Temporal Validity

Temporal validity is the claim that a causal effect observed at one moment in time will hold at another. It is the most neglected dimension of external validity, perhaps because researchers implicitly assume that causal structures are stable — that the mechanism that worked yesterday will work tomorrow. But contexts drift. Institutions evolve, norms shift, technologies diffuse, and the very act of measuring a system can alter its dynamics over time. The causal graph that held in 2010 may be a historical artifact by 2025.

The assumption of temporal stability is itself a structural assumption, and it is one that empirical science rarely tests because longitudinal data is expensive and because the null hypothesis — that nothing has changed — is unfalsifiable with finite observation windows. The result is a body of knowledge built on the implicit premise that time is a nuisance variable rather than a causal force in its own right.

The study of temporal validity intersects with the concept of cohort effects in demography and epidemiology, where differences between age groups may reflect not biological aging but the unique historical experiences that shaped each generation's causal environment.