Realpolitik of the Blame Game

After BP’s numerous failed attempts to stop oil contamination in the Gulf of Mexico its corporate image is in ruins.  As BP’s public image reached its saturation limits for criticism, the political debate redirects public attention from the oil company to the Federal government.  With the accident rolling into its sixth week, the Obama administration is forced to answer  “Who is to blame?” and “What to do?” In short, political pressure and public anger demand swift satisfaction.

In this respect, Obama's decision to begin criminal investigations into the Deepwater Horizon explosion and the oil spill may give such satisfaction. Unfortunately, it is short-lived and may prove counterproductive in the long-run. Criminal inquiries into the cause of the oil spill is likely to narrow down the scope of post-mortem analysis to near physical and temporal proximity of the accident. In other words, the investigation will focus on people who made decisions as the accident emerged.

After the Exxon Valdes accident, the public was given a scapegoat, Capitan Hazelwood, who was singlehandedly blamed to cause the accident. Although, he was later cleared of the charge being drunk at the time of the accident, he had never cleared his name from the stigma of causing of the biggest oil spill in US history (until current accident).   There were more systemic problems such as functional and interoperable navigation systems on board the ship and on the shore of Alaska, the culture of tolerance to safety problems among the oil executives and public official, etc. However,  systemic factors tend to “spread the blame” and lack the luxury of targeted blame, thus satisfaction of punishment.  Simple and straightforward causes are always easier to describe and accept. I hope to be wrong, but it is likely that we can expect several Capitan Hazelwoods from BP and its subcontractors who could be “scarified” to satisfy public outcry.  However, it is unlikely that such narrow view can help us understand the complexity of the accident.

It is in our human nature to go back in time and look for a single event, which could have prevented to the catastrophe from happening.  The bigger the accident, the stronger is our temptation to reverse time. Inability to reverse the situation results in anger, dissatisfaction and need for revenge. These emotions have to be channeled one way or the other. What make the case with the oil spill unique is the fact that the public is given a rare opportunity to be part of the accident as it emerges.  The public feels helpless to reverse time and to contain the accident. Old and new media channels enables people to get first-hand accounts of the environmental impact on the ocean's seabed, the coast and wildlife. As the result, it is natural for people to ask who is in charge and who is to blame. If those who are in charge do not act swiftly and find those who are to blame, the public could blame those who are in charge.  If those who were in charge yesterday do not blame those who are in charge today, then those who were in charge yesterday are to get blamed themselves and vice versa. The same is true with BP and its numerous subcontractors.

Unfortunately realpolitik of the blame game creates an environment in which each player is concerned with own political and legal liabilities. Criminal inquiry adds more gasoline to the flame.  Starting the investigation before accident’s containment is almost like investigating causes of an air crash with airplane still in the air. Perhaps, such approach could give unique perspective, but more likely it will downgrade itself to simplified description of immediate events with operators and decision makers who were unfortunate enough to be in close proximity as the accident emerged.

"The Logic Of Failure: Recognizing And Avoiding Error In Complex Situations"

Great book written in 1989 by West German Professor Dorner. This is a good introduction into System dynamics and System thinking. The book gives several mental models of how people "attack" complex problems and why they often fail. Dorner describes several psychological experiments that help to differentiate "bad decision maker" from "good decision makers," as well as mental traps that lead to failures. Based on this analysis he presents his model of decomposition and planning. At the end of the book Dorner quotes Clausewitz that "...War is not an infinite mass of minor events... War consists rather of single, great, decisive actions, each of which needs to be handled individually." Such "strategic thinking" requires far greater expenditure of mental energy argues Dorner.

Dorner ends his book with the thought that "...If we cannot form a picture of a temporal configuration, we cannot adjust our thinking and actions to take that temporal pattern into account... We human being are creature of the present. But the world today must learn to think in temporal configuration."

The book is easy to read and comprehend. I gave it to my 15 years old son to read it. Hopefully he will use my advise to read it earlier in his life.

"The Field Guide to Understanding Human Error" by Sidney Dekker

I met Professor Dekker couple of weeks ago during his visit to Engineering System Division at MIT.

He is definitely one of the most intelligent and sharp thinking persons I have ever met. I was impressed with his clarity and speed of thought. I could catch with his speed of thinking only after he made an iterative loop in his argument.

There are two major themes that are reflected throughout his book. First is that human error is not the cause but a symptom of a trouble. Accepting this notion is first step towards New view of human error, which can never be the conclusion of an investigation, but rather its starting point. Understanding error means reconstructing the context in which such decision was made. "Human error problem is an organizational problem.  This means that understanding human error hinges in understanding the organizational context in which people work."

People tend to react post factum and backwardly reconstruct sequence of events in a linear manner.  In the aftermath of an accident, one can easily list logical arguments how and why people should have foreseen and prevented the upcoming events. We easily judge people, who failed to take proper actions. We focus on their personal shortcomings such as absence of proper training and experience, health conditions and hours of proper sleep, etc.  We usually tend to focus our attention on people who happened to be closest to the accident in term of time and space.

