A complex system, by virtue of its structure, behaves exactly as it should, but its behavior still mystifies us. “What goes around comes around,” was a favorite phrase of the ’60s, but we still haven’t learned that life moves in circles, with confusing delays between cause and effect. We keep expecting the world to be linear and predictable.
Systems create archetypes. These are behavioral patterns that commonly crop up in systems, and therefore can be understood and managed—provided we have mental models to give us a framework. While the archetypes may seem counterintuitive, they’re counterintuitive in ways that we’ve expressed in our culture for centuries.
When you need help with a granny knot, who better to ask than Granny? Let’s examine nine of the most common archetypes and the ways we characterize them everyday language.
1. Information delay, or as Grandma would put it, “closing the barn door after the horses are gone.” This is the most basic systems trap, the problem of late-arriving feedback (as illustrated by the unfamiliar shower). In this archetype, a person or group is pursuing a long-term goal, and they make decisions according to the information that returns from the last decision they made. If they don’t take the information delay into account, every new decision will be wrong.
For example, a consulting firm finds that, after a protracted economic slump, the number of its client engagements finally starts to increase. Yet based on its most recent experience, it holds off on hiring staff until it can confirm that the uptick will continue. The number of engagements grows to the point where the firm can only accept a fraction of the work it’s being offered. So it begins an aggressive hiring program and leases additional office space, which it then builds out with expensive furniture, equipment, and interior design. As soon as the new space is ready, the flow of client work drops off a cliff, and the new space remains empty while the firm focuses on cutting staff.
The problem is latency. If the firm had had advance information about the upturn, it could have begun the hiring process earlier and taken advantage of growing volume of work. And if it had had advance knowledge of the downturn, it could have held off on additional office space and gotten by with contractors instead of full-time employees. Advance knowledge, of course, is in short supply. No one can predict the future, at least not with any accuracy. But a basic understanding of systems would have taught the firm to look for pertinent signs. Maybe business trends in this firm’s industry show up in another industry first. Maybe downturns can be seen first in weaker firms, the “canaries in the coal mine.” Without using these kinds of signals, the firm’s moves will always be too late, causing it to lose money on both the upturns and the downturns.
Rick Perry, a Republican candidate in the 2012 U.S. presidential primary, was asked by a reporter what he might do about climate change. His answer was the political equivalent of comfort food: “The science of global warming is not yet settled, so we shouldn’t risk our economy by addressing it.” Of course, by the time the science is settled, the time for action will have passed and the economy will be the least of our problems. Granny would have reminded Rick that “a stitch in time saves nine.”
2. Addiction, or “when the cure is worse than the disease.” Whenever we use a short-term fix for a long-term problem, we’re in danger of addiction, because we begin to depend on the temporary fix instead of solving the root problem. Let’s say you experience a sleepless night, so you stop at Starbucks for a grande latte on the way to work. This helps in the short term, but by the afternoon you’re even more tired, so you toss down a venti latte in lieu of lunch. At night you’re wired from the caffeine, so you calm your nerves with a couple of glasses of wine and wander off to bed. You go to sleep right away, but by three in the morning you’re tossing and turning. You struggle up the next morning and repeat the whole process, hooked on a downward cycle of quick fixes while the root problem grows worse.
Where’s the feedback delay in this system? It’s between drinking the coffee or wine and seeing the results. A jolt of caffeine works quite well in the first two hours, but later makes us feel more tired. If coffee led immediately to tiredness, none of us would think twice about rejecting it. Instead, we tell ourselves that now is more important than later, turning the “solution” into an additional problem. What systems thinking does is help us see the whole problem by thinking through—in Grandma’s terms—“how one thing leads to another.”
The cure for addiction requires making changes to the false solution—in this case, the cycle of caffeine and alcohol. If the long-term problem is insomnia, then the long-term solution is to address the root cause of insomnia. It may be anxiety, stress, sleep apnea, or alcoholism (especially if it’s more than two glasses of wine), each of which dictates its own solution. One thing the medical profession knows—and everyday experience confirms—is that the caffeine-alcohol cycle only compounds the problem.
3. Eroding goals, or “lowering the bar.” We’ve seen this phenomenon in school, where a teacher gives up measuring students against a broader standard and begins “grading on the curve.” The students, instead of competing as a team against the outside world, are now pitted against each other. This solution creates two new problems: The first is that learning becomes stressful instead of joyful, and the second is that the students lose confidence in their ability to succeed in the real world.
The feedback delay in this archetype, between the lowered goals and the resulting lack of mastery, can take years to notice. It becomes very easy to settle for lowered expectations rather than struggle against a higher standard.
Take the example of software development. The design of software is often a complex and lengthy undertaking, requiring the interaction of many people and many parts. While the working time can easily stretch out, the deadlines are usually less flexible. The quick fix, then, is to meet the deadlines by lowering the quality, while vowing, of course, to clean up the design on the next version. Over time the company creates a culture of “good enough” instead of a culture of excellence, leaving the door open to competition.
