Knowledge that Transforms

[This study describes the behavior of audit fees during a period of apparent increasing competition in the market for independent audit services. Academic researchers (e.g., Danos and Eichenseher 1986; Kinney 1988) and the business press (e.g., Journal of Accountancy 1984; The Wall Street Journal 1985, 1987; Work 1985) have noted increasing competition in the market for audit services among public accounting firms, but there has been no documentation that audit fees have decreased. The purpose of this study is to determine whether real audit fees decreased between 1977 and 1981. This interval started with federal investigations of anticompetitive behavior in the accounting profession, included several changes in the accounting profession that, coupled with the economic downturn of the late 1970s and early 1980s, could have increased competition in the market for audit services. It ended after the Federal Trade Commission announced it was closing its investigation of the profession's anticompetitive rules because many of the restrictions against competition had been dropped (The Wall Street Journal 1980). The study sample was developed from a set of publicly traded companies reporting external audit fees in a University of Michigan database. Of the 98 companies reporting these data for 1977 and 1981, the two years of interest, 20 were excluded to minimize the potential confounding effects of major changes in internal auditing, the impact of auditor turnover, and changes in regulation of the banking industry on external audit fees. The study finds a significant decrease in real audit fees between 1977 and 1981. These findings were not sensitive to alternative specifications of the audit fee model and were not driven by any particular industry or audit firm. The results of this study are consistent with claims of increasing fee competition in the market for independent audit services. In view of the number of changes occurring in the audit profession and the market during that period, it is difficult to make causal inferences about the effects of particular changes in the profession on audit fees. Thus, this article should be viewed as a descriptive study of the behavior of audit fees in a time when the market for audit services was allegedly becoming more competitive.]

[Since the late 1970s, oil and gas firms have been required to include (unaudited) estimates of proved reserve quantities and valuations based on these estimates in their financial statements. As defined by the SEC, proved reserves are those that will be produced with "reasonable certainty" under existing economic conditions. In addition, firms must disclose estimated quantities of proved developed reserves, which will be produced from existing wells. At the time the proved reserve determination is made, substantial uncertainty exists as to the amount of oil ultimately recoverable, suggesting that the reserve estimation process is unreliable (Magliolo 1986b). Although certain constituencies have indicated interest in having firms disclose reserve quantities (see, e.g., Deakin and Deitrick 1982), the relevance of the disclosures for any given firm is an open question. It is to this issue that our research is directed. Prior research has demonstrated a weak association between security prices and oil and gas valuation disclosures required by SFAS No. 69 (see, e.g., Harris and Ohlson 1987; Magliolo 1986a). At least two explanations could account for the weak association: (1) reserve quantity estimates underlying the valuation disclosures are unreliable, and (2) the valuation model used to attach value to reserve quantities is flawed. Although the valuation approach used in these "reserve recognition accounting" disclosures has been widely disparaged, much criticism also concerns problems in the estimation of oil and gas reserve quantities. This article narrows the focus of prior research and specifically examines the value-relevance of reserve quantity disclosures required by SFAS No. 69 (FASB 1982). This more narrow focus permits us to investigate, for the first time, one of the explanations for why the reserve valuation components do not appear to be particularly informative. It also allows us to compare the informativeness of management claims about reserve quantities in the SFAS No. 69 disclosures with quantity estimates that we infer from management actions (i.e., production of oil). The article explores two empirical questions concerning reserve quantity disclosures. First, we examine the required SFAS No. 69 disclosures for proved reserves and proved developed reserves and ask whether these reserve estimates are value-relevant, given a benchmark estimate of reserves based on firms' current oil production levels. The second question is whether the association between market valuation and firms' reserve disclosures differs across firms according to characteristics of the disclosed data. Specifically, we determine whether investors' reliance upon SFAS No. 69 reserve quantity disclosures is related to variation in the reliability of such disclosures across firms, as indicated by the absolute size and direction of reserve estimate revisions, as well as the ratio of proved developed to total proved reserves. Each of these factors potentially provides information on the reliability or usefulness of reserve disclosures. In our first tests, which do not allow for firm-specific variation in the informativeness of SFAS No. 69 quantity disclosures, we find no evidence that disclosures based on proved reserves and proved developed reserves provide additional value-relevant information to market participants once production is known. However, when we partition our sample, we find that the proved reserve information is informative for a subset of firms whose reserve quantity estimates appear more reliable. The importance of our results extends beyond the oil and gas accounting area. The differential reliability results represent some of the first capital-markets-based evidence that the informativeness of a disclosure varies in accordance with the accuracy of management projections. Stated differently, our evidence suggests that investor reliance on disclosures varies as a function of their "quality." Also, our initial results that disclosures of proved reserve quantities provide no information once production is known suggest a reliance by investors on the "objective" information provided by management actions (production decisions) rather than the (potentially) subjective estimate of proved reserves.]

