Consensus? No, Good Decisions Require “Respectful Disagreement”

What Warren Buffett, Talmudic Law, and Toyota Can Teach You About Decisions

When Warren Buffett is deciding whether to go forward with an acquisition or a merger, he hires two investment banks, rather than just one. One bank gets a bonus if the deal goes through, while the other gets a bonus if it doesn’t go through.

And in Orthodox Judaism, the Talmud specifies that when a jury is unanimous in its verdict, with no dissenting votes, the death penalty cannot be carried out. Ancient wisdom recognizes that absolute consensus is a hallmark of poor decision-making, because in real life there are always conflicting facts and mitigating factors.

Both Talmudic Law and Buffett’s merger evaluation process are examples of how to set up a good decision-making process, not by striving for consensus, but by probing for conflict and counter-examples. Buffett is ensuring that both sides of an argument are represented by strong advocates, motivated to make the best case for their own side, while the Talmud recognizes that consensus is the enemy of careful deliberation and a hallmark of mob rule.

But now think back to the last time your management team had to make a difficult decision. Perhaps it had to do with where and how to cut the budget, or whether to keep going with a poorly performing initiative, or which of several strategies might be the most effective. Once you zeroed in on a particular decision, did you push hard to gather the evidence to support it, in order to build consensus?

That’s what most decision-makers would do. But the right way to make sure it’s the right decision in the first place would be to push hard to find any evidence that might prove it wrong.

Too many business managers have conned themselves into thinking that it’s a good thing to cultivate a consensus within the team about the right course to take. To be sure, once a decision has been made and is being implemented, getting everyone on the same train is important. But the process of making that decision in the first place will benefit greatly from being influenced by a wide range of diverse views.

At Toyota, “respectful disagreement” is one of the tenets of the Toyota Way. Toyota’s goal is to foster a culture in which diverse views get freely expressed, because they consider every worker’s own individual perspective to be an important ingredient for strengthening corporate decisions. The saying at Toyota is:

If two workers always agree, then one of them is redundant.

While you want everyone to agree with the central mission of your organization, making good decisions about how to accomplish that mission requires surfacing all the different and conflicting ways in which it might be accomplished, and to consider objectively all the reasons why any particular method might fail. You need a lot of respectful disagreement to make solid decisions.

To make good decisions, don’t spend all your effort to dig up data and develop the analysis. Instead, concentrate on improving the decision-making process.

In one five-year study of more than a thousand different business decisions, two researchers evaluated the process followed to make each decision and the amount and detail of data used. They also tracked the actual business outcomes of these decisions, in terms of market share, profit, and revenue.

Know what they found? It didn’t really matter how much data went into a decision or how accurate and detailed the analysis was. What really mattered, in terms of producing the best business decisions, was whether the decision-makers themselves had engaged in a sound decision-making process, including an explicit effort to uncover dissenting views, contrary facts, and dis-confirming data. 

In predicting actual business results, this study found the decision-making process six times more significant than extensive data, sophisticated analytical tools, and consensus.

In pointing out the wisdom of seeking out dissenting voices rather than trying to develop consensus, one of the researchers drew an analogy to the modern legal system, which relies on a judge hearing from both a prosecutor and a defense counsel. If the only presentation to the court came from one side of the dispute, we certainly wouldn’t consider it to be a fair system.

And yet, he said, a large number of business decisions are made in exactly this fashion. There is only one side of the story represented, as if the only argument worth paying attention to were the prosecutor’s. And that argument, unrefuted, is constantly strengthened with data and analysis, until a consensus is achieved.

But the quality of the decision is compromised, because in the real world life is messy, absolute truth is rare, and certainty about anything will often obscure the facts.