TruIndex economic research · Version 1.0
TruIndex allocation model
Version 1 estimates the marginal benefit of additional supply and allocates a finite budget in proportion to observed requests fulfilled. The reference case uses two kitchens and a single service.
Data are synthetic. Each round updates the local shortage forecast and transfers simulated credits between three accounts. The civic pilot has a separate protocol for private testimony.
Interactive model
Two-kitchen simulation
Each kitchen has five meals ready. Move the sliders to decide how many people ask for lunch and how many extra meals arrive. Try setting requests to five at one kitchen.
Up to four units per additional person served, shared proportionately if the offer runs short.
This lunch would serve 3 additional people
The sponsor would pay 12 trial HVT. 0 of the offer would stay with the sponsor.
- River supplier 5 → 13
- Hill supplier 15 → 19
- Sponsor 80 → 68
The mathematics
Model specification
Expected reduction in unmet need
The letter a stands for an action. L is a loss the community chooses to measure, such as a lunch someone needs and cannot get. The large E means to average possible outcomes using what we know at time t. Missing evidence leaves the value unknown.
Realized service benefit
D is the number of people requesting lunch, including those turned away. S is the supply already available and q the extra meals delivered. With five meals ready and seven requests, a delivery of three serves two additional people. The third meal earns zero under this rule.
One extra meal helps exactly when requests exceed supply, assuming demand stays the same and meals are interchangeable. Its expected benefit is the chance of that shortage. Recording a lunch adds its demand to the kitchen’s history for the next forecast.
Budget allocation
Each supplier’s share is the number of extra people they served (si) divided by the total extra people served (T). The sponsor pays up to the budget B, at the agreed rate r. With zero extra people served, the payout is zero. Whole units left after rounding go to the largest fractional shares. The River account wins exact ties in this edition.
A saved balance earns no additional share. A wealthy supplier can still win every useful opportunity. The simulation gives you the balances so you can check who gained. It measures one lunch outcome. Readiness, displaced volunteer work and private care remain open questions.
Reproduce the experiment
Download the Python engine and the example into the same folder. Run python3 engine.py example.json. Historical lunches produce the initial forecasts. The engine also handles unknown demand and deferred settlement.
Working manifesto
Research principles
Communities define the outcome, the evidence used to measure it and the resources available for compensation. A payment must be traceable to a funded contribution. An unobserved outcome remains unknown.
Read the version 1 manifesto in full
Someone washes the dishes after everyone has gone to bed. In the morning another person walks into a kitchen ready to use. The work already had value in the interval when nobody knew who had done it. That is where TruIndex begins.
We want an economy in which the people who sustain everyday life can share more fully in its wealth. A person's ability to pay tells us little about how urgently they need help. A person who gives that help has a claim to our attention even when the work has never earned a wage.
TruIndex asks what an action makes possible for others, here and over time. The same delivery can feed a whole table in one place and spoil in another. A kitchen that runs well every day may owe that success to the person who quietly keeps it running. Both the shortage somebody relieves and the shortage somebody prevents deserve to be understood.
The people affected should choose what counts as an improvement. Where ordinary, consented records can help, the index can learn from them. Its evidence, assumptions, and errors should be open to challenge. Private care deserves room to stay private. An unanswered question about an act's value should stay open.
The Human Value Token is our proposal for connecting useful work to material support. Someone may love the work, love earning tokens, or both. We have no reason to test the purity of their motives. We do have reason to ask whether another rewarded act still helps somebody. Five fresh meals for five hungry people and five claims on the same meal call for different decisions.
People should be able to save what they earn and use it later. Holding a balance alone should earn no further claim on the common pool and give its holder no greater say over the rules. Communities should be able to see what resources back an offer and whose resources actually moved after it was fulfilled.
Our test is whether people receive something they need, whether the people doing overlooked work gain a greater share, and whether the work still happens freely where there is no reward. We invite collaborators to bring one ordinary example where these aims come apart. That example can teach us more than agreement on the principle.
Collaborate
Replication and pilot proposals
If you cook, volunteer, or organize a service, try one lunch and tell us which work the model misses. If you keep service records, suggest one anonymous count that would change the result. If you study economics or build models, propose a case where credits concentrate despite the useful work.
For a pilot, tell us what service you know, what people need, and what records already exist. A fictional example is welcome. Please leave individual names and service records out of your message.
Suggested citation
TruIndex. TruIndex and the Human Value Token: economic research model, version 1.0, 19 November 2026. Original concept by Emilio Castellanos. humanvaluetoken.org/research/v1