Cultural Transformation at NUMMI

U.S. automobile manufacturers have begun to heed the wakeup call that sounded more than a decade ago when Japanese automakers cut deeply into American markets. As General Motors, Ford, and Chrysler began to take stock of what had happened, they were surprised to discover that they had much to learn from the Japanese, not only about the processes of production, but also about the use of human resources. Today, the Big Three have begun to rebound. In 1993, all three companies posted gains simultaneously for the first time since 1984.1 The turnaround stems partly from exchange rates that now favor U.S. exports, and partly because output from Japanese transplants is included in the U.S. figures. But improvements in the auto industry also stem from what U.S. automakers have learned from Japan.

In this paper, we examine a notable case, New United Motor Manufacturing (NUMMI), in which such learning produced significant cultural change. NUMMI is a joint venture formed by the world’s two largest automakers — Toyota and General Motors. Located in Fremont, California, in the shell of a huge GM assembly plant that had closed in 1982, NUMMI opened two years later. Now in its tenth year of successful operation, it has survived the trials that cause most joint ventures to founder.2 NUMMI demonstrates how the introduction of a new production system and a foreign culture transformed one of the worst GM plants into a world-class assembly operation in a unionized environment. NUMMI’s experience, we think, has implications for other organizations that are trying to learn from each other, especially those that work across international boundaries. Of equal importance, findings from NUMMI reveal the central role that organizational culture plays in enabling an organization to adapt to a continuously changing environment.

In 1989, five years after the joint venture was created, we formed a team of five UCLA faculty members and graduate students to work in the plant and try to understand how such a radical transformation was possible.3 We knew that to capture the essence of the change, we would have to become insiders. Because there was little theory to guide us, we would have to develop theories or hypotheses from data we collected. The few other studies of NUMMI had not examined the company from the inside.4 Accordingly, we decided to work as ethnographers using participant-observation, a method anthropologists and other social scientists use to study people in their natural settings.5 While participant-observation has been successful in industrial settings, most studies have been limited to blue-collar workers or a single researcher’s efforts.6 Our study traversed the company from the assembly line to top executives to track changes over time. We spread through the company to tap into day-to-day life. Our multiple viewpoints enabled us to cross-check findings, which increased their reliability and validity.

In October 1989, five of us began working in different groups on the assembly line that represented a range of performance from high to low. Wearing NUMMI shirts and employee badges, we slowly became insiders, learning the Toyota production system firsthand, while establishing relationships with team members who taught us our jobs. For four months, we worked on both first and second shifts at various stations ranging from stamping to chassis and final assembly, where we stacked body parts and installed engine components, bumpers, carpets, mufflers, door panels, and seats. During downtime and breaks, we talked with team members, team leaders, and group leaders and recorded notes (with everyone’s knowledge). Through our constant presence, we were able to document critical incidents as they occurred, capturing unpredictable events like a group meeting that suddenly blew up, or the plant manager stopping the line to dramatically illustrate a point. The incidents provided a window into the organization from which we could see how it responded to stress.

Later, after we were accepted on the factory floor, we observed managers and engineers as they worked. As trust built, we gained access to meetings where we learned NUMMI’s decision-making style and saw how managers, labor leaders, and team members handled conflict. During the initial four-month period, we interviewed hundreds of people ranging from team members to the president. The interviews, coupled with more than a thousand hours of participant-observation, produced nearly 3,000 pages of typed fieldnotes, which we cross-checked with each other each week and coded into a database. We met periodically with a research committee of Japanese and American managers, union leaders, and team members, which validated the findings and directed our work in the plant. Our intent was not to evaluate “good” or “bad” practices, but rather to piece together a composite picture of NUMMI from multiple viewpoints. We also presented our findings to the company in periodic meetings with managers and team members. For the next five years, we maintained a lower-level but constant presence at the company to track changes as the study’s scope widened.7

We discovered that the company’s transformation stemmed largely from its ability to create a new “third” culture, a hybrid of the best of its American and Japanese parentage. What is especially interesting is that Toyota and General Motors did not intentionally set out to develop a new culture in the abstract like so many other companies do. Rather than creating slogans and the image of a new culture, they aimed to build a new organization designed around the Toyota production system. That production system, however, rested on a radically different set of assumptions than did mass production.8 In turn, the new assumptions and beliefs placed different demands on U.S. production workers and union leaders, most of whom came from the old GM plant. Likewise, members of the management team, who were hired from Ford, Chrysler, and other GM plants, also had to learn new ways. In this paper, we reveal how such a conversion was possible and some of the key principles that we can generalize from NUMMI’s experience.

The Guiding Role of Organizational Culture

Every organization contains a set of assumptions and beliefs for guiding individuals’ day-to-day working behavior, which together compose its organizational culture.9 Beliefs grew from scientific management, which rationalized the division of labor and centralized authority, and formed the basis for an inflexible culture, making large companies vulnerable to more nimble, decentralized competitors. Assumptions that led management and labor into generations of conflict similarly put many of America’s unionized smokestack industries at a competitive disadvantage with foreign companies with more cooperative labor relations. Such assumptions may have been rational when the United States monopolized world markets. However, in today’s unpredictable and fast-changing economy, they impede organizations’ abilities to adapt, often making them unproductive. For example, findings from our larger study of NUMMI, USS-POSCO, Douglas Aircraft, and Hewlett Packard show how America’s industrial growth led to the formation of a pervasive set of beliefs that have proven dysfunctional in the rapidly changing world economy.10

Studying organizational culture is difficult because there is no commonly accepted theoretical framework to explain what it really means. To some, it is the glue that holds an organization together.11 To others, it is a system of beliefs, clues to deeper issues. However, at deeper, less visible levels, organizational culture is formed from individuals’ beliefs, values, and unconscious assumptions.12 Its invisible quality also renders it elusive and difficult to grasp, much less to change. Managers frequently focus on surface-level alterations, leaving deeper beliefs and assumptions untouched.13 The beliefs and assumptions, however, truly guide day-to-day behavior. Researchers differ on whether transforming an organization’s culture requires a crisis or more peaceful, incremental steps. Most agree, however, that in mature organizations, little progress can be made in altering fundamental beliefs without intense stress.14

NUMMI: A Cultural Laboratory

The case of NUMMI affords useful insight into how a cultural transformation can occur. It reveals the critical role NUMMI’s production system played in directing and maintaining the change in behavior and beliefs once the joint venture was launched.

