**14. On the hindsight objection and the problems with risk management**

Before departing from this topic, we must address an objection commonly brought forth, which in the designation of the author may be labeled, the 'Hindsight Objection'. The following is the 'Hindsight Objection'. There are always warnings of disasters and they are commonly ignored. When no disaster occurs, these warnings are forgotten. If one went back to every successful venture; e.g. a space flight, one would find the ignored warnings. Therefore, if one infers from warnings that an unsuccessful flight occurred, one is inferring from unwarranted premises. Whenever one goes back and traces the causes of disasters to unheeded warnings, one is justifying warnings through hindsight which is always 20-20. Hence, the label, the Hindsight Objection.

There are two major replies to this objection. The first reply is that it is entirely hypothetical. In order to make good on the Hindsight Objection, one must bring forth evidence to support it. In other words, one must take a successful venture; e.g., a space flight, and show first of all that there were red-flagged warnings that the flight should not have taken place. If the 'red-flagged' designator is not in use, the warnings must be shown to be of high-priority status, i.e., it should be demonstrated that such warnings were equivalent in status to the warnings not to launch disastrous flights, such as the dire warnings that were lodged in the attempt to stop the launch of the *Challenger*. Warnings must be vetted in terms of the criteria spelled out above, in terms of source, form and content. What is commonly brought forth as "hindsight evidence" in the case of the Challenger are operational parts which are designated as Criticality 1, of which there were many such on the Challenger which did not fail. A Criticality 1 designation indicates that the failure of such a part would result in the loss of the vehicle and human life. But a Criticality 1 designation is not equivalent to a redflagged warning that the part designated as Criticality 1 was likely to malfunction. One cannot point to parts with a Criticality 1 designation that did not fail as evidence that there were warnings that failure was imminent. A part with a Criticality 1 designation could, theoretically, be extremely safe. To count as a legitimate warning, there must be a specific, high priority (red-flagged) warning concerning the faulty design or operational capacity under certain weather conditions, etc., of a part which possesses a Criticality 1 designation.

<sup>35</sup> *Cf*., Peter Lattman and Jenny Anderson, 'For 92nd St. Y, a Break from Wall Street Worry,' *NY Times*, November 29. 2011. In this article, this guarantee is offered by the hedge fund manager, John A. Paulson.

This first condition, the prior presence of red-flagged warnings that meet the above criteria in terms of source (knowledgeable expert witness), form (red-flagged), and content (spelling out the specific flaw in the part), to the best of my research and knowledge, have never been met. The Hindsight Objection is always made as a purely hypothetical objection without any evidence for its truth value ever put forth. In the case of space flights (this example is used because it is best known to the author), one must show that there were other launches in which the senior scientist issued repeated red-flagged alerts and all of the engineers and managers voting on the planned launch voted unanimously against the launch. Has anyone ever brought forth any evidence in support of the Hindsight Objection claim?

150 Risk Management – Current Issues and Challenges

Hence, the label, the Hindsight Objection.

on Wall Street that guarantees any loss of any client's investment up to four million US dollars.35 How does this company do this? While the author of this chapter cannot answer this question, it would be assumed that this company is sufficiently profitable such that it can make this guarantee. What is notable is not that it is capable of making this guarantee; what is notable is that it makes the guarantee. By making such a guarantee this company is making the statement that it regards the potential loss of its client's funds as completely unacceptable. This standard of ethics is apparently commercially viable. This is an example, not of practicing risk management in the context in which it is most commonly accepted (that of a fund manager),

but rather of practicing not taking risk in the first place (with other people's money).

**14. On the hindsight objection and the problems with risk management** 

Before departing from this topic, we must address an objection commonly brought forth, which in the designation of the author may be labeled, the 'Hindsight Objection'. The following is the 'Hindsight Objection'. There are always warnings of disasters and they are commonly ignored. When no disaster occurs, these warnings are forgotten. If one went back to every successful venture; e.g. a space flight, one would find the ignored warnings. Therefore, if one infers from warnings that an unsuccessful flight occurred, one is inferring from unwarranted premises. Whenever one goes back and traces the causes of disasters to unheeded warnings, one is justifying warnings through hindsight which is always 20-20.

