99% Invisible - 379- Cautionary Tales

Episode Date: November 20, 2019

Galileo tried to teach us that adding more and more layers to a system intended to avert disaster often makes catastrophe all the more likely. His basic lesson has been ignored in nuclear power plants, financial markets and at the Oscars... all resulting in chaos. At the 2017 Academy Awards, Warren Beatty and Faye Dunaway famously handed the Best Picture Oscar to the wrong movie. In this episode of Cautionary Tales, Tim Harford takes us through all of the poor design choices leading into the infamous La La Land/Moonlight debacle, and how it could have been prevented. Cautionary Tales Subscribe to Cautionary Tales on Apple Podcasts

This is 99% Invisible. I'm Roman Mars. There are lessons to be learned in our failures. Touching a hot stove once will teach you to never touch a hot stove again, but when it comes to the failures of complex systems, it's sometimes hard to know what went wrong and when. This is where Tim Harford, also known as the undercover economist and his brand new show Coshnery Tales, comes in. It's a podcast that brings you stories of awful human error, tragic catastrophes,

daring heists, and hilarious fiascoes. His stories will delight you, and if you're smart, they will scare the bejesus out of you. But they'll also make you wiser. The episode we're hearing today is about a fiasco that is often blamed on bad typography hence my specific interest in it, but the full explanation turned out to be much weirder and more surprising. Here's Tim Hartford with cautionary tales. You've heard this story before.

You might even have watched it happening live. 33 million people did. There they stand, together on stage, fade on away and Warren Beatty, Bonnie and Clyde together again after 50 years. Even for a pair of veteran Hollywood stars, it must have been a nerve-wracking moment. Yet their tasks seemed simple. Open a red envelope, take out a card, and read out the title of the film that had won the Academy Award for Best Picture.

Bati opens the envelope, so far so good. Then he looks at the card in his hand. He hesitates. Then he looks inside the envelope again as though checking to see if there was a cover note. He looks over at Dunnoway, then raises both eyebrows. She reaches over and touches his arm affectionately. He pauses again. He looks at the card again. He slips his hand inside the envelope one more time. The audience laughed

at the moment of maximum tension that old rogue is playing with everyone's emotions. He continues, for a best picture, he stops again. Faye Dunnway thinks he's goofing around too. You're impossible she says, come on. He shows her the card. She doesn't hesitate for a moment. The audience erupts in applause and the producers of La La Lounge take to the stage to give their acceptance speeches. Meanwhile, off in the wings, an accountant named Brian Cen, knows that the biggest screw up in the history of the Academy Awards is in full flow. The 2016 winner of Best Picture isn't La La Land. It's moonlight.

So, yes, you've heard this story. But have you understood what it really means? You're listening to another cautionary tale. There's a simple way to tell the story of this fiasco, as Warren Beatty walked on stage, Brian Cullenen's job was to hand him the envelope for best picture. Instead, Cullenen handed over the envelope for best actress. A few moments earlier, that award had been won by Emma Stone for her performance in La La Land. And so, when Beaty opened the envelope, he saw Emma Stone, La La Land. Jimmy Kimmel, the host of the Oscars that evening, jokingly blamed it all on Bati.

Warren, what did you do? I wanted to tell you what happened. But it really wasn't Bati's fault. He didn't understand what he was looking at, and rather than say the wrong thing, he hesitated. Then he turned to fade underway. She was the person who actually uttered the title of the wrong film, but it wasn't her fault either. Imagine her situation. She's on stage in front of the most star-studded audience imaginable, plus tens of millions watching live on television. Beaty seems to be messing

around, and she doesn't know why. The first thing she sees on the card is La La Land, and straight away that's what she says. Some people blame Beaty, some people blame Dunnoay, most people blame Brian Cullenen, the accountant who handed over the wrong envelope. But all of those people I think are making a mistake. Almost a decade ago, years before this fiasco, I interviewed one of the world's most important thinkers on how accidents happen. He's a sociologist named Charles Perot, and he told me something that stuck in my mind, that we always blame the operator. It's always, he says, a case of pilot error. Charles Perot is a wise old man.

