Corporate Practice Management in the 21st Century: Why Democratic Governance No Longer Works
As a manifestation of their core DNA, physicians dislike being led. As a corollary, physicians dislike, to a greater or lesser extent, their leader, particularly those leaders actually empowered to lead. To circumvent the leader problem, physicians tend to select a democratic form of governance for their practice, thus diluting the position of the leader. The process of consensus building is the functional crux of the democratic governance model. It permits the partners to maintain an appearance of equality while, at the same time, allowing decisions to be made with only minimal direction by the leader. Developing a consensus opinion is essentially a mathematical exercise: Members alter the proposal until the math works. The achievement of majority support for the proposal necessarily becomes the primary goal of the argument when, in reality, the primary goal should be the development of an opinion that improves the competitive stance of the group in its marketplace. The very nature of the math of consensus development ensures that the content of the proposal is diluted until it is acceptable to a majority of voters. While proponents are looking for enough residual content to make the desired change, opponents seek enough dilution to limit any undesirable effect. The proposition is watered down to a level that is acceptable to the majority of voters, and in the end, the corporate value of the proposal is lost. Consider this hypothetical partnership meeting. Physician A (CEO and senior partner) says, “The clinicians in our office building want us to open our office on Saturday, from 9 am to 2 pm. Most of the physicians’ offices are open until 1 pm, but they are usually running late, so they want us there until 2 pm.” Physician B replies, “It will cost us a fortune: $400 in technologists’ salaries, plus additional nontechnologist expenses. We will lose all kinds of money. Besides, many of these physicians send their patients to our competitor down the street during the week. I vote no.” Physician C (who has a Saturday-morning golf game in the summer and who skis on Saturdays in the winter) adds, “I vote no, too.” Physician D (who is busy on Saturday mornings with family and church activities) says, “I vote no.” Physician A says, “Rather than turning the physicians down completely, let’s work out something. We need a compromise.” After much haggling and complaining, a compromise is reached in which the group will provide limited services (radiography and mammography) from 10 am to noon. The vote is carried with five in favor and four opposed. Physician A informs the clinicians of the group’s decision. The physicians are furious; making mammography and radiography available until noon is insufficient. In the end, the extended hours fail. After six weekends of losing money every weekend and squandering the radiologists’ weekend time, they are stopped. How could a consensus decision—the product of such an elegant, human political process—produce such a bad outcome? Mathematics to Nowhere Consensus building is simply a cold, objective mathematical process. Of itself, it has no capacity to ensure the propriety or validity of any decision. Even worse, our infatuation with the perceived wisdom conferred by consensus building allows us to select inappropriate conclusions, as we smugly envelop those conclusions with the crowd’s aggregated wisdom. If the math is not damaging enough, two intrinsic weaknesses of the radiology partners are what ultimately sink the whole Saturday-hours endeavor. First, incomplete knowledge of the background issues is a devastating malady. The CEO is uncertain of the politics. Many of the other partners might have no understanding of the political implications of withholding needed clinical services from the referring physicians. If the knowledge is incomplete, it is very easy to get off track during the discussion phase, diluting the analysis with unsubstantiated opinion. Second, the partners’ personal biases will readily fill any void created by incomplete knowledge. The discussion becomes all about the other issues: the Saturday golf outing, social and family issues, or any other competing pet project. The importance of Saturday office hours is lost, and the analysis is doomed. It is readily apparent that consensus building is not a strong, useful management process in the competitive business marketplace. The math (and the inherent personal biases of the group) will inevitably compromise the group’s best deliberative efforts. When the group chooses to manage its affairs by consensus building, it must understand that while democratic consensus is a charming and elegant concept, it does not perform well in a competitive business marketplace. The basic mathematical challenge of the process and the characteristics of the individual partners are compelling weaknesses in the consensus model. There are, however, several additional elegant, subtle weaknesses in the consensus process. These characteristics, since they reflect more generalized human malfunctions, are difficult to recognize and control. They are characteristic of the human species at large, and are not