Following her committee’s latest recommendations to IEEE, Marconi Society Board Member and Stanford Engineering professor Andrea Goldsmith shared her thoughts about changing the diversity and inclusion equation in engineering and technology.

Q: Is there a commonly accepted set of metrics that we can use to set goals and understand our progress around diversity and inclusion?

A: Commonly accepted metrics and goals are one of the critical things that we need to benchmark our current state and to measure our movement forward. That is a first step that I am advocating for with IEEE.

There is certainly data, but it is not uniform. For example, through the IEEE’s Women in Tech Survey we know that nearly 60% of women did not think men and women working in technology fields are treated equally and over 70% reported the negative workplace experiences of having questions that should be addressed to them being addressed to a male colleague and of having a male colleague make disparaging comments about them.

McKinsey & Co estimates that women comprise 26% of the computing workforce while “women of color hold only 4% of technical roles at tech companies and are almost completely absent at the senior leadership level.” Many tech companies publish their diversity data, as well, but we must ask ourselves whether we have the right, actionable data to work from.

That right, actionable data will vary by organization and goal. At the IEEE, we are focused on diversity data across gender, geographic region, work sector, and career stage since that membership data is currently available. Other diversity dimensions are important as well, but we don’t currently have mechanisms to collect data around them. To meet our goals, we are currently focused on determining the percentage of women and people from different geographic regions across our awards, leadership positions, plenary speakers and authors. We need to define the metrics, collect them and then transparently share them.

Some of the data we need is easy to collect because it is already in the IEEE database. However, different data sits in different places. My committee is trying to get most, if not all, of the data we need so that we can publish an annual IEEE diversity report that shares the data, discusses how it’s changing and describes the best practices that will we use to get all groups represented and experiencing the full benefits of IEEE membership.

Q: What will success look like short term and long term?

A: In the short term, my committee is completely revising the IEEE ethics violation processes as well as merging the existing codes of conduct and ethics. It may seem surprising to have ethics combined with diversity and inclusion, but the three areas are inextricably linked. At the IEEE, we have a powerful code of conduct and code of ethics, but few members are aware of these codes nor of their responsibility to uphold them. We also have no effective mechanism to hold people accountable for violating these codes, particularly in areas such as harassment. In the short term we plan to raise awareness about the revised and streamlined code, how we will hold ourselves responsible to honor it, and what the IEEE will do about egregious behavior that violates it.

Diversity and inclusion itself is a marathon, not a sprint.

We have a lot of awareness building to do. Some people do not think that there is a problem at all. Others believe that there is a pipeline issue, but that does not entirely explain the low numbers. According to an HBR study, women make up 20% of engineering graduates but nearly 40% of those women who earn an engineering degree either leave or never enter the profession.

We need actionable measures to improve diversity and inclusion. These include:

• A report with benchmark numbers so that people can see the lack of diversity and the missed opportunity that accompanies that lack
• A public website with resources and best practices for organizations, universities and companies, information on how people from underrepresented groups can find mentors and sponsors, and the questions to ask when looking for a job or research group
• Making the face of the IEEE more diverse and highlighting the accomplishments of its diverse membership
• Most importantly, talking about diversity and inclusion at the global and national levels and providing tools and resources for these conversations at chapter levels.

Q: In your experience, what has actually worked to improve diversity and inclusion?

A: What works is highlighting the problem through diversity data, and explaining both diversity and inclusion so that people can understand that they are different things. Often there is a sense that it’s not so bad – people think it’s a pipeline issue and we’re trying but we really can’t do any better because of the supply. That’s just not correct and our lack of diversity is partly due to a lack of inclusion such that diverse people have a bad experience in our field and leave as a result.

What also works is getting women and underrepresented groups into leadership positions, honoring them with high-level awards and featuring them as authors and speakers. This does not mean that we change the criteria for these achievements. It means that we give all IEEE members a fair shot at the opportunities.

Best practices include discussing diversity and inclusion at each relevant decision point in awards, honors, and leadership positions, and ensuring diversity on selection committees.

Best practices for improvement also include having and sharing data. The objective of the data is to inform and to show where we can improve, not to punish or reprimand. We can argue about why the data is what it is and, if it’s not where we want it to be, we can figure out how we can move it in the right direction.

Corporations and universities have been collecting diversity data and working on improvement strategies for a while and it’s time for professional organizations like the IEEE to do the same.

One of the challenges for the IEEE is that there is not yet critical mass for some underrepresented groups. How do you define underrepresented in China? What about disabled, LGBTQ and other groups that the IEEE does not track at all yet? We need to look across the entirety of IEEE to find a critical mass of different groups so that we can be truly inclusive.

We have already seen places where we can have a quick and dramatic impact. The IEEE Awards are a great example.  A 2016 document on implicit bias led to unprecedented diversity in this year’s awards where we were proud to honor five women and nine people from outside North America with IEEE medals and high honors. We had the most diverse group of winners ever. We will build from here.