I investigate incentives and the order of play in a collaborative experimentation model with heterogeneous agents. Consider two teams of differing abilities in a research firm, who sequentially attempt to develop an innovation. I ask which team should try first and how incentives to innovate depend on abilities. Two opposing forces are at play. Since feasibility is unknown, a failure from the first team discourages those who come after. This discouragement effect suggests that a high-ability team should work later. However, the first team can always free-ride on a later team and this free-riding incentive implies that a high-ability team should go first. Given these subtleties, I propose a simple dynamic model of innovation and demonstrate that high efforts are achieved when both teams are of medium-ability. I prove that when teams’ abilities are similar but extreme, it is optimal to have the first attempt made by a relatively lower ability team. However, when abilities are both intermediate, the higher ability team should go first. If abilities are far apart, sequencing has no effect sans discounting. I further ask how a principal should structure teams, collaboration, and prize incentives to best elicit effort.

You can sign up for a 30-minute meeting with the speaker on Friday 11 December using this spreadsheet: docs.google.com/spreadsheets/d/1Vrk4zZ1ZHgoNeUiLxU2_zbAamcKhgNqxTGky1e51oYU/edit#gid=0. You must sign up before noon on Monday 7 December.