This paper presents empirical evidence which at first glance appears to show that firms are not pricing to maximize profits. We then present a model to explain how this behavior is in fact optimal when we account for additional constraints faced by firms in certain product markets.