The original GM plant in Fremont, California, was built in 1962 as a state-of-the-art facility. However, angry workers called it the “battleship,” partly because of its drab greenish-tan color, partly because of the endless conflict that raged there for twenty years. Sick-outs, slowdowns, and wildcat strikes frequently disrupted production, and daily absenteeism usually reached 20 percent. Alcohol and drugs were freely available on the premises.

As Japanese automakers cut deeply into the U.S. market, causing GM to shut down its unprofitable operations, Fremont became a leading candidate for closure. As in other strife-torn GM plants, Fremont had a long history of heavy-handed management and union conflict, which had created an “us versus them” attitude. The union’s leadership had learned to use conflict to dig deeply into GM’s pockets. The only time the two sides came together was to negotiate a new contract. Detailed job descriptions and wage rates, negotiated in an atmosphere of mistrust, defined human roles impersonally and contractually. Most of the union’s energies went to preserving these agreements and extracting new ones with each new contract. In 1982, when GM closed the plant and laid off the workforce, more than 6,000 grievances remained backlogged in the system.

Negotiating the Joint Venture

Toyota and GM both had much to gain from a joint venture. Toyota needed a foothold in the U.S. auto market, and its executives wanted to learn whether its production system could be successfully implemented in the United States. GM wanted to add a new small car to its line and also hoped to learn Toyota’s production system.

As Toyota began to negotiate with GM, its representatives were skeptical about working with a militant union like the United Auto Workers (UAW). General Motors’ executives knew, however, that it would be politically impossible to reopen the Fremont plant without the UAW, and ultimately persuaded Toyota to recognize the union. GM and Toyota retained Bill Usery, a former Secretary of Labor and head of the Mediation and Conciliation Service, to represent their mutual interests.15 Usery, no stranger to adversarial relationships, had reservations that he confided to then-UAW president Douglas Fraser:

Doug, I need to know how you feel about this deal with the Japanese because we’ve got the biggest blue-water ocean in the world between us. We speak a different language and we have a different culture. We don’t have as much working experience with the Japanese as we had with the Europeans. The UAW has some strong feelings about Japanese cars.

Fraser acknowledged the risks of working with the Japanese but also recognized the significance of the new venture and committed the union’s support. Usery shuttled back and forth between Detroit, Washington, and Toyota City, crafting the details of an agreement that would set the stage for the joint venture’s creation. He had met with the chairmen of both companies, Roger Smith and Eiji Toyoda, and was confident that they were each committed to the joint venture.

Wrangling broke out within the International union over whether to insist on the traditional work rules in the master contract or to give in to Toyota’s demands for abolishing seniority as a base for hiring and collapsing the myriad job classifications. Many UAW veterans resisted giving up seniority and job classifications, which represented the accomplishments of a lifetime of struggle. Bruce Lee, the UAW’s western regional director, and others in the UAW argued that retaining these protections would only slow the recovery of the U.S. auto industry. Lee commented:

The traditionalists within the union are so tunnel-visioned, they would have insisted on a mountain of stuff to climb over. And it wouldn’t have made one bit of sense to the working person. It would also have killed the Japanese interests and any hope of reopening the plant.

With support from outgoing UAW president Fraser, incoming president Owen Bieber, and Don Ephlin, head of the UAW’s GM department, the union was able to give Toyota the assurances it needed. After more negotiations, Toyota took a critical step. Its executives agreed that the joint venture would rehire the majority of its workforce from the pool of laid-off GM workers, and that it would recognize the UAW as the bargaining agent for the workforce. The joint venture also promised to pay union-scale wages. The UAW in turn agreed to accept the Toyota production system, to increase greatly the flexibility of work rules, and to simplify the job classifications. Part of the agreement letter indicates the tone of the new relationship: “Both parties are undertaking this new proposed relationship with the full intention of fostering an innovative labor relations structure, minimizing the traditional adversarial roles, and emphasizing mutual trust and good faith.”

The New Company

In 1985, its first year of operation, NUMMI assembled 64,766 Novas — a car that has been hailed by leading consumer magazines as one of the highest-quality small cars.16 Workers’ daily attendance averaged 98 percent in 1985, and only a handful of grievances were filed. NUMMI was cited in union magazines as having “changed into a worksite unlike any other in the UAW’s experiences.” Newsweek called NUMMI “a model of industrial tranquility,” while the Wall Street Journal reported that NUMMI “has managed to convert a crew of largely middle-aged, rabble-rousing former GM workers into a crack force that is beating the bumpers off Big Three plants in efficiency and product quality.”17

Tatsuro Toyoda, son of Toyota’s founder, headed NUMMI’s management team. Kan Higashi, also a Toyota executive, became executive vice president; Dennis Cuneo, whom Higashi hired from a Washington firm that assisted Toyota in the joint venture negotiations, became head of NUMMI’s legal affairs. Cuneo joined NUMMI partly because he was convinced that it represented a watershed case in American antitrust law:

Just a few years ago, the FTC and the courts would have never allowed this joint venture. Here you had the two largest automakers in the world coming together to form a joint venture, while Chrysler and Ford were hemorrhaging, near bankruptcy. NUMMI offered the chance to advance a more contemporary interpretation to American antitrust law.

