When Less Isn’t More: The Hidden Risks of Simplification Bias

Everything should be made as simple as possible, but no simpler. – Albert Einstein

Every day, we make decisions that shape our lives and the world around us. But what happens when the complexity of those decisions is underestimated? Simplification bias makes us ignore important details and choose easier or familiar options, even if they are not always the best for us. How often do we overlook grave dangers just because they seem unlikely? The decisions we make are often shaped by the simplification bias, with implications ranging from historical tragedies like the Titanic to everyday choices such as forgoing a helmet while riding a motorcycle.

Historical Impacts: The Titanic and Beyond

The RMS Titanic was fitted with fewer than half the lifeboat capacity needed to save her passengers and crew. The ship was constructed with the intention of being unsinkable, leaving everyone convinced that it would never meet its demise beneath the waves. On April 14th, 1912, Titanic collided with an iceberg on her maiden voyage, and sank with the loss of more than 1500 lives.

Earlier on the fateful day, the radio operators aboard the Titanic were startled by the reception of six urgent messages from nearby ships. These messages warned of treacherous drifting ice, which had already caught the attention of concerned passengers on the Titanic earlier in the afternoon.

Although the crew was aware of ice in the vicinity, they did not reduce the ship's speed, and continued to steam at 22 knots (41?km/h; 25?mph).

Titanic's?captain, Edward Smith, had earlier famously declared in an interview that he could not "imagine any condition which would cause a ship to founder, Modern shipbuilding has gone beyond that."

The Titanic's sinking created a lasting memory for all who bore witness to the tragic event. Thousands of people died either when the disaster occurred or shortly afterward. The impact of that single event resonated through society for years, leading to inquests and new laws of maritime safety.?

It is worth noting that it took a significant amount of time for both passengers and crew to react to the fact that a major disaster had occurred on the ship. In his landmark review of the sinking, the journalist Walter Lord notes:?

“For the first hour, few of the Titanic’s passengers took the collision very seriously—which is another reason why the early boats, at least, didn’t contain more people. They far preferred the warmth and comfort of the brightly lit ship to the prospect of spending a dark, cold night bobbing about the Atlantic in a rowboat. When the first boat to be loaded, No. 7, was swung out and First Officer Murdoch called for passengers, only a trickle responded…and at 12:55, No. 5 was lowered with just 41 people…meaning there was still room for another 24. As the boat jerked down toward the sea 70 feet below, Karl Behr wondered whether this precaution he was taking was worth the risk. Any idea that the Titanic might sink was “preposterous.” (Walter Lord)

What could have been the underlying thought process that guided the ship’s captain, crew and then passengers to ignore a consistent stream of warning signals of clear and present danger?

Contemporary Consequences: From Health to Finance

In modern times, consider the tendency for motorcycle riders to forgo wearing helmets, even though there is overwhelming evidence that it can save lives? According to the CDC, every year there are thousands of deaths and injuries to motorcyclists involved in crashes. More than 5,500 motorcyclists died on our nation’s roads in 2020, and more than 180,000 were treated in emergency departments for crash injuries. Motorcycle helmets are 37 percent (for riders) and 41 percent (for passengers) effective in preventing deaths. Helmets reduce the risk of head injury by 69%.

In the digital age, simplification bias leads us to choose convenience over security, prompting us to use simple passwords or ignore software updates, exposing ourselves to significant cyber risks.

In the same vein, as we have already discussed in previous editions of this series, there is usually a very low willingness to pay for insurance that covers individuals against low probability high impact catastrophe events, such as hurricanes, floods and earthquakes.

The common underlying thread under each of these choices is the simplification bias.

The Psychological Roots of Simplification Bias

When making decisions that require deliberate consideration of many different factors, a System 1 shortcut that the human mind takes is to reduce the number of factors – that is to simplify the problem. While this is often a great way to make quick decisions, when faced with the decisions regarding vague low probability scenarios, there is a high likelihood the factors that are eliminated from consideration include the most important one.

Thresholds of Concern

One such key factor is the low but very real likelihood of the event – since the probability is low, it is easy for the mind to ignore it altogether. If the probability of an event is below a certain threshold of concern, it is completely ignored. In the case of the Titanic, the ship’s captain, the crew and eventually the crew, woefully underestimated the likelihood of the worst case scenario. In each case, they chose to ignore the risk altogether rather than more thoughtful investigation of the warning signs and alerts. Similarly, motorcyclists weigh the experiential high of riding a motorcycle without helmets against the perceived low probability of an accident. In doing so, they end up ignoring the risk altogether. And, finally, since the probability of a disaster is well below their threshold of concern, the homeowner in the catastrophe exposed zone often forgoes insurance cover, a phenomenon also known as probability neglect.?

Another related consideration is that decision makers are often not equipped to appreciate the difference between small probabilities. Our brains categorize the broad range of probabilities into three broad buckets.