[Inventory accounting rules may have a purchase (production) repercussion on the firm that affects stockholders' wealth, because managers, paid by an accounting income-based contract, can potentially distort the operation while maximizing their payoffs. Stockholders, by directly or indirectly dictating the accounting policy choice, can use accounting rules to implement their preferences and control the managers. Analysis of such a scenario indicates that the stockholders' preferences towards accounting rules are driven primarily by three factors, the tax rate, the acquisition price, and the rate of real holding gain. With no taxes, and absent inflation, FIFO (first in, first out) implements the first-best decision rule. When taxes are present, the preferences of owners and managers diverge, which may lead to operating distortions. An optimal accounting policy choice decision must consider this distortion (the incentive effect) as well as the tax gain possible with LIFO (last in, first out). Further analysis shows that FIFO may still be preferred over LIFO under certain conditions because an incentive effect may exceed an "obvious" tax advantage with LIFO. Abdel-khalik (1985) has provided empirical evidence suggesting that there are incentives for corporate executives to stay on FIFO despite the apparent LIFO tax advantages possibly because their incentives are income-based. Larcker (1983) provides independent evidence that often management compensation contracts are indeed based on accounting income. Thus, when incentive considerations are present, the cash flow effects as shown by Sunder (1976) may not be obvious, and our analysis questions the conventional wisdom that firms tend to switch to LIFO at the time of rising prices. Although the effect of LIFO versus FIFO accounting rule on firm cash flow is well-known, many firms continue to follow FIFO in spite of tax advantages with LIFO. Granof and Short (1984) reported that more than 30 percent of firms in their non-LIFO sample rejected LIFO because of "excessive cost" or "other adverse consequences." We argue that investors choose an inventory accounting method to optimize the trade-off between incentive effect and tax gain, and align the managers' interests as much as possible with those of the stockholders.]

[This study provides evidence that market segmentation affects the strength of the association between financial disclosure and net interest cost for new issues of municipal bonds. Previous studies (see, e.g., Amershi and Ramamurtie 1990; Cook 1982; Hendershott and Kidwell 1978; Kidwell et al. 1983) show that the geographic segmentation of primary markets for municipal bonds along regional and national lines is characterized by different sources and costs of information. Because little information is available from alternative sources for the less marketable bonds of smaller issuers which use regional markets, financial reporting variables are hypothesized to be associated more strongly with interest costs for these issues than for those of larger municipalities, whose bonds are typically issued in the national bond market. Cook (1982) argues that the costs of obtaining information about small issuers from alternative sources are relatively high. For these issuers, who also do not issue bonds frequently, regional underwriters may be the only informational intermediaries (other than rating agencies if the bonds are rated) between the issuers and investors (Kidwell et al. 1983). Thus, financial and other entity-specific information contained in the offering statement are expected to be weighted more heavily in pricing these bonds than for bonds of larger issuers. Although this differential-information hypothesis is analogous to Atiase's (1985) argument regarding differential predisclosure of information, our study assumes differential availability of information across market segments, and size alone is not likely to be an adequate proxy for segmentation. Our main analysis uses the nature of the underwriting syndicate, whether it is managed by a national or regional underwriter, to proxy for the relevant market segment. Previous studies of municipal accounting (see, e.g., Ingram and Copeland 1982; Wallace 1981; Wilson and Howard 1984) have provided evidence of an association between bond measures and accounting and auditing variables. However, the results were mixed and inconsistent, which may be attributable to differences in the entities examined (cities, counties, and school districts), the time periods examined, or definitions of variables. Also, differences in the mix of bonds issued in each segment may have contributed to the inconsistent results. By examining the relationship between accounting and auditing variables and bond interest costs separately for each segment of the primary bond market, our study provides useful evidence on the potential effect of different information environments on this relationship. Regression results based on a sample of 119 new municipal bond issues, partitioned on a measure of segmentation (regional vs. national underwriter), are consistent with the hypothesis that the association between the quality and quantity of financial disclosure and interest costs is stronger for municipalities that issue bonds in local or regional markets than for those that issue bonds in the national market. Our results suggest that future research in this area should consider the potential effects of market segmentation when testing the association between accounting or auditing variables and bond interest costs.]

