We quantify and compare the size of the welfare losses arising from the use of alternative 'imperfect' welfare indicators as substitutes for the conventionally preferred consumption indicator. We find that the size of the welfare losses associated with different indicators varies considerably....