10 Rally the Herd

1.

Think of the last time you were in a situation where you weren’t totally sure how to behave. Maybe it was your first time in a new church, or your first time in another country, or maybe it was a dinner party where you didn’t know many of the guests. What did you do to try to fit in?

You watched other people, of course.

In ambiguous situations, we all look to others for cues about how to behave. Maybe you’ve had the experience of scanning the table frantically at a fancy dinner, trying to figure out which fork is for dessert. (If you haven’t had that experience, we hope you know your forks, because the rest of us copied you.) When the environment is unfamiliar, we sprout social antennae that are exquisitely sensitive.

In the fancy dinner situation, our antennae work great, because someone at the table knows what to do, and we can just copy that person. But sometimes in times of change, nobody knows how to behave, and that can lead to problems. For instance, if you ever find yourself in an emergency situation, we pray that there’s only one person in the vicinity who can help you, rather than a crowd. To see why, consider some research conducted by Bibb Latané and John M. Darley.

Columbia University students, having volunteered for a research study, were asked to sit in a room and fill out a survey. Some were left alone; others were put in rooms with two other students. As they filled out their surveys, a “crisis” emerged. Smoke began to pour into the room through a wall vent. The smoke continued to flow, in irregular puffs, until eventually the room was filled with haze. Of the students sitting in a room by themselves, 75 percent got up and found someone to alert about the smoke. But when three students were placed in the room at the same time, only 38 percent of the groups of three ever reported the smoke. They just sat there, inhaling the smoke, each individual’s inaction signaling to the other two people in the room that this smoke cloud isn’t such a big deal.

In a similar study, individuals or pairs who were completing a survey heard what sounded like a woman falling down loudly on the other side of a room divider. Of the lone bystanders, 70 percent went to help her, but only 40 percent of the pairs helped. Even when the pairs helped, they acted more slowly than the individuals.

Why do groups fail to respond as well as individuals?

In ambiguous situations—smoke pouring into a room, the apparent sound of a fall—people look to others for cues about how to interpret the event. If you see a man suddenly collapse at the mall, your brain races through possible interpretations: It’s a heart attack! Or, wait, maybe he tripped and fell down. Or what if he’s playing a gag on someone? You’re reluctant to rush over immediately, because if he simply tripped, your alarm-bell behavior will leave you both embarrassed.

If you’re the only person around to react, you’ll probably make your best guess—heart attack—and rush over. But if there’s a crowd, you’ve got two stimuli to process: the collapse itself and the crowd’s reaction to the collapse. You might pause briefly to study the crowd. Are other people acting like he’s had a heart attack? You stand there, idling, ready to spring into action at the first sign of crisis. But as you wait, other people are looking back at you, and when they see you idling, your behavior becomes data for their theory that it’s not an emergency. And that’s why three people can sit in a room filling with smoke and not make a peep.

We all talk about the power of peer pressure, but “pressure” may be overstating the case. Peer perception is plenty. In this entire book, you might not find a single statement that is so rigorously supported by empirical research as this one: You are doing things because you see your peers do them. It’s not only your body-pierced teen who follows the crowd. It’s you, too. Behavior is contagious. Let’s take a quick epidemiological tour of behavior.

We start with a mind-blowing finding: Obesity is contagious. A groundbreaking study led by Dr. Nicholas Christakis of Harvard Medical School, which followed 12,067 people for thirty-two years, found that when someone became obese, the odds of that person’s close mutual friends becoming obese tripled! Remarkably, proximity didn’t seem to matter. Obesity seemed to “spread” between friends even when they were in different parts of the country. In explaining these findings, Dr. Christakis said, “You change your idea of what is an acceptable body type by looking at the people around you.”

Drinking is contagious. A study showed that when college males were paired with a dormitory roommate who drank frequently in high school, they saw their GPAs go down by a quarter point on average. There’s an endless list of other behaviors that are contagious, as well: marriage; shaking hands to greet someone; wearing fashionably fluffy boots; and investing in Google. And you might want to avoid hanging around with baseball players, lest you start compulsively spitting.

It’s clear that we imitate the behaviors of others, whether consciously or not. We are especially keen to see what they’re doing when the situation is unfamiliar or ambiguous. And change situations are, by definition, unfamiliar! So if you want to change things, you have to pay close attention to social signals, because they can either guarantee a change effort or doom it.

