[1003] in Public-Access_Computer_Systems_Forum
GPO GATEWAY/WINDO
daemon@ATHENA.MIT.EDU (James P Love)
Mon Aug 17 09:25:26 1992
Date: Mon, 17 Aug 1992 08:19:02 CDT
Reply-To: Public-Access Computer Systems Forum <PACS-L%UHUPVM1.BITNET@ricevm1.rice.edu>
From: James P Love <LOVE%PUCC.BITNET@ricevm1.rice.edu>
To: Multiple recipients of list PACS-L <PACS-L%UHUPVM1.BITNET@ricevm1.rice.edu>
----------------------------Original message----------------------------
The following is the Taxpayer Assets Project (TAP) statement
submitted for the record for the July 23, 1992 (joint Senate
Rules and House Administration) hearing on the GPO Gateway/WINDO
legislation (S 2813/HR 2772).
The second half of the TAP statement addresses concerns about the
public cost of providing free Gateway/WINDO access to the 1,400
member federal Depository Library Program. These comments are
largely a response to OMB's July 29, 1992 statement opposing the
Gateway/WINDO.
OMB's opposition to the legislation was not unexpected. (OMB has
worked for years to prevent agencies from providing value added
information products and services to the public. Of course, the
prime sponsor of the Senate bill, Senator Gore, is currently the
democratic nominee to replace Dan Qualye, and this may also be
relevant.)
OMB's reservations aside, the Gateway/WINDO received good support
at the July 23, 1992 hearings, including an enthusiastic
endorsement from presidential appointee Robert Houk, the GPO
Public Printer.
Congress failed to schedule a markup on either bill before the
August recess. Senator Ford, however, has indicated that he may
schedule a markup for September. Obviously time is running out,
but the bills are still alive.
jamie
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Comments on bills that would provide
public online access to taxpayer
funded databases and information systems
--------------------------------------------------------------
S. 2813: GPO Gateway to Government
HR 2772: GPO Wide Information Network for Data Online (WINDO)
--------------------------------------------------------------
Submitted to the Senate Committee on
Administration and Rules
and the
House Committee on Administration
James Love, Director
Taxpayer Assets Project
P.O. Box 19367, Washington, DC 20036
voice 202/387-8030; fax 202/234-5176
internet tap@essential.org
August 5, 1992
The Taxpayer Assets Project is pleased to offer comments in
support of S. 2813, the GPO Gateway to Government, and H.R. 2772,
the GPO Wide Information Network for Data Online. These bills
would revolutionize public access to publicly funded and owned
information resources. The Gateway/WINDO proposals would use
modern computer technology to allow citizens from every community
in America to obtain timely and low cost information about nearly
every aspect of federal government activities.
Senator Gore, Senator Ford, Representative Rose, and the
members and staffs of the Senate Committee on Administration and
Rules, the House Committee on Administration and the Joint
Committee on Printing are to be commended for their diligence and
thorough understanding of the issues relating to the
dissemination of information in computer formats, and their fine
work in drafting the legislation.
Our organization was invited to testify before the Joint
Committee on Printing on April 25, 1991, and we are pleased that
the proposed legislation addresses many of the concerns that were
raised in those hearings. The Gateway/WINDO proposals would make
it easier for citizens to locate, purchase, and use government
information. The bills would ensure that the important role of
the federal depository libraries in providing universal access to
government information would not be eroded by the introduction of
new technologies. Finally, the bills recognize that the prices
for publicly owned information resources should be no more than
the costs of the dissemination of the information, a principle
that is increasingly being challenged by those who do not
appreciate the importance of a well informed citizenry to our
economy and our democratic institutions.
The Gateway/WINDO proposals present a practical program for
achieving ambitious goals. The legislation allows GPO to start
small, undertaking projects that are within its immediate
competence to carry out, while building the expertise and
experience to undertake more sophisticated tasks. It commits the
agency to a public process, receiving comments from agencies,
data users and other affected parties every year, on topics
ranging from the Gateway/WINDO product line to the many issues
that will be addressed on topics such as standards. The
legislation is the start of an open process that will allow all
affected parties to work together to provide better service to
the public, who, after all, is paying billions of dollars to
create these government information resources in the first place.
The potential government savings from the Gateway/WINDO are
large. Many agencies are finding it far cheaper to disseminate
press releases, bid notices, and other important public
information electronically. The existence of centralized low
cost online access to government information will make these
options more attractive to agencies, savings millions of dollars.
Government agencies are also large consumers of government
information, often from other agencies, and the Gateway/WINDO
legislation will provide a long overdue mechanism to avoid cases
where one agency has to buy access to another agency's data from
commercial data vendors.
The Gateway/WINDO legislation sets out a two track
dissemination program which parallels the current systems used by
the GPO for paper publications. The sales program will charge
fees for the service, while free access will be available through
the 1,400 member federal depository library program. We believe
the sales program will generate substantial revenues, easily
defraying the government's costs of providing public access to
these databases. Consider the types of database systems that the
government owns which are now closed to the public. The SEC's
EDGAR system, the Congressional LEGIS systems, the Patent and
Trademark Office's $300 million Automated Patent System, the
Department of Justice's JURIS system, the Department of
Commerce's National Trade Databank, the CIA's Foreign Broadcast
Information System, and other government funded information
systems are valuable national treasures that should be available
to citizens who are willing to pay the costs of acccess.
