Lack of available data is a major challenge in building a business case for gender equality. Iris Bohnet is a behavioural economist who takes a rigorous data-based approach. Her work as director of the Women and Public Policy Program at the Harvard Kennedy School has built up evidence around unconscious gender biases affecting individuals and organisations. Her new book, What Works, presents numerous inexpensive and easy to implement behavioural ‘nudges’ for removing these biases from individuals and organisations. Women Effect spoke with Iris in London in April 2016.
Do you think there's more we could be doing to get investors to consider gender in their portfolios?
The easy answer - which I don't think will solve your problem - is the need for a business case. There is work going on at banks which builds on the Credit Suisse study [where Iris is on the board] showing that companies with diverse boards and management are correlated with better outcomes. It's not what we would consider scientific proof because to do that we would need to see the comparative achievements of a control group - it might just be amazing people doing an amazing job in a company which generally might be more egalitarian and diverse. The Credit Suisse report is written in such a way that we can't argue causality and it would be key to do so. So the business case which would resonate with investors is not as strong as we wish it were - we’re getting there but more research work needs to be done. In my chapter on diversity in teams I'm trying to unpack what we know.
There's a very good meta analysis of 120 studies looking at the correlation between gender diversity on boards and company performance - there's a weak positive relationship. But there are studies that didn't find anything, there are studies that go the other direction. We haven't had huge success with that business case - we've had some success with talent in Japan. With decreasing fertility and place in the workforce it was clear that if they don't allow women to work and promote women they'll die out at some point and GDP is going to decrease.
There's some data on corporate social responsibility from HBS which I discuss in the book. Companies with more diverse senior management and boards invest more in CSR. The financial business case is slim - is there a market for the moral case? I don't know.
Do you feel like there's a gap in the data?
Absolutely. But also it's an unprovable case because all we'll ever be able to do is establish correlations. We'll never be able to say 'because they have more women they're performing better'. The language would have to be correlating the relationship between A and B - doesn't mean that A is causing B.
Having said this, [it would be worth looking] at the gender diverse funds, or funds that focus on women. Do they attract clients? I don't know what the uptake is [Women Effect estimate that about $500m has been deployed with a gender lens in the last five years]. That's another way to prove the case - what do people think is worth their investment?
Which dataset would you love to see?
I don't want to see a dataset, I want to create one. I want to run an experiment in an organisation and have the organisation experiment with blind performance evaluations. [One group] where they don't share self evaluations and then have a control group where they do, and measure the impact. And many more experiments with companies, helping them be less risk averse.
How could we, and other seemingly female-focused organisations, adapt their language to make men feel included?
Even with the word gender, right, which is such a surprise in many ways - men don't feel included. I've certainly had colleagues suggesting to me, just call it 'Equality by design'. But it's not, i'm not talking about rich and poor, I'm talking about women and men - it has to be called gender. Inclusion seems to be a better word, talent is a very good word. So in Boston we created something that tried to level the playing field and we called the initiative '100% talent'. I have no proof [it works] - that's just how I'd answer your question.