BIG IDEAs: Using Randomised Controlled Trials to Reduce Bias in the Workplace

Discrimination and social inequalities continue to pervade the workplace. The UK's Gender Pay Gap website, Equal Pay Day and the global #MeToo movement are but a few examples that highlight the growing awareness of deeply-embedded structural inequalities against women at work. These inequalities come about in part because of conscious and unconscious biases held against women - but it isn't just women who suffer. Bias is rife in the workplace towards other disadvantaged groups such as ethnic and racial minorities, people from lower socioeconomic backgrounds and the LGBTQ+ community. Although recognition of these problems is growing, businesses struggle to find solutions to level the playing field. What can we do to help reduce inequality and discrimination at work? Through the UKRI Future Leaders Fellowship (FLF) hosted at the University of Exeter, my goal is to use causal interventions to reduce bias workplace to help create more equality at work. This project leverages the latest insights from across economics, psychology and management to design interventions that will be tested in large-scale randomised controlled trials (RCTs) within several UK organisations. These interventions will lead to a better working knowledge of how we create more fairness in the workplace, disseminated through the public-facing Centre for Behavioural Insights: Gender, Inclusivity, Diversity, Equality and Access ("BIG IDEAs") at the University of Exeter. Gender bias is now a well-recognised phenomenon and many organisations try to tackle this problem through "unconscious bias" training. However, there is little evidence that this training works towards the advancement of women and, if anything, it can backfire, giving those completing it the false impression that they are "cured" of unconscious bias (Dobbin & Kalev 2016; Atewologun et al. 2018). The problem in fact goes much deeper: bias is often deeply ingrained in our minds and affected, even amplified, by the institutions and social environment in which we live (Thaler & Sunstein 2008). A combination of conscious and unconscious bias have likely negatively affected the career developments of women (Bohnet 2016; Sheltzer & Smith 2014). But these biases extend also to other disadvantaged groups: workers from racial or ethnic minorities often face discrimination in the workplace, too; and so do individuals from a lower socioeconomic background (Bertrand & Mullainathan 2004; Stephens et al. 2019). My research agenda takes a comprehensive view of bias at the workplace, studying and aiming to reduce bias against these disadvantaged groups. This project is the first to test novel interventions (possible solutions) through multiple, longitudinal RCTs inside organisations. It goes beyond the study of surveys which have looked at correlations of bias and work outcomes; instead, I propose to actually introduce and causally test interventions to reduce bias to create a more equal workplace on the ground, observing the actual hiring, promotion, and retention outcomes of disadvantaged groups. This kind of research thrives through collaboration and co-creation with organisations. Some of my partner organisations are among the largest UK employers, including Unilever, a pioneer in addressing gender inequality. Others are start-ups and consultancies, including the Behavioural Insights Team, GapSquare and MoreThanNow, who work with influential clients who want to overcome bias against disadvantaged groups. In sum, my research aims to test co-created interventions in the workplace to reduce bias against disadvantaged groups and contribute to a more equal society. References: Atewologun et al (2018), Equality Human Rights Commission. Bertrand & Mullainathan (2004), AER 94(4), 991-1013. Bohnet (2016), What works, HUP. Dobbin & Kalev (2016), HBR. Thaler & Sunstein (2008), Nudge, YUP. Sheltzer & Smith (2014), PNAS 111(28) 10107-12. Stephens et al (2019), Curr Dir Psych Sci 28(1), 67-73.