Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Causes and consequences of error and bias in management accounting systems

Causes and consequences of error and bias in management accounting systems China Journal of Accounting Studies, 2014 Vol. 2, No. 1, 1–12, http://dx.doi.org/10.1080/21697221.2014.893774 COMMENTARY Causes and consequences of error and bias in management accounting systems Ranjani Krishnan* Broad College of Business, Michigan State University, 632 Bogue Street, N207; East Lansing, MI 48824-1122, USA The purpose of this paper is to provide an overview of management accounting sys- tems, particularly with respect to their influence on error and bias in decision mak- ing. It also discusses the implications of such errors and biases for organizations and for policy and suggests avenues for future research. Keywords: management accounting system design; error; bias 1. Introduction The goal of management accounting is to provide information for decision making and control to managers within the firm. The literature is awash with suggested frameworks to analyze the different roles of management accounting. One popular framework is Demski and Feltham (1976), which distinguishes between the decision facilitating role and the decision influencing role of management accounting. The decision facilitating role of management accounting information is its use in problem solving and belief revision. In this role, the management accounting system assists managers to evaluate whether or not performance is proceeding as planned, identify the reasons for depar- tures from plan, and determine appropriate corrective actions. Budgeting and variance analysis is an example of the decision facilitating role of management accounting infor- mation. The decision influencing role is the use of accounting information to motivate employees and achieve goal congruence. Contracting, incentive design, and perfor- mance measurement systems are examples of the decision influencing role. A similar framework is used by Zimmerman (2009) in his book where he identifies the role of management accounting as serving the purposes of decision making (similar to the decision facilitating role) and control (similar to the decision influencing role). The decision facilitating/influencing framework uses a largely economic lens to view the role of management accounting, although there is scope to incorporate judg- ment and decision-making in both these roles. An alternative framework is used by Macintosh and Quattrone (2010), which characterizes management accounting as a social system with a goal not only to motivate and manage employees, but also to sanction the actions of managers. They view management accounting as a part of a wide array of other systems such as information and communication systems. As per this view, management accounting system design cannot be considered in isolation from other systems; rather, their features only have relevance within the network of relationships and other systems in place in the organization. *Email: Krishnan@bus.msu.edu Paper accepted by Jason Xiao. © 2014 Accounting Society of China 2 Krishnan Regardless of the perspective, in order to provide useful information for decision making, the management accounting and control system (MCS) should be designed to reduce error and bias in the numbers and reports that are produced by the organization. Such design should at the same time take into consideration the cost versus the benefit of reducing the errors and biases. One important driver of error is poorly designed MCS, which causes the reported performance to differ from the actual economic perfor- mance of interest. One important driver of bias is the agency problem, i.e. decisions within firms are made by managers whose goals are not aligned with those of the shareholders (Fama & Jensen, 1983; Jensen & Meckling, 1976). Goal incongruence is a problem in decentralized firms because of the inability of shareholders to observe managerial effort, which could encourage managerial shirking. Furthermore, relative to shareholders, managers often have better information about operations. This can cause managers to misrepresent their private information and maximize their own interest rather than the interest of the firms (Baiman, 1990). Incentive compensation is often used as a tool to address this goal incongruence problem between managers and share- holders. However, poorly designed incentive compensation systems can encourage managers to build bias into the numbers. For example, excessive incentive compensa- tion combined with poor quality and incongruent performance measures can encourage budgetary slack, earnings management, and fraud. In the next section, I provide examples of several types of errors that can arise from poor quality MCS and the implications of these errors. This is followed by a discussion of the potential for bias in reported accounting numbers and the consequences thereof. The final section concludes. 2. Errors in accounting reports Errors in accounting reports arise from several sources, the most prominent of which are incorrect allocation of overhead cost and incorrect identification of capacity costs. Traditional management accounting systems with crude, volume based allocations cause departures of reported costs from actual costs. Absorption-oriented accounting systems force the costs of excess capacity to be allocated to current production. Similarly, accounting systems that focus on the appropriate capture of aggregate costs rather than appropriate allocation of costs based on cause and effect, result in a number of impor- tant costs to be invisible to the decision makers. These hidden costs can be substantial in many industries and can cause flawed decisions about current and future resource allocations. Another contributor of accounting error is flawed judgment and psychological biases. Most decisions in firms are made by non-expert managers who do not have the cognitive resources to accurately apply complex economic models or analytical tech- niques. Similarly, accounting systems are designed by non-accountants such as plant managers, branch managers, and other front-line employees. These individuals often face challenges with respect to both analysis and judgment skills. While sometimes these gaps in knowledge result in unsystematic errors in decision making, which thereby increases the noise in the system, at other times these knowledge gaps result in systematic departures from optimal decisions leading to biased results. This section provides examples of errors in management accounting systems that result from three sources: (1) use of traditional accounting systems that focus on appro- priate recording of aggregate costs and revenues, resulting in hidden costs; (2) inappro- priate allocation of excess capacity costs; and (3) use of subjective decision tools that China Journal of Accounting Studies 3 can be qualitative, incomplete, and substitute informal and familiar attributes for the formal attributes of scientific models. 