Peter DiCola

Sequential Musical Creation and Sample Licensing

All musical creation builds on previous works. But using fragments of existing musical works in a new work can often constitute copyright infringement. Copyright law, in cases like Bridgeport Music v. Dimension Films (6th Cir. 2005), has recently increased its restrictions on musicians who wish to engage in sampling, defined as the practice of using other creators’ sound recordings to create new music. The paper describes a model of copyright holders’ and samplers’ incentives to create in light of the need to negotiate licenses for sample-based works to avoid violating copyright law. Even in the absence of traditional transaction costs or royalty stacking, a distinct kind of inefficiency emerges. Green and Scotchmer (1995) have shown that, in the patent context, bargaining may not divide the profit from the sample-based derivative work between upstream and downstream creators in a way that provides both groups with sufficient incentives to create. This paper builds on and extends Green and Scotchmer’s theory by showing that innovation occurring in sequence presents a reciprocal problem. Both upstream and downstream creators have incentive constraints; pure theory cannot say which incentive constraint is less likely to be satisfied. This problem is exacerbated in the sample-licensing context because ex ante agreements are not usually possible. An optimal system for regulating sequential creation would account for the incentives of both upstream and downstream creators, to the benefit of both groups and the public. Congress and the courts have probably failed to achieve this balance, since the economic analysis of courts (especially the Sixth Circuit) has focused mainly on upstream creators’ incentives.