Abstract Reviews the book `1991 Tax Penalties: The Complete Guide to Penalties Under the Internal Revenue Code,' by Pauline White and Consuelo M. Ohanesian.

Abstract Examines participative budgeting in the context of the psychology of risk. Theoretical background on risk preferences; Empirical evidence of domain-specific risk preferences; Study design; Dispositional risk attitude measurement; Preference ratings of risk-averse versus risk-taking groups; Limitations of the study.

[This study examines participative budgeting in the context of the psychology of risk. As Young (1985) and Waller (1988) report, there is some preliminary evidence that risk-averse workers create more budgetary slack than risk-neutral ones. They show that "truth inducing incentive schemes" (e.g., Soviet incentive schemes; see Weitzman 1976) reduce budgetary slack for risk-neutral subjects but not for risk-averse subjects. If this is true, it means that resource allocations within organizations are mediated by perceptions of risk. Young (1985) and Waller (1988) define risk preference as a dispositional variable, which presumes that it is a stable personal trait, a latent variable, traditionally inferred from observed behavior of risk propensity in such settings like lotteries. This study tests whether risk preferences are domain-specific; that is, latent risk preferences translate into differing manifest risk preferences according to the context. Domain-specific risk preference can be understood as a manifest psychological variable that may well be the result of the combination of latent risk propensity and the situation. Kahneman and Tversky's prospect theory (1979) suggests that manifest risk preferences depend upon whether the subject frames his or her task in the context of gain or loss prospects, where gains and losses are defined in relation to a neutral reference point. If risk preferences are domain-specific, then past studies' suggestion that incentive schemes should be designed in consideration of dispositional, or latent, risk preferences needs to be reexamined (Waller 1988; Kaplan 1982). The question before this study is: If subordinates are influenced by prior period performance in setting current period budgets for themselves, will that influence take the form predicted by prospect theory and thus lead to riskier preferences (tight budgets) when the subordinates perceive themselves in a losing situation? This is an important question considering Young's (1985) alluding to the possibility that inducing subordinates to less risk-averse behavior may be a way to favor tight budgetary standards and reduce slack. This prediction is consistent with prospect theory's implication that losers who are slow to adjust their reference point act in a more risk-seeking manner. Thus, the induction of losing prospects might be a way to minimize budgetary slack. An additional concern in this study is to jointly test domain-specific risk preferences and dispositions toward risk as influences over budgetary decisions. Whereas prospect theory explains risk preferences as domain-specific contingencies, other theories construe risk preferences as dispositional. It is likely that both domain-specific and dispositional factors influence budgetary decisions. This study deploys a conventional lottery procedure to elicit and test dispositions toward risk. An experiment simulating the public accountants' budgeting of billable hours was designed to test the hypothesis that subject preference for tight or safe budget behavior depends on the performance of coworkers and domain-specific risk preferences. The hypotheses were tested in an experiment employing 81 students. The results generally support the view that subordinates' risk preferences are influenced by a situation-dependent variable. The reversal of risk preferences around a neutral reference point is statistically significant for both dispositionally risk-averse and dispositionally risk-seeking subjects. The dispositional variable also contributes to the explanation of variations in subjects' manifest risk preferences. Thus the propensity to induce budgetary slack seems to be a joint function of situations and dispositions.]

