Back in the very late 90's, something occured that I still have nightmares about. I was payroll clerk for the smallest Kmart store in Australia. Smallest in floor area, and smallest in staff numbers. Mine was a part time job then, ensuring that "I paid people right." Calculating sick leave entitlements correctly, copping the flack when someone who switched from 12hrs/week to full time (38hrs/week), just after their anniversary, got paid only 12hrs a week annual leave because that's the rate they accrued it at, and having someone angrily knock on my window to loudly complain that I "ripped them off that payrun" - only to discover that their car payment came out the same day, and they'd only done a bank balance check, or they'd serially failed to clock off or on - on days they'd actually worked extra hours- and their manager had filled in the daily "odd clockings" report that they'd just worked to roster. Or they'd taken three days off sick, but had used up their sick pay accrual months ago.
It was a tough, thankless, and often very intensely anxiety-inducing job. But I was good at it. I still have awards from Kmart Payroll services, that weren't presented in staff meetings, but slipped through the little sliding window between the manager's office and my little shoebox of an office (that I thankfully shared with a great person), saying I was consistently (based on the few number of times I had to request payment adjustments, and how well I did with external audits) one of the most accurate payroll clerks in the nation. But within-store, it was quite a thankless task - mostly only having interactions with my fellow workers when, with a sudden loud tap on my window, I'd get a scathing "you didn't pay me right!" (On more than one occasion I had to politely ask my fellow office worker to leave the room whilst I delicately had to explain that either the debt collectors, or the department in charge of Child Support, should have informed them that their payroll clerk would be receiving a court enforced garnishee order (that they should have also received a copy of) to recoup payments directly from their wages!)
So, yes. Thankless for the most part - even when, in the background I made enemies of managers when I pulled THEM up for trying to pull shifts and practices against the Award/Enterprise Agreement they were required to follow - which is why I got little thanks from certain management too!
But I digress. I'm avoiding the nightmares of the late 90s. The nightmares over and above the ones I've just described. The horror of Y2K. I have shared this with only a few people, but it is, I think, timely that I speak up now.
Our store, and every single other Kmart store in Australia, had a problem looming as the year 2000 approached. I am going to only focus on my little part in dealing with my little segment of the big Kmart machine, but keep in mind that my story would be similar in the hundreds of Kmart stores in Australia and New Zealand.
Obviously, in payroll, you need to keep track of when people worked; days and times. So clocks play a surprisingly crucial part in the systems I used as lowly payroll clerk, but the ramifications in my job if those clocks suddenly "broke" would have far reaching ramifications on staff who expected their anniversary dates to register that they now have annual and sick leave again. That the fact they worked "yesterday" is still recorded, because it - the system - knows what day it was then, and what day and time it is now. That "clock" - which was actually numerous clocks; in the actual clocking-on clock near the staff entry door, in the computer that I had on my desk that would grab those clockings into ancient (but at that point still perfectly functional) software, which relied on its own clock to be synchronised and correct so that at the end of the pay week I could ensure that absolutely every staff members' (we had about 150 people on the books then) electronic timesheet was accurate, before having that piece of ancient (yet functional) software transfer all hours worked to another piece of ancient (yet functional) piece of software. That software allowed things like annual leave and the like to be entered (based on its own time-and-date-crucial accrual calculations and birthday/payrate adjustments for its own list of 150 odd staff). Again, this piece of software totally relied on an internal clock within the computer it was running on, to be right. Once all the additional payments had been entered into this ancient (yet functional) piece of software, I'd - along with all the other payroll clerks in all the stores - "press the button" to send the payroll file off to the Payroll department's ancient (yet functional) systems, to send everyone's weekly payments into their various bank accounts; into banks that were all running ancient (yet functional) systems themselves.
But the clock was ticking.
All the clocks that I - and the software I had direct investment and reliance in - were heading towards a day where the information they gave was not going to be correct. They essentially knew the date in the form of "ddmmyy" which worked fine in the sixties, but as the year 2000 loomed ever closer, the very real FACT (not possibility) was my systems were suddenly going to see days marked as 120100 as Friday, the 12th of January, 1900 rather than Wednesday, the 12th of January, 2000!
The Sign In/Sign Out clock had no software option to fix it. On the stroke of midnight, Friday, December 31, '99, it would tick over to Monday, January 1, '00 - the Year of Our Lord, Nineteen Hundred AD. Even if we were to tell staff "just ignore the Day it says it is", we'd hit more problems when it hit midnight on the 28th of February (because 1900 was NOT a leap year, but 2000 was); it would skip the 29th entirely! So, really, just with the problems of the clock not being able to be software-fixed for it to deal with it being 2000 (and four digit years rather than two) meant it absolutely needed replacing. Along with every other store's clock-in/out clock. At between $600-$1000 a pop, plus labour for installation including new cabling between clock and new payroll computer. The old one remained in place and continued to be used until we officially changed over.