2.1. Traditional accounting systems and hidden costs Management accounting research has long advocated the use of refined cost allocation systems that systematically track, pool, and allocate overhead costs to cost objects in an accurate manner (Cooper & Kaplan, 1992). Although overhead cost allocation does require subjectivity, judgment, and assumptions, as long as there is an attempt to match overhead resource consumption with the appropriate cost drivers, errors can be mini- mized. Almost all undergraduate and graduate accounting programs, as well as MBA programs teach activity-based costing (ABC) and other refined cost allocation systems. Nevertheless, most firms continue to use traditional cost allocation systems with volume-based cost drivers. Adherence to outdated accounting systems often stems from an absorption costing mentality, where the goal is to report aggregate costs accurately. When accounting systems focus on the aggregate, they fail to isolate the impact of important exogenous factors that affect their costs such as competition and regulation. Although managers acknowledge that these costs can be substantial, most firms do not have adequate systems for identifying and measuring hidden costs. An example of an important cost that is often hidden in firms is environmental costs. White, Savage, Brody, Cavander, and Lach (1995) present a three-part nested scheme for distinguishing between different types of environmental costs as shown in Figure 1. Conventional accounting costs are directly traceable environmental costs such as end-of-pipe emission treatment costs, which are captured by the accounting systems of most firms. Hidden costs are those that are not explicitly identified by the accounting system, such as changes to the production processes, and input substitutions attributable to environmental regulations. External costs include costs for which the firm is not cur- rently accountable but which may become material in the long run. These costs include contingent environmental liabilities such as those related to Superfund sites. It can be seen from the figure that conventional accounting costs identify only a small fraction of the total costs. A study by Joshi, Krishnan, and Lave (2001) reveals that in industries such as steel, these hidden costs can be substantial. Joshi et al. (2001) use a translog cost function External Costs Hidden Company Costs Conventional Company Costs Figure 1. Types of regulatory costs. Source: White et al. (1995). 4 Krishnan and data from 55 steel mills to explore whether accounting systems identify all the costs of environmental regulation. They econometrically distinguish the ‘visible’ cost of regulatory compliance that are appropriately classified in firms’ accounting systems, and ‘hidden’ costs, i.e. costs that are not visible because they are aggregated and included in other accounts. Seemingly unrelated regression estimations reveal substan- tial hidden costs. For every dollar of increase in the visible cost of environmental regu- lation, the total cost increases at the margin by $10–11. Therefore, $9–10 of environmental costs are hidden in other accounts and mislabeled as non-environmental costs and result in distorted costs across the board. The econometric estimates in Joshi et al. (2001) revealed substantial hidden costs. To ascertain whether managers in steel mills considered these estimates to be reason- able, the authors conducted a series of field interviews with senior managers of seven steel mills. All the managers interviewed acknowledged that they were aware of the substantial hidden environmental costs. Some even stated that these costs could be in the magnitude of 400% of the reported costs, on average. The managers offered several reasons for not attempting to isolate the hidden costs, including the technical difficulty in isolating these costs from other accounts, and the focus of the accounting system on accurately recording aggregate costs rather than capturing disaggregate costs with some noise. The controller of a Fortune 500 company had the following comment regarding identifying hidden environmental cost: There has been so much turnover in our Accounting and Environmental Affairs Depart- ments that the people who would have prepared the cost reports are no longer with the company. So I can’t tell you much about how that information was gathered and compiled. Often the information is requested in a form different from what we use to report under GAAP. This means that we provide what we think is correct but it may not have the con- sistency or reliability of other numbers, just because we don’t have a system in place to routinely define and gather the data. Our intent is to report the information at the lowest cost with the least amount of effort. We would identify direct costs that we believe are environmental but would not spend a lot of time trying to identify overhead costs or opportunity costs or really exploring the economics of the cost. This company declared bankruptcy within the next two years! Other comments from the field interviews included the following: Obtaining more detailed estimates of environmental costs would require a major change in the cost accounting systems. Collecting such level of detail on all regulatory costs does not pass the cost-benefit test. In any case, these costs are eventually captured under other cost categories, so total cost of production is not misstated. Moreover, the auditors have signed off on these estimates. The Joshi et al. (2001) study indicates that there are likely to be substantial errors in the accounting reports of firms that face significant environmental regulations such as steel, chemicals, paper, and oil refineries. These hidden costs arise because of reliance on traditional accounting systems whose main focus is the accurate reporting of aggre- gate costs. Omission of important cost drivers (such as environmental costs), will not only lead to less accurate decisions related to the management of these costs but also reduce the effectiveness of control systems by providing misleading information for use in flexible budgeting systems, variance analysis, and responsibility accounting systems. Hidden costs have implications for important decisions such as pricing, performance China Journal of Accounting Studies 5 measurement, product profitability analysis, plant closure, and investment decisions. Additionally hidden costs reduce the ability of the industry to engage in a debate and dialogue with regulators and policy makers regarding the consequences of regulation on productivity and performance. To understand and manage environmental and social costs, firms could explore refined accounting systems such as Activity-Based Costing, Resource Efficiency Accounting (REA), or Full Cost Accounting (FCA). 