A Theoretical and Empirical Analysis of the Impact of the Digital Age on the Music Industry

We present an in-depth analysis of the music industry and use our findings to judge the practicalassumptions and design of an original theoretical model. The model is in three stages, where, in aHotelling-type framework, the last agents to act are consumers who choose between copying, purchasing,or staying out of the market for music. Prior to the last stage, the record label chooses its profitmaximizing price and, in the first stage, we incorporate the artist-label bargaining agreement into atheoretical framework using the Nash cooperative bargaining solution. The current structure of the musicindustry is a combination of the oligopoly and monopolistic competition models, consisting of five majorlabels and many independents. Despite major labels’ advantage in large-scale distribution, we argue thatdigital downloading has the potential to radically alter the current industry structure, and that artists wouldbe unable to sell their music in such an environment without enforceable copyrights. Our model assumesthat the most important determinants of CD and copy demand are consumers’ tastes and transaction costsof copying, CD prices, and the substitutability between CDs and copies. We hypothesize that Internetfile-sharing has been undertaken by both consumers who were previously not in the market, and by thosewho decided to copy rather than buy. In regard to firm strategy, the model suggests that labels couldincrease the sales of CDs by trying to increase consumers’ taste for music, perhaps by reducing the priceof CDs. Our model also predicts a positive relationship between artists’ optimal share of album sales andtheir bargaining power, as well as a negative relationship between artists’ optimal share and their riskaversion. Since lowering the reliance on labels for distribution would increase artists’ bargaining power,our model predicts that artists’ share of profits should increase as legitimate digital distribution gainsprominence. We also provide empirical testing of our hypothesis that some music file-sharing has beendone by consumers frequently not in the market. After examining consumers’ expenditures and aggregateindustry sales, we are unable to reject our hypothesis