelmann & Winkelmann (1998), Economica 65 · Ryan & Deci (2000), American Psychologist 55(1) · Case & Deaton (2015), PNAS 112(49) · Durkheim (1897), Le Suicide · Kocherlakota (1998), J. Economic Theory 81(2)
The score is already moving. Left alone, it moves to attention.

One thing makes this urgent rather than philosophical. In our digital lives, money has already started to lose its grip as the scoreboard — and we can watch what is replacing it. Online, copying anything costs nothing. When information became abundant, the scarce thing became the human capacity to notice it. Herbert Simon saw this in 1971: "a wealth of information creates a poverty of attention."

So a new score emerged on its own: followers, views, engagement. And we now have a decade of evidence about what that score rewards. Each moral-emotional word in a message raises its share rate by about 20%. Engagement-ranked feeds amplify the content that makes people angriest at their out-group — content users themselves, when asked, say they don't want. One platform, at the peak of this logic, weighted an angry reaction five times a like. None of this is a conspiracy. It is just what happens when the yardstick is noticeability: the loudest rises, not the best.

To be careful with our own claim — and an economist would push here, so we'll push first: attention online is not free of money. It pays, through ads and sponsorships, which is part of why people chase it. The precise claim is this: wherever standing gets scored apart from income — feeds, followers, fame — the score that actually emerged is attention. That is the best evidence we have about what fills the seat by default, and extending it to a whole post-money world is an extrapolation. We'd rather build the alternative now than test the extrapolation live.

Go deeper — the evidence for step 3

Simon's passage is from "Designing Organizations for an Information-Rich World" (1971) — he identified attention as the scarce resource; the phrase "attention economy" came later from others. The ~20%-per-moral-emotional-word finding is Brady et al. (2017), an analysis of 563,312 tweets (correlational diffusion data, not an experiment). The causal piece is Milli et al. (2025), a preregistered audit of Twitter's engagement ranking against a chronological control: the ranked feed amplified emotionally charged, out-group-hostile content — and users' stated preferences disagreed with what the algorithm chose for them. The angry-emoji weighting (5× a like, 2018 configuration, later cut to zero) comes from Facebook's own internal documents reported in the Wall Street Journal's Facebook Files.

Simon (1971), in Computers, Communications, and the Public Interest · Brady et al. (2017), PNAS 114(28) · Milli et al. (2025), PNAS Nexus 4(3) · Washington Post (2021), on Facebook's MSI ranking (cf. WSJ's Facebook Files)
There is a better candidate for the next yardstick: impact.

By impact we mean something a twelve-year-old can check: the difference your existence makes in other people's lives. Did someone eat because of you, learn because of you, get well, get free, get heard? That is impact. It is not a brand value or a corporate report. It is the oldest measure there is — it's how we already talk about the dead at their funerals. We just stopped using it for the living.

Attention rewards being noticed. Impact rewards being useful to other lives. Both can be pursued. Only one of them, pursued at scale by millions of ambitious people, leaves the world better than it found it. That asymmetry is the whole movement.

And underneath the economics, be clear about the end. The yardstick is a means. The end is lives that are actually worth living — and the evidence in step 2 says what makes lives feel worth living is precisely contribution, purpose, and being needed. A society that measures people by impact is not asking them to sacrifice for the common good. It is pointing their ambition at the one game where winning and mattering are the same thing.

This idea has a lineage. Bhutan committed a state to a non-GDP yardstick. Amartya Sen and Mahbub ul Haq dethroned GDP-as-sole-measure at the United Nations with the Human Development Index. Eiichi Shibusawa — who helped found some five hundred Japanese companies — spent his life arguing that morality and economy were one thing. And Ken Suzuki's PICSY worked out, in equations, what a currency would look like if purchasing power flowed from contribution. We stand on those shoulders and we differ from each of them — the next page explains exactly where.

Go deeper — the lineage, with honest footnotes

Bhutan: Gross National Happiness was articulated by the Fourth King in the late 1970s (the widely repeated "1972" date is a later retrojection, as Bhutan scholarship has shown). The GNH Index, with nine domains, dates from 2008, the same year Bhutan's constitution directed the state to promote it. Honest footnotes: Bhutan still uses GDP alongside it, and the same era saw the expulsion of roughly a hundred thousand Lhotshampa — a state can adopt a happiness yardstick and still do terrible things. A yardstick is not a conscience.

The HDI: first published by the UNDP in 1990, led by Mahbub ul Haq, anchored in Sen's capability work. Sen reportedly resisted compressing human development into one number — calling such an index vulgar — and ul Haq's reported answer was that he needed "an index as vulgar as GDP but standing for better things." We take both sides of that exchange seriously: measures of flourishing work, and collapsing them into a single number is where the danger starts. That tension shapes our whole design.

Shibusawa: 道徳経済合一 — the unity of morality and economy — from『論語と算盤』(1916). His 合本主義 assembled people and capital to advance public benefit, with profit subordinated to it, not rejected. PICSY: Ken Suzuki's 伝播投資貨幣 (presented 2009; book『なめらかな社会とその敵』, Keiso Shobo, 2013) computes each person's purchasing power as their measured contribution propagating through the transaction network — the closest prior art to making contribution the unit of value, and never deployed at scale. Where we turn from it matters: PICSY computes contribution centrally, from transactions. We think the unit has to be conferred by people, not computed about them. See The Impact Standard.

