There’s a moment in your product you probably treat as the finish line, and the brain treats as a permission slip.
It’s the instant right after a win. The workout is logged. The lesson is done. The savings goal is hit. The inbox is at zero. Your interface throws confetti, and somewhere behind the user’s eyes, a quiet piece of arithmetic runs: I’ve been good — I’ve earned a little slack. That arithmetic has a name. Psychologists call it moral licensing, and it’s one of the stranger facts about how we keep score of ourselves.
The classic demonstrations are almost too clean. In work going back to Monin and Miller in 2001, people who first got a chance to establish their good credentials — to show they weren’t biased, weren’t stingy, weren’t lazy — then felt freer to do the very thing the credential was supposed to rule out. A good act doesn’t just sit there being good. It gets deposited, like credit, into an account the brain keeps. And a balance that goes up is a balance you can spend.
For a builder, the uncomfortable part is where the spending happens. It happens inside your product, at the exact surface you designed to feel like an achievement. The fitness app that says “You crushed it!” may be handing out a coupon good for skipping tomorrow. The finance app that celebrates a saved dollar may have just made the next impulse purchase feel pre-forgiven. The learning app that awards a badge for the completed module may have quietly closed the loop the user came to keep open. The completion screen is a licensing screen. You built the thing that grants the permission.
This is the mirror image of an effect I’ve written about before — the what-the-hell moment, where one lapse (“I already broke the diet”) licenses a collapse (“so the day’s a write-off”). Same accountant, opposite ledger. There, a failure gets over-charged. Here, a success gets over-credited. Both come from treating behavior as a running moral balance instead of a direction you’re heading. The fix, in both cases, is to change what the moment means rather than to add more willpower on top of it.
So the build decision lives in one sentence of copy and one choice of framing. Celebrate the act as a transaction — “Nice work, you’ve earned a break” — and you’re underwriting the withdrawal. Frame the same moment as trajectory — “Three in a row. You’re becoming the person who does this” — and there’s nothing to cash in, because identity isn’t a balance you spend down; it’s a direction you’d be contradicting. The goal-met state is the most dangerous surface you own, precisely because it feels like the safe one.
One more wrinkle worth designing around. A 2026 meta-analysis by Rotella and colleagues found that being observed changes the effect — licensing behaves differently when the good act is public than when it’s private, which fits the idea that some of this is about the credit you bank in other people’s eyes, not just your own. Practically: a visible streak, a shared badge, a “posted to your friends” win leans on the reputational ledger. That can cut either way — a public commitment can also make the next lapse feel costlier — so if you’re going to make virtue visible, decide which of those two forces you’re pulling on.
A caution before you rebuild every success state: the effect is real but modest, and its replication record is mixed — a 2015 meta-analysis put it at small-to-moderate, and some direct replications came up empty. Treat moral licensing as a lens, not a law. You don’t need to strip the joy out of a win. You just need to notice that the most satisfying screen in your product is also the one most likely to quietly grant permission — and to decide, on purpose, what permission it grants.
The science, to look up: moral licensing / self-licensing (Monin & Miller, 2001, on moral credentials; Khan & Dhar, 2006, on licensing in choice); the observation moderator (Rotella, Jung, Chinn & Barclay, “Observation Moderates the Moral Licensing Effect,” Personality and Social Psychology Bulletin, 2026); and the replication caution (Blanken, van de Ven & Zeelenberg, “A Meta-Analytic Review of Moral Licensing,” PSPB, 2015).
Sources
- Monin & Miller 2001
- Rotella, Jung, Chinn & Barclay 2026
- Blanken, van de Ven & Zeelenberg 2015
Liked this? Get the next one in Working Theory.
Going weekly in August (it's in beta now). One genuinely interesting read on building, the brain, and the science most people missed.
Subscribe →