When the System Changed, the Workforce Changed

What the Toyota-General Motors joint venture teaches leaders about culture, management systems and operational performance

Central Lesson

The workforce did not suddenly become capable. The management system finally allowed its capability to matter.

In organizational life, weak performance is often explained by weak people. NUMMI offers a more demanding interpretation: before judging the workforce, leaders should examine the system in which that workforce is being asked to perform.

One Plant, Two Operating Realities

General Motors closed its Fremont, California, assembly plant in 1982 after years of troubled labor relations, poor quality, weak productivity and absenteeism that frequently exceeded 20 percent. Approximately 5,700 workers were idled. Within GM, Fremont had come to represent the kind of operation that appeared almost beyond repair. [2]

Two years later, the same facility became the setting for an unusual industrial experiment. Toyota and General Motors established New United Motor Manufacturing, Inc. (NUMMI) as a 50:50 joint venture. Toyota would manage daily production; GM would contribute the factory, market part of the output and gain direct exposure to Toyota’s production knowledge. Toyota, meanwhile, would test whether its operating model could succeed with an American workforce, an American union and an established U.S. facility. [1]

The experiment was especially revealing because NUMMI did not solve the problem simply by replacing the workforce. Researchers reported that 99 percent of assembly workers and 75 percent of skilled-trades workers at start-up were former GM-Fremont employees. The entire union hierarchy was rehired. The people who had been associated with one of GM’s weakest operations were now being asked to produce at world-class levels. [2]

What Toyota Changed

NUMMI did not begin with a campaign asking employees to adopt a more positive attitude. It changed how work was designed, how problems were surfaced and how managers were expected to respond. Culture followed repeated experience.

Problems became visible at the source

Under the Toyota Production System, quality was not something inspected into the vehicle at the end of production. Employees were expected to identify abnormalities while the work was being performed. Visual controls and the andon process made problems easier to see and escalate. A worker encountering a defect or an obstacle could call for help and, when required, stop the line. [3]

This changed the meaning of speaking up. Raising a problem was no longer an interruption to production or an admission of failure. It was part of the job. Management, in turn, made an operating commitment that support would arrive quickly. Employee voice and management response were designed as two sides of the same control system. [3]

Standardization enabled improvement

NUMMI used standardized work, but standardization was not treated as a permanent instruction imposed on passive employees. Team members and team leaders participated in defining work methods and were trained to improve them. Once an improvement was accepted, it became the new standard until a better method was developed. [2]

The distinction matters. In a rigid organization, standards can preserve the status quo. In a learning organization, standards establish a visible baseline from which improvement becomes measurable.

Management behaviour reinforced the new expectations

Toyota invested heavily in developing supervisors and team leaders. Early NUMMI managers and supervisors learned through direct experience, mentoring and training in Toyota plants. John Shook, who helped develop training for NUMMI employees, later emphasized that new behaviours were defined, taught and reinforced before leaders expected attitudes to change. [3]

The message delivered through daily practice was more powerful than a statement of values: expose the problem, ask for help, solve the cause and improve the standard. When employees experienced that sequence consistently, trust became operational rather than rhetorical.

The Performance Shift

The results were significant. By 1986, researchers found that NUMMI’s productivity was almost twice the level achieved by the former GM-Fremont operation in its best years, approximately 40 percent better than the typical GM assembly plant and close to Toyota’s Takaoka plant in Japan. NUMMI was also producing the highest quality levels of any domestic automobile plant at the time. [2]

Shook reported that absenteeism, which had regularly reached 20 percent or more under the former operation, fell to a steady level of about 2 percent. He described the plant’s quality as moving from GM’s worst to its best within a year. These figures should be read as evidence of a system-level turnaround, not as proof that culture alone produced every improvement. [3]

A Critical Qualification

NUMMI is sometimes simplified into an inspirational story about keeping the same workers and changing the culture. The reality was more substantial. The plant had been closed and reopened. Management was new. Products, production engineering, training, work routines and labor-management arrangements changed. Employees returning after a long shutdown also had powerful reasons to make the venture succeed. [2]

That qualification strengthens rather than weakens the case. It shows that culture did not change through speeches, slogans or employee engagement events. It changed because leadership redesigned the conditions under which people worked and then reinforced those conditions over time. The system made the desired behaviour easier, visible and consequential.

