Nice in theory, just doesn't work in practice. There are two main problems with company or even divisional stacking: One, inter-team comparative rankings are e

Nice in theory, just doesn't work in practice. There are two main problems with company or even divisional stacking: One, inter-team comparative rankings are essentially meaningless. Imagine you have two teams of roughly equal competence. When asked to rank employees one manager is a bit of a soft touch and gives all his team members good-high rankings. The other is a bit honest and ranks most in the middle, with some low and some high. Some managers might be weak, others lazy, or may simply have a difference understanding of the rankings to other managers. The result is that the group people who get let go are arbitrarily determined by incomparable manager rankings - when you look at it this way, forcing every manager to fire their least effective employee is a blunt, but fairer approach. Of course if you have some magic formula to truly discover the bottom 10% of an entire company...maybe in a call center where everyone can be measured by calls/day...but in a software company what would you do? Measure people by KLOCs? That craziness ended in the 80s (except maybe in IBM). At the end of the day manager rankings (and perhaps peer rankings like MS) are all you have - and as they are essentially incomparable, what else can you do? If you have a better way, you should start a management consulting firm. Two, when a company is engaging on a total staff reduction, it needs to spread the cuts otherwise it might see entire crucial areas and teams disappear. I think we can all agree here that sales/corporate/IT are generally far less effective employees than engineers, but a company can't just fire its entire sales division. That fact alone, by definition, implies that some more-competent staff will need to be fired while less-competent ones are kept, simply to maintain a necessary skill balance. Firing by team is an effective, if crude way to reach that goal. I think this is one of those areas where you have to accept there are no easy answers, and the "least worst" solution is the best one.