Gary Convis left a ten-year career with Ford in Cleveland, Ohio, to head manufacturing. Bill Childs was hired from General Dynamics to establish NUMMI’s human resources and labor relations departments. Though regarded as a hard-liner by many in the plant, Childs saw NUMMI as an opportunity to alter the prevailing model of antagonistic labor-management relations.

After ten years, NUMMI continues to improve. By 1987, NUMMI matched Toyota’s quality and nearly matched its productivity. Today, the plant represents a $1.6 billion investment and has already had three major expansions, most recently a new truck line. In 1992, NUMMI produced 104,407 Toyota Corollas, 75,271 pickup trucks, and 75,493 Geo Prizms, and its truck line is scheduled for yet another expansion. In 1994, it produced 102,114 Corollas and 108,000 Geo Prizms. The number of workers has grown from 2,200 to more than 4,300 and is projected to increase in 1995.

NUMMI produces the same number of cars as GM did in the same plant, but with much higher quality and half the workforce. In 1993, J.D. Powers rated NUMMI’s pickup truck as the highest-quality compact pickup truck. Similarly, in 1993, its Geo Prizm and Toyota Corolla were ranked the best compact cars built in the United States. In May 1994, the company won J.D. Power’s “Silver Quality Award” for its Geo Prizm, above every other American-built car ever rated by J.D. Power. The Toyota truck and Corolla received high ratings as well, exceeding the quality found at their Japanese sister plants.18

At NUMMI’s tenth anniversary, General Motors president Jack Smith, head of the GM side of the negotiations, commented, “Ten years ago a lot of people thought NUMMI would never get off the ground. But today, with the quality and efficiency that NUMMI has achieved, it is the mecca of the auto industry in North America.” Tatsuro Toyoda, president of Toyota, remarked, “I have a special place in my heart for this company. NUMMI’s success proves that labor and management can work toward mutually beneficial goals that serve the best interests of everybody.”

Establishing a New Culture

There have been a variety of explanations for NUMMI’s achievements, none based on first-hand research in the plant. Even people inside NUMMI were unsure about the reasons for its success. Bill Childs quipped, “We know what goes in and what comes out. But what goes on in the middle remains somewhat of a mystery to us.” Some think NUMMI’s success resulted from workers being unemployed for two years and willing to accept almost anything to regain their high-paying jobs. Others attribute its achievements to the careful job-application screening process. Our research indicates that the company’s success stems from the creation of a third culture that was neither American nor Japanese. The hybrid, by combining the best of both parents, enabled the new company to break away from the old conflict-ridden culture and start anew.

Our evidence suggests that if GM had set out to recreate the plant from its old workforce, it would have failed, because old beliefs were too deeply etched into the psyches of managers and workers to accommodate the necessary radical change. Gary Convis explained how working with the Japanese was a powerful form of training:

The Japanese have been a critical factor in what we’ve accomplished at NUMMI. It’s not until you work with them, learn the principles from them, that you begin to understand how they see the world and how they feel. My understanding keeps deepening and my commitment to managing in a more open way gets stronger. It has evolved over time from seeing why they do things how they do them. So the idea that we could have learned what we have without the Japanese is absurd.

GM’s seasoned, experienced workforce provided a hospitable environment for the Japanese contributions. Most of the people we interviewed said they would not go back to work for GM, even if it reopened across the street. Most NUMMI employees take pride in the quality of the cars they build under the new system. In a recent survey, 93 percent said they were proud of their work.19 A team member commented:

I’d be talking to people and they’d find out I’m from GM. Man, I’d be embarrassed. But you know, they made us build cars that way. One day I found a bolt missing. I called the supe over and he said, “What’s the matter with you boy, you goin’ to buy it? Move it!” Then when the plant failed, they blamed us.

NUMMI’s tolerant atmosphere, in which differences of opinion are accepted and valued, provided another important ingredient. The joint venture inherited a work-force of people who had known each other for many years. Friendships cut across lines that might divide others, creating an open atmosphere. Though racism and sexism can be found in any organization, NUMMI’s employees (more than 50 percent minorities) get along with relatively little friction. Their tolerance helps unite the multicultural, multiethnic workforce, often bridging serious conflict. Employees speak proudly of the plant’s diverse workforce, commenting that differences between people are a source of strength. Jesse Palomino, a former UAW committeeman explained, “This place is a blend. We have Mexicans, blacks, whites, Japanese, men, and women. And we all think a little differently. That’s good.” Sherry Ward, a seasoned GM autoworker, commented on how she felt when a Fortune article reported that NUMMI’s workers came from the barrio and the ghetto:20

I look at some of these people — people I’ve known most of my working life — and I don’t see them that way. Hell, after reading that article, I’d be afraid to come in this place! We’re the pride of America — autoworkers, middle class. We’re a culturally diverse group. Isn’t that the way it should be?

Principles of a New Culture

As the company worked to fuse the Japanese and American beliefs into a single unit, it discovered four principles without which little progress could have been made. And, as we will show, each principle enabled the company to develop a new set of beliefs around which an emergent culture is constructed:

• Both management and labor recognized that their futures were interdependent, committing them to a mutual vision.
• Employees felt secure and trusted assurances that they would be treated fairly, enabling them to become contributors.
• The production system formed interdependent relationships throughout the plant, helping to create a healthy work environment.
• The production system was managed to transform the stress and conflict of everyday life into trust and mutual respect.

Also, by concentrating on areas where Japanese and American culture naturally converged, while openly acknowledging their differences, NUMMI was able to establish a common ground on which to create a new culture.

A Mutual Vision between Management and Labor

The management team knew from the beginning that unless it overcame the old adversarial management-labor relationship, the new venture would founder. Workers were skeptical about Bruce Lee’s decision to appoint the old GM bargaining committee as the new venture’s bargaining agent because of its leadership’s reputation for militancy; these twenty-five individuals were now expected to guide the union into a new era of cooperative labor-management relationships. However, by being included in decision making from the outset, they could test management’s commitment to teamwork, trust, and respect. Because the committee members were also influential in establishing the mutual vision between management and labor, they recognized that their futures were intertwined, a first step in creating a model of cooperative labor relations.

