[sokoloff]

My experience is that a vast majority of managers would prefer to not face the fact that some of their employees are underperforming.

There are a variety of psychological or procedural "tricks" to get managers to be more discerning in evaluating that. Forced curve stack ranking is just one of them (procedural).

Another (psychological) is to ask managers to rate their employees and separately, to answer the question "Knowing everything that you know now, would you hire this person again?" There is a fair percentage of "this person is meeting expectations/highly valued/<whatever the middle ranking is>" combined with "yeah, in retrospect, we made a mistake by hiring them."

If you have absolutely no controls over the curve, many managers of modestly performing teams will earnestly report that their team is a mix of 25% superstars and 75% stars.

I don't know what the best answer is here, and the best answer certainly varies by company and might even vary by department within a company.

[AtlasBarfed]

A lot of that is that weird headcount thing that translates to how important you are as a manager.

- if you have a lot of reports YOU ARE IMPORTANT

- if you have a big budget (because of lots of reports) YOU ARE IMPORTANT

wait, firing someone might take away a req/headcount. I AM NO LONGER AS IMPORTANT

To the point that managers that don't have more than, like, 4 or 5 reports basically should plan and aren't a "I do a workers job plus management", regardless of company, should know they are going to get laid off or fired at some point.

You're right, it is a forced culling. I can see it as being done for a couple years to cull fat, but as a permanent fixture of the business practices? Well then it's your culture, and your culture is backstabbing, machiavellian, ass kissing, undermining, outright sabotaging, juking stats, etc.

[tomohawk]

Peer review.

After WW II, there was a large rif (reduction in force) in the US military. The Marines decided to use this to their advantage

At each level they did peer reviews, where the results were one of a) retain at current grade, b) retain, but at lower grade, c) do not retain.

This was all done with peer reviews of people at the same level. The result was they retained most of the best people they needed to retain.

I've never seen promotion boards or managerial review produce very good results. In largish orgs, you might as well just hold a lottery.

[rwmj]

The Apollo astronauts also had to rank each other, with that feeding into who got assigned to missions. (I'm not saying this was or was not a good idea - probably too small a sample size and they were all superstars - only that it was a thing that happened). [Source: https://space.stackexchange.com/a/23149]