In a pioneered paper, Chamley (1986) studied an otherwise standard Ramsey growth model with identical and infinitely-lived households, capital accumulation, and perfect markets without failures such as externalities and monopolistic competition. Using this model, Chamley (1986) investigated the second-best problem of using factor income taxes to finance exogenously given government spending and established that the optimal tax rate on capital income is zero in the long run.
The result emerges because the capital income taxation distorts capital accumulation and thus creates dynamic inefficiency. Several studies have re-examined the issue and found the result to be robust; see, for examples, Lucas (1990), Jones, Manuelli and Rossi (1993, 1997), Chari, Christiano and Kehoe (1994), Chari and Kehoe (1999) and others. In particular, Atkeson, Chari and Kehoe (1999) found that the result of a zero capital tax rate in the long run was robust when they unified the work and relaxed all Chamley’s assumptions by allowing for either heterogeneous consumers, or endogenous growth, open economies and overlapping generations.
The reality in the taxation of capital income, however, is quite different from the above theoretical prediction. Capital income includes corporate profits, capital gains, dividends, interest income and others. Just taking the corporate income tax as an example, the top statutory tax rates on corporate income stayed as high as 43% in the US, 41% in Japan, and 23.5% on average in 27 European countries in 2009. Indeed, capital income taxes remain one of the major sources of the tax revenue across countries. For example, according to the Eurostat, total tax revenues as a percentage of the gross domestic product were 39.8% on average in 27 European countries in which over a quarter (28.7%) came from capital tax revenues in 2007.
The tax practice indicates the need for extensions beyond those made by these above-mentioned authors. This paper studies the standard Ramsey growth model and investigates under what conditions the optimal tax on capital is positive. We set up a Ramsey growth model with the factor tax incidence wherein the government uses the capital income tax and the labor income tax to finance exogenously specified expenditure. We argue that the elasticity of substitution (hereafter ES) between consumption and leisure in a given period plays a pivotal role in the determination of a positive optimal capital tax rate. The reason is that the factor taxation produces two conflicting effects.
First, a lower capital income tax generates an intertemporal efficiency gain toward savings and thus future consumption and hence is against the capital income tax in favor of the labor income tax. Second, a lower labor income tax creates an intratemporal substitution effect away from leisure and toward labor supply and thus output and consumption in a given period and is thus against the labor tax in favor of the capital tax. A sufficiently large ES between consumption and leisure creates an adequate intratemporal efficiency gain and thus a positive capital tax is optimal.
Download
Taxing Capital Income Is Really Not a Bad Idea When the Elasticity of Substitution Is Large