2.2. Excess capacity costs The Joshi et al. (2001) study indicated that traditional accounting systems that focus on the accurate reporting of aggregate costs can lead to substantial hidden costs with adverse consequences for a variety of important decisions. Another source of error in traditional accounting systems is the treatment of capacity costs. For decades, manage- ment accountants have argued that traditional, absorption-based accounting systems that allocate all costs, including fixed overhead costs to current production can lead manag- ers to produce in excess of demand and hide overhead costs in inventory (Balakrishnan & Sivaramakrishnan, 2002; Dickhaut & Lere, 1983; Turner & Hilton, 1989). This ‘real earnings management’ behavior has been empirically observed in a variety of industries (Gupta, Pevzner, & Seethamraju, 2008; Roychowdhury, 2006). However, in spite of the economic losses to multiple stakeholders from real earnings management, financial accounting systems continue to ignore the costs of excess capacity. Management accountants have long argued that the tendency for managers to produce in excess of demand will be curtailed if the cost of excess capacity is separated from current pro- duction (Cooper & Kaplan, 1992). The responsibility for excess capacity should not only be reported separately, but also the responsibility for excess capacity should be assigned to the individuals that caused such excess capacity to occur in the first place, and have control over such excess capacity (typically, the top management). In most firms, not only is excess capacity cost not identified separately, but also high powered incentives are attached to the inaccurate accounting numbers that result from these flawed capacity costing systems. Consider the following example: a plant has the practical capacity to produce 100,000 units. When the capacity was installed, it was expected that actual production would be less than the 100,000 units. The installed capacity decision was based on: (a) the lumpiness of capacity acquisitions; (b) the long-term horizon of investment decisions; and (c) to incorporate a buffer for future growth in demand. Further, assume that the fixed cost of maintaining a facility that can produce 100,000 units is $10 million. For simplicity, assume that the fixed cost of $10 million is comprised of costs such as depreciation, interest, and other long term commitments. Suppose during the year 2012, the expected demand is 50,000 units. Because expected demand can rarely be a point estimate, say, the marketing department estimation of expected demand is between 40,000 and 60,000 units. If the plant man- ager produces 50,000 units, the fixed cost per unit is $200. However, if the manager is provided with a bonus contract based on absorption costing using expected production as the denominator, the manager has the incentive to produce in excess of demand and ‘hide’ some of the fixed costs in ending inventory. For example, producing 60,000 units reduces the unit cost to $167 per unit, albeit at the cost of excess inventory of 10,000 units. These excess inventory units absorb $1.67 million of fixed costs, and thereby increase current reported income at the cost of future income. A study by Brüggen, Krishnan, and Sedatole (2011) examines the effect of absorp- tion costing systems on excess production in the auto industry. Additionally, their study 6 Krishnan investigates the association between excess production and tangible costs such as adver- tisement, inventory carrying costs, and discounts, as well as intangible costs such as brand image. The authors conduct archival as well as field analysis of the drivers and consequences of inappropriate allocation of excess capacity costs. Their field and archi- val analysis indicates that the combination of improper accounting for the fixed over- head cost of excess capacity and a performance measurement system that focuses on short-term costs and profits results in excess production. They find that for each per- centage point of excess capacity there is a 0.49-percentage-point increase in excess pro- duction. They also find a positive association between excess production and customer incentives, indicating that to sell the excess production, firms are required to spend on advertising as well as use discounts and other incentives. Importantly, their study indi- cates that excess production has implications for intangible costs such as brand image: for every additional 1% of rebate, there is a two-point decline in the JD Power APEAL index – a popular measure of brand image. Brüggen et al. (2011) conduct field interviews at a big-three auto maker. The fol- lowing remarks by senior managers reveal the extent of the problem caused by absorp- tion costing systems: ‘Excess capacity costs are not separated.’ ‘We build more to reduce unit costs.’ ‘And then a little more – to fill capacity.’ (Brüggen et al., 2011, p. 92) ‘The issue is that when the executive committee approves those volumes, they have been overly optimistic — extremely overly optimistic. And this is where the truth comes out. And again, this is the crux of the problem – in order to make the money – the profit tar- gets – you have to build more units. So, even though the [marketing department managers] come back and say, “Listen, we really can’t sell that many units”, they are told: “You have to sell more units, because otherwise we can’t hit the profit number.” And so we find a way to sell more units.’ (Brüggen et al., 2011, p. 93) ‘You don’t necessarily want to fall on your sword for the sake of a long-term profit down the road, because you may not be the one that’s in the chair when those long term profits come to roost. So, we get into this short-term cycle.’ (Bruggen et al., 2011, p. 95) In sum, Brüggen et al. (2011) provide systematic field and archival evidence that three characteristics of traditional accounting systems have implications for error in accounting systems. First, absorption costing systems force the inclusion of excess capacity costs into current production. This provides a convenient opportunity for man- agers to increase reported profits by hiding fixed costs into the ending inventory. Sec- ond, traditional costing systems have the goal of reporting transaction-oriented costs accurately and therefore ignore intangible costs such as brand image. When these two factors are added to performance measurement systems that attach high-powered incen- tives to short-term accounting profits, a dangerous cocktail of incorrect decisions emerge. These incorrect decisions can perpetuate for a long time, because managers do not want to ‘fall on their sword’ for the sake of long term profit. In addition to the need for appropriate identification and allocation of excess capac- ity, the Brüggen et al. (2011) study also demonstrates that it is possible to calibrate the value of an intangible asset such as brand image. Accounting research (e.g. Lev, 2001) China Journal of Accounting Studies 7 has argued that intangible assets such as patents, brands, and a unique organizational structure, are drivers of future firm value. These intangible drivers are not completely reflected in the physical and financial assets reported in the balance sheet, nor are they captured in the income statement where only the costs of depleting physical and finan- cial assets are reported. It is important for firms to understand how to measure the value of intangible assets because these assets are often the first to reflect the implica- tions of a company’s strategy (such as customer relationship management, which is reflected first in customer satisfaction and then only makes its way into the financial accounting system). A strategy of building to fill capacity to meet short term goals leads to overproduction and can erode brand image, which is an important intangible asset (Ashton, 2005). 2.3. Judgment errors in accounting Research in accounting provides ample guidance for the design of incentive contracts in firms to mitigate agency problems (Banker & Datar, 1989; Feltham & Xie, 1994). For example, research recommends that the weights placed on performance measures used for managerial incentive contracts should be in proportion to their precision (i.e., lack of noise), sensitivity (the extent to which the performance measure responds to managerial effort), and congruence (the association between the performance measure and the outcome desired by the firm, such as market value). When managers perform more than one task, they have to be evaluated based on multiple performance measures. In the presence of multiple performance measures, the optimal bonus contract needs to not only induce managerial effort (i.e., reduce moral hazard), but also ensure the appro- priate allocation of effort amongst the multiple tasks (Feltham & Xie, 1994; Holmström & Milgrom, 1991). Although analytical models provide guidance for the design of bonus contracts in the presence of multiple performance measures, the task of assigning weights to each of the performance measures is challenging because of the complexity of these models. Krishnan, Luft, & Shields (2005) experimentally study how individuals cope with the task of assigning weights to performance measures in multi-task situations. They study two properties of performance measures, i.e., precision and error covariance. In their setting, the manager is responsible for both cost and quality performance and therefore needs to be evaluated based on two performance measures, one for cost per- formance and the other for quality performance. As per the analytical model of Feltham and Xie (1994), if the precision of one of the measures (say, the cost measure) changes, then not only does the weight on the cost measure need to be changed, but the weight on the quality measure also needs to be changed. This is because if the weights on both measures are not changed, it will cause a sub-optimal allocation of effort. For example, if the precision of the cost measure reduces, then the firm has to reduce the weight on the cost measure as well as the quality measure. But if the firm does not reduce the weight on the quality measure, then the manager will shift effort towards the quality measure (that is more precisely measured) to an extent that is not optimal. Experimental results in Krishnan et al. (2005) indicate the following: first, many individuals do pay attention to the decrease in the cost measure precision. However, a significant number of individuals place a higher weight on the lower precision cost measure, contrary to the analytical model. The reason for this judgment error is that individuals do not have analytical ability to compute complex models. When con- fronted with complexity, individuals use qualitative and incomplete models, and 8 Krishnan substitute familiar attributes for formal models. When the precision of the cost measure decreases (i.e., it becomes more noisy/risky), instead of decreasing the weight on the cost measure, individuals use incomplete rule of thumb principles such as ‘risk and return should be positively associated’ and therefore increase the weight on the more risky measure. Many individuals also do not adjust the quality measure because they are unable to see the association between the two measures. The Krishnan et al. (2005) study indicates that individuals’ decisions systematically depart from model predictions because of errors of judgment and decision making. Research in accounting has demonstrated the presence of other types of judgment errors. For example Vera-Munoz (1998) finds that accounting knowledge is positively associated with the neglect of opportunity costs in business decisions, i.e., individuals with higher accounting knowledge are more likely to ignore opportunity costs. Vera- Munoz (1998) attributes this to the GAAP orientation of accounting graduates. When individuals with a high accounting knowledge are confronted with a resource allocation problem, they use their knowledge of GAAP-based rules to solve the problem. Because GAAP rules do not allow for the inclusion of opportunity costs in most instances, they tend to ignore opportunity costs while making business decisions. In sum, judgment errors can cause systematic departures of actual managerial reporting and decision making from optimal. There is a significant body of research in accounting that provides evidence of judgment errors and provides guidance on how to mitigate these errors. To the extent this research is not communicated in accounting undergraduate and graduate programs, these errors will continue to occur, causing the reported accounting performance to depart from the economic optimal. 