Generation-adjusted discounting in long-term decision-making

Stefan Bayer
2003 International Journal of Sustainable Development  
The usage of benefit-cost-analyses in long-term decision-making is common sense in neoclassical models. It is assumed that there is no difference between short-term and long-term projects. Therefore, no special adjustments have to be made in the analytical framework. This is especially true for the choice of the discount rate, the question of what is to be discounted, and the discounting procedure. Most commonly, economic theory analyzes long-term costs and benefits in the framework of optimal
more » ... amework of optimal growth models in tradition of Frank Ramsey: 1 A representative agent is supposed to maximize its lifetime-utility subject to a given budget constraint. The agent lives as long as the planning horizon of a specific project wants him to live. Taking natural resources theory as example, Solow (1974) as well as Stiglitz (1974) assume, that the planning horizon is infinitely long which means that the agent lives infinitely as well. 2 However, individual lifetimes are, of course, limited. This implies that we should use models with finite lifetimes of all agents for more realistic findings, OLG-models for instance. 3 But these models all assume the existence of long-term optimal growth paths. 4 Ad-hoc policy -for example to prevent further anthropogenic climate change due to carbon emissions -is not necessary because of perfect foresight and all-time rational behavior. Certainly, reality cannot be depicted as assumed in these models. New insights in climatic interdependencies for instance force policy-makers to react, especially when there exist (intertemporal) externalities. Therefore, our analysis goes as follows: We want to concentrate on discounting single projects. We do not investigate the discounting process within a first-best world and the necessity of all time optimality. Our discountiung considerations relate to calculations of societal present values to determine whether a long-term, market failure correcting project is favorable or not. 1 See Ramsey (1928), extended and applied in Barro/Sala-i-Martin (1995) . However, most applications do not exactly refer to Ramsey's original work because he strictly refuses utility discounting as "ethically indefensible". 2 Ironically, in an answer to critical remarks made by ), both Solow (1997 and Stiglitz (1997) stress that the planning horizon in fact is not infinitely long. They only wanted to get approximately "good" results for a planning horizon of 60 to 70 years. The long-held discussion about the convergence of the utility integral with conventional assumptions, especially positive utility discount rates until forever, could have been avoided if they had made this point clearer in their 1974 articles or in following publications.
doi:10.1504/ijsd.2003.004187 fatcat:czc3asioqzdwbm5wvumslzlxiu