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

==== ### ====

Giuntella, Nicodemo & Vargas‑Silva (2015) link Hospital Episode Statistics (2003–2012) with local immigration and find:
* reduced outpatient waiting times on average,
* no significant effects on A&E and elective care waiting times,
* but increased outpatient waiting times in more deprived areas outside London in the period right after the 2004 enlargement, which vanished in the medium run (3–4 years).

This is a good example of what you asked for: a non-pecuniary welfare channel (service quality / waiting time as rationing “price”) where the average effect is not strongly negative, but distribution across places matters.

===== Geay, McNally & Telhaj use the National Pupil Database (England) and study effects of non-native English speakers in classrooms. Their results suggest small and often inconclusive effects for reading/writing and a small positive effect in maths in some specifications, consistent with the idea that primary-aged students catch up in English proficiency and do not impede peers much. =====

Again: not obviously a large negative welfare externality on average, but local resource constraints could still matter.