GSA has been given more of the government’s contracting authority

As Nextgov/FCW reports, the General Services Administration is likely to take over the two most important — and longstanding — GWAC’s outside the agency’s control: NASA SEWP and NIH CIOSP. This is a further expansion of the philosophy outlined in a Trump executive order directing agencies to centralize procurement at GSA.

As I have written recently, getting GSA more involved in contracting is a back to the future approach that harkens back to the period between the establishment of GSA in 1949 and the procurement reforms of the 90’s, when GSA did a lot of the buying for the federal government. The changes in the 90’s were designed to give the agencies and their contracting officials more power and to tame GSA.

Although I was involved in these changes as administrator of the Office of Federal Procurement Policy from 1993 to 1997, I was actually less enthusiastic about them than other officials I worked for because I was worried that neutering GSA would make it harder for government to get price discounts and better service from large buys.

But another change in the 90’s worked to alleviate these problems: the growth of non-GSA governmentwide acquisition contracts. The first of these was proposed by the Transportation Department, whose procurement officials there came to me, as OFPP administrator, to bless. They came up with the innovative idea that an agency outside GSA could award a contract open to the whole government. I approved the idea, and the GWAC was born. 

As it happened, Transportation’s GWAC never took off. But two of the contracts appearing at that time did: SEWP and CIOSP. They actually pioneered the innovation, which has since become standard practice, for a two-stage procurement process for IT buys. In stage one the government held a full and open competition for prime contract holders. Then the limited number of winners of the first-stage competition would compete using simplified and fast procedures for task orders under the prime contract. The idea was to mix streamlining with competition.

The system did not work perfectly. The government often chose too many winners in the first stage, usually most of those who bid. But in the second stage, they went from hyper-competition to not enough competition, with the government often making task order awards with only one bid. Towards the end of my government service, I spent a lot of my personal time trying to get the original intent of the two-stage competition to become reality, and my impression is that this gradually more or less happened.

The ideal is to get a balance between the pricing and efficiency advantages of few awards and the customer service and choice advantages of many awards. I actually think the Trump administration had landed in a pretty good place in this balance.

But now I’m worried that folding SEWP and CIOSP into GSA will unbalance the system by putting too much centralization in GSA. Both SEWP and CIOSP have historically pioneered various customer-centric innovations. And these central contracts that are outside GSA are helpful for agencies with insufficient in-house contracting staff or staff that is not customer-centric enough.

We had landed in a good place, and I am worried recent changes might screw this up.