Ph.D proves DOD on right procurement track

Doctoral dissertations on government procurement are not exactly a dime a dozen. I was therefore interested to see that a Ph.D thesis in public policy recently completed at Carnegie Mellon University, one of the country's most academically rigorous public-

policy programs, was an empirical study of how the Defense Department purchased commercial off-the-shelf software and electronics parts. I was even more interested to see the thesis conclude that the federal work force does a much better job negotiating good prices than most people believe.

Air Force Maj. Joseph Besselman, who works at the Standard Systems Group in Montgomery, Ala., authored the study, so some may accuse him of bias. But Besselman said he started the study skeptical about DOD buying practices, and his thesis passed muster with a passel of professors sensitive to good methodology— professors who initially shared the view that DOD is an incompetent buyer.

Besselman started his study by noting that accounts of DOD's supposed incompetence are based on anecdotes and that even if the anecdotes are true as represented (which they often turn out not to be), DOD makes literally millions of purchases each year, so determining if the anecdotes are commonplace is crucial. It would be rational for DOD, just as it is rational for any consumer, to pay significantly greater attention to big-ticket buys than small ones; such a (rational) approach might produce atypical anecdotes that show overpayments for small items.

Besselman researched actual samples, not just anecdotes, of purchases of electronics parts and COTS software. For parts, he collected data on all commercial-parts purchases made by the Defense Electronics Supply Center in Columbus, Ohio, over an l8-month period during l995-96. He compared DOD prices with catalog prices and with actual prices paid by large commercial customers he contacted.

For software, he looked at all orders for software packages from January to September l996 under a larger contract with an integrator; the contract included services and hardware as well.

Intriguingly, Besselman found that when considering each item in his sample, without weighting the item by price or by how many were bought, DOD overpayed for more items compared with commercial buyers and underpayed for fewer items. However, if the sample were weighted by the amount of money DOD spent on the items, DOD did extraordinarily well on big-ticket items. On a dollar-weighted basis, in l996— several years into acquisition reform— DOD paid 22.6 percent less than the commercial retail price and 5.9 percent less than large private customers for electronics parts.

For software, DOD paid 53.7 percent less than vendor catalog prices, 34.2 percent less than General Services Administration schedule prices and even 30.9 percent less than prices listed on the contract with the integrator. The higher-paid prices were for the nickel-and-dime buys— hence the anecdotes of overpayment.

Negotiating Discounts

How did DOD get such good prices? For large buys, DOD negotiates significant quantity discounts. Discounts are greatest when DOD makes a minimum-purchase commitment. For software buys, the key to the astonishing finding that actual prices were much lower than prices on the contract itself was the Air Force's aggressive use of practices that unfortunately are not typical for government as a whole. In particular, the Air Force took advantage of vendor end-of-fiscal-quarter discounts, when salespeople are trying to make quota, and of other short-term buying opportunities.

These findings should provide a real tonic for federal employees and should have implications for procurement policy. The benefits of quantity discounts and the high prices DOD pays for small buys compared with the commercial world, show how important it is to resist the misguided efforts of those who, in the name of fighting "contract bundling," would require the government to buy in small quantities.

It also underscores how crucial it is to develop electronic malls, which offer products from quantity-discounted contract vehicles and which are accessible to credit card users. Such malls allow the government the best of both worlds— the convenience, quick turnaround time and administrative cost savings of the credit card combined with the great prices available from contract vehicles, thereby reflecting government buying power. Without electronic malls, the government risks achieving the convenience and administrative savings while still paying retail.

There is finally a lesson for how government users can benefit from the business skills of contracting professionals. In the situations Besselman described, contracting folks not only negotiated good contracts but also made use of tricks of the trade, such as taking advantage of special discounts to get great deals. (GSA, take note: I'd love to see the Federal Supply Service publish a list of when fiscal quarters of publicly traded schedule holders end so that customers can negotiate, as the Air Force did, to get better deals.) In times when program dollars are scarce, that partnership is good business.

-- 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.