Second theme of the book is forward looking metal model, which focuses on preventing accidents in the future rather than analyzing the accident with a hindsight bias. Hindsight gives the investigators better and more complete information in comparison to people who made the decision prior to the failure.  It provides with facts that become midpoints in the linear logic flow that the investigator reconstructs. As we walk backwards, each facts is perceive not as an "intersection point" with a list of equally valid choices, but rather as point of the process when/where incorrect decision was recorded. Hindsight bias exaggerates the importance of the recorded facts versus other events that are not directly related to the specific accident.

The author compares three accident models (1) the sequence-of-event, (2) the epidemiological model and (3) the system model. He argues that the latter is a holistic approach that looks at accidents as emerging form interactions between system components and processes, rather than failures within them.

Dekker warns against "quick fixes" and misuse of technical labels, which do not describe the gap between reality and our judgments, whereas "safety improvements come from organizations monitoring and understanding the gap between procedures and practice." He concludes that "a safety culture is a culture that allow the boss to hear bad news."

Congress, Goldman Sachs and the Hindsight Bias

Being a student of System safety I could not ignore the public hearing that is taking place in the US Senate on Goldman’s handling of high-risk mortgage business in 2007.

The bigger is the failure or the devastation caused by an accident the more difficult it is to reconstruct the preceding events, processes and systematic factors.  In the case of Goldman, its handling of highly securitized mortgage deal called Abacus, is overshadowed by even larger devastation of the financial crisis that went far beyond the Wall Street.  The public outrage with the economic crisis that has been boiling for the last two years has suddenly found actual addressee, Goldman Sachs Group, thanks to the SEC’s investigation and the independent inquire by the Senate’s Permanent Subcommittee on Investigations.

Goldman’s interaction during the hearing fuels the outrage even further as it treats testimonies of its executives as part of the group’s upcoming legal battle with the SEC.  By fighting against possible legal litigations and denying any wrongdoing, Goldman gets deeper into the trap of becoming the villain, who in public’s eyes is blamed for the financial crisis.  On top of that, the tug of war between two parties in the Congress makes the Goldman’s hearing major strategic battle that will allow the ruling party push its bill on financial regulations.

The goal of this essay is not to defend Goldman, but to explain the phenomena of hindsight bias. As people try to make sense in such confusing environment, it is easier to focus one’s attention on specific individuals and become bias towards the end result.  Sidney Dekker, the Swedish system safety expert, describes the hindsight bias in form of four different, but interrelated reactions: (1) Retrospective, (2) Counterfactual, (3) Judgmental and (4) Proximal (The Field Guide to Understanding Human Error, 2006).

People tend to react post factum and backwardly reconstruct sequence of events in a linear manner.  In the aftermath of an accident, one can easily list logical arguments how and why people should have foreseen and prevented the upcoming events. We easily judge people, who failed to take proper actions. We focus on their personal shortcomings such as absence of proper training and experience, health conditions and hours of proper sleep, etc.  We usually tend to focus our attention on people who happened to be closest to the accident in term of time and space.  In the case of Goldman Sachs, Fabrice Tourre was designated as the sacrificial lamb, who will be slaughtered to satisfy the crowd.

Hindsight gives the investigators better and more complete information in comparison to people who made the decision prior to the failure.  It provides with facts that become midpoints in the linear logic flow that the investigator reconstructs. As we walk backwards, each facts is perceive not as an “intersection point” with a list of equally valid choices, but rather as point of the process when/where incorrect decision was recorded. Hindsight bias exaggerates the importance of the recorded facts versus other events that are not directly related to the specific accident.

Specific sequence of events shapes its post-factum interpretation. It connects events in specific post factual manner. What Sidney Dekker calls tunnel vs. zigzag.  The tunnel vision that the bias creates, oversimplify the history of the accident, with linear logical flow and binary (right and wrong) choices.  As part of the oversimplification process, causality of events are oversimplified as well. As the result, the hindsight prevents us from differentiating between what is know after the accident and what was known by people prior and during the accidents.

What are the objectives of the Congress and the SEC? Get more popularity with the public; re-direct public dissatisfaction toward the Wall Street; take down Goldman Sachs; push for new regulations; etc? As in any other accident investigation, the public, the Congress and the SEC will focus on assigning blame on particular individual instead of focusing on learning the root causes, especially if the objectives of the blame are so vague.

How Safe Is Safe Enough?

Safety is a public good because the “consumption” or enjoyment of safety by one individual does not diminish the “availability” of safety for other individuals (assuming that the value of each human life is equal to the value of any other human life).  Government is elected by citizens of the society and is given official authority to enforce rules and regulations. Thus the government is given official power through enforcement of laws and regulations to maintain safety for the benefit of all citizens within its jurisdiction.  Magnitude of an accident (and its consequences) may affect each and every individual in the society or a random group of individuals. Theoretically, it is in general interest of all citizens to oblige with the rules and maintain safety.  In reality, “the picture” has many conflicting nuances.

