When Congress passed the American Recovery and Reinvestment Act (Recovery Act) in early 2009, the legislation's transparency provisions represented a significant step forward for government openness. While select agencies and programs have been using recipient reporting for years, the Recovery Act represented the first time such reporting had been attempted across all agencies at once and presented to the public online. Thus, bumps in the road toward transparency and accountability, including data quality problems, were inevitable. A new Government Accountability Office (GAO) report shows that while there are fewer reporting errors as time passes, there is still room for improvement in both data quality and implementation details.

Understandably, the first battles over the Recovery Act revolved around implementation details, as the Obama administration tried to turn the law into a working framework for reporting. After the first round of recipient reports were released, though, the public's attention turned to data quality, as high-profile mistakes, such as "phantom" congressional districts, dominated news coverage.

The GAO report moves beyond these issues, with the main focus being an examination of how states and local governments are using and administrating their Recovery Act funds, with emphasis on education, transportation, and public housing funds. The GAO publishes a report like this every two months to help give a sense of how Recovery Act money is being used, whether the spending is meeting the act's dual goals of recovery and helping those most in need, and how federal agencies are managing the act's transparency provisions.

In this report, the GAO also fact-checked a selection of third-round recipient reports, looking for errors. In previous examinations of recipient reports, the error rate was not exceptionally high, but it was enough to add caveats to any analysis of Recovery Act data, according to GAO. And while previous GAO analyses, and select data from Recovery.gov, indicated that the error rate was dropping, having only two sets of recipient reports made it difficult to establish a trend.

In its examination of the third round of reports, GAO confirmed this trend, finding that while recipients are still reporting flawed information, they are doing so with far less frequency than in the previous two quarters. For instance, compared to the second quarter of recipient reporting, reports incorrectly marked as finished fell by almost 60 percent, and errors in reporting the Treasury Account Symbol (TAS) code and Catalog of Federal Domestic Assistance (CFDA) numbers fell by 40 percent.

In other words, recipients are learning how to report. It appears that the more often recipients report, and the more time agencies have to communicate to recipients how to report, data quality improves. Unfortunately, the report does not give overall numbers for data quality, making it difficult to gauge how serious the data quality problems are.

In an effort to address the lingering data quality problem, the GAO met with a handful of outside groups to "solicit feedback from data users about their use of recipient reported data and suggestions they had for improving the recipient reporting process and the data." Groups interviewed included transparency organizations such as OMB Watch, news media such as ProPublica, quasi-governmental groups such as the Council of State Governments, and research entities such as the Federal Funds Information for States. The GAO report conveys a wide range of critiques and improvements suggested by the groups.

The most prominent critique from these groups concerned data quality, which fits well with the report's focus on the quality of the recipient reports. The groups were particularly concerned with the "capacity of recipients to report correctly, difficulty with determining the flow of awarded funding through state capitals or state agencies down to the local level, and the difficulty using the data across quarters because of the FTE calculations." To help fix these problems, the groups suggested that Recovery Act information be provided at the county level instead of at the congressional district level; agencies allow more time to review reports so that they can spend more time correcting errors; and a move toward ultimate-recipient reporting instead of the current two-tier model, to give a more comprehensive picture of where the money is going.

It is important to note that the GAO is not officially recommending any of these changes and has decided to not evaluate the effectiveness of any of the suggested improvements. Most of the report's recommendations are either agency-specific, such as recommendations for the Department of Education on how to report certain education jobs, or are focused on improving agencies' single audits. Since the GAO is an auditing office, it is not surprising that its recommendations are so narrowly focused. However, observers say instituting something such as ultimate-recipient reporting would help the GAO accomplish its mission of investigating the uses of Recovery Act funding.

Despite lingering data quality problems, the GAO found that recipient reporting is relatively well received by states. Although some state officials complained that the reporting requirements demand precious resources, many expressed optimism. According to the GAO, "when asked about the perceived costs and benefits of the recipient reporting exercise, state officials reported benefits resulting from the reporting requirements." States found that the recipient reports gave them a trove of new data, which they could then use to help educate their citizens, revamp websites, and overhaul internal cost controls. While data quality may still be an issue, this positive reception indicates that recipient reporting has been a positive development and will likely continue to be an aspect of federal spending transparency in the future.