Cross-Domain Stress Tests of Historical-Depth Non-Redundancy Strong-Baseline, Seasonal, and Cycle Controls Beyond the Fertility Panel

This paper tests whether historical depth retains non-redundant explanatory and predictive value across domains once stronger contemporary baselines and temporal controls are introduced. Building on an earlier fertility-panel analysis, it extends the design to multiple domains in order to distinguish genuine historical dependence from effects that can be absorbed by recent-state information, seasonal structure, recurrent cycles, or other short-horizon predictors. The study compares historical-depth variables against strong baselines under controlled specifications that include seasonal and cycle adjustments, and evaluates whether the historical contribution survives across datasets, targets, and operating regimes. Historical depth is treated as non-redundant only when its removal produces a measurable loss that cannot be recovered by the competing baseline or temporal controls. The paper therefore separates mere temporal correlation from historically persistent constraint and tests the external validity of historical-depth effects beyond a single panel. Its broader methodological claim is that claims about path dependence or historical influence should be supported by cross-domain stress tests in which recent-state, seasonal, cyclic, and domain-specific alternatives are made maximally competitive. The resulting framework provides a general procedure for determining when historical information changes reachable prediction or explanation rather than merely redescribing structure already captured by stronger contemporaneous variables.