Success Breeds Success: Evidence from Field Experiments

• In Event: 272 - Regular Session. Rational Choice

Abstract

Successes tend to accrue to people who have been successful before. Two mechanisms can explain the relationship between past and present success. Pre-existing differentiation along fitness dimensions may repeatedly direct successes toward certain subpopulations. Alternatively, success may breed success through positive feedback. Observational studies have not been able to evidence the operation of a success-breeds-success mechanism because correlations between successes in longitudinal records may also be produced by unobserved fitness advantages. To overcome this problem of empirical confounding, we conducted randomized experiments by intervening in four live social systems across the domains of funding, status, social support, and reputation. In each system we consistently found that bestowed success upon arbitrarily selected recipients produced significant improvements in subsequent rates of success as compared to the control group of non-recipients. Experiments in which money was donated to random projects reveal that the average benefit of the treatment exceeded its cost, suggesting that under certain conditions investments may pay for themselves through the subsequent contributions they trigger, even when made in disregard of target quality.

Authors

• Arnout van de Rijt, State University of New York-Stony Brook

• Soong Moon Kang, University College-London

• Michael Restivo, State University of New York-Stony Brook

• Akshay Patil, State University of New York-Stony Brook