Ura, Alkire, Zangmo & Wangdi (2012), A Short Guide to GNH Index · UNDP (1990), Human Development Report · Shibusawa Memorial Foundation, on 合本主義 · Suzuki (2013), なめらかな社会とその敵
The hardest objection is one we raise ourselves: rewarding goodness can ruin it.

Stop and feel the force of this before we answer it. The people we admire most are the ones who made a difference without counting it — the anonymous donor, the caregiver nobody thanks. Now here we come, proposing to recognize and honor exactly those people. Doesn't the recognition corrupt the thing it touches? Doesn't an impact measure, chased for its own sake, stop measuring what we meant?

This is not a rhetorical worry. There is a whole research literature on it, and parts of it bite. Pay people for a deed they did from the heart, and you can watch the heart withdraw. A daycare that fined late parents got more lateness — the fine turned a guilty obligation into a cheap price (a famous result, though a 2020 replication attempt failed to reproduce it — details below). Villagers willing to accept a burden for their country refused it when offered money.

But read the same literature to the end and a precise pattern appears. What corrupts is the price: rewards that are tangible, promised in advance, and contingent — do this, get that. What does not corrupt — and in the best evidence we have, strengthens giving — is honor: recognition that is symbolic, freely given, after the fact. The meta-analysis of 128 experiments found expected tangible rewards undermine intrinsic motivation while verbal recognition enhances it. In a randomized trial, a purely symbolic award — a digital badge worth nothing — made volunteer Wikipedia editors about 20% more likely to keep contributing. Honor is not a watered-down price. It runs on different rails — and we're honest below about how much thinner the honor evidence is than the price evidence.

So the paradox dissolves into a design rule, the strictest one we have: impact recognition must work like honor and must never become a price. Never promised for a specific act. Never convertible to money. Never a contract. The moment "do good, get X" exists as a deal, X has re-created the problem money had.

One more honest step. The admirable anonymous donor does not actually argue against us — she argues against the current scoreboard. She is admirable because she gives while the existing yardstick calls her a fool. A civilization cannot run its kindness on heroes alone. And the limit of the promise we make her: a measure built on witnesses cannot see the unwitnessed — the anonymous donor stays uncounted by us too, by her own choice and our design. What changes is the culture around her: what the visible scoreboard tells everyone else a good life is. The unwitnessed-contribution problem is on our open-problems list, where it belongs.

Go deeper — the crowding-out literature, including what doesn't replicate

The canon: Titmuss (1970) warned that paying for blood would degrade voluntary donation. Mellström & Johannesson (2008) found a Swedish payment offer nearly halved willingness to donate — but only among the women in the study (an effect not robustly replicated since), and letting them redirect the payment to charity restored it. Frey & Oberholzer-Gee (1997): Swiss villagers' acceptance of a nuclear-waste site fell from 50.8% to 24.6% when compensation was offered. Gneezy & Rustichini (2000): the Haifa daycare fine roughly doubled late pickups, and lateness stayed high after the fine was removed. Bénabou & Tirole (2006) supplied the mechanism — rewards spoil what a good deed signals about the doer — and Ariely, Bracha & Meier (2009) confirmed it: cash incentives helped private prosocial effort and neutralized public effort.

And the corrections, because we promised honesty: a 2013 meta-analysis of controlled studies found no overall negative effect of incentives on blood donation — Titmuss's conjecture, as a blanket claim, did not hold up. A 2020 replication attempt did not reproduce the daycare result. Field experiments by Lacetera, Macis & Slonim across ~14,000 American Red Cross drives found modest non-cash gifts reliably increased donations without harming safety. The honest summary is narrower and more useful than the legend: crowding-out is real but conditional — it lives where rewards are cash-like, expected, and contractual; it disappears, and often reverses, where recognition is symbolic, unexpected, and expressive of gratitude.

The moderator map (Deci, Koestner & Ryan, 1999; 128 experiments): expected, tangible, contingent rewards undermine intrinsic motivation (effect sizes −0.28 to −0.40); verbal recognition enhances it (+0.33); unexpected rewards leave it intact. Even the meta-analysis's strongest critics (Cameron & Pierce) agree that praise enhances and that expected tangible rewards can undermine — what they dispute is how broadly the undermining generalizes. Scope honesty: these are mostly short-horizon laboratory tasks (puzzles, drawing) measuring task interest, not lifelong prosocial motivation. The field evidence is thinner and points the same way: Gallus (2017, Management Science) randomized a purely symbolic award across ~4,000 newcomer Wikipedia editors and raised one-month retention ~20%, with effects persisting for quarters — one strong proof-of-concept on one platform, and we weight it as that, not as a law.

