Transparency on government spending: The missing links
- By Alice Lipowicz
- Feb 15, 2012
When I started reporting in Washington in the pre-Internet days, the first Monday in February brought a familiar ritual: Journalists rushed around town trying to squeeze in as many federal budget briefings as they could and returned to their offices with a huge stack of paper.
Even with thousands of budget figures made available on a single day, it was often difficult to get basic information for comparing numbers from year to year or agency to agency. Organizational divisions and programs would change, merge, be renamed or eliminated. Counting personnel was tricky (we worked the phones on that). Comparing certain types of spending from agency to agency required a fair amount of guesswork.
Since then, documents have become digitized, and the Web has made data accessible around the clock. Hard work has gone into USAspending.gov and Recovery.gov to make spending details available to the public.
But even with all the advances in IT in the past two decades, when I searched the Web last week to find out how much the Smithsonian Institution spent on digitization projects in recent years, I was out of luck. A helpful staffer gave me the available information over the phone.
For now, the idea of a fully transparent federal budget publicly accessible online is just that — an idea. But a few people are trying to raise it to the next level.
Jim Harper, director of information policy studies at the Cato Institute, suggested that I visualize the federal money flow as a five-layer process. The appropriations come in at the top layer, which consists of the departments. Then the money flows to the bureaus in the second layer, programs in the third layer, projects in the fourth layer, and finally out at the bottom in spending, including payments to contractors.
Currently, we have pretty good information from the president’s budget on layers 1 and 2, and from USAspending.gov and Recovery.gov, we can report on layer 5, albeit with some shortcomings, Harper said.
What is missing is the ability to track the money through layers 3 and 4, Harper said, which makes it impossible to follow an expenditure all the way through the system.
For the system to be transparent, Harper said, the government needs to make its spending data machine-readable and organized into a data model. That big but doable job would incorporate existing unique numeric identifiers for each department and its bureaus and would create new unique identifiers for programs and projects. The result would be a machine-readable organizational chart that could be used to tag and track the flow of money through the system.
“There currently is no authoritative organizational chart online,” Harper said. “It has never been done.”
The White House’s Appendix C to the annual budget contains the closest thing, he said. But it is not machine-readable, and many of the categories are fuzzy, catch-all listings rather than true organizational units.
The usability gap
Another sticky wicket is the system of numeric identifiers for contractors, managed for years by Dun and Bradstreet. Ideally, experts say, agencies would own and use their own corporate identifiers or engage in some type of open-source data exchange. Some experts are already tackling the problem.
Meanwhile, although Harper is eager to get started on the federal budget data model, he acknowledges that the massive amount of data to be generated would require expertise to chart. At the Project on Government Oversight (POGO), creators of a groundbreaking database on federal contractor misconduct, one of the areas of emphasis is effective tools for making the data understandable for the average person.
Otherwise, even a great transparency idea — such as the Federal Awardee Performance and Integrity Information System (FAPIIS) database on contractor performance — can stumble.
“What we have noticed is that the federal government often falls short on execution,” said Joe Newman, director of communications at POGO. “FAPIIS is an example of great intentions resulting in something not very usable.”
And not being usable means it is not very transparent either.
For groups trying to move transparency to the next level, I hope that doesn’t mean all the pushing will be uphill.