One antidote to this archetype is not to lower the goals but to raise the performance. If there’s a performance gap in the classroom, the solution might be to go more slowly, work with students one-on-one, get parents engaged, or institute new rules for classroom behavior. If the software team is performing poorly, the solution may be to clarify the goals, rebalance the workload, collaborate in a different way, or look for better models. The long-term fix could also be a new policy of realistic deadlines, which could be addressed by placing the problem in a larger frame.
Another antidote is to reward people who embrace rising goals. Instead of punishing those who don’t meet the standards, encourage those who surpass them. Let the high achievers serve as inspiration for everyone else, and mine their successes for best practices. In this way, the archetype of eroding goals is flipped to the positive.
4. Escalation, or “an eye for an eye, plus interest.” While an eye for an eye may seem like justice, it usually comes with a side of revenge or dominance. “If you execute one of our soldiers, we’ll execute two of yours,” says General One. “Oh, really?” says the General Two. “If you execute two of our soldiers, we’ll execute ten of yours.” And so on. It’s easy to see how this becomes a trap for both sides. But it takes an understanding of systems to escape it.
The US and the USSR lived through forty years of escalation, caught in a mindless competition to build the largest arsenal of nuclear weapons. Meanwhile, a whole generation learned that life could end any moment in a fiery mushroom cloud, and that planning for the future was futile. The best one could hope for was to live for the moment—a recipe for suboptimization if there ever was one. The situation came close to solving itself in a fireball of mutual destruction during the Cuban missile crisis. Luckily—or unluckily—the USSR finally ran out of money and could no longer fund its side of the game. The US didn’t “win” the Cold War, it simply outlasted it, and is less robust for hav
ing engaged in it.
This dangerous game is not confined to the past. We’re still playing it today against factions in the Middle East. Beginning with the Bush administration and continuing more aggressively with the Obama administration, the CIA has been sending drones over borders to kill perceived enemies, reportedly killing 2,000 militants and an unknown number of civilians. “Is this the world we want to live in?” asks Micah Zenko, a member of the Council on Foreign Relations. “Because we’re creating it.”
The best way to exit an escalation trap is to find a way for both sides to win. It’s not about compromise so much as common ground. By using empathy to understand the needs of the other side, you can design a third solution that changes the original positions for mutual benefit. It helps enormously when both sides see the picture as a system, but even one side can release the trap by unilaterally disengaging.
There’s another kind of escalation—reverse escalation—that can happen with price wars. It’s a race to the bottom in which companies leapfrog each other down the slope to lower and lower prices. Of course, this is a game many can play, so each company lowers its prices to match. They soon compete away their profits, thus hampering their ability to invest in innovation or serve their customers at the former level of quality. As customer loyalty erodes, the companies get stuck in the low-price trap until one or more go out of business. Customers win in the short term, but in the long term they suffer from fewer options in the marketplace.
The way out of a price war is not to engage in the first place. It’s one thing to build a business on low prices, and quite another to sacrifice your profit margins to keep up with the Joneses. As Grandma would say, “Just because your friends jump off the Empire State Building doesn’t mean you should do it.” A better answer is to invest the potential loss of profits into a new product, an improved service, or a change in strategy to reposition the brand. In the short term there might be some pain, but in the long term you’ll have something to show for it.
5. The tragedy of the commons, or “don’t be selfish—take turns!” This trap is related to the escalation trap, except that it takes place in a fragile commons. A commons can be any shared resource that becomes endangered by overuse. A highway is a commons that can accommodate only so many cars before it becomes gridlocked. A park is a commons that can become trampled by too many picnickers. A company home page is a commons that becomes less effective with every inessential item that’s added to it. A fishing area is a commons that benefits no one after it’s fished out.
This archetype is named for a 1968 article by ecologist Garrett Hardin. In it he told the story of a village with a publicly owned pasture on which herdsmen were encouraged to graze their animals. Since the privilege was free and the pasture was large, each herdsman in the village began to think, What harm could there be in adding more animals to the pasture? After all, Farmer Jones is doing it, isn’t he? But every herdsman had the same idea, and before long the pasture was ruined by overgrazing, thereby wrecking it for everyone. “Ruin is the destination toward which all men rush,” said Hardin, “each pursuing his own best interest in a society that believes in the freedom of the commons.” Irresponsible freedom always leads to less freedom.
The main problem, once again, is feedback that never arrives. If no one can see that the resource is being ruined, there’s no reason to limit their use of it. But like the straw that finally breaks the camel’s back, the resource looks perfectly healthy until one day it tips into the danger zone. By then it’s too late to turn around.
Author Donella Meadows, in her excellent primer, Thinking in Systems, suggests three ways out of this trap. The first is to educate and persuade participants to be temperate, appealing to their human decency. The second is to privatize the commons so that each person reaps the consequences of his or her own actions. And the third is to regulate the commons by agreeing on certain rules, then enforcing them.