[This article relates belief functions to the structure of audit risk and provides formulas for audit risk under certain simplifying assumptions. These formulas give plausibilities of error in the belief-function sense. We believe that belief-function plausibility represents auditors' intuitive understanding of audit risk better than ordinary probability. The plausibility of a statement, within belief-function theory, measures the extent to which we lack evidence against the statement. High plausibility for error indicates only a lack of assurance, not positive evidence that there is error. Before collecting, analyzing, and aggregating the evidence, an auditor may lack any assurance that a financial statement is correct, and in this case will attribute very high plausibility to material misstatement. This high plausibility does not necessarily indicate any evidence that the statement is materially misstated, and hence, it is inappropriate to interpret it as a probability of material misstatement. The SAS No. 47 formula for audit risk is based on a very simple structure for audit evidence. The formulas we derive in this article are based on a slightly more complex but still simplified structure, together with other simplifying assumptions. We assume a tree-type structure for the evidence, assume that all evidence is affirmative and that each variable in the tree is binary. All these assumptions can be relaxed. As they are relaxed, however, the formulas become more complex and less informative, and it then becomes more useful to think in terms of computer algorithms rather than in terms of formulas (Shafer et al. 1988). In general, the structure of audit evidence corresponds to a network of variables. We derive formulas only for the case in which each item of evidence bears either on all the audit objectives of an account or on all the accounts in the financial statement, as in figure 1, so that the network is a tree. Usually, however, there will be some evidence that bears on some but not all objectives for an account, on some but not all accounts, or on objectives at different levels; in this case, the network will not be a tree. We assume that all evidence is affirmative because this is the situation treated by the SAS No. 47 formula and because belief-function formulas become significantly more complex when affirmative and negative evidence is combined. This complexity is due primarily to the renormalization involved in Dempster's rule for combining belief functions. The variables in the network or tree represent various audit objectives, accounts, and the financial statement as a whole. We assume these variables are binary. For example, we assume that an account either is or is not materially misstated. This assumption is clearly too restrictive for most audit practice. Often, for example, an auditor must consider immaterial errors in individual accounts that could produce a material error in the financial statement when they are aggregated. We derive formulas for plausibility of material misstatement at three levels: the financial statement level, the account level, and the audit objective level. The formula at the audit objective level resembles the SAS No. 47 formula, but the formulas at the other two levels are significantly different. Because our model does distinguish evidence gathered at the three different levels, audits based on our formulas are sometimes significantly more efficient than audits based on the SAS No. 47 model or on the simpler Bayesian models.]

