There are many ways to think about people’s “moral behaviour” (from donating to charity to telling the truth) and many different types of competitive scenarios that could be considered (from competing for a job to subtle competition for status in social contexts).

The study of the impact of competition on behaviour, particularly on pro-social and moral aspects, has aroused growing interest in a wide variety of disciplines. Some experts argue that market activity can reduce conflict and violence, while others claim that it can bring out the worst in human beings and encourage unethical practices.

Gert Cornelissen: “Just one study cannot answer broad research questions. Conversely, it is important to perform several studies to find out the conditions under which an effect occurs”

To test whether competition affects moral behaviour, an ambitious collaborative research project was carried out, involving 45 independent research teams from around the world, including a team led by Gert Cornelissen, a professor at the UPF Department of Economics and Business and the Barcelona School of Economics (BSE). Each team applied its own research design (45 different experimental designs) based on a common protocol, in an online experiment.

“In general, the aggregate results of the study show that competitive environments undermine people’s moral behaviour, although the effect is relatively small”, says Gert Cornelissen, who has enjoyed the collaboration of Helena Palumbo, a doctoral student with the University’s Department of Economics and Business.

However, the UPF professor highlights the “meta-analytic” aspect of the study, which has recently been published in the journal PNAS: “Since some individual studies found substantial effects and others found no effects, one of the main messages of this work is that only one study cannot answer broad research questions. Conversely, it is important to perform several studies to find out the conditions under which an effect occurs”.

Along the same lines, the authors assert that by imposing a degree of standardization, such as requiring an experimental economics protocol for all designs (with monetary incentives and no deception), it was thought that this, a priori, would reduce heterogeneity, but this was not so: the variation in effect size estimates across the 45 designs was substantially larger than the variation expected due to sampling errors.

This “design heterogeneity” highlights that the generalizability and informativeness of individual experimental designs are limited: “Our findings provide an argument for moving toward much larger data collections and more team science to improve the informativeness and generalizability of experimental research in the social sciences”, the authors conclude.

More than 18,000 participants in an online experiment

The project invited 102 teams from around the world to present their own research design, 95 of which submitted a proposal: among them, 50 were selected at random to participate in the study, of which 45 delivered the experimental software and, therefore, were included in data collection.

18,123 complete observations (around 400 per design) were collected, via the Prolific survey platform, in January 2022. The mean age of the participants who completed the experiment (from the United Kingdom, USA and South Africa) was 32.5 years, and 55.5% were women.

Reference work: Christoph Huber, Anna Dreber, Jürgen Huber, Felix Holzmelster et al. (May 2023) “Competition and moral behavior: meta-analysis of forty-five crowd-sourced experimental designs”, PNAS

https://doi.org/10.1073/pnas.2215572120