5.6 Faulty Decision Making
Learning Objectives
- Recognize overconfidence bias and how to avoid it.
- Describe hindsight bias and how to avoid it.
- Define anchoring and how to avoid it.
- Recognize framing bias and how to avoid it.
- Describe escalation of commitment and how to avoid it.
- Describe what a project premortem is.
No matter which model you use, you need to know and avoid the decision-making traps that exist. Daniel Kahnemann (another Nobel prize winner) and Amos Tversky spent decades studying how people make decisions. They found that individuals are influenced by overconfidence bias, hindsight bias, anchoring bias, framing bias, and escalation of commitment.
Potential Challenges to Decision Making
Overconfidence Bias
Overconfidence bias occurs when individuals overestimate their ability to predict future events. Many people exhibit signs of overconfidence. For example, 82% of the drivers surveyed feel they are in the top 30% of safe drivers, 86% of students at the Harvard Business School say they are better looking than their peers, and doctors consistently overestimate their ability to detect problems (Tilson, 1999). Much like a friend who is always 100% sure he can pick the winners of this week’s football games despite evidence to the contrary, these individuals are suffering from overconfidence bias. People who purchase lottery tickets as a way to make money are probably suffering from overconfidence bias. It is three times more likely for a person driving 10 miles to buy a lottery ticket to be killed in a car accident than to win the jackpot (Orkin, 1991). To avoid this bias, take the time to stop and ask yourself whether you are being realistic in your judgments.
Hindsight Bias
Hindsight bias is the opposite of overconfidence bias, as it occurs when looking backward in time where mistakes made seem obvious after they have already occurred. In other words, after a surprising event occurred, many individuals are likely to think that they already knew this was going to happen. This may be because they are selectively reconstructing the events. Hindsight bias becomes a problem especially when judging someone else’s decisions. For example, let’s say a company driver hears the engine making unusual sounds before starting her morning routine. Being familiar with this car in particular, the driver may conclude that the probability of a serious problem is small and continue to drive the car. During the day, the car malfunctions, stranding her away from the office. It would be easy to criticize her decision to continue to drive the car because, in hindsight, the noises heard in the morning would make us believe that she should have known something was wrong and she should have taken the car in for service. However, the driver may have heard similar sounds before with no consequences, so based on the information available to her at the time, she may have made a reasonable choice. Therefore, it is important for decision makers to remember this bias before passing judgments on other people’s actions.
Anchoring
Anchoring refers to the tendency for individuals to rely too heavily on a single piece of information. Job seekers often fall into this trap by focusing on a desired salary while ignoring other aspects of the job offer such as additional benefits, fit with the job, and working environment. Similarly, but more dramatically, lives were lost in the Great Bear Wilderness Disaster when the coroner declared all five passengers of a small plane dead within five minutes of arriving at the accident scene, which halted the search effort for potential survivors, when, in fact, the next day two survivors walked out of the forest. How could a mistake like this have been made? One theory is that decision biases played a large role in this serious error; anchoring on the fact that the plane had been consumed by flames led the coroner to call off the search for any possible survivors (Becker, 2007).
Framing Bias
Framing bias refers to the tendency of decision makers to be influenced by the way that a situation or problem is presented. For example, when making a purchase, customers find it easier to let go of a discount as opposed to accepting a surcharge, even though they both might cost the person the same amount of money. Similarly, customers tend to prefer a statement such as “85% lean beef” as opposed to “15% fat” (Li, et. al., 2007)! It is important to be aware of this tendency because, depending on how a problem is presented to us, we might choose an alternative that is disadvantageous simply because of how it is framed.
Escalation of Commitment
Escalation of commitment occurs when individuals continue on a failing course of action after information reveals this may be a poor path to follow. It is sometimes called sunk costs fallacy because the continuation is often based on the idea that one has already invested in this course of action. For example, imagine a person purchases a used car that turns out to need another repair every few weeks. An effective way of dealing with this situation might be to sell the car without incurring further losses, donate the car, or drive it without repairing it until it falls apart. However, many people spend hours of their time and hundreds, even thousands of dollars repairing the car in the hopes that they will justify their initial investment in buying the car.
