The Challenges of ‘Data Science for Social Good’ in a New Era of ‘Alternative Facts’

SEATTLE — Although the bulk of a two-day workshop on data-driven municipal governance last week at Seattle City Hall focused on case studies and best practices in the delivery of human services, there were very timely moments of discussion on some related, but vital, questions facing anyone in the public sector who deals with data, analytics and performance.

What can data analysts, technologists and public officials who rely on facts, statistics and shared knowledge to inform decisionmaking and improve operations do to combat anti-establishment sentiments and misinformation that sow the seeds of distrust and skepticism in their important work?

At a few points during Tuesday morning’s opening sessions for the MetroLab Network Big Data and Human Services workshop—which drew approximately 100 representatives from 20 U.S. cities plus civic stakeholders, academic partners and non-profit organizations to Seattle—speakers referenced some of the divisions that emerged from last year’s bitter presidential campaign. Donald Trump’s road to a White House victory featured the Republican nominee regularly distort and mischaracterize facts and was accompanied by trolls peddling fake or misleading news online and critics of the establishment, including the real estate mogul himself, skewering elites and questioning the integrity of public institutions.

As of Friday, the Trump campaign’s rhetoric and rancour has found a new home in the White House. On Sunday, presidential counselor Kellyanne Conway characterized an inaccurate statement about the size of Friday’s inauguration crowd made by press secretary Sean Spicer as “alternative facts.”

Facts are supposed to be just those, facts. But facts now seem to live in a grey area in an era of deepening distrust, which puts policy professionals who deal with facts and data in an uncomfortable spot.

It’s a dynamic that’s not confined to the United States, as The Atlantic noted last week. According to a global survey by the communications consulting firm Edelman that’s now in its 17th year, there’s a “widening trust gap” around the world between what the firm calls “mass populations” and “informed publics.”

The group that gathered at Seattle City Hall last week for the MetroLab Network workshop, just a few days before Trump took office, definitely fell into the category of “informed publics.” That included chief data officers, chief technology officers, public health officers and municipal policy analysts—people who rely on facts and data to inform important decisions and have high faith in the abilities of public institutions.   

Former Maryland Gov. Martin O’Malley, a Democrat recognized for his data-driven governance efforts that began when he served as Baltimore’s mayor and later adapted by other city halls, reiterated an important point when it comes to making informed decisions: “Grounding our public discourse on the data.”

That’s been an ongoing challenge in the public sector, but it’s far more difficult than it once was thanks in part to social media and other communication methods that can amplify and disseminate inaccurate information. “The technology got ahead of the data science,” said O’Malley, who now serves as a MetroLab Network senior fellow.

“Mass populations,” and even public officials who would ordinarily be considered “informed publics,” can simply dismiss facts or confuse them when they’re inconvenient or run counter to their beliefs or is otherwise politically expedient.

Trish Millines Dziko, the executive director of Technology Access Foundation, a non-profit organization that works to help students of color in Washington state succeed through STEM education, noted an example regarding the disparity between facts and rhetoric that emerged during the presidential campaign: Myths around public assistance, including, a talking point in some conservative circles that it’s mostly people of color who rely on welfare and other government benefits—which is not the case.

“The story already got out so the data doesn’t really matter,” Dziko said.

The use of data and analytics can be hamstrung in the larger sense by difficulties among “informed publics” to communicate with “mass populations,” to use Edelman’s terminology.

“Elected officials need to become much better storytellers” about insights that come from data, O’Malley said.  

Despite the sour mood in the nation and a political environment that now casts a skeptical eye at informed expertise from civic-minded professionals, it’s critical that their important work continues—including those who are working to improve the delivery of human services at the municipal level.

“It really is absolutely necessary and important for this group to come together,” University of Washington President Ana Mari Cauce, a trained clinical psychologist, said in her welcoming remarks to the group. Cauce described their work as “data science for social good,” a theme that was repeated by O’Malley.  

“This work could save lives of people you may never meet,” O’Malley said.  

Having attended multiple workshops and seminars on data-driven government in recent years, there are certainly many good stories—based on facts, case studies and sound research—to tell the public. So there’s no reason to pull back, even when skeptics armed with misinformation have a bullhorn.

“Get the new stories out, get the new data out,” Dziko told last week’s gathering.

But the overall challenge remains, regardless of how many compelling case studies that data and policy analysts can produce and publish: Getting the public to better understand the data while trusting it and, hopefully, embracing it to create positive change in their communities.