Conference Agenda

This paper studies optimal taxation when marginal contributions to an externality differ between urban and rural regions (non-atmospheric externalities), and individuals can migrate from rural to urban regions to earn higher wages. I show that when the government is constrained to set a uniform commodity tax across regions, the presence of externalities alters the structure of optimal redistribution. In particular, if, for example, urban residents impose higher marginal external damages, the optimal policy features more redistribution, relative to a benchmark without externalities. This additional redistribution reduces the attractiveness of higher urban wages and thus discourages migration into the urban region.

We analyse optimal fiscal policy in an overlapping generations model with endogenous labour supply and savings, and two consumption externalities: aspirations and habits. Aspirations cause each generation’s consumption to influence the utility of the next, while habits imply that higher early-life consumption reduces the perceived utility of consumption in old age. To implement the social optimum, the government uses a positive tax on wages, and a subsidy on investments, received by the old. An additional positive (negative) transfer to the young (old) is required to attain the optimum when there is under-accumulation of capital at the decentralised equilibrium. When there is over-accumulation of capital, the sign of the transfers becomes unclear, and the possibility that both are positive arises. Numerical exercises reveal conditions under which equal transfers to all individuals, a Universal Basic Income (UBI), can be optimal. Our results reveal when UBI can effectively enhance welfare.

This paper explores whether heterogeneity of tastes invalidates Deaton's (1979) theorem in support of uniform commodity taxation. By extending the model of Boadway and Cuff (2022), I allow for a piecewise linear income tax and a population with both heterogeneous skills and tastes. It is shown that uniform taxes remain (i) efficient and (ii) optimal from a Rawlsian perspective if conditional consumption demand and earnings are independently distributed. In special cases, this criterion translates into independence between expenditure shares and earnings, or skills and tastes.

We study the optimal taxation of corporate and dividend income when entrepreneurs can use retained earnings to reduce their tax burden. We show that eliminating dividend taxes while increasing the corporate income tax (CIT) to keep investment unchanged raises total tax revenue. Our simulations suggest net revenue gains of 0.1-0.4% of GDP. In an infinite-horizon model, the optimal policy sets dividend taxes to zero in every period. As the discount factor approaches one and when the planner values only workers' welfare, the optimal steady-state CIT converges to a standard inverse-elasticity rule.