The Harder Lesson: Learning Did Not Travel Easily

NUMMI demonstrated what was possible inside one plant, but transferring that capability across General Motors proved much slower. Academic analysis of the alliance describes knowledge transfer as difficult and initially constrained by skepticism. GM eventually used managerial assignments, plant visits, liaison structures, leadership involvement and internal learning networks to spread the lessons more widely. [5]

This is the less celebrated but equally important part of the case. An organization can observe a successful practice without absorbing the management philosophy that makes it work. Tools such as visual boards, standardized work or improvement workshops can be copied quickly. The deeper commitments – how leaders respond to bad news, whether employees are trusted to expose problems, and whether management treats learning as part of production – are harder to reproduce.

What Leaders Should Take from NUMMI

Do not diagnose people before diagnosing the system

Persistent underperformance may reflect incentives, work design, weak support, fragmented accountability or a history of management responses that taught employees to remain silent.

Make problems safe to reveal and difficult to ignore

Operational excellence depends on the early visibility of defects, delays and risks. Employees will surface them only when the process is clear and the leadership response is constructive and reliable.

Connect authority with management obligation

Giving employees the authority to stop or challenge work is insufficient if supervisors are not equally obligated to respond, investigate and remove the underlying cause.

Treat standards as the beginning of learning

A standard should create a common baseline, ownership and traceability. It should not become a shield for preserving practices that no longer serve the business.

Transfer systems, not isolated tools

Sustainable improvement requires aligned leadership behaviour, operating routines, measures, skills and governance. Copying a visible technique without its supporting system usually produces compliance without transformation.

The COGMERIX Perspective

NUMMI remains relevant because many organizations still separate culture from operations. Culture is discussed through values and engagement, while operational excellence is pursued through projects, technology and performance targets. In practice, the two are inseparable.

At COGMERIX, we view culture through the operating evidence it produces: how quickly problems become visible, whether their causes can be traced, how leaders respond, whether lessons change the standard, and whether improvements survive beyond the individuals who introduced them. The objective is not to replace the people, systems and knowledge a business already has. It is to create the organizational fit that allows them to work as one accountable, learning system.

References

1. Toyota Motor Corporation. “Joint Venture with GM.” 75 Years of Toyota, company history. Accessed August 3, 2026. Open source

2. Adler, Paul S., Barbara Goldoftas and David I. Levine. “Flexibility Versus Efficiency? A Case Study of Model Changeovers in the Toyota Production System.” Organization Science 10, no. 1 (1999): 43-68. Open source

3. Shook, John. “How NUMMI Changed Its Culture.” Lean Enterprise Institute, September 30, 2009. Open source

4. Toyota Motor Corporation. “Toyota Production System.” Official Global Website. Accessed August 3, 2026. Open source

5. Inkpen, Andrew. “Learning Through Alliances: General Motors and NUMMI.” California Management Review 47, no. 4 (2005): 114-136. Open source

6. Glass, Ira, and Frank Langfitt. “NUMMI (2010).” This American Life, episode 403, March 26, 2010. Included as additional journalistic background. Open source

Publication and Attribution Note

This case study is an independently written synthesis and analysis of publicly available historical, corporate and academic sources. It does not reproduce the text, graphics or proprietary case materials of the cited publishers. Factual claims are attributed through the numbered references above. Toyota, General Motors, NUMMI and associated names and marks are the property of their respective owners. Their use is for identification, commentary and educational analysis only and does not imply endorsement of COGMERIX.

Editorial note: Before publishing, COGMERIX should retain the references and attribution note. Any third-party photographs, logos, charts or extended quotations should be used only with the appropriate licence or permission.

* This case study is an independently written synthesis and analysis of publicly available historical, corporate and academic sources. It does not reproduce the text, graphics or proprietary case materials of the cited publishers. Factual claims are attributed through the numbered references above. Toyota, General Motors, NUMMI and associated names and marks are the property of their respective owners. Their use is for identification, commentary and educational analysis only and does not imply endorsement of COGMERIX.

Editorial note: Before publishing, COGMERIX should retain the references and attribution note. Any third-party photographs, logos, charts or extended quotations should be used only with the appropriate licence or permission.

Businesses can begin an ERP program without fully deciding what should adapt: the enterprise or

Explore Other Successful Projects