[The antitrust laws of the United States prohibit monopolies or attempts to create a monopoly in any unregulated line of business. In the past, the two agencies that enforce these laws, the Department of Justice and the Federal Trade Commission, have relied on accounting profits in prosecuting such antitrust violations. The agencies argued that high accounting rates of return were "excessive" and indicative of monopolistic power on the part of the firm. Thus, to reduce the possibility of an unfavorable ruling and the costs associated with it, managers in firms investigated for monopoly-related violations would have incentive to use accounting procedures (e.g., accounting methods, accruals) that produce abnormally low levels of income. Because the incentive to reduce income will increase as the threat of an unfavorable ruling becomes more imminent, it is expected that managers will take additional steps to lower income while being actively investigated, compared with periods of non-investigation. The differences in political costs caused by the antitrust investigation are used to provide a refined test of the political-cost hypothesis. In particular, the study examines, on a longitudinal basis, whether managers respond to these investigations by adjusting their discretionary accruals. The discretionary accruals for 48 firms that were investigated for monopoly-related violations were estimated over a 15-year period by using the residuals of a fixed effects covariance model that regressed total accruals on the change in sales, the fixed asset balance, and dummy variables representing each firm and year. The specific hypothesis tested is whether the estimated discretionary accruals over the period of investigation (typically more than one year) were lower, or more income-reducing, than those in other years. The hypothesis was tested by using a nested design with a dummy variable, coded 1 for the years of investigation, included in the accrual model. This variable was significant and negatively signed, which indicates that discretionary accruals were lower while the firm was being investigated, as expected. The discretionary accruals for a control group of firms that were not investigated were also estimated, and these were found to remain the same during the periods of investigation and non-investigation for the matched sample firms. The results from these tests support the political-cost hypothesis and are consistent with the view that managers adjust earnings in response to monopoly-related antitrust investigations.]

[During 1974-75, more than a fourth of the manufacturing and merchandising firms listed on the New York Stock Exchange and the American Stock Exchange adopted or extended their use of the last-in, first-out (LIFO) method of inventory accounting. Because a change to LIFO accounting can materially affect a firm's cash flows, the stock price behavior of firms adopting LIFO during fiscal year 1974 has been studied extensively. However, existing studies provide little evidence that stock price movements during the period were directly linked with LIFO adoption. In this article, we report the results of new tests for a stock price response to disclosures of 1974 LIFO decisions by both adopting and non-adopting firms. Our tests assume that, prior to disclosure, prices reflect available information about the likelihood of LIFO adoption. This implies (for both groups of firms) that the stock price change in response to disclosure of the LIFO decision depends on investors' revisions in their beliefs regarding the probability of adoption caused by the disclosure. We use this implication to design tests of the null hypothesis of no association between stock price movements and the disclosure of 1974 LIFO decisions. Our tests are based on a sample of 487 firms that could have adopted LIFO during fiscal 1974. For each firm, we first identify an interval that includes the disclosure of its LIFO decision. We then estimate the probability of LIFO adoption at the beginning of that interval on the basis of information that was available at the time. Finally, we regress disclosure-interval stock returns on revisions in the probability of adoption resulting from disclosure of the decision. Our results provide strong evidence of a price response for nonadopting firms and weaker evidence of a price response for adopting firms.]

