Impactism
日本語 Join Impactism
Manifesto The Impact Standard Foundations Objections Read For organizations 日本語
Join Impactism
A hand writing a letter with a fountain pen
Photo by Kelly Sikkema on Unsplash

Essay

Honor Is Not a Price

June 18, 202610 min read

The people we admire most made a difference without counting it. The anonymous donor. The nurse who stays late. The caregiver nobody thanks. They are admirable precisely because they gave while no scoreboard was watching. And here we come, proposing to honor exactly those people — to make impact, not money, the measure of a life.

Feel the force of that before we answer it. If you reward goodness, do you not corrupt the very thing that made it good? Does an impact measure, once chased for its own sake, stop measuring what we meant by it? This is the hardest objection to everything we are building, and we did not get it from a critic. We raised it against ourselves. We owe you the most honest answer we have.

The short version: there is a real research literature on this, and parts of it bite hard. But read it to the end and the corruption has a precise location. What corrupts is the price — a reward that is tangible, expected in advance, 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 paradox does not refute us. It hands us our strictest design rule.

Where the price corrodes

Start with the evidence against ourselves, at full strength.

Pay people for a deed they did from the heart, and you can watch the heart withdraw. The economist Richard Titmuss argued in 1970 that paying for blood would degrade the impulse to give it freely. In a Swedish field experiment, Carl Mellström and Magnus Johannesson offered women a small payment (about $7) to donate blood, and willingness fell by nearly half — until they let the women redirect the payment to charity, which restored it (Mellström & Johannesson 2008). Bruno Frey and Felix Oberholzer-Gee asked Swiss villagers whether they would accept a nuclear-waste repository in their community; just over half said yes — until compensation was offered, at which point acceptance fell from 50.8% to 24.6% (Frey & Oberholzer-Gee 1997). Offering money turned a civic duty into a bribe to be refused.

The mechanism is not mysterious. Uri Gneezy and Aldo Rustichini sent teenagers out to collect for charity: those paid nothing raised the most, and a small commission produced less than no payment at all (Gneezy & Rustichini 2000b). A price can sit below the value of the motive it replaces, so naming it shrinks the act. Roland Bénabou and Jean Tirole worked out why in theory: a reward spoils what a good deed signals about the doer, because an observer can no longer tell whether you acted for the cause or the cash (Bénabou & Tirole 2006). Dan Ariely, Anat Bracha and Stephan Meier confirmed it in the lab: monetary incentives raised prosocial effort in private but not in public — cash muddies the audience’s read of the motive (Ariely, Bracha & Meier 2009). And the philosopher Michael Sandel gave the family of results its name: pricing a good can degrade its meaning. This is the corruption objection, and it is distinct from any worry about fairness (Sandel 2012).

So far the case against us looks strong. Now the honesty we promised.

The legend is bigger than the evidence. A 2013 meta-analysis of controlled studies found no overall negative effect of incentives on blood donation; Titmuss’s warning, taken as a blanket claim, did not hold up (Niza, Tung & Marteau 2013). The Swedish effect appeared only for the women in the study, and has not been robustly replicated. And the most famous story in this whole literature is shakier than its fame suggests. Gneezy and Rustichini studied ten Haifa daycares; introducing a small fine for late pickup roughly doubled lateness, and lateness stayed high even after the fine was removed — a guilty obligation, once priced, became a cheap purchase (Gneezy & Rustichini 2000a). It is a perfect parable, and we have to tell you that a 2020 replication attempt did not reproduce it. We will not lean a movement on a result that wobbles when it is pushed.

The honest summary is narrower than the legend, and more useful. Crowding-out is real but conditional. It lives where rewards are cash-like, expected, and contractual. It thins, and sometimes reverses, where recognition is something else.

Where honor strengthens

Here the literature turns, and it turns on a distinction the headlines usually smear over.

The cleanest map comes from a meta-analysis of 128 experiments by Edward Deci, Richard Koestner and Richard Ryan. Expected, tangible, contingent rewards undermine intrinsic motivation, with effect sizes from about −0.28 to −0.40. Unexpected rewards do not undermine it. And verbal recognition — praise, being seen — enhances it, by about +0.33 (Deci, Koestner & Ryan 1999). This is the same line Deci drew in his original 1971 experiment, where paying students for a puzzle they already loved reduced how much they played it for free, while praise increased it (Deci 1971). The meta-analysis’s sharpest critics, Cameron and Pierce, dispute how broadly the undermining generalizes — but not the moderator structure itself. Everyone agrees praise enhances; everyone agrees expected tangible rewards can undermine. The fight is about magnitude, not about which way the arrow points.

We owe you the scope honestly: these are mostly short-horizon laboratory tasks — puzzles, drawings — measuring task interest, not a lifetime of giving. The lab is where the moderator map is cleanest and where its reach is smallest.

The field evidence is thinner, and points the same way. Jana Gallus ran a randomized trial on German Wikipedia: a purely symbolic peer award — a “barnstar,” a digital badge worth nothing, convertible into nothing — made newcomer editors about 20% more likely to still be contributing a month later, with effects that persisted (Gallus 2017). One platform, around four thousand newcomers, and it measured whether people stayed, not whether their work got better. We weight it as one strong proof-of-concept, not a law. Back on blood: across roughly fourteen thousand American Red Cross drives, Nicola Lacetera, Mario Macis and Robert Slonim found that modest non-cash gifts — announced as appreciation, not as cash-for-blood — reliably increased donation without harming safety (Lacetera, Macis & Slonim 2012). And Italian blood donors, asked directly, say they accept medals and a day off gladly but that cash would make them stop (Lacetera & Macis 2010). That last is partly a statement of preference rather than a measured behavior, and we mark it as such — but it says out loud what the experiments imply: a gift of honor and a payment of money are not points on one scale. They are different things, and people feel the difference.

