This is my last update of my original Y2K essay written in May 1999 dealing with the Y2K computer bug. That essay provides essential background for this update and should be read first. My September 1999 update and December 1999 update are also background for this essay.
After months of reading nearly a dozen books as well as countless articles, websites and newspaper stories -- after months of straining to know the truth about the Y2K computer bug with as much objectivity as I could muster -- after months of hands-on Y2K remediation at my workplace -- and after months of sounding the alarm far & wide -- I must acknowledge that I was SPECTACULARLY WRONG.
I am delighted & relieved that there was no Y2K computer bug disaster. The result of such a disaster could have been devastating to life, liberty, wealth & progress -- as well as to life-extension and cryonics. At the same time I am distraught that I could have sought the truth so aggressively and been so wrong. I am disturbed at the prospect that I may have upset people -- causing them to spend time, money & effort preparing for an hallucination. And I am shamefaced at having discredited myself to those who trusted & believed me. I had felt at times like Paul Revere and at times like Chicken Little. Now I am Chicken Little with egg on his face.
But life goes on -- especially for those who ardently want life to go on -- and on and on... I groan, "Oh No, Not Another Learning Experience" (to quote a button). And then, hopefully, do learn to be smarter about distinguishing real problems from illusory ones -- and about making the right preparations for the problems that are real.
Although not widely publicized, there were quite a number of Y2K computer problems, but most have been glitches that have been quietly fixed without public fanfare -- much like the computer glitches that occur chronically in everyday life.
There were problems with heart monitoring equipment and defibrillators reported in Sweden & Malaysia. E-mail systems failed in Qinghai branches of the People's Bank of China and in Russia's press service. Machines processing credit-card transactions in many Chinese banks failed on January 1st.
The first baby born in Denmark in January was registered as being 100 years old. The IRS sent demands for payment by 1900 to many taxpayers. Ten percent of cash registers in Greece printed receipts with the year 1900.
Computer controls on prison cell doors failed in British Columbia. Computerized prison records in Italy gave incorrect dates for birthdays, trial-dates and release-dates.
Highland Community Bank in Chicago was unable to electronically transfer Medicare funds. The Federal Reserve Bank in Chicago reported a Y2K failure associated with the transfer of $700,000 in tax payments. Three mission-critical systems failed at the Federal Housing Administration. 100,000 people in Sweden were unable to access their bank accounts over the Internet.
Emergency phones on the Adirondack Northway in New York went dead because they weren't Y2K compliant. Cash register/inventory systems were so malfunctional at many Washington State liquor stores that some stores were forced to temporarily close. Hundreds of slot machines failed at racetracks in Delaware.
Y2K computer problems at the Hong Kong Futures Exchange forced manual compiling of options prices, whereas more serious problems forced Pakistan's stock exchange to close on January 4th.
A hydroelectric plant in Kazakhstan has been forced into manual control due to Y2K problems encountered on January 1st. Manual operations are also in place for an income tax system in Gambia which was not Y2K compliant. Y2K bugs affected aluminum manufacturing in Korea & Venezuela.
Problems described as somewhat serious were failures to process data from US miliary reconnaissance satellites and a problem at the main US uranium storage site for nuclear missiles. Both problems occurred at midnight GMT and both problems were dealt-with within four hours -- although the satellite photos for the few hours after midnight were irretrievably lost.
A survey of 1,750 technology workers by CMP Media (a publisher of high-tech trade publications) revealed that 25% of organizations experienced Y2K computer problems, more than half of which were serious enough to cause a brief interruption of service. Peculiarly, hotels & restaurants have reported the largest adverse effects of any industry.
Problems in the nuclear power industry are difficult to cover up because the reporting requirements are very stringent. In the United States, only one nuclear plant was shut down due to a possible Y2K-related incident. Seven Y2K-related non-critical nuclear power plant incidents were reported in total for the US. Japan reported 5-10 minor nuclear power Y2K problems, such as failing radiation monitors and temperature gauges. One similar problem was reported in Korea and for two of Spain's nine nuclear reactors. But in the countries where the worst nuclear power problems were feared -- Russia, Ukraine and elsewhere in Eastern Europe -- there were few reported problems. Prior to rollover officials in those countries had anticipated no problems because their equipment was too old to contain the kind of computer-controlled automation that could lead to disaster. In retrospect, these claims appear to have been reasonable.
