This article reviews the now extensive research literature addressing the impact of accountability on a wide range of social judgments and choices. It focuses on 4 issues: (a) What impact do various accountability ground rules have on thoughts, feelings, and action? (b) Under what conditions will accountability attenuate, have no effect on, or amplify cognitive biases? (c) Does accountability alter how people think or merely what people say they think? and (d) What goals do accountable decision makers seek to achieve? In addition, this review explores the broader implications of accountability research. It highlights the utility of treating thought as a process of internalized dialogue; the importance of documenting social and institutional boundary conditions on putative cognitive biases; and the potential to craft empirical answers to such applied problems as how to structure accountability relationships in organizations. Accountability is a modern buzzword. In education (Fairchild &

We report two experiments which compared individual and group behavior in the twoplayer centipede game (Rosenthal, 1981). In Experiment 1 the joint payoff of the two players increased linearly as the game progressed whereas in Experiment 2 it remained constant. In both experiments the game was played either by two individuals or by two (three-person) groups. The members of each group could communicate freely to decide whether to exit or pass at each node of the game. As the game ended, each group member received an equal share of her group’s payoff (which was three times the payoff in the individual treatment). The game-theoretic solution for the two treatments is identical, namely for the first mover, whether an individual or a group, to exit at the first node. Nonetheless, we found that in both experiments groups chose to terminate the game at an earlier stage than individuals. Group behavior, in other words, was more rational than individual behavior. Possible explanations for this effect are discussed.

This paper develops and tests a model of how organizational structure influences organizational performance. Organizational structure, conceptualized as the decision-making structure among a group of individuals, is shown to affect the number of initiatives pursued by organizations, and the omission and commission errors (Type I and II errors, respectively) made by organizations. The empirical setting are over 150,000 stock-picking decisions made by 609 mutual funds. Mutual funds offer an ideal and rare setting to test the theory, as detailed records exist on the projects they face, the decisions they make, and the outcomes of these decisions. The independent variable of the study, organizational structure, is coded from fund management descriptions made by Morningstar, and the estimates of the omission and commission errors are computed by a novel technique that uses bootstrapping to create measures which are comparable across funds. The findings suggest that organizational structure has relevant and predictable effects on a wide range of

Abstract: In this essay, I discuss the difficulty of sustaining an inquisitorial system of policy research and analysis when it is embedded in a broader adversarial political setting. Conflicts of interest in public policy research exist on a continuum from blatant pecuniary bias to more subtle ideological bias. Because these biases are only partially susceptible to correction through individual effort and existing institutional practices (peer review, replication), I consider whether a more explicitly adversarial system might be preferable to the awkward hybrid that exists today. But there are important disanalogies between policy-relevant empirical debates and the kinds of conflicts we address with our adversarial legal system. If we are stuck with a muddled inquisitorial-adversarial hybrid, we need to encourage norms of “heterogeneous inquisitorialism, ” in which investigators strive for within-study hypothesis competition and greater clarity about roles, facts, and values. THE VARIETIES OF CONFLICT OF INTEREST In public policy research, as in other domains of professional life, conflicts of interest (henceforth, COIs) are legion. Most policy researchers can readily provide many personal war stories from their professional experience. Generically, the most blatant cases tend to fall into four categories: 1. Investigators with a commercial or proprietary interest in the research outcome, or the use of funding from sources with a commercial or proprietary interest in the research

A dynamic model of group performance is suggested that combines the group learning approach and the combination of contributions approach. Three hypotheses are tested in two experiments, comparing individual training conditions with mixed group and individual training conditions on subsequent nominal and collective group performance of rule induction tasks under identical time constraints. As predicted, collective group performance improves as a function of group experience, nominal group performance improves as a function of improved individual resources for performing the task individually, and group experience reduces process losses by improving individuals’ ability to collaborate efficiently. Several experiments from the literature that address issues of group learning are analyzed in the light of the propositions made in the dynamic model. Overall, theoretical and experimental approaches to investigating group learning phenomena are discussed.