limited to any individual radiology group or, for that matter, to the specific circumstance of weekend coverage. They are factors in virtually every partnership meeting (in every venue). Wisdom of The Crowd Ever since the heyday of the ancient Greek city-states, humans have been attracted to the democratic governance format, with its attendant consensus management. The elegance and appeal of democracy, however, do not—in any fashion—imply that consensus is an effective and functional management process. Specifically, can the democratic model identify and select the best answer to a difficult question even when there is no single popular option and when all choices will be distasteful to a significant segment of the partnership? The wisdom of the crowd affirmatively declares that it can make that choice. Just imagine that all the partners are bright, experienced individuals. When they convene as a partnership, they bring together not only all their knowledge and experience, but their wisdom as well. Magically, when they are gathered as a group, their wisdom is aggregated into a superwisdom of the crowd. It seems that when the group gathers for a meeting, two heads are better than one—and multiple heads are even better than two. It is this attribute—the wisdom of the crowd—that allows the partnership to select the democratic process as its governance model. Without the hope of assumed wisdom, the group would hesitate to choose consensus management. Intrinsically, the democratic model has no specific capacity to resolve complex issues, and no one seriously makes that assertion. In reality, the wisdom of the crowd rarely achieves fruition. Rather, the wisdom is diluted by imitative behavior, human emotions (peer pressure and the herd instinct), incomplete knowledge, personal bias, private agendas, and excessive homogeneity of the crowd. When so altered, the wisdom of the crowd becomes groupthink. Groupthink and Information Cascade Irving Janis (1918–1990), a psychologist who studied disasters in US foreign policy (such as failure to anticipate Pearl Harbor), identified groupthink as a mode of thinking that people engage in when the group’s pursuit of consensus overrides its interest in appropriate resolution of the question. Compared with wisdom, groupthink is frightening indeed. More frightening is the subtle and slippery slope leading from wisdom to groupthink. Those involved in the process (for example, attending a partnership meeting) might feel only their collective wisdom and fail to perceive the looming groupthink. Conditions leading to groupthink are homogeneity of members’ social backgrounds and ideologies, along with isolation of the group from outside sources of information analysis. Symptoms of groupthink include illusions of invulnerability (producing excessive optimism); rationalizing warnings that challenge group assumptions; unquestioned belief in the moral superiority of the group; stereotyping those opposed to the group as weak, evil, or stupid; and illusions of unanimity (consensus) in the decision (where a partner’s silence is viewed as tacit agreement). An information cascade can be considered a variant of game theory. An information cascade develops in an open forum such as a partnership meeting. In the meeting, a decision is to be made by the group. The proposal is a simple binary question: Should we open the office on Saturday or not? Each member of the group, in sequence, must make a decision on the question. Since the meeting is open, each person knows the decision (and the opinions) offered by each of the preceding individuals. The new information garnered by observing the preceding choices alters the participants’ preexisting knowledge base and leads them to follow their predecessors, believing they might be correct—even if their choices are contrary to the participants’ firmly held preexisting knowledge/bias. The herd instinct is a powerful one. Even if most of the individual partners have primary information favoring the correct conclusion, the aggregated partners might reach the wrong conclusion because of the cascade effect of observing the sequential comments and votes of the other partners. If one of the early speakers is particularly powerful, intimidating, or animated, that person can start the cascade. An information cascade is a compelling challenge to the knowledge and beliefs of the individual partner. An information cascade can overcome all but the strongest and best-prepared partners. Keep in mind that the individual partners believe that they are acting rationally and making the right judgment. They fully trust the evolving wisdom of the crowd and ultimately deny their own privately held beliefs. They tell themselves that it feels good to see that so many of their bright partners share the same opinion. They really trust these knowledgeable and wise people. Pluralistic ignorance is an additional influence: All those opposed to the proposition think they are alone in their beliefs because they see the cascade developing—when, in fact, several (if not all) members, based on their primary knowledge before the meeting, are opposed to the cascade-based conclusion. Indecision Analysis paralysis and paradox of choice are common problems in