Scott Alexander, curated
← Back to curation

Is There An Illusion Of Moral Decline?

Quality
74
Strong
Claude Shift
53
Moderate
RWI
2
of 10

Summary

Critiques Mastroianni & Gilbert's 'The Illusion Of Moral Decline' on two levels. Methodologically: their 'objective morality' proxies are weak -- mostly 2000-2015 (too short for a decline people mean across generations), the one 1960s series (fear of walking at night) tracks neither violent crime nor morality, and the 'negligible change' verdict rests on percent-variance-explained framing that hides visibly large declines in social-trust series. Substantively: they measure too FEW things -- zero polls on sex, marriage, religion, drugs, patriotism, hard work (cf. Haidt's Moral Foundations). Scott's positive move is a generational-imprint model: each cohort absorbs its birth-era morality and later years score worse on it because they are aiming at different values -- which fits MG's own 'morality peaked around your birth year' data better than their rose-coloured-glasses bias. Declines to say whether morality is really declining (wealth makes industry/chastity less necessary, compassion/pacifism more convenient).

Why this score

Quality 74 · Strong. Strong, upper end. A sharp methodological takedown PLUS an original positive model that explains the target paper's own data better than the paper does -- the hallmark of a good Contra piece. Held just below Excellent because it is a response to one paper rather than a standalone synthesis.

Claude’s paradigm shift 53 · Moderate. Notable. The methodological skepticism is routine, but the birth-year-imprint account of perceived moral decline (vs a memory bias) is a fresh, non-obvious reframing.

Real-world impact 2 · Minor. Minor. Lives in the moral-psychology / replication-critique discourse; no institutional or material consequence. Within-niche.