Editing Openai/69640cff-e5e0-8003-a8d1-a78b118ca18b (section)

==== 4) “You ignored long-run institutional/trust effects (Garrett Jones / non-WEIRD immigration).” ====

Two things can be true at once:
# Cultural persistence / intergenerational transmission is real. There’s serious evidence that traits like trust have an inherited/cultural component among descendants of migrants (in multiple settings).
# The leap from ‘persistence exists’ → ‘UK institutions degrade in the long run from today’s migration composition’ is not empirically settled. The diversity–trust literature is mixed and effect sizes are often small, and results depend a lot on segregation, deprivation, measurement (generalized vs neighborhood trust), and causal identification.

What does the best synthesis say? A narrative + meta-analytic review of the ethnic diversity–trust literature concludes there is substantial heterogeneity; negative associations exist in many studies but are often small and sensitive to design/measurement.

And the UK-specific evidence is not one-sided:
* A well-cited London neighborhood study finds ethnic diversity is positively related to perceived social cohesion once deprivation is accounted for, while segregation predicts lower cohesion.
* Another UK-focused study (Citizenship Survey data) finds a negative association between ethnic diversity and trust overall, but does not support the claim that language fluency is the mechanism for whites; for non-white groups it finds more complex patterns.
* Micro-context causal-ish work from Denmark finds negative trust effects when exposure is extremely local (within ~80m), and that effects can vanish at more aggregated levels—again pointing to “how diversity is encountered” rather than a simple country-of-origin story.

So if your claim is: “institutional/trust externalities could swamp the tiny wage triangle,” that’s conceptually possible. But if your claim is: “the literature shows large, durable institutional harm from non-WEIRD immigration in rich democracies,” that’s not a safe read of the evidence.

My own take: the non-pecuniary channel is where the sign could flip, but it’s also where the empirical base is weakest and most context-dependent. In the UK, segregation + housing scarcity + stressed local services are plausible mechanisms for negative welfare effects even if wages/GDP look fine. The “country-of-origin institutions transplant” story is harder to validate cleanly in UK data compared to those more immediate channels.