Iris Bohnet’s What Works: Gender Equality by Design is probably the most pragmatic books on changing organizational culture that I’ve ever read. Over the course of the book, Bohnet offers 36 research-grounded design suggestions for achieving gender equality in the workplace. The use of the word “design” here is intentional: Bohnet is a behavioral economist, and the book offers much in the way of behavioral design from that perspective. Bohnet presents a careful review of current research — occasionally, this can be a bit dizzying as she walks through conflicting studies, but in the end she consistently delivers evidenced-based recommendations for how to achieve gender equality in the workplace and beyond.
This dizzying coverage of so much research and so many ideas can be a bit overwhelming. I had an incredibly difficult time narrowing down my selection of quotes for this summary. The behavioral design approach means that the solutions Bohnet offers in the book are incredibly practical. Given my own focus on hiring and interview processes over the last few years, I loved the checklist for comparative, structured interviews, and it has evolved my thinking on interviewing management candidates. The book also offers thought provoking insights into opportunities to improve diversity when hiring like this one:
The police force in the United Kingdom substantially increased the talent pool by adopting a friendlier tone and asking people a simple question before they took the entry exam: Why do you want to join the police and why does it matter to your community? Hearing a different tone and being nudged to think about what motivated them increased the pass rate of minority group applicants by 50 percent. Making a small change that had a big impact required no more than the insights and experimentation of a few creative thinkers at the UK Behavioral Insights Team.
Bohnet reviews research into all aspects of gender differences without shying away from the fact that there are measurable differences between the genders in different societies. For example, she digs into gender differences in competitiveness, pointing out that, “For MBA graduates at Booth School of Business at the University of Chicago, interpersonal differences in willingness to compete translate into significant differences in earnings, explain part of the gender gap in earnings, and have long-lasting effects: competitive graduates are more likely to work in higher-paying industries nine years after graduation.” When discussing these ideas, Bohnet covers both the long term need to explore the ways society might create or amplify those differences, while suggesting how companies might address it in the meantime.
Bohnet really digs into every aspect impacting gender equality in the workplace and society. While I was somewhat aware that female leaders can sometimes be the cause of discrimination against women, I was not familiar with much of the research this book presents. Bohnet gets right into interesting stories like this and goes on to address how we might correct for it.
In Spain, academic promotions from assistant professor to associate professor and then to full professor are determined by randomly created evaluation committees, leading to random variation in the gender composition of the committees. It turns out that female associate professors evaluating assistant professors for promotion were more likely to be in favor of promoting male junior colleagues. But this effect was only evident when evaluator and candidate were at the same institution. It appears as if the evaluator again feared same-sex competition, perhaps assuming some implicit gender quota. In fact, if a male evaluator was replaced by a female evaluator of the same institution, this decreased the likelihood that a junior female candidate was promoted by 38 percent. It did not affect female candidates from other institutions. Nor did the researchers observe any of this behavior for promotions to full professor. When full professors evaluate associate professors to join their ranks, they need not fear same-sex competition. Full professor is the highest rank possible in academia, and at that point, evaluators, looking for friends and people like themselves, exhibited an in-group preference, favoring candidates of the same sex and academic network.
The book also digs into other areas of diversity that, while not directly related to addressing the gender gap, struck me as fascinating. Given the huge trend anti-PC rhetoric in the U.S. at the moment, this research-driven call in favor of political correctness for creative endeavors stood out:
In addition, and somewhat counter-intuitively, groups might want to constrain themselves from speaking freely. The study “Creativity from Constraint” presents experimental evidence on how imposing a norm of political correctness (PC) that specifies how men and women should interact with each other enhances creativity in mixed-sex groups. The PC norm increased the exchange of ideas by clarifying the rules of engagement and providing assurance to those, predominantly women, for whom speaking up was associated with counterstereotypical behavior.
As organizations work to hire diverse employees into largely homogeneous teams, they face challenges. I’ve heard women describe having a single woman in an otherwise-male software team as an anti-pattern. Bohnet offers pointers into when diversity matters the most, and when homogeneity might actually be a good thing. This will take me a while to absorb and think through.
If a task involves coordination, say the provision of a public good like clean water or better health care, homogeneous groups can be helpful. All-women teams, for example, outperformed mixed and all-male teams in Friend or Foe because they correctly believed that women would be more likely to cooperate than men, leading to a virtuous cycle.
If a task involves individual problem-solving, say test-taking, be aware of peer effects. Diversity might produce spillovers (or, more formally, “externalities”) affecting, for example, how students perform in a class. If one group is more likely to work hard or disrupt less, as has been found to be true of girls, having them over-represented can help others, in this case boys, perform better.
Like many, I pay careful attention to diversity reports coming out of major tech companies each year. I often find them discouraging, and after finishing the chapters on behavioral modifications that might work, I find myself wondering if they’re actually doing a disservice by normalizing the current levels of diversity.
Making public and visible how well a company or country does in terms of gender equality compared to others might also promote convergence on a new norm. Indeed, a number of organizations now provide social comparisons or explicit rankings based on gender equality. In 2006, the World Economic Forum (WEF) launched its annual Global Gender Gap Report measuring the existing gender gaps in four categories: economic participation and opportunity (pay, participation, and leadership), political empowerment (representation), education (access), and health and survival (life expectancy and sex ratio at birth). Since then, the WEF has annually published a report measuring how the gaps are changing over time. It ranks countries on their overall performance, as well as on how well they do in all four categories. Over nine years, the Nordic countries have been leading the pack with Iceland having closed the overall gap by 87.3 percent (with 100 percent indicating gender equality) as of 2013. Generally, Middle Eastern and North African countries have fared worst, with Yemen having closed only 51.2 percent of the overall gap.
At the least, the book has me convinced that collecting the numbers that companies release and putting them in ranked order with a visual and easy to understand score might be a helpful step in the right direction.
Improvements in a college’s rank on US News and World Report Best College Rankings immediately translate into a larger number of applications. Interestingly enough, however, the authors only find this effect when the magazine presents the colleges ordered by rank. When the colleges are listed alphabetically, with their rank included in the body copy describing the institution, no effect on applications could be detected. Only easily understandable and highly visible comparisons mattered, something gender equality designers hoping to influence behavior need to keep in mind.
Beyond the huge volume of research surveyed in this book, Bohnet also provides tons of other resources for people and organizations working towards gender equality. It’s a tremendous resource of practical and thought-provoking ideas. I especially appreciated the depth of coverage. In fact that very depth occasionally had me worried that this book could become a resource for the James Damores of the world to cherry pick isolated arguments out of context. But I suppose it would be a bit too much to expect this book to offer solutions to that. Overall, I can’t recommend this book highly enough. It’s sure to be something I refer back to often, and if you’re interested in understanding and correcting gender inequality, you should read this book.