Abstract
I develop a model which nests both Ricardian equivalence and an alternative non‐Ricardian theory. The alternative is a stochastic variant of Blanchard's [1985] model, and its empirical implications are derived. I then test Ricardian equivalence against this alternative using data from nineteen countries. Ricardian equivalence is resoundingly rejected. Moreover, the estimated deviation from Ricardian equivalence is roughly what one would expect if households faced perfect insurance markets and did no have altruistic bequest motives. For many purposes, however, the estimated deviation from Ricardian equivalence has no practical import.

This publication has 19 references indexed in Scilit: