Learning From Failure in an Organization Designed to Develop Success

Beginnings of NTL Through the Early 1960s

In 1946, in a story very often told (Back, 1972; Freedman, 1999; French, 1982 Hirsch, 1986; Jones & Brazzel, 2014, Kleiner, 2008; Marrow, 1967), the Connecticut State Inter-Racial Commission created a conference to address racial and religious prejudice. They invited Kurt Lewin and others, including Leland Bradford, Ronald Lippitt, and Kenneth Benne, to lead it. Standard training activities took place during the day, but in the evenings staff members discussed the days’ events, including participant behaviors they had observed. The participants started joining in these meetings, heard others’ feedback and responded to it. This was very energizing for the participants and staff alike, and led to the creation of sensitivity training, or T-groups (see Highhouse, 2002, for more information on T-groups).

Following on this experience, Lewin, Benne, Lippert, and Bradford founded NTL in 1947. It was incorporated in 1950 as a division of the American Education Foundation, which gave NTL a somewhat stable financial foundation (Kleiner, 2008). NTL established a program headquarters (Freedman, 1999) at Gould Academy in Bethel, Maine, where it conducted T-groups and trained NTL members to conduct them. (Its administrative headquarters were in Washington, DC.) Lewin died in 1947, but Benne, Lippert, and Bradford continued the work, with Bradford as Director.

T-groups became a flourishing enterprise in Bethel. For example, Warren Bennis and Biederman (2010, p. 86) wrote that

In 1955 Lee Bradford wrote to invite me to join Bethel’s summer staff. By then, almost everyone studying group dynamics wanted to go there. NTL had refined the T-group and made it its principal tool for studying how groups work. . . . It was an honor to be invited to Bethel. An inner circle, called fellows, decided who would lead the groups and made other major decisions. Getting fellow status at Bethel felt like a social science knighthood.

In the mid-1950s, NTL started adapting T-Groups for work with organizational leaders through “Key Executives” labs conducted for presidents and vice-presidents from industry (Freedman, 1999; Highhouse, 2002; Hirsch, 1986). Furthermore, in the 1950s and 1960s several consultant-academics associated with NTL began to formulate a conceptual basis for a new type of consulting to organizations that had some commonalities with T-Groups and was more process oriented than expert oriented (cf. Freedman, 1999; French, 1982; Kleiner, 2008). This group included Douglas McGregor, Chris Argyris, Roger Harrison, Herbert Shepard, Robert Blake, and Richard Beckhard, among others. The name it became called, “Organization Development.” emerged independently from work by Shepard and Blake at Esso and Beckhard and McGregor at General Mills (Freedman, 1999, p. 131).

Following on these developments, in 1965 NTL “established a Presidents’ Conference on Human Behavior for top-level executives” (Hirsch, 1986, p. 77). It also began publishing The Journal of Applied Behavioral Science (JABS). Burke (2015, p. 20) commented that

No doubt the greatest period of growth for the field of OD was the decade of the 1960s and NTL was highly instrumental in spearheading that expansion. Much of the learning in those days came from sharing failure experiences. Much of the learning was also learning from scholar practitioners who served as staff from the Program for Specialists in Training and Development (PSOD)

This group included not only Burke, but also Richard Beckhard, Fritz Steele, and visiting staff members such as Herbert Shepard, Chris Argyris, Sheldon Davis, and Warren Bennis.

Burke (2015, p. 18) had “joined NTL in the summer of 1966 and was immediately sent to Bethel, Maine, to run T-group laboratories.” Over time he also became responsible for additional NTL initiatives, including Management Work Conferences, the OD Network, the Key executive conferences and, later, the Presidents’ conference for CEOs. During the late 1960s, NTL’s work with managers was substantial and successful. The entire work of NTL, not only OD but also personal development training, was successful enough that in 1967 NTL was independently incorporated, with Lee Bradford continuing as Executive Director.

Internal Challenges at NTL

Despite its successes, by the late 1960s there were multiple internal conflicts within NTL. Freedman (1999, p. 133) claimed that “The year 1966 was the beginning of a moral, ethical, and financial crisis within NTL.” One of its crises had to do with the composition of the Board of Directors. The Board “was overwhelmingly male, white, and over 40” (Hirsch, 1986, p. 86). Another had to do with civil rights issues. Freedman (1999, p. 133, see also Back, 1972) described how in the summer of 1969, non-White participants in NTL programs at Bethel confronted white NTL staff about what they claimed to be the Institute’s “institutionalized racism and sexism.” Furthermore, in 1970 women also attacked all-male privilege at NTL Hirsch (1986, p. 87).