There are two major replies to this objection. The first reply is that it is entirely hypothetical. In order to make good on the Hindsight Objection, one must bring forth evidence to support it. In other words, one must take a successful venture; e.g., a space flight, and show first of all that there were red-flagged warnings that the flight should not have taken place. If the 'red-flagged' designator is not in use, the warnings must be shown to be of high-priority status, i.e., it should be demonstrated that such warnings were equivalent in status to the warnings not to launch disastrous flights, such as the dire warnings that were lodged in the attempt to stop the launch of the *Challenger*. Warnings must be vetted in terms of the criteria spelled out above, in terms of source, form and content. What is commonly brought forth as "hindsight evidence" in the case of the Challenger are operational parts which are designated as Criticality 1, of which there were many such on the Challenger which did not fail. A Criticality 1 designation indicates that the failure of such a part would result in the loss of the vehicle and human life. But a Criticality 1 designation is not equivalent to a redflagged warning that the part designated as Criticality 1 was likely to malfunction. One cannot point to parts with a Criticality 1 designation that did not fail as evidence that there were warnings that failure was imminent. A part with a Criticality 1 designation could, theoretically, be extremely safe. To count as a legitimate warning, there must be a specific, high priority (red-flagged) warning concerning the faulty design or operational capacity under certain weather conditions, etc., of a part which possesses a Criticality 1 designation.

<sup>35</sup> *Cf*., Peter Lattman and Jenny Anderson, 'For 92nd St. Y, a Break from Wall Street Worry,' *NY Times*, November 29.

2011. In this article, this guarantee is offered by the hedge fund manager, John A. Paulson.

The second major objection is that even if one could offer an example of a flight which met the equivalent set of warnings that were issued in the case of the *Challenger*, one might miss the point that such warnings are not intended to be construed to be 100% reliable predictions concerning the particular flight in question. Such warnings are not meant to be on the level not of, if this flight will take place, but of when this flight, or a similar flight, takes place, such a likelihood of such a flight ending in disaster is a horrific eventuality. If the warnings are not justified on the basis of one flight, they might well be on the basis of the third or the seventh flight. When one leaves the auto mechanic's shop with faulty brakes, the service manager might warn the customer that the brakes may fail. If the brakes do not fail on the first hill, that does not mean that her or his warning was without value. They may fail on the ninth hill. Thus, even if one could find a case in which dire warnings that fit the above criteria were present and the flight or sail went without incident, that would not prove that all such flights were to be considered safe.

It is informative to reflect on the fact that Boisjoly's warnings cover both the likelihood of the possibility of the *incidence* of the occurrence of the horrific failure of the mission and the *consequence* of the death of the crew and passengers. Thus, both aspects of risk are covered: the possibility of the likelihood of the incidence and the gravity of the consequence. This example of risk taking fits both the criteria of the likelihood of the possibility of the incidence and the enormity of the consequences. Such risk taking is entirely incompetent and unconscionable. *That it need not be restricted in its eventual occurrence in this flight is evidenced by the fact that it was warned against eight years previous.* The resistance to launch on this flight was based on the fact that the weather conditions compounded the already existent risk. *The problem was that the case of the decision to launch the Challenger was an iconic case where it was thought that risk could be managed. In other words, to some extent, the very concept of risk management was at fault.* A better example of a safeguard would have been never to have installed this unsafe part in the first place. That would have been an example not of risk management, but rather of not taking risk in the first place. (After that fatal flight, one of the original designs that was originally rejected was chosen to be used). More obviously, *not launching in adverse weather conditions would be an example of not taking risk*. *Launching in dangerous weather conditions is an example of attempting to manage risk*. It comes under the thinking of, 'the weather is not good, but we can manage it'. It is not clear what this means. It seems to suggest the belief that 'the weather is not good, but we can chance it.' 'Risk management', when properly analyzed, seems to be equivalent to 'risk chancing' or gambling'. The eminent Nobel laureate physicist, Richard Feynman likened the decision to launch the *Challenger* space shuttle to playing Russian roulette. The proper way of

