The Restraint Bias

Most of us have found ourselves in the scenario described above, in some shape or form. Maybe we pledge to become vegetarians after watching an impactful documentary, but later find ourselves unable to resist the bacon cheeseburger on a restaurant menu. Or maybe we frequently smoke cigarettes but believe that we can easily quit at any time. Temptation is powerful, and often we inaccurately gauge the power it has over us. This exaggerated view of our own restraint can cause us to make poor decisions and put ourselves in scenarios where our restraint is tested. We then have a greater chance of slipping and going against our prior commitments.

While indulgence every now and then is normal, the restraint bias can cause significant unintended negative outcomes. With drug addiction, individuals often demonstrate a strong restraint bias.¹ The inflated sense of impulse control causes individuals to enter situations where they are exposed to addictive substances because they think their self-control is greater than it is. These behaviors increase the chances of relapse.

Clearly, the restraint bias affects our individual self-control strategies but the question of how its effect is magnified on a social scale is still unclear.

Let’s look at smoking for example. As previously stated, when recovering smokers significantly overestimate their impulse control, it impacts their avoidance of temptation, and consequently, increases their likelihood of taking up smoking again.¹ One study questioned ex-smokers about their beliefs and habits in order to assess levels of restraint bias and then checked in with the participants months later. The experimental data shows that recovering smokers with high levels of restraint bias had significantly higher rates of relapse after 4 months.¹

While smoking, addiction, and their consequent health problems are incredibly complex, the restraint bias does play a role in encouraging impulsive behaviors that have sizable consequences to our population. Through understanding the cognitive mechanisms directing our individual behaviors, we can gain insight into factors of larger public health trends.

Our affective states, or our moods and feelings, have a major impact on our judgment and decision-making. Research on the transformative ability of emotion also has provided insight into the psychological mechanisms of the restraint bias. Behavioral scientist George Loewenstein found that we have a general inability to predict our own wants and behaviors in future states, especially in different affective states.⁴ He refers to this discrepancy as the empathy gap.

The “cold-to-hot” empathy gap

Loewenstein uses the terms “hot” and “cold” to differentiate states of being where we are highly effected by our emotions or visceral impulses versus our more neutral and rational states of being. This terminology matches up with how we refer to our emotions in everyday life. We can think of statements made in the “heat of the moment” that we don’t necessarily mean. Or, people that remain “cool as a cucumber” during stressful situations.

When we are in a “cold” state, we underestimate how our decisions are swayed by our visceral impulses when we are in a “hot state”. For example, when we are not hungry we might make inaccurate assumptions of how we will act in the future when struck by hunger pains. Loewenstein states that this tendency can be traced to how our memory stores our visceral impulses. It’s easy to remember circumstances where we felt incredibly emotional or consumed by your bodily urges. We can conjure images of long car rides where we were famished or conversations where our emotions got the best of us. However, it is really difficult to truly remember what the sensations and impulses felt like when we are in a “cold state”.

This gap in understanding of our own impulses drives the restraint bias. When we underestimate the power of our impulses, we overestimate our own control over them.

We are faced with temptation and impulses every day. It is universal to act impulsively at times, or let our emotions and urges push us towards behaving in ways we otherwise wouldn’t. However, when we repeatedly overestimate our restraint and fail to take measures to limit our temptation, there can be life-shaping consequences. This tendency can severely affect our health by encouraging unhealthy eating, drug use, smoking, and more. It can impact our relationships by nudging us towards infidelity or rash decisions. It can even affect our finances, as we give in to temptations of spending and gambling. In its strongest form, the restraint bias can push us back into cycles of addiction. Overall, matters of restraint are no small thing.

For most of us, building will-power and restraint is a lifelong task. Navigating the restraint bias plays a critical role in our general pursuit of strong habits and resoluteness. To minimize the strength of the restraint bias is to have a more accurate perception of our impulse-control and make better decisions accordingly. Basically, how well we know ourselves can improve our choices and behaviors.

Collect personal data on our relationship with temptation and impulse

According to Stanford health psychologist Kelly McGonigal, “the best way to improve your self-control is to see how and why you lose control”.⁵  McGonigal encourages using a data-based approach to improving will-power and increasing our self-knowledge. First, we can take inventory of the areas in our lives that feel most influenced by impulsiveness or lack of restraint. Maybe we procrastinate or maybe we spend too much money on take-out. Whatever it is, we can start to document the circumstances where we feel more inclined to give in to temptation. For example, we might notice that we spend more money on food when we are upset or busy.

We can also keep track of how we think about restraint. Are we acting without thinking or are we rationalizing certain actions to ourselves? How do we judge our own acts of restraint? For example, we might say “Oh, getting take-out for one is cheaper than cooking”, but use this rationalization so often that it isn’t true anymore.

