Local Rationality

What in the World Were They Thinking?


As I write this, the Houston area is dealing with the aftermath of a 500-year flood that has left several feet of water in areas that have never flooded before.  Some areas received 15- to 20-inches of rain in less that 6-hours which left all of the creeks and bayou’s overflowing their banks and inundating residential areas, displacing several thousand people and shutting down travel in much of the area.  As I watched live television coverage of this event from my non-flooded home I was saddened by the impact on the lives of so many, but initially struck by the “stupidity” of those who made decisions that put their lives at risk and in a few cases cost them their lives.  I began to try to make sense of why these individuals would make what appeared to be such fool-hearty decisions.  What could they have been thinking when they drove past a vehicle with flashing lights right into an underpass with 20 feet of water in it?  What could they have been thinking when three people launched their small flat-bottom, aluminum boat to take a “sight-seeing” trip down a creek that was overflowing with rushing waters and perilous undercurrents only to capsize, resulting in them floating in the chilly water for 2+ hours before being rescued by the authorities?  As I reflected on it, and after my initial incredulous reaction, my conclusion was that it made perfect sense to each of them to do what they did.  In the moment, each of their contexts led them to make what to me seemed in hindsight to be a very foolish and costly decision.  You may be asking yourself….” What is he talking about?  How could it make sense to do something so obviously foolish?”  Let me attempt to explain. Context is powerful and it is the primary source we have when making decisions.  Additionally, it is individual-centric.  My context, your context and the context of the individual who drove around a barricade into twenty feet of water are all very different, but they are our personal contexts.  In my context where I am sitting in my living room, watching TV, sipping a cup of coffee, with no pressure to get to a certain location for a specific purpose is most likely completely different from the man who drove around a police vehicle, with flashing lights, in a downpour, with his windshield wipers flashing, on his way to check on someone he cares about and who could be in danger from the rising water.  What is salient to me and what was salient to him are very different and would most likely lead to different decisions.  His decision was “locally rational”, i.e. it made perfect sense in the moment.  We will never know, but it is very likely that his context precluded him from even noticing the flashing lights of the police vehicle or the possibility of water in the underpass.  It is also possible that “human error” was present in the tragic deaths of at least 6 people during the flood, but human error is not a sufficient explanation.  We can never really understand what led to their decisions to put themselves at risk without understanding the contexts that drove those decisions.

This is what we really need to focus on when we are investigating incidents in the workplace so that we can impact the aspects of contexts that become salient to our workers.  The greater impact we have on minimizing the salience of contextual factors that lead to risk taking and increasing the salience of contextual factors that minimize risk, the greater opportunity we will have to end “senseless” injury and death in the workplace, and on rain swollen highways.  This approach will have a lot more positive impact than just chalking it up to “stupidity”!

Why It Makes Sense to Tolerate Risk


Risk-Taking and Sense-Making Risk tolerance is a real challenge for nearly all of us, whether we are managing a team in a high-risk environment or trying to get a teenager to refrain from using his cellphone while driving.  It is also, unfortunately, a somewhat complicated matter.  There are plenty of moving parts.  Personalities, past experiences, fatigue and mood have all been shown to affect a person’s tolerance for risk.  Apart from trying to change individuals’ “predispositions” toward risk-taking, there is a lot that we can do to help minimize risk tolerance in any given context.  The key, as it turns out, is to focus our efforts on the context itself.

If you have followed our blog, you are by now familiar with the idea of “local rationality,” which goes something like this: Our actions and decisions are heavily influenced by the factors that are most obvious, pressing and significant (or, “salient”) in our immediate context.  In other words, what we do makes sense to us in the moment.  When was the last time you did something that, in retrospect, had you mumbling to yourself, “What was I thinking?”  When you look back on a previous decision, it doesn’t always make sense because you are no longer under the influence of the context in which you originally made that decision.

What does local rationality have to do with risk tolerance?  It’s simple.  When someone makes a decision to do something that he knows is risky, it makes sense to him given the factors that are most salient in his immediate context.

If we want to help others be less tolerant of risk, we should start by understanding which factors in a person’s context are likely to lead him to think that it makes sense to do risky things.  There are many factors, ranging from the layout of the physical space to the structure of incentive systems.  Some are obvious; others are not.  Here are a couple of significant but often overlooked factors.

