Computing Research Policy Blog

Federal Spending Database

Senator Tom Coburn (R-OK) wants there to be a single source of information explaining where federal money is spent, and it appears the rest of the Senate Homeland Security and Government Affairs Committee agrees, approving legislation on July 27th that would require the Office of Management and Budget to create and establish a database of government spending. According to an article in the Washington Post, the database would include contracts, subcontracts, grants, subgrants, loans and other forms of financial federal spending. The bill also specifies that the data must “be searchable by agency, geography, industry, congressional district and types of funding.” Apparently, it is meant to be a Google-like search engine that would track a trillion dollars in federal grants, contracts, earmarks, and loans .

Other agencies have their own databases that serve the same purpose and are utilized in the same manner (such as the RAND/NSF RaDIUS and the GSA’s Federal Procurement Data System- Next Generation), but there is not one central funding database that is interoperable among all of the agencies. There have also been problems collecting data for these systems, and a centralized database under the department charged with financial oversight may help to streamline this process. This system could provide a strong analysis tool that could be used by agencies, and not simply those who audit them, to see how money is spent.

CRA's Peter Harsha adds:

While transparency in government is generally a good thing, there are some issues with the use of this kind of database. It sounds like an interesting tool for higher-resolution reviews of federal funding – “how much did institution ‘A’ receive from all federal sources?” for example (though that kind of question makes a lot of industry types very nervous…one reason the bill might not go anywhere). But for the more meta-level review – “how much is agency ‘X’ spending on computing research?” – I’m not sure it will be particularly useful. I haven’t seen a database implementation that’s able to handle that kind of query particularly well, mainly because the definition of something like “computing research” can be so variable. Last year , the House Government Reform Committee asked a version of that question to the folks at RAND to figure out using their RaDIUS database (actually, they asked how much the federal government is spending, not just one agency) and got back an answer that, depending on how you defined “information technology,” was between $2 billion and $30 billion annually. So the danger in having this sort of database, I think, is in a user / policymaker not knowing what kind of questions are answerable and what really aren’t, despite the fact that the database will be happy to spit out an answer in either case.

Anyway, it’s not clear that the bill has much of a future, but we’ll post any further developments here.

Posted by EricaCameseToo at August 4, 2006 09:03 AM
Posted to Policy