In 2015, open-records advocate Carl Malamud filed a lawsuit to force the IRS to release electronically filed nonprofit tax forms — called the Form 990 — in a machine-readable format. The court ruled in his favor, and in 2016, the IRS began making troves of searchable data available for the first time.
Since then, an increasing number of investigative journalists, academics studying social-sector trends, regulators looking to root out charity fraud, and many others have used the Form 990 to better understand and scrutinize the operations of nonprofits. The annual tax return contains details on nonprofit finances, programs, compensation, governance, and more.
On Tuesday, five organizations announced a new effort to further enable civic and government scrutiny of nonprofits, making it easier to access and analyze large sets of their data. GivingTuesday, the Aspen Institute’s Program on Philanthropy and Social Innovation, Charity Navigator, CitizenAudit, and the Urban Institute have teamed up to pool, clean, and standardize raw nonprofit data in a central clearinghouse. GivingTuesday will host the open-source repository as part of its collaborative Data Commons.
“Right now, a high level of technical competence has been needed to access this information,” says Cinthia Schuman Ottinger, Aspen Institute’s deputy director for philanthropy programs. “This effort will help make it easier for others to tap.”
The effort is also expected to increase analysis of nonprofits of all stripes and even strengthen the field.
The nonprofit sector “is huge and frequently underscrutinized, and 990s are probably the single best tool for forcing public scrutiny,” says Dick Tofel, former president of the nonprofit investigative-journalism organization ProPublica. “Wider and deeper access to them is a really good thing.”
The tax returns have been critically helpful in the world of investigative journalism, he says. “Basically every investigation of a nonprofit that’s ever been done relies on 990s.”
Take, for example, David Fahrenthold’s Pulitzer Prize-winning Washington Post investigation of former President Donald Trump’s exaggerated claims about his philanthropy. Or research from the Kaiser Family Foundation and reporting by Kaiser Health News, which used the forms to investigate to what extent nonprofit hospitals were fulfilling their requirement to provide charity care and other community benefits in exchange for tax-exempt status.
While journalism and research that relied on 990s have been possible for a while, the back-end efficiencies this new initiative enables could make it easier for even more people to dig into the data. Aspen’s Schuman Ottinger says having more data at their fingertipscould make it easier for charity regulators working at the state level to identify patterns of fraud or suspicious activity.
“That strengthens the nonprofit sector as a whole, when we can rely on our regulators to identify problems more quickly than they have in the past,” she says.
A Yearslong Fight
The new data clearinghouse is just the latest step in a yearslong fight to increase transparency around nonprofit data.
Since 2016, organizations like Charity Navigator, Candid, and ProPublica’s Nonprofit Explorer have drawn on a common IRS database of 990s to share information with the public. But that database isn’t user friendly. The data is often complex and requires cleaning and organizing.
Before 2016, groups like GuideStar spent millions of dollars each year to convert images of public filings into usable data. More than a decade after the IRS began allowing nonprofits to submit their tax return electronically, a third were still submitting the paper forms, a 2018 Chronicle review found.
For years, nonprofit watchdogs and transparency advocates pushed the IRS to require all but the smallest nonprofits to file their tax information electronically. In 2019, a coalition of nonprofits spearheaded the passage of a law with that requirement.
But the unwieldy IRS database continued to get in the way of broad access, and people continued to work on an ad hoc basis to make sense of the information.
Those challenges were described in a 2022 report by scholars at the Dorothy A. Johnson Center for Philanthropy at Grand Valley State University, which highlighted the need to create a more coordinated data ecosystem. The Aspen Institute commissioned the report, and the Bill & Melinda Gates Foundation provided funding.
The latest effort came out of a 2022 survey of the Aspen Institute’s Nonprofit Open Data Collective, a group of nonprofit researchers, foundations, data platforms, charity regulators, journalists, and software experts focused on open 990 data use and access.
Collective members that store 990 data can upload their files to the new central repository, which has been developed by staff at the GivingTuesday Data Commons and consultants. Already, groups like ProPublica’s Nonprofit Explorer, Charity Navigator, and Candid have donated their historical data. Soon, for example, a researcher who wants access to raw data on nonprofit executive pay could pull a large set of information from across the field, rather than one organization at a time.
Members of the collaborative will begin testing the clearinghouse’s functionality next week, and by November, GivingTuesday hopes to make the information publicly available for anyone who can access its application programming interface, or API, says Woodrow Rosenbaum, the nonprofit’s chief data officer.
“Now others can really focus on adding value to the data because we will have this standardized, clean, normalized, up-to-date resource,” Rosenbaum says. “Almost anybody who wants to build any kind of tool needs some access to those data.”
GivingTuesday, for one, is already building a prototype large language model on top of the 990 data. The hope is that a user will be able to type a question and unlock information contained in sections of the form that aren’t standardized, such as descriptions of nonprofit missions.
Every industry needs to have a good understanding of the ecosystem in which it operates, Rosenbaum adds. “There’s a lot of rich information locked up in the 990 that we’re not getting at,” he says. “This hopefully allows all of these various stakeholders to devote more resources so we can get more out of it and be a lot more efficient in what we’re already doing.”
Reporting for this article was underwritten by a Lilly Endowment grant to enhance public understanding of philanthropy. The Chronicle is solely responsible for the content. See more about the Chronicle, the grant, how our foundation-supported journalism works, and our gift-acceptance policy.