thinking would have been, 'the weather is not good; we are not going to *take* the risk'. *When we frame the decision to be taken in term of risk taking rather than risk managing, it is far more likely that we will act not to take the risk rather than to act to take the risk and then attempt, somehow, to "manage" it.* 

On the Very Idea of Risk Management: Lessons from the Space Shuttle *Challenger* 153

The idea of "risk management" implies that risk must be present. We must ever guard against complacency. If we keep uppermost in mind that the preciousness of life is our highest priority, we will ensure that the presence of the possible incidence and the consequence of risk is kept at the lowest possible point. *If we change our language from the language of "risk management" which implies that somehow there must always be a risk present, and it is our task to manage it, to risk taking, we will be more alive to the ethical responsibility which is involved in taking risks with our own or other people's lives.* We will be more inclined to work sincerely to minimize

The notion of providing enough safety boats for half of the passengers (the model of the *Titanic*) fits perfectly into the concept of "managing risk". When we do this, we have performed some kind of cost-benefit analysis, or, to speak more strictly, some kind of probability-benefit analysis, and have reached a decision that by providing life boats for only half of the passengers we are fulfilling the responsibility of "managing risk". It is not clear how this decision was reached. Perhaps, it was reached by assuming that only half of the passengers would make it to the life boats, hence, by providing half of the needed boats,

In conclusion, if we use the language of risk taking, the provision of life boats for half of the passengers on board is still a case of *taking risk with half of our passengers' lives*. On the other hand, if we take the language of risk taking seriously and consider that human life is precious, we would not sail, that is, we would not take risk, unless there were enough life boats provided for 100% of the passengers. (If we rely upon life preservers, for example, those in ocean waters would only live for a few minutes because of the icy temperature of the sea).

*Everything in the area of risk management is a matter of ethics*. Do we value human life? What do we mean by the phrase 'acceptable risk'? Such a concept can only be tolerated if there are safety provisions that will fully protect and preserve life if the consequences of the risk taking threaten human life. Otherwise, there is no such concept as an 'acceptable risk'. Was the decision to provide life boats for only half of the passengers aboard commercial vessels based on past performance data? In the case of the *Titanic*, lifeboats for only half of the passengers were provided and, as a result, the lives of 1,523 men, women and children were lost.36 1,523 men, women and children were killed based on risk management. The *Titanic* disaster was in 1912. Now, 100 years later while the custom is to provide enough lifeboats for all passengers, the lifeboats are prepared with engines that will leave their occupants at

It is hoped that this introduction to the new idea of risk taking as opposed to risk management will create reflection and commitment to a greater ethical sensitivity when we consider how our decision making may affect other people's lives. Do we ever have a right to take decisions that risk other people's lives? If the message of this chapter is heard, then we do not ever have the right to risk other people's lives unless we provide full and adequate protection for the consequences of the risk that we are taking with other people's lives. If we change our language habits from commonly using the phrase 'risk management'

the possibility of risk and to reduce the effect of the consequences of risk.

we have "managed" the risk.

the risks of the high seas.

<sup>36</sup> *Ibid*., p. 87.

It is therefore worthwhile to consider abandoning the concept of risk management and replacing it with the concept of risk taking. When one removes the euphemism of risk management and replaces it with risk taking, it becomes abundantly evident that it is actual human life with which one is taking risk. One is playing G-d with human life. *The very concept of risk management is itself too risky*. The seemingly objective social science language of 'risk management' is in reality a license to treat human life lightly.
