|
In everyday life, we make intuitive decisions without consciously attributing either quantitative or qualitative values to the
risks involved. But in public policy and the deployment of modern technology, decisions need to be more objectively informed; and for this, risk analysis is used.
Risk analysis is often assumed to be objective, and its results - risk values and the decisions based on them - to be correct. Yet
all stages of the process, including the techniques used, involve subjectivity. Always there is uncertainty, the need for judgment, considerable scope for human bias, and
inaccuracy. The results obtained by one risk analyst are unlikely to be obtained by others starting with the same information.
There is also a natural impediment to arriving at "correct" risk values. The future is implicit in risk, which does not
define a current problem, or a future certainty, but the potential for future harm. Thus, risk may be estimated but not measured, and its values cannot be assumed to be
correct. The United Kingdom Interdepartmental Liaison Group on Risk Assessment recognizes this in saying that risk assessment is "a tool for extrapolating from
statistical and scientific data" to arrive at "a value which people will accept as an estimate of the risk attached to a particular activity or event" [Ref. 1].
Pertinent questions are whether the estimate is a sufficiently good approximation for the purpose in hand, and what confidence there is in it.
Speaking of risk determination in technological systems, Lowrance [Ref. 2] points out that estimates of risk, whether made by
scientists or lay people, cannot escape containing elements of subjectivity. These, he said, enter into "the very defining of the questions, and into the designing of the
experiments used in assembling evidence, and then into the weighing of the social importance of the risk." It is often claimed that the greatest value of risk analysis
lies not in the values derived but in the fact that the process forces us to think deeply about, and therefore better understand, the risks.
Wharton [Ref. 3] advises us that, "Failures to cope with uncertainty in the management of technological risk abound. Their
causes include overconfidence in scientific knowledge, the underestimation of the probability or consequences of failure, not allowing for the possibility of human error and
plain irresponsibility concerning the potential risk to others." He goes on to say, "And yet to avoid such risks by adopting an overly conservative attitude to
technological innovation may be to deny the potential benefits to shareholders, employees and society." Indeed, achievement depends on accepting risks, and this is the
reason for risk analysis - we need to understand the risks so as to make informed decisions. Thus, we should recognize and allow for the subjectivity inherent in our analysis
and decision-making.
|
|
The purpose of this article is to show the subjectivity in the process of risk analysis. This is not intended as a destructive
dissection, for a major benefit of risk analysis is its subjectivity - its requirement for thought and judgment. It should not be the mere application of a set of rules.
However, where subjectivity is arbitrary and could be reduced, or where its better understanding could improve accuracy, its exposure could be beneficial. Understanding their
own subjectivity and scope for error could lead risk analysts to recognize their assumptions and consider more fully the confidence that they can reasonably have in their
results. It could also lead to research into the processes and techniques of risk analysis.
The Process of Risk Analysis
In most literature, risk analysis is divided into three stages or sub-processes:
· Hazard identification;
· Hazard analysis;
· Risk assessment (or evaluation).
The purpose of the first stage is to identify the hazards that could lead to breaches of safety. That of the second is to analyze
the identified hazards, estimate the frequency and severity of potential harm, and thus define the risks that they pose. That of the third stage is to assess the risks against
defined criteria so as to determine their tolerability.
The terms "risk analysis" and "risk assessment" are not consistently defined. Different authors use them
differently, and sometimes synonymously or interchangeably. Here, "risk analysis" refers to the total process and "risk assessment" to its final stage.
Risk analysis is generic and may be applied to any situation and any form of decision-making, from determining policy and strategy,
through all levels of planning, to tactical decision-making. The nature of the application and the purpose of the analysis influence the level of formality, the techniques
used, and whether a quantitative or a qualitative approach is taken. Thus, it is both useful and important to define a planning or "definition of scope" stage to
precede the three technical stages defined above.
Definition of Scope
The definition-of-scope stage necessarily involves the judgment of those planning the analysis. It influences the nature and
direction of the analysis, and it is a predisposing factor on its results. Decisions must be made about the study itself and the system to be studied, and both sets of
decisions involve considerable discretion. Kasper [Ref. 4] says, "The very choice of questions to be asked, issues to be considered, and methods to be used involves
judgement."
The terms of reference of a study place limitations on the admissible sources of information and on where information may be sought.
How a study is conducted (regardless of the terms of reference) can have the same effects. For example, if the public is not consulted, certain perspectives and opinions, and
perhaps the main sources of opposition to a proposal, may be precluded. Wynne [Refs. 5, 6] reports on how the Windscale Public Inquiry in 1977 was predisposed toward the
evidence of the "experts" and how this set the project's opponents at a disadvantage.
The physical and logical boundaries of the system to be analyzed, whether it is a tangible system or a proposed policy, need to be
defined. Then the analysis should identify and analyze the risks that may be posed to people, property or the environment on the other side of the boundaries. These, in turn,
depend on the inputs and outputs across the boundaries, some of which may only be observable from some perspectives, or acknowledged by some participants, so any constraint on
what data is acceptable, where it is sought, or who should provide it could severely affect the results.
The terms of reference may also place an outer boundary on the scope of the analysis. If, as in many cases, the system is an
industrial plant, and the study is limited to the risks posed within the factory, then, by definition, any risks to the public are not considered. Similarly, risk analyses may
be limited to one type of risk - such as financial risks that are of interest to management - while excluding other types, such as safety risks that the public may be more
concerned about.
However accurate the technical aspects of a risk analysis may be, the results are predisposed, or distorted, by the definition of
the terms of reference, the exclusion of certain types of evidence, the definition of the study and system boundaries, and, in general, by the strategic planning of the
analysis itself. Risk-tolerability decisions to be taken in the fourth (risk assessment) stage of the analysis are also influenced by the subjective first stage, for the
criteria to be used for these decisions should be defined then. Decision-makers need to be identified, the decision-making process defined, and its mechanisms put in place. If
politicians, the public or other non-experts are to be involved, plans need to be made about what risk information will be communicated to them, and how and when it will be
communicated. Subjectivity cannot be avoided.
|
