America’s Most Successful Export to Japan: Continuous Improvement Programs

Another area in which U.S. firms have often lagged behind their overseas competitors is in exploiting the potential for continuous improvement in the quality and reliability of their products and processes. The cumulative effect of successive incremental improvements and modification to established products and processes can be very large and may outpace efforts to achieve technological breakthroughs.

—Made in America

MIT Commission on Industrial Productivity1

CONTINUOUS IMPROVEMENT Programs (CIPs) unleash employee experience and I creativity to improve both products and processes. They are often cited as the most important difference between the Japanese and Western management styles and as a major factor in Japan’s economic success.2Yet the CIP was conceived, developed, and brought to maturation in the United States. After World War II, the U.S. government helped to export it to Japan, where it was well received and promptly flourished. Despite the long history and well-documented benefits of such systems, few U.S. companies have invested effort in CIPs equivalent to that of their Japanese competitors. Japanese companies have put almost forty years into the development and refinement of CIPs, or kaizen programs as they are known in Japan, and have brought the art and science of managing them to new levels of sophistication. The aim of these programs is precisely to design and implement a system whose natural equilibrium is constant improvement and change. How can a company that does not have such a program compete with one that does?

In this paper we give the historical background of CIPs, which we believe is essential for a useful understanding of these programs. We document their export across the Pacific immediately following World War II and illustrate their power as a competitive weapon when fully and properly deployed. We then identify and discuss requirements for successful implementation and suggest reasons why many U.S. firms find them difficult to start and maintain. We conclude by arguing that what is commonly perceived to be the “best practice” of CIP management is itself open to improvement. Throughout the article, we identify promising directions for companies to pursue so that all minds in the company, not just those of a few at the top, are actively solving problems, reducing costs, and eliminating waste.

The Industrial Revolution

For most of its history, manufacturing was a craft performed by skilled artisans. Apprenticeships were used to pass production methods and skills onto the next generation. The artisans or workers largely determined how work was to be done.

Two important developments during the nineteenth century altered this state of affairs. The first was the increasing standardization of parts to make them interchangeable, championed in particular by Eli Whitney, then a successful rifle manufacturer. The tightly controlled uniformity necessary to make interchangeable components required manufacturing to measurements and standards, which naturally led to a reduction in the artisan’s latitude and authority. The second development was the increased mechanization of skills and division of labor begun during the Industrial Revolution. Tasks that had required highly trained artisans could now be performed on machines by workers with little training or education.

The tradition of the worker determining the method of production remained largely in place, however, even though the workforce was now less skilled and trained. As the pace of change quickened, the level of industrialization rose, and production methods became more complex, the limitations of this state of affairs soon became painfully apparent.

To correct this problem, Frederick Taylor and Frank Gilbreth advocated that the “one best way” to produce should be found through the application of an objective “scientific” method; they placed responsibility for this search firmly on management. The science of work measurement, labor standards, and human engineering began to replace worker determination of production methods. New avenues were thereby established for change and improvement. The scientific method also encouraged standardized documentation of production knowledge for training and dissemination.

Despite offering a powerful method to production managers, Scientific Management, as practiced, did not slow the established trend toward disenfranchisement and deskilling of the worker begun by standardization and the Industrial Revolution.3 The disenfranchisement continued because the source of methods improvement and cost reduction was now (quasi-officially) restricted to management. Those performing the work were told what to do and, often, not to think; the jobs requiring “thinking” were separated from those requiring “doing.” Unfortunately, much of the direct knowledge of the process and shop floor was thus eliminated from decision making.

Henry Gantt, a protégé of Frederick Taylor, noticed this shortcoming. In a 1901 paper on the advantages of Scientific Management presented before the American Society of Mechanical Engineers, Gantt concluded with some thoughts about reintroducing the worker into the improvement process:

Although the results [of scientific management] are very far in excess of that anticipated, the writer does not feel that we have yet taken advantage of all our opportunities. As before stated, the system is one of education, with prizes for those who learn; but the prizes have so far been awarded for learning and doing only what our experts already knew. The next and most obvious step is to make it to the interest of the men to learn more than their cards can teach them. So far nothing has been done in this line, not because the need for some such provision has not been felt, but for the reason that no entirely satisfactory method has suggested itself.

The writer believes in paying a liberal compensation for improved methods of work, and in offering special inducements to a workman to make out instruction cards, by which others are enabled to carry out these methods. The compensation should be sufficiently liberal not only to induce him to part with what information he may have, but to use his ingenuity to devise better methods.4

The reintroduction of the worker into the improvement process marked the beginning of modern CIP management.

Early Continuous Improvement Programs

Denny of Dumbarton

One early system was the “awards scheme” started in 1871 by Denny of Dumbarton, a Scottish shipbuilder.5 He claimed it was the first industrial suggestion system in Great Britain. The following rules for the scheme describe how it worked:

Any employee (exclusive of head foremen, officials of Awards Committee, and heads of departments) may claim an award from the committee on the following grounds:

1. That he has either invented or introduced a new machine or hand tool into the works.
2. That he has improved any existing machine or hand tool.
3. That he has applied any existing machine or hand tool to a new class of work.
4. That he has discovered or introduced any new method of carrying on or arranging work.
5. That he has invented or introduced any appliance for the prevention of accidents.
6. That he has suggested some means by which waste of material may be avoided.
7. Or, generally that he has made any change by which work is rendered either superior in quality or more economical in cost.6

All claims were evaluated by a two-person committee consisting of a member of the engineering department and a person from outside the company. In addition to the cash awards for each suggestion, a premium often pounds (then a substantial sum) was awarded for every fifth suggestion a worker contributed. (In 1921, this scheme was made even more generous.)

