Recent models of 'endogenous growth'allow savings and investment behaviour to have permanent effects on not only the level but also the rate of growth of per capita income – for example through their effect on technological change, which more traditional models view as exogenous. In Discussion Paper No. 606, Research Fellow Willem Buiter considers the growth effects of the level of savings, abstracting from the efficiency with which they are invested. With a production function that is linear in the aggregate capital stock, he develops an overlapping generations model of private saving behaviour in which the long-run growth rate of capital - and hence that of output - is the product of the private savings/output and output/capital ratios (net of the capital depreciation rate).

Buiter then uses this model to examine the effects of changes to savings levels on account of private attitudes towards intertemporal choice, demographic effects, the decline of labour productivity with age, and a variety of fiscal policy instruments. The latter include public consumption spending, taxation of wage and non-wage income, deficit financing and balancedbudget intergenerational redistribution - for example through an unfunded social security retirement scheme. He finds that the long-run growth rate is positively related to the time-preference rate, birth rate and life expectancy. On the other hand, increases in the tax rate on capital income or in public consumption financed by lump-sum taxation, the postponement of lump-sum taxation by borrowing, or a balanced-budget income redistribution from young to old will all reduce the long-run growth rate.

Buiter discusses three implications of these findings. First, while endogenous growth clearly allows changes in fiscal policy or the parameters that govern private behaviour to have long-run effects on the rate of output growth as well as on its level, their welfare implications are less straightforward than those in exogenous growth models, but policy is likely to matter more. Second, since the real interest rate is constant, government borrowing will reduce total saving and hence displace private capital formation: one-sector, endogenous growth models of this type cannot identify the effect of government financing on private capital formation. Third, since the analysis of the growth rate's response to private or government shocks is much more straightforward than in even the simplest models of exogenous growth, teachers of growth theory may do well to begin with the simple special case provided by the endogenous growth model developed here.

Saving and Endogenous Growth: A Survey of Theory and Policy
Willem H Buiter