Example: Motorola’s Iridium Project
A classic example of escalation of commitment from the corporate world may be Motorola’s Iridium project. In 1980s, the phone coverage around the world was weak—it could take hours of dealing with a chain of telephone operators in several different countries to get a call through from, say, Cleveland to Calcutta. Thus, there was a real need within the business community to improve phone access around the world. Motorola envisioned solving this problem using 66 low-orbiting satellites, enabling users to place a direct call to any location around the world. At the time of idea development, the project was technologically advanced, sophisticated, and made financial sense. Motorola spun off Iridium as a separate company in 1991. It took researchers 15 years to develop the product from idea to market release. However, in the 1990s, the landscape for cell phone technology was dramatically different from the 1980s, and the widespread cell phone coverage around the world eliminated a large base of the projected customer base for Iridium. Had they been paying attention to these developments, the decision makers would probably have abandoned the project at some point in the early 1990s. Instead, they released the Iridium phone to the market in 1998. The phone cost $3,000 and it was literally the size of a brick. Moreover, it was not possible to use the phone in moving cars or inside buildings! Not surprisingly, the launch was a failure and Iridium filed for bankruptcy in 1999 (Finkelstein & Sanford, 2000). The company was ultimately purchased for $25 million by a group of investors (whereas it cost the company $5 billion to develop its product), scaled down its operations, and modified it for use by the Department of Defense to connect soldiers in remote areas not served by landlines or cell phones.
Why does escalation of commitment occur? There may be many reasons, but two are particularly important.
First, decision makers may not want to admit that they were wrong. This may be because of personal pride or being afraid of the consequences of such an admission. Second, decision makers may incorrectly believe that spending more time and energy might somehow help them recover their losses. Effective decision makers avoid escalation of commitment by distinguishing between when persistence may actually pay off versus when persistence might mean escalation of commitment. To avoid escalation of commitment, you might consider having strict turning back points. For example, you might determine up front that you will not spend more than $500 trying to repair the car and will sell the car when you reach that point. You might also consider assigning separate decision makers for the initial buying and subsequent selling decisions. Periodical evaluations of an initially sound decision to see whether the decision still makes sense is also another way of preventing escalation of commitment. This becomes particularly important in projects such as the Iridium where the initial decision is not immediately implemented but instead needs to go through a lengthy development process. In such cases, it becomes important to assess the soundness of the initial decision periodically in the face of changing market conditions. Finally, creating an organizational climate where individuals do not fear admitting that their initial decision no longer makes economic sense would go a long way in preventing escalation of commitment, as it could lower the regret the decision maker may experience (Wong & Kwong, 2007).
Perform a Project “Premortem” to Fix Problems Before They Happen
Doctors routinely perform postmortems to understand what went wrong with a patient who has died. The idea is for everyone to learn from the unfortunate outcome so that future patients will not meet a similar fate. But, what if you could avoid a horrible outcome before it happened by identifying project risks proactively—before your project derails? Research suggests that the simple exercise of imagining what could go wrong with a given decision can increase your ability to identify reasons for future successes or failures by 30% (Mitchell, et. al., 1989). A “premortem” is a way to imagine and to avoid what might go wrong before spending a cent or having to change course along the way (Breen, 2000; Klein, 2003; Klein, 2007; Pliske et al., 2001).
Gary Klein (2003), an expert on decision making in fast-paced, uncertain, complex, and critical environments, recommends that decision makers follow this six-step premortem process to increase their chances of success.
- A planning team comes up with an outline of a plan, such as the launching of a new product.
- Either the existing group or a unique group is then told to imagine looking into a crystal ball and seeing that the new product failed miserably. They then write down all the reasons they can imagine that might have led to this failure.
- Each team member shares items from their list until all the potential problems have been identified.
- The list is reviewed for additional ideas.
- The issues are sorted into categories in the search for themes.
- The plan should then be revised to correct the flaws and avoid these potential problems.
The premortem technique allows groups to truly delve into “what if” scenarios. For example, in a premortem session at a Fortune 50 company, an executive imagined that a potential billion-dollar environmental sustainability project might fail because the CEO had retired.
Exercises
- Describe a time when you fell into one of the decision-making traps. How did you come to realize that you had made a poor decision?
- How can you avoid escalation of commitment?
- Share an example of anchoring.
- Which of the traps seems the most dangerous for decision makers and why?
- How might you use the premortem technique to be more effective within groups at school or work?
- Imagine that your good friend is starting a new job next week as a manager. What recommendations would you give your friend to be successful with decision making at work?
Key Takeaways
Understanding decision-making traps can help you avoid and manage them. Overconfidence bias can cause you to ignore obvious information. Hindsight bias can similarly cause a person to incorrectly believe in their ability to predict events. Anchoring and framing biases show the importance of the way problems or alternatives are presented in influencing one’s decision. Escalation of commitment demonstrates how individuals’ desire for consistency, or to avoid admitting a mistake, can cause them to continue to invest in a decision that is not prudent.
“Faulty Decision Making” and “Developing Your Personal Decision-Making Skills” in Principles of Management by University of Minnesota are licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License, except where otherwise noted.