As individuals we react to immediate pain.  Pain simplifies our point of view and helps us to survive (Minsky, 1985). Pain is build-in alarm system that informs human brain about immediate threats to its body’s existence.  As humans we tend to react to most immediate threats. When threats (and/or health hazards) are remote in terms of time and impact – such as smoking, breathing poisonous air, eating trans-fat food, etc – our reaction is not as swift or as rational.

A societal equivalent of human pain is loss of human life.  As a community or a society, people tend to react to accidents that involve loss of human lives with great attention.  The risks that are not visible or well understood do not get public attention until fatal accident actually takes place. As individuals, human society in general does not have high attention to threats and problems that are far in time and space.

Nothing can be valued more than human life. However, there is a distinction between the value of an exiting human life and the value of a deceased human being.  The later case was a substitute for blood feud or what is called vendetta, where attaching price to diseased life was the way to avoid further bloodshed. In fact this approach has been considered as more civilized conflict resolution for centuries (what is know as blood money). The moral line is drawn between the world of alive and the world of deceased. Those who attempt to take “the price tag” from the world of deceased and use it in our world make rational analysis, which is immoral at the same time.

In my point of view, there are two important trends worth noticing:

(1) As human society become more dependent on complex systems, there seems to be slow convergence of system safety and public health. It is clear that the fundamental of two fields is the same – the value of human life. However, it is not clear whether such oversimplified approach can be beneficial to either system safety or public health. This is just an intuitive feeling which I hope to understand during the course.

(2) Availability of information and its widespread and fast dissemination leads to greater public awareness. Easily available instruments and tools (email, matlab, google earth, skype, youtube, simple chemical tests and pH strips, individual radioactivity sensors, etc) lead to increased role of individuals in dealing with safety, which in turn influence behavior of governments and corporations. Availability of information also builds much richer postmortem picture of an accident and different accident scenarios, where ordinary citizens are able to question and challenge official reports and causes of an accident.

The simplest answer to the question “who should be responsibility for risk management?” is government and its regulatory agencies.  However, we know that the government agencies are managed and run by individuals. These individuals, as many of us, can manage only a certain level of complexity. There is no guarantee that they can integrate all the pieces together and maintain public safety all the time.

I think (“which implies that I do not as yet know so”), since safety is a public good, it has to be collective responsibility of each and every citizen in the society to think in cohesive and responsible when it comes to safety.

The role of the legal system depends on the end–result that society expects from its courts.  There must be a reason why the US legal system adopted the civil (tort) approach towards safety hazards (besides obvious corporate interests). It is my understanding that such approach was in part chosen to shift attention from responsibility of individuals to corporate responsibility, and eventually to accidents/hazards causes. It is difficult to judge if the system is functioning effectively or not since we have no information how many accidents/hazards were prevented just by potential possibility and danger of civil litigation.  We can only see the cases that are floating on the surface, in which corporations are using complex legal procedures to avoid, mitigate or postpone expensive settlements.

Criminal charges, on the other hand shift the focus from the accident to particular individual responsibility. It is often the case where both the government (which could be under public pressure) and the public itself are eager to “teach a lesson”. The expectation is shifted from understanding wholesale list of accident’s causes to finding specific “target” or “villain”.  Other potential “targets” are giving limited information about the accident reinforcing the initial bias, redirecting the blame and sacrificing the least protected “target.”

In the case of tainted milk, the Chinese government decided to teach a lesson. Three individuals were sentenced to death and 21 others, “mostly dairy producers and middlemen, were given terms ranging from two years to life in prison” (http://www.nytimes.com/2009/01/22/world/asia/22iht-milk.3.19601372.html).  Such lesson is likely to influence behavior of Chinese businessmen and even prevent potential food poisoning. One can argue that the trial may bring limited benefits to the society, even if it does not address the structural problem of food safety.

Here I would like to address the issue of Human error and Learning from Mistakes described (Flatch et al.). I think that we need to decouple <understanding the accident> from <learning from the accident>. I argue that it could be possible to learn from an accident even if complete description of an accident is not determined and vice versa. In any of the four options, the end result of the accident is the same.

	Learning form an accident	Failing to learn from an accident
Understanding the accident	X	X
Failing to understand the accident	X	X

So in the case of tainted milk, even if full picture of the food safety hazard has not been found or at least publicly recognized (i.e. export of the poisoned food products), there is still a possibility that appropriate lessons are learned. In the case of uncontrolled acceleration (Audi, Toyota, GM), even if the cause(s) of the fault(s) was known, the management has not learned the lesson.

This takes us back to individual reaction to physical pain.  Perhaps criminal prosecution of individual executives (in this case imprisonment) could be more effective approach in achieving public safety at the cost of objective accident investigation.

The biggest doubt here is that by focusing on individual responsibility we do not solve the challenges of potential accidents, which could be greater in power and magnitude as our society becomes more complex and interdependent.