Researchers Find Data Limitations in USASpending.gov, FAADS

December 10, 2013 | By admin | Post a Comment

untitled-1282930-mWhile we “enjoy” our first week of snow here in the D.C. area of the winter (um, can someone let Mother Nature know that it is not officially winter yet?), we’ve come across a recent report by two university researchers whose reaction to USASpending.gov was just as chilly as our recent temperatures.

A recent FierceGovernmentIT article discussed a Nov. 25 report by researchers Jesse Lecy of Georgia State University and Jeremy Thornton of Samford University that found that data fields on USASpending.gov can be subject to considerable interpretation. The researchers sought to determine how easy it was to find information on government grants to nonprofit organizations on USASpending.gov. When trying to study 2012 data, they found numerous data limitations.

The article states that the researchers aimed to identify nonprofits within data sets about federal grant recipients, but the Federal Assistance Award Data System and the Federal Procurement Data System record entities according to DUNS numbers. However, the Internal Revenue Service tracks entities according to Employer Identification Numbers.

The researchers wanted access to EINs to match USAspending.gov data taken from FAADS and FPDS to nonprofit financial information held by the National Center for Charitable Statistics, which uses the publicly available EINs as its database unique identifier. The center classifies charities according to the National Taxonomy of Exempt Entities; correlating federal recipients to their NTEE major category would allow the researchers to identify which sectors receive the most federal support, according to the article. While the System for Award Management contains both DUNS and EIN records, the EIN isn’t made publicly available.

The researchers also sought to bridge the databases through an algorithm that would match nonprofit names recorded in FAADS or FPDS to their known EINs listed by the National Center for Charitable Statistics, but could attain only a 70 percent match rate with reasonable levels of confidence.” The researcher also said they found errors within FAADS reports, such as a single organization sometimes listed as a nonprofit, a for-profit or as a government agency,  adding that unlike the reporting field for prime awards, FAADS lacks a field to indicate whether a subaward grantee is a nonprofit.

“For example, a particular award to an organization may be categorized as a nonprofit one year, but a government agency the next,” the report states. “Significant inconsistencies in award categorization over time limit the usefulness of the FAADS data for statistical analysis.”

The report shows that with all the resources placed on getting these governmental systems up and running to provide data on federal funding, much more needs to be done to ensure that the data is accurate. Having a system such as USASpending.gov available is critical for transparency and accountability, but we need to find ways to make sure the data is reliable.

What has been your experience with USASpending.gov? Have you found similar problems with unreliable data?

 

LinkedInShare

Post a Comment

Your email is never shared. Required fields are marked *

*
*