When a monthly economic indicator series contracts sharply for a few months and then starts to recover, the published annual and monthly growth rates can give conflicting signals: the annual growth rate can indicate a decrease and the monthly growth rate an increase or vice versa. This is well known to the seasonal adjustment community, see, for example, Shiskin (1957). In this paper, we revisit, illustrate and then explain this potential for conflict more analytically. For example, the annual differences lag the monthly differences by five and a half months because the same-month-year-ago difference is the sum of the current and eleven preceding monthly differences, and the annual sum has a phase shift of five and a half months. Illustrative examples are followed by an elementary formal mathematical derivation using the gain and phase functions of the annual sum.