Don’t panic. Collaborate.

Covid-19 threatens local journalism, just when we need it most. Here’s how the Covid Public Info project is responding.

John S. Knight Journalism Fellows prototype journalism projects, Winter 2020. (Pamela Chen)

Cross-posted with the Covid Public Info team.

Few journalists could have planned for 2020. A life-threatening pandemic, economic recession and deep-seated racial inequality have collided with virtually every storyline and beat at the dawn of the decade.

Meanwhile, a decimated news industry has weathered waves of job losses at a moment when the public depends on accurate, enterprising local journalism.

At Stanford University, John S. Knight Journalism fellows were midway through an academic year exploring solutions to the industry’s most urgent problems when coronavirus began to spread across the globe. It became clear that this was not a scholarly exercise. The threat was real, communities needed information fast and the situation was changing from one moment to the next.

A group of current and former fellows—Sarah Alvarez, Garance Burke, Matt Kiefer and Michael Morisy—decided to combine our efforts to support journalists at the local level, where we feel the need is most urgent. We focused on our areas of expertise: using SMS surveys to identify community information needs, reaching out to local newsrooms to address coverage gaps, planning and scaling local public records requests and identifying angles to hold governments accountable for algorithmic decision-making.

The Covid Public Info project launched this spring after winning a grant from the John S. Knight Journalism Fellowships. We’ve collected input from thousands of individual community members, filed hundreds of public records requests and compared notes with dozens of journalists. Here are some highlights of what we’ve learned so far, and ways for you to get involved.

Step 1: Listen

Even before the grant officially kicked off, the Outlier Media team already was conducting community surveys in Detroit and Milwaukee, focusing on the information needs of low-income communities that media outlets were often overlooking, even before the pandemic. The results were eye opening, showing that while every community was impacted, they were all impacted in very different ways. Some through lines became clear.

One thing we identified early on was that Covid-19 was really three crises at once: The devastating health crisis; an economic crisis that most impacted those already struggling; and an information crisis, as local news organizations were collapsing while misinformation ran rampant.

We also took time to listen to newsrooms and potential partners, including organizations in the social services and healthcare sectors, about what information needs they had, surveying newsrooms and having one-on-one conversations to hear where information providers were struggling.

What we learned: The crises facing communities across America impact each community differently, so reporting resources need to be flexible enough to serve those varied interests. At the same time, reporters are looking for resources and a chance to compare notes, particularly as everyone’s beat has — to an extent — become the healthcare beat, whether or not they trained for it.

(Brian Wangenheim/Unsplash)

Step 2: Prototype

When working with local news partners, it’s important to prototype and refine story ideas to understand how they might serve readers.

Case in point: In early April, we identified a critical data source for reporting on the local impacts of Covid-19. The data portal for the Cook County Medical Examiner’s Office published real-time, detailed individual death records that we used to identify disproportionate Covid-19 mortality among Black Chicagoans, among other findings. We shared this info publicly as local newsrooms picked up on this line of reporting in Chicago and elsewhere.

These data sets proved to have broad applications for local-level reporting across the country. We started a sign-up sheet for journalists to participate in data collection efforts and planned a coordinated approach to prioritize our public records request campaign, focusing first on the Midwest and Bay Area and expanding from there.

What we learned: In conversations with journalists at newspapers, radio stations and online outlets from coast to coast, we’ve identified at least 10 major story angles that could be derived from death records, including: racial disparities, missing data points, clusters in nursing homes, outbreaks in jails, job-related Covid-19 fatalities, impacts on the homeless, medical examiner backlogs, investigations into the earliest coronavirus deaths and overall counting accuracy, among others.

We do our best to listen to our newsroom partners, find common threads and think of ways to scale the work. We have sent out more than 100 public records requests and plan to share training materials and other information on how journalists can get involved in this effort. We’re hosting a public webinar on this project on July 30. Email mkiefer@stanford.edu if you’d like to participate.


Step 3: Anticipate

Even as we keep pace with a news cycle that eclipses anything we’ve seen, we’re looking ahead for what newsrooms nationwide will need to be asking and thinking about going forward. That involves a deep dive into all the ways — often just behind the scenes — that algorithms play a major role in government decision-making, whether or not the public knows about it.

We have filed dozens of public records requests into how algorithms and other predictive tools decide who qualifies for unemployment benefits or early release from prison, as well as how facial recognition is being applied in new and disturbing arenas. Currently, we’re reporting on the use of algorithms in state benefit fraud detection, an issue that is impacting millions of Americans who have recently applied for unemployment. We believe that investigating algorithmic decision-making power in society is crucial, and that revealing the ways that these tools are impacting our lives is ever more important in this tumultuous time. Email gaburke@stanford.edu if you’d like to participate in this effort.

We’re also hoping that our work builds ongoing partnerships between newsrooms, and with reporters and the communities they serve. The impact of Covid-19 on health, wellness, inequality and so much more will be with us for years to come and newsrooms must adapt to better serve their communities in direct and creative ways.

Get involved

Sign up for the Covid Public Info newsletter to stay up to date with the project and find ways to get involved.


  • Sarah Alvarez

    Sarah Alvarez is founder of Outlier Media, a journalism service that delivers data reporting and information to low-income news consumers over SMS and message apps. She is a former lawyer and former public radio reporter and was a 2015 John S. Knight Journalism Fellow at Stanford University.

  • Garance Burke

    Garance Burke is a 2020 John S. Knight Journalism and Human-Centered Artificial Intelligence Institute Fellow at Stanford University, where she is researching algorithmic accountability. She is on leave from her position as a national investigative reporter at The Associated Press, where her stories and collaborations were honored as a finalist for the Pulitzer Prize in 2019.

  • Matt Kiefer

    Matt Kiefer is a 2020 John S. Knight Journalism Fellow at Stanford University, where he designed and prototyped tools to monitor civil and human rights. He is also a project lead for the journalism collaborative, Covid Public Info, and the creator of FOIAmail, an open source framework for automating public records requests.

  • Michael Morisy

    Michael Morisy is the co-founder and chief executive of MuckRock, a journalism tools non-profit that runs DocumentCloud, oTranscribe, and other open-source ways to better inform the public. He was previously a John S. Knight Journalism Fellow at Stanford University and an editor at the Boston Globe.


Current page