Managers in the music industry closely monitor both radio airplay of an album as well as the album's
sales. Their interest in radio airplay is due to the belief that airplay can increase an album’s sales. Therefore it
is natural for managers to attempt to influence radio airplay so as to subsequently impact album sales and
ultimately profits. Over the past several years the concept of “pay-for-play” has resurfaced. If direct
payments for radio airplay are to be made, then a precise understanding of the dynamic relationship between
sales and airplay is needed. Typically radio airplay and album sales both show an exponential declining
pattern. It is natural to ask whether both series are evolving concurrently–but independently–or is there
some type of dependence? If there is a causal relationship, what is the direction of causality, or is there be a
feedback relationship where both series influence each other? The purpose of this paper is to address these
modeling questions using vector autoregressive models (VARMA), and show how these models can be used
to answer the substantive question of whether the music industry should pay for airplay.