Morgan Stanley Dean Witter quietly settled out of court when faced with a lawsuit from Sunquest Information Systems for helping software maker Compucare sell a Y2K-bug ridden database division to Sunquest.
I myself dealt with three Y2K-associated problems in the first week after the Y2K rollover. One was due to a user running a job too early in order to preclude a Y2K problem, one was a true Y2K glitch that was nothing more than an inconvenience and one is a puzzle that has neither recurred nor been explained. Three weeks later another problem appeared when I discovered we had been failing to get some data because of a Y2K bug (which was easily fixed). The bugs which I had found & eliminated early in 1999 in the Y2K lab would have been much more serious.
Reviews of many other worldwide Y2K glitches can be found at "www.yfiles.com/Y2K.html" and "www.iy2kcc.org/Glitches2000.htm"
There was no shortage of compelling evidence that disaster might occur, much of which I have previously reported in depth -- and will not attempt to repeat. Last Spring MCLEAN'S magazine (19-April-1999, page 44) published a Gartner Group world map which showed best & worst prepared countries as having 10-15% to over 50% (respectively) companies & government agencies experiencing critical Y2K system failures. The May 1999 issue of CONSUMER REPORTS (page 23-27) gave warnings about possible failures of power, water, phones, cash and food-supply.
The 29-November-1999 issue of TIME magazine (page 49, Canadian Edition) repeated warnings from the CIA concerning Russia, Ukraine, Egypt, India and Indonesia. The US State Department had not only issued travel warnings about China and other countries, but had authorized embassy employees to leave Russia, Belarus, Moldova and Ukraine due to concerns about health & safety associated with "potential disruptions to energy supplies". The State Department did not lift its travel warnings for those four countries until the end of the first week of January.
In December the Japanese people were warned by their own Prime Minister to stock-up on food, water, flashlights, etc. -- just in case. The national Y2K co-ordinator of Paraguay had predicted widespread power outages, water shutdowns and phone disruptions. Nigeria officials warned that up to 80% of the country's computers and most of its banks were not ready for Y2K. Lawrence Gershwin, the top US intelligence officer for science & technology stated that the intelligence community expected there would be "calls on the U.S. military to intervene in humanitarian crises".
According to one survey, 40% of all companies postponed purchasing some computer equipment until after Y2K. IBM attributed an 11% drop in fourth quarter income to a Y2K freeze on purchasing mainframes.
Del Monte Foods reported a 3% increase in sale of canned goods, and Ely Lilly reported $90 million in pharmaceutical sales associated with stockpiling. Campbell Soup company reported a 12% increase in soup shipments in the second quarter of 1999 -- the first shipment increase in five quarters -- attributed to Y2K worries. Peanut butter sales were up 2-3%, although with this and many other products, it is hard to prove that Y2K stockpiling was the explanation.
At Peterson Air Force Base in Colorado, top Russian & American military leaders established the Center for Year 2000 Strategic Stability to ensure that no accidental missile launchings occurred on the rollover. The military officials assured the public that there was no chance of such an event occurring, and that the Center was just "insurance". But insurance only makes sense against an event that could possibly occur.
British Telecomm (BT) had over 5,000 staff working on New Year's Eve. Deutsche Telekom, the largest telecom company in Europe had 7,500 staff working that evening -- directed from an underground command centre in Bonn. KLM Dutch airline grounded all but 3 of its 120 aircraft and had 250 Y2K staff on alert. By contrast, many Asian countries did nothing -- and reported no problems.
The city government of Toronto, Canada (where I live) employed 450 people to remediate computer systems against Y2K, and 300 people were working on New Year's Eve to monitor the systems. At the bank where I work, approximately a third of the computer staff -- including me -- were told to be at work on New Year's Eve or New Year's Day, and most of the rest of the staff was expected to be on-call. The freeze on release of new applications extended to February 15th.
Among the general public, however, the much-feared possibility of panic never materialized. The stock market continued to climb until the last trading day in December, although the trading volumes dropped dramatically as December 31st approached. Citibank officials stated that cash withdrawals did not exceed those that would be expected for a long holiday weekend. Instead, the media went apoplectic about the possibility of terrorist attacks -- which never materialized. Hoax & hype?
The most important issues about the Y2K computer bug -- along with all the other potential problems in life -- were/are: "What is the truth?" and "How can the truth be known?". Concerning the Y2K computer bug, there was an appalling silence on these questions from academic scientists -- to whom we usually look for answers. That in itself raises questions about the way academic science actually works to produce relevant knowledge.
