Steve Kelman contends that the recent GAO report shows the practice has truly arrived
You know that a new practice has "arrived" in government when it becomes the subject of its very own GAO report. This is now the case for the use of reverse auctions in contracting, which was illegal until the Federal Acquisition Regulation was changed in 1997 and now has become an accepted part of the federal contracting environment. (According to GAO, at the four agencies they examined, there was a 175 percent increase in the use of reverse auctions between 2008 and 2012.)
The title of the GAO report -- Reverse Auctions: Guidance is Needed to Maximize Competition and Achieve Cost Savings -- is a bit of a parody of itself; along with "Progress Made, Problems Remain," GAO reports love to call for more regulatory guidance. However, the report is generally a balanced one, clearly recognizing the positive role reverse auctions play as a cost-savings driver in the federal procurement environment but also making suggestions for improvement. (Full disclosure: I am on the board of advisors of Fedbid, the leading reverse-auction provider to the government.)
One noteworthy element of the report was the information, which I don't think has been published before, that roughly a quarter of reverse auctions attract only one bidder. It is really not possible to imagine this is the fault of the procurement method, which widely advertises the availability of bidding opportunities and typically attracts many bidders. My guess is that these situations are produced by very restrictive specs, perhaps combined with set-asides, or sometimes by excessively low government target prices (maybe contingent on funding availability). However, this question deserves further research. As with the data generated by government credit card transactions, the procurement method does not create problems so much as it creates transparency that sheds light on problems that existed all along.
Similarly, the report notes that savings numbers are hard to specify because they depend on the government cost estimate, which may be too high (or too low, again perhaps because of funding limits). This is a problem, but it is not unique to reverse auctions.
As for the report's request for "guidance," I guess on balance I think some FAR language would be a good idea -- not the least for those very conservative contracting officials who, despite the language in FAR Part 1 clearly stating that a practice is allowed if it is in the government's interest and is not forbidden, believe that if a practice is not explicitly authorized in the FAR, you can't use it. However, I would urge that the language be as unrestrictive as possible. It should cover only the most-important issues, and emphasize use of "may" rather than "shall."
For example, I disagree with the report's suggestion that reverse auctions be limited to buys under $150,000 or to "simple" services. One could imagine a larger price-driven buy that would be appropriate for reverse auctioning; indeed, guidance from the Defense Logistics Agency recommends use of this technique mostly for buys over $150,000. Also, I believe that in the future we will and should see the use of reverse auctions among shortlisted bidders on large-dollar best-value contracts, which might involve very complex services, to determine the price part of a bid. That pricing could then be placed into a larger evaluation including other factors. So we don't want "guidance" to be unnecessarily restrictive.
Reverse auctions continue to have real potential for the government. To the extent that GAO's suggestions add to the attractiveness of this technique for government customers, that is good for everybody.