When you’re leading an Elephant on an unfamiliar path, chances are it’s going to follow the herd. So how do you create a herd?

2.

The Elephant constantly looks to the herd for cues about how to behave. This is why baristas and bartenders seed their tip jars—they’re trying to send signals about the “norm” of the herd. It’s a time-honored tactic. In fact, opera companies used to plant stooges in the audience to laugh and applaud at the appropriate times. (If that seems quaint, remember that the “stooges” are alive and well on the laugh tracks of your favorite sitcom.)

But sometimes social cues are hidden. For instance, hotel bathrooms often display little cards asking guests to use their towels more than once, usually appealing to a pro-environmental objective such as the conservation of water. (Also conserved, coincidentally, are labor costs in the hotel laundry room.) So should you reuse your towel? It’s not a clear-cut situation. At home, you probably use your towel more than once, but in a hotel, you might expect a bit more pampering, including a clean towel every time you need one. There’s no obvious social norm to consult, because you can’t peek into the other guests’ bathrooms.

Well aware of the power of contagious behavior, a group of social psychologists persuaded a hotel manager to test out a new sign in the hotel bathrooms. The sign didn’t mention the environment at all; it simply said that the “the majority of guests at the hotel” reuse their towels at least once during their stay. It worked—guests who got this sign were 26 percent more likely to reuse their towels. They took cues from the herd.

Note the downside, though. If the sign had said, “About 8% of our guests decide to reuse their towels,” new guests would have been less likely to reuse theirs. (The same effect explains why you don’t tip the person who bags your groceries, even though that person is doing as much valuable work as your barista. The herd told you not to.) In situations where your herd has embraced the right behavior, publicize it. For instance, if 80 percent of your team submits time sheets on time, make sure the other 20 percent knows the group norm. Those individuals almost certainly will correct themselves. But if only 10 percent of your team submits time sheets on time, publicizing those results will hurt, not help.

When the norms are against you, what can you do to rally the herd? That, in essence, was the problem facing Gerard Cachon, a Wharton School operations professor. In 2006, Cachon became the editor of the journal Manufacturing and Service Operations Management (MSOM). Here are the titles of some of the articles featured in MSOM:

If you just felt your pulse quicken, you are definitely an operations person.

Needless to say, MSOM isn’t the kind of mass-appeal periodical that will be shelved between Maxim and People at the local newsstand. Its role is to showcase the latest thinking in the field of operations. Professors compete strenuously to get their articles published in journals like MSOM, because in order to get promoted within their university departments, they need a solid track record of publication. (You’ve probably heard the expression “Publish or perish.”)

Getting articles published is a long process. First, you do a lot of research—often several years’ worth. Then you write an article describing the research and submit it to a journal. The journal editor farms out your article to “peer reviewers”—other professors who agree to critique your piece (anonymously). The editor then summarizes the opinions of these reviews and delivers a verdict—yes, no, or revise and resubmit. If you get the R&R verdict, you’ll often have to launch some new research to plug the holes in your work, and if you do that, you can eventually resubmit your paper for a new cycle of reviews. It’s an exhausting process—it routinely takes years to get a paper published.

A key bottleneck is the peer review stage. If it takes a long time for other professors to read the article and give feedback, the whole process grinds to a halt. The standard review times for many academic journals range from three to six months. A frontline scientific journal like Science or Nature might get reviews done in a few weeks.

When Gerard Cachon took over MSOM, most peer reviews were taking from seven to eight months, and many were taking over a year. Early in his tenure, Cachon got e-mails from professors who were hesitantly checking in with him about a paper. A typical note might say, “I submitted this paper two years ago and I just wanted to check into progress.” In several cases, when Cachon checked on the paper in question, he learned that the journal had no record of its ever being submitted. Imagine writing that e-mail response. (And take a second to contemplate an environment in which professors feel bad about checking in after two years! Talk about learned helplessness.)

Ironically, operations people are supposed to be the folks who make the trains run on time—they deal with logistics and bottlenecks and supply chains and cycle times. For a psychology journal to outperform an operations journal—that’s as disgraceful as Michael Phelps being trounced in the 100m freestyle by Dr. Phil.