Prior to the Y2K rollover there was a bewildering mass of conflicting evidence over what would happen. After the event it is still difficult to discern what actually did happen, but many false claims have fallen-away and a few gems come shining through. I have not given-up on attempting to evaluate the evidence.
By far, the largest amount of money was spent doing remediation in countries with the greatest amount of modern computerization -- the countries most vulnerable to a problem (implying that much of the computerization in the former Soviet Union was not modern). Nonetheless, several times as much money was spent in Canada for Y2K remediation as was spent in Italy. Italy has nearly twice as many people and an economy nearly twice the size as Canada's, yet Italy has not suffered noticeably more problems. An even greater contrast is seen in the $4 million spent by the Russian military as compared with the $4 billion spent by the American miliary in Y2K preparation. As another example, British Telecommunications PLC and South Korea bought similar telephone hardware in the 1980s. BT spent millions on preparation whereas South Korea spent much less -- and did not suffer (noticeably) more. (But this is only one system -- it would be dangerous to overgeneralize.)
In my mind, the real wild-card of the Y2K computer bug problem lay in embedded systems -- computer chips controlling industrial machinery, infrastructure switches and weapons systems. My efforts to get at the truth about this problem were maddeningly frustrating. Now I can point to two URLs (one post-Y2K) which provide the most sensible explanations about why there was no embedded systems disaster.
As is indicated in an analysis of the problem posted at the National Institute of Standards and Technology website www.nist.gov/y2k/embeddedarticle.htm, the most serious embedded chip problem involves control systems calculating a time interval on two sides of the rollover, eg, 01/01/00 00:05 minus 31/12/99 23:55. If 00 was recognized as 2000 (as on newer chips) the calculation would be correct, but otherwise there could be a problem. There could also be problems where an embedded system passed an incorrect date to another system.
One solution to the interval-calculation problem would be to not have equipment running during the rollover. This was done in the case of ATMs in China, plants manufacturing hazardous chemicals in the US and trains & subways worldwide. Another solution would be to ensure that the most critical systems had newer chips controlling them. Problems would not be anticipated in countries and companies using pre-1990s equipment that did not use microprocessor embedded controlling chips or in post-1996 equipment.
The most insightful analysis of the embedded chip issue can be found at www.albany.net/~dmills/fallback/chapt5.htm. If I had read, understood and believed this piece before the Y2K rollover I would have had a very different perception of possible problems.
In general, process control systems are divided between highly separated regulating devices and protection devices. Regulating devices typically calculate elapsed time by number of ticks (like clock ticks) rather than by date-subtraction calculation. Most power blackouts are due to protection devices which respond to problems such as excessive current -- not to date-sensitive functions.
In response to critics claiming that the billions of dollars spent on Y2K remediation was wasted, the US State Department's Chief Information Officer made the statement: "We should be careful not to confuse the lack of catastrophic disruptions with unnecessary preparations by the federal government." Y2K guru Peter de Jager said, "There was never any doubt, right from the beginning that, if we had any success at all in fixing Y2K, then the critics, true to form and supported with the impunity of 20/20 hindsight, would claim that the problem never existed in the first place."
John Koskinen, Clinton's chief Y2K Officer, gave three reasons why Y2K problems were minimal world-wide: (1) the most extensive date-sensitive computerization was found in the countries & industries that spent the most money on remediation, (2) standardized software was more common than customized software for smaller countries & businesses, and (3) those that started late on remediation were able to benefit with lessons & software from those that started early. Two or three years before the rollover there were fears that elevators might malfunction. But after a year of testing it was established conclusively that elevators were not at risk.
Much has been written about the benefits of massive Y2K remediation work associated with forcing organizations to inventory & upgrade their entire hardware & software systems. But rationalizing the benefits of misperceptions is a poor substitute for accurate perception as the basis of rational control.
People commonly believe only what they want to believe and think-about only what they want to think-about. Many of those who believed that there would be a Y2K computer-bug crisis were those who wanted there to be a crisis, either because they wanted to see damage done to a civilization they hate or because they wanted to believe that they were making preparations that would give them an "edge" over others. Most people did not want to believe that there would be a problem, and often preferred to bury their head in the sand about the matter. If I raised the question of Y2K problems, I would be told "I don't want to hear about it." This approach to dealing with potential problems is bound to lead to disaster for anyone who lives long enough.
