[mmcky]

New lecture gradual_liberalization_network ("Gradual Liberalization in a Global Production Network"), placed after organizational_equilibrium: a three-country, five-sector version of the Bassetto-Huo-Rios-Rull (2026) trade model with intermediate inputs, solved by exact hat algebra from illustrative shares. It decomposes the effect of tariffs on worker welfare, derives a first-order condition on cross-sector wage inequality under which the Markov and best constant tariffs are positive, computes Markov, Ramsey, and organizational equilibrium tariff paths and their response to a protectionist shock, and discusses trade blocs (Bonadio et al. 2026). Three exercises with dropdown solutions.

Additions to organizational_equilibrium: two decompositions of worker welfare, a table matching the growth and tariff models, a three-step computing recipe, workers' consumption paths across equilibria, and a table comparing the equilibria after a temporary shock.

Adds bib entries Armington1969, HuoLevchenkoPandalaiNayar2025, and BonadioEtAl2026.

[netlify]

✅ Deploy Preview for lustrous-melomakarona-3ee73e ready!

Name	Link
🔨 Latest commit	02c1ef6
🔍 Latest deploy log	https://app.netlify.com/projects/lustrous-melomakarona-3ee73e/deploys/6ae31d53808781aefb98a162
😎 Deploy Preview	https://deploy-preview-399--lustrous-melomakarona-3ee73e.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify project configuration.

[Copilot]

🔵 Needs a closer look

Verifying the solver's many hardcoded numeric claims in the 1140-line lecture requires a CI notebook execution that cannot be run here, leaving quorum unmet.

No open issues 🎉

What changed in this PR

This PR adds a new QuantEcon lecture, gradual_liberalization_network ("Gradual Liberalization in a Global Production Network"), and extends the existing organizational_equilibrium lecture with additional welfare decompositions, comparison tables, and worker-consumption analysis. The new lecture builds a small three-country, five-sector version of the Bassetto–Huo–Ríos-Rull (2026) trade model with intermediate inputs, solved by exact hat algebra from illustrative shares, and reuses the dynamic policy machinery of organizational_equilibrium to compute Markov, Ramsey, and organizational-equilibrium tariff paths plus their response to a protectionist shock. It fits into the "tariff/trade-cooperation" cluster of lectures, serving as the multi-country, multi-sector sequel to organizational_equilibrium.

Changes:

• New 1140-line lecture gradual_liberalization_network.md (model setup, separability, solving-in-changes solver, welfare decompositions, a first-order condition gln-condition, dynamic tariff paths, trade-blocs discussion, and three exercises with dropdown solutions).
• Extends organizational_equilibrium.md: static() now also returns real wages ω_i, two welfare decompositions, a growth-vs-tariff model mapping table, a three-step organizational-equilibrium recipe, workers' consumption paths across equilibria, and a post-shock comparison table, with cross-links to the new lecture.
• Registers the lecture in _toc.yml and adds three BibTeX entries (Armington1969, HuoLevchenkoPandalaiNayar2025, BonadioEtAl2026).

I reviewed the Python code, the mathematics, and the cross-references in detail. All static() call sites in organizational_equilibrium.md were correctly updated for the new 6-tuple return (e.g., [2:] → [2:4]), the decomposition and consumption-path cells are algebraically consistent, and I numerically confirmed that ψ_i stays positive over the wider [0, 1.5] grid (min ≈ 0.34) so the np.log/argmax are safe and the "peaks at about 85 percent" claim matches the computed value. In the new lecture, the revenue_system/solve_changes einsum dimensions and exponents match the stated equations, calibration arrays (λ, Φ, each Ξ row, α_k+α_l+α_x) sum to one, the first-order condition derivation matches gln-condition, all cross-cell variables are defined before use, and every {eq}/{ref}/{doc}/{cite} reference resolves. I did not find concrete defects to flag.

Note: the lecture quotes many hardcoded numerical results in prose that are generated by the heavy exact-hat-algebra solver; their correctness ultimately depends on a successful CI notebook execution, which I could not run in this environment.

File	Description
lectures/​gradual_liberalization_network.md	New lecture: multi-country/multi-sector tariff model, exact-hat-algebra solver, welfare decompositions, dynamic tariff paths, exercises.
lectures/​organizational_equilibrium.md	Adds real-wage output to static(), two welfare decompositions, model-mapping and post-shock tables, consumption-path analysis, and cross-links to the new lecture.
lectures/​_toc.yml	Registers gradual_liberalization_network after organizational_equilibrium.
lectures/​_static/​quant-econ.bib	Adds Armington1969, HuoLevchenkoPandalaiNayar2025, BonadioEtAl2026.

_{Review effort: Balanced}

[mmcky]

Review

The economics and the code hold up. I've pushed one commit of style fixes and small cleanups (54d54a7). The remaining questions are about claims concerning other papers and are collected in #400.

What I checked

• Execution: every code cell in both lectures runs without errors. gradual_liberalization_network takes about 4 seconds and organizational_equilibrium about 1 second.
• Numbers in the prose: every number quoted in the text matches the code output. Examples include the 14.5% Markov and 1.2% best constant tariffs, the organizational path 11.8 → 7.9 → 4.8 → 3.7%, Ramsey reaching free trade in year 31, the 17-year delay after the shock, and the exercise thresholds. In organizational_equilibrium they include the 85% welfare peak, the 0.23 dip in logs, and the four periods (28 years) before import-sector workers gain. I also checked claims the code doesn't print, such as the return to capital falling from 5.2% to 4.5% and the transition-map slopes (0 at the Markov tariff, about 1 at the best constant tariff).
• Derivations: the separability argument, the hat-algebra price equation, the revenue system (the code matches the equations), the Markov and best constant objectives, and the first-order condition with its thresholds of 4.48 and 7.95 are all correct.
• Ramsey solver: L-BFGS-B stops after 6 iterations, which looked premature, but re-solving with tight tolerances gives the same path.

Changes in 54d54a7 and 02c1ef6

Lecture	Change
both	Shortened six figure captions to six words or fewer (qe-fig-004)
gradual_liberalization_network	Bolded the defined terms "static separability" and "exact hat algebra" in the Overview (qe-writing-005)
gradual_liberalization_network	The 5, 10 and 30 percent grid indices are now found by value instead of hard-coded, so they stay correct if τ_grid changes
gradual_liberalization_network	Renamed the Ramsey horizon from N, which the text uses for the number of countries, to horizon
gradual_liberalization_network	Corrected the description of the sector bar chart: gains fall with baseline consumption for the three lowest-paid groups, but the two best-paid groups are not in order
gradual_liberalization_network	Dropped an f-string with no placeholders
organizational_equilibrium	Added a print line showing how the gain in equality splits between wage changes (0.113) and the rebate (0.061), which the text asserts but the cell did not compute
organizational_equilibrium	Reworded the Ramsey import-sector recovery to match the printout: back to the Markov level in period 5, above it from period 6

After these edits, both lectures produce the same output as before, plus the one new print line.

Open questions for the author

See #400. It covers an inference in the new lecture about what drives the G7 tariff in Bassetto, Huo, and Ríos-Rull (2026), two figures and the bib year for Bonadio et al., and three figures from Bassetto, Huo, and Ríos-Rull (2026) that I couldn't check against the paper. None of them affects the code or its results.

Recommendation

Merge once the CI tests job passes. #400 can be settled before merging or as a follow-up.

[github-actions]

🚀 Deployed on https://6ae31fb040327fbcb14700e2--lustrous-melomakarona-3ee73e.netlify.app