Cachon’s goal was obvious—make things move faster. But what could he do? He had no power over any of the reviewers. They didn’t report to him. Reviewers are volunteer labor who perform a difficult task for free. How would you like to verify the logic of someone’s mathematical derivations in a paper on optimal serial transport systems?

Cachon’s game plan provides a great review of the Switch framework. First, he spoke to his constituents’ Riders by pointing to the destination. “I knew there was a collective goal that I could appeal to,” he said. “Every author wants fast cycle time and is willing to provide it if everyone else does. But no one wants to be the one sucker who provides the fast lead time, and then when they submit their papers it takes forever.” Cachon announced that MSOM would review papers within sixty-five days—that was 72 percent faster than its previous average!

Second, he appealed to identity. We’re operations people, for Pete’s sake. We should be leading the way on efficiency and turnaround time! Third, he defined a clear behavior: Every reviewer had to submit feedback within five weeks. Cachon got the reviewers to commit up front that they could meet the deadline.

Finally, Cachon found a way to rally the herd. Every Friday, he posted an Excel spreadsheet on the internet that showed the status of every paper submitted to the journal. Every reviewer could see what the other reviewers had done (and when). If they violated their five-week commitment, the tracking sheet created powerful pressure, especially when Cachon called them and said, “Look, other people are doing this on time, and, by the way, here’s the data.” When people saw the data, they realized, Whoops, I’m the bottleneck.

With the online tracking sheet, Cachon was using the hotel-towel strategy. He was publicizing the group norm. Other people are getting their work done on time. Why won’t you?

Cachon set out to make good behavior contagious, and he succeeded. As a result of Cachon’s brilliant plan, MSOM now has the fastest turnaround time of any journal in the field of management science. And, because of his work, Cachon was asked to take over the flagship journal of the whole field, Management Science.

Cachon said, “Now, when people get their reviews in fifty days, they come back and say, ‘Wow! I can still remember the paper!’”

3.

We’ve seen that behavior is contagious at the individual level (obesity and tip jars) and at the group level (Cachon’s on-time reviewers). It probably will not surprise you that behavior also is contagious at the societal level (see bell-bottoms and organic food and the phrase “at the end of the day”). But what may interest you is that there’s a particular behavior, now ubiquitous in America, that we can trace back to its origin. What follows is the story of a guy who, at the end of the day, changed the way a society behaved.

In the 1980s, Jay Winsten, a public health professor at Harvard, got interested in the idea of a “designated driver.” He’d picked up this concept from Scandinavian countries, where it was already a norm. At the time, the concept did not exist in the United States. No one here knew what a “designated driver” was.

Winsten and his team at Harvard made it their goal to create a social norm in the United States: If you were going out drinking, you would pick a designated driver who would commit not to drink for the evening. How do you create a social norm out of thin air? Winsten’s inspiration was that you could make the behavior contagious by repeatedly exposing people to it, in many different contexts, even if those contexts were fictional.

Winsten and his team collaborated with producers, writers, and actors from more than 160 prime-time TV programs, sprinkling designateddriver moments naturally into the plots. Segments featuring designated drivers appeared on Hunter, The Cosby Show, Mr. Belvedere, and Who’s the Boss? On an episode of the smash-hit 1980s legal drama L.A. Law, the heartthrob lawyer played by Harry Hamlin asked a bartender to call his designated driver. A designated-driver poster appeared in the bar on Cheers.

“Jay’s crusade was one that we could do something about fairly easily, unlike a lot of other worthwhile causes,” said Grant Tinker, then a vice president of NBC, who introduced Winsten to dozens of writers at all the major networks. Winsten always requested just “five seconds” of dialogue featuring the designateddriver idea, not a whole episode or even a whole scene. “Considering the simplicity of it all,” said Tinker, “it was very hard for us to feel our independence was being challenged.”

Notice how smart Winsten was: He used the power of the Path to change the public’s behavior, but he used the power of the Rider and the Elephant to change the network executives’ behavior. With his five-second requests, he was directing the Rider by describing a simple action that could help on a complex problem, and he was motivating the Elephant by shrinking the change.

In 1991, three years after the campaign launched, nine out of ten people were familiar with the term designated driver. And they were behaving differently as a result. Thirty-seven percent of all Americans reported having acted as designated drivers, and 54 percent of frequent drinkers had been driven home by one. The behavior change saved lives. Alcohol-related traffic fatalities declined from 23,626 in 1988 to 17,858 in 1992.