Most of us have enough problems & interests so that the intrusion of a new problem into our lives may be greeted with irritation. This was my first response to the Y2K computer bug. But "gloom & doom" is not a criterion for accepting or rejecting truth. People on the Titanic, in Pompeii or in Pol Pot's Cambodia ignored "gloom & doom" warnings at the cost of their lives. If we are not open to believing both the positives & negatives of life, we cannot successfully cope with reality. Too many people refuse to face the reality of death. Too often, people respond to warnings of danger by shooting the messenger. False alarms, of course, are a nuisance.
I have tried to look into my own psychology to find answers as to why I would have so grossly misperceived the Y2K computer bug problem. I don't have to probe too deeply to find evidence of my being an anxiety-ridden worry-wart. I grew up believing that a nuclear war was inevitable and later came to believe that ballooning government debt would lead to depression -- if not economic collapse. As potential disasters came and went uneventfully, I became more skeptical. Claims of environmental or population disasters have not terrified me -- although these problems gnaw on me like a dull ache.
In the face of so much conflicting evidence about the Y2K computer bug, psychological bias has a potential to play a significant role in the way evidence is evaluated. There were many assurances by government and corporate authorities that there would be no problem. The vast majority of the population took these assurances at face-value and believed them. I am distrustful of authority and believed that such assurances were motivated by self-interest and by a fuzzy understanding of technical systems on the part of management.
Nonetheless, I gave too much credence to statements by corporate consultants who had a vested interest in remediating the problem. In the last days of December, the most widely quoted of these consultants -- the Gartner Corporation (with little to gain by further alarmism) -- began saying problems would be minimal. I remained skeptical. I wasn't alone. An estimated 80% of Fortune 500 corporations established command centers to monitor the Y2K rollover. These are not organizations that frivolously waste resources.
There were editorializing journalists who wouldn't know the difference between a computer program and a medical diagnosis who claimed that the Y2K bug was all hoax & hype. Such people gloat that they were "right" -- but how can they be right about something they have no understanding-of, simply on the basis of outcome?
Results are important, but they are not really "the only thing". A person who spends a lot of money gambling is not proven shrewd by the fact that he or she happens to have a big win. The ancient Greek philosopher Democritus is given credit for his claim that matter is composed of small, indivisible "atomos" particles -- in contrast to Plato & Aristotle who said that matter is infinitely divisible. But Democritus had no evidence for his belief, so it is silly to give him credit for something less consistent with experience than divisibility. People who are right for shallow reasons do not deserve more credit than those with deeper understanding who make mistakes. Nonetheless, if understanding does not minimize mistakes, something is being misunderstood.
Some IT (Information Technology) professionals have been feeling like housewives -- whose considerable accomplishments in doing their work went unnoticed because it was so successful. Only if the work had not been done would the severity of the problem have been appreciated. But because there was no serious problem, much of the public believes there was never a problem and that the Y2K computer bug was a hoax.
It is very difficult to establish proof that severe problems were prevented rather than would never have occurred in the first place. Why did not the 30% of small-to-medium size businesses with no preparation for Y2K not suffer more? How could a problem of such magnitude have been so perfectly fixed that there was not a single major disaster somewhere in the world? Technological malfunctions and disasters occur daily in normal life on this planet. The Y2K computer bug ended-up looking like the world's greatest refutation of Murphy's Law. It seemed that so many things could have gone wrong.
It is easy to latch-on to an explanation such as "hoax", "hype", "problem-fixed", etc., but it is not so easy to find an explanation that fits all the facts. I am still left with the disquieting feeling that I cannot understand the Y2K computer bug problem or why events transpired as they did. As I said in my initial Y2K essay (May 1999), "The Y2K problem can be very frustrating for someone in search of hard facts".
Although the Y2K-turnover did not produce an immediate computer-related crisis of civilization, there may have been more long-term effects. First, because of concern about a possible crisis due to psychological factors, the Federal Reserve (U.S. central bank) pumped large amounts of money & credit into the economy and then instituted a sharp reversal in this policy early in the year-2000. This may have contributed to a stock-market/investment bubble along the classic patterns of Austrian Business Cycle Theory. Second, the huge disproportionate spending that went into computers & technology to prevent a possible Y2K-problem -- and the following sharp pull-back in such spending in the year-2000 -- may have had devestating consequences on the economy due to the sudden reallocation of resources. Even the so-called &DOT-COM bubble may have been indirectly fueled by technology spending (it was certainly fueled by the credit-expansion).