Lack of Equitable Data is Harming Underserved Communities, U.S. Chief Data Scientist Says
Enhanced data collection efforts and outreach are needed to improve outcomes for underrepresented communities across the country.
Policymakers need to do a better job of identifying disparities in available datasets in order to develop more equitable policies and practices, the U.S. chief data scientist said during a webinar on addressing the data divide that was held by the Center for Data Innovation on Tuesday.
Denice Ross, chief data scientist in the White House Office of Science and Technology Policy, said that policymakers, researchers and companies must move away from the trend of taking existing data for granted, since “using only the data we have can mask important aspects of our society.” To do so, companies and policymakers must focus their efforts on collecting more equitable data that better accounts for the experiences of underrepresented groups. The lack of accurate data and data collection methods for these marginalized communities has led to a growing data divide, in which high-quality data on a variety of issues often does not reflect the experiences or needs of vulnerable or underrepresented populations.
“Equitable data are those that allow for rigorous assessment of the extent to which government programs and policies yield consistently fair, just and impartial treatment of all individuals,” Ross said. “Most importantly—and this is what really drives me professionally—these data illuminate opportunities for targeted actions that will result in demonstrably improved outcomes for underserved communities.”
Ross said that President Joe Biden has prioritized efforts to bridge the data divide since his first day in office, when his administration released an executive order on advancing racial equity and support for underserved communities that included the establishment of an Equitable Data Working Group.
“A first step to promoting equity in government action is to gather the data necessary to inform that effort,” the order said.
In July, the Biden administration released another executive order creating a new federal subcommittee on sexual orientation and gender identity—or SOGI—data under the working group’s authority “to advance the responsible and effective collection and use of data on sexual orientation, gender identity and sex characteristics.” In a Federal Register notice posted on Aug. 24, OSTP asked for public feedback to help guide these efforts, including input on where potentially useful data is lacking and what privacy policies federal agencies should follow when collecting SOGI data.
Ross cited the administration’s efforts to collect these types of data—and the ongoing call for public feedback—as “one really concrete place where we can address an extensive data gap.”
Panelists during the webinar cited the need for high-quality data to better target services and support to underserved communities, many of whom are already affected by a lack of accurate data collection or data models that minimize or ignore their experiences.
Ioana Tanase, accessibility program manager at Microsoft, said that people with disabilities often feel invisible because of “the lack of data to serve their needs.”
“We simply don't have a grasp of how big the scale of disability is,” Tanase said. “We presume and we estimate, but there’s no accurate number. And obviously that has a real life impact—how governments make decisions when it concerns policy, how nonprofits choose to focus on certain issues and how companies operate. It all gets impacted.”
Tanase cited the growing trend of companies using artificial intelligence hiring tools in their recruiting processes as a use of data that does not always reflect the realities or lived experiences of disabled individuals.
“They can actually sustain issues and even deepen the disability divide when it comes to the employment of people with disabilities,” Tanase added.
While the panelists all agreed on the importance of more accurate collection methods to bridge the data divide, they expressed concern that the historical uses of collected data to target and harass minority groups could limit enhanced data collection efforts. To overcome some of these fears, they said that researchers and policymakers should partner with community organizations and nonprofits to show the value of accurate data in promoting more equitable policies and services.
Dr. Dominique Harrison, the director of racial equity design and data initiative at Citi Ventures Innovation, said that “we’ve seen the harmful effects” of data used by companies and others, but added that “it’s important to understand what those communities may be experiencing in their access and use of technologies.”
“It matters to have a diversity of folks involved in this research process,” Harrison added. “And I think it’s important because you need some type of cultural competence as you’re going along and doing the research.”
The webinar came after the Center for Data Innovation released a report on Aug. 22 that discussed the impact of the data divide on equitable access to critical public and private services. The report included a number of policy recommendations, including calling for Congress to establish a bipartisan federal commission to “study the data divide, including its causes and impacts, and provide recommendations on opportunities to close the data divide to promote more economic and social opportunities for all Americans.”