Loading…
0:00
6:49

Over a year ago, I was in a meeting that devolved into a debate about which software platform we should use to develop a custom tool for our most important group of clients. I recommended a platform that was particularly well-suited to the job, but our lead programmer didn’t have experience with it. He advocated for a platform that was less suitable to our clients but more familiar to him.

As the conflict rose, the meeting became unproductive. So, I changed the topic by posing a question: How do decisions get made in our organization?

The response: silence.

Nobody on the team really knew. One of our principal managers was out on medical leave, and it created an authority vacuum. But even without that extenuating circumstance, we likely wouldn’t have been able to describe an explicit decision structure for our organization.

Typical.

“Well,” one of the meeting participants finally offered, “we usually talk about things, and then I think Stan decides.”

“Well,” I answered, “Stan’s not here. So that decision-making process is gone. What shall we use instead?”

This time the answer came fast.

“Consensus!” everyone agreed.

So I asked another question. “What constitutes consensus?”


Many American organizations default to “consensus” in the absence of a clear decision-making hierarchy. Maybe it’s a result of America’s egalitarian ideology (“all people are created equal”) or our natural aversion to conflict and criticism. Often, people gravitate toward consensus without even knowing what consensus is. And if consensus fails, they propose a vote.

Both of these seem (to me) like crappy decision processes. Why?

  • They’re too slow.
  • It isn’t always clear when the time for a final decision has come.
  • The diffusion of responsibility makes accountability almost impossible.
  • They provide insufficient feedback on the quality of the decision.
  • They tend to reduce the decision criteria to the lowest common denominator.
  • They fail to foster creativity and innovation.
  • They are too risk-averse.

Since that conversation, I’ve realized most Americans are poor decision-makers. Although people who lead meetings and discussions often aspire to become managers, many managers don’t actually make the call on difficult decisions. In fact, the typical American system of bureaucracy is actually designed to remove discretion from managers.

Bureaucracy substitutes policy for judgment and algorithms for experiments.

That’s the purpose of bureaucratic policies. They serve as a substitute for judgment. Lolly Daskal makes a critical point in her piece, “Dumb Rules That Make Your Best People Quit”: Bureaucracy is about minimizing mistakes by establishing decision rules for as many conceivable conditions as the company can imagine. It substitutes policy for judgment and algorithms for experiments.

One advantage of bureaucracy is efficiency. It reduces the time and energy required to weigh and evaluate alternatives, and it typically guides the organization toward systems of specializing labor that avoid duplicating effort, output, and communicative confusion. But bureaucracy affords few opportunities to practice and improve decision-making skills. Thus, when an organization encounters a surprise, anomaly, or vacancy in authority, no one knows how to make a decision.

And sometimes, no decision is worse than any available alternative.


My undergraduate students in engineering business practices typically fall into the same consensus-based no-decision trap that my software development team did. Without an explicit assignment of decision rights, they’re bogged down in a self-recursive loop of indecision.

To avoid this, members of an organization have to understand how decisions are made.

Ray Dalio, founder of Bridgewater Associates — the world’s largest hedge fund — chronicles his company’s decision-making processes in his book, Principles. According to Dalio, decisions that can be reduced to algorithms are codified in computer code rather than in bureaucracy. For every other decision, there’s a credibility-weighted consensus-based process. If it sounds complicated, that’s because it is.

Ray Dalio on decision-making.

Your company doesn’t have to do it like Bridgewater. There are other models of decision-making. The point is that the people in your company or on your team absolutely must know how decisions get made. In his TED Talk (above), Dalio embraces mistakes as a source of knowledge that will improve future decisions. He believes in creating a culture of accountability and evaluation of decision quality.

Similarly, in my undergraduate class, I require students to form teams and file a management plan. The plan assigns roles and responsibilities to each of them and describes their decision-making processes.

Most importantly, I ask my undergraduates to decide how to allocate a finite number of grade points to their team members. I call it their “compensation plan.” At the end of the assignment, they must divide 100 additional experience points among the members of the team. They can use any decision process they want for this allocation, but they must describe the process before beginning the assignment.

Most students say, “We are going to allocate the grade points equally,” which fails to create the incentive structure necessary to motivate team members to do good work. In fact, it creates a free-rider problem, in which students who contribute minimally still receive the full benefit of being a member of the team.

Decision-making is a skill, and it improves with deliberate practice.

There are alternative decision processes. The most successful teams I’ve had thus far have appointed a Chief Executive Officer. The responsibility of assigning points falls to the CEO, to whom the other team members report. The CEO allocates or delegates decision rights to other team members but remains accountable for the team’s performance.

Under this management structure, it’s important for the team to know how who the CEO is accountable to. Teams that do best with this process create an advisory board, like a board of directors, that includes the other team members. The advisory board has the power to remove the CEO and appoint a new one.

This system is much less complicated than Dalio’s believability-based collective approach — but it’s much more complicated than the typical approach to student teamwork. And it works much better.


Other systems can work, too — including egalitarian, consensus-based approaches. My classroom is a good place for students to experiment and compare different approaches before they enter the workforce. Decision-making is a skill, and it improves with deliberate practice.

Only by making their decision processes explicit, conscious, and accountable (including a way to track decision quality) will students — and real-world employees — improve their performance.