Difficult questions RE race & ethnicity data: Let's not be France

A few weeks ago, I sat down to write a follow-up to the first edition of this newsletter–in which I had written about the lack of diversity in the data industry, and what I see as a? reigning attitude of complacency towards that lack of diversity.?

First, here’s a quick recap of that article:?

• I had called for data professionals to start paying attention to a new metric, which I called the #wholookslikeme score.?
• This would consist of measuring and reporting on how many of the people with whom we interact on professional social media platforms look like us.
• Why focus on professional social media platforms? Because these virtual spaces–like Linkedin, Twitter, and Slack– can be as powerful and influential as corporate boardrooms and conference speaking panels. And yet, if you look at the conversations and comments on these platforms, there’s a glaring lack of diversity.?

In this follow-up article, I had planned to start this diversity-metrics-measuring movement by looking at my own metrics. I had designed a simple initial methodology:

• Choose the professional social media platform on which I’m most active (in my case, Linkedin);
• Choose a random day and time (I chose 7 P.M. on Wednesdays);
• Look at the fifty most recent interactions on my profile, and list whether or not those people look like me,? in the categories of gender and race.

Sounds tedious (manually looking at each interaction), but it’s a straightforward process that I should haven been able to complete in 15-30 minutes.

Unfortunately, I got stuck.?

This is what happened: I got deeply uncomfortable.

I realized that I DID NOT LIKE categorizing my interactions in this way.

It felt reductive. I would stop and look at the profile of someone who commented on my page, and try to identify their ethnicity. Do I judge according to someone’s name? According to their skin color? It felt creepy and invasive to start examining profiles in this manner.?

And gender, much like ethnicity, is not always obvious on someone’s Linkedin page. I love living in a world in which no one has to clearly define any of these characteristics about themselves, particularly in a professional setting. And yet–here I was–trying to force people into those categories, for my purposes.?

So what did I do??

I stopped measuring.?

Before I got any further with my first attempt at my ambitious #wholookslikeme metric, I had given up.

But I knew I couldn’t stop there. I didn’t want to just… abandon… an initiative I believe is urgently needed.?

I sat on the issue for a few WEEKS.?

Eventually, an answer came to me.?

In the form of data.

More specifically, an answer came to me in the form of a story about data.?

I suddenly remembered another time I’d witnessed discomfort around collecting data like this.

It was during my time in France (where I lived from 2003-2011). The French don’t collect data on ethnicity. Their reasoning is, essentially, that they are “beyond” race; they like to think that anyone living in France benefits from a “universalist” perspective that views and treats all human beings as equals.?

You won’t be shocked to hear that, from my perspective, the French politicians and people who believe that are dangerously delusional.?

During my time studying and working in France, I witnessed overt racism everywhere from the dinner table to the lecture hall to the office.

My understanding of the racism I witnessed is limited and biased by the fact that I’m a white woman. But the thought that came to me, over and over again, was: “This feels like stepping back in time 30-40 years.” Like, maybe this is how Americans acted before affirmative action policies were widely implemented? And before the advent of political correctness? It seemed that progress that had taken place in the U.S. simply… skipped over France.

I dug into it. The historical and political differences between the U.S. and France that would lead to differing levels and types of racism are multiple and complex, of course. But there’s one glaring difference I kept coming back to: France’s ongoing refusal to collect race and ethnicity data.

In the U.S., the collection of data on race and ethnicity is something we grew up with–when our parents first sign us up for school, when we apply for jobs and financial aid, and when we fill out our census forms. In my personal and professional experiences, I haven’t come across anyone loudly calling for an end to this data collection. I assume many people of many different backgrounds opt to “not disclose,” for a diversity of reasons. But–among all the topics that have become flashpoints in the polarizing culture wars over the past 30 years–the question “should we stop collecting race and ethnicity data” has not made the list. It’s like, one of the few things… we all just agree on?

Meanwhile over in France, it’s almost the opposite. Most people I spoke to about the topic firmly and proudly defended the French perspective. They would argue that it’s racist to ask people to have to define themselves according to race, a category many of them say simply “does not exist.”?

I say the French attitude is “almost” the opposite of the U.S. perspective because there is a growing number of French voices calling for the collection of this type of data.?

This includes former government spokesperson Sideth Ndiaye, who in 2020 formally called for a nationwide constructive debate around “the collection of statistics on ethnicity” (this was a bold move, as her own very powerful boss-President Emmanuel Macron–immediately expressed his disagreement with her).?

Ndiaye’s push is bolstered by years of research about how the lack of data on race and ethnicity prevents France from tackling racism– a challenge summarized succinctly by Patrick Simon, research director of France’s National Institute of Demographic Studies, with this example: “We don’t know the salary gap between Blacks and whites in companies, because we refuse to speak about Blacks and whites.”

It’s important to note that France is not an outlier. Twenty out of 38 of the OECD countries don’t collect racial and ethnic data. This despite calls to do so from activist and human rights groups, and the United Nations.

So–what does the French take on this have to do with the #wholookslikeme metric for the data industry?

Reflecting on the French discomfort with ethnic and racial data, and their adamant refusal to collect it, made me think differently about my own discomfort with collecting that type of data from the profiles I interact with on Linkedin.?

In my view, it’s high time that the French just get past their discomfort.?

And–why should this be any different when it comes to my #wholookslikeme metric?

After all, this call to action is not about me. The purpose of this race and gender data effort is two-fold. First, it’s to benefit the people who are being excluded from the data community. Second, it’s for the good of the data community itself–since everyone in it benefits from it becoming more diverse.?

If it’s not about me, then my feelings simply… don’t matter.?

Yes, it’s still somewhat uncomfortable for me.?

Undoubtedly, I’ll make mistakes and missteps.

But it needs to be done.?

Next edition, I’ll finally publish my #wholookslikeme score.?I invite you to collect your own data, and to join me in the conversation about how we can push for greater diversity in our industry.?

Mary MacCarthy is journalist and a Data Advocate at Hightouch. You can DM her hear on Linkedin, follow her on Twitter @MaryMacCarthy, or email her [email protected]?

NOTES:

* I’m aware that this might seem like a painfully simple data collection and measuring methodology. I could certainly automate part of it by ingesting the Linkedin data, and I could also devise a more complex metric (for example, weighing the quality of interactions–with more importance given to commenting on a post rather than just “liking” a post). But I think it’s worthwhile starting with something extremely straightforward that everyone can do on their own, in confidence that the results are accurate.?

** You don’t have to take my word about the state of racism in France. Here’s some solid? reporting in English on the topic. And this article from Le Monde just last week describes what I saw first-hand working in France’s television industry.

*** It’s important to note that both the U.S. and French policies on collection of race and ethnicity data are part of complex, fraught histories when it comes to these issues. The U.S. has been collecting race information since the first census in 1790–which seems, like, incredibly forward-thinking! But a look at the categories being collected shows that the data collection was in service of maintaining the inhumane racial status quo of the time: Americans were categorized only as white, slaves, or “all other free persons” (free Blacks and some Native Americans).?Conversely, the French aversion to collecting this type of data is often discussed in relation to their memories of the murderous Nazi occupation–when the French witnessed collection of racial, ethnic, and religious data for the worst possible purposes.?