He's older than the Academy Awards themselves. He was four back in 1929 when they were first presented. And Perot is absolutely right. We inevitably look for someone to blame, and that inclination leads us astray. Our instinct is to blame the accountant, Brian Colullinan, he did give Beaty the wrong envelope. He was distracted. He'd tweeted a photograph of Emma Stone holding her Oscar statuette.

At a time he should have been focusing on giving Beaty the right envelope, he was on his phone, enjoying being close to one of the world's most beautiful people at her moment of triumph. We can all learn a lesson from that. Get off your phone, get off your phone when you're supposed to be having dinner with your family, get off your phone when you're supposed to be driving, and get off your phone when you're supposed to be giving the envelope containing the best picture card to Warren Beatty. But the reason that it's a mistake to simply blame Colonyn is because Colonyn was just

being human. Humans are always getting distracted. Humans are always making mistakes. If our systems can't cope with those mistakes, then it's hopeless to demand better humans. We need better systems. When La La land, fleetingly one moonlight oscar, some rich and successful people suffered some embarrassment and some heartache, but nobody died. Yet the lessen of the fiasco is far from trivial. The same kind of mistake in different situations can be catastrophic.

Such mistakes can lead to nuclear accidents and financial meltdowns, and when I say, can lead, I mean, have led such disasters have already occurred. So, let's try to understand what really happened that strange night at the Oscars. Perhaps we can use that insight to prevent far more serious calamities. Let's talk about problem number one, bad typography. I know it sounds strange, but it's true, the card that Bati took out of the envelope had nine words on it, and the largest word is, oscars. Really? Is that really the most important piece of information? Are we worried that without it, Warren Beaty's going to think he's at the Iowa State Fair handing out a medallion for the best hog. The words La La Land and Emma Stone are printed with equal

weight, even though the winner is Emma Stone, the rest is detail. Meanwhile, the important words best actress are tucked away at the bottom of the card, and they're tiny. If best actress had been prominent, Warren Beatty wouldn't have been confused he would have known that he had the wrong card in his hand. And if Emma Stone had been in larger type than La La Land, Faye Dunnoay wouldn't have blurted out the name of the wrong film. It would have been awkward for Beatty and Dunnoay to walk off stage to get the right envelope, but not nearly as awkward as not walking off stage to get the right envelope.

So the Academy should have hired a designer, but they're not the only ones. Imagine that you're the night shift supervisor at a nuclear power plant. It's 4 o'clock in the morning, and you don't know it yet, but the turbine system that draws away the heat from the reactor core has just shut down. You're going to have to make some quick decisions, assuming you can figure out what's going wrong. And why? What's that? Let's see. I think it's a key noise! Can we shut it down?

I think...no. Can you? During the first few minutes of the accident, more than a 100 alarms went off, and there was no system for suppressing the unimportant signals so that operators could concentrate on the significant alarms. Information was not presented in a clear and sufficiently understandable form. That's from the overview of the official inquiry into what was then arguably the world's most serious nuclear accident at 3 Mile Island in 1979.

It destroyed the reactor, came close to a serious release of radioactive material on the eastern seaboard, and shattered the reputation of the American nuclear industry. It's striking how quickly the inquiry focused in on the question of design. But to anyone who studies how accidents happen, it's not surprising at all. The plan's control rooms were so poorly designed that error was inevitable. Don Norman, the director of the design lab at UC San Diego,

was asked to help analyze the problems at Three Mile Island. Design was at fault, not the operators. The control panels were baffling. They displayed almost 750 lights, some next to the relevant switches, or above them, or below them, sometimes nowhere near the relevant switch at all. Red lights indicated open valves or active equipment. Green indicated closed valves or inactive equipment, but since some of the lights were typically red and others were normally green,

the overall effect was dizzying. So, yes, the operators made mistakes at 3 mile island, but with better design, they might not have. Warren Beatty and Faye Dunnoay would sympathise. But something else went wrong that night at the Oscars, something deeper and more surprising. It's a strange effect and it was first observed by a man less famous for looking at why things go wrong, and more famous for looking at the stars. Galileo Galilei is known for his astronomy, and because his work was consigned to the Church's indexly brawrum-prohibitorum, the list of forbidden books. But the Great Man's final work

opens with a less provocative topic, the correct method of storing a stone column on a building site. There with me, this book from 1638 is going to explain the Oscar Fiasco and much more. I must relate to circumstance which is worthy of your attention, as indeed are all events which happen contrary to expectation. Especially when a precautionary measure turns out to be a cause of disaster. A precautionary measure turns out to be a cause of disaster, that's very interesting Galileo. Please go on.