The hardest version of the objection — we'd rather raise it than wait for a reviewer to: the meta-analysis's safe cell is rewards that are unexpected. But a society-wide standard of recognition can't stay unexpected — once everyone knows good work gets seen, recognition is anticipated in general. So our defense does not rest on surprise. It rests on non-contingency: no specified criteria, no entitlement, no if-this-then-that. The experimental literature locates the damage in the contingent deal — the felt contract that relocates the reason for acting — not in living in a world where honor exists. Societies with strong honor cultures around courage or scholarship did not extinguish the underlying motive — though they bred pathologies of their own, which is what the design rules upstream are for. But we flag it honestly: no study has tested symbolic recognition as a standing system that all participants know about in advance. That experiment hasn't been run, and we are proposing to run it carefully.

Honor has its own failure modes, and we carry them openly: awards inflate when handed out cheaply (Frey & Gallus catalog this); media-conferred celebrity corrupts — award-winning CEOs went on to underperform and manage earnings (Malmendier & Tate, 2009); criterion-based symbolic awards invite gaming and demoralize the quietly excellent (Gubler, Larkin & Pierce, 2016); public token gestures can substitute for real help (the slacktivism and moral-licensing literatures — see Objection 10); and recognition that converts too reliably into material advantage starts behaving like a price again. On that last one, full honesty: standing always leaks some advantage — trust, invitations, who people choose to work with — the way honor always has. The firewall is not zero consequence; it is no contracts, no prices, no exchange rate. Whether that difference of degree is enough is argued at length in Objection 09. Our design rules — peer-conferred, retrospective, decaying, plural, non-convertible — exist because of these findings, not despite them. The full design is on The Impact Standard.

Deci, Koestner & Ryan (1999), Psychological Bulletin 125(6) · Gallus (2017), Management Science 63(12) · Gneezy & Rustichini (2000), J. Legal Studies 29(1) · Frey & Oberholzer-Gee (1997), AER 87(4) · Mellström & Johannesson (2008), JEEA 6(4) · Niza, Tung & Marteau (2013), Health Psychology 32(9) · Lacetera, Macis & Slonim (2012), AEJ: Policy 4(1) · Bénabou & Tirole (2006), AER 96(5) · Ariely, Bracha & Meier (2009), AER 99(1) · Frey & Gallus (2017), Honours versus Money · Malmendier & Tate (2009), QJE 124(4) · Gubler, Larkin & Pierce (2016), Organization Science 27(2)
Start before we need it.

In 1930, with the world in depression, John Maynard Keynes wrote an essay about his grandchildren. He predicted that within a century the economic problem — scarcity itself — would be largely solved, and that the real challenge would arrive after: "how to use his freedom from pressing economic cares, how to occupy the leisure, which science and compound interest will have won for him, to live wisely and agreeably and well."

The people building today's AI expect it to push much further in that direction, and the serious ones are already proposing income floors to distribute what it produces. We treat that as their forecast, not our certainty. But notice what's at stake if they're even half right. When money stops being the thing that decides whether you eat, its grip on the scoreboard loosens — not vanishes; wealth can stay a ranking above any floor, which is why the seat is contested rather than simply vacated. The contest opens. Attention is already auditioning.

The income floor matters for one more reason: it is what keeps an impact yardstick humane. A measure of standing that sits above a guaranteed floor is a game of meaning — you can ignore it, opt out, walk away, and still live. A measure that decides whether you eat is a leash. Everything we propose assumes the floor; nothing we propose may ever reach beneath it. The early evidence on unconditional income — Finland, Kenya, Alaska — says floors don't collapse the will to work and do improve well-being. Who pays for a full floor in a rich country is genuinely unsolved, and we keep it on our open-questions list rather than waving it away.

So Impactism runs in two phases. Today, while money still binds: make contribution visible and honored — culture first, no mechanism required. In the future, if abundance arrives: the rails we built become the yardstick people actually live by. A curve, not a switch. You don't have to believe the second phase to join the first. You only have to believe the dead are eulogized by the right measure — and the living deserve it too.

Go deeper — the evidence for step 6

Keynes: "Economic Possibilities for our Grandchildren" (1930). His growth prediction was roughly right; his 15-hour-week prediction was wrong — people kept working, for reasons (status, meaning, inequality) that this site is partly about. The AI-abundance premise: Sam Altman's "Moore's Law for Everything" (2021) is the clearest statement by a frontier-lab leader — an interested forecast, which is how we cite it. Income floors: Finland's experiment (2,000 unemployed people, €560/month, 2017–18) found clear well-being gains and small, slightly positive employment effects; GiveDirectly's Kenya RCT (~23,000 adults) found no work disincentive and rising enterprise; Alaska's Permanent Fund Dividend shows no aggregate employment effect. The defensible claim is exactly this: unconditional cash does not collapse work and improves well-being. None of these tested a full livable floor in a rich country — that experiment hasn't been run.

Keynes (1930), Economic Possibilities for our Grandchildren · Altman (2021), Moore's Law for Everything · Kangas et al. (2020), Finnish basic income final report · Banerjee, Egger, Faye, Krueger, Niehaus, Suri et al. (2023), Kenya UBI results · Jones & Marinescu (2022), AEJ: Policy 14(2)