Spectral Auditing of Cliometric Data: Separating Systemic Phase Transitions from Preprocessing Artifacts
Quantitative history and economics (cliometrics) rely heavily on linear dimensionality reduction, standardization, and spectral clustering to detect historical market integration and structural regime shifts. However, archival data matrices (N × P) frequently suffer from non-stationary secular trends, nested regional duplicates, and localized single-entity shocks ("lightning-rod" outliers). When processed through standard pipelines without prior spectral diagnostics, these pathologies generate false discoveries or mask genuine historical crises. In this paper, we apply a Two-Echelon Spectral Audit Protocol to three benchmark cliometric datasets: the Maddison Project Database (1820–1950 and 1500–1820) and the Allen–Unger Global Commodity Prices archive (1500–1800). We demonstrate that standard Z-score Singular Value Decomposition (Bypass mode) produces an illusion of cross-country structural coherence (normalized Participation Ratio PR(v1) = 0.998) by conflating shared exponential growth with synchronous business cycles. Conversely, naive application of the symmetric normalized graph Laplacian (L_sym) without Leave-One-Out (LOO) feature stability checks misidentifies localized peripheral shocks—such as Peru's post–War of the Pacific debt restructuring culminating in the 1890 Grace Contract, or the 1866 Australian financial deflation—as global phase transitions (p = 1.6 × 10⁻⁶ collapsing to p = 0.652 upon single-country removal). Finally, on primary European wheat prices (1650–1800), we show how First-Echelon participation diagnostics expose nested regional duplicates (34.5 percent of Mode 1 energy) and extreme temporal localization (PR(u1) = 0.155), whereas log-differencing combined with L_sym restores temporal uniformity (PR(u2) = 0.585) and isolates major 18th-century climatic and political shocks. Full reproducible Python code, Jupyter/Colab notebooks, and benchmark datasets are openly available at: github.com/avtonomov-petr/cliometric-spectral-audit
Authors
- Petr Avtonomov (ORCID: https://orcid.org/0009-0003-2816-6782)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-20
- DOI
- https://doi.org/10.5281/zenodo.24398830
- Primary Topic
- Historical Economic and Social Studies
- Type
- preprint