WHY FRANCE NEEDS A PRESIDENT WITH BUSINESS EXPERIENCE
Analytical supplement, Dr. Driss Effina, Global Strategy Files, 25 October 2026

This supplement contains the original data transcription, explicitly illustrative
models, figure source files and manuscript used in the 30-page article.

Observed data: data/insee_fiscal_2020_2025.csv, sourced to INSEE's 28 August 2026
annual accounts. Forecasts quoted in the article are external Banque de France
projections. They are not outputs of the illustrative model.

All scenario assumptions: data/assumptions.json. These are not estimated policy
effects and must not be used as forecasts of a named candidate's performance.
Source URLs and bibliographic details: research/sources.json.

Reproduce from the project root with Python 3:
  python research/analyse.py
  python text/manuscript.py
  python research/tables.py
  python research/build_article.py

Packages: numpy, pandas, matplotlib, python-docx. PDF conversion requires a
compatible Word or LibreOffice installation. The supplied PDF was rendered from
the supplied DOCX and visually reviewed. All eleven figures are also supplied
as PNG and SVG files. OMML equations remain editable in the Word file.

Debt accounting uses end-2027 starting debt of 120% of GDP, a 2.5% initial
effective rate, 15% annual repricing, 1.8% annual GDP-deflator inflation and
zero stock-flow adjustments. Paths for growth, primary balance and the rate
on new borrowing are explicitly supplied for 2028 to 2032. No endogenous
tax, inflation, monetary or financial-market feedback is fitted.

The Shapley decomposition averages marginal contributions over all six
orders. The stress grid contains 4,851 equally enumerated combinations,
not a probability sample. Its range is not a confidence interval.

Project values hold the last benefit year fixed at year 10. Delays therefore
shorten the benefit period. The probability of political durability is
conditional on technical completion. The CES model uses equal weights,
rho = -2 and elasticity 1/3. None of these parameters is estimated for France.

The article's recommendation is an argued judgement about transferable
operating competence. It is not a causal claim that businesspeople always
govern better, nor an endorsement of a named candidate.
