Governing Magazine/August 1994

TECHNOLOGY COLUMN

REINVENTING TECHNOLOGY PROCUREMENT

By Jerry Mechling

Jerry Mechling is director of the Program on Strategic Computing and
Telecommunications in the Public Sector at Harvard University's
Kennedy School of Government.

Most people agree that government does a poor job buying computers.
The process is overly costly and detailed, it takes too long, and, too
often, it gets poor results. The recent $44.3 million fiasco involving
a California Department of Motor Vehicles' computer system is just the
latest example.
 To minimize the chances of such failures, the conventional wisdom
calls for even more control--stronger purchasing rules, additional
layers of bureaucracy, more auditing. Unfortunately, this usually
doesn't work, and it often makes things worse, while it also squeezes
out desperately needed innovation.
 The core problem with procurement arises from a vicious cycle:
Inappropriate controls lead to poor performance, which leads to more
controls, and so on. In the typical approach to procurement, extensive
rules require prior specification of needs and competitive bidding;
provide numerous opportunities for external reviews and appeals; and
restrict vendor selection to detailed criteria mandated by the agency
and published in advance.
 That is the way most government purchasing is done, and it is not
always the wrong way. It can sometimes succeed, as long as what is
being purchased is an off-the-shelf commodity--shrink-wrapped
software, for example, or standard-model personal computers. It often
does NOT succeed for "non-commodities": complex, innovative, custom-
designed goods and services such as emerging technologies or massive
new software systems. In non-commodity settings, value cannot be fully
measured or specified in advance.
 The best approach for reform is not a frontal attack on all
procurements, but a flanking attack that hits hardest at the pure
commodity and non-commodity flanks of the problem, expanding later to
the mixed middle territory. Here's what needs to be done:
 First, distinguish between commodities and non-commodities.
Commodities are simple, well-known, off-the-shelf goods and services,
readily procured on a transaction-by-transaction basis. Non-
commodities are complex, novel, and/or customized goods and services,
where contextual knowledge gained from one transaction adds
substantial value to subsequent transactions, as in complex systems
development services. While these distinctions are sometimes not
clear-cut, they can and should be used to distribute procurements in
rough order from most to least commodity-like.
 Next, attack the commodity flank with electronic commerce. Especially
at the easy-to-measure commodity end of the spectrum, use computer
networks rather than paper for procurement communications. Such
networks will, among other things, make market pressures more visible
and effective; encourage just-in-time delivery relationships (saving
the government on warehousing costs); and supply quicker and better
data for procurement management. Electronic commerce is widely
supported in the vendor community, as shown by the success of such
recent government initiatives as Oregon's widely replicated online
Vendor Information Program.
 Next, attack the non-commodity flank with more empowering and
innovations-oriented procurement procedures. Give front-line managers
more discretion, then hold them accountable for results--just as
private businesses do.
 Unfortunately, while a strong case can be made for giving managers
more discretion, it faces tough resistance in today's political
environment. Americans support trust and empowerment, but not for
government.
 Because empowerment is so controversial, it should be expanded
initially on the extreme non-commodity end of the spectrum. Here is
where the benefits and uncertainties of technology use are largest and
the need for managerial empowerment strongest. For example, a simpler
procurement process could be authorized for a pre-specified list of
technological innovations, perhaps limited at first to 5 percent of an
agency's total procurements. The new process might--like that used for
procurements in Sweden--require only that product and vendor
selections be made by a selection committee and explained in writing.
Other simplified procedures are in place in the city of Albuquerque
and under consideration for "advanced technology" procurements in a
variety of states.
 Finally, work from the flanks toward the middle. For better
procurements, the primary attacks should be targeted at the commodity
and non-commodity flanks of the problem most suitable for electronic
commerce and technological innovation. As these approaches prove
themselves, the attacks can expand more aggressively into the middle
territory of mixed commoditites and non-commodities.
 In the information age, procurement reform is not merely a nagging
problem. It's a necessary and strategic opportunity for better
government.

----------------------------------------------------------------------

Copyright 1994, Congressional Quarterly, Inc. Reproduction in any form
without the written permission of the publisher is prohibited.
Governing, City & State and Governing.com are registered trademarks of
Congressional Quarterly, Inc.
http://governing.com