1. no chance,
2. some chance and
3. absolutely certain

Decision makers are willing to pay a “certainty premium” only if the action will move the probability from one category to another. For instance, even if the risk of a deadly side effect from a cheaper drug is 1 in 10,000, patients are willing pay a significantly higher price for a drug that has Zero known side effects. On the flip side, if a different drug reduces the risk from 1 in 10,000 to 1 in 100,000 it will not command a premium price. In the case of disaster mitigation efforts, more actions can’t totally eliminate the disaster potential. Consequently, if an action reduces the likelihood of loss from a 1 in 100 year to a 1 in 250 year event, it is considered not worth taking since there is still some chance that the event will happen, even if the action is economically justified based on an expected value analysis.

On the flip side, once the probability exceeds a certain threshold of concern, people tend to overestimate the likelihood of the event. Risk to personal health is something for which individuals have a very low threshold of concern. This explains vaccine hesitancy even when the probability of adverse effects is documented to be low. Consider a statement like,

1 of 100,000 vaccinated kids will be permanently disabled.

This will likely create a strong visceral reaction and cause people to avoid vaccinating, since the probability is above their threshold of concern and is over weighted compared to the benefits of the vaccination.

Single-Action Bias

A related tendency to simplify is the "single-action" bias. Typically, risk mitigation actions require a series of concrete actions for complete protection. For instance, when faced with an impending hurricane, a homeowner needs to stock up on supplies, remove debris, board up the house, ensure a source of backup power, and ensure that evacuation arrangements are made, etc. Often, it is found that only one or two actions are taken, and the rest are ignored. When a risk event is brought to one’s attention, there is a “cognitive dissonance” and unease; however, as soon as the first set of actions is taken, this dissonance gets quickly alleviated. The problem is considered solved, and the motivation to persist and follow-through on the rest of the actions is lost.

Among people who have initiated climate mitigation behaviors, they may suffer from single-action bias, a tendency to perceive a reduced risk of climate change after taking one single climate action (say, recycling or planting some trees), although a combination of actions is more effective. Single-action bias can lead to negative spillover effects, where adopting one pro-environmental behavior reduces the probability of taking additional actions. Moreover, after engaging in one climate mitigation action, people may experience a heightened moral self-image and, consequently, are less likely to perform another climate action, known as moral licensing.

Have you ever experienced a situation where simplification bias influenced your decisions? How might recognizing this bias change your approach in the future?

Navigating Complex Decisions: Strategies for Clarity

Lowering the Threshold of Concern

To overcome the simplification bias for people who treat low probabilities of experiencing a disaster as zero, communication strategies could frame low probabilities over long time horizons, so individuals are less likely to perceive these risks as being below their threshold level of concern. For example, communicating the probability of flooding over a longer time horizon (e.g. a one in four chance of a flood in 30 years) instead of an annual time horizon (a one in 100 chance of a flood per year) can increase demand for protective measures against flooding.

Empirical evidence has also shown that communicating the consequences of not preparing for climate change risks, such as the damage one would experience from a flood, can make people pay attention to the risk and demand protection against it. Focusing on such worst-case scenarios may trigger individual concern for a risk and overcome the simplification bias.

Further, the “certainty premium” of loss mitigation action, can be conveyed by framing the reduction in loss likelihood in more tangible terms – for instance a reduction from a 1 in 100 year to 1 in 250 year loss likelihood is similar to odds improvement by 2.5x. Finally, where possible, express risk in frequency terms (as opposed to probability) to leverage “denominator neglect”. For instance, to encourage vaccine adoption.

Overcoming single-action bias

Single-action bias impacts not only individuals but also risk managers and policymakers. Decision makers who have experienced or been made aware of a certain risk scenario, might tend to get quickly satisfied by a token nominal action, while in fact, the call of the hour might be a sustained series of actions to mitigate future loss potential. There may be a tendency to believe that the first action is often enough. One way to counter this tendency is to anticipate this bias and work hard to condense the number of loss mitigation/service recommendations to the bare minimum, and to clearly identify the most impactful recommendation(s).

Lastly, a debiasing tool for the single-action bias is reinforcing the connection between mitigation actions (such as for Climate Change) and a person’s identity or values by reflecting on how the behavior is related to their values, identity, or views on how people in society should behave. Such intervention has been shown to increase the adoption of subsequent climate actions.

Conclusion

While often problematic, we must acknowledge that simplification bias can sometimes serve us in emergency scenarios where swift action outweighs perfect knowledge. However, as we navigate a world rich in information but poor in attention, recognizing and mitigating simplification bias is crucial. When faced with low probability and high consequence decisions, even when we consciously try to reason through the problem, our cognitive calculus tends to be simplified. We end up ignoring the problem. In cases, where we do engage with the problem and identify a series of required actions, we end up being satisfied after taking the first in a set of actions. We can start by challenging our assumptions and seeking diverse perspectives, especially in high-stake decisions.