Making Organizations Adaptive to Change: Eliminating Bureaucracy at Shenandoah Life

John B. Myers

Today there is an epidemic that threatens to destroy or, at the minimum, severely cripple many business organizations. The offending pathogens have been isolated and identified. They are well known in the business community and are typically referred to as “rapid change” and “increasing complexity in the environment.”

It is not as clear why some organizations fall prey to these pathogens, while others flourish or seem quite unaffected in their presence. A common and significant error made in attempting to understand this phenomenon is to dissect and analyze the pathogens’ structure and physiology without examining their unique interactions with the structure and physiology of the organizational organisms they infect.

The organizations that are the most susceptible to these pathogens appear to be those with the generic properties of a bureaucracy. In the U.S., the term bureaucracy has taken on a negative connotation suggesting the anxious visions of endless red tape and bungling often associated with government agencies. While most business organizations would be insulted to be called bureaucratic, many in fact do operate bureaucratically. This point can be seen more clearly by looking at a bureaucracy in descriptive rather than evaluative terms.

Characteristics of Bureaucracy

Descriptively, a bureaucracy’s structure and physiology can be characterized by several traits:

A hierarchically arranged, functional, and highly specialized structure. Accountants work with accountants, engineers with engineers, manufacturing people with manufacturing people; and in the life insurance industry, which I will be focusing on, underwriters work with underwriters, premium accounting people with premium accounting people, and policyholder clerks with policyholder clerks. Within these functional units, employees are most often arranged in highly specialized, distinct jobs. Each function is directed by one person who, in turn, reports to a boss who manages similar functions.

A defined chain of command, usually on the vertical plane. Communications are supposed to flow up and down in each functional organization. Except in the most routine matters, horizontal and diagonal communications across functions are not supposed to occur, as such contacts would destroy the order that bureaucracy is designed to maintain.

Clearly defined and evaluated job duties and accountabilities. Nearly everyone has a written job description that has been rank-ordered against other job descriptions. Each rank order is assigned a salary range. Employee behavior is directed toward carrying out the duties specified in the job description. In some organizations, performance standards or measurable objectives have been established for many of the specified duties. The addition of duties and responsibilities generally increases a job’s value and prestige. Since most rewards are linked to the enhancement of the job, behavior usually is oriented toward the best interest of the job and the incumbent rather than the organization.

Extensive use of rules, records, and precedents. To help maintain order and consistency, lists of standard operating procedures, employee handbooks, and various policy manuals proliferate in every comer of the organization. When someone does something novel, there is a knee-jerk reaction in which bureaucrats lunge for their manuals, hoping to find guidance therein.

An emphasis on length of service rather than competency. The amount of verbiage in employee handbooks describing the programs that reward length of service far exceeds the amount of text that describes reinforcements for competency-based behavior, such as creativity or extraordinary service to a customer.

An authority system based on position rather than expertise. Authority resides basically in the position rather than the person. There is an assumption that the incumbent in a particular position has expertise because the person has been screened for the qualifications specified in a job description. This assumption is, however, more wishful thinking than reality. Qualification specification and screening processes are not perfect. Errors are made and are not corrected quickly in most bureaucratic organizations. The net result is that one often finds oneself working on a problem with a person who has the assigned authority and responsibility for solving the problem but has neither the expertise nor the ability required.

The preceding is not intended to be an exhaustive description of the bureaucratic organism but rather a rough sketch of its morphology. Nevertheless, these characteristics do indicate the essential reason why environmental change and complexity pose such a threat: The bureaucratic organization’s interests are so vested in “what was” and “what is” that it cannot become flexible enough to meet the challenges of change and complexity.

What is needed is a vaccine that will reduce the morbidity and mortality rates for bureaucratic organizations. Shenandoah Life Insurance Company has been working with a particular vaccine for the past two years. So far, it is proving to be an effective immunizing agent against the ravages of the change and complexity pathogens. The vaccine is known at Shenandoah Life as “organizational experimentation.”

