Paradoxes in Lucas's 1988 model with variable returns

The point of this paper is that the introduction of variable returns to scale in Lucas's 1988 model can produce paradoxical results: there can be negative relationships between the rate of growth of income and the rate of growth of human capital, between physical and human capital, and between the rate of growth of human capital and the rate of growth of wages. Moreover, the author argues the following "ontological paradox": the main equations of Lucas's 1988 model depend on the cardinal measurability of the external effects of human capital represented by a variable that is not cardinally measurable.