We are excited to feature Prof. Iris Bohnet, a valued member of the EDGE Certified Foundation Academic and Scientific Council. In this interview, Prof. Iris Bohnet shares insights from her recently published paper on behavioral design and diversity training, as well as thought-provoking perspectives on workplace fairness from her upcoming book, Make Work Fair: Data-Driven Design for Real Results. The interview concludes with an inspiring call to action that challenges us to rethink how we foster equity in our environments.
We hope you enjoy this interview as much as we did.
In your work, you focus on behavioral design and eliminating gender bias in the workplace. What do you think are the main opportunities that your recent research has unveiled in this area?
In a recent paper that has just been published in Science, we apply insights from behavioral design to diversity training. Such trainings are prevalent but unfortunately, the evidence does not suggest that they are able to change people’s behaviors. At best, they can raise short-term awareness; at worst, they can lead to backlash. So, many, including I, had given up on trainings arguing that to make change happen, we need structural interventions. That is still true—but given that trainings are quite possibly the most popular tool in use, my coauthors and I set out to see whether we could make them more impactful. There are various behavioral design elements in the training that we collaborated on with a multinational telecommunications and engineering company but the two that might stand out most are the timeliness and the focus of the training. Most diversity trainings are rather generic and take place at a random moment in time. In our randomized controlled trial, we exposed hiring managers to a video message about the value of diverse teams shortly before they shortlisted job candidates. The message was timely and tailored to the decision at hand. For example, it reminded managers of the importance of skills and skills complementarities in teams encouraging them to not just look for the talent they were used to. It worked. In this multi-national company active in over 100 countries, the impacts of the training were particularly strong for women who applied from outside of the country where the job was posted when compared to a control group that did not receive any training. They may well have been overlooked beforehand given that humans tend to prefer members from their “in-group”–and hiring managers overwhelmingly were men and nationals. The hope is that our findings will stimulate a new approach to diversity training. One that starts with the desired behavior change and then, designs the training accordingly. And one that uses more rigor to evaluate what works and what doesn’t.
For organizations just starting out in measuring gender equality in the workplace, where should they start?
That’s a good question. I don’t know there is any research showing that, for example, measuring pay gaps is more or less important than focusing on representation. Surely, equal pay for work of equal value is crucial and so is leveling the playing field to make sure people have equal opportunities to succeed. In terms of representation, I would encourage organizations to look at the data in a relatively granular way, by occupation, by seniority, by time status, and if the sample size allows it, intersectionally. Time status is particularly important in countries where how much people work is very gendered such as, e.g., the German-speaking countries.
We’re excited to hear about your new book, “Make Work Fair: Data-Driven Design for Real Results”, co-authored with Siri Chilazi, which explores the concept of fairness in the workplace and presents innovative, data-driven solutions for professionals. For our audience, who are passionate about and invested in fostering equitable work environments, could you share an example of one of the most impactful solutions outlined in the book, and explain why you believe it is pivotal for transforming fairness into a reality in the workplace?
It depends on your final goal. If the goal is to increase women’s workforce participation rates, e.g., to reduce the gender gap in poverty after retirement, then making work more attractive might be your best strategy. That includes fair pay and fair working conditions but also fair taxation. In a recent OECD-report, seven countries reported gender bias in their tax systems. Switzerland was one of them. As long as the personal income tax system is based on the taxation of the household rather than the individual, there will be a disincentive to work for the “second income earner.” With a progressive income tax schedule, these people are confronted with a higher marginal tax rate. And given that women are more likely to be the “second earner,” they are more likely to be hurt by unfair tax systems. Fair pay is self-explanatory but there is lots to be said about fair working conditions. In fact, this is what our book focuses on.
Something I had only partially understood before writing it is the importance of control over one’s time. It turns out, often people have little control over their work schedule, which can be changed at the last minute. Think hospitality, care, and other service jobs where shift work is common. Predictive scheduling is an increasingly important topic in the United States where employees don’t tend to be protected from unexpected schedule changes. Not surprisingly, such lack of predictability has huge impacts on people’s lives, and it is more likely to affect people of color and women.
More generally, the question of when, where and how we work will remain important going forward and will affect people differentially depending on their responsibilities outside of formal work. This also plays an important role in who is even invited to a job interview. A recent FT-review of our book focused on an experiment by two of my former fellows, Oliver Hauser and Ariella Kristall, and their colleagues that I also really like. It turns out if job seekers simply list the number of years worked instead of the actual dates on their CVs, employers are more likely to invite them to a job interview. Why? Because this masks career gaps, a data point often used by human and algorithmic screeners alike, without much evidence that gaps are predictive of future performance.
Finally, I remain concerned about occupational segregation and am encouraged by the evidence of programs enabling more men to become teachers and nurses and more women to become engineers and mechanics. Role models, it turns out, can play an important role here, and as such, this might also be a call to action for your readers: make sure the portraits on your walls represent the people you serve and become a role model yourself, perhaps, visiting a school showcasing that men can care and women can lead.
A big thank you to Prof. Iris Bohnet for sharing these insights with us.
This interview is a part of our series, where we speak with members of our Global Advisory and Academic and Scientific Councils to hear their insights on diversity, equity, and inclusion. Each interview highlights their unique expertise and the important work they’re leading. Visit our reading corner for more interesting insights from our Global Advisory and Academic and Scientific Council members.