[Accounting information plays an important role in a decentralized economy. It is a primary way for managers to make assertions about the past performance, current condition, and future prospects of their firms. The auditing of these accounting disclosures is purported to provide value to the economy for several reasons (Baiman 1979; Baiman et al. 1987; Blazenko and Scott 1986; Evans 1980; Scott 1984). Without accurate firm-specific information, investors may align their portfolios in a less-than-optimal fashion, resulting in an inefficient allocation of resources in the society. The public-good feature of auditing may prevent the less efficient private search for firm-specific information. Additionally, auditing is argued to assuage the divergent preferences of managers and investors. Included in these divergent preferences is managerial motivation to manipulate the accounting information to hide perquisite consumption or imply appropriate production, investment, or financing decisions. Investors anticipate these conflicting goals and price protect themselves, forcing managers to bear the residual loss (Jensen and Meckling 1976). Because of managers' motivation to avoid that loss, auditing can be purchased to provide credibility to managerial disclosures necessary to distinguish among firms of differing quality. With auditing in place, firm managers can reap the benefits of those actions expected to improve the firm's prospects. Research into managerial incentives to disclose and the effects of disclosures on other managerial actions is at an early stage. The analytical assertions of full disclosure when fraudulent disclosures are not possible are well known (Dye 1985; Grossman 1981; Grossman and Hart 1980; Milgrom 1981; Milgrom and Roberts 1986), and the body of experimental research is growing (Forsythe et al. 1989; King and Wallin 1990a, 1991). Only recently has either analytical or experimental research been conducted in environments permitting fraudulent disclosure (Dopuch and King 1991, 1992; Dopuch et al. 1989; Kachelmeier 1991; King and Wallin 1990a, 1990b; Wallin 1990). In particular, progress on the formulation of models of the demand for auditing has only recently been made. These models are argued to be at best embryonic (Baiman 1979; Scott 1984) and have typically ignored the possible effects of other mechanisms that may solve the problems that auditing addresses. An important aspect of this article is an investigation of the effect of legal recourse, which allows investors to sue when disclosures are believed to have been fraudulent. The analysis presented here concludes that the threat of lawsuit will cause less frequent fraudulent reports. Importantly, legal recourse also allows the manager to benefit from costly effort, such that the level of effort expended will be that which maximizes societal benefit. The central purpose of this study is to investigate the demand for auditing in environments both with and without legal recourse. Formal models are developed and tested that investigate the extent to which managers and investors can achieve a cooperative solution when no mechanism exists to aid that cooperation. The analysis and tests are performed in an environment in which the time horizon is not known. This extends the work of Dopuch et al. (1989) by constructing a base-line environment where the demand for auditing is not derived from the backward induction of a single-period equilibrium. Thirty-two experimental markets were conducted. The results show a demand for auditing, regardless of whether legal recourse was present. The availability of either auditing or legal recourse induced a higher level of managerial effort, the highest occurring when both options were available. Both auditing and legal recourse reduced a tendency for investor overbidding, but only legal recourse reduced the proportion of fraudulent disclosures.]

[Relative performance evaluation (RPE) is the process of comparing performances across workers. Relative performance information (RPI) allows a superior to better infer a particular worker's unobservable effort level than would otherwise be possible, and analytical studies have demonstrated that RPE may be an optimal strategy for mitigating the effects of moral hazard if the workers face some common uncertainty (Baiman and Demski 1980; Holmstrom 1980, 1982; Wolfson 1985). Although these analytical studies provide important insights into the role of RPE, they ignore the intrinsic value of comparing workers' performances as asserted (Locke 1968) or demonstrated (Beck and Seta 1980; Harkins and Jackson 1985; Klinger 1969) in behavioral studies. Holmstrom (1982, 325) specifically states that "inducing competition among [workers] by tying their rewards to each other's performance has no intrinsic value." That is, formal models do not assign any intrinsic value to RPE because they emphasize the economic rather than behavioral factors that influence workers' responses to RPE. Consideration of the direct effects of both behavioral and economic factors on effort can provide a better understanding of the potential benefits from RPE. In this study, variables suggested by formal models-that of Holmstrom (1982) in particular-are used to examine the importance of economic and behavioral factors in explaining the motivational effects of comparing workers' performances. Specifically, the direct effects on effort of (1) the degree of common uncertainty among the workers and (2) compensation as a function of RPI are examined in a setting where workers know that they and their supervisor will receive RPI. Several assumptions of agency theory are used to develop hypotheses about the importance of economic factors, and social influence research is used to develop hypotheses about the importance of behavioral factors. The hypotheses were tested in a laboratory experiment that required subjects to act as managers and make production decisions for a hypothetical company. Half the subjects worked under a profit-sharing contract that based compensation solely on their absolute performance, and the remaining subjects worked under an RPE contract that based compensation on both their absolute performance and performance relative to the RPI. There were three levels of common uncertainty, and it was manipulated by varying the number of sources of uncertainty that subjects had in common. Subjects' risk and effort preferences were induced experimentally. The experimental results support the importance of both economic and behavioral factors, depending on the type of contract examined. Subjects' effort levels increased significantly as the degree of common uncertainty increased with the RPE contract, but not with the non-RPE contract. In addition, effort levels were higher under the RPE contract than under the non-RPE contract. These results imply that behavioral factors can be important determinants in motivating effort and that future attempts to model behavior analytically may need to consider these factors. The results also provide weak evidence that economic factors, such as contract type, may enhance or mitigate the importance of behavioral factors in motivating effort.]