[In the last two decades of accounting research, many studies have investigated the relation between accounting variables and risk-adjusted security returns. The earliest studies (e.g., Ball and Brown 1968) used simple, nonparametric methods and focused mainly on the question of whether accounting earnings are associated with residual equity returns. Subsequent studies made methodological refinements in both the measurement and the statistical techniques. One important statistical refinement was the "unexpected earnings response regression model" (UERRM), a linear statistical model that uses the unexpected earnings variable as a regressor to explain risk-adjusted returns. The UERRM has become well-known, and many recent studies (e.g., Cornell and Landsman 1989, 686; Daley et al. 1988, 580; Doran et al. 1988, 392; Landsman and Damodaran 1989, 107; McNichols 1989, 15) have used some form of it as a "benchmark" model, against which to compare more complicated models. In one paradigm (so popular that it has become almost standard practice), various accounting variables are added to the UERRM, and their "incremental information content" is assessed by testing the statistical significance of their coefficients. The inferences derived from this procedure are, of course, conditional on the degree to which the UERRM is correctly specified. A critical problem caused by using a misspecified UERRM is that its least squares estimator can lead to erroneous inferences in the research design. Despite the UERRM's popularity, researchers have expressed concerns regarding its specification. For example, Lev (1989) recently surveyed a large number of research papers that used some form of an UERRM and found that for the most part R2 s were low and often bordered on "the negligible." His table 1, which includes statistics from 19 studies, indicates that most reported R2 s are less than 10 percent. Although low R2 s are not proof of major specification problems, Lev does suggest that they are a cause for concern and may be the result of specification problems. Investigation of multiple specification problems is an important aspect of the present study because the various specification issues are interrelated. For example, nonnormality can be associated with nonlinearity, which, in turn, can be associated with heteroscedasticity or variation in the coefficients (Judge et al. 1985, 455, 814, 839). Thus, ad hoc tests for a single specification problem can be misleading and can fail to identify the fundamental problems. To our knowledge, no studies have comprehensively and formally evaluated the specification of the cross-sectional, ordinary least squares model that relates unexpected earnings to risk-adjusted security returns. Therefore, the purpose of this study is to test such a model systematically and empirically for specification problems. Specifically, this study tests for nonlinearity, heteroscedasticity, residual nonnormality, omitted variables, and interfirm systematic and random coefficient variation. Also, when appropriate, adjusted R2 -statistics are included to indicate the degree of misspecification-information that is not directly observable from the tests themselves. A high degree of generality is obtained by using three samples of earnings forecasts as proxies for expected earnings. These were obtained from IBES financial analyst consensus forecasts, Value Line financial analyst forecasts, and COMPUSTAT-based time-series forecasts. Daily and monthly security returns are considered for short and long event-windows, respectively. In addition, one study recently published in The Accounting Review (Cornell and Landsman 1989) is replicated. The findings from all of the samples and the replication indicate that the specification error is large enough to affect conclusions regarding economic relationships. For example, in the replication, the specification problems are shown to lead to substantial instability in inferences from the model.]

Abstract In the last two decades of accounting research, many studies have investigated the relation between accounting variables and risk-adjusted security returns. The earliest studies (e.g., Ball and Brown 1968) used simple, nonparametric methods and focused mainly on the question of whether accounting earnings are associated with residual equity returns. Subsequent studies made methodological refinements in both the measurement and the statistical techniques. One important statistical refinement was the "unexpected earnings response regression model" (UERRM), a linear statistical model that uses the unexpected earnings variable as a regressor to explain risk-adjusted returns.' The UERRM has become well-known, and many recent studies (e.g., Cornell and Landsman 1989, 686; Daley et al. 1988, 580; Doran et al. 1988, 392; Landsman and Damodaran 1989, 107; McNichols 1989, 15) have used some form of it as a "benchmark" model, against which to compare more complicated models. In one paradigm (so popular that it has become almost standard practice), various accounting variables are added to the UERRM, and their "incremental information content" is assessed by testing the statistical significance of their coefficients. The inferences derived from this procedure are, of course, conditional on the degree to which the UERRM is correctly specified. A critical problem caused by using a misspecified UERRM is that its least squares estimator can lead to erroneous inferences in the research design. Despite the UERRM's popularity, researchers have expressed concerns regarding its specification. For example, Lev (1989) recently surveyed a large number of research papers that used some form of an UERRM and found that for the most part R2s were low and often bordered on "the negligible." His table 1, which includes statistics from 19 studies, indicates that most reported R²s are less than 10 percent. Although low R²s are not proof of major specification problems, Lev does suggest that they are a cause for concern and may be the result of specification problems. Investigation of multiple specification problems is an important aspect of the present study because the various specification issues are interrelated. For example, nonnormality can be associated with nonlinearity, which, in turn, can be associated with heteroscedasticity or variation in the coefficients (Judge et al. 1985, 455, 814, 839). Thus, ad hoc tests for a single specification problem can be misleading and can fail to identify the fundamental problems. To our knowledge, no studies have comprehensively and formally evaluated the specification of the cross-sectional, ordinary least squares model that relates unexpected earnings to risk-adjusted security returns. Therefore, the purpose of this study is to test such a model systematically and empirically for specification problems Specifically, this study tests for nonlinearity, heteroscedasticity, residual nonnormality, omitted variables, and interfirm systematic and random coefficient variation. Also, when appropriate, adjusted R²-statistics are included to indicate the degree of misspecification information that is not directly observable from the tests themselves. A high degree of generality is obtained by using three samples of earnings forecasts as proxies for expected earnings. These were obtained from IBES financial analyst consensus forecasts, Value Line financial analyst forecasts, and COMPUSTAT-based time-series forecasts. Daily and monthly security returns are considered for short and long event-windows, respectively. In addition, one study recently published in The Accounting Review (Cornell and Landsman 1989) is replicated. The findings from all of the samples and the replication indicate that the specification error is large enough to affect conclusions regarding economic relationships. For example, in the replication, the specification problems are shown to lead to substantial instability in inferences from the model.