National Cash Register

One of the earliest examples of a CIP in the United States was that of the National Cash Register Company (NCR) in Dayton, Ohio, under the leadership of its legendary founder, John H. Patterson7 In 1894, a large shipment of defective cash registers was returned to the company from England. When Patterson, who until then had been almost entirely involved in sales, found that acid had been poured into the mechanisms (obviously by his own employees), he moved his desk onto the factory floor to find the problem. He quickly discovered that the working conditions were unsafe, dark, dingy, and unpleasant, and had lowered morale to the point of hostility. It didn’t take him long to determine what had to be done.

From this experience, he pioneered the “daylight” factory—a well-lit and pleasant facility whose walls and ceiling were 80 percent glass. He installed safety devices, ventilation, bathrooms for use on company time, lounges, and individual lockers. He also kept doctors and nurses on the premises and started a laundry service in the plant to provide clean towels and aprons for employees. A subsidized hot meals program reduced absenteeism enough to pay for itself. A period of mandatory daily exercise was instituted long before this became a fashionable trademark of a “team-oriented” company.

At the same time, Patterson set up a program aimed at soliciting written suggestions for improvement from the factory workers. His idea was that the company pyramid should be supported by the base, not the top, in which case it could be viewed as having a “hundred-headed brain? In 1894, prizes of $30 were awarded for the best ideas, and the factory magazine published the names and photos of the winners. Award amounts grew rapidly in size, reaching $500 in gold by 1897. By the 1940s, the company was receiving an average of three thousand suggestions each year. Patterson had discovered that by recognizing and rewarding improvement suggestions, many problems were resolved with minimal intervention by top management. Some, in fact, were resolved even before management became aware of them.

But NCR’s continuous improvement program was more comprehensive than a good work environment and a suggestion system. Extensive evening (“owl”) classes were offered, for one dollar each, to further the education and development of employees. Subjects included mathematics, blueprint reading, accounting, dressmaking, public speaking, home economics, mechanical drawing, salesmanship, shorthand, typing, machine shop training, and tool design. As employees learned, they became more eligible for advancement.

Although Patterson created a progressive company that provided excellent wages and benefits for his personnel, his strongly dictatorial style of management often made him difficult to work for.8 He readily fired employees; it was said that the words “It can’t be done” were grounds for instant dismissal. Another of his sayings was, “When we get to the point where all depends on one man, let’s fire him.” He even once fired his newly hired assistant controller after just a few hours on the job because during his welcome dinner the hapless employee mentioned that his soup was getting cold as a polite way of turning the conversation over to Patterson. Patterson fired him for showing more interest in his food than company business. Ironically, this side of Patterson’s character accounts for much of his influence on U.S. business. It is estimated that between 1910 and 1930 one-sixth of the nation’s top executives had been trained and fired by Patterson.9 This select but large group included his son-in-law and one-time sales manager, Tom Watson, who later became famous for his prominent role in building IBM.

Lincoln Electric Company

James T. Lincoln of Lincoln Electric Company (the world’s largest manufacturer of arc-welding supplies and equipment) also understood the value of the “hundred-headed brain.” In 1946, he wrote:

Management, if it is to be the best obtainable, must be the collective intelligence of the whole organization. No one man, or even a small group of men, can have sufficient knowledge, experience, and wisdom to make decisions mat can be as sound as they would be if these decisions represented the collective intelligence and experience of the group. The problem is to get this collective intelligence and experience to bear on decisions as they are made.10

To harness this “collective intelligence,” Lincoln designed an “incentive management” system to promote continuous improvement. Its purpose was to foster the ongoing development of the worker’s capabilities and to create desire on the worker’s part to use these abilities cooperatively for the success of the enterprise. Lincoln achieved this through company policies and operating procedures that rewarded workers with compensation in proportion to output and gave them increased status for achievement and publicity for contribution.

In 1915, Lincoln began a piece-rate system, but one with a clever twist. He used the best known work methods to fix job rates, turned these rates into long-term contracts between workers and management, and did not change them when workers discovered more efficient ways to perform operations.11 Therefore, when workers initiated improvements in methods, they were directly rewarded for their ingenuity and increased output. At the same time, the company profited because its fixed overhead could now be spread over the increased production.

In 1929, the company implemented a suggestion system that remunerated employees with half of the savings resulting from any improvement idea during the first year of its implementation. This program was later subsumed by a profit-based employee bonus system begun in 1934. Each year the board of directors set the bonus pool, which was distributed in proportion to each worker’s wages, adjusted by a merit multiplier for dependability, quality, output, ideas, and cooperation. Employee bonuses often approximated the worker’s annual wages. Such a large part of compensation in the form of performance bonuses meant that employees were severely penalized for unanticipated absences, low productivity, and poor quality. On the other hand, they were generously rewarded for methods improvement, high quality, and high productivity.

