Cumulative Innovation, Sampling and the Hold-Up Problem

Rufus Pollock

[Cambridge University]

Examples

Music

Software

Documentary Films

[The] 90-minute documentary [Wanderlust] ... was also a window into the frustrations of making a clip-intensive film dependent on copyright clearance, which has become hugely expensive in the past decade. Initial quotations for the necessary sequences came to more than $450,000, which would have raised by half the cost of the IFC film. ... "Paramount wanted $20,000 for 119 seconds of 'Paper Moon,' " Ms. Sams said. "The studios are so afraid of exploitation that they set boundaries no one will cross. Even after the prices were cut, we were $150,000 in the hole." (from the NY Times No Free Samples for Documentaries: Seeking Film Clips With the Fair-Use Doctrine)

Framework

Structure of this talk

  1. Simple (standard) model of 2 Stage Cumulative Innovation
  2. Sampling model (sampling by second-stage firms)
  3. Conclusions

Model 1:
Simple 2-Stage Cumulative Innovation

Solving

Plot of optimal policy as function of q and p

Prefer IP to No IP if:

$$\textrm{%tage extra 1st stage innovation} \geq \frac{\textrm{Expected Holdup Cost}}{\textrm{Net Surplus From 2nd Stage} + \textrm{Expected Holdup Cost}}$$


NB: net surplus gets smaller the more of license revenue used to pay for 1st stage innovation costs

Model 2:
Cumulative Innovation with Sampling

cumulative_innovation_discrete_choice_transaction_costs_game.png

Results

IP preferable to No IP if and only if:

$$\textrm{%tage Extra 1st Stage)} \geq \frac{ \textrm{Benefit of Higher Sampling w/o IP} + \textrm{Holdup Cost}} { \textrm{Net Surplus 2nd Stage} + \textrm{Benefit of Higher Sampling w/o IP} + \textrm{Holdup Cost}}$$

Numerator increases as:

Conclusion

Basic Model

Sampling Model

Take Away

Thank you!

Comments to: rufus.pollock@thefactz.org