The American highway system is a frustrating experience for many drivers, not because of the quality or capacity of the roads, but because of the mental models people use for driving them. The reigning model is that, on a four-lane highway, there’s a slow lane and a fast lane in each direction. These are subject to a speed limit that’s sporadically enforced. The result is that each driver is free to decide what constitutes slow and fast. One driver may want to exceed the speed limit to shave some time off his trip, and therefore chooses the fast lane. Another driver feels that since she’s adhering to the speed limit, she should be able to use the fast lane, too. Yet another driver feels that the speed limit is too fast for safety, but still wants to go faster than the trucks in the slow lane, so he moves to the fast lane and simply drives slowly. The result is “clotting,” in which vehicles in the fast and slow lanes often go the same speed side by side, thus blocking all the vehicles behind.
European drivers avoid this problem by using a different mental model. Instead of a slow lane and a fast lane, they conceive of them as a driving lane and a passing lane. They stay in the driving lane until they need to pass, then move into the passing lane, then immediately back into the driving lane, leaving the passing lane open for others. Everyone can go the speed they want without slowing the flow of traffic. The result is very high throughput with very little frustration. While European highways certainly have speed limits, there’s little need for police enforcement since accidents are rare. If we Americans want more freedom on the road, we could do worse than learn from the European system.
6. Rule beating, or “obeying the letter of the law, but not the spirit.” If there’s one quick way to subvert a system, this is it. The rules that govern a system are rarely perfect, since it’s difficult to predict its behavior until it has some operating history. Even then, some problems won’t show up right away (due to latency), and others may be too difficult to address without making the rules ornate.
In 2005 the United Nations began issuing carbon credits to companies that make a coolant for air-conditioning that’s known to be harmful to the ozone layer. The companies receive 11,000 credits for destroying one ton of HFC-23, an even more harmful gas that’s left over from the manufacturing process. They can then sell those credits on the open market, adding revenues that fall straight to the bottom line. The goal of the program is simple: to discourage companies from dumping HFC-23 into the atmosphere. What could go wrong?
Well, 19 international companies, which account for more than 40 percent of the United Nation’s credits, have begun to increase their manufacturing of the harmful coolant. Why? To get the credits from destroying the waste gas. The more coolant they make, the more credits they get for destroying the waste. Not only that, but the higher manufacturing volume lowers the price of the unfriendly coolant so that friendlier alternatives can’t get traction in the marketplace. The companies win while the planet loses.
Some everyday examples: A driver slows down in the presence of a patrol car, and then breaks the speed limit to make up for lost time. A salesman meets his sales quota by keeping the remaining supply of products hidden from the other salespeople. A politician says he didn’t have sex with an intern—technically speaking. A moviegoer sidles up to an old acquaintance to get a better position in the line. A wealthy American investor keeps a large part of her fortune in the Cayman Islands to avoid US taxes.
These are snapshots of rule beating, often excused by the rule beater with the old saw: “Well, everyone does it.” The problem is that everyone does do it, so the effect is multiplied by the number of people in the system. When the system becomes damaged enough to cause concern, the finger of blame is often pointed at the moral character of the participants. But the problem can be partially solved by redesigning the rules so that participants find more advantage in following them than in breaking them.
For example, maybe the computerized speed traps that result in hefty traffic fines could be altered to reward safe driving as well, tracking cars in order to bestow “responsibility points” on their owners. The p
oints could then be used to lower their insurance premiums. Or maybe the largest taxpayers could be awarded “status,” not unlike frequent flyer programs, that confers symbolic benefits such as special parking privileges or invitations to White House events. In designing rules for systems, it pays to reward the behavior you want. As Grandma always said, “You’ll catch more flies with honey than vinegar.”
7. Limits to growth, or “what goes up must come down.” Every form of growth eventually encounters limits. A wildfire that runs out of trees to burn, a city that runs out of buildable land, a virus that runs out of victims, and a business that runs out of customers are all examples of bumping up against limits. When the growing entity hits the wall, it often reverses itself and contracts as quickly as it had expanded.
When a growth line is heading straight up, it’s difficult to imagine that it will ever stop. The leaders of a rapidly expanding business will feel little incentive to change the formula. The phrase, “If it ain’t broke, don’t fix it,” will often be heard. As success feeds upon success, the growth line soars and the company keeps it accelerating by hiring more staff, opening new offices, launching subbrands, and layering complexity upon complexity. The faster the company grows, the faster it hurtles toward its unseen limits. When it finally hits them, the very inputs that kept it accelerating now turn against it, weighing it down with costly complexity that accelerates its fall back to Earth.
The latency in this archetype is the lag between fanning the flames and burning down the forest. The company should have known the party would soon be over, but they couldn’t see it. They’re shocked and confused by their new situation. They blame the market, the economy, their competitors, their customers, and even each other. But they failed to see that the system held the seeds of their demise. If they had channeled Grandma, they would have remembered that “all good things must come to an end.”
Metaskills- Five Talents for the Robotic Age Page 11