A Changing Environment

The change and complexity pathogens began their attack on Shenandoah Life during the late seventies and early eighties. Among the most virulent of these pathogens were the rapidly rising interest rates, which caused a massive drain of the company’s cash assets. Policyholders borrowed from their whole-life policy cash values at guaranteed, relatively low interest rates and placed these funds in instruments offered by other financial institutions that paid significantly higher interest rates.

Continuously rising interest rates led the company to develop and market a new base-life product with greater interest-earning potential than whole-life insurance. This product, known generically in the industry as “universal life,” has effectively countered the interest rate and borrowing problem. However, the product carries with it some side effects that we have not been able to ignore. The universal-life product has much narrower profit margins and is more difficult to administer than whole-life insurance. Since universal life now accounts for eighty-five percent of the company’s new individual policy business, the pressures to control costs relative to sales have taken on new and more intense meanings than those experienced prior to the late seventies.

Our efforts to control costs have been complicated by the competing need to improve our service to agents. Competition from life insurance companies, as well as from nontraditional sources, such as full-service financial institutions, has been increasing. Countering these moves on our markets has required an acceleration in product development and the acquisition of state-of-the-art computer technology. Without these measures we would not have been able to offer our agents competitive products on a timely basis. Acquiring product-development specialists, computer hardware, computer software, and computer specialists has been a sizeable investment relative to present sales.

We have also noticed an increasing demand for improvement in the paper-processing service we provide to our agents and policyholders. Timeliness and accuracy in paperwork saves them valuable time and reduces their frustrations. Since competitors can easily examine our products and quickly develop their own competing ones, the provision of high-quality and distinctive service in this dimension may prove to be one of a few significant ways we can differentiate ourselves on a continuing basis in the marketplace.

Like our markets, our employees’ needs and values have also been changing. A 1979 employee-attitude survey revealed that our home-office clerical personnel were less satisfied with supervision, job duties, workgroup climate, communication, and training than they were in 1969 and 1972. In fact, personnel reported greater dissatisfaction with job duties, work-group climate, communication, and supervision than their peers in other insurance companies.

These changes in economic and market conditions and in employee attitudes, combined with a rising consumerism and a more difficult regulatory environment, prompted the Shenandoah Life management to start searching for something that would help it manage its business and productivity more effectively.

A Sociotechnical Analysis

Between 1979 and 1982 we identified the need for improved management training and communication skills. During these three years we implemented various managerial and supervisory training programs, as well as an active and successful quality-circle program. In the fall of 1982 we launched a series of two-day seminars on sociotechnical organization design for the president, vice presidents, managers, supervisors, and various technicians. Similar one-day seminars on the same subject were provided to many of the clerks who would eventually be affected by a sociotechnical organization design.

One of the key parts of these seminars was a sociotechnical analysis of one Shenandoah Life case: the processing of a universal-life application that involved the termination of several whole-life policies, refunding a portion of the cash values to the policyholder, rolling over the other cash values into the universal-life policy, setting up an automatic bank draft, underwriting the increased amount of life coverage, and adjusting commissions. This situation was considered to be a fairly typical one, similar to approximately twenty to thirty percent of the cases our agents were sending us at that time.

The results of the sociotechnical analysis of this case indicated that in the technical realm an on-line computer system was available for processing the case. Thus, it was feasible for any single employee (given the proper training and authority) to alter those parts of the computer records that were necessary to complete this transaction. In the social realm the analysis revealed that the clerks responsible for processing this case were arranged in highly specialized, assembly-line types of jobs in premium accounting, policyholder services, general accounting, marketing services, policy records, underwriting, receiving, and reinsurance. The management structure above and around these clerks was essentially bureaucratic. Before this case was completed, it passed through thirty-two sets of hands, across nine sections and three departments. The processing required twenty-seven working days to complete.

The following conclusions were reached in the seminars:

The current and projected future computer technical systems do not require sequential processing along an assembly-line in a highly specialized, fragmented work force. Indeed, such a social system cannot utilize effectively the total capacity of these technical production systems.

