Personnel turnover is a structural preservation event for research data

Research is performed predominantly by temporary workers who leave by design. The median time from grad school start to PhD is 7.3 years, ~43% have not completed within ten years, the average postdoc lasts ~4.5 years, and only ~15-23% of postdocs eventually secure tenure-track positions (S-0026). The person who understands what a dataset is, how it was generated, and where it lives is always within a few years of leaving the institution that holds it.

The structural exposure this creates is measurable. In 133 popular GitHub projects, 65% have a bus factor ≤ 2 — two departures would leave the project effectively unmaintained (S-0027). At HLRS Stuttgart, 57 of 262 user accounts on the tape archive were de-registered, leaving ~619 TB of dark data without active stewardship (S-0028). The same pattern operates at every research institution at every level of granularity.

Institutional infrastructure treats every departing researcher as a preservation event it is not equipped to handle, then responds to the accumulated consequence as if it were unexpected. The architectural fix — preservation that survives independent of any individual's continued participation — is the M-0001 Tier 3 property. Every other "fix" (better onboarding/offboarding, stronger DMPs, mandatory deposit) is layered on top of the same single-point-of-failure architecture.