Recent highly cited research uses time-series evidence to argue the decline in interest rates led to a large rise in economic profits and markups. We show the size of these estimates is sensitive to the sample start date: The rise in markups from 1984 to 2019 is 14% larger than from 1980 to 2019, a difference amounting to a $3000 change in income per worker in 2019. The sensitivity comes from a peak in interest rates in 1984, during a period of heightened volatility. Our results imply researchers should justify their time-series selection and incorporate sensitivity checks in their analysis.