Recent models of `endogenous growth' consider reasons why different countries grow at different rates for extended periods, which may include the long-term effects of tax rates and public expenditure levels. In Discussion Paper No. 630, Robert Barro and Research Fellow Xavier Sala-i-Martin discuss these fiscal effects for a variety of models of a closed economy. In their simplest model, production is linear in a broadly defined concept of capital. There are no transitional dynamics, and the economy remains on a path of constant, steady-state growth. If the social rate of return exceeds the private rate, steady-state growth falls below the Pareto-optimal rate. Such an excess of the social over the private rate of return may reflect learning-by-doing with spillover effects, the financing of government consumption purchases with an income tax, or monopoly pricing of new types of capital goods. Tax policies to encourage investment may raise both growth and utility.

Barro and Sala-i-Martin construct three models incorporating different types of public services as productive inputs to private production. First, if each producer has property rights to a specified quantity of such services (and can neither trespass on nor congest those provided to others), lump-sum taxation is superior to income taxation. Second, if each producer's production function includes the aggregate quantity of such services, lump-sum taxation is again preferred. Third, for public goods that are subject to congestion, lump-sum taxation (with no user fee) leads to excessive private use of the public good. Income taxation may approximate a user fee and therefore outperform lump-sum taxation, which entails incentives for investment and growth that may be too high. Barro and Sala-i- Martin suggest that this `congestion model' may apply to a wide array of public services, including transport, public utilities, courts and possibly national defence and the police.

These models all predict a positive correlation between movements in the interest rate and growth, which is difficult to detect empirically for per capita growth of consumption, output or capital. This reflects an asymmetry in the model: consumers require a premium on future consumption that is increasing in the growth rate, of which the private rate of return on investment is independent. But the private rate of return on investment that entails adjustment costs may be diminishing in the growth rate.

Public Finance in Models of Economic Growth
Robert J Barro and Xavier Sala-i-Martin

Discussion Paper No. 630, March 1992 (IM)