Despite his belief in management’s ultimate authority, Lincoln succeeded in creating a climate of openness and respect between workers and managers. Workers had full responsibility for their workstations and were held accountable for the quality of their output. Open communication between management and workers was encouraged by managerial open door policies and an employee advisory board elected by the workforce, which met twice monthly with top management to pass on ideas and to identify problems.

Another way Lincoln tried to lower barriers between management and workers was through uniform treatment of all employees. Management had no special benefits such as assigned parking spaces or an executive dining room. Fifty percent of the company’s stock was owned by 70 percent of its employees through a special program. Management offices were intended to be strictly functional and only the president and CEO had private offices, which also doubled as conference rooms. All promotions came from within the company; any employee was entitled to apply for an open job, and jobs were posted on employee bulletin boards. Employees understood that increased productivity would not result in their termination, for Lincoln Electric assured lifetime employment.12

Lincoln also had a unique view on how profits should be distributed. He believed that the people responsible for creating profits should receive the largest share of them. Since customers, more than absentee stockholders, provided funds for company growth and profitability, they were rewarded with low prices and high quality. Because productive workers created products and profits, employees were rewarded with compensation that was often twice that of workers performing similar operations at other companies. It was with some justification that in 1946 Lincoln could boast that “the wage rates of Lincoln Electric Company were higher than those in any other manufacturing activity in the world.”13 Even with such highly paid workers, Lincoln Electric’s high productivity resulted in lower labor costs per dollar of sales than any other company in the electric equipment manufacturing industry.

But Lincoln’s system, perhaps, was too far ahead of accepted business practices of the time. Despite his willingness to share his management methods with competitors to help the war effort, the company’s unique practices and high productivity brought him considerable trouble from the U.S. government during World War II:14

• The Price Adjustment Board and the FBI investigated Lincoln for alleged overpricing even though Lincoln charged less than the government-set prices, and competitors charged full price.
• The Federal Trade Commission brought antitrust proceedings against Lincoln for alleged under-pricing to drive competitors from the market.
• The Vinson Naval Affairs Investigating Committee examined Lincoln Electric’s wage structure, charging that workers were being overpaid.
• The Labor Department claimed Lincoln workers were underpaid because year-end bonuses did not properly reflect overtime and warned Lincoln that he was open to lawsuits from employees. The War Labor Board disagreed and threatened Lincoln with jail if he complied with the Labor Department ruling.

It is ironic that the same government that was persecuting Lincoln for his success with CIP was shortly thereafter exporting and even mandating it to Japanese industry through a U.S. occupation forces-sponsored extension of the wartime Training Within Industries (TWI) service. Imposing the CIP on Japanese industry, while aggressive, was quite definitely not an act of war. The war was over.

The Export and Foreign Development of CIPs

Employee suggestion programs have become a mainstay of many corporate kaizen programs in Japan. Although rare before World War II, suggestion boxes were not totally new there. In 1721, the eighth shogun, Yoshimune Tokugawa, had a small box called the meyasubako placed at the entrance to his casde.15 On this box was written, “Make your idea known. Rewards are given for ideas that are accepted.” Citizens were allowed to make suggestions only in this way, for making policy suggestions directly to the shogun remained punishable by decapitation. Japan’s rulers were not always so open to new ideas, however. During the period of the Kansei Reforms (1789–93), one of the antimerchant decrees was, “You must not invent anything new.”16

An early industrial application of the suggestion box was made in 1905 at Kanebuchi Boseki, a textile firm. A management team had observed NCR’s suggestion box program during an earlier visit to the United States.17

Prior to World War II, Japanese suggestion systems mostly took input from a handful of elite workers thought to have the capability to offer ideas. After the war, however, suggestion systems proliferated rapidly; they expanded to include the entire workforce and were integrated closely with broader continuous improvement efforts. There were two reasons for this surge in popularity. First, they proved an inexpensive means to improve production and reduce costs during a time of severe resource shortages. Eiji Toyoda, the long-serving CEO of Toyota Motor Corporation, explains how his company adopted such a system after his 1950 visit to the United States:

Soon after my return to Japan, payment for the procurements had not come in yet, so management got together and discussed what internal changes the company should be making that didn’t require an input of cash. We decided then and there that we could streamline operations and cut transportation costs. Both could be accomplished without further expenditures. All we had to do was use our know-how. While at Ford, I had seen how considerable savings in manpower could be had in materials handling by judiciously making even minor changes, so we decided to begin there. That’s how Toyota’s suggestion system got started.18

The second reason for the surge in popularity was that the U.S. military occupation authorities, fearing widespread starvation and unrest in Japan if industry was not rebuilt quickly, contracted with TWI, Inc. (owned and managed by Lowell Mellen, who had been in charge of TWI’s Northern Ohio district during the war) to run programs in Japan using ex-instructors from the wartime U.S. TWI service19.” TWI had played a large role in bringing U.S. industrial production from postdepression to full wartime levels in a very short time through the successful training of almost two million supervisors who, in turn, had trained over ten million workers.20 It was hoped that TWI could successfully boost Japanese industry from within.