Winsten used the power of television to simulate a social norm. But you don’t need Hollywood to create a herd. When Steven Kelman, whom we discussed in the “Shrink the Change” chapter, took over the government’s procurement reform efforts, he realized that one of his best strategies was to “unleash” change. By that, he meant that he needed to send signals to the people who already supported procurement reform. Kelman knew he needed to tell the supporters, “It’s safe to get vocal now.” In the beginning, he didn’t need to create new believers so much as he needed to unleash the believers he already had.

In the fall of 2007, a group of public-health and AIDS experts had a chance to get involved in a situation where change needed to be unleashed. The setting was Tanzania, and the subject was sugar daddies.

4.

In Tanzania, “sugar-daddy” relationships are common. You know the drill: An older man pursues a younger woman; they start having sex; and as part of the “deal,” the young woman receives gifts or favors—a cell phone, money for school tuition, clothes, and so on. This is hardly a unique feature of Tanzanian culture, as evidenced by Hugh Hefner and his sextuplet blond housemates.

But the sugar-daddy relationships in Tanzania are more troubling. First, the women are often underage girls—15, 16, 17 years old. Second, the power dynamics in the Tanzanian sugar-daddy relationships often lead women to engage in unsafe sex. (It’s not easy to insist that your partner put on a condom when he’s your elder and your benefactor.) This power dynamic is universal, of course. American teens who date men who are six or more years older are almost four times more likely to get pregnant than teens who date within two years of their own age.

The reality is that when older men want unsafe sex, they tend to get their way. In America, that means the girls get pregnant. In Tanzania, it means they get AIDS and die. And that’s why sugar daddies in Tanzania are a serious public-health problem.

Health experts call these sugar-daddy relationships “cross-generational.” In the age group 15 to 24, women in sub-Saharan Africa are three times more likely to be HIV-positive than men in the same age range. It’s the relationships that these younger women have with older men that explain the disparity. The cross-generational relationship opens up a bridge for HIV to travel between populations that ordinarily wouldn’t intersect.

The other thing that’s different about sugar-daddy relationships in Tanzania is that, despite the health risks, there’s no strong social taboo against the behavior. In the United States, 50-year-old lechers who chase college girls are punished for it socially. Can’t you just hear the man’s sister telling him, “You’re pathetic”? Can’t you just see the eyes rolling at his office? But there’s no real equivalent of this social stigma in Tanzania. It’s accepted that high-status men will display their status in this way.

Yet the great majority of Tanzanians—89 percent in one poll—believe that cross-generational relationships are wrong. Unfortunately, their opposition tends to be quiet and private; it’s an uncomfortable issue to discuss.

In August 2007, Pamela White and Mike Gehron of USAID, a relief organization that’s part of the U.S. State Department, called together a diverse team of experts (including the two of us) in a hotel in Dar es Salaam. The mission: to develop a campaign to fight cross-generational sex. Leading the team was a group from Johns Hopkins University’s Bloomberg School of Public Health. Others included AIDS experts and about a dozen local artists and creatives (producers, actors, writers, and at least one Tanzanian soap-opera star, who was interrupted at one point for an autograph).

The discussions were difficult. The issue was complicated, and it wasn’t even clear where to start. For example, no one believed that scolding the sugar daddies would be effective in stopping their behavior. And the public-health experts thought that we were unlikely to convince young women to refuse the advances of these men, because the social and financial pressures on them were too intense to be countered by a campaign. So we started thinking: If we can’t change the main characters in this story, can we change their environment?

Recall from Chapter 8 that Rackspace changed individuals by changing their environment, their culture. But that was one company. Was it possible to change the social atmosphere in an entire country? The team knew that Tanzanians objected to cross-generational relationships but, for whatever reason, remained quiet about them. Could the team give a voice to that resentment?

We needed to find a way to make Tanzanians feel comfortable discussing something uncomfortable, a way to disarm the conversation. And someone blurted out: “We need people to be able to laugh at this! We need humor!”

Inspired, the team hatched the idea for a villain. He’d be a villain people would love to hate, like J.R. Ewing, the eternally scheming oilman of the old TV show Dallas. As the group kicked around the idea, a portrait of the villain began to emerge: He’d be an older man who encapsulated the sugar daddy—someone who hit on young girls constantly, relentlessly, shamelessly. He’d approach them anywhere he could find them, and he’d offer favors—free meals, free drinks, clothes, or time on his cell phone.

