I'm not sure this is the argument the author is making, but you could claim that the rest of the argument is true for any (or most) reasonable measure of employee performance that a company actually cares about.

▲

nradov 10 months ago | parent [-]

You could claim anything, but is there hard quantitative data to support such a claim? Or are we just guessing?

	▲	michaelmior 10 months ago \| parent [-]
		The author presents some data in the article. Also, the absence of hard quantitative data doesn't necessarily make it a complete guess. (At least not any more than starting with the assumption of a Gaussian distribution.)