Managing GWACs: Procurement reform's biggest challenge

We are entering a phase where continuous improvement efforts will be central to procurement reform. Right now government and industry must apply this logic to the management of governmentwide acquisition contracts (GWACs) for services whose abrupt appearance not much more than a year ago - along with the growth of blanket purchase agreements (BPAs) for hardware - has been the most dramatic procurement change for the information technology community in more than a decade.

GWACs and BPAs have produced sensational improvements for federal buyers. Procurement cycle time has been slashed frequently by 90 percent or more. Resources that in the past were poured down a rat hole of bureaucratic paper exercises now have been freed to give managers more time to deliver better results from IT investments. Furthermore GWACs were a grassroots initiative based on provisions in the l994 Federal Acquisition Streamlining Act (FASA) not a top-down fiat. That's good news itself - of a re-energized federal work force that is not simply a passive subject in the command of others.

By contrast GWACs have created significant adaptation challenges and some potentially serious public policy problems. The first is assuring vigorous competition for task orders under GWACs. The second is the ability of price- and quality-competitive small firms to have access to task orders.

Task-order contracting has long been common. Traditionally vendors competed for a contract with a fairly broad scope of work one award was made and the winning vendor executed task orders for specific assignments.

Task-order contracting was criticized as insufficiently competitive because the winning vendor gained a monopoly on a significant stream of business based on winning a competition for a very general scope of work. Some even argued that task-order contracting was illegal under the Competition in Contracting Act. FASA responded by specifically authorizing task-order contacting and establishing a preference for multiple awards in such contracting so competition could continue during the life of the contract at the task-order level.

The FASA framework established streamlined procedures for selecting task-order awardees and did not permit bid protests of award decisions - except if the order was alleged to exceed the scope of the underlying contract - but it also required that each vendor receiving a multiple-award contract be given a "fair opportunity to be considered" for individual task orders.

The first danger in the emerging GWAC environment is that many task orders become de facto sole-source awards. GWAC vendors appropriately are marketing themselves to federal customers. Customers are increasingly able if they choose to create a situation where even though a task order as a formal matter is advertised on a GWAC World Wide Web site vendors other than the one who marketed the agency are unlikely to bid. This is done by making little effort to let other vendors know about upcoming task orders and by giving minuscule turnaround times to bid. Prior to changing its policy on the Chief Information Officer Solutions and Partners (CIO-SP) contract in September the National Institutes of Heath allowed customers to specify one "preferred" vendor - about as strong a signal as you can give that others need not apply.

Some defend this situation by arguing that if a firm can persuade a customer that it's the right one to do the job why shouldn't it be able to just bag the work? However FASA struck a good balance between giving vendors the chance to show their abilities and the virtues of competition in keeping vendors on their toes. As a practical matter if de facto sole-source task orders become the normal way of doing business under GWACs it will only be a matter of time before Congress eliminates this authority altogether. Therefore agencies and GWAC managers must move urgently to stop development of a culture in which vendors assume that if they weren't first to get to a customer on a requirement they shouldn't bother bidding. How? First customers need to do much more to give vendors earlier warnings about upcoming task orders. Second minuscule bid times frankly are greedy on the government's part. In a world where cycle time is down 90 percent it's not the end of the world to give vendors l0 or l5 days to respond instead of five days. And if de facto sole sourcing leads to a return to where it took a year or two to get to contract such greed will have proved to be extraordinarily foolish.

Finally customers must communicate clearly to vendors that task-order competition is for real by carefully considering proposals and by making awards to vendors other than the first one they heard from. Otherwise policy changes such as the recent decision on CIO-SP to require agencies to name at least two vendors if they wish to specify a preference quickly will come to be seen as a sham.

Another crucial GWAC issue involves small business. Many small firms unfortunately have reacted with paranoia to procurement reform. But even paranoids have enemies and there is evidence that GWACs are indeed making it harder for qualified small firms to get business. One problem is that GWACs are often too big for small companies to win as primes. Small firms also complain that primes make it difficult to work as subcontractors.

These problems can be solved but agencies and vendors need to take measures to do so. The Transportation Department's Information Technology Omnibus Procurement worked to get small firms into the system as primes by dividing the overall scope of the contract into smaller bits and setting aside some prime-level multiple awards for small businesses and 8(a)s. This ensures that such firms will be in the mix to compete on task orders. Agencies should continue to consider - as many are already doing especially outside the IT arena - awarding multiple-award task-order contracts focused on narrower kinds of requirements such as data security that will allow high-quality niche players to win.

When GWAC contracts are recompeted I believe the government should require primes to commit to awarding a certain dollar percentage of the total volume of their work to their small-business subs. Primes must move as well. They need to make it easier for subs to bid on entire task orders rather than relegating them to providing bodies to round out the contingent the prime is bidding. And I urge industry associations with large-business prime members such as the Information Technology Association of America to organize pledges where GWAC primes agree to increase the proportion of work they sub to small firms during l999.

Again these changes make good policy sense but the possibility of a hanging should also concentrate the minds of agencies and large-business primes. If GWACs don't become more small-business-friendly they may well become history.

-- Kelman was the administrator of the Office of Federal Procurement Policy from 1993 to 1997. He is now Weatherhead Professor of Public Management at Harvard's Kennedy School of Government.