Someone suggested that the villain be called “Fataki,” and everyone’s eyes lit up. Fataki is a Swahili word that translates loosely as “explosion” or “fireworks”—something dangerous and a bit unstable. In other words, Fataki is someone you want to stay away from.

The plan was to start telling Fataki’s story in radio commercials, because radio comes closest to being a universal medium in Tanzania. The team started dreaming up dozens of situations to put Fataki in, and across these situations there would be a common element: Despite his status and wealth advantages, and despite his smooth talk, Fataki never succeeds. He fails again and again because an outsider intervenes to disrupt his seductive efforts. Like Wile E. Coyote stalking the Road Runner, Fataki chases and chases and chases but never gets his prey. He’s pathetic. He’s someone you can laugh at.

Here’s the translation of a Fataki spot called “Chicken and Chips”:

(Noisy sounds of a restaurant)

GIRL: Oh so many choices …

FATAKI: Please pick one …

GIRL: Ah, waitress, is this chicken or …?

WAITRESS: I would recommend this one …

FATAKI: That’s too expensive! Listen. Give her chicken and chips with some sauce. That’s good, right, my love?

GIRL: Hmmmm … OK.

(Waitress acknowledges their order and walks away.)

FATAKI (to Girl): Oh, please go and tell her it is a take-away!

GIRL: Okay …

(Background voices of waiters ordering food)

WAITRESS (whispering): I’m glad that I have a chance to talk to you alone. Don’t you feel ashamed to have a relationship with such an old man? Here, take your chips and leave quickly through this back door …

(Sound of door opening)

FATAKI: Waiter, what is going on there?

WAITRESS: Sir, the girl just went away. Would you like something?

FATAKI: What?! I came with her. She went through which door?

WAITRESS: But you have not paid for the chips, sir …

FATAKI: Ehhh???

WAITRESS: Sir, this kid is not your age …

FATAKI: Ah, take the money and leave me alone!!

WAITRESS: Hee hee hee! Got him!

ANNOUNCER: Protect your loved ones from a Fataki!

Radio spots like that one became part of a unified campaign that was piloted in a rural region called Morogoro. The campaign consisted of 10 different radio spots that were played on 3 stations, and 170 banners that were hung on stores and public buildings. The typical person heard about one radio spot a day. The campaign had two goals. The first was to create a mocking label for sugar-daddy behavior. The creative team’s objective was to someday walk into a Tanzanian nightclub and overhear one patron telling another, “That guy is such a Fataki.” By making it OK to mock the Fatakis, the team would help to counteract the natural status advantages of older and wealthier men. The second goal of the campaign was to encourage “interventions” by outsiders—friends, relatives, teachers, even waitresses—by modeling the behavior in the radio spots. The message was: “It’s your responsibility to look out for these young women. Protect your loved ones from a Fataki!”

The results from the pilot campaign were unexpectedly strong. By the end of the four-month campaign, 44 percent of people who were asked “What would you call a 50-year-old man who is always trying to seduce younger women?” spontaneously replied “Fataki.” Seventy-five percent of Morogorans reported discussing Fataki with others. And the percentage of people who said “I can do something about cross-generational sex” increased from 64 percent before the pilot to 88 percent afterward.

Having succeeded in Morogoro, the campaign was rolled out nationally in Tanzania, and the word began to spread nationwide. A health care worker setting up HIV clinics in the outer reaches of Tanzania reported that even in remote villages, Fataki had managed to bring a conversation out of the closet. Within weeks of the national launch of the campaign, the banner front-page headline of a Tanzanian tabloid accused Kanumba, a popular actor, of being a Fataki. (He’d been spotted checking into the Lambada Hotel with a young girl.) The Tanzanian public had taken ownership of a name and a character who symbolized the bad behavior they’d resented, quietly, all along.

—————————— CLINIC ——————————

How Can You Stop John’s BlackBerry Addiction?

