Many banks have a formal or informal expectation that certain decisions need to be unanimous. We often see this in credit committees, strategic planning sessions, hiring decisions and operational moves. If you find your bank committees often achieving votes where 100% of the members agree, chances are you have a flawed system. While unanimity in decision making seems like a positive attribute, we will show you how it counterintuitively leads to more risk, not less. We will also show you that complete agreement on many topics also can point to systemic error, an error that can cost your bank millions.
For the sake of discussion, let’s remove any decision that is a false dichotomy where there isn’t a decision to make. The decision to hire a needed person to solve a BSA issue may not have much variability around it as you may not have any other options or choice. However, absent of these easy decisions a split decision is not only expected, but it is also a sign of a healthy process.
The Problem with Unanimous Decisions
Bankers are conservative by nature, and unanimous decisions are reassuring and self-validating. This notion is supported by society-at-large as if the majority rules, the complete agreement must rule even more. Reaching a total consensus is a common desired outcome in many business situations. However, if this happens too often or happens in situations where there should be more disagreement, then complete agreement points to a problem.
The key to understanding this paradox is to understand the inherent level of uncertainty in the type of decision that is before the group. These decisions involve outcomes that could have a varied distribution.
Let’s say that all ten people on the board want to hire Sally as CEO. We don’t know Sally and we’re sure that she is probably qualified but just given the unanimous consensus, we would say that Sally may not be the right hire. We know this because the quality of CEO candidates not only follows a normal distribution, but people’s opinions on CEO candidates also follow a normal distribution. It may be that Sally is just an outstanding candidate, but the odds are against it.
In Sally’s case either three things are going on. Chances are the board has not seen enough candidates to achieve a normal distribution, so that Sally presented a false choice such as depicted in the diagram below. If only three candidates are seen, then Sally represents a selection bias. For it is possible that if candidates four and five were interviewed, then Sally wouldn’t look so good.
In addition to a selection bias, a unanimous decision in hiring could be as a result of an influence bias. Here, one particular board member could hold an outsized amount of power, so the nine other members support whatever decision that is made by this highly influential board member. This also happens when meeting members are apathetic and just don’t care enough to understand the complete issue. In order to save time, these members may just vote with an influential member or go with the majority.
Third, systemic errors could be as a result of incomplete information, such as data supporting the negative side of the decision. A derivation of a systemic process error could also include outright fraud that while rare in community banking, unfortunately, has occurred in the past.
No one doubts that the 100% vote counts that were in favor of North Korea’s Kim Jong-un or Iraq’s Saddam Hussein were rife with problems. These are extreme examples but highlight the notion that by just observing a vote’s outcome, you know can recognize when the process is flawed.
When You Should Expect Non-unanimous Votes
As a rule of thumb, in situations where you can expect more varied results, then 100% consensus usually points to some hidden factor influencing the system.
Another common example is credit. Like CEO candidates, credit is also evenly distributed so if 100% of your credits are getting approved by your process or loan committee, then something is likely flawed in the system.
A Problem With Votes On Credit
Decisions on credit quality is also a common problem with banks, as it means that either there are not enough credit grades, members influence each other too much, there is not enough information known about a credit, or analytical rigor is not being applied.
This is highlighted by the graphic below. There should be a positive selection bias when working with credit. Credits similar to Credit #1 below should never make it to a credit decisioning group. However, if the mid-line on the graphic below is where the bank strives for acceptable credit, then only bringing credits similar to Credit #5 uncovers a flaw in the system. This means that there is a large spectrum of credit that the bank is missing out on, such as those illustrated by Credits #2, 3 and 4 below.
It is this middle area, where the chief credit officer, loan approval committee or board can best leverage their expertise. If only credits like #5 are making it to Committee, then the bank is wasting the group’s time.
In situations similar to the above distribution of credit, we might expect that Credit #2 would result in a 70% vote against, Credit #3 might be closer to a 50%/50% vote and Credit #4 would get approved with an approximate 70% approval vote. Non-unanimous votes would be indicative of a healthy loan approval process.
Not uncovering these flaws in the system can lead to sub-optimal and inefficient decisions. If you doubt this, ask anyone that was part of a failed bank loan decisioning process. In most every case that we have reviewed, unanimous decisions on credit were the norm and split votes, the exception.
The Psychological Impact of Wanting Unanimous Decisions
Sometimes banks have a formal rule that all members must agree, and sometimes it is an unspoken rule. Here, to keep business moving, members raise their threshold of when they voice a dissenting opinion, and the result is that more poor outcomes get approved.
Creating a Process of Good Decision Making: Learning from The Bay of Pigs
Dissent and disagreement should be naturally expected and also encouraged. In a healthy process, members should disagree and split votes should occur on a regular basis. This should not only be acceptable but desirable as well.
After a review had uncovered a flawed decision making process that resulted in the Cuban Bay of Pigs debacle, President Kennedy assigned a “devil’s advocate” to all decisions so that a dissenting opinion could always be heard and the probability for a counter-popular vote was more likely.
The military often does this by creating a “Red Team” whose sole job is to create problems with a plan or mission. This Red Team helps uncover information and flaws that might be missed so that the dissent can have equal resources as the proposing team. Creating an opposition group helps reduce the probability of selection bias.
Putting It into Action
If you are getting consistent unanimous consent at your bank, it is time to review your decision making process. Since banks are six to ten times leveraged, every decision a bank makes runs the risk of being amplified, so it is important to get decisions right. By having more dissenting opinions and votes where variability of outcomes is expected, banks can take more comfort in their process.
Submitted by Chris Nichols on July 11, 2016