Applied Identity Physics™ · SNSFT Foundation

Policy Impact Check

Function over Fallacy. Step 6 or it didn't happen. A policy's stated intent and its measured effect are two different things. This tool checks the effect: every affected group, over time and on the ground, placed against the Torsion Limit TL = 0.136899099984016. Nothing you enter leaves your browser.

Learner mode: plain words, rounded numbers. Switch to Researcher for the formulas, full precision and the working for every row.

Researcher mode: τ to full precision, TL = 0.136899099984016, TL_IVA = 0.88 TL = 0.120471207985934, and the working for every row.

1 The policy

The group the policy says it helps. The stated-allegiance check tests this group's measured outcome.

2 Affected groups

One row per group and measure. Before is the group's situation now. After is the projection at each horizon, as the policy text reads. On-the-ground loss is the extra loss from how the policy is actually carried out (delays, paperwork, enforcement), as a percent. ± is the uncertainty in the projection, as a percent of Before. People is how many people the row covers, used for the aggregate check.

GroupMeasureBetter whenPeopleBeforeAfter · nowAfter · 1 yrAfter · 3 yrOn-the-ground loss %± %

3 Result

No check run yet

Fill in the groups and press Run the check.

NOBLE τ = 0, no loss LOCKED 0 < τ < TL_IVA IVA PEAK TL_IVA ≤ τ < TL SHATTER τ ≥ TL GAIN the group is better off
Group · measureNow1 yr3 yrNow, on paper onlyPlacement holds within ±?

In plain words

The working

τ = max(0, −s), s = (X − X_ref) / X_ref when higher is better s = (X_ref − X) / X_ref when lower is better X on the ground = X·(1 − ℓ) when higher is better, X·(1 + ℓ) when lower is better (ℓ = on-the-ground loss) Shatter line: X* = X_ref·(1 − TL)/(1 − ℓ) (higher is better) X* = X_ref·(1 + TL)/(1 + ℓ) (lower is better)
Group · measureX_refX now (ground)X 1 yr (ground)X 3 yr (ground)τ nowτ 1 yrτ 3 yrShatter line X*

The four ways harm hides

Net effect across everyone

4 Payments over time

For a payment that repeats every year (a dividend, a grant, a stipend, a wage supplement). Each year is placed against the year before, because last year's payment is what a household budgets on. Then the same history is run through two payout rules to see which one keeps every year out of Shatter, and how big a reserve that takes.

One year per line. The third number is optional: what the formula or the law said the payment should be. When it is there, the year is also placed against it (paid versus promised).

Years

Each year pays the average of the last N years' amounts.

Max drop %
Max rise %

Each year pays the amount, but never less than last year minus the max drop, and never more than last year plus the max rise. Good years put the difference into a reserve; lean years draw it down. A max drop of 12.0% sits just under the IVA Peak edge (TL_IVA = 12.047%).

Window years

Each year is also placed against the highest payment of the previous N years, so a slow slide that never trips the year-to-year check is still caught. To guarantee no N-year stretch reaches Shatter, the max drop must stay below 1 − (1 − TL)^(1/N): 4.79% per year for 3 years.

No series run yet

Payout ruleVerdictShatter yearsIVA Peak yearsWorst drop (τ)Worst multi-year drop (τ)Multi-year Shatter yearsPaid in total, per personStarting reserve needed, per personReserve at the end

Year by year

The working

τ_t = max(0, 1 − P_t / P_(t−1)) drop against last year's payment τ_paper = max(0, 1 − paid_t / promised_t) paid versus promised Rolling average: P_t = mean(A_(t−N+1) … A_t) Bounded change: P_t = min( max(A_t, (1 − d)·P_(t−1)), (1 + u)·P_(t−1) ), P_first = A_first Reserve: R_t = R_(t−1) + (A_t − P_t); starting reserve = max(0, −min_t Σ(A − P)) Multi-year slide: τ_W,t = max over k = 1…W of max(0, 1 − P_t / P_(t−k)) Highest max drop that stays out of Shatter: d < TL = 13.6899%; out of IVA Peak: d < TL_IVA = 12.0471% Guaranteed for every W-year stretch: (1 − d)^W > 1 − TL ⇔ d < 1 − (1 − TL)^(1/W) (W = 3: 4.7890%)

5 Compare two designs

Boards choose between options. Compare the design on screen (A) with a module or a saved file (B), group by group.

6 Share and verify

Share link: the whole setup is packed into the link itself. Anyone who opens it sees exactly the same inputs; nothing is stored on a server.

Independent verification: a check run only by the party being checked is not enough (Magna Carta, Article 61: no authority is the sole judge of its own compliance). Save this run, have a second party with a different interest run the same inputs, and load both here. When two independent runs match the same inputs and the same verdict, the result is marked independently verified.

How the check works

The measure. Each group's torsion is its adverse change against its own reference, the reduction of the Applied Identity Physics Economics paper: τ = |X − X_ref| / X_ref, counted only in the direction that harms the group. X_ref is the group's situation before the policy; X is the projected situation after. A group that ends up better off has τ = 0 and is shown as a gain.

The placement. TL = 0.136899099984016 and TL_IVA = 0.88 × TL = 0.120471207985934. A group below TL_IVA is Locked: it carries the change. Between TL_IVA and TL it is at IVA Peak: under strain at the edge. At or above TL it is in Shatter: the change exceeds what the group can carry. The Economics paper places real events the same way: a 2% inflation target at τ = 0.020 (Locked), US output in 1932 at 0.129 (IVA Peak), a 20% bear market at 0.200 (Shatter).

Bounded uncertainty. The ± band is applied to each projection. If the whole band stays inside one phase, the placement holds. If the band reaches a boundary, the placement is marked open, and better data decides it.

The verdict. The policy passes when no group reaches Shatter at any horizon, either as written or on the ground. The net effect sums each group's signed change weighted by its people: Justice as a net gain across every party (APPA NOHARM Lossless Kernel [9,0,1,1]).

Payments over time. A repeating payment is placed year against year: τ = 1 − this year ÷ last year, counted only when it drops. The same history is then run through a rolling average and through a bounded-change rule with a reserve. A bounded rule with a max drop below TL can never place a single year in Shatter; the question it answers is how large a reserve that takes, given the real history. Each year is also placed against the highest payment of the previous N years, because a run of drops that each pass can add up to Shatter: the deferred-harm check from Function over Fallacy, applied to the payout rule itself. Smoothing rules of this kind are standard for endowments; Yale's spends 80% of the prior year's spending plus 20% of a target rate on the endowment, for intergenerational neutrality.

The four checks come from Function over Fallacy: the aggregate mask (the overall number improves while a group is harmed), deferred harm (fine now, Shatter later), compliant on paper (fine as written, Shatter on the ground), and stated allegiance (the named beneficiary is harmed).

This tool measures effect, not motive. It does not render a verdict on any person, party or intention, and a policy can fail through honest miscalculation. Its output is a starting point for scrutiny, to be confirmed by more than one independent party with real data. The module examples use illustrative numbers; replace them with your own.