Leadership Succession

In 1968, Bradford, who had been the Director of NTL since its beginning, told the Executive Board that he would retire as of August 1, 1970. This set into motion an elaborate process of selecting his successor. Kleiner (2008, pp. 200-201) noted that there were many reasons for Bradford to retire in addition to the internal issues surfacing at NTL. One was that he had several medical problems. Another was that “Income was high, but expenses continued to rise, in part because the increased enrollments had prompted $250,000 worth of renovations to [NTL headquarters] up in Bethel.”

A nominations committee named three possible successors to Bradford. From these possible successors, Warren Bennis, who had been involved in important NTL initiatives for years, and who also had considerable university administrative experience (Back, 1972) was the overwhelming first choice of everyone. Thus, the Executive Committee did everything it could to hire Bennis as the next Executive Director of NTL.

As summarized by Hirsch (1986, p. 81), Bennis expressed some willingness to take the job, contingent on several conditions. The most important was that “the Executive Committee and the Board agree to the creation of an NTL University of Applied Behavioral Science to be located in or near Washington DC.” Building such a university would require between 5 and 15 million dollars. A million dollars must be raised right away, earmarked for an endowment. Bennis told the board that “they could not think small” (Kleiner, 2008, p. 203).

The executive board accepted his conditions, in part because Bennis provided a broad and exciting vision of NTL. This vision allowed NTL to focus on something bigger “that would allow for subsuming differences among its internal constituencies” (Hirsch, 1986, p. 85). “To show Bennis that the million dollars could be raised, the board began recruiting financial pledges within a few days” (Kleiner, 2008, p. 204). It also took on considerable debt to purchase land for the University (Freedman, 1999). Bennis gave an inspirational speech to the Board during its November 13-14, 1969 meeting about how the University “should be the center for social change in our society” (Hirsch, 1986, p. 83).

While this was transpiring, Warner Burke (2015) “was asked by Warren to be his #2, more or less a role of chief operating officer. I was excited to say the least. What excited Warren was the promise by the Board of purchasing property near Dulles Airport to build an NTL campus.” (p. 2)

At the same time he was expressing outward enthusiasm, however, Bennis was privately developing doubts about whether NTL was “sufficiently committed to a new program” (Back, 1972, p. 73). "I felt that NTL and T-Groups were never going to really make a difference,” he said years later (Kleiner, 2008, p. 205; see also Bennis & Biederman, 2010, Ch. 4). Furthermore, Kleiner (2008, p. 205) reported that “while Bennis was being recruited, ‘Privately, staff members and trainers complained that the organization was sliding downhill fast. Then the same people stood up in board meetings and said,’ The organization is in great shape.” As Kleiner noted, “This was a terrible warning sign, because people came to T-Groups to cure exactly that sort of duplicity.”

It seemed to at least some on the executive committee that Bennis had formally accepted the board’s offer. Bradford stated that Bennis signed a letter of acceptance (and kept a record of it; Hirsch, 1986). But on November 24, 1969, Bennis sent a telegram to Bradford saying that “he was turning down the job for personal reasons. At the NTL offices, the people who received it were so angry that they remembered it years later as an endless document” (Kleiner, 2008, p. 206). Burke (2015, p. 2) later said that

For reasons I never fully understood (insufficient funds was no doubt a large factor), this promise (of a university) never really materialized, and consequently Warren decided to withdraw and not accept the Board’s offer for him to become NTL’s President.

Fiscal and Leadership Crisis

In part, at least, because of the debt the board had already taken on, NTL experienced a fiscal crisis in addition to the leadership crisis associated with Bennis’s declining to take the position. “The Board of Directors scrambled to meet NTL’s operating expenses, while simultaneously, at least for a time, straining to keep the vision of an NTL university alive” (Hirsch, 1986, p. 94). NTL’s fundraising efforts were not successful. “For its operating expenses, the institute depended on the money from training seminars, which had unexpectedly plummeted. The organization had to take out loans to pay its staff salaries that fall” (Kleiner, 2008, p. 205). In addition, “Bradford had to postpone retirement. Corporate enrollments stagnated. The black caucus continued to raise objections and press for training. Bills piled up from renovations and other halfhearted projects begun in preparation for the renaissance that had been aborted” (Kleiner, 2008, p.207). “I handled what I did very badly,” Bennis would later say, “but I was absolutely correct in what I did” (Kleiner, 2008, p. 206).