The suggestion system, and the drive for continuous improvement, grew naturally out of the TWI programs, which consisted of three ten-hour courses known as the “J” modules: Job Instruction Training (JIT), Job Methods Training (JMT), and Job Relations Training (JRT). Each of these programs was structured to produce a “multiplier” effect, which was expressed in TWI’s stated objective to “develop a standard method, then train the people who will train repeated groups of people to use the method.”21

JIT taught supervisors the value of good training and how to provide it. During the war, over one million U.S. supervisors had been JIT-certified. One notable success of JIT was the role it played in reducing the critical shortage of skilled lens grinders early in the war. TWI was able to reduce the time to train lens grinders from five years to two months, thus eliminating the shortage. JIT also passed onto supervisors another valuable tool: the training timetable, a schedule for each worker to acquire needed skills.

JMT taught supervisors how to generate methods improvement ideas, and how to make sure they were implemented. TWI taught managers to write up improvement proposals with expected cost reductions, improved quality, and increased safety all clearly presented. In fact, JMT required a written improvement suggestion from each participant before he or she could be JMT-certified.

The JRT course, developed at the recommendation of the National Academy of Sciences, provided the requisite human relations training to make JMT work more effectively. It taught supervisors both how to deal with job relations problems and how to prevent them from arising in the first place. TWI also found a strong demand for a version of the JRT course from labor unions. Accordingly, it released a Union Job Relations Training (UJRT) course in early 1945. (UJRT was identical to JRT, except that the word “supervisor” was replaced by the word “steward” throughout.)

When the TWI programs were introduced into Japan after the war, it was hoped that the same multiplier effect that had proved so successful in the wartime United States would work in Japan as well and would spread the methods to a wide cross section of industry extremely rapidly. It did. It is estimated that, in 1952, over one million Japanese supervisors were actively undergoing TWI training.22

In this manner TWI had a large impact in training Japanese workers and managers in methods analysis and laid a solid foundation for continuous improvement. Many Japanese today credit TWI with a major role in their country’s success, most particularly with regard to introduction of the notion of continuous improvement.23 The Japan Human Relations Association, in a primer for supervisors on how to generate job improvement ideas, states:

TWI (Training Within Industries), introduced to Japanese industry in 1949 by the U.S. occupation forces, had a major effect in expanding the suggestion system to involve all workers rather than just a handful of the elite. Job Modification constituted a major part of TWI and as foremen and supervisors taught workers how to perform job modification, they learned how to make changes and suggestions.24

The exposure of Japanese management to TWI, the low cost of CIPs, and the experiences Japanese executives brought home from U.S. tours combined to give kaizen programs a good start. Companies that adopted such programs early included Toshiba in 1946, Matsushita Electric in 1950, and, as mentioned above, Toyota in 1951.

Although suggestion systems and boxes became widespread in Japanese companies during the 1950s, at first they achieved only moderate success because of the difficulty in getting group-oriented Japanese workers to contribute individual ideas. Furthermore, many workers had neither the time nor the inclination to write up their suggestions. However, quality control circles, a group-based oral forum for suggestions, began to appear in Japan in the early 1960s. Indeed, QC circles are credited with starting the Zero Defects (ZD) movement at NEC in 1965, where workers were required to make contracts agreeing to produce no defects whatsoever.

During the 1973 oil crisis, kaizen programs again proliferated as companies looked for proven methods to reduce costs without making new investment. (Existing kaizen programs also reflected the crisis atmosphere—Toyota received more than six times the number of suggestions in 1973 than it did in 1970.25) The experience of Canon, Inc. is evidence of this trend.26 There was great consternation when, in 1975, the company posted its first loss since 1949. A full-blown kaizen program was started as a component of a long-range strategic plan to make Canon a world-class company. Over the next ten years the company saved an estimated $200 million in direct costs from employee-generated ideas, for a paltry corporate investment of $2.2 million, of which 96 percent was prize money.27 The indirect benefits of such a program are difficult to assess, but are probably much greater. It would be hard to put a value on reduced defects, higher morale, and a workforce alert for opportunities to improve production methods and eliminate waste.

CIPs Return to the United States

Direct Japanese investment in U.S. companies, coupled with the desire to imitate successful Japanese competitors, has resulted in the widespread introduction (or, often, reintroduction) of CIPs into U.S. companies. Not all of these programs have been successful, leading some critics to argue that they are inappropriate outside of Japan or cannot be effectively introduced into ongoing U.S. manufacturing operations. After all, in foreign countries Japanese corporations are careful to hire young and inexperienced workers and to indoctrinate them weE.28 Consequently, conclusions drawn about the performance of CIPs in these new operations may not apply to their potential in existing companies. An opportunity to observe the impact of a Japanese-style CIP with an existing U.S. facility and workforce arose with the incorporation of New United Motor Manufacturing, Inc. (NUMMI) in February 1984.

General Motors and Toyota established NUMMI, a joint venture, to produce the Chevrolet Nova and Toyota Corolla. It currently produces the Geo Prizm. GM wished to gain experience with the Japanese management style; Toyota wanted an introduction to U.S. workers, suppliers, and management. The CEO of the new company was Tatsuro Toyoda, son of the founder of Toyota. The COO was Kan Higashi, a member of Toyota’s Board of Directors. The plant was to be operated in the Toyota style.