Plan according to your data

The more data we have on our impulsive behaviors, the better equipped we are to find solutions. If we find that our take-out spending goes up around busy weeks, we might look for ways we can meal prep pre-emptively. Or if we see that our spending on food is  higher when we study at the library because of its proximity to higher-priced amenities, we might choose to study at home instead. McGonigal also suggests making barriers and commitments for ourselves in order to be more accountable for our impulses. This can include removing junk food and cigarettes from our houses or changing the limits on our credit cards. Changing habits is tough and won’t happen overnight. But with a dedication to growth, we can learn more about our own habits and make incremental strides towards better choices.

In a 2009 study, psychology researchers Loran Nordgren, Frenk van Harreveld, and Joop van der Pligt explored our self-perceptions of impulse control and the consequences of these beliefs. Working off of Loewenstein’s empathy gap research, they proposed that people have a tendency to “overestimate their capacity for impulse control” and coined this phenomenon the restraint bias.¹ They tested this hypothesis through four experiments on impulse-control beliefs surrounding fatigue, hunger, and smoking/addiction.

In the first study on smoking impulses, Nordgran, van Harreveld, and van der Pligt manipulated the participants’ beliefs about their levels of impulse control by administering a “bogus ‘implicit-measures’ test” with fake results. The participants were then asked to watch the film Cigarettes and Coffee and had to choose a level of smoking temptation they would be exposed to. The higher the temptation, the larger the monetary compensation the participant would receive. For example, they would be given €4 to keep a cigarette on their desk and €6 for holding an unlit cigarette in their hand while watching the film.

The results supported their hypothesis: smokers who received results for high-control of their impulses were more likely to choose a higher level of temptation than the low-control group. Consequently, the high control group had a significantly higher rate of succumbing to their impulses and smoking a cigarette.

We all know what it is like to go grocery shopping while we are hungry. Often our eyes are bigger than our stomachs and we purchase an overabundance of food or items we normally wouldn’t. Clearly different levels of hunger yield different choices. As a part of their 2009 study, Nordgran, van Harreveld, and van der Pligt tested how hunger states affect our perception of our own impulse control and self-exposure to temptation. ¹

The researchers conducted this study by questioning people in two conditions: “as they entered a cafeteria (the hungry condition) and as they exited the cafeteria (the satiated condition”. The participants were then asked to rank a series of snacks from most favorite to least preferred and select one snack to keep for a week without eating. If they could restrain from eating their chosen snack they would receive 5 euros.

They found that those in the satiated condition chose a snack that was more tempting because they overestimated their impulse control (due to not being hungry at the time of choosing). This exposed them to more temptation throughout the week and made them more likely to give in to their visceral urges.

In Loewenstein’s research on the empathy gap, he discusses how this discrepancy in our self-perception can affect major decisions we make for ourselves and our health. As previously stated, the foundation of the restraint bias lies in the cold-hot empathy gap where we underestimate the power of our affective states on our impulses.

In addition to corroborating work on smoking and addiction, Loewenstein discusses how our inability to understand our future affective states can play a role in treatment for bipolar disorder and depression. When individuals are in a healthy mental state, they can forget what it is like to feel depressed and the lack of control they have over certain feelings. This can cause people to fail to take the necessary measures for treatment in order to preempt depressive states. This phenomenon also can provide insight on the low rates of adherence to mood-stabilizing drug regimens for bipolar patients.

What it is

The restraint bias refers to our tendency to overestimate our control over our impulses.

Why it happens

This happens due to a phenomenon called the empathy gap, in which we are unable to predict how we will act in future affective states. With the restraint bias, when we are in neutral states we underestimate the influence of our urges and reactions to temptations.

Example 1 –  How the restraint bias affects hunger-driven actions

In a study, researchers Nordgran, van Harreveld, and van der Pligt found that people that were in a satiated state overestimated their control over food temptation, in comparison to people who were hungry. Because of this bias, they exposed themselves to greater levels of temptation and were more likely to give in.

Example 2 – How the restraint bias affects medical decision-making

Psychologist George Loewenstein discusses how the empathy gap underlying the restraint bias can contribute to the choices we make about our health. Often when patients with depression or bipolar disorder are in more neutral mental states, it can be difficult to fully remember the sensation of those previous low states. Because of this, patients may not take medications or necessary treatment precautions.

Projection bias – Biases & Heuristics

The projection bias refers to our inability to accurately predict the wants and needs of our future selves. In this piece, the author discusses how this tendency affects our everyday decisions, including spending habits and time management.

Dunning–Kruger effect – Biases & Heuristics

The Dunning-Kruger effect describes our inability to recognize when we are not as good at something as we think, resulting in overconfidence. This article demonstrates why the loudest person in the room is not always the most capable, as well as why startups are so prone to failure.