A large marble column was laid out so that its two ends rested each upon a piece of beam. I can picture that in my mind, support the column while it's being stored horizontally, ready for use. If you lay it on the ground, it may get stained, and you'll probably break it when you try to get ropes underneath it to pull it upright. So yes, store it flat, but propped up by a support at one end and a support at the other. But what if the column can't support its own weight like that and simply snaps in half? Galileo has thought of that.

A little later, it occurred to a mechanic that, in order to be doubly sure of its not breaking in the middle, it would be wise to lay a third support midway. This seemed to all an excellent idea. Yes, if two supports are good, surely three supports are better. It was quite the opposite. For not many months passed before the column was found cracked and broken, exactly above the new meter support.

How did that happen? One of the end supports had, after a long while, become decayed and sunken, but the middle one remained hard and strong, thus causing one half of the column to project in the air without any support. The central support didn't make the column safer. It pressed into it like the central pivot of a seasaw snapping it in half. Galileo's tale isn't really about storing columns, and neither is mine. It's about what I'm going to call Galileo's principle. The steps we take to make ourselves safe sometimes lead us into danger. The problem Galileo described is well known to safety engineers. There's an article in the Fine Scholarly Journal process safety progress.

Titled, no good deed goes unpunished. Case studies of incidents and potential incidents caused by protective systems. These case studies are magnificent. There's a chemical plant where two pressure relief systems interact in a way that means neither of them work. There's a flare designed to destroy pollutants, which ends up causing the release of toxic gases. There's an explosion suppression device that causes an explosion. This kind of thing happens more than you might think. An example is one of the earliest nuclear accidents at Fermi-1, an experimental reactor in Michigan.

It's barely remembered now, except in Gil Scott Heron's song, we almost lost Detroit this time. The operators at Fermi-1 lost control of the nuclear reaction, for reasons that were baffling to them. Some of the reactor fuel melted, it was all touch and go. Eventually they got the reactor under control, shut it all down, and waited until it was safe to take it all apart. It was almost a year before the reactor had cooled enough to identify the culprit. A piece of metal, the size of a crushed beer can had blocked the circulation of the coolant in the reactor core.

It was a filter that had been installed at the last moment for safety reasons at the express request of the Nuclear Regulatory Commission. It had come loose and caused the entire problem. Galileo's principle strikes again. Why do safety systems sometimes backfire? The wise old sociologist Charles Perot's most famous book is called Normal Accidents. It's about how certain kinds of system are vulnerable to catastrophic failure. So vulnerable, in fact, that we should view accidents in those systems as inevitable. The vulnerable systems have two features.

The first is that they're what perot calls tightly coupled. In a tightly coupled system, one thing leads to another and another and another. It's like domino toppling, which is actually a great example of a tightly coupled system. Once you start, it's hard domino toppling, which is actually a great example of a tightly coupled system. Once you start, it's hard to hit the pause button. A second feature is complexity. A complex system has elements that interact in unexpected ways. A rain forest is a complex system, so is Harvard University. But Harvard University usually isn't tightly coupled.

If there's a problem, there's also time to find a solution. A system that's both complex and tightly coupled is dangerous. The complexity means there will occasionally be surprises. The tight coupling means that there will be no time to deal with the surprises. The tight coupling means that there will be no time to deal with the surprises. Charles Perot's theory explains Galileo's principle. Every time you add a feature that's designed to prevent a problem, you're adding complexity. The middle support for the column added complexity. So did the safety filter that damaged the Fermi-1 reactor.

Safety systems don't always make us safe. So did the safety filter that damaged the Fermi-1 reactor. Safety systems don't always make us safe. More of Tim Harvard's cautionary tales are 99% invisible after this. Here's the question we should be asking about that bizarre evening at the Oscars. How was it even possible for the distracted accountant Brian Cullenon to give Warren Beatty the wrong envelope? A few minutes earlier, the envelope for best actress, the envelope containing the card that read Emma Stone La La Land, that envelope had been in the hands of Leonardo DiCaprio, as he stood on stage announcing her win.