The chosen facility was an old GM plant in Fremont, California, which had been closed in1982 in no small part because of its history of labor-management conflict, poor quality, low productivity, and high absenteeism. At the time of plant shutdown, absenteeism exceeded twenty percent, over sixty firings were in dispute, and there were more than one thousand unresolved grievances.

By agreement with the United Auto Workers Union (UAW), the workforce was to comprise as many members as possible of the original five thousand workers laid off in 1982. NUMMI was able to hire over eighty percent of its workforce from the ranks of former GM employees. In addition, the 1987 NUMMI-UAW contract specifies that “together we are committed to building and maintaining the most innovative and harmonious labor-management relationship in America. The parties are committed to. . . constantly seek improvement in quality, efficiency, and work environment through kaizen, QC circles, and suggestion programs.”29

With this emphasis on kaizen and continuous improvement, the contract also stipulates that it be difficult to lay off workers—it even forbids layoffs owing to productivity increases. Indeed, management must take pay cuts and call in subcontracted work to be performed in the factory before any workers are laid off. A 1984 Fortune article commented:

As a cooperative endeavor between a symbol of Japanese efficiency and a powerful U.S. union, New United Motor is the most important labor relations experiment in the United States today. If Toyota can succeed in producing a car to Japanese standards at near-Japanese cost using unionized U.S. workers, the venture could force profound changes on the rest of the U.S. auto industry—and perhaps on other industries as well.30

Participation in the employee suggestion system at NUMMI grew rapidly, as illustrated in Table 1.31

Dale Buss, an auto industry reporter for the Wall Street Journal, summarized the results in 1986: 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 the Big Three plants in efficiency and product quality.”32

Successful CIP Management

Why have CIPs not been deployed as successfully over the last forty years in the United States as they have been in Japan? Japanese companies received almost 48 million improvement proposals from their employees in 1986, while their counterparts in the United States received about one million. Toyota alone received 2.65 million suggestions from its employees that year.33 Below, we identify some of die requirements for successful CIP management and also discuss why the CIP has not been deployed as successfully over the last forty years in the United States as it has in Japan.

• It must be clearly understood that improvements at first cause dislocation and almost always require time before they prove worthwhile. It is not always easy to link the positive impact of minor improvements to the bottom line. But even if such links could be made, the effect from each improvement is, generally, small, and frequently reduces performance during the learning period. Shigeo Shingo, one of Japan’s foremost manufacturing experts, explains the problem:

Since improvement to a greater or lesser extent demands new procedures, a certain amount of difficulty will be encountered and a certain amount of time will be needed to decide how to carry out those new procedures and then to memorize them. Initially, then, new methods will be difficult. Old procedures, however, are easy just because they are familiar. . . . For instance, as long as it is unfamiliar, even an improved procedure will be more difficult and will take more time than the old procedure because decisions and hesitation will occur and because mental anticipation is absent. Thus, no improvement shows its true worth right away. Its real efficacy will become apparent only after a certain period of practice. This means that 99 percent of all improvement plans would vanish without a trace if they were to be abandoned after only a brief trial. People in charge of plant improvement must grasp this fact.34

Lillian Gilbreth designed a clever experiment to communicate this phenomenon.35 A subject repeatedly writes the words “production engineering” on an index card. Then, the subject is told to write only every second letter, that is, the words “pouto egneig.” This immediately results in a lower rate of output. After ten or so tries, however, the subject acquires the necessary mental anticipation, and productivity doubles.

The experience of Kobayashi Kose, the fourth largest cosmetics manufacturer in Japan, reflects this productivity-reducing initial effect of change.36 In January 1977 the company introduced a Total Quality Control program that, naturally, included continuous improvement. In the first year costs rose by 7 percent and the number of customer complaints rose by almost 25 percent. Three years later, however, costs were down by 12 percent and complaints by 60 percent.

All this does not mean that any proposed improvement should be given years of trial before a final evaluation is made. It does mean that, during the window of learning time, managing the process will require both the ability to look beyond the costs that increase in the short term, and sufficient technical manufacturing knowledge to lead the changes in the proper direction. These two requisites for success are not always present when they are needed. For example, when a Just-in-Time (JIT) production system was recently implemented at a large plant in Western Massachusetts, costs soared, numerous quality and vendor problems surfaced, and top management quickly declared the experiment a failure. Indeed, the person in charge of the program later gave several lectures on why JIT cannot work, despite considerable evidence of its success elsewhere. Regrettably, the firm’s management did not fully understand the costs of learning, or that organizationwide improvement efforts nearly always expose previously hidden or ignored problems.

It is unfortunate that current standard accounting methods, while well suited to measuring the tangible direct costs of improvement, fail to reflect less tangible (but often more significant) benefits such as better quality, higher morale, lower inventories, and shorter lead times.

The wrong signals sent by accounting methods can pose a problem if the company is led by “generic” managers.37 Generic managers are people with “managerial” skills that supposedly enable them to run any company well, regardless of its line of business. Since their expertise is industry-and product-independent, such managers obviously must get their signals from financial and accounting data to a much greater extent than if they had a solid technical understanding of the process.