a medical group. A yes-or-no choice is certainly the most efficient format. If there are possibilities other than aye and nay, however, failure to reach any decision is a common outcome. The group is paralyzed by too many choices. Depending upon the difficulty of the question, resolution of the issue may be beyond the capacity of the consensus governance model. In this circumstance, making no decision is, in fact, a decision. It is firmly established that making no decision constitutes a decision—and, on occasion, the worst possible decision. Making no decision is a compelling indictment of the inadequacy of the consensus management process. Lurking in the background in these prolonged indecisive meetings, look for incomplete knowledge, personal bias, and dysfunctional management personnel. Groupthink and information cascade are other extremely subtle developments in a partnership meeting. Peer pressure and the herd instinct obviously will dilute the quality of discussion and impede a rational voting procedure. The wisdom of the crowd demands full participation by all attendees. Preparation beforehand, with complete and exhaustive research conducted by the partners, is a necessity. Consensus management is a truly elegant management system that is very attractive to the educated, thoughtful physician. It has variously been celebrated for being inclusive, participatory, collaborative, cooperative, and egalitarian. The only problem is that consensus management doesn’t work, particularly when the issues are complex and difficult. The concepts of wisdom of the crowd, groupthink, and information cascade are difficult and challenging. A medical group should not choose consensus management or any of its variants unless the partners fully understand these questions:If democratic consensus governance doesn’t work, how can we ever acquire the benefit of the wisdom of the crowd? Is there such thing, or is it a myth? Is democratic governance valid if it is not accompanied by the wisdom of the crowd? Is groupthink an inevitable contaminant of the wisdom of the crowd and, consequently, the democratic governance process? Can we get beyond groupthink? No matter what governance process we choose, whether democratic or otherwise, can we somehow benefit from the wisdom of the crowd? In its purest form, how could the wisdom of the crowd fail to be helpful? Is it possible to build enough safeguards to forestall the apparently inevitable degeneration into groupthink? Furthermore, can we assure ourselves that our knowledge is complete, rather than incomplete? Can physicians assure themselves that when they vote, they will select the option that best serves the partnership, rather than ceding—under the protective cloak of a secret ballot—to their personal wants and needs? By the time the hypothetical partnership’s meeting opens to discuss Saturday office hours, the wisdom of the crowd has descended upon the partners. They firmly believe that they can work out a solution that will best serve the purposes of the partnership. Although the aggregated partners believe that they possess superwisdom, their opinions are not based on wisdom, but on incomplete knowledge—further corrupted by personal agenda and bias. As regularly happens in these circumstances, an unstoppable information cascade develops. No one in the meeting possesses enough knowledge to turn back this cascade. By the end of the meeting, the group’s wisdom has morphed into groupthink, leading the partners to choose to open the office, with limited services, from 10:00 am to noon, thereby offending the clinicians. The Saturday office-hours project is doomed from the beginning. Even when the project fails, six weeks later, the partners have no insight whatsoever. Instead, they blame the clinicians. Such is the power of the wisdom of the crowd. Corporate Management In the usual corporate management system, the partners meet for the same purpose. The process of the corporate meeting, however, is completely different. In typical corporate settings, CEOs are fully authorized to act and are fully responsible and accountable for their actions. It is their responsibility alone to analyze and act upon a proposal. To assist them in carrying out their responsibilities, CEOs call meetings of their expert advisors, seeking their knowledge and opinion. The purpose of these meetings, therefore, is to educate CEOs, not to develop policy (and certainly not to vote). These meetings are educational events rather than voting exercises. The experts are interrogated aggressively, elevating the knowledge base of all in attendance, including CEOs. Note that there is negligible incomplete knowledge because only experts are allowed to speak, and those experts are thoroughly grilled to unmask any incomplete knowledge. Personal biases and personal agendas are similarly exposed. An appropriately conducted meeting is managed ruthlessly, and the experts are held accountable for the strengths and weaknesses of their opinions. In The Effective Executive, Peter F. Drucker (1909–2005) stated that effective decision makers deliberately create dissent and disagreement rather than consensus. “Decisions of the kind the executive has to make are not made well by acclamation,” he wrote. “They are made well only if based on the clash of conflicting views.”