Being in a Position of Relative Power

If you have a chemistry set and a few willing test subjects, give this experiment a shot.  Have two people sit in submissive positions (heads downcast, backs slouched) and one person stand over them in a power position (arms crossed, towering and glaring down at the others).  After only 60 seconds in these positions, something surprising happens to the brain chemistry of the person in the power position.  Testosterone (risk tolerance) and cortisol (risk-aversion) levels change, and this person is now more inclined to do risky things.  That’s right; when you are in a position of power relative to others in your context, you are more risk tolerant.

There is an important limiting factor here, though.  If the person in power also feels a sense of responsibility for the wellbeing of others in that context, the brain chemistry changes and he or she becomes more risk averse.  Parents are a great example.  They are clearly in a power-position relative to their children, but because parents are profoundly aware of their role in protecting their children, they are less likely to do risky things.

If you want to limit the effects of relative power-positioning on certain individuals’ risk tolerance - think supervisors, team leads, mentors and veteran employees - help them gain a clear sense of responsibility for the wellbeing of others around them.

Authority Pressure

On a remote job site in West Texas, a young laborer stepped over a pressurized hose on his way to get a tool from his truck.  Moments later, the hose erupted and he narrowly avoided a life-changing catastrophe.  This young employee was fully aware of the risk of stepping over a pressurized hose, and under normal circumstances, he would never have done something so risky; but in that moment it made sense because his supervisor had just instructed him with a tone of urgency to fetch the tool.

It is well documented that people will do wildly uncharacteristic things when instructed to do so by an authority figure.  (See Stanley Milgram’s “Study of Obedience”.)  The troubling part is that people will do uncharacteristically dangerous things - risking life and limb - under the influence of minor and even unintentional pressure from an authority figure.  Leaders need to be made aware of their influence and unceasingly demonstrate that, for them, working safely trumps other commands.

A Parting Thought

There is certainly more to be said about minimizing risk tolerance, but a critical first step is to recognize that the contexts in which people find themselves, which are the very same contexts that managers, supervisors and parents have substantial control over, directly affect people’s risk tolerance.

So, with that “trouble” employee / relative / friend / child in mind, think to yourself, how might their context lead them to think that it makes sense to do risky things?

The Human Factor - Missing from Behavior Based Safety


Since the early 1970’s, there has been an interest in the application of Applied Behavioral Analysis (ABA) techniques to the improvement of safety performance in the workplace. The pioneering work of B.F. Skinner on Operant Conditioning in the 1940’s, 50’s and 60’s led to a focus on changing unsafe behavior using observation and feedback techniques. Thousands of organizations have attempted to use various aspects of ABA to improve safety with various levels of success. This approach (referred to as Behavior Based Safety, or BBS) typically attempts to increase the chances that desired “safe” behavior will occur in the future by first identifying the desired behavior, observing the performance of individuals in the workplace and then applying positive reinforcement (consequences) following the desired behavior. The idea is that as safe behavior is strengthened, unsafe behavior will disappear (“extinguish”).

The Linear View

Traditionally, incidents/accidents have been viewed as a series of cause and effect events that can be understood and ultimately prevented by interrupting the chain of events in some way. With this “Linear” view of accident causation, there is an attempt to identify the root cause of the incident, which is often determined to be some form of “Human Error” due to an unsafe action. The Linear view can be depicted as follows:

Event “A” (Antecedent) → Behavior “B” → Undesired Event → Consequence “C”

Driven by the views of Skinner and others, Behavioral Psychology and BBS have been concerned exclusively with what can be observed. The issue is that, while people do behave overtly, they also have “cognitive” capacity to observe their environment, think about it and make calculated decisions about how to behave in the first place. While Behavioral Psychologists acknowledge that this occurs, they argue that the “causes” of performance can be explained through an analysis of the Antecedents within the environment. However, since they also take a linear view, they tend to limit the “causal” antecedent to a single source known as the “root cause”.

Human Factors

The field of Human Factors Psychology has provided a body of research that has demonstrated that many, if not most, accidents evolve out of complex systems that are not necessarily linear. Some researchers call this a “Systemic” view of incidents. The argument is that incidents occur in complex environments, characterized as involving multiple interacting systems rather than just simple linear events. That is, multiple interacting events (Antecedents) combine to create the “right” context to elicit the behavior that follows.

