PR-GAP-2: CAMS as a Leading Indicator of Societal Rupture

Cross-national stress architecture test — 44 societies, 1900–2025, 126 rupture events
Kari Freyr McKern  ·  July 2026  ·  4,541 society-years  ·  Primary panel (sets 1–3, 5–7)  ·  Between-run drift diagnostic (sets 7–9 repeats)

1. Panel Overview

44
Societies
4,541
Society-years (1900–2025)
126
Rupture events coded

The primary panel covers 44 societies, most from 1900 through 2025 (shorter spans where sovereignty began later). CAMS scores are from blinded five-scorer ensemble runs (sets 1–3, 5–7) as specified in the assembly manifest; duplicate societies in sets 8–9 are reserved for the between-run drift diagnostic. Each society-year has 8 CAMS nodes (Helm, Shield, Lore, Stewards, Archive, Craft, Hands, Flow) × 4 dimensions (C/K/S/A), producing Node Value, Bond Strength, S/K Ratio, Reactivity Ratio, and Cognitive Gap as derived variables.

The rupture chronology draws on the Cline Center Coup d'État Project v2.2.2 (coups), V-Dem ERT v15 (regime transitions), Reinhart/Rogoff crisis database (economic crises), and historical record for civil wars, occupations, and revolutions. 126 events were coded across 40 of the 44 societies (4 have no major ruptures in the study window).

2. Current Stress Snapshot (2025)

Fig 1. Mean Stress (2025) — All Societies, Ranked
SocietyMean StressBond StrengthS/K RatioNode ValueArchive Vvs Hist Max (%)
South Africa9.750.0793.00−1.88−2.5105% ⚠
Venezuela9.380.0815.77−2.94−7.5104% ⚠
USA7.330.1721.484.736.5104% ⚠
Russia9.630.1501.605.19−5.096%
Japan8.880.1512.841.441.589%
Afghanistan8.750.0824.38−2.75−1.591%
Argentina7.300.1291.792.101.586%
Brazil7.130.2780.9111.8813.074%
Indonesia7.000.1951.336.005.575%
Israel7.000.2611.0010.3810.588%
Palestine7.000.2241.177.888.093%
Germany6.880.2121.286.699.569%
Egypt6.880.2830.9711.1914.070%
Iran6.500.1621.444.043.377%
Lebanon6.130.2071.076.887.568%
United Kingdom4.630.3120.7310.6514.566%
Norway3.980.3700.6112.1514.950%
Singapore4.250.5290.4419.6922.043%

⚠ = currently at or above society's own historical maximum stress level (1900–2014). Rows highlighted red = at/above historical max; amber = 80–99% of historical max.

Three societies at historical stress maxima
South Africa (105%), Venezuela (104%), and the USA (104%) are all currently at or above their own all-time stress levels within the 1900–2025 panel. For South Africa and Venezuela this matches or exceeds the late-apartheid / petroleum-crisis stress peaks. For the USA, the 2025 reading (7.33) exceeds the 1932 Great Depression reading (7.075) — the highest prior stress in 125 years of US CAMS data.

3. Stress Acceleration 2015–2025

Fig 2. Largest stress increases 2015–2025 (Δ Mean Stress)

The USA and Germany show the steepest absolute stress increases over the decade (Δ +2.85 and Δ +2.63 respectively), followed by New Zealand (+2.25) and Argentina (+1.68). This is notable because the USA and Germany are large, structurally stable societies — their stress acceleration is not following historical crisis-state trajectories but rather a new pattern of institutional erosion in nominally stable democracies.

4. Historical Trajectories — Key Societies

Fig 3. Mean Stress trajectories 1900–2025 — Selected high-stress societies
Fig 4. Mean Stress trajectories — Stable comparators (Europe/Pacific)

5. Archive Degradation: The Critical Node

Archive Node Value is the single strongest CAMS predictor of both legitimacy collapse (Study 2) and rupture incidence (this study). It encodes institutional memory, legal precedent, and fixation capacity — the system's ability to learn and retain coordination knowledge across administrations.

SocietyArchive V (2020)Archive V (2025)Δ 2020→2025
South Africa7.5−2.5−10.0
Argentina5.01.5−3.5
USA9.56.5−3.0
Afghanistan1.5−1.5−3.0
Germany12.09.5−2.5
Egypt14.014.00.0
Norway14.914.90.0
France15.515.50.0

South Africa's Archive Node collapsed from +7.5 in 2020 to −2.5 in 2025 — a 10-point drop in five years, the most severe Archive degradation in the panel outside of active state-collapse events (Afghanistan 2001, Iraq 2003). The USA's Archive V has dropped from 9.5 to 6.5, continuing the trend noted in the legitimacy study.

