The equally big flaw is that it focuses on the _average_ employee rather than the _marginal_ employee. If you want to decide to hire another employee, your con

The equally big flaw is that it focuses on the _average_ employee rather than the _marginal_ employee. If you want to decide to hire another employee, your concern is not how productive your average employee is. Your concern is whether the next employee will be a net positive or a net negative. I'll bet that Wal-mart's marginal employee profit is almost exactly $0. In other words, Wal-mart has hired so many people that there is no benefit to hiring one more, and has been so efficient in getting lean that there is nothing to be gained from firing somebody. (Of course, that ignores individual performance, but on average.) The same analysis might be true for Goldman. Even though the average Goldman employee is raking in hundreds of thousands for the company, it's not clear that Goldman would profit an extra $200,000 just by hiring the next average employee. It's very likely that Goldman has a limited number of opportunities to invest and that the current staff is enough to handle those and no more.