In such complex environments, individuals are constantly evaluating multiple contextual factors to allow them to make decisions about how to act, rather than simply responding to single Antecedents that happen to be present. In this view, the decision to act in a specific (safe or unsafe) manner is directed by sources of information, some of which are only available to the individual and not obvious to on-lookers or investigators who attempt to determine causation following an incident.

Local Rationality

This is referred to as “Local Rationality” because the decision to act in a certain way makes perfect sense to the individual in the local context given the information that he has in the moment. The local rationality principle says that people do what makes sense given the situation, operational pressures and organizational norms in which they find themselves.

People don’t want to get hurt, so when they do something unsafe, it is usually because they are either not aware that what they are doing is unsafe, they don’t recognize the hazard, or they don’t fully realize the risk associated with what they are doing. In some cases they may be aware of the risk, but because of other contextual factors, they decide to act unsafely anyway. (Have you ever driven over the speed limit because you were late for an appointment?) The key here is developing an understanding of why the individual made or is making the decision to behave in a particular way.

A More Complete Understanding

We believe that the most fruitful way to understand this is to bring together the rich knowledge provided by behavioral research and human factors (including cognitive & social psychological) research to create a more complete understanding of what goes on when people make decisions to take risks and act in unsafe ways. We believe it is time to put the Human Factor into Behavior Based Safety.

Why Rule Breaking Makes Sense

Complexity & Rationality Why do employees decide to break the rules?  Do it their way?  Resist change?  It doesn’t make any sense!

It can be frustrating, and often perplexing, when employees fail to adhere to company policies and procedures, especially when those policies and procedures are in their best interest. There is a useful way to think about this issue: What employees do makes sense...to them; but the complexity of work environments makes it hard to understand why it makes sense to them.

We live and work in complex environments. It helps to think of our environments as systems with overlapping and interacting components - including people, things, rules, values, etc. - which are, in turn, complex sub-systems. One of the principles of complex systems is that the “people” component tends to respond only to the limited information that they are presented with locally. We make decisions based on our knowledge of what makes sense at the local level, which is called “local rationality”.

The policies and procedures contained in the corporate manual are only influential if they are brought to bear on the daily lives of people in the workplace. If those policies and procedures only exist in the manual and are not made a part of the local workplace, then they don’t exist in reality and will not have an impact on performance. They will lack influence.

Companies have policies and procedures for a reason - to create good, reliable results; so it is the responsibility of supervisors to bring those policies and procedures to life in the workplace. By intentionally incorporating formal policies and procedures into the “local” work environments of employees - through conversation, feedback, modeling, etc. - supervisors make it “rational” to follow the rules.

Can you work incident free without the use of punishment?

I was speaking recently to a group of mid-level safety professionals about redirecting unwanted behaviors and making change within individual and systemic safety systems.  I had one participant who was particularly passionate about his views on changing the behaviors of workers.  According to him, one cannot be expected to change behavior or work incident free without at least threatening the use of punitive actions.  In his own words, “you cannot expect them to work safely if you can’t punish them for not working safely.”  He was also quite vocal in his assertion that it is of little use to determine which contextual factors are driving an unsafe behavior.  Again quoting him, “why do I need to know why they did it unsafely?  If they can’t get it done, find somebody that can.”  

What an Idiot!

I meet managers like this from time-to-time and I’m immediately driven to wonder what it must be like to work for such a person.  How could a person like this have risen in the ranks of his corporate structure?  How could such an idiot...oh,wait.  Am I not making the same mistakes that I now, silently scold him for?  You see, when people do things that we see as evil, stupid, or just plain wrong, there are two incredibly common and powerful principles at play.  The first principle is called the Fundamental Attribution Error (FAE) and, if allowed to take over one’s thought process, it will make a tyrant out of the most pleasant of us.  The FAE says that when we see people do things that we believe to be undesirable, we attribute it to them as being flawed in some way or to them having bad intentions.  They are stupid, evil, heartless, or just plain incompetent.  If we assume these traits to be the driving factor of an unsafe act and we have organizational power, we will likely move to punish this bad actor for their evil doings.  After all, somebody so (insert evil adjective here) deserves to be punished.  The truth is that most people are good and decent people who just want to do a good job.

Context Matters

This leads us to our second important principle, Local Rationality.  Local Rationality says that when good and decent people do things that are unsafe or break policies or rules, they usually do it without any ill-intent.  In fact, because of their own personal context, they do it because it makes sense to them to do it that way; hence the term “local rationality”.  As a matter of fact, had you or I been in their situation, given the exact same context, chances are we would have done the same thing.  It isn’t motive that normally needs to be changed, it’s context.