Strong pressure to produce, when combined with the short-term downturn in performance that any improvement brings, can induce a bias to seek immediate and dramatic improvements, rather than the stream of small improvements that result from a CIP. The effect of a major innovation, after all, can be more cleanly and directly measured than the effects of a large number of incremental improvements. According to Masaaki Imai, a leading Japanese improvement expert, these pressures have led to considerable differences in the way change is managed in the United States and Japan.

Western managers worship at the altar of innovation. … Innovation is dramatic, a real attention-getter. Kaizen, on the other hand, is often un-dramatic and subtle, and its results are seldom immediately visible. While kaizen is a continuous process, innovation is generally a one-shot phenomenon.

In the West, for example, a middle manager can usually obtain top management support for such projects as CAD (computer-aided design), CAM (computer-aided manufacture), and MRP (materials requirements planning), since these are innovative projects that have a way of revolutionizing existing systems. As such, they offer ROI (return on investment) benefits that managers can hardly resist.

However, when a factory manager wishes, for example, to make small changes in the way workers use the machinery, such as working on multiple job assignments or realigning production processes (both of which may require lengthy discussions with the union as well as reeducation and retraining of workers), obtaining management support can be difficult indeed.38

While major innovations can improve performance dramatically, they offer little opportunity for sustained competitive advantage. Their visibility and rapid commercialization by vendors often allow them to be quickly copied by rivals. The incremental improvements of a CIP, on the other hand, originate and remain within a company, and can thus develop into a considerable cumulative advantage that remains safely proprietary.39

• Operating practices that restrict the flow of ideas must be eliminated. One of the aims of a CIP is to allow ideas to flow freely, and to make everyone, regardless of level or function, into a “thinker.” Any corporate scheme or management behavior that divides people into “thinkers” (such as top management) and “doers” (such as workers) will therefore undermine improvement efforts. Although initiatives to bring workers and managers closer together often use words like “trust,” “respect,” “teamwork,” and “worker empowerment,” some of the most prized perquisites and trappings of managerial power have the opposite effect. Imai explains:

The typical Western manager’s behavior is every cultural anthropologist’s nightmare. His behavior is just the opposite of everything a person wishing to establish rapport with strangers is supposed to do. Since the manager views the workplace as a hostile jungle, his office is a well-fortified, plush outpost where he entrenches himself and shuns communication. If communication exists at all, it is at best one way. The manager feels protected by walls he builds between himself and the workers. He often flaunts his status and power in the face of the less privileged workers.

From the color of the shirts they wear to the separate dining rooms they use, workers are constantly reminded that they are a different breed of animal. And yet, management today loves to talk about such ideals as personal fulfillment and the quality of the work environment!40

Compensation practices can also create barriers. One of die clearest ways to communicate the value placed on employees is through the relative size of their paychecks. Excessive disparity in wages and benefits between workers and managers has long been recognized to be counterproductive. In Book 744 of The Laws, Plato proposes that the wealth disparity between the top and bottom layers be no more than five to one, as a means to avoid discontent. Even John Pierpont Morgan observed the disruptive effect of excessive disparity. He noted that his client companies that were doing poorly all were characterized by excessively high disparity in salaries, and therefore they suffered from a lack of teamwork.

It is unfortunate that the trend among U.S. firms is one of increasing wage disparity. A recent study by Sibson & Company, summarized in Table 2, shows that while the average compensation for hourly production workers has grown 57 percent (in nominal terms) between 1979 and 1988, the average compensation of their top managers has grown by 179 percent.

The trend in Japan has been in die opposite direction. In the last sixty years, the disparity between the average compensation of the company president and the lowest-paid employee has been steadily decreasing. In 1927, the corporate president earned, on average, over 100 times the total compensation of an entry level college graduate. By 1980, this ratio had dropped to less than eight to one.41

The way in which management captures the benefits from improvements can also separate workers and managers and reduce the flow of ideas. This can happen in two ways. First, a company can cash in on productivity gains by decreasing the number of its employees. While it is, of course, hard to imagine a manager saying, “Thanks for the improvement, here’s your pink slip,” the practice of rewarding productivity gains by cutting the workforce is quite common. Why should workers make improvement suggestions when these may make their lives harder, or even, in extreme cases, cost them their jobs? Successful CIPs are often found where some form of employment guarantee is in place. This guarantee ensures that employees do not feel that their job security is based on their continued inefficiency.

The second practice common in large organizations is for a company to reward improvement efforts by reducing allocated resources. For example, by basing next year’s budget upon the resources consumed this year, a company rewards increased efficiency with a lower budget. When an organization falls on hard times, it looks for “fat” to cut. Departments that are furiously busy, apparently understaffed, and consuming all the resources available to them cannot, of course (!), be cut, but those that operate at a less frenetic pace with apparent overstaffing and no need for additional resources can, of course, be cut! Yet the former are often those departments that have made much less effort to improve than the latter. The message for those who would implement improvements is clear—it may well cost them resources.