SITUATION John has a BlackBerry addiction. His body twitches every time his BlackBerry goes off. He can’t stop himself from checking every message that comes in. It’s become a serious distraction. In meetings at work, he finds himself sneaking the device under the table to read messages covertly. (It annoys John’s colleagues that he thinks he’s fooling them.) John’s wife becomes increasingly irritated that he can’t focus on their dinner conversation, and one day he almost has a fender-bender because he is trying to e-mail someone while driving. John knows that he needs to cut it out, but every time he resolves to stop, the BlackBerry buzzes. [John is fictitious, but we all know a John.]

WHAT’S THE SWITCH AND WHAT’S HOLDING IT BACK? This is the last Clinic in the book, and by now, we hope you’ll make easy work of this situation. The behavior change is clear enough: John needs to stop using his BlackBerry all the time (and especially while driving). What’s holding him back? His Elephant, of course. In any addiction situation, the Elephant is the culprit. To rein in John’s Elephant, we’ll use all three parts of the framework. Take a moment to generate an action plan for John, and then compare your notes with ours.

HOW DO WE MAKE THE SWITCH?

• Direct the Rider. 1. Find the bright spots. Are there times when John doesn’t feel the BlackBerry compulsion? What’s different about those times, and can we find a way to replicate those conditions? 2. Point to the destination. John needs a B&W goal, like BP’s “No dry holes.” Recall that B&W goals are particularly useful in situations where people are prone to rationalize. (John is constantly telling himself, “I just need to check and see if this one specific e-mail has come in.”) John could experiment with different B&W goals: No BlackBerry during dinner, or No BlackBerry after 6 p.m., or The BlackBerry comes out only when I’m traveling. He needs to eliminate his own wiggle room.

• Motivate the Elephant. 1. Find the feeling. Have John’s wife force him to read aloud the last ten e-mails that he received and then ask whether any of them is really worth getting twitchy over. A little embarrassment might do him good. 2. Find the feeling. Car accidents are happening more and more because of people like John. Confronting John with a particularly sobering news story—for example, “Cute Puppy Run Over by E-Mailing Driver”—might be a good idea. 3. Build identity. John’s wife (or his colleagues) could highlight how his BlackBerry habit doesn’t fit with his character: “John, you’re usually such an ‘in control’ guy. It’s weird to see you so flaky.” 4. Build the growth mindset. Some smokers quit successfully on the seventeenth try. If John truly wants to curb his habit, his friends shouldn’t let him give up if he “relapses” a few times.

• Shape the Path. 1. Break the environment. John’s wife can simply smash his BlackBerry with a hammer. Problem solved. 1a. Tweak the environment. If smashing it isn’t possible, John could lock his BlackBerry in the trunk of his car every time he drives. That way, he wouldn’t have to fight the Elephant when it buzzes. 2. Tweak the environment. The siren song of the BlackBerry is its buzz (or ring or flashing red light). Can you turn off the sound? Cover up the light? (Paint over the light with Wite-Out if necessary.) 3. Rally the herd. John’s colleagues should make sure he knows he’s not fooling anybody in meetings. They should make a pact that, each time John sneaks a look at his BlackBerry under the table, they will all stare at him until he meets their (disapproving) eyes.

5.

The Fataki case study may have felt foreign to you. It probably doesn’t bear many superficial similarities to the changes you’re contemplating. But look at the underlying dynamics: You want certain people to act differently, but they are resistant to the change. So you rally the support of others who in turn could influence those you hope to sway. In essence, it’s an attempt to change the culture, and culture often is the linchpin of successful organizational change. As former IBM CEO Lou Gerstner said, “I came to see, in my time at IBM, that culture isn’t just one aspect of the game—it is the game.” But organizational culture is a slippery, abstract concept. How do you change it? Where do you start?

In 1984, Libby Zion, an 18-year-old freshman at Bennington College, at home visiting her parents in Manhattan, died in a New York teaching hospital. She’d been given the wrong medication by a medical resident who’d been working for over 19 hours. Her death sparked an outcry over the excessive hours worked by medical interns. (Interns are first-year residents. They’ve completed three years of medical school and are beginning full-time work in hospitals.) Traditionally, interns have worked an astonishing 120 hours per week.

The story of Libby Zion became the centerpiece of a campaign to limit the workweeks of medical residents. Almost two decades later, in 2003, Congress finally seemed ready to move. Then the American Council for Graduate Medical Education—which accredits medical schools—made an effort to preempt congressional legislation by requiring 80-hour workweeks for residents starting in July 2003. That might seem to be the end of the story. A switch happened in medical schools because someone with fiat power mandated the change.

