Best practices for reproducibility, research assessment reforms, and implications for experimental economists

• Scientists need to adopt best practices for enhancing reproducibility. • This tendency will only increase with the new developments in researcher assessment. • A novel taxonomy shows considerable heterogeneity in the amenability of different domains to employing practices for enhancing reproducibility. • This raises issues of fairness, exemplified in the recently expressed concerns by experimental economists. • Adjustments in research assessment are needed to account for this heterogeneity. • In the long run, practices for improving reproducibility need to expand further in non experimental domains. Scientists are under pressure to adhere to best practices for enhancing reproducibility, such as preregistration and data sharing. This tendency will certainly increase with the unfolding reforms in researcher assessment, and it brings new challenges. Heterogeneity in the amenability of different domains to reproducibility-enhancing practices raises an issue of possible inequity: will different scientific domains bear disparate adjustment costs? Is this justified and efficient? To illustrate the problem, we consider recent concerns expressed by experimental economists, namely that they are unfairly burdened relative to other economics domains. Our analysis indicates that such fairness concerns may have merit, but only insofar as research assessment does not fully internalize the costs of adjusting to new practices.