This experiment tested predictions derived from a social contingency model of judgment and choice that identifies 3 distinctive strategies that people rely on in dealing with demands for accountability from important interpersonal or institutional audiences. The model predicts that (a) when people know the views of the audience and are unconstrained by past commitments, they will rely on the low-effort acceptability heuristic and simply shift their views toward those of the prospective audience, (b) when people do not know the views of the audience and are unconstrained by past commitments, they will be motivated to think in relatively flexible, multidimensional ways (preemptive selfcriticism), and (c) when people are accountable for positions to which they feel committed, they will devote the majority of their mental effort to justifying those positions (defensive bolstering). The experiment yielded results supportive of these 3 predictions. The study also revealed some evidence of individual differences in social and cognitive strategies for coping with accountability.Many writers have criticized cognitive social psychology for its apparent indifference to the interpersonal and institutional settings within which people make judgments and choices (e.g., Gergen, 1982;Sampson, 1981). Although these critiques make important points, it is necessary to advance beyond metatheoretical position papers and to specify the particular ways in which interpersonal and institutional variables interact with cognitive tendencies of the perceiver to shape how people actually make up their minds.Recent research on accountability points to one possible answer. Tetlock (1985a) has argued that accountability is a universal feature of everyday decision-making environments. Accountability, from this standpoint, is a critical rule-and normenforcement mechanism: the social psychological link between individual thinkers on the one hand and the social systems to which they belong on the other. The fact that people are ultimately accountable for their decisions is an implicit or explicit constraint on virtually everything they do. Failure to behave in ways for which one can construct acceptable accounts leads to varying degrees of censure-depending, of course, on the gravity of the offense and the norms of the organization (Schlenker, 1982;Scott &Lyman, 1968;Tetlock, 198 la).A good deal of experimental evidence indicates that accountability pressures can affect both what people think (the beliefs and preferences they express) and how they think (the reasoning

This research demonstrated that accountability can not only reduce judgmental bias, but also exacerbate it-in this case, the dilution effect. Ss made predictions from either diagnostic information alone or diagnostic information plus mixtures of additional data (nondiagnostic information, additional diagnostic data pointing to either the same conclusion or the opposite conclusion). Relative to unaccountable Ss, accountable Ss (a) diluted their predictions in response to nondiagnostic information and (b) were more responsive to additional diagnostic information. The accountability manipulation motivated subjects to use a wide range of information in making judgments, but did not make them more discriminating judges of the usefulness of that information.Cognitive social psychologists have painted numerous portraits of the person as information processor. Early work emphasized the rigorous rationality with which people analyzed and drew inferences from evidence: the correspondent inference model, the causal schemata model, the covariation model, and the Bayesian model.

No abstract

This study assesses the impact of accountability, the status quo, and anticipated costs and benefits on judgement of the acceptability of a drug in the US pharmaceuticals market. Several effects are documented: (1) subjects were more accepting of a drug, the lower the anticipated risks of side effects and the greater the anticipated benefits; (2) subjects were especially unwilling to accept risk when the drug was not yet on the market and when they felt accountable for their decisions;( 3 ) accountable subjects confronted by an off-the-market drug that posed moderate or high risk were also especially likely to procrastinate, to buckpass. and to think in integratively complex ways about the problem, notwithstanding the fact that many more lives would be saved than lost. We explain these results by positing that perceptual-cognitive processes (loss aversion) and political processes (blame avoidance) mutually reinforce each other when decision makers are accountable for choices that raise the possibility of changing the status quo in ways that impose losses on identifiable constituencies. We conclude by commenting on the complex normative issues that arise in labelling response tendencies as biases. Choices that look irrational within one ethical or political framework sometimes appear quite reasonable with another. KEY WORDS Accountable Risk Rationality PharmaceuticalsPsychological research on judgement and choice addresses fundamental questions of human rationality: When do people draw inferences or make decisions in ways that are biased, flawed, or maladaptive? What normative models -Bayes' theorem, expected utility maximizationshould we use in gauging whether people have indeed made mistakes? How robust are the observed deviations from rational actor standards? (See Fiske and Taylor, 1991; Payne et al., 1992;Kahneman et al., 1982.) The social contingency model of judgement and choice both complements and contradicts this burgeoning research literature (Tetlock, 1992). The model begins with the uncontroversial claim that people make the vast majority of decisions in social settings in which they feel accountable (to varying degrees and in varying ways) to others. The model then goes on to make a series of more complex and controversial claims. It maintains, for example, that under certain conditions, accountability pressures will motivate people to engage in pre-emptive self-criticism and to anticipate objections 08943257f94JO 10001-23s 16.50 Not Aoaountable, S t a t u s QuoAooountable, New Dnrg LOW WSK M B M M m H w m Exhibit 1 . Acceptability of drug as a function of risk level, accountability and market statusin the other three conditions: M(I5) = 5.78 versus M(45) = 3.40. This interaction was again highly significant, F(2, 1 12) = 6.50, p < 0.0001. Again, .the results of the Turkey HSD test were unequivocal: the medium-and high-risk judgements of accountable subjects evaluating an off-the-market drug were not significantly different from each other, forming their own grouping, and were significantly different...