- Pivotal Ventures
- Each year, Melinda Gates and her husband Bill publish an annual letter outlining the things that cheer, worry and surprise them.
- The 2019 letter was published on Tuesday. Business Insider met with Melinda Gates to discuss one of their most surprising assertions in it: that data can be sexist.
- Business Insider sat down with Gates to learn more about sexist data and how it damages women in the US and the world.
When you are one of the world’s wealthiest women and you’ve dedicated your life to solving poverty, how do you know which projects are worthwhile to fund and which are not?
If you are Melinda Gates, a software engineer by training, you do what engineers always do: look at the data.
But Gates and her husband Bill were disturbed when they discovered that data can be sexist. It can also be biased. It might even be racist.
She recently sat down with Business Insider at her office outside Seattle to discuss the issue, which is highlighted in the Gates’ 2019 annual letter.
Gates is a champion for gender equality, one of the UN’s sustainable development goals. She has pushed for better family leave policies in the US and has spoken out about problems women face in the tech industry.
Now Gates is taking on data, a sacred cow of that industry.
“We think data is objective and that’s one of the things that surprised Bill and I the most,” she told Business Insider.
Gates began to see the problem when she learned just how little data is collected on the lives and deaths of women worldwide.
An example: Medical professionals often didn’t collect all pertinent details surrounding a woman’s cause of death in the US until 2017 – including whether or not she was pregnant.
That’s alarming, especially when you consider the US has a surprisingly high rate of maternal deaths from pregnancy- or childbirth-related causes.
“That, to me, is sexist data,” Gates says.
The Gates’ annual letter also discussed a surprising finding about premature births. DNA testing company 23andMe found a correlation between preterm labor and low blood levels of a mineral called selenium.
But without enough historical data on women, it’s tough to turn that finding into an action.
“Even in the preterm birth area, you have to look at Hispanic women vs. Caucasian women. Well, Hispanic women are having more preterm births than Caucasian women. Why is that? And where is that happening? Is that true across the nation or is it different in different pockets?” Gates said. “Until you get that granular in the data, you won’t know what to go look at to say, ‘Where do we need to go to invest money and create solutions?'”
Gates also noted the well-documented history in the US of under representing women in drug testing and clinical trials.
That’s changing, “but we used to test drugs with men and very few women in the trial,” Gates said. “We even used male mice and very few female mice,” even though women and men have different metabolic systems.
Some researchers say that women are still under-represented in clinical research, even for illnesses that commonly plague women, like heart disease.
“You have to collect the right data. And then you have to disaggregate it and look at it,” Gates says.
Data can be racist, too
While the annual letter talks at length about sexist data particularly in developing countries, US researchers have been known to overlook important information about other populations as well.
- REUTERS/Tami Chappell
“We don’t collect all the same data about people across the United States,” Gates said, noting that the foundation has supported work from the groundbreaking Harvard economics professor Raj Chetty on poverty in America.
More details need to be gathered on the ethnic backgrounds of people, particularly those living in low-income areas, to know what their communities need.
Do they have a good gym, solid after-school programs or decent parks? Such research helps policy makers decide how and where to invest in projects that “help people be safe and thrive and lift themselves up,” Gates said.
Does this mean that female researchers or people of color are less likely to create sexist or racist data?
“We don’t know, because we haven’t had men and women equal in those positions for a long enough time,” Gates said.
“But I know when you have diverse viewpoints at the table – women, people of color – the questions change.”