Bruno Frey and Jana Gallus pulled it together in 2017: awards are the one incentive with evidence of crowding in rather than out, because an award is expressive and relational, not read as a price (Frey & Gallus 2017). Honor is not a watered-down price. It runs on different rails.

The rule the paradox hands us

So the objection does not destroy the project. It writes its strictest law:

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 instant “do good, get X” exists as a deal a person can count on in advance, X has re-created the exact problem money had.

We have to be careful here, because there is a tempting and wrong way to draw this line. The safe cell in the Deci meta-analysis is rewards that are unexpected. So why not just rest on surprise? Because a society-wide standard of recognition cannot stay unexpected. Once everyone knows that good work gets seen, recognition is anticipated in general, even when no one can predict its particular arrival. Surprise is not a firewall a movement can build on.

The firewall is non-contingency and non-convertibility. The experiments locate the damage in the contingent deal — the felt contract that relocates the reason for acting from the thing itself to the payout — and in the convertibility that turns standing into a spendable balance. Not in the mere fact that honor exists in the world. You can know, in general, that a society honors contribution, and still do the work for its own sake — the way people have always lived inside honor cultures around courage or scholarship without the underlying motive being extinguished. What you cannot do is sign a contract for your own virtue.

We will say plainly what hasn’t been proven: no study has tested symbolic recognition as a standing, society-wide system that all participants know about in advance. That experiment has not been run. We are proposing to run it, carefully, with the firewall built in from the start.

Honor’s own failure modes

If we stopped here we would be doing the thing we accuse others of — telling you only the half of the evidence that flatters us. Honor has its own pathologies, and the same book that documents honor crowding in documents how it fails.

It fails by inflation. Hand awards out cheaply and they stop meaning anything; Frey and Gallus catalog this as the first failure mode of honors. It fails through envy and error — the wrong people honored, the right people passed over, resentment where there should have been gratitude.

It fails when celebrity replaces judgment. Ulrike Malmendier and Geoffrey Tate tracked CEOs who won prestigious media awards and found that, afterward, they underperformed, extracted more pay, and managed earnings more aggressively than matched peers who did not win (Malmendier & Tate 2009). The honor there was media-conferred, in weak-governance settings — exactly the conditions our design fights — and it is a matched-sample study, not a randomized trial. But the lesson stands: recognition that flows to a name rather than to a deed corrupts the name.

It fails when you attach criteria. Timothy Gubler, Ian Larkin and Lamar Pierce studied a firm that introduced a symbolic, criterion-based attendance award. Workers gamed the eligibility line, and — worse — the previously punctual, who had been showing up for their own reasons, became about 8% less efficient once their reliability was turned into a contest they could win or lose (Gubler, Larkin & Pierce 2016). This is the anti-criterion lesson, and we take it directly into the design: recognition cannot run on a published checklist, because a checklist is a target, and a target is gamed.

It fails as performance. Katherine White and colleagues found that public token support — the visible, low-cost gesture — reduces the substantive help a person gives afterward, while private token support increases it (Kristofferson, White & Peloza 2014). The contrast is between public and private token acts, not a blanket verdict on small gestures. And the broader worry behind it — moral licensing, the way one good act can grant permission for a later lapse — is real: a meta-analysis of 91 studies put the effect at about d = 0.31, while the authors themselves caution that the figure is likely inflated by publication bias (Blanken, van de Ven & Zeelenberg 2015). We carry that caveat every time we cite the number.

And honor fails, finally, when it converts too reliably into material advantage — when standing buys so much, so dependably, that it starts behaving like a price again. Here we owe the most uncomfortable honesty of all. Standing always leaks some advantage: trust, invitations, who people choose to work with. Honor has always done this. We are not promising zero consequence; that would be a lie. We are forbidding the contract and the exchange rate. Whether a difference of degree — no prices, no convertibility, but real soft advantage — is enough to hold the line is a fair question, and we argue it at length, not in a footnote.

What this commits us to

We did not collect these failure modes to scare you off. We collected them because every one of them is the reason for a specific rule. Recognition that is peer-conferred, not central. Retrospective, not promised. Decaying, so no one banks it. Plural, never a single global score. And non-convertible — carried, never cashed. Each rule answers a documented failure above it, not a hypothetical one.

That is the difference we are insisting on, and it is the whole of our defense. A price says do this and you will get that. Honor says what you did mattered, and we saw it. The first is a contract; the second is a witness. One relocates the reason you act; the other simply refuses to let the good go unrecorded. We are betting a civilization can be built on the second without sliding into the first — and we have shown you, above, exactly where the slide happens and what we have placed in its way.

The full mechanism — every design rule and the documented failure it answers — is on The Impact Standard. The objection this essay raises is the fifth and hardest of the serious challenges we put to ourselves; the rest, from social credit to Goodhart’s law, are steelmanned and answered on Objections.

Make impact, not money, the measure of a life. Honor it. Never price it.

Sources

← All essays