We recently came across a stark reminder of the difference between Western and Japanese attitudes toward people. The retiring president and owner of Eastern Specialties, Inc., a U.S. specialty paper company, recently sold his firm to Marubeni, a Japanese trading company. Speaking at the University of Massachusetts, he was asked why, when there were many bids for his company, did he sell to a Japanese company? His answer: the only bidder that would commit to continued growth and retention of his long-serving and loyal workforce was Marubeni. The retiring president was concerned that all the other bidders wanted only to strip the company of its assets to make a quick profit. Four years after the sale, he told the audience, Marubeni had kept its word.

• Employees must be continuously trained and developed, particularly in techniques of methods improvement. Even companies that consider their employees to be their greatest asset often give more care to the improvement and maintenance of equipment than to the development and training of their workers. The workforce, if not constantly upgraded and improved, can become a company’s greatest liability.

Unfortunately, we often think of training as teaching people to perform a particular job in a prescribed way. Employee development, however, involves much more than this. Employees need to be able to improve the methods by which their work is done. This requires that they learn problem-solving techniques that distinguish among goals, means, and ends.42 The end is the important thing; goals help achieve ends, and means help achieve goals.

The following story illustrates how a focus on the end can lead to a radical change in the means. An NCR salesman in South America was finding it difficult to persuade businesses to buy his cash registers because of different recording customs for monetary transactions. Rather than failing to meet his quota, the salesman took a somewhat unconventional tack. He persuaded the government to pass a law requiring cash registers in all businesses in order to make it more difficult for them to hide revenue from taxation. Needless to say, this change in means resolved the problem and the salesman easily achieved his goal of selling more cash registers, and, presumably, his end of being promoted.43

Extensive training in job methods alone can make employees resistant to methods improvements for fear of making their skills obsolete. Shingo has a name for people who use their knowledge and experience to shut down valid improvement proposals:

Nyet is the Russian word for “no.” A Soviet ambassador to die United Nations was once nicknamed “Mr. Nyet because he invariably said nyet to proposals put forward by Western nations. There are many people like that ambassador who say, “No, that’s impossible,” “No, that’s too difficult,” or “No, that won’t work” to almost all proposed improvements without giving them any real thought.

In any plant there are usually several of these nyet-sayers who always have their reasons for claiming that things will not work. Nyet engineers tend to be well educated and to hold relatively high positions, which makes the problem all the thornier. Any proposed improvement is bound to entail problems, some minor, some major. But saying “this won’t work” all the time will never lead to progress.44

It is no simple task to overcome either entrenched patterns of thinking or the natural tendency to resist change. It is possible, however, to facilitate change through structured group problem-solving techniques (such as brainstorming) and other mechanisms to generate new ideas and give them a fair hearing. This was done, for example, by Shin-To Plastics in Japan, where a special brainstorming room was set up with procedures and rules for idea generation posted on the walls. Included is a notice that a fine of 100 yen (802) is due for each criticism levied in the room, since fear of criticism lowers the willingness to share even the most creative ideas.

Another method that can very effectively break down resistance to change is to expose employees, as much as is feasible, to different ways of thinking and a more complete view of the whole organization. In fact, Toshiko Yamashita, former CEO of Matsushita Electric Industrial Company, calls personnel shifts “the miracle drug of productivity,” and regards them as central to his company’s program of continuous improvement:

To cite an extreme example, when I was running the air conditioner division, I switched the chiefs of manufacturing and sales. I wanted to broaden them and thought they both could handle an entirely different situation. The sales chief, who had regarded his counterpart as some kind of grease monkey utterly ignorant of how hard it was to sell his finished product, now had to deal with on-time delivery and tight specifications. The engineer discovered the world wasn’t full of customers lined up to buy his machines. It was a very successful move; soon I had two paragons of cooperativeness working for me.45

To give its transfer policy teeth, Matsushita requires managers to have served in three divisions for each promotion, To prevent superficial “ticket-punching,” transfer policies stipulate an A-rating in annual evaluations for eligibility to transfer out of a division. The requirement for high ratings also keeps managers from transferring their poor employees and retaining their best. In 1978, three thousand managers and staff members changed jobs under this program at Matsushita.

• A continuous improvement effort needs an efficient mechanism to handle improvement ideas. Without a well-planned means to gather, evaluate, implement, and reward improvement ideas, no CIP will succeed. Many companies operationalize their CIPs using some form of employee written suggestion system.

Several characteristics seem to be common to effective suggestion systems. They are simple to participate in, and workers are actively encouraged to submit suggestions. All ideas, no matter how trivial they appear, are subjected to a structured evaluation by technically knowledgeable judges who then give employees rapid feedback. Accepted ideas are implemented as quickly as possible. Rewards, whether in the form of recognition or merit, are clearly linked to the quality of ideas. Ideas that are not accepted are used as opportunities to develop employees by explaining why in a manner that expands the contributor’s knowledge of die operation.