After Bennis’s withdrawal, as Hirsch (1986, p. 98) summarized,

A search committee was formed to find a successor to Bradford and submitted the names of five possible replacements. . . . All refused . . . After the Executive Committee meeting in July, 1970, responsibilities were split between Vladimir Dupré, Charlie Seashore, and Cecil Hannan. . . . In September (1970), Vladimir Dupré and Cecil Hannan were proposed as candidates for Executive Director. Dupré was elected unanimously for a term of four years.

Burke (2015, p. 2) noted that “Vlad had been instrumental in establishing the Midwest Center associated with NTL, was a bit older than myself, and had some administrative experience. Vlad accepted the position and asked me to stay on,” which Warner agreed to do for 2 years.

Hirsch (1986, pp. 99-100) recounted that Dupré had to cope with several serious problems which together threatened the very existence of NTL. These included, among others, the debt for refurbishing NTL headquarters at Bethel, dealing with several cancelled contracts, which created an “income squeeze,” addressing problems of accreditation of the NTL network, dealing with members who were conducting consulting “in direct competition with NTL,” reforming the governance structure, reconstituting the board of directors to include underrepresented NTL members, and responding to the demands of the Black Caucus.

These serious problems were apparently not handled well. Burke (2015, pp. 2-3) stated that

Well into the first year of Vlad’s presidency, the situation at NTL had deteriorated financially and administratively, and we had a potentially serious lawsuit facing us. Chris Argyris, also an NTL Board member at the time, contacted me and asked for a private meeting. . . . For about 2 hours, Chris grilled me about what was going on at NTL. I told him everything and added that I was worried about NTL’s future. Chris then said he was going to call an emergency meeting of the Board and that I must attend and tell the entire eight-member Board what I had just told him. I swallowed hard and said that I would . . .

Argyris chaired the meeting and began with asking me to tell the Board what I had told him. I did so. There were questions that I attempted to answer. I was nervous yet forthright. Then Argyris began to grill other NTL central office members at the meeting especially my professional staff members. The silence was deafening. Argyris kept questioning. When staff members did speak, they said something to the effect of while we had problems things weren’t so bad. I felt abandoned, “hung out to dry.” I had blown the whistle but it turned out to be only my whistle; no one else seemed to have one.

Subsequently, the Board voted on whether Vlad should remain in office. Four said “no”—Argyris, Marrow, (the) President of the New School, and one other CEO—two said yes and two abstained. Thus, the Board was deadlocked, and Vlad continued as President. Two weeks later, he came to my office and fired me.

Burke (2015, p. 5) added that

Even though my NTL colleagues stated in advance that I was doing the right thing’ and they would be in my corner,’ when the time came, psychologically they were nowhere to be found. . . . And I lost respect for each and every one of them.

Problems continued during Dupré’s administration. Hirsch (1986, p. 102) noted that “In his attempts to shore up the organization, Dupré had expanded the staff. . . . That is, during the Dupré tenure, NTL was shrinking, while the central staff (in Washington DC) was increasing.” Furthermore, (p. 103) Dupré,

in an effort to solve the financial crisis, hired a consulting firm to manage its finances for $5000 a month. According to Ray Gibbons (the NTL Comptroller), the benefits were nil . . . In 1973-1974, the Oil Crisis hit, creating a sharp recession that drastically cut into the income of NTL. By 1975, NTL was $250,000 in debt. Attempts to broaden membership on the Board of Directors did not change its composition.

Meanwhile other organizations had begun to carry out some of the work previously associated with NTL. For example, Freedman (1999) wrote that the inauguration of University Associates’ Annual Handbooks in 1972 had an adverse impact on NTL’s publishing business, as did the Addison–Wesley OD series. NTL kept sinking deeper into debt. When Dupré’s term ended NTL was on the verge of bankruptcy.