The assembly-line social system could perhaps be improved—but not enough, in management’s view—to serve the marketplace competitively or to utilize fully our technical systems. Further, this type of social system requires a large number of supervisory personnel to direct and coordinate all of the specialists. This is a relative waste when one considers that the supervisors could be devoting their time to projects with a higher potential for adding value to the organization.

The rapidly changing conditions that prompted the company to develop and market a universal-life product would continue to require development and marketing of new products as well as enhancement of existing ones. At the same time the company would have to do a better job of processing these new and improved products because of policyholder and agent demands for better service.

The specialized, fragmented organization is a slow learner and therefore cannot adapt rapidly and effectively enough to meet the challenges that the environment imposes on the company.

A Self-management Team

A key part of the two-day seminar was the introduction of the concept of self-managing teams as a structural form that could be more efficient and adaptable than the specialized, fragmented assembly-line approach. In 1983 Shenandoah Life formed its first experimental self-managing team.

The Team’s Characteristics

The team was composed of six nonexempt clerical personnel. They were responsible, as a group, for processing all universal-life business and health insurance generated in Virginia. The phrase “as a group” is stressed here to emphasize that management stipulated no individualized job assignments or accountabilities. The team’s processing responsibilities included all tasks related to premium accounting, policy issues, and policyholder services. In addition, the team was responsible for constructing a model office for a computer-system conversion project and for building the model and systems to be used in developing future teams. Members of the team designed their own office layout, for example, as well as aspects of their salary system.

The team members were also instrumental in the design of what is commonly referred to as a pay-for-learning system (although it was management that decided upon the use of such a system). In this system a member is eligible to receive a defined pay increment after demonstrating competency at a particular task. The team as a whole determines which members are competent at particular skills, and the team initiates salary increment requests. Members whom the team deems to be unsatisfactory learners and contributors do not receive salary increments. All team members are permitted and encouraged to achieve the top pay rate, which signifies the ability to process any piece of business from beginning to end without guidance or consultation. Team members are responsible for training one another.

To facilitate learning and cross-training, the team designed its office layout so that the members could face each other, in conference style, throughout the workday and could communicate informally and rapidly.

The team members’ jobs immediately before joining the team were in premium accounting, policy issue, and policyholder services, with two members coming from each area. They were elected for membership in the team by their peers.

Management expected the team to be as self-regulatory as possible. For example, a supervisor was not appointed. The team did have several advisors with whom they could consult. These advisors included the second vice president, human resources; the personnel director; the manager of methods and procedures; the managers of premium accounting, policy issue, and policyholder services; and a methods and procedures analyst. Team members were also permitted to contact directly the external consultants who did most of their group-dynamics training.

The self-regulating activities performed by the team included deciding who would know what work, formulating a vacation schedule that minimized the negative impact on production, final selection of new team members, designing the office layout, training one another or arranging for training, determining the basics of the salary system, managing the interface with other units and their managers, and correcting members who were not contributing their fair share to the team’s performance.

As an aid in carrying out managerial responsibilities, team members received extensive training in how to teach and how to be good learners. Group dynamics training in brainstorming, conflict resolution, decision making, negotiations, planning, and team functioning were also provided.

The Rationale for the Team Structure

This approach to structuring the experimental team has engendered a range of reactions. One person angrily labeled the team “social elitism”; others praised it for being “highly innovative.” But since labels and evaluation terms contribute little to knowledge and understanding, it is important to explain the rationale for the way the team was structured.

Jobs were not assigned because management wanted to reduce as much as possible the probability that members would serve the interests of their individual jobs rather than the organization’s goals. If there are no jobs, job-oriented behavior becomes impossible. A common mission—service to a geographic market—replaced the concept of job.

