Policy Symmetry versus Trade Integration: A Multilayer Network Theory of Monetary-Union Sequencing in ECOWAS

Analyses reproduced

The master program reproduces:

1. Dataset and panel-structure validation
2. ECOWAS cluster assignments
3. Unordered country-dyad identifiers
4. Period classifications
5. Average external-hub exposure
6. Descriptive statistics
7. Full-sample cluster-pair results
8. Period-specific results for:
   • 1990–1999
   • 2000–2009
   • 2010–2019
   • 2020–2023
9. The manuscript’s 2,000-replication observation-level bootstrap
10. A supplementary dyad-cluster bootstrap
11. Year-fixed-effects diagnostic regression
12. HC1 robust standard errors
13. Coefficient confidence intervals and model-fit statistics
14. Layer-correlation matrix
15. Annual integration-depth trajectories
16. ECOWAS-wide layer trajectories
17. Centrality summaries
18. Automated numerical and structural tests

Main code files

• run_all.py, master replication program
• src/analysis.py, complete reusable analytical functions
• run_all.sh, one-command execution script
• tests/test_replication.py, automated verification tests
• requirements.txt, required Python packages
• README.md, complete execution and result-mapping instructions
• SHA256SUMS.txt, integrity checksums

Generated outputs

The package already contains the reproduced outputs:

• complete_replication_results.xlsx
• Cluster-result tables
• Bootstrap estimates and confidence intervals
• Period-specific results
• Diagnostic regression results
• Full regression summary
• Descriptive statistics
• Layer correlations
• Centrality summaries
• Three publication-quality figures
• Machine-readable run manifest

Execution

After extracting the ZIP, run:

Alternatively:

The replication was executed successfully against the supplied data. All four automated tests passed, including exact checks for the 3,570-observation balanced panel, 105 dyads, principal cluster means, reported coefficients, and diagnostic regression R2=0.750126R^2=0.750126. The ZIP archive also passed integrity testing.