That's the problem with just reading data reports at scale.
Hiring 30k people so recently then firing almost half of them the next year, even filtering a tiny quarter of a percent is going to create outliers who get affected with serious life changing events.
> Doesn't have to inevitable, though. They literally have all the data.
Even if you have all the data, lets say there is a 4 person team being reduced to 3 during this redundancy - What do you do? Fire Anne instead of Peter purely because Peter joined the team more recently? Sure that sounds more fair for Peter, but it's less fair for Anne.
Someone else suggested putting Peter at risk - inform him that in 3 months he will be part of a layoff, so that he has those 3 months to find a new job.