One reason for substituting a pay-for-learning system for the traditional job-evaluation approach is the belief that creativity is an essential ingredient in effective adaptation to a rapidly changing, complex environment. By arming all employees with increasingly broader knowledge and skills, one increases creative potential—particularly in solving problems affecting the larger systems of the organization. By contrast, the traditional approach, which usually involves training in increasingly narrow specialties, makes people incapable of using their expertise appropriately within the context of larger system needs. Further, in traditional organizations there are usually no rewards for expanding one’s knowledge and skills beyond the assigned specialty.

Another reason for implementing a pay-for-learning system is that it helps people become more flexible, which in turn permits the more effective utilization of manpower—another important factor in adapting effectively to a changing, complex environment. The highly specialized assembly-line approach to processing is efficient as long as the documents being processed flow evenly and predictably.

But assume that a company has an assembly-line approach to the functions of policy issue, premium accounting, and policyholder services and that at a particular time the volume of work increases significantly in premium accounting and policy issue while the volume in policyholder services declines. The traditional organization can work overtime or hire temporary help from outside to try to move the extra volume of work in premium accounting and policy issue. Hiring temporary help is usually not a viable alternative, though, because it is not economical to train these people to do anything more than the simplest work. Rarely would it be feasible to move people from policyholder services to the other two areas, because transferees present the same problem as temporary help—they are not trained to do the more complex among the tasks in premium accounting and policy issue. In addition, their salaries are usually too high for them to do the simpler work within those areas economically. Furthermore, they may perceive the change to such work as a decrease in status. Finally, the supervisor of policyholder services might not wish to release any staff for fear that someone might think his or her area is overstaffed. The supervisor’s rewards come from enhancing his or her own job performance, not the performance of others.

The mature, self-managing team does not encounter these difficulties because the number of people available to work on any single task is not limited by relative salaries and status, supervisors’ interests, and, most important, training. For example, the number of people working on premium accounting tasks in Shenandoah Life’s team of six can range from zero to six in a matter of seconds. As the market environment changes ever more rapidly and unpredictably, an organization’s effectiveness will depend on its ability to shift people from regular tasks to unforeseen problems imposed by the changing environment.

The bureaucratic organization is designed to prevent mistakes and to maintain the status quo. One follows the chain of command in these organizations regardless of whether doing so helps to get the job done. For the rules say that this must be so, and everyone has a vested interest in maintaining these rules. The notion of trying to make this form of organization more adaptable is a contradiction in terms—the structure is built to resist change. Therefore, it seemed reasonable to set up the Shenandoah Life team without supervisors, rules, or defined communication channels so that (1) the team and individual members could change as they grow and learn, (2) management could discover the extent to which people can manage themselves, and (3) rapid adaptation to various work situations could be made.

Since the bureaucratic organization is designed to maintain the status quo, it will attack and attempt to destroy an experiment, for successful innovations are generally perceived to lead to change. Thus, any organization attempting to innovate should enlist as broad a participation as possible in implementing the change in order to protect the experiment. The organization will then come to own and embrace what it would normally attack and reject. Shenandoah Life’s decisions to have the peers select the team members, to assign a group of advisors, and to train many employees in sociotechnical organization-design concepts were a way of asking for participation in the change and the consequent decision making.

A final point in the team structure rationale is that direction and control of activities by functional and technically competent experts (supervisors) are wasted actions if the people being directed and controlled can manage themselves. Managerial and supervisory training was provided to increase the probability that the team would have the skills required to manage itself and its boundary interfaces.

Results

The team has now (1986) operated as an experiment for about a year. The problems that have developed have been relatively minor and, on the whole, expected. The source of these problems has been identified and corrected. In late 1984 the remaining employees in policyholder services, premium accounting, and policy issue were placed into teams serving geographic markets. The experimental team continues to service Virginia.

