Kant on Happiness as the Agreeable (Not Good) End of Prudential Rationality

For Kant, prudential rationality is governed by hypothetical imperatives telling us what actions would be good to take as means to the end of happiness. Commentators commonly suppose that such means are good only if they are means to a good end, namely happiness as a non-instrumental good whose goodness is conditional on good willing. But goodness is assessed by reason alone, while our attraction to happiness as enduring agreeable feeling is merely sensible. Prudential goodness is thus restricted to instrumental goodness as assessed by reason and, since happiness is not a good end, to merely instrumental goodness: the means to the end of happiness are good simply because they are good for happiness as the end one happens to have, not because they contribute to the realization of a good end. Denying that happiness is a good reflects Kant’s dualistic conception of the human being and human agency.