How could that envelope have made its way into the hands of Warren Beatty? The answer...it didn't. There were two envelopes. Every envelope, for every category, had a duplicate version waiting in the wings. These duplicate envelopes were there, as a safety measure. And that safety measure is what made the fiasco possible. Galileo's principle had bitten hard. Charles Perot's argument is that when systems are both complex and tightly coupled, we should

expect catastrophic accidents. Does the Academy Award ceremony fit that description? It's certainly tightly coupled, you can't easily interrupt a live TV spectacular in front of millions of people to ask for advice. The show must go on. Yet, the ceremony doesn't have to be complex. Giving an envelope to Warren Beatty doesn't have to be complex. But you can make it complex, if you try.

Brian Cullenen's partner in crime that evening was Martha Ruiz. Like Cullenen, she was a senior accountant. The pair of them carried identical briefcases with an identical set of envelopes. On the day of the show, we'll get the ballots and Brian and I will go to the theatre on two separate roads. He'll go one route and I'll go another route. That's how Martha Ruiz helpedfully explain things to journalists just before Oscar night. Both she and Colonnon have been proudly giving interviews about the foolproof system.

We do that to ensure that in case anything happens to one, the other will be there on time and delivering what's needed with the full set. We do have security measures up until we're at the theater and delivering that envelope to the presenter just seconds before they walk on stage. We'll be in two different locations, Brian will be on stage right and I'll be on stage left. It all sounds sensible and in many ways it is sensible. It's also complicated. The system of twin envelopes meant that every time an envelope was opened on stage, its duplicates in the wings had to be set aside.

Martha Ruiz, stage left, handed the envelope for best actress to Leonardo DiCaprio, leaving Brian Cullinan, stage right, with a job of discarding the duplicate, the job he failed to do. If there hadn't been that set of twin envelopes, Warren Beatty could never have been given the wrong one. So, bad design helped cause the Oscar Fiasco. It also helped cause the accident at Three Mile Island. And a complicated safety system was at the root of the Oscar fast. It was also at the root of the Fermi-1 accident.

But it's not just Oscars and Nuclear Power. I promised you a financial catastrophe too. And in this one, both safety systems and bad typography at a blame, you might even start to see the banking crisis in a very different light. In September 2008, at the height of the financial crisis, an insurance executive named Robert Willemstad, requested a meeting with Tim Geithner. Geithner would later be the Treasury Secretary. At the time, he was the President of the New York Federal Reserve. That meant he was responsible

for supervising Wall Street's banks, including Lehman-Rubbers, which was on the brink of collapse. Robert Willemstad was the boss of an insurance company called AIG and since AIG wasn't a bank, it was far from obvious why Willemstad was Geithner's problem. In his book Too Big To Fail, the journalist Andrew Ross-Sorkin reports the intimate details of this ill-fated meeting. I'm really sorry, Mr. Will instead. Mr. Gatner is going to be a few minutes. No problem. I have time. I'm sorry, I know you've been waiting a long time. Mr. Gatner is on the phone to the

boss of Lehman Brothers. He's up to his eyeballs and Lehman. Tim Geithner was also exhausted. He'd been on an overnight flight from a banking conference in Switzerland. He must have felt completely overwhelmed. Who wouldn't have? Uh, Bob, is it? Sorry to keep you waiting, come in. Willemstadt got his moment. He badly needed to be able to borrow from the Fed, not normally something AIG would be allowed to do, but he also didn't want to panic Geithner. He needed to walk a tight rope to suggest that AIG could use some help, but wasn't actually bankrupt. Is this a critical or emergency situation, Bob? Well, you know, let me just say that it would be

very beneficial to AIG, Mr. Geigner. Perhaps I can leave this with you. Willemstadt handed Geigner a briefing note. Berid, deep within it, was a fast ticking time bomb. The largest firms on Wall Street were relying on AIG to pay out on insurance against financial trouble. The total sum insured was a truly ludicrous $2,700 billion. AIG couldn't possibly pay if all the claims came in, and it was starting to look as though

they might. But that meant that the big Wall Street banks wouldn't get the insurance payments they were relying on. AIG was both a bigger threat to the financial system than Lehman Brothers and a far more surprising one. If AIG was a safety net, it was one that wasn't going to break anyone's fall. But to realise that, Tim Geithner would actually have to read and absorb the information in the note. And he was busy. Really busy.

