This distributional analysis of the PC, however, ignores the effects of immigration on domestic skill-supply decisions. For example, the PC (2006) finds that a 50 per cent expansion of skilled immigration causes a reduction of 13 per cent in the salaries of Medical Practitioners, relative to the base case. An expectation of this slower growth in salaries should reduce the demand for medical degrees, with negative consequences for the size of the education sector.^[13]

In thinking about the issue of investment decisions, it is useful to consider a familiar textbook Ramsey growth model.^[14] In this model, there are two factors: capital and labour. There is also a unique steady state in which the capital–labour ratio is fixed. A one-off increase in the capital stock puts the economy on a transition path that leads back to the steady state. Along this path, consumption is higher than the steady-state level and investment is lower. An analogous argument in a model with human capital and labour suggests that a one-off increase in skilled immigrants would result in transition with higher consumption but lower human capital investment. Hence we have a ‘crowding-out’ of education through skilled immigration.

A similar intuition lies behind the theoretical results of Shea and Woodfield (1996), who consider the steady-state effects of skilled migration on skilled and unskilled wages.^[15] The intuition is also similar to the mechanism that underlies recent simulation results presented by Harris and Robertson (2007). They solve the perfect foresight path for education decisions under the scenario of a temporary expansion of the skilled-migrant intake. According to their estimates, a temporary increase of 20,000 immigrants per year over 15 years (which accords roughly with the increase in migrants between 1999–2005 whose stated occupations was 'Professionals') results in an 8 per cent fall in student enrolments in the short run. Though many caveats would apply to this figure, nevertheless it contrasts sharply with the modelling studies discussed above.^[16]

These considerations potentially reverse the earlier conclusions about the distributive effects of skilled migration. With a fixed supply of skilled labour, the burden of adjustment falls upon the wages and salaries of skilled workers. Recognising that incoming workers can substitute away from education, however, changes this story. In this case, skilled workers need not experience slower wages and salary growth to accommodate the increased supply of migrants. Rather, the burden of adjustment falls mainly on the education sector.