Goodhart & crowding out

To make money flow toward love at any scale beyond acquaintance, love must be made legible — counted, compared, ranked. Donation totals, contributor counts, engagement, impact scores, testimonials. And the moment any of these becomes the thing that unlocks money, it will be produced directly, without the love it was meant to indicate. Not necessarily out of malice: Goodhart pressure bends honest actors too — you start doing the version of your work that photographs well.

This is not hypothetical for this project's lineage. A decade of Gitcoin taught it empirically: quadratic funding made "breadth of small contributions" the measure of community love, and immediately got sybils (fake lovers), collusion (coordinated pseudo-love), and airdrop farming (simulated affection at scale). See 🧮quadratic-funding. Charity metrics have the same disease: overhead ratios produced starvation of infrastructure; "lives saved per dollar" produced portfolio-tilting toward the countable.

The deep version of the problem: love may be precisely that which resists metrics. Its defining properties — unconditionality, particularity, inner transformation (🪷metta) — are the properties measures cannot capture. A perfect metric of love would be a contradiction.

• Titmuss, The Gift Relationship (1970). Comparing blood systems: paying donors did not add supply on top of voluntary giving — it degraded the whole system (fewer altruistic donors, worse blood). Introducing price didn't price the gift; it replaced it.
• Gneezy & Rustichini, "A Fine Is a Price" (2000). Israeli daycares fined parents for late pickup; lateness increased and stayed high even after the fine was removed. The fine reframed a moral relation (I'm inconveniencing a teacher) as a market one (I'm buying overtime) — and the reframing was a one-way door.
• Self-determination theory (Deci & Ryan). External rewards can corrode intrinsic motivation, especially when they feel controlling rather than informational. The same dollar can be experienced as "you did this for money" (corrosive) or "your work mattered; here is recognition" (nourishing). The framing, timing, and expectation structure decide which.

The entire project is an attempt to route money toward love. Crowding-out is the documented failure mode of exactly that move: money doesn't just fail to carry love — under the wrong framing it deletes the love that was already there. Get this wrong, and it makes the world less loving, with good intentions and real money.

1. Money as response, never as incentive. Pay after, unprompted, for what was already given freely — never announce "do X, receive Y." This is dana's structure (🎁dharma-gift) and retroactive funding's (🧮quadratic-funding), and it is the single strongest crowding-out defense: an unexpected gratitude cannot have motivated the act it thanks.
2. Informational, not controlling. Money should arrive wrapped in meaning — the story of what the work meant to the giver (🌀gift-economies) — so it reads as recognition. Bare transfers with KPIs attached read as control.
3. Never convert an existing gift relation into a paid one. The daycare door swings one way. Where care already flows unpriced (mutual aid, volunteering, maintenance done from love), the loving intervention may be infrastructure and thanks, not wages. First, do no reframing.
4. Watch the self-story. The experiments should track how givers and receivers describe what they're doing before and after money enters. The moment language shifts from "I do this because I love it" to "I do this for the support" — that's the canary, measurable in words.