Lee knew that the venture would fail if the old leaders sabotaged it. But he also knew these people were important allies:

These guys were tough as nails. They’d strike GM as fast as you could snap your fingers. They hated General Motors, and they didn’t want to change. But I knew that if we didn’t bring in the leadership of the old GM local, they’d be settin’ across the street pitchin’ firebombs at the place. There was no way Toyota could have opened the plant without them.

As an elected union official, Lee had a great deal at stake. If the union cooperated and gave away some of its traditional job protections to provide the flexibility Toyota demanded, and if the workers found the conditions intolerable, Lee would quickly become the target of union opposition. He gathered the leaders of the old GM bargaining committee and told them: “What I want you to do is give it a fair hearing. If you come back and say, ‘Bruce, this just doesn’t work, then I’ll say, ‘Fine! We’ll try something else.’ ” Lee admits that bringing in the old UAW leadership at the outset was of great value.

At the startup in early 1984, union leaders and newly hired managers together stuffed envelopes with job applications. Five thousand former GM workers received applications with a letter explaining how new employees would be expected to contribute to an atmosphere of trust and cooperation and that poor quality workmanship and absenteeism would not be tolerated. About 3,000 applications were returned. In the end, 85 percent of the newly hired men and women had worked in the old GM plant. UAW members and NUMMI executives made hiring decisions on a case-by-case basis, screening and testing each of the applicants. Some applicants withdrew after they were told what was expected of them. According to Bill Childs, only 300 applicants were rejected outright — because of unusually poor work histories or drug or alcohol problems.

Many of the UAW local’s leaders had difficulty setting aside old antagonisms to work with the company. Rank-and-file union members were suspicious of collaboration. Tony DeJesus, Local 2244 president, and George Nano, chairman of the bargaining committee, were frequently accused of being backsliders and sellouts. However, both DeJesus and Nano knew that the joint venture offered a chance for their members to go back to work. Nano described conditions at GM, where he headed the bargaining committee. “We had a lot of unchecked power,” he said, referring to the union. “And we used it. But we forgot that there’s got to be smoke coming out of the stack or there’s nothing to fight about.” Some union leaders saw that the adversarial relationships had to give way to a new form, but no one knew what it was.

Toyota sent 400 trainers from Japan to work with American counterparts to teach the production system that would become the leverage for cultural change. During the intense startup, the Americans and the Japanese searched for points of agreement, establishing valuable relationships. At the same time, NUMMI invested more than $3 million to send 600 of its new employees to Toyota for training. At Toyota, Tony DeJesus learned the system, while getting to know the Japanese. At first, they communicated in sign language:

We learned how close-knit the managers and workers are on the floor. They have to be to work that hard! American workers always complain when things don’t go right. But the Japanese have a respect for authority — it’s not a dictatorship, it’s not forced, it’s just part of their culture. The Japanese are reserved and they didn’t invite us to their houses the first time. But when they came here, we took them to our houses and some other places I won’t discuss!

The bonds of friendship and mutual understanding became more important as Americans and Japanese felt the stresses and strains of starting the new company.

A Crucible of Security and Fairness

The second major principle in establishing a new culture based on fairness, trust, and mutual respect was NUMMI’s no-layoff policy, which Toyota initiated. Large Japanese companies had offered employment security since World War II, though the policy is now being reconsidered in the wake of Japan’s current recession. The UAW quickly embraced NUMMI’s policy of no layoffs unless the long-term viability of the company is threatened. The policy also stipulates that before anyone is laid off, all outside contract work will be dropped and executives will take substantial pay cuts. At first, workers distrusted the policy because it was so unusual, but it has now been tested at least four times — and each trial has made the company more credible. In March 1988, NUMMI was forced to reduce production 40 percent due to a sales slump. Most U.S. auto companies would have laid off the second shift, but NUMMI’s management stood behind its policy and fired no one. Team members not needed on the line were retrained in the basics of the new production system, teamwork, and problem solving.

Not surprisingly, NUMMI employees value their security, especially in the climate of rising unemployment elsewhere. A recent company survey showed that 80 percent of the team members agreed that job security was the most important aspect of working at NUMMI. The policy also fostered the workers’ belief that they are valued as assets rather than as costs to be trimmed during downturns and reinforces managers’ respect for their employees. Gary Convis remarked:

One of the key concepts is respect for the worker, for the team member. The Japanese know that to make things more waste-free and streamlined, they have to work with the people on the line. They have to work with their people, to listen to them for their ideas, and to work with them to support theirs.

Convis explained that when something goes wrong on the line, most U.S. managers look for a culprit —someone who is slacking off or doing something wrong. In contrast, Convis said:

Japanese managers go looking for the problem. They trust their team members are doing their best. When something breaks down, managers feel it’s their responsibility, and they’re apologetic, out of respect for their team members. It’s that mindset that the Japanese have helped teach us.

Kan Higashi explained that, from the beginning, Toyota understood the need to establish the concept of fairness. He wanted a general leveling — fewer levels of management, no executive perks, and a blurring of lines between Japanese and Americans, managers and team members. Higashi admitted that, after a fifteen-hour day, he wanted nothing more than to fraternize with his Japanese colleagues, but he resisted:

I knew we would be misunderstood if we ate separately, so we all ate together. The things we did at NUMMI, like creating open offices and a communal cafeteria and getting rid of reserved parking spaces, we did out of necessity. Your academic friends say we imported Japanese management techniques, but what we did at NUMMI is not found here in Japan! These were not gimmicks. Rather they were symbols of our concept and of our intent.

