U.S. Patent Reforms Needed to Spur Innovation | The Worldwide Economics Department Network
Against Better Judgment: Prequalification
in Procurement Auctions
Whether it is a household purchasing cleaning
supplies or the U.S. Department of Defense
purchasing a weapons system, procurement is
one of the most basic commercial activities.
For some large organizations, procurement
expenditures can total billions of dollars. It
makes sense, then, that managers and academics
alike devote substantial time and attention
to finding the most cost-effective ways to
implement the procurement process.
Traditionally, organizations have used a simple
RFP process: a firm issues a request for proposals
(RFP), potential suppliers submit proposals
in response, and the firm reviews the
proposals to select the supplier who offers the
best combination of price and quality.
As shown by Che (1993) and Rezende (2006),
the problem with the RFP process is that it
does not apply enough pressure on suppliers
to lower prices. In an effort to address this
problem, various organizations have been
experimenting with auction-style mechanisms,
which apply a greater weight to price in the
selection process and provide greater transparency.
These efforts have been quite successful,
generating cost savings of 5% to 15% in many cases (Gehrke et al., 2007).
One auction mechanism that is particularly
prevalent employs a two-stage process. In the
first stage, potential suppliers are prescreened
on the basis of documentation they are asked
to submit. Those qualified move on to the second
stage, where they participate in the auction
itself. In some cases, the auction’s selection
mechanism is purely price-based: the bidder
who offers to supply the product or service
at the lowest price is awarded the contract.
In other cases, the auction selects the winning
supplier by applying a “scoring rule.” A scoring
rule uses a weighted average of price and
quality to assign a score to each bidder. The
bidder with the highest score is awarded the
contract. Scores tend to be heavily weighted
toward price so that in most cases, the winner
is the bidder offering to supply the product or
service at the lowest price.
This two-stage auction scheme is practiced
across a wide variety of organizations.
Pharmaceutical giant Pfizer prescreens potential
suppliers on the basis of quality criteria
such as global capacity, financial stability, risk
management, and past history and performance.
Those suppliers who make the initial cut
are invited to participate in an online auction.
The final selection is made by applying a scoring
rule that features “bottom-line price as a heavily weighted factor” (Gehrke et al., 2007).
The U.S. Navy has been known to administer a
prequalification stage when procuring more
complex items, such as storage tanks, ejection
sequencers, and turbo engine blades. Engineering
data collected as part of the RFQ
(request for qualifications) process is used to
certify potential bidders (Mabert and Skeels,
2002). Indeed, the two-stage scheme is practiced
across so many industries and for so
many different products that it would appear
to represent an optimal way to make procurement
decisions, were it not for one thing: most
economists do not think this two-stage scheme
is optimal.
The academic literature suggests that auctions
which include a prequalification stage are suboptimal
for two reasons. First, prequalification
reduces the number of participating bidders,
which in turn reduces competitive pressures.
If bidder quality is a concern, one could simply
use a scoring rule with a greater weight
applied to the relevant quality dimensions.
Doing so would retain the effect of prequalification
without explicitly removing bidders
from the pool. Second, it has been shown that
the optimal procurement mechanism need
not include a prequalification stage. Theoretical
work in auctions has demonstrated that
the optimal outcome can be achieved by instituting
a scoring rule that understates the
importance of quality (relative to price) and
establishing a minimum score that must be
met in order for a bid to be viable.
It is against this backdrop that the following
question may be asked: why is the two-stage
procurement scheme so prevalent in practice
when the academic literature seems to counsel
against it? Moreover, if the optimal auction can
be implemented in a single step as described
above, why do organizations go to the extra
effort of instituting an additional prequalification
step?
An important assumption in the theoretical literature
is that quality can be easily quantified
and “plugged into” a scoring rule. That is, it
must be the case that scores are assigned in an
unambiguous and transparent fashion. The literature
also assumes that the procuring organization
is able to act in a manner which belies
its true inclination. It must, in other words, be
able to commit to a scoring rule that downplays
the importance of quality and overstates
the importance of price.
Clearly, one or both of these conditions fail to
hold in many cases. When any of these failures
occurs, the procuring organization may be
unable to implement the optimal auction mechanism prescribed by the academic literature.
If, for example, the application of the
scoring rule is not transparent, then it is difficult
to hold the procuring organization to it.
The organization may be tempted to discard
the announced scoring rule and select the bidder
he most prefers. Sophisticated bidders will
understand the organization is likely to apply
a scoring rule that reflects its true preferences.
That is, sophisticated bidders will
understand that the organization will apply a
higher weight to quality than what was
announced. These bidders will respond by
bidding less aggressively, which results in a
suboptimal outcome for the organization.
Alternatively, consider a procurement mechanism
in which bidders are pre-qualified on the
basis of quality in the first stage and participate
in a purely price-based auction in the second
stage. Under such a mechanism, the ability
to easily quantify quality is less important
because the auction does not take quality into
account. Decisions in the second stage are
made on the basis of price alone—which is
entirely verifiable and therefore easy to commit
to. It is still the case that the intensity of
price competition in the second stage is dampened
by reducing the number of participating
bidders but this effect is counterbalanced by
the benefit associated with reducing the risk of
contracting with a low quality bidder.
Moreover, because the second-stage auction
operates on the basis of price alone, price
competition will intensify for those bidders
who do make it past the initial screening.
My hypothesis is that the two-stage mechanism
delivers a better outcome than the RFP process
and may be the best the procuring organization
can do in light of the constraints imposed
by the failure of the two conditions. The mechanism
outperforms the RFP process because
the price-based auction mechanism in the
second-stage stimulates price competition
while the pre-qualification process in the first
stage guards against contracting with a low quality
supplier. This hypothesis would explain
why the mechanism is so prevalent in commercial
arrangements in spite of the academic
findings.
For more information about this project,
please contact Professor Jennifer Lamping,
lamping@colorado.edu.
References
Che, Yeon-Koo, “Design Competition through
Multidimensional Auctions,” RAND Journal of
Economics, vol. 24, no. 4, Winter 1993, pp. 668-
680.
Gehrke, Allen R., Elizabeth Faulkner, and Cindy
van Dijk, “Reverse Auctions: Crusade or Curse?”
Applied Clinical Trials, January 1, 2007.
Mabert, Vincent A. and Jack A. Skeels, “Internet
Reverse Auctions: Valuable Tool in Experienced
Hands,” Business Horizons, July-August 2002, pp.
70-76.
Rezende, Leonardo, “Biased Procurement,”
mimeo, University of Illinois, 2006.