Abstract Examines the influence of annual earnings announcements on financial forecasts. Convergence of beliefs; Earnings surprise; Suggestion that annual earnings announcement should increase the convergence of analysts' forecasts of future earnings.

[One indication of information usefulness is its ability to increase the precision of individuals' estimates of events of interest (FASB 1980; ljiri and Jaedicke 1966). In this context, earnings reports are useful if they increase the precision of investors' forecasts of future earnings when the latter proxy for the event of interest, future cash flows. The cross-sectional variance of analysts' earnings expectations often is used as a proxy for the unobservable precision of their earnings estimates (Ajinkya and Gift 1985; Brown et al. 1987; Imhoff and Lobo 1992). We show that, when combined with the time-series properties of accounting earnings and prior research in analyst forecasts, Bayesian revisions suggest that year t earnings reports should, on average, increase the convergence of analysts' year t + 1 earnings forecasts. Operationally, we examine whether the information contained in year t earnings decreases the cross-sectional variance of analysts' year t + 1 forecasts. Morse et al. (1991) use I/B/E/S Summary data, and conclude that the information contained in year t earnings announcements increases the cross-sectional variance of analyst forecasts of year t + 1. This is a surprising result. One feature of the I/B/E/S Summary data is that they do not contain dates of the analysts' earnings forecasts. Thus, researchers who use these data do not know the set of information upon which the analyst's earnings forecast is based. In contrast to the I/B/E/S Summary data, the I/B/E/S Detail data are precise regarding the date that the individual analyst's earnings forecast entered the I/B/E/S system. We use I/B/E/S Detail data to reexamine the relation between annual earnings announcements and convergence of beliefs. Using the Detail data, we show that the information contained in year t earnings decreases the cross-sectional variance of analysts' earnings forecasts of year t + 1. Moreover, our finding is insensitive to year of study. Using the Summary data, we find that the information contained in year t earnings increases the cross-sectional variance of analysts' earnings forecasts of year t + 1, but this finding is sensitive to year of study. Morse et al. (1991) hypothesize and provide evidence that reduction in variance is less likely to occur when "standardized" surprise is large. Using the Detail data, we show that significant decreases in variance occur for the seven smallest deciles of standardized surprise, and that significant increases in variance occur only for the largest decile. Using the Summary data for the same "window" as the Detail data, we find no deciles of standardized surprise associated with significant decreases in variance, and we observe significant increases in variance for the three largest deciles of standardized surprise. In sum, our results using I/B/E/S Detail data suggest that, on average: (1) annual earnings announcements increase convergence of analysts' forecasts of firms' future earnings; (2) annual earnings announcements decrease convergence only for the largest decile of earnings surprise. In contrast, we do not obtain consistent results with I/B/E/S Summary data.]