Previous research has shown an ‘interindividual-inter-group discontinuity effect’: inter-group interactions generally lead to less cooperative outcomes than interindividual interactions. In this dissertation, I replicate the discontinuity effect in the deterministic prisoner’s dilemma, but find that groups are more cooperative than individuals in a stochastic version of the game. Three major factors that underlie the usual discontinuity effect, were reduced in the stochastic environment: greed, fear, and persuasion power. Two group mechanisms are proposed to explain the reversed discontinuity effect: the motivation to avoid guilt and blame when making decisions that affect others ' welfare, and the social pressure to conform to certain norms when one is in a group setting. Follow-up studies reject the social pressure mechanism, but confirm that guilt aversion and blame avoidance drive groups to be more risk concerned than individuals and more likely to invest to reduce risks when uncertainty is present. There is also evidence thatiii uncertainty reduces inter-group distrust that usually exists in group interactions, and may even increases inter-group trust to be higher than inter-individual trust. iv Contents ii

Psychologists have used laboratory experiments to demonstrate that individual judgment and choice behavior systematically violates the assumptions underlying rational choice theories. In principle, these cognitive and motivational biases have important implications for economic theory and policy. But critics contend that various aggregate-level processes (group discussion, market transactions) attenuate or compensate for individual biases (e.g., Kagel & Roth, 1995; Page & Shapiro, 1992. Thus, recent experiments by social psychologists and experimental economists have tested whether interaction with others attenuates individual biases, sustains them, or amplifies them. Are groups more or less biased than individuals? No simple conclusion has emerged from this research (see reviews by Kerr, MacCoun, and Kramer, 1996a; Tindale, 1993). Some studies find a clear reduction of bias at the aggregate level, while others find that groups are significantly more biased than individuals acting alone. To identify some of the conditions that determine relative bias (group bias- individual bias), and to help clarify discrepant research findings, my colleagues Norb Kerr and Geoff Kramer and I (Kerr, MacCoun, & Kramer, 1996a, 1996b) have conducted theoretical "thought" experiments. Our work suggests that relative bias is a function of the type of bias, the applicable

Experimental studies consistently indicate that human information processing and decisionmaking violate basic precepts of rationality. Yet rational choice theory is increasingly used to model organizations, politics, and international relations. Experimental evidence of cognitive bias is often discounted as a methodological artifact because analysis, organization, specialization, and the presence of strong incentives are presumed to eliminate bias outside the laboratory. A controversy in the historiography of the Paris Peace Conference of 1919 provides an opportunity to address that assumption. Did “the

The authors selected 8 ordered quantities from smallest (1st) to largest (8th) from each of 36 domains, such as population of countries and production of commodities. Given the 1st and 8th (broad domain boundaries), 2nd and 7th (medium boundaries), 3rd and 6th (narrow boundaries), 2nd and 3rd, 6th and 7th, or none of the quantities, participants estimated the 4th and 5th quantities from each domain. They then repeated the estimations as 3-person groups or individuals. The groups performed at the level of their best members and better than the independent individuals. All 5 domain boundaries improved estimation for both groups and individuals. Estimations were most accurate given the narrow (3rd and 6th) boundaries, suggesting processes of assimilation rather than averaging. Quantity estimation is an important aspect of both daily life and occupational activities. People frequently estimate quantities such as the population of a country, revenues of a company, length of a river, beef production of a region, mileage between cities, or area of a state. Moreover, many of these quantity estimations are made by formal and informal groups, such as securities analysts, weather forecasters, construction crews, travelers, or sports fans. Group versus individual quantity estimation is a basic and enduring topic in the social psychology of small group performance (for reviews see, e.g.,

Prior research by Kaplan and Miller (1978) suggested that juries are generally influenced less by extralegal, biasing information than individual jurors are. A social decision scheme (SDS) analysis of this question by Kerr, MacCoun, and Kramer (1997) suggested (a) that Kaplan and Miller’s conclusion should hold only for relatively extreme legal cases (i.e., cases where the probability of conviction, without biasing information, was either very high or very low) and (b) that the opposite pattern should hold for moderate cases (with moderate conviction rates)—i.e., juries should show even greater sensitivity to biasing information than should individual jurors. An experiment is reported that compared juror vs jury sensitivity to biasing information (viz., prejudicial pretrial publicity) for versions of a legal case with a moderate and an extreme conviction rate. Consistent with the SDS analysis, juries were more biased than jurors for the moderate-case version, but the reverse was true for the extreme-case version. The implications of these findings and the more general utility of the SDS model for studying group processes are discussed.