3. Bias in accounting reports The previous section discussed the causes and consequences of errors in accounting reports. This section briefly summarizes the causes and consequences of bias in accounting reports. One of the primary drivers of bias is faulty accounting systems combined with managerial incentive contracts that assign reward based on flawed per- formance measures that result from these faulty systems. There is plentiful research in accounting that examines earnings management behaviors and calibrates the types of earnings management and magnitudes thereof. There is also general agreement that bonus contracts can encourage earnings management and slack-building behaviors. Considerable research in management accounting has studied the design of accounting and control systems to minimize slack. Prominent examples of such control systems include: participative budgeting (Baiman & Evans, 1983), and hurdle contracts (Antle & Eppen, 1985; Antle & Fellingham, 1995). If incentive compensation is a driver of bias, then it raises the question – why do firms use incentive contracts with managers? An important purpose of incentive com- pensation is to mitigate agency problems arising from motivational and informational problems that exist in decentralized organizations. These problems arise because the principal (shareholder or owner) is not able to observe the actions of the risk-averse and effort-averse agent (manager). Because of unobservability of effort, the agent may shirk instead of exerting effort (the moral hazard problem). Accounting literature has explored the design of compensation contracts such that the agent’s actions are in the best interest of the principal thereby achieving goal congruence. This is a major area of research in accounting because the design of contracts and the choice of performance measures require information from the management accounting system. Noisy, China Journal of Accounting Studies 9 incongruent, or insensitive performance measures can make incentive compensation ineffective. Noisy performance measures impose risk on the agent without commensu- rate return; therefore, the organization has to compensate the agent for the additional risk, which increases the cost of incentive compensation. Performance measures that are not sensitive to the agent’s effort are less efficacious at motivating effort. Incongruent performance measures allow the agent to improve the performance measure but not the underlying economic performance (Banker & Datar, 1989; Feltham &Xie, 1994). Although incentive compensation can be used as a tool to align the interests of the manager and the owner, considerable accounting literature has found that incentive compensation can cause dysfunctional behaviors including gaming, budget biasing, and earnings management. For example, Healy (1985) finds evidence that earnings-based incentive compensation is associated with earnings management. His results indicate that managers choose income-increasing accruals when their bonus plans do not contain an upper bound on earnings. Similarly, Holthausen, Larcker, & Sloan (1995) find that managers perform earnings decreasing manipulations after they achieve the maximum limit on their bonuses. A large body of literature in management accounting has found that incentive com- pensation based on a budgetary target can lead to budgetary biasing and budgetary slack (Dunk, 1993; Schiff & Lewin, 1970; Young, 1985). Often, the primary assump- tion in the economics based literature that influences the design of budgeting systems is that employees will try to create budgetary slack and misreport to the extent possible to maximize their earnings. Considerable research in management accounting has explored the design and implications of budgeting practices, using a variety of theoreti- cal perspectives. Research has examined the role of budgets in performance measure- ment, budgetary target setting, budget biasing and gaming, and the causes and consequences of budgetary slack (see Covaleski, Evans, Luft, & Shields, 2003, and Luft & Shields, 2009, for reviews). At the same time, research in accounting also shows that individuals often do not misreport, even when they have the opportunity to do so. The pioneering research in behavioral decision theory (BDT) by Daniel Kahneman, which won the 2002 Nobel Prize in economics, challenges the basic assumptions of neo-classical economic theory. Research in accounting and finance has begun to explore the implications of BDT on the behavior of financial markets and design of accounting systems. In the design of control systems, it is important to take into consideration individuals’ inherent prefer- ences for honesty, fairness, reciprocity, and other social norms. For example, Evans, Hannan, Krishnan, and Moser (2001) test how preferences for wealth and honesty affect managerial reporting, through a series of experiments. Contrary to predictions from conventional agency theory, which assumes that people obtain utility only from monetary wealth, they find that subjects often sacrifice wealth to make honest or par- tially honest reports, and they generally do not lie more as the payoff to lying increases. Evans et al. (2001) also find less honesty under a contract that provides a smaller share of the total surplus to the manager compared with a contract that provides a larger share, suggesting that the extent of honesty is likely to depend on how the sur- plus is divided between the manager and the firm. Incorporating such preferences for honesty and reciprocity can assist in the design of contract and control systems that lead to improved outcomes compared with those based exclusively on conventional economic assumptions. 