The question of inter-municipal migration is interesting for both empirical and theoretical reasons. Migration has grown more and more extensive in Finland and concerns a wide variety of elementary choices, both individual and collective, which are of immense importance to the functioning of the market economy. This thesis provides theoretical clarification of the issue of spatial resource allocation with special emphasis on the migration of people. The thesis examines the effects of migration into and out of municipalities on the welfares of their current and potential residents and on social welfare. The focus is on economic efficiency, and the main intuition is that migration should be a vital element in the real-world mechanism of resource allocation. The thesis consists of six essays integrated into a single work, the logical structure of which is given by three approaches called the Partial Equilibrium Approach, the General Equilibrium Approach and the Club Theoretic Approach. The Partial Equilibrium Approach is based on a synthesis of the classical model of inter-regional labour migration and the model of trade policy. It is found that the short-run effects of labour income taxation on migration and welfare are asymmetric in out- and in in-migration municipalities: taxes imposed in an out-migration municipality cause effects on migration and welfare only if the tax is decidedly high, but taxation in an in-migration municipality causes out-migration and causes welfare losses even for a small tax. However, the effects remain minor if the out-migration municipality is small and rich in capital, and/or if the in-migration municipality is big and rich in capital. Thus, at least in Finland, the short-run effects remain quite insignificant in practice. The General Equilibrium Approach applies the seminal neo-Keynesian macroeconomic model to the analysis of inter-regional factor migration. Four types of market adjustment are specified, namely price adjustment, nominal wage adjustment, opposite adjustment and parallel adjustment. The effects of labour migration in particular on local interest rates are shown to depend on the specific adjustment path. From a theoretical point of view, an interior solution of spatial resource allocation is not in general granted. The model is used to examine the effects of local policy on migration, and an AEH-type explanation is used to describe people’s perceptions of the policy. A fair policy programme, which does not distort the real wage comparisons between localities, does not affect migration. On the other hand, an unfair programme is shown to encourage migration the more the farther the anticipations concerning the repayments are from the true nature of the program. In the longer term, when people correct their perceptions concerning the repayments, the effects are mitigated and eventually reduced to zero, provided that the programme is actually fair. In the Club Theoretic Approach, welfare in a city depends on the size of the city because of agglomeration economies and diseconomies. Technological and pecuniary externalities are internalised in the city and explain migration and the formation of cities. It is shown that migration may lead to non-optimal solutions, even when the city itself acts as a true market agent in choosing its optimal size from a within-club viewpoint. The results of the effects of national level policy somewhat contradict the conventional wisdom. First, administrative and economic policy measures are shown to differ in their effects, and second, if cities are initially outside their within-club optima, lump sum transfers rather accelerate than stabilise spatial evolution. The competitive dynamics of the system is a product of people’s exit-type choices (migration) and their collective voice-type choices. The use of local efficiency enhancing measures depends on the pressure of exit. A necessary condition for dynamic efficiency is that the migration equilibrium is non-stable. Dynamic efficiency is more likely to emerge between initially small than initially large cities. The pressure for efficiency is weaker in the most relevant case of asymmetric cities.