Wilfred C. Uecker

A Laboratory Investigation Of The Moral Hazard Problem In An Agency Relationship

Recent developments in agency theory have focused on the moral hazard problem in single-period and multiperiod models. In particular, the literature has emphasized how alternative institutions (e.g., liability rules and long-term contracts) can mitigate the adverse effects of the moral hazard problem. In this study, we use laboratory markets to examine the ability of two such institutions to remedy the adverse effects of moral hazard. We begin by presenting one set of markets with no liability rule from DeJong, Forsythe, and Lundholm [1985], hereafter DFL, where the effects of moral hazard were examined in a market for a service with a hidden characteristic; that is, the principal could not distinguish between shirking and chance as causes of an unfavorable outcome.1 The study showed that moral hazard led to shirking, and while

A laboratory investigation of alternative transfer pricing mechanisms

Abstract In this study, we conduct a laboratory experiment to evaluate three alternative mechanisms for pricing transfers between a buyer and a seller. The three mechanisms are: (1) a direct negotiation mechanism with which either party suggests terms of trade that the other can accept or reject; (2) a traditional mechanism where the transfer price and quantity are determined by the intersection of the reported demand and supply schedules of the buyer and seller, respectively; and (3) the Ronen-McKinney mechanism which computes the transfer quantity in the same way as the traditional mechanism, but is designed to give different transfer prices to the buyer and seller in order to induce them to truthfully report their respective schedules. For all periods of the experiment, the direct negotiation mechanism is the least efficient of the three due to the high frequency of bargaining impasses. The efficiency of the traditional mechanism and the efficiency of the Ronen-McKinney mechanism are indistinguishable from one another. However, in the last four periods of the experiment where the data exhibit less volatility, the efficiencies of all three mechanisms are indistinguishable from each other. When truthful reporting is considered, the Ronen-McKinney mechanism had the least amount of misreporting followed by the traditional mechanism and then the direct negotiation mechanism.

Judgmental evaluation of sample results: A study of the type and severity of errors made by practicing CPAs

Abstract This paper examines the extent to which practicing CPAs may employ two heuristics, representativeness and protectiveness, which lead to systematic errors in the judgement evaluation of sample outcomes. The results indicate that both heuristics are used, but to a lesser extent than suggested by previous research. Implications of the results for audit firm policy are discussed.

A Behavioral Study of Information System Choice

Most problems in accounting reduce to one of choosing among two or more information systems which could be provided to a decision maker (user) (American Accounting Association [1972b, p. 317]). In recent years, researchers have developed a normative theory for the evaluation of information system alternatives (Feltham and Demski [1970]; Demski [1972]; Feltham [1973]; Demski and Feltham [1976]). The normative theory has provided important insights into the nature of the accountants choice problem; however, little is known about the accountants ability to apply the theory correctly. The purpose of this study is to investigate the ability of the accountant to apply the normative theory of information evaluation correctly in choosing among information system alternatives. An important feature of the normative theory of information evaluation is its explicit recognition that the accountant must often make information decisions concerning what information system to provide another decision maker who will make an action decision. With this separation of the accountant and the decision maker, the decisionmaking behavior of the user should become an important consideration in the accountants application of the normative theory of information evaluation. Therefore, this study focuses upon the accountants ability to apply correctly the normative theory of information evaluation with different models of user decision-making behavior. Specifically, this research investigates the ability of the accountant (1) to learn the optimal information system for a decision maker and (2)

The Quality Of Group-Performance In Simplified Information Evaluation

In recent years, accounting researchers have investigated the extent to which individuals can make information choice decisions consistent with normative models proposed for the evaluation of information system alternatives. For example, Uecker [1978] observed that while subjects did, on average, adapt their information system choices for different users in the direction prescribed by normative models, their choices were not optimal and showed no improvement over repeated trials with outcome feedback. Employing a modification of Ueckers experimental task, Hilton, Swieringa, and Hoskin [1981] found that only a third of their subjects correctly perceived that information value was monotonically increasing in information accuracy. In a second experiment, Hilton and Swieringa [1981] found that only a fifth of their subjects correctly perceived that information value was monotonically increasing in the degree of initial uncertainty in the particular decision setting employed. These results suggest the desirability of developing ways to improve the performance of information evaluators in applying normative models of information evaluation. As one attempt, Uecker [1980] investigated the effectiveness of providing information evaluators with a detailed description of a users decision model. Presumably, knowledge of the users decision model would improve performance since evaluators would know with certainty the users action choice in response to a signal from an information system. Unfortunately, the performance of subjects with

The Methodology of Laboratory Markets and Its Implications for Agency Research in Accounting and Auditing

Accounting and auditing reflect many relationships that can be characterized as principal-agent relationships. Concern over these arise within several contexts. For example, the Securities and Exchange Commission and Congress have maintained an ongoing concern about the price and quality of audit services and the effect of these services on investor welfare (see, for example, Williams [1979], Committee on Government Affairs [1977], and Staff Study of Subcommittee on Reports, Accounting and Management [1977]). Similarly, a large and growing literature has developed which considers the conflict between owners and managers (see, for example, Jensen and Meckling [1976], Fama [1980],

The value of information in decision making

Abstract This paper reports upon an experiment conducted to test a normative model of information evaluation as a descriptive representation of human behavior in judging the value of information. The test was conducted under a more general set of assumptions than those used in previous research. Subjects judged the maximum amount that they would be willing to pay for a single marble drawn from one of two urns containing many black and white marbles. Five levels of priors and accuracy of the information source, and three types of decision maker models were experimentally manipulated. The results confirm previous research which has suggested that normative models are not appropriate descriptive representations of information evaluation behavior. Among alternative models tested, observed behavior was best represented by one containing a multiplicative rule to combine the subjective levels of accuracy, priors, and type of decision maker, and an averaging rule to combine the manipulated factors.