The economics of providing public access to government
databases are very attractive. In those cases where the
government has already designed an online information system, the
largest expenses have already been incurred, and the incremental
costs of adding additional users is often a trivial expense. For
example, according to a 1989 SEC study, it would cost less than
$5 million in hardware and software to provide more than 2.2
million hours of prime time public access to the SEC's new EDGAR
system. In addition to the one time cost of about $2.27 per hour
of capacity, the SEC estimated operating costs of less than 30
cents per hour. Compare these figures to the hundreds of dollars
per hour that commercial vendors charge for online access to SEC
information.
The only difficult economic question will be the financing
of the free access by the federal Depository Libraries Program
(DLP). Our own investigations suggest that the data processing
costs for free access by the DLP will be a small number. For
example, the Department of Commerce's Economic Bulletin Board
(EBB) has a budget of about $150,000 per year. However,
practically none of this is sensitive to usage levels. The one
time capital costs for the computer, software and
telecommunications equipment were less than $20,000. The major
expenditures are for the staff which maintains the database. The
cost of additional users to the EBB is close to zero, since the
computer and telecommunications inputs are so inexpensive, when
compared to the fixed costs of maintaining the data.
The situation with the EBB is typical. Once the GPO or the
agency provides a system that can serve the sales program, it
will cost next to nothing to allow the DLP to "ride" the program.
Incidently, this will be true for agency support staff as well.
Indeed, the DLP will likely reduce agency costs for support
staff, since librarians will quickly become highly trained users
who can train their own patrons. For a good example of this,
look what law school librarians have done for Lexis and Westlaw,
private companies that sell access to legal information. The
librarians act as free resources for the vendors, training the
users. The DLP will do the same thing for the Gateway/WINDO,
resulting in lower agency costs for online support services.
The only signficant cost for the DLP will be the
telecommunications charges. If GPO or the agency pay the DLP
telecommunications charges, just as GPO has historically paid for
the postage for paper products sent to the DLP, there will be
significant costs to finance. Congress, of course, does not have
to provide unlimited telecommunications support, despite OMB's
ill informed assertion to the contrary. It can pick a number it
can afford, and allow the DLP libraries to spend their own money
if they want expanded access. This is what Congress should do,
if the funding of the DLP telecommunications charges becomes an
issue.
But, the telecommunications cost issue is really only a
short term problem. Senator Gore has wisely required GPO to use
NREN and other computer networks to disseminate the Gateway/WINDO
services. Thus, once the members of the DLP obtain NREN access,
which many already have, they will use an existing program which
is designed for precisely this type of task.
Finally, it is necessary to point out some of the more
egregious flaws in OMB's poorly researched July 29, 1992
statement. In particular, OMB has referred to three poorly
researched estimates of the cost of DLP access. These three
reports are included in the printed record of 1989 hearings held
by the House Subcommittee on Government Information. While these
estimates have become part of the folklore about the economics of
online services, they are deeply flawed. For example, consider
the EDGAR study, which was done by the private contractor. It
assumes the government would have to buy 2,000 terminals for the
libraries at a cost of $5,800 per terminal. Not only is the
estimated cost of a terminal ten times higher than it need be,
but even more important, the libraries themselves have always
bought their own computer equipment, just as they provide book
shelves for paper products and microfiche readers for microfiche.
The estimate also included an unnecessary $1.44 million for the
maintenance of the libraries terminals, plus huge fees to the
contractors. BDM wanted a fee of $4.6 million to supervise the
set up of the DLP access program, and it also budgeted an
additional fee of $7 million for Mead Data Central, who is
already receiving more than $13.5 million from the SEC. And on
top of everything else, it budgeted huge telecommunications
charges, based upon nearly constant usage of the EDGAR system
from each library. Of the $87.8 million the contract said it
would cost to provide access to the DLP over a five year period,
less than $5 million were related to the costs of the computer
and telecommunications hardware. And, by 1992 the estimates for
these costs have surly fallen dramatically.
The 1989 estimates for the PTO access and the Library of
Congress (LC) access are also deeply flawed. For example, most
of the LC fees are related to support, and as we have indicated,
it is likely that the DLP will actually lower agency support
costs, rather than add to them. And, the PTO case study focused
on the high technology image based version of its database, which
requires expensive work stations and high speed data connections.
But, if the library buys its own workstations and uses the NREN,
these costs disappear from the analysis. Moreover, many citizens
simply want access to the PTO's text files, which are far cheaper
and easier to access.
There are also many other flaws in the OMB statement.
Indeed, it is curious that an agency with responsibility for the
government's budget didn't notice any of the cost saving
opportunities that the Gateway/WINDO will present. Perhaps OMB
could be asked to present the Congress with its own estimates of
the amount of money federal agencies spend every year to buy back
government data from private vendors. Maybe they could start
with the federal government's purchases of Landsat data from
EOSAT, the GM/GE joint venture that did a brisk business selling
the government its own Landsat data during the Gulf War.
-------------------------------------------------------------------
James Love, Director VOICE: 609-683-0534
Taxpayer Assets Project FAX: 202-234-5176
P.O. Box 19367 INTERNET: love@essential.org
Washington, DC 20036