10 Krishnan 4. Conclusions This paper examines how accounting systems can include error and bias, resulting in the output from the accounting system departing from true economic performance. It discusses three sources of error in accounting systems, i.e., the use of aggregate, vol- ume-based allocation systems, absorption costing systems that incorrectly include excess capacity costs in current production, and judgment errors by managers, primarily arising from cognitive limitations and a GAAP mentality. Bias arises primarily from an excessive bonus attached to flawed performance measures that result from poorly designed accounting systems. Moreover, compensation contracts are typically based on the assumption that individuals have utility primarily for monetary wealth. I discuss how control systems that disregard the preferences of individuals can lead to the design of accounting and control systems that can crowd out preferences for honesty and reci- procity. A well-designed MCS should provide managers with the ability to anticipate errors and biases, make assessments about tolerance limits around the errors and biases, and ensure that the sub-systems built around the overall MCS (such as budgeting and variance analysis, compensation systems, etc.) can function effectively even when there are errors and biases. There are a number of research opportunities for scholars in the area of MCS. Error in MCS arises from (a) inappropriate identification of costs; (b) inappropriate aggrega- tion; and (c) inaccurate cost prediction models. Future research could explore decompo- sition of errors along these dimensions and analyze the relative contributions and causes. However, the major challenge in error analysis is establishing the counterfactual – that is, what is the error-free baseline? This often means going down to the transac- tion-level analysis and systematically re-analyzing and re-aggregating the costs into appropriate cost pools, which may require field-based studies. Research is also needed to explore the tradeoffs between the costs and benefits of reducing errors. These trade- offs are ultimately endogenous to the design of MCS. Regarding bias, future research could examine how firms contract with managers. The question of whether high-pow- ered incentives crowd out managers’ preferences for honesty has not been studied. Human resource policies that build organizational commitment could weaken the need for using incentive compensation for goal congruence. There is an entire array of behavioral drivers of managerial motives for misreporting such as trust, reciprocity, gift-exchange, social distance, identity, other-regarding behaviors, and so on, which can be incorporated into optimal contract design. The competitive and regulatory landscape in which companies operate is constantly changing, which places pressures on even the best designed MCS. In addition, organi- zations across the world are exploring new types of performance measurement and con- trol systems, information systems, and stretching the boundaries of their firms by collaborative contracting with suppliers and customers. These changes create challenges in designing appropriate contracting and performance measurement systems. Research could explore the implications of these changes using data from public or proprietary sources. This is an important topic because inappropriate design of MCS can have serious implications for firms and managers as well as for policy. Acknowledgements This paper was the based on a talk delivered at the annual conference of the China Journal of Accounting Studies in Dalian in May 2013. I thank Jason Xiao, Satish Joshi, and two anonymous reviewers for their helpful comments. China Journal of Accounting Studies 11 Note 1. See Sprinkle and Williamson (2007) for a comprehensive review of the decision facilitating and decision influencing roles of management accounting information and their implications for experimental research in accounting. References Antle, R., & Eppen, G.D. (1985). Capital rationing and organizational slack in capital Budgeting. Management Science, 31, 163–174. Antle, R., & Fellingham, J. (1995). Information rents and preferences among information systems in a model of resource allocation. Journal of Accounting Research, 33,41–58. Ashton, R.H. (2005). Intellectual capital and value creation: A review. Journal of Accounting Literature, 24, 53. Baiman, S. (1990). Agency research in managerial accounting: A second look. Accounting, Organizations and Society, 15(4), 341–372. Baiman, S., & Evans III, J.H. (1983). Pre-decision information and participative management control systems. Journal of Accounting Research, 22, 371–395. Balakrishnan, R., & Sivaramakrishnan, K. (2002). A critical overview of the use of full-cost data for planning and pricing. Journal of Management Accounting Research, 14,3–31. Banker, R., & Datar, S. (1989). Sensitivity, precision and linear aggregation of signals for perfor- mance evaluation. Journal of Accounting Research, 27,21–39. Brüggen, A., Krishnan, R., & Sedatole, K. (2011). Drivers and consequences of short-term production decisions: Evidence from the auto industry. Contemporary Accounting Research, 28,83–123. Cooper, R., & Kaplan, R.S. (1992). Activity-based systems: measuring the costs of resource usage. Accounting Horizons, 6(3), 1–13. Covaleski, M., Evans III, J.H., Luft, J., & Shields, M.D. (2003). Budgeting research: three theo- retical perspectives for selective integration. Journal of Management Accounting Research, 15,3–49. Demski, J., & Feltham, G. (1976). Cost determination: A conceptual approach. Ames, IA: Iowa State Press. Dickhaut, J., & Lere, J.C. (1983). Comparison of accounting systems and heuristics in selecting economic optima. Journal of Accounting Research, 27(2), 495–513. Dunk, A.S. (1993). The effect of budget emphasis and information asymmetry on the relation between budgetary participation and slack. The Accounting Review, 68(2), 400–410. Evans III, J.H., Hannan, R.L., Krishnan, R., & Moser, D. (2001). Honesty in managerial reporting. The Accounting Review, 76, 537–559. Fama, E., & Jensen, M. (1983). Agency problems and residual claims. Journal of Law and Economics, 26, 301–325. Feltham, G.A., & Xie, J. (1994). Performance measure congruity and diversity in multi-task prin- cipal/agent relations. The Accounting Review, 69, 429–453. Gupta, M., Pevzner, M., & Seethamraju, C. (2008). The implications of absorption cost account- ing and production decisions for firms’ future performance and valuation. Working paper. Healy, P. (1985). The effect of bonus schemes on accounting decisions. Journal of Accounting and Economics, 7,85–107. Holmström, B., & Milgrom, P. (1991). Multitask principal-agent analyses: Incentive contracts, asset ownership and job design. Journal of Law, Economics & Organization, 7,24–52. Holthausen, R.W., Larcker, D.F., & Sloan, R.G. (1995). Annual bonus schemes and the manipulation of earnings. Journal of Accounting and Economics, 19,29–74. Jensen, M.C., & Meckling, W.H. (1976). Theory of the firm: Managerial behavior, agency costs, and ownership structure. Journal of Financial Economics, 4, 469–506. Joshi, S., Krishnan, R., & Lave, L. (2001). Estimating the hidden costs of environmental regulations. The Accounting Review, 76, 171–198. Krishnan, R., Luft, J., & Shields, M.D. (2005). Effects of accounting-method choices on subjective performance-measure weighting decisions: experimental evidence on precision and error covariance. The Accounting Review, 80, 1163–1192. 