As Toyota’s production system was introduced, the management team and the union compressed job classifications and created a flat wage structure that equalized work and rewards while fostering fairness. GM’s former eighty job classifications were collapsed into only three —one production class and two higher-paid skilled trades classes — each of which falls into the same $19 to $21 hourly wage rate. More recently, NUMMI instituted a bonus system, based on gainsharing, that benefits all employees equally.

NUMMI’s flat wage structure also reinforced the feeling that the company’s fortunes depended on everyone’s efforts. Many American managers believe that employees are motivated to work hard only by incremental pay increases, not by flat wage rates. To the contrary, Convis explained:

Our team members are ready and willing to change as long as they feel they are being treated fairly and equitably. We’ve tried to avoid favoritism and to level out the harder jobs. A single pay level is as fundamental to the success of this company as is security of employment. We have learned from the Japanese the importance of tying the company’s success, and the success of the individual, to things they can control.

Generating Interdependence from the “Pull” of Production

As Toyota’s production system has diffused to NUMMI, some adaptations have been made, but its key features have endured. One is that each employee must be responsible for product quality — a significant departure from tradition at GM. By standardizing work and by removing excess inventory through just-in-time delivery, Toyota’s “pull” system creates a tight link with the customer.21 It demands that individuals work together interdependently and cooperatively for the common good.

The system runs at a preset speed known as takt time (the German word for meter or musical rhythm) that determines the length of time a team member has to perform his or her task, currently set at sixty seconds. Because the time standard is observed throughout the plant, it levels the work — known as heijunka — while keeping the line under continuous tension.

By running the system with few buffers and using human motion sparingly, problems can be traced to their root causes through kaizen, or continuous improvement. Toyota’s jidoka principle, that jobs are to be done correctly the first time, minimizes the chance that defects will be passed down the line. However, workers are obligated to stop the line when they see a problem by pulling the overhead andon cord. Empowering workers to stop the line was a radical change from the old GM plant. Gus Billy, a veteran union leader recalled, “At GM, you just didn’t stop the line. The superintendent would have your ass. No one had that authority except him.

Nevertheless, some NUMMI team members try to avoid stopping the line because of their old habits. Some say they worry that they might be punished or that their supervisors will be disciplined. Other team members pull the cord just long enough to get the team leader’s attention, then pull it again to shut off the lights and music without actually stopping the line. Managers, however, want the line stopped because they have learned from Toyota that perfection is unattainable; it is the process that is important. One manager remarked, “When there’s no downtime, I know that my people are sending junk through, or they’re trying to be super-stars.” His comment reflects NUMMI’s philosophy that each group should generate a few minutes of downtime each day because each imperfection represents an opportunity for improvement.

Another element that reinforces interdependence is making performance data available. Standardized work and kaizen charts hang in the team areas, along with attendance boards and defect records. Visible information keeps everyone abreast of changes, and individual contributions are easily seen. Nowhere is this principle more evident than in NUMMI’s daily Assembly Quality Review. After lunch, about forty to fifty team leaders, group leaders (NUMMI’s first-line managers), and assistant managers convene to discuss the defects found in a random sample of cars. Managers explain to their peers the reasons for defects and the planned countermeasures. Managers say they do not mind being singled out, but one group leader commented, “It’s okay to be up there once in a while. I mean we’re all human. But if you’re up there more than once for the same defect, watch out!”

Tension, Conflict, and the Development of Trust

Managing a system like NUMMI’s is difficult because it is dynamic and requires constant adjustment to change. When NUMMI reaches its goals, team members and managers naturally tend to become complacent, which causes quality to suffer. During several months, we began to detect predictable cycles of effort, revealing that every drive for perfection must ultimately be accompanied by a leveling off and downward cycle. Some veteran auto-workers interpret these cycles as a return to old GM days, which generates fear and uncertainty. On a cycle’s downside, defects increase, causing management to exert corrective pressure that stresses the entire organization.

Osamu Kimura, NUMMI’s third president, described a constant need to “renew the spirit” by finding problem areas and accelerating improvements. For instance, in 1991, when complacency had set in and quality began to suffer, Kimura launched the J-1 program: “We made it a campaign with buttons and banners, so that employees could see our commitment each day.” By reinforcing the basics, Kimura’s J-1 program refocused energies to break the downward cycle, leading NUMMI to win the highest J.D. Power rankings for U.S.-built compacts and number one for compact trucks.

However, when the pressure gets too great, managers sometimes fall back on old ways. For instance, Mike Mulleague, manager of assembly, explained how pressure to meet production quotas tempts him to get cars out of the plant at any cost. He learned to trust that productivity would follow quality by following the system’s principles. Mulleague commented that it was difficult at first to grasp what Higashi meant when he said, “I don’t care if you build only six cars a day, as long as they meet our quality standard.” Mulleague remarked: “When the pressure’s on, half of the people in this place will revert to their old ways — pushing cars out the door, and I’m no exception.”

Discipline is critical to NUMMI’s system. When managers fail to follow the system’s principles, order breaks down, quality suffers, and relationships become strained. For instance, team members interpret managers’ demands for productivity as a signal that NUMMI’s quality is slipping, which raises fears of regression. However, when management consistently observes the system’s principles, quality and productivity achieve a natural balance. For example, in trial runs in 1993, team members found that some new parts were difficult to install on the new models. Convis described how the problem held up production for weeks, inceasing pressure as the schedule slipped:

We stuck to our guns as far as building the car in station, holding to the principles of the Toyota production system, adding help where we needed to keep that system in place. We were always in control. We knew that although we had a lot of problems, we weren’t violating the system. We weren’t producing bad quality cars, so that’s number one. You stick to quality first. It’s not just a word, it must be in your heart.

