Abstract of Meeting Paper

Society for Risk Analysis-Europe 1997 Annual Meeting

How Should Governments Solve Ambiguous Decision Problems? Theory, Food Scares and Organisational Biases. Paul Anand, De Montfort University; and Milton Keynes (http:/www.mk.dmu.ac.uk/risk)

The central theme of the paper is that a number of the decision problems involving hazards associated with, or alleviated by, new technologies are appropriately modelled as decisions made under uncertainty (with no probabilities) as distinct from risk. This distinction, which can be found in work by statisticians and economists gives to a representation of decision problems in which actions can only be evaluated in terms of their consequences. Furthermore, there is a set of rational decision rules (maximin, maximax, insufficient reason) which can be applied to decisions under uncertainty and which, in general, can lead to different rankings of actions. The framework (for uncertainty or ambiguity) has been used extensively by theorists and experimentalists but there are relatively few empirical investigations that use it to understand policy decision making. This paper represents an attempt to address that gap.

The framework/paper is structured as follows. Section I provides an introduction to relevant normative frameworks that might be applied to social choice problems characterised by uncertainty. Section 2 reports on a comparative analysis of decision-making under risk (Salmonella in Eggs) and under uncertainty (BSE in Beef for the period 1987-93). Section 3 provides a discussion of the empirical evidence in the light of normative theory.

In the discussion we seek to establish an understanding of some of the links between the normative ideal and institutional reality. In particular, we ask whether the risk-uncertainty distinction makes sense in practice, whether it is recognised by practitioners and what the consequences are. We make remarks on the institutional context in which ambiguous decisions are made in the UK and identify organisations biases (organisational corollaries to individual biases that are now well understood in the behavioural decision-making literature) that may impair the quality of decision-making. The biases identified and discussed are:

Organisational Biases

1. Problem Misspecification: incorrect use of hypothesis testing with sparse data and failure to give sufficient weight to questions of (non-statistical) significance
2. Theory Aversion: inappropriate use of arguments based on claims concerning the
3. Digging In: difficulties that organisations have revising beliefs in public settings,
4. Over Production of Casual Information: reactions to uncertainty lead to diversions of research funds into projects which may have little impact on decisions
5. Exclusive Use of Expertise: narrow definitions of appropriate expertise lead to unstable decisions
6. Idealistic Models of Media Behaviour: a variety of misconceptions prevent organisations from making the most of their relations with the Press

Go to:

    Table of Contents of the 1997 SRA-Europe Conference
    Program of the 1997 SRA-Europe Conference
    RiskWorld Welcome Page
    Tec-Com, Inc.