HBM 02: Why Knowledge Does Not Always Change Behavior

Every licensed driver knows that speeding increases crash risk. Seat belt campaigns have been running for decades, and awareness of their life-saving effect is near-universal. Most commuters in car-dependent cities are aware that public transit options exist. Residents in flood-prone areas generally know that evacuation orders are issued for their safety.

Yet speeding persists. Seat belt compliance varies widely across regions and demographics. Transit ridership stays flat despite new investments. Some households refuse to evacuate even under mandatory orders.

The problem is clear: knowing what to do is not the same as doing it. This gap between knowledge and action is one of the oldest and most important puzzles in behavior research. It is also the puzzle that motivates this entire series.

The Knowledge–Action Gap in Transportation

The assumption behind many public campaigns, safety programs, and awareness efforts is simple: if people know the facts, they will act accordingly. Inform drivers that texting while driving is dangerous, and they will stop. Tell commuters that transit reduces emissions, and they will switch modes. Show residents the flood risk map, and they will prepare.

This assumption has a name in health and social research: the information-deficit model. It treats behavior change as a knowledge problem. If people behave badly, it must be because they lack information. The solution is more information.

The information-deficit model is not entirely wrong. Knowledge matters. People cannot act on risks they have never heard of. Awareness campaigns do shift some behaviors, especially when the behavior is new and the barriers are low. But in most cases, information is necessary and insufficient. It opens the door, but it does not walk through it.

Consider four cases where the knowledge–action gap is stark.

Case 1: Speeding

Speeding is the most common traffic violation in nearly every country. The facts are not in dispute: higher speeds increase stopping distance, reduce reaction time, and dramatically increase the severity of crashes. Most drivers know this. Many have seen graphic public safety campaigns. Many have personally witnessed or experienced speed-related incidents.

Why do people speed despite knowing the risk? The reasons are varied:

• Habit and automaticity. Driving at a customary speed is largely automatic. Conscious knowledge about risk does not interrupt automatic behavior.
• Perceived control. Many drivers believe they can handle higher speeds safely. The risk is abstract; the driving skill feels concrete.
• Social norms. When traffic flow moves above the speed limit, individual drivers conform. Going slower feels unsafe or antisocial.
• Time pressure and context. Running late, familiar roads, open highways — these contextual factors override abstract risk awareness.
• Low perceived enforcement. If the probability of being caught is low, knowledge about the law does not produce compliance.

No amount of additional safety information will address these factors. They require different kinds of models and different kinds of interventions.

Case 2: Public Transit Use

Cities invest heavily in transit infrastructure: new bus lines, light rail, fare subsidies, real-time information apps. Awareness campaigns advertise service improvements. Commuters are generally aware that transit exists and that it can reduce congestion and emissions.

Yet mode shift is often disappointing. In many North American cities, transit ridership has stagnated or declined even as service has improved.

The right-hand column reveals factors that have nothing to do with information. They involve travel time, reliability, habit, social norms, built environment, and activity scheduling. These are the factors that behavior models try to capture — and that information campaigns alone cannot address.

Case 3: Seat Belt Use

Seat belt effectiveness is one of the most well-established facts in traffic safety. Public awareness campaigns have been running since the 1970s. In many countries, knowledge of seat belt benefits is close to universal.

Compliance, however, varies dramatically: by region, by age, by gender, by vehicle type, by seat position, and by trip purpose. Rear-seat compliance is significantly lower than front-seat compliance. Compliance in pickup trucks is lower than in sedans. Young men are less likely to buckle up than older women.

The variation is not explained by variation in knowledge. Almost everyone knows seat belts save lives. The variation is explained by:

• Perceived invulnerability. Young drivers underestimate their personal crash risk.
• Comfort and inconvenience. Some people find seat belts physically uncomfortable.
• Social norms. In some peer groups, wearing a seat belt is seen as unnecessary or even weak.
• Trip characteristics. Short trips, familiar routes, and low-speed roads reduce the perceived need.
• Enforcement context. Primary enforcement laws (where police can stop a vehicle solely for a seat belt violation) are associated with higher compliance than secondary enforcement laws.

Case 4: Evacuation Compliance

When a hurricane approaches, emergency managers issue evacuation orders based on storm track, surge modeling, and vulnerability zones. The information is disseminated through emergency alerts, media, social media, and direct communication. Most residents in evacuation zones are aware of the order.

Yet compliance is far from universal. Studies of hurricane evacuations consistently find that 30–50% of residents in mandatory evacuation zones do not evacuate.

These factors — risk perception, social influence, past experience, resources, trust — are precisely the constructs that behavior models are designed to represent. They cannot be addressed by repeating the evacuation order more loudly.

What Fills the Gap?

If knowledge alone does not produce behavior change, what does? The research literature identifies several categories of intervening factors that sit between what people know and what they do.

These are not just “other factors.” They are the mechanisms through which knowledge is (or is not) translated into action. Each behavior model in this series emphasizes a different subset of these mechanisms:

• KAP (Post 3) acknowledges the gap but does not explain it. It measures knowledge, attitude, and practice separately and documents the discrepancies.
• The Theory of Planned Behavior (Post 4) adds subjective norms and perceived behavioral control to explain why intention forms — and why intention does not always produce action.
• The Health Belief Model (Post 6) focuses on threat perception, perceived benefits, and perceived barriers.
• Protection Motivation Theory (Post 7) adds coping appraisal: even when people perceive a threat, they may not act if they doubt their ability to respond effectively.
• Social Cognitive Theory (Post 9) emphasizes self-efficacy and learning from others.
• COM-B (Post 20) takes an entirely different approach: it asks whether the person has the capability, opportunity, and motivation to perform the behavior, and diagnoses which of these is missing.

Why This Matters for Practice

The practical consequence of the knowledge–action gap is significant. If behavior change were simply a matter of information, then the primary tools of planners and policy makers would be:

• public awareness campaigns
• warning signs and safety messages
• educational programs
• information brochures and websites

These tools are inexpensive, easy to implement, and politically popular. They are also, for most behaviors, insufficient.

The right-hand column — behaviorally-informed strategies — requires knowing why people behave the way they do. And that requires models that go beyond awareness assessment.

Speed feedback signs work not because they provide new information (drivers already know the speed limit) but because they make speed salient at the moment of driving and activate self-regulation. Transit pass defaults work not because people did not know about transit but because opting out of a default requires effort. Road diets reduce speeding not by informing drivers but by physically constraining the choice environment.

These interventions are designed around behavioral mechanisms, not information deficits. They are the practical payoff of behavior modeling.

The Problem That Drives the Series

This post has established the central problem that motivates the rest of the series:

1. Knowledge is necessary but insufficient. People need information, but information alone rarely produces sustained behavior change.
2. The gap between knowledge and action is not random. It is structured by beliefs, norms, habits, context, efficacy, and system design.
3. Behavior models exist to represent those structures. Each model highlights different mechanisms and leads to different intervention strategies.
4. Choosing the right model matters. A model that explains intention may not capture habit. A model that captures individual psychology may miss social and environmental forces. A model that describes gaps may not explain them.

The next post introduces the KAP framework — the simplest and most widely used tool for assessing the knowledge–action gap. KAP does not solve the gap, but it makes it visible. Understanding both its power and its limits is the first step toward more explanatory models.

In the next post, we introduce the KAP framework — a simple, widely used tool for measuring the gap between what people know, what they believe, and what they do. KAP does not explain the gap, but it makes the gap visible and measurable. Understanding its value and its limits is the gateway to more powerful models.