Generalizing the Pareto principle with endogenous populations Draft (submitted)
previously circulated as When is a life worth living? Generalizing the Pareto principle with endogenous populations
Abstract: This paper introduces the G-rule and G-dominance which extend the Pareto rule and Pareto dominance when populations are endogenous. Societal evaluations of lives are required. The G-rule relies on these evaluations of lives and on individual preferences to compare pairs of allocations that do not require evaluating the gains of some agents (utility gains, benefits of being born, or benefits of not being born) against the losses of others. The G-rule is the largest social preference relation that never makes such inter-individual evaluations. G-dominance and G-efficiency naturally ensue. My G-efficiency encompasses the A-efficiency and P-efficiency of Golosov, Jones, and Tertilt (2007). Within the set of G-rules only the transitive ones i) can be extended by rational social preferences, and ii) do not impose any inter-individual evaluations to their rational extensions. Transitive G-rules can be conveniently represented with G-nondecreasing endogenous Critical Levels. Some well known Social Welfare Functions extend transitive G-rules. In a simple economy, the laissez-faire equilibrium is G-efficient. I study when fertility taxes/subsidies and bequest constraints yield G-(in)efficient equilibria. Finally, in presence of childbearing fiscal or environmental externalities, Pigouvian fertility taxes are necessary for G-efficiency.
