The reason for the diverging results concerning the relation of profits in the two cases lies in the determination of wages. Here, a regionalisation of UI leads -compared to central UI - to higher (lower) equilibrium benefit rates in the rich (poor) region. A higher (lower) benefit rate unambiguously leads to higher (lower) equilibrium wages in the context of the assumed bargaining setup. This causes labour demand and profits to shrink (increase). The effects on expected utilities and thus on migration are contradictory, though. On the one hand, workers from the rich region have higher wages and a higher benefit rate. On the other hand, the probability of becoming employed is smaller in equilibrium, and vice versa for the poor region. With the assumed functions and parameter values, however, equilibrium migration from the poor to the rich region increases, if UI is regionalised. The costs connected with this additional migration have a negative impact on the efficiency measure, which supports thus central UI. Migration equilibrates the differences of the expected utiliy between workers because it increases the expected utility in the origin region, and it lowers the expected utility in the destination region. Hence, workers from both regions prefer central UI. The latter result is thus independent of which parameter of the UI serves to equilibrate the budget.