🇬🇧 English · 🇪🇸 Español
Why the tariff brief has positive consumption EV
The default brief gives US consumption EV of +6,427.19 million USD (+0.0383%). A separate Leontief equilibrium solver reproduces it within 0.000001 million USD. It is a consumption result in a static model with investment as another final use.
| US final use | Quantity change valued at baseline purchaser prices, million USD |
|---|---|
| Consumption, C | +6,427.19 |
| Investment, I | −7,095.64 |
| Residents' purchases abroad, Cx | −466.70 |
Only the C row is consumption EV. Adding these rows does not produce an intertemporal or social welfare measure: future returns to investment are not modeled, and the table's C aggregate includes government and NPISH consumption.
Budget and resource explanation
The model allocates fixed expenditure shares to final uses. In the default experiment US nominal disposable income, including fiscal rebates, increases 0.0660%, while its consumption price increases 0.0277%. Consumption therefore increases 0.0383%. Investment and purchases abroad face larger price increases (0.1972% and 0.3304%); their quantities fall 0.1417% and 0.2635%. Investment spending also includes the prescribed foreign-balance adjustment. These are equilibrium budget identities, not separately identified causal contributions.
For fixed sourcing, let A be intermediate input coefficients, F final basket coefficients and v unit factor-resource requirements at baseline factor prices. Then factor-resource accounting prices satisfy
Fixed factor endowments imply \(\sum_j P_j^R\Delta c_j=0\). The independent calculation gives world changes of +6,761.06 for C, −6,493.69 for I and −267.37 for Cx, in million baseline factor-value units; their sum is less than 0.000001 in magnitude. These tax-free resource values differ from consumer prices and are not welfare weights. The world consumption increase uses resources released by the other final categories.
The positive US C result persists when both equilibria are recomputed without domestic taxes (+6,500.05 million), with zero foreign balances (+5,892.06), or with both changes (+5,923.46). Neither domestic-tax distortions nor nonzero foreign balances alone explains the sign. These are separate model closures, not an additive decomposition of the headline.
Independent checks and scope
tools/reference_validation/tariff_reference.py shares calibration arrays
only. Conditional on country wages, it solves separate price, quantity and
rebated-income linear systems, then finds relative wages from factor-market
clearing. It does not call the production equilibrium, residual, accounting,
Jacobian or welfare helpers. State comparisons pass the registered scaled
1e-8 tolerance, and US EV agrees within 1e-5 million USD.
A separate evaluator reconstructs goods flows, zero profit, labor and capital
use, purchaser prices, household and government budgets, external balances,
and expenditure-function EV for both sigma=0 and sigma=2. Maximum normalized
equation errors are below 7e-11; EV errors are below 1.5e-8 million USD.
This independently checks the sigma=2 solution's equations, without claiming
a second nonlinear equilibrium solver for that variant.
The default tariff applies to all 20 modeled sectors, including services.
The alternative sigma=2 composite substitutes across all country-sector
intermediate inputs, including different sectors; it is not restricted to
origins within the same input sector. Final-use baskets remain fixed. Its US C
gain is +23,497.93 million, alongside I of −2,840.78 and Cx of +1,470.69 at
baseline purchaser prices. The fixed-A resource identity above is specific to
sigma=0 and is not applied to the changing-input technology.
Frozen inputs, source hashes, tolerances, state errors and closure comparisons
are in tests/fixtures/first_brief_tariffs/independent_validation.json.
Regenerate from the development checkout:
No retaliation, transition dynamics or estimated substitution elasticities are included. The numerical sign is validated for this stated model and closure; it does not establish that the tariff benefits real-world households.