Using an overlapping-generations small open economy with endogenous lifetime ` a la Chakraborty (2004), we show that an increase in public investments in health is beneficial to life expectancy but can actually reduce both saving and domestic income per worker. Moreover, using the notion of A -efficiency introduced by Golosov et al. (2007) in a context of endogenous population (i.e., Pareto efficiency applied to people alive in every state), which contrasts with the P-efficiency criterion with respect to which the preference profiles of both born and unborn agents are evaluated in every state, some normative conclusions can be drawn: although the financing of health investments reduces both the disposable income and marginal propensity to consume of young individuals through increased labor income taxation, it also increases the length of life and can represent an A -Pareto improvement when the world interest rate is sufficiently high, because consumption when old becomes more attractive in such a case. Indeed, there exist: (i) a whole range of health tax rates that can effectively be used to increase social welfare, and (ii) an A -Pareto-efficient allocation. Moreover, the numerical simulations presented for some actual small open economies reveal that raising public health spending above the existing level causes both saving and domestic income to fall in all countries, while also finding that developed and underdeveloped (transition) economies are currently underinvesting (overinvesting) in health.