• optimal income taxation
• fiscal policy
• innovation-led growth

Chamley (1986) and Judd (1985) showed in a standard neoclassical growth model that taxing capital income cannot be optimal in the long-run. This conclusion has been challenged on various grounds. In particular, I will focus on the critical contribution by Aghion et al. (2013), who analyze the optimal use of labour and capital income taxes within a Schumpeterian growth model. By introducing innovation-led growth, they find that a zero tax rate on capital income may become suboptimal due to a market size effect. More specifically, the higher the level of public expenditure, the more should capital income be taxed. In fact, not taxing capital income means that labour income must be taxed at a higher rate. This, in turn, has a detrimental effect on labour supply and, thereby, on the market size for innovation. Also taxing capital income reduces innovation incentives. However, for sufficiently high public spending, the former effect dominates. The main limitation of the Schumpeterian analysis is that, by making the assumption of a single representative household, it ignores the effects of personal income distribution.