¹ Since this kind of meeting does not develop a formal opinion, groupthink is nonexistent. Rather, such meetings only develop facts and fact-based opinions. The final judgments are developed by CEOs alone, after they listen to the spirited, dissenting, and often contradictory arguments arising from the knowledgeable experts. There is no information cascade because the opinions of the experts are based on fact and are subject to withering challenge by those attending the meeting. Furthermore, sequential comments (as spoken in consensus meetings) leading to an information cascade are not part of the usual corporate meeting. Rather, comments in the corporate meeting are data driven and are not sequential. They are not derived from the prior speaker’s opinion and do not influence the knowledge or comments of the following speakers. Fully accountable CEOs develop the final opinions and policies based on the facts and fact-based opinions they gather at the meeting. The partners and experts at the meeting do not make the final judgment. There is no vote. Rather, they submit facts and fact-based opinions, leaving the decisions to the responsible and authorized CEOs. The CEOs, not the aggregated partners, implement the new policies. The CEOs alone are responsible for the initial analyses, the final decisions, and the implementation of the new policies. Responsibility and authority to develop and implement policy are hugely important. Corporate CEOs are fully authorized to act and are fully accountable for their actions. Unfortunately, it is commonplace, in organizations managed by consensus, for no single individual to hold these responsibilities. Rather, accountability and authority to act are, as a signature of equality, intentionally diffused throughout the partnership. Thus, the group of equals is paralyzed with indecision and inaction during its most difficult challenges. When all options are difficult, consensus decisions and consensus actions are simply beyond the aggregated wisdom and knowledge of the arguing partners and their management process. They are crippled by incomplete knowledge, personal goals and wants, and a voting process that demands—absolutely demands—a mediocre, suboptimal outcome. A Better Model Because of increasing challenges by government entities, Byzantine behavior by enlarging hospital corporations, aggressive postures by insurance companies, and increasing intergroup and intragroup competition, it is in the best interest of the partnership that the group be managed as a typical corporate entity with an authorized and responsible CEO. Only then will the group be spared the alluring imperfections of consensus management. Be aware, though, that any decision can occasionally become a failure, despite the best efforts of the management team. In that circumstance, it is easier to fire the dysfunctional CEO than it is to fire your own dysfunctional consensus opinion. Who takes the hit if the consensus opinion is wrong? Do you fire the group? Do you fire the management (consensus) process? Because you, the partner, are really at fault, do you simply fire yourself? Maybe you just move along, doing none of these. Choosing the usual corporate management process is a large and bitter pill for the group. The comparison of corporate management with consensus management can be accomplished more easily if it is viewed from a broader perspective. The group should schedule an evening meeting for this purpose, with a limited agenda (and unlimited food and drink). At this meeting, review the vision statement and the mission statement as they were written at the time of the group’s founding. Understand their purpose. Understand their cultural and spiritual sources. Partners should acknowledge that their professional careers and livelihoods (food and clothing, mortgages, children’s tuition fees, and retirement savings) depend on their fulfillment of the vision and mission statements. Put your heart and soul into these statements. Give them real meaning. Have a knowledgeable discussion of the best management process to use to achieve those goals. Along the way, discard all preconceived notions. Put all options on the table. When stuck, reread the vision statement and the mission statement. Consider hiring outside, knowledge-able experts to keep you focused on your goal of developing a functional and competitive group. Consider having a facilitator act as a referee—because it may be a difficult meeting. What about the wisdom of the crowd? Is it only a myth? Since knowledge is a substrate of wisdom, and knowledge is incomplete in the consensus-managed partnership, wisdom can only be a myth in the consensus-managed group. Does wisdom exist in the group with corporate management? Possibly; we trust that the knowledge level is adequate in the corporate setting. We know, however, that wisdom has other components, in addition to knowledge. All of us can only hope that those other segments come together in the corporate CEO. That’s why the CEO is given a contract, not a crown. He welcomes your comments at David F. Hayes, MD, is a radiologist practicing at Windsong Radiology, Williamsville, New York.