In 1975, Edith Seashore succeeded Dupré as president. She, Barbara Bunker, Hal Kellner, and Peter Vaill formed the “four horsepersons” committee to save NTL (Hirsch, 1986; Kleiner, 2008). There is a lengthy discussion of their work in Kleiner (2008, pp. 211-214). To summarize, as Bunker put it (Bunker & Gorelick, in press):

In 1975, NTL was in a financial crisis. Four of us decided to see if we could rescue the organization (Edie Seashore, Peter Vail, Hal Kellner and I[Barbara]). We offered to pay off the $275,000 debt if the board would resign and turn the organization over to us. The board was mostly businessmen and they were relieved to not have to declare bankruptcy. They resigned and we took over the board. We reconstituted the membership so that it was more inclusive. We mandated that the board be 50% women and 50% men; 50% people of color and 50% white. Seventy-five new members worked for free as NTL lab staff trainers for two weeks over the next two years and we paid off the debt. As a result, NTL became a different organization.

“By the end of Edith Seashore’s tenure as President in 1979, NTL had become stabilized as an organization, it was developing programs for individual, group, and organizational development and its financial problems were behind them” (Hirsch (1986, p. 106). It has not returned to the prominence it held in the 1960s, but it continues to operate, especially as a training organization for human interaction training and OD, and to publish JABS.

Challenges to NTL Practice From a Business Perspective

One obvious challenge from a business perspective was finances. The board was apparently not adequately aware of what was needed for NTL to be financially stable. This is not unique to NTL. Organizations with a social mission, including some profit making organizations, are sometimes slow to pay careful attention to finances (e.g., Edmondson, 2014). Yet a stable financial situation is necessary for an organization’s mission to be accomplished.

A second obvious challenge from a business perspective was the loss of first mover advantage. “First mover” discussions refer to ways that innovative companies can secure the benefits of their innovation (Teece, 1986). The assumption is that if knowledge cannot be easily appropriated by others, then the originators will maintain it. But NTL did not do this. Several years after these events at NTL, the Society for Organizational Learning had a similar experience. It did not maintain its innovative advantage, in part because it published materials that taught others how to carry out many of its activities (Bartunek et al., 2007). The fact that consultants who had been NTL members were able to carry out as independent entrepreneurs the work they had developed at NTL meant that it was very difficult for NTL to maintain a first mover advantage.

In contrast to standard business practices, Freedman (1999, p. 129) noted, “One of NTL’s core values was to give what was learned away to people who could use the knowledge. There was no sense of proprietary information, and copyrights were not used.” In this sense, NTL did exactly what it was aspiring to accomplish. What was not questioned was what aspirations like this would do for the sustainability of NTL, a term not then used in the context of organizational change.

Challenges of Theory and Practice

At least initially in NTL, theory and practice were somewhat synonymous. In the 1950s, 1960s, and early 1970s, the OD practitioners were, by and large, the academics, and the academics were, by and large, the OD practitioners. They were developing the theoretical foundations of OD based directly on their experiences leading T-groups and other NTL programs (cf. Bennis & Biederman, 2010), including those for executives.

These theoretical foundations were quite pronounced in the Eddy et al. (1969) book for managers that NTL published. That book includes chapters by several NTL members who played roles in the crises describe above. It includes, for example, a chapter by Bennis (1969, p. 188) on Organizational Revitalization, which involves “a deliberate and self-conscious examination of organizational behavior and a collaborative relationship between managers and scientists to improve performance.” It includes a chapter by Burke (1969) on Interpersonal Communication that includes many steps speakers and listeners can take to improve listening and communication between them. It also includes a chapter by Argyris (1969, p. 216) on Today’s Problems with Tomorrow’s Organizations that sketches out, among other things, the need for trust and risk taking in organizations, and creating places where individuals “do not fear stating their own views.”

But Bennis did not seem to be operating out of a collaborative relationship with the NTL Board when he turned down the Director job by means of a telegram. NTL board members did not seem to listen to Burke in a way that fostered appropriate communication when he was describing problems he encountered. Argyris did not seem to create a climate of trust in which individuals are not afraid of stating their views before inviting Burke to speak to the Board. At least from the documents I have been able to find, there is no evidence of an approach consistent with OD theory being during throughout the leadership and fiscal challenges that began with Bennis being offered and refusing the job of director.