In this case, though, fiat power failed. A subsequent study by the Journal of the American Medical Association found that only a third of general surgery residency programs complied with the new workweek limits. Why would hospitals fail to change their ways in spite of the fact they were putting patients at risk (as well as their own accreditation)?

Katherine Kellogg, an ethnographer at MIT’s Sloan School of Management, wanted to understand why some organizations complied with such regulations and others didn’t. She decided to study two northeastern teaching hospitals, nicknamed Alpha and Beta, that were well matched in size, sector, structure, and other factors. She was given full access to the staffers of the two hospitals, and she observed the two hospitals for an average of twenty hours a week, each, for fifteen months. Kellogg was the perfect eyewitness to a real-world cultural change effort.

6.

In the quest to reform, the fiercest battle would be fought over a surprisingly mundane practice called the “daily signout.” The daily signout is the point, usually about 9 or 10 p.m., when surgical interns hand off their patients to the on-call night resident. During the signout, the intern briefs the resident on each patient’s status, so that the resident is prepared to take care of any emergencies that happen overnight.

But interns weren’t really handing off their work, and that’s one reason why they were working 120 hours a week. On every third night, there was no handoff at all—the interns were on duty all through the night. And on the other nights, they stayed late to finish the paperwork they should have handed off. Worse, night residents usually refused to do any paperwork overnight, so the interns had to show up early the next morning—at, say, 4 a.m.—to complete it before beginning their normally scheduled rounds at 6 a.m.

To enforce an 80-hour workweek, the hospitals needed to start taking the daily signouts seriously, but that wasn’t quite as easy as it sounds. Emotional resistance to reform was entrenched in both Alpha and Beta and, indeed, in most hospitals across the country. Signing out conflicted with long-standing cultural practices. For instance, one opponent said, “You learn by being here. There is a huge amount of information passed on in an ad hoc fashion at 2 o’clock in the morning when the senior resident and you are trying to get a patient’s blood pressure up or an IV in. You need to be the one managing it, doing it, in order to learn.” (In other words, interns would be putting their medical education at risk by, er, sleeping.)

Other opponents stressed the importance of “continuity of care”—minimizing the number of times that patients were passed among doctors. As one resident said, “The problem with handoffs is that things fall through the cracks and things get missed.” A final objection, in essence, was that if the interns didn’t pay their dues, it would be unfair to everyone who’d already put in time as the paperwork lackeys: “These guys have been there and done that…. You can see how it would be tough as a fourth-year resident to be told to do the intern job all over again.”

Fortunately, at both Alpha and Beta, there was substantial support for signout change among the interns’ superiors, such as senior residents and chief residents. At Alpha, 13 out of 31 superiors were “reformers” who supported the change; at Beta, 12 out of 18 were reformers. Because of this support, circumstances looked ripe for real change. All signs were “Go.” At long last, the 120-hour workweek would be abolished.

There was just one problem: The interns wouldn’t sign out. As one intern said, “Being considered a good intern has nothing to do with what you know or how well you manage your patients. It is totally based on working hard and not handing stuff off. It’s your attitude, not your ability…. This is where I live. I only go home to sleep. It sounds sick, but these people are like my family. The worst thing would be not to be respected by these guys.”

For the interns, social status was at stake. The interns felt they wouldn’t be respected if they embraced the signout. Change was coming into conflict with culture, and let’s face it, a new rule is no match for a culture.

Could the hospitals change their culture? Here’s where their paths diverge: Fifteen months into Katherine Kellogg’s research, Alpha had managed to win the culture battle, and Beta had lost it. Anyone who wants to create organizational change needs to understand why.

7.

Kellogg discovered that the change hinged on the smallest work teams at the hospital, which met each day for “afternoon rounds.” Each team was composed of three or four residents (both interns and more senior people), and during afternoon rounds, they discussed the patients under their care and other important issues. The afternoon rounds at the two hospitals differed a great deal:

Teams were shuffled about once a month, and every so often, by the luck of the draw, all the people on a team would be reformers who supported the shorter workweeks. At Alpha, the reform-minded teams found great strength in their rounds. They were spending an hour in private discussions with a group of their fellow believers. But at Beta, the format of the afternoon rounds blocked any momentum for reform. The teams met for only a short time, and members often were missing. Worse, teams were meeting in the computer lounge, where lots of people who opposed reform could overhear their conversation, which led reform-minded members to self-censor.