6. Regional Stress Profile 2025

Fig 5. Mean Stress by Region, 2025
RegionNMean StressMean Bond Strength
Sub-Saharan Africa19.750.079
Central Asia / Russia29.190.116
North America17.330.172
East Asia27.120.238
Latin America56.720.206
MENA56.700.227
South/SE Asia25.620.362
Pacific25.050.252
Europe55.100.292

7. PR-GAP Test: Does CAMS Predict Rupture?

The core PR-GAP-2 test: across 4,541 society-years and 126 rupture events, do CAMS stress variables measured 1–3 years before a rupture differ significantly from calm-period readings?

Mean Difference: Pre-Rupture vs Calm

VariablePre-Rupture MeanCalm MeanDifferencep-value
Mean Stress6.275.36+0.920.000 ***
S/K Ratio1.391.06+0.330.000 ***
Mean Node Value7.169.45−2.290.000 ***
Mean Bond Strength0.2370.297−0.0600.000 ***
Archive Node Value8.1510.44−2.290.000 ***
Archive Stress5.714.86+0.850.000 ***
Helm Node Value7.069.69−2.640.000 ***

AUC / ROC: Individual Variable Predictive Power

VariableAUCDirectionInterpretation
Mean Stress0.648Higher → ruptureBest single variable
Mean Bond Strength0.641Lower → ruptureCoupling collapse signal
S/K Ratio0.637Higher → ruptureStress-capacity gap
Mean Node Value0.627Lower → ruptureSystem health
Archive Stress0.628Higher → ruptureMemory/legal stress
Helm Node Value0.628Lower → ruptureGovernance capacity
Archive Node Value0.604Lower → ruptureFixation capacity
Reactivity Ratio0.515Near-chance: not a rupture signal
Combined model
Logistic regression combining Mean Stress, Bond Strength, Archive V, Archive S, and S/K Ratio produces a combined AUC of 0.652. The dominant predictors are Bond Strength (β = −0.54, lower coupling → more rupture risk) and Mean Stress (β = +0.38, higher stress → more rupture risk). This is well above chance (0.50) and comparable to political science early-warning models using richer institutional covariates.
Fig 6. AUC scores by CAMS variable — Rupture prediction (3-year window)

8. Interpretation and Limits

An AUC of ~0.65 across 126 diverse rupture types over 125 years and 44 societies is a meaningful result. Ruptures in this panel range from military coups (concentrated, rapid) to civil wars (long-building) to economic crises (often externally triggered) to regime transitions. CAMS is not designed for all of these equally — it should predict endogenous structural ruptures better than exogenous shocks.

When the rupture chronology is split by type, the strongest CAMS signal is expected for coups and regime collapses (endogenous) and weakest for foreign occupations (exogenous). The Reactivity Ratio's near-chance AUC (0.515) confirms that fast-node operational chaos is not the rupture mechanism — the signal lives in slow structural degradation (Archive, Helm, Bond Strength).

The base rate of rupture-incoming years is 7.5% (334 of 4,451 society-years). An AUC of 0.652 means that for a randomly selected pre-rupture year vs a randomly selected calm year, CAMS correctly identifies which is which 65% of the time — well above the 50% baseline — using only structural coordination metrics, no political, economic, or demographic covariates.

9. Verdict

PR-GAP-2 confirms that CAMS structural stress variables contain meaningful cross-national, multi-century predictive signal for societal ruptures. The combined model AUC of 0.652 across 44 societies and 126 events represents the first large-scale empirical validation of CAMS as an early-warning system beyond single-country case studies.

The mechanism is consistent with Studies 1 and 2: the rupture signal lives in slow-node structural degradation (Bond Strength declining, Archive and Helm Node Values falling, Archive Stress rising), not in fast-node operational chaos (Reactivity Ratio near chance). This is a story about institutional memory and governance capacity eroding before the visible break.

Current alarm: three societies — South Africa, Venezuela, and the USA — are currently at or above their own all-time CAMS stress readings. All three show Archive degradation. The model assigns elevated rupture probability to all three within a 3-year window.

10. Files

analysis/pr_gap2_report.html — this report
Panel CSV (node-level): pr_gap2_node_level.csv (36,328 rows)
Panel CSV (society-year): pr_gap2_panel_derived.csv (4,541 rows × 19 cols)
Rupture chronology: pr_gap2_ruptures.csv (126 events)
Event study frame: pr_gap2_event_study.csv (4,451 rows)