So instead he filed it away and turned back to the Lehman Brothers problem. AIG would melt down a few days later. The parallels with Oscar Knight are uncanny. For one thing, Geithner, who's no fool, had no idea how to interpret what he was looking at. It was unexpected and the key information was buried in the small print. Fade on away and Warren Beatty know the feeling. Then there's Galileo's principle. Safety systems don't always make us safe. Those insurance

contracts were supposed to offset risk, not create it, right? But by now, we know that safety systems also introduce new ways for things to go wrong. The insurance contracts that were about to destroy AIG were called credit default swaps. They had become popular as a way to offset risk with the blessing of regulators. They seemed a smart idea. Just as the third support for the columns seems like a smart idea, and the metal filters at the Fermi reactor and the duplicate set of award envelopes. But they backfired. Wall Street banks were relying on these credit default swaps to keep them safe if there

was trouble. When it became clear that insurance companies such as AIG couldn't possibly pay out, for banks all scramble to sell off their risky investments at the exact same time, for the exact same reason. A few days after Willemstadt had met Geintner, officials and bankers worked through the weekend on the Lehman Brothers problem. Only on Sunday evening did one of those bankers receive a request from a treasury official to ask if she could drop everything and work on rescuing AIG instead. The surprising phone call was greeted with a response that was unsurprisingly unsuitable for family ears. Hold on, hold on. You're calling me on a Sunday night saying that we just spent the entire weekend on layman

and now we have this? How the f*** did we spend the past 48 hours on the wrong thing? How indeed. For the same reason, they gave the Oscar to the wrong movie, confusing communication, and above all, a safety system that created a brand new way to fail. The banking crisis of 2008 shook the world financial system and destroyed millions of jobs. More than a decade later we're still living with the consequences. It was, in its way,

a more serious crisis than any nuclear accident. It was, in its way, a more serious crisis than any nuclear accident. It was certainly far greater than a bungalow at the Oscars, yet the same problems were at the roots of all these accidents. After the La La Land Shambles, Vanity Fair reported. The Oscars have an intense sixth-step plan to avoid another envelope disaster. The six steps include getting rid of the two accounting partners, Brian Cullen and Martha Ruiz, and the twin sets of envelopes. Instead, says Vanity Fair, there will be three partners. A third partner will sit in the show's control room with the producers.

All three partners will have a complete set of envelopes. If having two sets cause the problem, having three sets is better, right? I'm not sure Galileo would agree. You've been listening to Corsionary Tales. I expand on some of the ideas in this episode in my book Adapt, You Might Like It. Corsionary Tales is written and presented by me, Tim Halford. Our producers are Ryan Dilly and Marilyn Rust. The sound designer and mixer was Pascal Weise, who also composed the amazing music.

In the time since this story was produced, Charles Perot, the brilliant sociologist and author of Normal Accidents, who was mentioned in the episode, passed away. Cautionary Tales is a production of Pushkin Industries, including the talented Julia Barton Heather Fane, Miello Bell, Carly McLeory, Jacob Weisberg, and Malcolm Gladwell. I don't know how you could have listened to more than three minutes of this episode and not subscribed to Cautionary Tales, but maybe you were driving or you were doing the dishes in your hands were wet. If so, you can remedy that right now by opening up the podcast app you're listening

to and subscribe to cautionary tales. There's supposed to be eight episodes, I think, and I'm hoping it's a monster hit because I want there to be dozens more for me to listen to. 99% of Isabel is a project of KALW 91.7 in San Francisco and produced on Radio Roe, in beautiful, downtown, Oakland, California. We are a member of Radio Topia from PRX, a fiercely independent collective of the most innovative shows in all of podcasting. Find more at radiotopia.fm.

You can find the show and join discussion about the show on Facebook. You can tweet at me at Roman Mars in the show at 99PI or we'll have a new episode produced by us next week. But if you need a fix in the meantime, you can always go to 99IPI.org. Radio til the end. From PRX. Thanks.