Team members and managers traced the problems to the parts suppliers, where they were remedied. Only after making corrections was the line brought up to speed. Convis reflected:

Consistency and the Japanese style of consensus making are the keys to better management. We learned from the experience. Just because our system’s nine years old doesn’t mean it’s not fragile. People will follow management’s direction. Whatever management does, and in what direction they push, and how hard they push dictates where this company eventually goes.

Trust is indispensable in a dynamic organization like NUMMI because it serves to insulate the organization from the shocks of change. It develops from working together productively under stress. For example, pressure from management caused an experienced group leader, Roger Gallet, to eliminate a job on the line without first consulting his group. During the morning break, Gallet informed his team members, with whom he had worked for a number of years, that he had already made the change. Pandemonium followed, as one team member shouted angrily, “You know, we’ve got a stake in this place too! Why weren’t we consulted? Aren’t we important?” Another team member blurted, “Management’s got to learn to play by the rules. It’s not GM, you know, a one-way street. I don’t want to see this place fold, but this is sure as hell no way to run a plant.” As music signaled the end of the break, the meeting broke up without resolution. On the way back to the line, one team member commented, “This isn’t like Roger. He must be being pushed from the higher-ups.” Another responded, “Yeah, but it doesn’t make it right. He’s got to know how we feel.”

At the end of the shift, Gallet called an impromptu meeting and explained, “I was under pressure this morning, and I didn’t want to hear what you had to say. I’m sorry.” The group’s mood shifted, becoming more relaxed and positive. A team member whispered, “Well, at least he’s honest. You know, Roger’s really okay. We cut him some slack.” After the meeting some of the team members tried to soften the morning’s harshness by tapping Gallet on the arm, making friendly comments as they went to their lockers. It became apparent how the trust that had developed between Gallet and his group over the years played an important role by absorbing conflict and redirecting energy.

When a manager gives in to pressure, however, and turns authoritarian, he or she usually gets an angry response that distances team members and diminishes trust. For instance, management announced a policy that groups should rotate jobs every two-and-a-half hours to reduce injuries, level the work, and break boredom. In one meeting, as the manager discussed the policy, members voiced heated objections. As the tension built and he began to lose control, he pointed at the fifteen team members and blustered, “We’re going to rotate whether you like it or not! Management is not going to tolerate the defects coming from this group. And if you don’t perform, you might not be around much longer!”

Hands waved as team members asked to be recognized, but he ignored them. Anger spread as one woman shouted, “Why not find out what the majority wants? Let me tell you, sir, you’re gone a good deal of the time, and you don’t know what goes on here.” A chorus of angry voices yelled, “That’s not fair! Who the hell do you think you are?” The manager backed behind a table and put his hands on his hips. Facing the group, he said, “There are some things you vote on and some things you don’t. There is a time to get input, but I don’t have to take a vote on it or get your blessing. You’d never get consensus from this group anyway. My role is to give you information! Rotation! We’re going to have it!”

He stalked off, but team members stayed behind talking angrily in small groups. One man said, “You just can’t get the American out of the manager. It’s getting just like GM. I thought the Japanese owned half this plant. Must be the other half!” What little trust that may have existed was quickly obliterated in this angry exchange. His manager and a member of the human resources department counseled the group leader in more effective ways to foster participation and handle conflict. His performance has improved, and he still works at the plant.

A large element of building trust lies in NUMMI’s participative style of management. NUMMI does not have a democratic system in which workers and managers vote on issues, but rather open decision making, which enourages individuals to contribute. Convis explained its importance:

The Japanese taught us how to manage from data. Data don’t lie. The consensual form of decision making insures that the whole picture emerges, that each piece of data is included and analyzed. That’s a far cry from the old American way of shouting “We’re all screwed up,” and getting into a shouting match about who’s at fault and what to do about it! The Japanese grind data. They grind it and grind it until they come to a decision that’s right.

Every voice in a decision has value and must be heard; management’s job is to honor opposing forces, draw them together, and reconcile them in the decision-making process.

Forging a Third Culture

NUMMI has now survived for nearly a decade.22 Our evidence suggests that the establishment of a third culture accounts for the company’s transformation. However, this culture required that the Japanese and Americans find common ground.

Points of Convergence . . .

Employment security was a concept that both the Japanese and Americans could embrace. Secure workers are more productive, and as a company’s productivity increases, its market share expands. In 1991, Toyota selected NUMMI to build its pickup truck largely because of the company’s record for high productivity and quality. Toyota invested another $340 million in the plant and expanded the workforce by about 900 employees. Team members witnessed how this principle translated into opportunity, as the plant grew and new jobs were created.

NUMMI’s form of consensual decision making was yet another area where both cultures could agree. Its goal is to insure that all individuals can contribute and that decisions include multiple viewpoints. Japanese coordinators taught their American counterparts how to extend decision making to workers. They gained the respect of most NUMMI employees because of their detailed job knowledge and willingness to ignore status differences and work on solving problems.

Tony Fisher, who worked for GM and now heads NUMMI’s environmental affairs, described the power of consensual decision making, “The Japanese ask questions and seem to talk endlessly about solutions. It may take a long time to make a decision, but once it’s made, that’s it. Pity the person who tries to undermine the decision once it’s made.” He described how an American manager tried to intervene once a decision had already been made: “The Japanese sat the guy down and straightened him out. ‘That’s our decision,’ they said, ‘And we’re going to put our heart into it.’ Boom! That’s it! You don’t learn this kind of leadership in the classroom — you learn it from coaching and by trial and error.” Fisher recounted his introduction to Toyota’s beliefs, which he found could be easily integrated with his own: “I always believed in including people in important decisions. I’m a Christian, a Catholic, and I believe in the Golden Rule — the essence of respect and trust. But I had never seen it practiced in business until I came here. The Japanese are non-Christian and look what they do!”

