# Phase 72 — who gains and who loses?

**Phase73 extension:** The fixed saving/financing closures below are historical scenario assumptions. [SAVING_ASSET_SENSITIVITY_V73.md](SAVING_ASSET_SENSITIVITY_V73.md) now permits saving, equity purchases/sales and principal repayment, with matched no-transition controls. Zero worker wealth accumulation and signed government-claim behavior below must not be read as invariant wealth predictions.

**Status: a four-cohort accounting extension of phase71. All group sizes, endowments, valuation rules and behavioral closures are hypothetical. No empirical deciles, estimated parameters or welfare theorem.**

## Fixed cohorts and ownership counterfactual

The normalized workforce1 is split into workers who remain employed70%, future displaced workers without personal loans10%, future displaced workers with nominal loans10%, and an initially capital-rich/financial-claim holder group10%. All groups initially work. The two displaced cohorts lose employment gradually, in equal proportions, until their combined unemployment is20%. The financial group remains employed. Group figures are within-cohort averages with pooled debt-service capacity, not individual unemployment histories.

Two alternative **initial** equity ownership systems are compared:

- Concentrated: all productive equity belongs to the financial group.
- Broad: productive equity shares equal the population weights, so initial productive equity per person is equal.

The latter starts households with different income and wealth; it is not a costless redistribution performed during transition. Private loan claims and government claims remain held by the financial group under both systems. The ownership counterfactual concerns productive equity, not all financial wealth. Concentrating equity in10% is an input, not evidence that this group will become the observed top decile.

## Production and the government budget stay linked

Production, wages, prices and tax revenue come directly from `coupled_thesis_v71.simulate`; phase72 does not independently grow the profit base. The reference and price assumptions are documented in [COUPLED_THESIS_V71.md](COUPLED_THESIS_V71.md). All monetary quantities are annual normalized units, not country/currency estimates.

All profit surplus after cash profit tax is distributed according to productive equity ownership, without a second personal capital-income tax. The potentially overlapping profit base is not taxed twice. Government interest is paid to the financial group; if signed government net debt becomes negative, that group's signed claim becomes a liability and it pays interest instead. This financing closure is a strong assumption, not a forecast of which assets governments will hold.

Each group's reference real disposable income includes its initial after-tax wage, dividends, financial interest and contractual private-loan interest obligations. Unemployed members' required transfer is:

**B_g* = max(0, P × reference disposable_g − current nonlabor income_g + current contractual debt interest_g).**

Current wages are zero for an unemployed member. Cohort-average transfers equal the unemployed fraction times this benefit. Employed members continue to earn wages and capital income. A group's100 index refers to its own reference;100 does not imply equal living standards across groups.

Government required transfer cost is the population-weighted sum across eligible groups. Under **protect**, the full transfer is paid and the deficit changes government net debt. Under **budget**, one common fraction min(1,max(0,available funds/required cost)) scales the group-specific required transfers. Available funds retain existing real government spending and financing costs. Each ownership/closure combination has its own government debt history. The government does not spend the full tax revenue a second time.

## Nominal debt, liquidity and wealth

Initial aggregate private loans are20:200 per person in the indebted cohort. They are interest-only at4%, producing an initial obligation8 per debtor per year. Cash interest paid cannot exceed the cohort's available current cash. Unpaid interest is capitalized into the loan, increasing debtor liabilities and the lender's contractual claim equally. The lender receives only interest actually paid as cash income; capitalized interest is not treated as cash received.

The main purchasing-capacity index subtracts **contractual** interest. Negative values indicate a liquidity shortfall, not negative physical consumption. Actual cash available after interest paid is separately recorded and is nonnegative for the debtor cohort.

Productive equity has fixed aggregate real value200; nominal value follows the consumption price index. This is an explicit valuation assumption, not a discounted-profit asset pricing model. End-of-year net worth includes equity, signed government claims, private loan claims and liabilities. Nominal private loans are valued at contractual par. Default losses and recovery are not modeled, so lender book wealth under arrears must not be interpreted as realizable wealth.