12 Krishnan Lev, B. (2001). Intangibles: Management, measurement, and reporting. Washington, DC: Brookings Institution Press. Luft, J., & Shields, M.D. (2009). Psychology models of management accounting. Foundations and Trends in Accounting, 4, 199–345. Macintosh, N.B., & Quattrone, P. (2010). Management accounting and control systems: An organizational and sociological approach. UK: Wiley. Roychowdhury, S. (2006). Earnings management through real activities manipulation. Journal of Accounting and Economics, 42, 335–370. Schiff, M., & Lewin, A.Y. (1970). The impact of people on budgets. The Accounting Review, 45 (2), 259–268. Sprinkle, G.B., & Williamson, M.G. (2007). Experimental research in managerial accounting. In C. Chapman, A. Hopwood & M. Shields (Eds.) Handbook of management accounting research. New York: Elsevier. Turner, M.J., & Hilton, M.G. (1989). Use of accounting product-costing systems in making production decisions. Journal of Accounting Research, 27, 297–312. Vera-Munoz, S. (1998). The effects of accounting knowledge and context on the omission of opportunity costs in resource allocation decisions. The Accounting Review, 73,47–72. White, A.L., Savage, D.E., Brody, J., Cavander, D., & Lach, L. (1995). Environmental cost accounting for capital budgeting: A benchmark survey of management accountants. US Environmental Protection Agency, Report no. EPA742-R-95-005. Young, S.M. (1985). Participative budgeting: The effects of risk aversion and asymmetric information on budgetary slack. Journal of Accounting Research, 23(2), 829–842. Zimmerman, J. (2009). Accounting for decision making and control. New York: McGraw-Hill http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png China Journal of Accounting Studies Taylor & Francis

Causes and consequences of error and bias in management accounting systems

Krishnan, Ranjani
China Journal of Accounting Studies , Volume 2 (1): 12 – Jan 2, 2014
Download PDF
12 pages

Loading next page...
 
/lp/taylor-francis/causes-and-consequences-of-error-and-bias-in-management-accounting-Tgfb4E0HiS

References (39)

Publisher
Taylor & Francis
Copyright
© 2014 Accounting Society of China
ISSN
2169-7221
eISSN
2169-7213
DOI
10.1080/21697221.2014.893774
Publisher site
See Article on Publisher Site

Abstract

China Journal of Accounting Studies, 2014 Vol. 2, No. 1, 1–12, http://dx.doi.org/10.1080/21697221.2014.893774 COMMENTARY Causes and consequences of error and bias in management accounting systems Ranjani Krishnan* Broad College of Business, Michigan State University, 632 Bogue Street, N207; East Lansing, MI 48824-1122, USA The purpose of this paper is to provide an overview of management accounting sys- tems, particularly with respect to their influence on error and bias in decision mak- ing. It also discusses the implications of such errors and biases for organizations and for policy and suggests avenues for future research. Keywords: management accounting system design; error; bias 1. Introduction The goal of management accounting is to provide information for decision making and control to managers within the firm. The literature is awash with suggested frameworks to analyze the different roles of management accounting. One popular framework is Demski and Feltham (1976), which distinguishes between the decision facilitating role and the decision influencing role of management accounting. The decision facilitating role of management accounting information is its use in problem solving and belief revision. In this role, the management accounting system assists managers to evaluate whether or not performance is proceeding as planned, identify the reasons for depar- tures from plan, and determine appropriate corrective actions. Budgeting and variance analysis is an example of the decision facilitating role of management accounting infor- mation. The decision influencing role is the use of accounting information to motivate employees and achieve goal congruence. Contracting, incentive design, and perfor- mance measurement systems are examples of the decision influencing role. A similar framework is used by Zimmerman (2009) in his book where he identifies the role of management accounting as serving the purposes of decision making (similar to the decision facilitating role) and control (similar to the decision influencing role). The decision facilitating/influencing framework uses a largely economic lens to view the role of management accounting, although there is scope to incorporate judg- ment and decision-making in both these roles. An alternative framework is used by Macintosh and Quattrone (2010), which characterizes management accounting as a social system with a goal not only to motivate and manage employees, but also to sanction the actions of managers. They view management accounting as a part of a wide array of other systems such as information and communication systems. As per this view, management accounting system design cannot be considered in isolation from other systems; rather, their features only have relevance within the network of relationships and other systems in place in the organization. *Email: Krishnan@bus.msu.edu Paper accepted by Jason Xiao. © 2014 Accounting Society of China 2 Krishnan Regardless of the perspective, in order to provide useful information for decision making, the management accounting and control system (MCS) should be designed to reduce error and bias in the numbers and reports that are produced by the organization. Such design should at the same time take into consideration the cost versus the benefit of reducing the errors and biases. One important driver of error is poorly designed MCS, which causes the reported performance to differ from the actual economic perfor- mance of interest. One important driver of bias is the agency problem, i.e. decisions within firms are made by managers whose goals are not aligned with those of the shareholders (Fama & Jensen, 1983; Jensen & Meckling, 1976). Goal incongruence is