Finally, though some team members complain about the system’s demands for discipline, most agree that, under GM, discipline had become too lax. Team members and managers respect the self-discipline the Japanese bring to their work. One manager marveled at how the Japanese analyze a problem: “They’ll sit there and spend hours on it, asking hundreds of ‘whys’ until it’s solved.” A team member on the final assembly line explained that he likes the Japanese discipline because it fosters orderliness and pride in the company. He characterized NUMMI as a family in which the parents have to insist on what they want: “When GM ran this place, it looked like a dump — chicken bones in the cars, broken glass in the parking lot. The Japanese insisted that it stop, and we don’t eat outside anymore. The parking lot’s clean, just like the rest of the place.”

. . . And Divergence

Japanese communitarian beliefs, however, frequently conflicted with American individualism. Hayao Kawai, a Jungian analyst and professor of education and Japanese studies at Kyoto University, explained the basis of the Japanese communal spirit:

In the East, we think of ourselves as related to everything — animals, plants, and humans. This table is no different from you or me. By saying “I,” we separate ourselves from each other. So rather than saying “I love you,” I might say only “love.” That’s enough. We don’t like to use “I” and “you.” Often if you ask a Japanese, “What do you think?,” he’ll say, “Don’t know.”

Kawai described how Japan’s ancient animistic religions conceived of the center, where Christians place God, as nothing. In this way, the center is kept open, allowing the entry of other religions, with which life can be harmonized and balanced. This principle extends to the inclusive concept of lifetime employment.

The Japanese concept of self as part of a larger community reveals itself in numerous ways at NUMMI. A Toyota senior coordinator explained how he first noticed the difference between Americans and Japanese —in sports competition:

Here in America, the competition is among individuals, but the Japanese compete as a group. If you look at how many gold medals Americans got from Olympic Games, you can easily see just how capable and skillful each American is. In contrast, Japanese as individuals cannot compete with Americans at all.

The most important thing is the extent to which each individual joins his or her effort in the company. Each Japanese individual is far less strong than each American, but when they join their efforts, they are stronger. American individualism is not bad if you look at it from the perspective of an individual. However, it is big trouble for a company or for industry.

Interestingly, the clash over communitarian and individualistic values occurs in the areas on which Japanese and Americans most easily agree — employment security, participative decision making, and plant discipline. For instance, when NUMMI started, music would play before each shift. At first, Toyota coordinators led the Americans in warm-up exercises to loosen their muscles and enhance group spirit. Over time, interest in the exercises died, though the music still played. One team member, who appreciates the Japanese influence, nevertheless quit the exercises: “They just went too far. I mean we’re Americans. We’re not robots!”

Some Americans also said that Japanese loyalty to their employers makes them uncomfortable. One manager explained that the Japanese develop loyalty early by interning youngsters in high school. The children know that if they work for Toyota, they’ll be taken care of. “We don’t have anything like that here. It’s not the same type of commitment. That’s why the Toyota model in its pure form wouldn’t work here at NUMMI.”

But Toyota coordinators say that, without a strong group consciousness, U.S. companies are at a disadvantage when competing with the Japanese. A young Toyota coordinator expressed concern with the distance between American workers and their employers: “Japanese workers make a lifetime commitment to their employers, but Americans lack this sense of ‘belonging.’ ” He explained how the Japanese family supports the collective effort; in the typical Japanese household, “It’s like the aircraft carrier that supports the fighter.” A U.S. manager commented, “Most of us put our families first. We work to live, not live to work. I don’t think the American worker will ever be like the Japanese. I just don’t think that model would ever work here.”

Sometimes tension surfaces, as Toyota coordinators press the fine points of the system and unknowingly transgress this imaginary line. For example, after the first shift, a senior Toyota coordinator was training a group of managers in standardized work. He squatted down to count some trays filled with screws, bolts, and nuts. According to the standardized work part sheet, six trays should have been on the left, and two on the right. He loudly counted the trays one by one, using exaggerated arm movements to draw attention to his point: “One, two, three, four, five, six, seven, eight, nine! Chart says six plus two equals eight, but I counted nine!” He exclaimed in mock disbelief, “How can this be?” As the managers listened, a team leader who overheard the conversation responded angrily, muttering under his breath about the Japanese’ attention to detail, “Picky, picky. They’ll always find mistakes. Works great in theory, but not down here!”

Some U.S. workers fear losing their individualism. One night at a party after work, a group from chassis discussed a team member, Ramon, who had left early after angrily denouncing the company and the team concept. Ramon was known as a complainer, doing as little work as possible and counting the days until retirement. Team members agreed that Ramon thought he would get attention by complaining. “Like the squeaky wheel gets the grease?,” a young new team member asked innocently. “Not exactly,” said the group leader. “In Japan, the sticky nail gets the hammer! Ramon’s lucky we’re Americans. If this were Japan, he’d get clobbered!”

Finding Common Ground

NUMMI has successfully steered around these intractable problems, focusing energy on critical issues where the Japanese and Americans must find agreement. This is most evident in how Toyota introduced the production system to the American workforce. At first, according to Kan Higashi, the Japanese coordinators at NUMMI were not sure how to approach the Americans and resisted the urge to transfer the pure form of the process. A senior Toyota coordinator reinforced Higashi’s view, saying that, at Toyota, he simply explained how to do something, and team leaders followed without question: “But, at NUMMI, I didn’t get the same reaction.” He had to compromise because he could not force his way on the American workers: “The new way had to be localized. It had to fit correctly with American ways. Otherwise, team members simply refuse to participate. It took a year for me to discover a way to get the Americans to agree with me.” Only after they reached mutual agreement, would NUMMI employees share their ideas and make suggestions on how to improve the system. The coordinator concluded, “What we are developing here is a mixed model of Toyota and GM.”

