Replication Package for "The Determinacy of Effective Demand: A Dynamical-Systems Generalisation of Marglin's Price-Adjustment Models"

This is the replication package for "The Determinacy of Effective Demand: A Dynamical-Systems Generalisation of Marglin's Price-Adjustment Models" (J. G. de A. Oliveira, 2026). It reproduces every numerical result in the paper from public data, with no proprietary software, in under ten minutes on a standard desktop. The code has two self-contained entry points. reproduce_tables.py recomputes the five-model determinacy taxonomy and every value in Tables 2–4 analytically, from the closed-form equilibrium Jacobians in the Appendix; it runs no simulation and is the canonical check that the paper is internally consistent. nur_core.py (mirrored in base R as nur_core.R) reproduces the near-unit-root identification of the output-adjustment speed theta2 on the FRED quarterly series — OLS AR(1), augmented Dickey–Fuller, the Andrews (1993) median-unbiased confidence interval, and the Anderson–Rubin identification-robust set. A base-R script, verify_fred_data.R, independently re-downloads the FRED series and re-runs the full identification at the source. Additional scripts regenerate the paper's phase portraits, impulse responses, empirical IS-gap series, natural-experiment charts, sensitivity map, and theta2 forest plot. Contents: documentation (README, CODEBOOK, LICENSE, requirements), analysis and simulation code in Python and R, and one public data file (data/core_quarterly_fred.csv) of quarterly U.S. macroeconomic series from FRED (Federal Reserve Bank of St. Louis). The FRED series are works of the U.S. Government and are in the public domain; fetch_data.R rebuilds the file from source. Requirements: Python 3.12 (numpy, scipy, matplotlib) and base R ≥ 4.0. The code is released under the MIT License and the compiled data file under CC-BY-4.0 (see LICENSE.txt).