How does personnel turnover function as a preservation event for research data?

A subsidiary question under Q-0019 (shared structural property). §3.1 documents the personnel mechanism in the four-category §3 taxonomy.

The answer is C-0017: research is performed predominantly by temporary workers who leave by design. Median time to PhD is 7.3 years, the average postdoc lasts ~4.5 years, only 15-23% of postdocs secure tenure-track positions. 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. 65% of popular GitHub projects have a bus factor ≤ 2; HLRS Stuttgart found 57 of 262 user accounts de-registered, leaving ~619 TB of dark data. Institutional infrastructure treats every departing researcher as a preservation event it is not equipped to handle, and then responds to the accumulated consequence as though it were unexpected.

The architectural fix — preservation that survives independent of any individual's continued participation — is the M-0001 Tier 3 property. Procedural reforms layered on top of single-point-of-failure architecture do not reach the underlying property.