Osamu Kimura noted that the cultural exchange has worked both ways: “Despite our differences as Japanese and Americans, we found that we share many important similarities.” Kimura has a personal style that managers and team members appreciate. During union negotiations, he often plays poker and drinks with the union leaders. He worries, however, that language obscures the fine points of communication so critical to building relationships:

Practical language, discussing quality or cost is not so difficult. But jokes at dinner, which are an important means of communication, are most difficult. Americans sometimes say that Japanese are very shy and do not like to joke. But that is not true. It’s just that most American jokes end very quickly with a fast sentence or two. Americans laugh, but I cannot understand! Because of the language, we cannot join in those areas!

Another Toyota coordinator explained how personal relationships with his U.S. colleagues helped him understand American culture. He described GM and Toyota like oil and water and NUMMI like a small salad bowl:

Unless you mix them well, the oil and water will never become dressing. NUMMI is like a small salad bowl — a big one would never work — and communication is the way to mix them. It takes a while to mix them to make a good dressing.

Conclusion

NUMMI has so far weathered pressures that have caused others to founder, and its culture continues to evolve. Since the late 1980s, Toyota has gradually recalled its management and technical team, which today includes Iwao Itoh (NUMMI’s fourth president) and about twenty-five managers and coordinators. However, Toyota began to scale back its presence at NUMMI only after being assured that its key principles had been successfully diffused, and that the American management team had learned to manage the dynamic system. Itoh, like his predecessors, is fully conscious of the continuous reinforcement needed to keep the system from slipping into a comfortable zone of complacency: “My greatest challenge is to make sure that both managers and team members truly internalize the concepts. That is the key to successful diffusion.”

With few exceptions, the original U.S. management team remains nearly intact. However, after recent elections, there have been changes in the union’s original leadership. Bruce Lee, interpreted the change in union leadership as normal:

It’s only natural that ultimately there’d be another caucus. Some of these guys had been in power before, but now they were out. So what do they do? First, they take on the system, but when that doesn’t fly, they take on the leadership. You know when the opposition says “This system stinks,” it falls on deaf ears. Nobody is trying to throw the system out.

As we noted earlier, NUMMI remains efficient, and its quality continues to rise. Internal quality audits done by Toyota, GM, and NUMMI over the last few years were validated in May 1994, when NUMMI’s Geo Prizm won the highest score ever for an American-built vehicle. Toyota continues to invest in the joint venture’s expansion. There is little doubt that Toyota, GM, and the UAW have succeeded in finding common ground on which to construct a new and productive culture. The Federal Trade Commission noted NUMMI’s importance when, in late 1993, it set aside its original order limiting the joint venture to twelve years. The commission explained that it made the decision because of evidence that NUMMI was helping General Motors “reap the benefits of gaining first-hand experience with an efficient production system.”23

The NUMMI case reveals three elements for transforming organizational culture. First, a force great enough to induce change, or “unfreeze” an organization, led the transformation.24 The force was powerful enough to overcome individuals’ natural fear of abandoning their core beliefs for new ones.25 For instance, union leaders and team members had to suspend hard-learned beliefs that management would take advantage of them in order to learn Toyota’s principles. When individuals learned that they would share fairly in the fruits of the company’s success, they began to embrace the system’s principles and let go of old, dysfunctional beliefs. The impetus came partly from the shutdown and resulting unemployment, but more so from two added factors: NUMMI’s commitment to treating its workforce fairly, and Toyota’s integrated production system that demanded participation, mutual respect, and trust. Without these twin incentives for change, little progress could have been made.

Second, Toyota assumed dominance in the joint venture and took full responsibility for teaching the Americans how to implement and manage its production system. The Japanese partner’s beliefs and management style were more in tune with market demands for quality and speed, as well as with employees’ needs for cooperation, trust, and respect. Agreeing on who was responsible for day-to-day management at the outset enabled Toyota to administer consistent direction from the top, which sent a message of consistency and stability to the rest of the company.26

Third, Toyota managers resisted temptations to forge ahead with a pure version of the system. Both the Japanese and Americans learned as they went. By adopting a “go slow” attitude, the Japanese and Americans remained open to points of resistance as they arose and navigated around them. By tolerating ambiguity and by searching for consensus, Toyota managers established the beginnings of mutual respect and trust with the American workers and managers.

Toyota’s attitudes contrast sharply with those of Mazda when it established a plant in Michigan in 1988.27 Mazda’s assembly plant was not a joint venture, although it had some characteristics similar to NUMMI in its agreement to hire an all-union workforce. However, Mazda failed to lay sufficient groundwork and establish a mutual cultural understanding with the U.S. workers. Mazda knew little about American unions, and the unionists were suspicious of their Japanese bosses’ motives. The UAW local confronted Mazda’s management and resisted the introduction of its production system. Grievances mounted and Mazda tried to exert control by ordering workers not to speak publicly in hopes of avoiding negative publicity. Mazda’s tactics only confirmed union leaders’ worst fears. Cooperation and teamwork broke down as the UAW’s New Directions, a wing of the union that opposes cooperation with management, won control. Thus, despite Mazda’s hopes for a cooperative, profitable venture with its unionized American work-force, the company’s failure to examine both its own and the UAW’s beliefs and assumptions led to a morass of conflict that could have been avoided.

NUMMI has created a new culture from the ashes of the old GM plant. Toyota has successfully diffused its principles, and the company has enjoyed nearly a decade of continuous expansion. It has thus demonstrated to its workforce that, by converting waste to value-adding activity through continuous improvement, market share and resulting employment will grow. Though there is growing evidence that U.S. manufacturers have begun to adopt some of the principles introduced by Toyota at NUMMI, it is still unknown just how far or in what form these lessons have permeated.28 As other manufacturers embrace Toyota’s principles, it is also unclear how the system will endure if economic conditions change, demanding cutbacks and layoffs. Also unclear is what new manufacturing and management paradigms will emerge and how lean systems like NUMMI’s will adapt. However, the adaptability and flexibility of NUMMI’s new culture suggests that it is prepared to adjust to an unpredictable future.