I agree with parts of the Robert Skidelsky argument supporting restrictions on immigration but not with others. I agree that the value of diversity are overstated (we want some but shared values also of value and we do not want ISIS) and that public sector tax benefits from immigration are like a Ponzi scheme. But I disagree with the Skidelsky analysis of labour market effects.
Skidelsky says that more migrants mean lower wages but that, in time, the benefits to capital owners will encourage more investment thereby offsetting the initial wage decline. Skidelsky argues that the lag between the initial wage and the final investment effects is likely to be excessively long.
This is generally a wrong view of standard immigration arguments. Wages do fall and profits do increase but the standard economic argument is that gains to capital exceed losses to workers. It is a trivial bit of microeconomics to show this. Thus the argument against migration is that while it realises efficiency gains – total incomes rise – it worsens the functional distribution of income. Thus the argument against immigration is a distributional argument.
I have neglected capital flows and the easiest assumption is that there is perfect capital mobility internationally. Then the initial wage fall creates a boost to the marginal product of capital that creates a capital inflow. Theis occurs until wages are driven back to their initial levels. Then wages for the original workers are as before but the economy has a larger capital stock. All gains from the initial migration accrue to the new migrant arrivals.
If capital is imperfectly mobile – the realistic situation – then the outcome is as somewhere between the situation of zero international capital flows and perfect capital mobility so that economic outcomes are, again, on balance, bad for local workers.
The economy gets bigger in this latter case and the new arrivals as well as local capitalists gain advantage but workers in the original pre-immigration economy lose out.