There is no voluntary saving optimization. Non-financial cohorts consume cash remaining after service; the financial group absorbs net government financing and consumes its cash residual. Consequently zero wealth accumulation among assetless workers is a closure assumption, not a finding that they cannot save. A negative residual is flagged as requiring financing beyond the model. The current preset paths have no such negative consumption residual.

Three accounting checks are enforced in tests:

1. Cohort employment aggregates to the macro employment path.
2. Government net debt changes by the overall deficit; private loan liabilities/claims change equally by unpaid interest.
3. Aggregate private real net worth equals productive equity200 plus government net debt/P; private credit cancels internally. Household cash remaining for consumption plus government goods spending equals nominal output, after net government financing is assigned to the financial group.

These identities do not solve investment, housing, voluntary saving, asset prices, default probabilities, demand feedback or market-clearing behavior. They are necessary accounting checks, not sufficient evidence of an economic equilibrium.

## Selected results at year20: reference transition, budget-limited benefits

The economic assumptions are the phase71 reference transition:20% unemployment,50% efficiency gain over10 years,75% price pass-through,25% terminal profit tax and a fully accessible modeled profit base. Price is85 and real output120 relative to initial100.

| Cohort | Purchasing-capacity index, concentrated equity | Index, broad initial equity |
|---|---:|---:|
| Workers remaining employed | 117.65 | 133.48 |
| Displaced, no personal debt | 49.88 | 91.52 |
| Displaced, nominal debt | 38.09 | 90.12 |
| Financial/capital-holder group | 150.68 | 129.52 |

Each index uses that ownership system's own initial income as denominator. In both systems initial aggregate disposable income is79.8 and initial aggregate real net worth260; their distributions differ. Broad ownership raises displaced groups' direct dividend income and lowers their required public top-up. It does not change aggregate production or tax revenue. Under this budget-limited example government net debt/output is58.82% in both systems.

In the concentrated case, debtor real net worth changes from−200 to−235.29 per person with unchanged nominal principal; financial-group real net worth changes from2800 to2941.18. This is balance-sheet revaluation and government financing under the stated consumption rules, not a prediction of saving or market returns. The phase71 representative debt-free index50.61 becomes49.88 here because one shared program now also serves an indebted cohort with a higher required nominal transfer.

With full protection, both displaced cohorts retain100 at the terminal date in the reference case by construction; the fiscal cost/debt path changes. This shows why income protection and wealth protection are different: real debt burdens can still rise under deflation.

## Stress cases and limits on the thesis claim

With only50% of the modeled profit base accessible for collection and concentrated equity, budget-limited transfers fall to zero at the terminal date. The indebted cohort has12 years with unpaid interest; aggregate private claims rise from20 to31.28. Its negative contractual purchasing index is a funding gap. The financial group's additional par-value claims carry unmodeled credit risk. Broad initial equity provides dividend income in the same stress, avoiding arrears in this example.

The strong-productivity/tax case also exposes sensitivity: sustained government surpluses turn government net debt into assets. Under the chosen signed-claim closure the financial group's government claim becomes a liability, and its book wealth can fall sharply even while some incomes rise. It therefore does **not** follow that initially wealthy, debt-free households always gain wealth in every scenario. That result depends on valuation, financing and saving assumptions.

The defensible thesis statement at this stage is conditional: **nominal obligations and access to capital income materially change the distribution of an automation transition in this model, even under the same aggregate output path.** It does not identify real-world winners, predict the top decile, establish feasibility of ownership redistribution or measure social welfare.

## Reproduce and interact

Run `python scripts/household_distribution_v72.py` to generate four economic cases × two initial ownership systems × two fiscal rules. No empirical data or external account is required. Run `python -m unittest discover -s calibration -p 'test_*v[67]*.py'`. Browser tests exercise ownership/rule/year selection, the control and arrears warnings. The website reads this same result JSON rather than recomputing a separate model.

Open `/thesis.html#research=scenarios`; the phase72 section appears first. The next substantive extension should replace or vary the imposed saving/valuation/financing closure and add housing/rent obligations before claiming long-run wealth accumulation. Fiscal calibration and empirical-decile classification remain unapproved.