With this knowledge, let’s look back at the two questions from our Safety Manager.

  1. “How can I be expected to change behavior or work incident free, without threatening to to punish the wrong-doers?” and
  2. “Why do I need to know why they did it unsafely?  If they can’t get it done, find somebody that can.”

Once we understand that, in general, people don’t knowingly and blatantly do unsafe things or break rules, rather that they do it because of a possibly flawed work system, e.g. improper equipment, pressure from others, lack of training, etc., then we have the ability to calmly have a conversation to determine why they did what they did.  In other words, we determine the context that drove the person to rush, cut corners, use improper tools, etc.  Once we know why they did it, we then have a chance of creating lasting change by changing the contextual factors that led to the unsafe act.

Your key take-aways: 
  1. When you see what you think is a pile of stupidity, be curious as to where it came from.  Otherwise, you may find yourself stepping in it yourself.
  2. Maybe it wasn’t stupidity at all.  Maybe it was just the by-product of the context in which they work.  Find a fix together and you may both come out smelling like roses.

Complexity and Local Rationality

Why do people - like employees and children - decide to break the rules?  Do it their way?  Resist change?  It doesn’t make any sense!  Or does it? It can be frustrating and often perplexing when employees fail to adhere to company policies and procedures, especially when those policies and procedures are in the employees’ best interest. Filing a required document can legally protect managers, but they don’t file it.  Locking out a machine that is being serviced can keep a technician safe from pain, injury and even death, but he regularly services the machine without locking it out.  Your children “know” the rules, but sometimes break them anyway. There is a useful way to think about this issue:  What employees and children do makes sense...to them.

We live, work, play and make decisions in complex environments. It helps to think of our environments as systems with overlapping and interacting components - including people, things, rules, values, knowledge, etc. - which are, in turn, complex sub-systems. One of the principles of complex systems is that the “people” component tends to be driven by the limited information that is available to and impressed upon those people within their local contexts. We make decisions based on our knowledge of what makes sense at the local level, at any given moment.  We call this the principle of “local rationality.”  In other words, our decisions are rational to us because they are based on the information available within the local context (which includes knowledge residing in our brains) at a particular point in time.

As supervisors and parents, we observe behavior that is driven by the principle of local rationality, but we only have limited information about what factors the individual is using to make their decision.  Why did it make sense to the employee to do such a dangerous thing?  Why did it make sense to the child to break the curfew rule?  After all, they know the rules and we have rules to make it clear how they should behave, don’t we?

Rules are only one component of the complex environments that we live and work in.  There are also pressures from other people - including superiors, peers and even you - to make a decision to act in a certain way.  Knowledge of past successes and failures, availability of resources needed to be successful, time pressures, workplace layout and numerous other kinds of factors make it ‘make sense’ for the person. Consider for a moment the last thing you saw a person do that “irked” you.  What kinds of factors could have led the person to do what he or she did?  Why would it have made sense...or did you assume it was because the person was lazy, rude, selfish, or in some other way had poor personal motivation?

As humans, we have a tendency to assume that people do what they do because of personal motivation , and then we treat their “failures” as an opportunity to motivate them to change and make better decisions in the future.  Research shows, however, that actions are often the result of the person’s evaluation of complex input from the environment and may have nothing to do with personal motivation.  For example, we tell employees to work safely, but at the same time push them for productivity, sometimes beyond their ability. The trade-off for the employee is to “cut corners” to be productive because he thinks in the moment that safety is really not at risk and showing you how productive he can be is what is really important.  Your teenager decides to come home late because her date had a couple of drinks and she didn’t trust him to drive safely...and her cell phone was dead, so she didn’t call. It makes sense in the moment, given the information that she had at her disposal.  It doesn’t make sense to you while you sit at home worried sick and imagining all kinds of terrible things.  It doesn’t make sense because you aren’t making decisions in her context!

As parents and supervisors, we need to ask a simple question before we punish undesired behavior: “Why did making that decision make sense to the person making it?”  Why was the decision “locally rational”? If we find out it was motivation, then we can deal with that; but if it is some other factor or combination of factors, then simply motivating won’t work. Look for which contextual factors actually are at play in the decision before you try to change the person’s behavior, and you will be much more successful at creating sustained change.