The Hamilton method for estimating CPI bias is simple, intuitive, and has been widely adopted. We show that the method conates CPI bias with variation in cost-ofliving across income levels. Assuming a single price index across the income distribution is inconsistent with the downward sloping Engel curves that are necessary to implement the method. We develop and implement the Translated Engel curve (TEC) method that disentangles genuine CPI bias from dierences caused by comparing changes in the cost of living across dierent income levels-non-homotheticity. The TEC method gives substantially dierent estimates of CPI bias prior to major reforms to the CPI in 1999 (post-Boskin), but both methods suggest very little CPI bias thereafter. * Almås acknowledges valuable support from Vetenskapsrådet (the Swedish Research Council), ESOP,