[Congressional Record Volume 142, Number 120 (Thursday, September 5, 1996)]
[Senate]
[Page S9960]
From the Congressional Record Online through the Government Publishing Office [www.gpo.gov]
THE YEAR 2000 COMPUTER PROBLEM
Mr. MOYNIHAN. Mr. President, on the 31st of July, I took the liberty
of writing to the President concerning a problem that could have
extreme negative economic consequences in the year 2000 when we will
have to make the transition of computers from the 20th to the 21st
century.
This is a matter that will necessarily concern the Congress. I ask
unanimous consent that my letter to the President and a summary of an
accompanying report by Richard M. Nunno be printed in the Record at
this point. Cost considerations prevent having the entire report
printed in the Record. The report can be obtained from the
Congressional Research Service.
There being no objection, the material was ordered to be printed in
the Record, as follows:
U.S. Senate,
Washington, D.C., July 31, 1996.
The President,
The White House,
Washington, DC.
Dear Mr. President: I hope this letter reaches you.
I write to alert you to a problem which could have extreme
negative economic consequences during your second term. The
``Year 2000 Time Bomb.'' This has to do with the transition
of computer programs from the 20th to the 21st century.
The main computer languages from the '50s and '60s such as
COBOL, Fortran, and Assembler were designed to minimize
consumption of computer memory by employing date fields
providing for only six digits. The date of this letter in
``computerese,'' for example, is 96-07-31. The century
designation ``19'' is assumed.
The problem is that many computer programs will read
January 1, 2000 as January 1, 1900. Computer programs will
not recognize the 21st century without a massive rewriting of
computer codes.
I first learned of all this in February and requested a
study by the Congressional Research Service. The study, just
now completed, substantiates the worst fears of the
doomsayers. (A copy of the CRS study is attached.) The Year
2000 problem (``Y2K'') is worldwide. Each line of computer
code needs to be analyzed and either passed on or be
rewritten. The banking system is particularly vulnerable. A
money center bank may have 500 million lines of code to be
revised at a cost of $1 per line. That's a $500 million
problem. (I learn from Lanny Davis that his client, the Mars
Company, estimates the cost of becoming Y2K date compliant at
$100 million to $200 million. Mars is only a candy company.)
One would expect that a quick fix of the problem would have
been found but it hasn't happened and the experts tell me it
is not likely.
There are three issues. First, the cost of reviewing and
rewriting codes for Federal and state governments which will
range in the billions of dollars over the next three years.
Second, the question of whether there is time enough to get
the job done and, if not, what sort of triage we may need. I
am particularly concerned about the IRS and Social Security
in this respect. Third, the question of what happens to the
economy if the problem is not resolved by mid-1999? Are
corporations and consumers not likely to withhold spending
decisions and possibly even withdraw funds from banks if they
fear the economy is facing chaos?
I have a recommendation. A Presidential aide should be
appointed to take responsibility for assuring that all
Federal agencies including the military be Y2K date compliant
by January 1, 1999 and that all commercial and industrial
firms doing business with the Federal government also be
compliant by that date. I am advised that the Pentagon is
further ahead on the curve here than any of the Federal
agencies. You may wish to turn to the military to take
command of dealing with the problem.
The computer has been a blessing; if we don't act quickly,
however, it could become the curse of the age.
Respectfully,
Daniel Patrick Moynihan.
____
The Year 2000 Computer Challenge
(By Richard M. Nunno)
summary
Most computer systems in use today can only record dates in
a two-digit format for the year. Under this system, computers
will fail to operate properly when years after 1999 are used,
because the year 2000 is indistinguishable from 1900. This
problem could have a serious impact on a wide range of
activities that use computers. Information systems must be
inspected, and modified, if necessary, before January 1, 2000
to avoid major system malfunctions.
Many managers initially doubted the seriousness of this
problem, assuming that an easy technical fix would be
developed. Several independent research firms, however, have
refuted this view, with the conclusion that inspecting all
computer systems and converting date fields where necessary
and then testing modified software will be a very time-
consuming and costly task. Research firms predict that due to
a lack of time and resources, the majority of U.S. businesses
and government agencies will likely not fix all of their
computer systems by the start of the new millennium.
Most agencies and businesses have come to understand the
difficulties involved, although some have not yet started
implementing changes. Several companies have emerged offering
services to work on the year-2000 conversion, and software
analysis products are commercially available to assist with
finding and converting flawed software code. Even with the
assistance of these products, however, most of the work will
still have to be done by humans.
Federal agencies are generally aware of the year-2000
challenge and most are working to correct it. Agencies that
manage vast databases, conduct massive monetary transactions,
or interact extensively with other computer systems, face the
greatest challenge. An interagency committee has been
established to raise awareness of the year-2000 challenge and
facilitate federal efforts at solving it. The interagency
committee has initiated several actions, such as requiring
vendor software listed in future federal procurement
schedules to be year-2000 compliant and specifying four-digit
year fields for federal computers. The shortage of time to
complete year-2000 computer changes may force agencies to
prioritize their systems. Agencies may also need to shift
resources from other projects to work on year-2000 efforts.
State and local governments, as well as foreign
organizations, will also have significant year-2000
conversion problems.
Congressional hearings have been held recently to
investigate the year-2000 challenge, and a legislative
provision was introduced directing the Defense Department to
assess the risk to its systems resulting from it. Several
options exist for congressional consideration. One option is
to provide special funding to federal agencies for year-2000
conversion. While agencies are reluctant to request
additional funds, some observers contend this may be
necessary. Another option is to give agencies increased
autonomy in reprogramming appropriated funds for year-2000
efforts. A third, less controversial alternative is to
continue to raise public awareness through hearings and by
overseeing federal efforts.
____________________