Perhaps the most apt conceptual approach for understanding these events was the single loop/double loop approach to learning that Argyris and Schön (1974) were developing at this time. To briefly summarize, they argued that most people and organizations make use of Model 1 defensive routines that enable them to protect themselves from challenge and maintain their control in situations that might be threatening (such as, say, a discussion of a leadership and financial crisis). It is quite difficult to move to a Model 2 approach that is open to new, disconfirming information. When Argyris asked Burke to be honest about the problems at NTL he seemed to be asking him to initiate a model II approach, but without any of its necessary preconditions, such as the climate of trust Argyris (1969) had identified as crucial for Model II learning to succeed. This was especially the case because NTL leaders and members had already been enacting defensive routines for some time, as indicated by their complaining in private but acting as if everything was fine in public while Warren Bennis was being recruited. Bennis also was complicit with this type of behavior, as his later statements acknowledged. ²

Learning From Failure

Some of those involved in the financial and succession decisions at NTL described how important failure was to their learning. Burke (2015) wrote that much of his learning from NTL “came from sharing failure experiences.” In his 1977 commencement address at the University of Cincinnati, when he was completing his presidency there, Bennis (cited in Kleiner, 2008, p. 225) stated that

I seem to have spent a good deal of my life trying to understand from failure and always believing that that’s what we are really here for: to learn from ourselves and others, and some anticipation of failure has to be a part of that.

Failure, if reflected on, can sometimes be very valuable, as Burke and Bennis both noted. Mirvis and Berg (1977) argued that it is legitimate and important for organizations to make errors and sometimes fail in the process of carrying out OD, as long as they treat the errors as sources of learning and improve based on them. Heracleous and Bartunek (2020) suggested that errors be “incorporated as part of a long-term appreciation of the nature of change.” However, statements like these leave unaddressed the question of what is learned from failure, and how much it truly leads to improvement.

Among other things, it is obvious from the presentation above that learnings from NTL’s failures during its succession and fiscal crises may differ considerably for different people. Burke (2015) lost respect for each of his colleagues. The four horsepersons lost confidence in the ability of the current board to resolve NTL’s problems. Bennis (1970, p. 602) lost trust in many of the methods of OD, saying shortly after turning down the NTL position, that

There was a time when I believed that consensus was a valid operating procedure, I no longer think this is realistic, given the scale and diversity of organizations. In fact, I have come to think that the quest for consensus, except for some microsystems where it may be feasible, is a misplaced nostalgia for a folk society as chimerical, incidentally, as the American search for “identity.”

The experience of NTL also raises questions about why particular learnings occur and about what is necessary for generative learning that truly leads to improvements to take place. Conceptual concerns like these were highlighted in Organization Science in 1996, in an exchange between Argyris, building on his double loop learning theory, and Anne Miner and Stephen Mezias, building on the learning theory originally developed by Cyert and March (1963) in their Behavioral Theory of the Firm. I will highlight only a few pertinent aspects of their complex exchange that address why what is learned from organizational events might not lead to improvement, and what successful learning might need if it is to happen.

Both agreed that sometimes what is learned from organization events is not helpful, though they gave different reasons for this. Argyris (1996, p. 80) argued that the primary reason that double loop learning does not occur in organizations is that “Model I theories in use and organizational defensive routines combine to create quasi-resolution of conflict, coalition groups each looking out for its own interests and, at the same time, competing and condemning other groups, and limited learning.” This limited learning is basically about how to protect oneself. Miner and Mezias, in contrast, emphasized that in the presence of noise it may be very difficult for organizations to detect the true signals of events; thus, what they learn may be incorrect. In superstitious learning, for example, “a firm may incorrectly conclude its own actions caused a valuable outcome and repeat that action, producing harmful outcomes (p. 93).” In other words, both Argyris and Miner and Mezias show how what is learned from organizational events may be harmful rather than lead to improvement.

In addition, both argued that some types of learning are effective. For Argyris, double loop learning is effective because it truly challenges and changes the status quo. Miner and Mezias (1996, p. 93) discussed generative learning as effective. This approach does not tacitly accept “a fixed world . . . to which organizations must adapt.” Rather, it “includes an active, creative component” that shares some commonalities with double loop learning, in that it may be quite radical. In other words, Argyris and Miner and Mezias agree that some types of learning are productive for organizations’ ongoing development. However, such learning is not easy to accomplish.

From the description here, it does not appear that the NTL board ever engaged in generative or double loop learning. Rather, the board exhibited defensive routines that resulted in a loss of respect for NTL colleagues, distrust of OD methods, and near bankruptcy.