Bottom line: At Beta, the afternoon rounds were irrelevant to the change. At Alpha, they became the spark, and the rounds became, in essence, underground resistance meetings.

8.

Researchers who study social movements call situations like these “free spaces”—small-scale meetings where reformers can gather and ready themselves for collective action without being observed by members of the dominant group. Free spaces often play a critical role in facilitating social change. Civil rights leaders, for instance, were able to use southern black churches as free spaces to prepare themselves for action.

Kellogg was present, every day, at Alpha and Beta as one reform movement blossomed and the other wilted. She attended thirty-one free-space meetings at Alpha and twenty-two meetings at Beta. The meetings at Beta had a comparable number of reformers but weren’t conducted in a free space.

At Alpha, 77 percent of the meetings included discussion about the legitimacy of the signout process, and 81 percent of the meetings drew boundaries between “us” and “them”—reformers versus resisters. At Beta, none of the meetings featured either kind of discussion.

During the Alpha free-space meetings, the reformers began to develop a language for talking about the advantages of the reform. For instance, during one meeting, the chief resident deconstructed the “continuity of care” argument made by change resisters:

It is important to take personal responsibility. But I think you can preserve this personal commitment without having one person there all the time. Some people are old school and say, “I’m going to do it myself.” For me, it is the team that is going to take care of everything. Each of us takes personal responsibility to make sure that patient care is the best it can be, but that doesn’t mean doing it all yourself as long as all of the pieces fit together. [emphasis added]

This resident was developing what Kellogg calls an “oppositional identity.” Every culture, whether national or organizational, is shaped powerfully by its language. Across the reform-minded Alpha teams, a new language was being incubated that reflected a new set of values. Old school versus new school. Trusting your team versus doing it all yourself. Being efficient versus living in the hospital.

At Alpha, the reformers had the space and the language needed to brew a new identity. At Beta, they didn’t. The lessons are clear. If you want to change the culture of your organization, you’ve got to get the reformers together. They need a free space. They need time to coordinate outside the gaze of the resisters.

Counterintuitively, you’ve got to let your organization have an identity conflict. For a time, at least, you’ve got to permit an “us versus them” struggle to take place. We know this violates our “we’re all on the same team,” Kumbaya-ish instincts. It’s not desirable, but it’s necessary. Think of it as organizational molting.

To encourage this molting in your culture, think of all the tools we’ve built up in the Path section. First, you need to tweak the environment to provide a free space for discussion. At Alpha, the rotational teams had a private place to meet, and that created a free space where the new identity could grow. Do your own “reformers” have a private place where they can meet and coordinate?

Second, you should build good habits. Recall the idea of action triggers—visualizing when and where you are going to do something important. The interns at Alpha were essentially setting action triggers. They thought about what they would say, and how they would act, when 9 p.m. came and the signout process was triggered. They mentally rehearsed how they would respond if an argument flared up with the night resident. Have members of your team rehearsed how they’ll react when they meet resistance from your organization’s “old guard”?

Finally, you should rally the herd. At Alpha, the leaders helped the reformers find one another, and the reformers began to create a language—as we saw in the examples of the designated driver and Fataki—that allowed them to talk about their values with others. As a leader, you can help prod them to create this language, to find ways to articulate what is different and better about the change you seek.

9.

We started the Path section by discussing the Fundamental Attribution Error—the tendency to attribute behavior to people’s character rather than to their environments. Look again at the teaching-hospital example. At Alpha, 42 percent of superiors supported the change; at Beta, 66 percent supported it. Almost all of us would have put our money on Beta. Not many of us, when confronted with that data, would have immediately thought, Yes, but what about the situational forces?

At the two hospitals, individual character competed with situational forces, and situational forces won. This brings us back full circle to the food drive study, where a jerk with clear instructions was more charitable than a saint with generic instructions. The line between saints and jerks wasn’t as clear as we might have thought. Neither was the line between supporters and re-sisters of hospital reform.

If the right Path can turn a jerk into a saint, then the right Path can also turn a change enemy into an ally.