The payroll computer also had a chip that tracked just the last two digits of the year. The possibility of rewriting the two pieces of (long unsupported and woefully inflexible) software to somehow deal with it was unrealistic, compared to purchasing far more modern, (we're talking "able to use a mouse with"!) off-the-shelf time and attendance software, and Payroll Software. This software would require a Payroll Computer to have a time chip with four digit years also. So, at huge, but if everyone wanted to "be paid right", necessary expense new systems and software were ordered, whilst other systems in the company and store were also expensively upgraded due to the determination that time/date critical functions WOULD (not "might") be impacted (right down to having the right purchase date on your receipt, and the manager being able to use a spreadsheet to do budget forecasts into the '00s!)
Whilst continuing to perform my usual duties on the old system through '99 (the year my first child was born), I was required to attend frequent external extensive training on the new systems. Head office internally produced software to migrate "some" of the personal data from both old software systems to the two new ones. However, there was still enormous amounts of manual data entry required, including all full and part-time staff rosters, people's pay grades, addresses and email and phone numbers that could be automatically migrated into the time and attendance, and finally doing a number of weeks where we were required to essentially do two parallel pay runs on the old and new systems to compare the two results, and assist head office in ironing out the causes of any discrepancies. It was grueling, total-anxiety-elevation-inducing, and it quickly became apparent the pitfalls of trying to save a buck on the tender for the new software! But we (and I say "we", because the payroll clerks throughout the state really teamed together, often with little support other than "are we under control and on schedule?" from direct management, and helped each other) finally switched from old to new systems a month or so before crunch day - but continued running and entering weekly data into both systems- with staff blissfully unaware that their most recent pay was produced using 21st century technology rather than tech from the eighties!
Even though, after all this vital trouble, effort, anxiety, and expense, I was confident (now) that payroll, at least in our store, would appear to survive the "dreaded threat of the Y2K bug", we were still required to stockpile enough cash in the vault, for that payrun and the next, to print the old fashioned pay envelopes we used to have to use pre pay-directly-to-bank-account, so that in the event the banks had glitches, we could still pay by cash. We didn't have to resort to that two-hour-plus procedure, fortunately!
It was interesting to see how well the old system coped when the day of doom ticked over: it didn't. If all that was done to get payroll to not skip a beat on Y2K-D-Day, hadn't been done, no-one would have been "paid right". I played a bit with date functions on the old machine's spreadsheet software. It couldn't tell what day it was to save itself! The exerted efforts and frustrations of underpaid, overworked, dedicated payroll clerks meant nothing "disasterous" seemed to happen when the big day finally arrived. But certainly, at least at a store level, there was no "Phew! Well, that all paid off! Well done everyone!" No. Instead there was a chorus of "Well, what a giant fuss about nothing! What a waste of money. What a waste of extra wages that could have been spent elsewhere (or not at all)!" In the staff room, as people watched the news announcing what a total non-event Y2K was, staff were heard to comment "It was just a ploy of the computer companies to get huge new sales! It certainly got Colin a new machine to play with! Why can't we go back to the old clock-on machine, now all this has blown over; it was easier to use!"
Y2K went down in history as a Non-Event, but as a lowly payroll clerk, I know the blood, sweat and tears that went on in the background in my department to make the OUTCOME appear a non-event.
Now, I'm no nuclear power station manager, or airline company operator, or bank interest-rate calculator, or rocket launch scientist, or any kind of life-date-critical-computer-controlled-system-type-person (although, thinking back at certain staff I had the joy of "not paying right", my life would certainly have felt in danger if my own massive, unseen undertaking hadn't been undertaken!) I can be VERY certain that they, too, conducted far more massive expensive undertakings to ensure that Y2K appeared a "non event" WHEN the Doomsday Clock ticked over. And much like myself, but to a far greater extent, they must have been flummoxed by the reaction to their unrecognised efforts and fumed at the comments of "Well, what a giant fuss about nothing! What a waste of money. What a waste of extra wages that could have been spent elsewhere (or not at all)!"
Why do I write all this, twenty years after the fact? The events and reactions playing out today regarding COVID-19 have caused this startling realisation of similar and frightening public reaction, where the greater the effort to contain potential disaster, the more successful that containment becomes, which bizarrely creates a reaction of "Well, that was a lot of fuss over nothing, because the disaster was far less than originally predicted!" It seems as if (and this is even when the actions to mitigate disaster are far more publicly visible than Y2K) there is an inability to make a correlation between actions and results. There are complaints that all the forecasted figures haven't eventuated, thus the experts were wrong - COMPLETELY failing to recognise the actions made to stop those figures eventuating! As the Australian comedian Charlie Pickering put it, it's like saying "because we put up pool fences and no one drowned, we wasted our money on pool fences."
The comparison with Y2K only goes so far though. Efforts, recognised or not, were effective, and the public can move on whether they believe any effort happened or was effective, or was worth it without giving it another thought - at least until December 31st, 9999! The problem with saying "The death toll from COVID-19 isn't what they predicted, so we're fussing (lockdowning, iso-ing, boarder-closing, whatever) over nothing!" is that it's that very fussing that has made it appear as if it's a fuss over nothing. Unlike the Y2K bug, this ain't fixed yet, no matter how successful we've been in stamping it out. We can't afford to be complacent yet. We can't.