"Open" Disclosure of Innovations, Incentives and Follow-on Reuse: Theory on Processes of Cumulative Innovation and a Field Experiment in Computational Biology

Most of society’s innovation systems – academic science, the patent system, open source, etc. – are “open”in the sense that they are designed to facilitate knowledge disclosure among innovators. An essential difference across innovation systems is whether disclosure is of intermediate progress and solutions or of completed innovations. We theorize and present experimental evidence linking intermediate versus final disclosure to an ‘incentives-versus-reuse’ tradeoff and to a transformation of the innovation search process. We find intermediate disclosure has the advantage of efficiently steering development towards improving existing solution approaches, but also has the effect of limiting experimentation and narrowing technological search. We discuss the comparative advantages of intermediate versus final disclosure policies in fostering innovation.

Policy implications

Intermediate disclosure of innovation progress in research is likely more useful when an effective solution is already known. When there is no existing viable solution, final disclosure rules could be more effective, as they can incentivise a wider variety of ideas being tested.

Additional information

Hypotheses/research question

Prediction 1: Implementing an intermediate rather than final disclosure policy leads to lower incentives but greater follow-on reuse. Prediction 2: Implementing an intermediate rather than final disclosure policy leads innovators to tend to converge towards successful solution approaches and to engage in a lesser degree of independent experimentation.

Study design

Factorial

Subjects were recruited from the TopCoder platform’s existing membership of software and algorithm programmers. The posting indicated that the challenge would feature an algorithmic optimisation solution related to genomics, the solution to which was sought by Harvard Medical School (from which the problem had been sourced), that the total prize pool would be $6,000.00 and that the exercise was also being used for research purposes. The call for participation did not describe what particular problem would be solved. 733 TopCoder members signed up for the experiment. 44% were professionals, the remainder were students. Participants from India (20%), the United States (16%), China (9%) and Russia (9%) accounted for more than half of a pool that represented 69 countries.

Cost benefit ratio

Not available.

Rounds of data collection

3: Baseline + continuous digitally automated data collection of dynamics during problem solving and development phase + final questionnaire

Baseline data collection and method

TopCoder members voluntarily signed up to participate in the contest. TopCoder collects demographic information on members as well as their historical quality of performance and volume of submissions in TopCoder's other contests.

Data collection method and data collected

Live, online data collected on user decision to participate, submissions, users' examinations of other disclosed submissions, and quality performance. A final survey was administered after all coding activity was completed and before final results were made public, asking to report the number of hours worked over the two-week period.

Unit of analysis

Individual

Stratification

Reference

Boudreau, K., & Lakhani, K., 2015. ''Open' Disclosure of Innovations, Incentives and Follow-on Reuse: Theory on Processes of Cumulative Innovation and a Field Experiment in Computational Biology'. Research Policy vol. 44(1), pages 4–19.

"Open" Disclosure of Innovations, Incentives and Follow-on Reuse: Theory on Processes of Cumulative Innovation and a Field Experiment in Computational Biology

Key information

Additional information

Key facts

Update trial