Figure 1 shows the organization of the policyholder services, policy issue, and premium accounting groups prior to the experiment and after everyone was placed in a market-service team. The Advisory Team consists of the former managers of premium accounting and policyholder services, the methods and procedures analyst who served the experimental team as an advisor, and a management trainee who had been working in policyholder services. The manager of policy issue resigned during the experimental phase and was not replaced. The general manager is the former manager of methods and procedures. He was also an advisor to the experimental team. At present the company does not have a methods and procedures section. Much of this section’s work involved solving work-flow problems between units and jobs. Since many units and jobs have been combined, there is much less flow of work across job and unit boundaries. Thus, the necessity of such a section has become questionable. Furthermore, by having the market-service teams report to the general manager, the supervisor/employee reporting ratio has been reduced from approximately 1:7 before the experiment to 1:37 in 1986.

The Advisory Team has two missions. One is to help the teams to become self-managing as quickly as possible. The other is to serve as role model for the market-service teams, demonstrating how teams should operate. Since the market-service teams do not report to any of the advisors, and the advisors and market-service team members do not have assigned duties and are cross-training one another, the stage has been set for making labor cost improvements when opportunities occur. For example, when a manager resigns, as did the manager of policy issue, the probability that the person would not have to be replaced is greater in this type of organization. Had traditional bureaucratic assumptions prevailed at the time this manager resigned, a replacement would, in all likelihood, have been hired. A bureaucracy assumes that people and units cannot manage themselves and must have an appointed leader.

In addition to positioning itself for direct cost improvements, the stage has apparently been set for productivity and service improvements. Compared with 1982, the number of people in these three functions has remained approximately the same while new business applications have increased thirteen percent. Beginning in September 1984, when everyone was placed into teams, there was a decline in the number of complaints about service and errors. Even though these data are encouraging, they must be viewed as indicative rather than conclusive. There are many intervening variables that could have a role—either positive or negative—in determining the figures. For example, during this period we have introduced new products and computer systems. Their impact on productivity has not been measured and evaluated.

The most important evaluation to be made is whether the experimentation vaccine has immunized our bureaucracy against the change and complexity pathogens. If a vaccine is to be effective, it must alter the chemistry of the organism into which it is injected. Therefore, the key question is whether experimentation is changing our bureaucratic structure and our ways of responding to environmental change and complexity.

The answer is a definite “yes.” Experimentation has become common. We are experimenting with several other exempt and nonexempt salary systems besides pay-for-learning, within the same home-office building. One professional group has formed into a team and has developed a unique peer-evaluation method and process. Other clerical teams, like the market-service teams in individual insurance services, have been developing in marketing services, group underwriting and issue, group administrative services, annuities, and underwriting services. At present an experimental team of professionals from different disciplines is being formed to manage all aspects of a particular product line. This represents a radical departure from the traditional specialized and functional approach to organizational structure.

The primary weakness of a bureaucracy is that it is prescriptive by nature. It defines its structure and methods as ideal. Thus, there is no reason for it to experiment with alternative forms of organization. The fact that Shenandoah Life is experimenting means that it is no longer a bureaucracy but an organization that is continually seeking new ideal states so that it can operate effectively in a rapidly changing, complex business environment.

Note

This article originally appeared in National Productivity Review, vol. 4, no. 2, Spring 1985, copyright Executive Enterprises, Inc.

~~~

When I read this paper I am reminded of a term that has recently entered the literature: learning organization. Any organization that, as described here, changes its structure and adjusts its activities based upon its own experience is worthy of the title.

Although John Myers is no longer with Shenandoah Life (he is currently vice president of human resources for Eclipse, Inc., an organization in the Midwest), the company still operates the teams, changing their functions as circumstances vary. And the company is still quite successful, People from other companies, and from other countries, continue to call. They ask to visit and learn how Shenandoah Life designs and trains its self-managing teams.

Jack Cochran, a vice president in customer services, has been responsible for these work teams since John left. He writes, “In 1987 our company was a runner-up for the United States Senate Productivity Award in the Private Service Sector category for Virginia. A recent customer survey rated our Customer Service teams on timeliness (team response time), accuracy (team error rate), and overall team services. Ninety percent of the customers rated Shenandoah very good. In 1988 … we were awarded first place by the U.S. Senate in the Private Service category.” SSG.