Labour, whatever its geographical origin or its other characteristics, is an input to the production process. Therefore, the issue of the impact of migration upon the utilisation of native workers and/or the rate of return of their services has to be studied within a production-theoretic approach. Ulrich Kohli has done this using more disaggregated data then those used in this chapter (Greenwood et al., 1996). In chapter 6, however, the author starts from the observation that several researchers have indeed studied the migration issue using a production-theoretic approach, but they have surprisingly not taken into account the fact that, rather than ‘importing’ workers, the domestic economy can import the goods that those workers contribute to produce in the home country. This remark brings up the issue of trade as a possible alternative (or actually as a complementary activity) to labour-factor mobility. Hence the interest in characterising the substitution possibilities in production between ‘imported’ labour and imported goods.

Two brief remarks come to mind at this point. First, the production-based analysis of migration invariably holds the number of hours worked per worker constant and concentrates on changes along the ‘extensive margin’ of labour.