This last point, unfortunately, is often forgotten. Recently, one of us was advising a company that wished to increase the involvement of its employees in improvement activities. At the initial meeting with first-line supervisors, a newly promoted supervisor announced to the president that, based on his workgroup’s preliminary research, the purchase of a new piece of equipment would pay for itself in less than a year and greatly improve performance. The president exploded, saying, “I told you already that that idea just won’t work!” While he was, in fact, correct, the president missed an excellent opportunity to develop a new supervisor’s broader understanding of the system. He could have taken the time to explain, or given the employee an assignment that would allow him to discover, that the actual price of the new equipment was only a small part of the total cost of putting it into use. Among other things, it had to be installed, staffed, integrated into the rest of the operating system, and hooded for ventilation. Rather, the result of the outburst was a supervisor who did not volunteer another word at the meeting and an environment of skepticism about management’s true receptivity to input. The lesson is that all suggestions represent opportunities and require response.

Canon’s kaizen program includes an exemplary employee suggestion system. Worker-submitted proposals are initially judged by a factory committee that immediately determines whether or not they merit a participation award (75 yen or 50¢) or one of the first four levels of awards (ranging from 150 yen to 3,000 yen or $20). Suggestions deemed wormy of higher ratings are then forwarded to a higher committee that grants awards ranging from 5,000 yen ($33) to 50,000 yen ($330) or even, in special cases, air tickets to travel abroad. At the end of each week, the supervisors pass down the production lines distributing checks for the previous week’s suggestions. In addition to cash, each proposal is awarded points that apply toward awards for annual and lifetime point totals. Each year the twenty people with the highest lifetime totals receive 300,000 yen ($2,000) and the twenty people with the highest annual totals receive 100,000 yen ($660). Over 90 percent of the Canon proposals are implemented; indeed, most are not formally proposed until they prove successful.

Offering monetary rewards is not the only way to elicit suggestions from employees, of course. In fact, some argue that it is not appropriate to offer extra money for suggestions. Almost fifty years ago, the TWI service argued that written suggestions should be part of the job, and not something voluntary that management should wait for and be grateful for. In fact, during the war, when the War Production Board was requested to rule on whether or not suggestions should be allowed to earn extra money, TWFs official position was that suggestions and improvement ideas should be part of the job. The board ruled against TWI, and allowed suggestions to be rewarded with money, as long as the rewards were commensurate with the value of the suggestion. This latter stipulation was to prevent companies using suggestion awards to pad the wage packets of their top producers and thereby circumvent the tight wartime wage controls.

Conclusion

Without some form of CIP, it is hard for a company to keep up with a competitor that has one. This seems, perhaps, an obvious assertion, but we have recently observed or participated in the installation of a number of CIPs that were implemented at great expense, with extensive retraining, and accompanied by much fanfare, only to wither into perfunctory exercises as soon as it became obvious to the workers and lower management that an uncommitted upper management was merely introducing the program to be fashionable, or in response to corporate or customer pressures. Frederick Taylor understood the reasons behind this type of failure:

The mechanism of management must not be mistaken for its essence, or underlying philosophy. Precisely the same mechanism will in one case produce disastrous results and in another the most beneficent. The same mechanism that will produce the finest results when made to serve the underlying principles. . . will lead to failure and disaster if accompanied by the wrong spirit in those who are using it.46

It is easy to adopt the tools and management mechanisms of a CIP. It is much more difficult to live up to the underlying philosophy. Without a solid foundation, a CIP can easily turn into mere superficial sloganeering, which can have the opposite effect to that desired.

A strong CIP is a key component for the long-term competitiveness of a company. Lincoln Electric Company, for example, remains the dominant manufacturer of electric arc welding equipment, with a 40 percent U.S. market share. The incentive management system, with its important component of continuous improvement, has changed little over the years. The company still considers its employees to be the highest paid production workers in the world. Yet its high productivity has enabled it to prevent some very aggressive Japanese competitors from gaining significant market share. In fact, Lincoln recently opened a cored welding wire manufacturing plant in Japan.

Although the nature of continuous improvement is to be patiently incremental, a company need not have a long-established program such as Lincoln’s to notice the resulting benefits. The turnaround of the Harley Davidson Company is largely the story of a group of managers who, upon competitive benchmarking of their production facilities against those of Japanese competitors (who were driving them out of business), realized that they needed to install a CIP immediately.47 In 1984 Harley Davidson came four minutes away from bankruptcy. Today, it is again the dominant manufacturer of large motorcycles in the world. The power of a successful CIP is, almost, as simple as that.

Apart from the fact that it is voluntary, and therefore immune to management by objective, the written suggestion system has one other obvious weakness. This weakness is the overhead it generates—paperwork, committees, and the administration of the reward scheme. There are ways to eliminate such overhead and still retain a viable and vigorous atmosphere of continuous improvement. For example, National Cash Register dropped its written suggestion system in 1974 when the company went through a serious crisis because its electromechanical technology became outdated. The speed of change demanded by the new environment required faster and more directed continuous improvement efforts. As the company rapidly globalized and decentralized to restructure its technological base, NCR pushed operating decisions down to the shop floor by giving increasing authority to its workforce. Cells of workers were made responsible for their tasks, including the continuous improvement of products and processes. With this responsibility and power now well in place, work cells generate, evaluate, and implement ideas rapidly. A large part of each worker’s compensation depends on the productivity of the process he or she is in charge of, or “owns.” The goal is, like TWFs was fifty years ago, to have a workforce that can respond rapidly to changing situations and take initiative independent of the central management structure, and without a lot of overhead. John Patterson, NCR’s founder, would have been proud.