Should the winner be always the same? Group bonuses to enhance competition and rotation of prizes
How to cite
Ezquerra, Lara; Ortiz, Jose M.; Olavide, Natalia Jimenez Universidad Pablo de. 2026. "Should the winner be always the same? Group bonuses to enhance competition and rotation of prizes." AEA RCT Registry. AEARCTR-0016921. https://doi.org/10.1257/rct.16921-1.0
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
This experiment investigates how different reward structures influence outcomes in competitive environments where the winner is always the same person. These type of incentives may be appropriate in groups where rankings are stable or where the abilities or the starting point of subjects are more important than effort.
Workers have different productivity rates. These rates are fixed throughout the experiment, which consists of 15 periods. In all periods we have the same groups of 3 participants with one having a high productivity rate and the other two sharing the low productivity rate. With this we ensure that the most productive worker is always the one with high productivity.
The experiment applies four treatments that manipulate how prizes are awarded in the competition. In Treatment 1, a “winner-takes-all” system is applied, with rankings determining a single winner each round. Treatment 2 uses the same rule, but with the restriction that the same participant cannot win consecutively, forcing changes in ranking outcomes. Treatment 3 introduces multiple prizes of unequal size, not directly proportional to performance, but maintains the possibility of repeated winners. Treatment 4 combines unequal multiple prizes with the restriction that winners must change between rounds.
Together, these treatments allow comparison between highly concentrated rewards (winner-takes-all) and more distributed but uneven incentives (multiple unequal prizes), while also testing the effects of enforced rotation in winning. This approach sheds light on how prize structures affect effort, strategy, and outcomes in creative or competitive labor markets.
Intervention
The intervention consists of four distinct incentive treatments that vary the way rewards are distributed among participants competing in a repeated production task. The underlying task and group composition remain constant across treatments; only the reward structure changes.
In all conditions, participants are assigned to fixed groups of three—one high-productivity worker and two low-productivity workers—and perform the same individual task during 15 periods. Their individual scores are used to rank them within their group each round. Based on this ranking, monetary prizes are distributed according to the treatment in place.
The intervention manipulates two dimensions of the incentive environment:
Prize concentration – whether all rewards go to a single winner or are distributed among multiple participants.
Winner rotation – whether the same participant is allowed to win consecutively or whether winners must change between rounds.
Specifically:
T1: Winner-takes-all, same ranking (single winner, no rotation).
T2: Winner-takes-all, change ranking (single winner, but the same person cannot win twice in a row).
T3: Multiple unequal prizes, same ranking (rewards for multiple ranks, not proportional to performance).
T4: Multiple unequal prizes, change ranking (unequal rewards and enforced winner rotation).
By systematically varying these two dimensions, the intervention isolates the effects of reward concentration and forced turnover on effort, motivation, and adaptation in competitive settings where productivity differences are fixed and known to exist.
Experimental design
The experiment will be conducted in a controlled laboratory setting using computerized individual decision tasks. Participants will be randomly assigned to fixed groups of three that remain constant throughout all 15 periods. Within each group, one participant will be assigned a high-productivity role, while the other two will be low-productivity workers. Productivity rates will be exogenously determined and remain constant across all periods to ensure that performance differences reflect the underlying productivity parameters rather than random variation or effort shocks. Participants will be informed of their own productivity level but not of the productivity levels of the other group members, although they will know that heterogeneity exists within their group.
Each period consists of a production task where participants perform a simple, measurable activity (count the number of 0 in a grid of numbers) that determines their score for that round. As an alternative instead of being productive they can select a leisure activity (e.g. read jokes or see reels) that does not contribute to production in any case. Based on individual performance, participants are ranked within their group and prizes are distributed according to the treatment condition.
The four treatments manipulate the structure and rotation of rewards:
T1: Winner-takes-all, same ranking (the top performer always receives the entire prize).
T2: Winner-takes-all, changing ranking (the same person cannot win consecutively).
T3: Multiple unequal prizes, same ranking (two or three prizes of different values, not proportional to performance).
T4: Multiple unequal prizes, changing ranking (unequal rewards with enforced rotation of winners).
This design allows causal identification of how different incentive schemes affect motivation and performance when productivity differences are known but immutable.
Design details
Cited by
No citation data available for this registration yet.