Evidence suggests that people understand qualitatively how tastes change over time, but underestimate the magnitudes. This evidence is limited, however, to laboratory evidence or surveys of reported happiness. We test for such projection bias in field data. Using data on catalog orders of...