I have wide experience with a very broad variety of companies' disaster recovery infrastructures, for over a decade. To date, without exception, not a single one of those DR infrastructures and plans include a provision for, "...and we lose X% of IT staff in a random distribution all at once during the DR incident, where X>30". This has implications.
1. The probability of 30+% of the IT staff suddenly not working the next day is considered vanishingly low, even in DR contexts where they routinely deal with low probability events.
2. Even if (1) was to come about, IT management pretty universally agrees with each other that IT staff are generally replaceable on short notice without irreversible consequences to the organization.
The case for organizational data disappearing and causing irreparable harm to the organization is pretty well established and accepted by IT management. It is not at all clear to me that even if an IT union entrenched itself in a major company, that a mass walkout would be taken that seriously, because I haven't yet seen the case made and accepted that IT organizational staff disappearing causes the same or greater harm.
The corollary implication is that IT organizations commit more process capital to retention of the bits sitting on spinning rust than to staff retention. Staffing spend remains an overwhelming percentage of the annual budget of most IT organizations, and if it isn't well-accepted that this staff disappearing all at once causes catastrophic damage, then I can't be surprised when there is so much assiduous effort at reducing that spend.
I don't agree with this mainstream IT management outlook on the matter, but most IT management is constrained to what they can demonstrate in financials and budget spreadsheets to senior management. Slowly degrading capabilities don't show up in those sheets until very late in the game.