a problem in decentralized firms because of the inability of shareholders to observe managerial effort, which could encourage managerial shirking. Furthermore, relative to shareholders, managers often have better information about operations. This can cause managers to misrepresent their private information and maximize their own interest rather than the interest of the firms (Baiman, 1990). Incentive compensation is often used as a tool to address this goal incongruence problem between managers and share- holders. However, poorly designed incentive compensation systems can encourage managers to build bias into the numbers. For example, excessive incentive compensa- tion combined with poor quality and incongruent performance measures can encourage budgetary slack, earnings management, and fraud. In the next section, I provide examples of several types of errors that can arise from poor quality MCS and the implications of these errors. This is followed by a discussion of the potential for bias in reported accounting numbers and the consequences thereof. The final section concludes. 2. Errors in accounting reports Errors in accounting reports arise from several sources, the most prominent of which are incorrect allocation of overhead cost and incorrect identification of capacity costs. Traditional management accounting systems with crude, volume based allocations cause departures of reported costs from actual costs. Absorption-oriented accounting systems force the costs of excess capacity to be allocated to current production. Similarly, accounting systems that focus on the appropriate capture of aggregate costs rather than appropriate allocation of costs based on cause and effect, result in a number of impor- tant costs to be invisible to the decision makers. These hidden costs can be substantial in many industries and can cause flawed decisions about current and future resource allocations. Another contributor of accounting error is flawed judgment and psychological biases. Most decisions in firms are made by non-expert managers who do not have the cognitive resources to accurately apply complex economic models or analytical tech- niques. Similarly, accounting systems are designed by non-accountants such as plant managers, branch managers, and other front-line employees. These individuals often face challenges with respect to both analysis and judgment skills. While sometimes these gaps in knowledge result in unsystematic errors in decision making, which thereby increases the noise in the system, at other times these knowledge gaps result in systematic departures from optimal decisions leading to biased results. This section provides examples of errors in management accounting systems that result from three sources: (1) use of traditional accounting systems that focus on appro- priate recording of aggregate costs and revenues, resulting in hidden costs; (2) inappro- priate allocation of excess capacity costs; and (3) use of subjective decision tools that China Journal of Accounting Studies 3 can be qualitative, incomplete, and substitute informal and familiar attributes for the formal attributes of scientific models. 2.1. Traditional accounting systems and hidden costs Management accounting research has long advocated the use of refined cost allocation systems that systematically track, pool, and allocate overhead costs to cost objects in an accurate manner (Cooper & Kaplan, 1992). Although overhead cost allocation does require subjectivity, judgment, and assumptions, as long as there is an attempt to match overhead resource consumption with the appropriate cost drivers, errors can be mini- mized. Almost all undergraduate and graduate accounting programs, as well as MBA programs teach activity-based costing (ABC) and other refined cost allocation systems. Nevertheless, most firms continue to use traditional cost allocation systems with volume-based cost drivers. Adherence to outdated accounting systems often stems from an absorption costing mentality, where the goal is to report aggregate costs accurately. When accounting systems focus on the aggregate, they fail to isolate the impact of important exogenous factors that affect their costs such as competition and regulation. Although managers acknowledge that these costs can be substantial, most firms do not have adequate systems for identifying and measuring hidden costs. An example of an important cost that is often hidden in firms is environmental costs. White, Savage, Brody, Cavander, and Lach (1995) present a three-part nested scheme for distinguishing between different types of environmental costs as shown in Figure 1. Conventional accounting costs are directly traceable environmental costs such as end-of-pipe emission treatment costs, which are captured by the accounting systems of most firms. Hidden costs are those that are not explicitly identified by the accounting system, such as changes to the production processes, and input substitutions attributable to environmental regulations. External costs include costs for which the firm is not cur- rently accountable but which may become material in the long run. These costs include contingent environmental liabilities such as those related to Superfund sites. It can be seen from the figure that conventional accounting costs identify only a small fraction of the total costs. A study by Joshi, Krishnan, and Lave (2001) reveals that in industries such as steel, these hidden costs can be substantial. Joshi et al. (2001) use a translog cost function External Costs Hidden Company Costs Conventional Company Costs Figure 1. Types of regulatory costs. Source: White et al. (1995). 4 Krishnan and data from 55 steel mills to explore whether accounting systems identify all the costs of environmental regulation. They econometrically distinguish the ‘visible’ cost of regulatory compliance that are appropriately classified in firms’ accounting systems, and ‘hidden’ costs, i.e. costs that are not visible because they are aggregated and included in other accounts. Seemingly unrelated regression estimations reveal substan- tial hidden costs. For every dollar of increase in the visible cost of environmental regu- lation, the total cost increases at the margin by $10–11. Therefore, $9–10 of environmental costs are hidden in other accounts and mislabeled as non-environmental costs and result in distorted costs across the board. The econometric estimates in Joshi et al. (2001) revealed substantial hidden costs. To ascertain whether managers in steel mills considered these estimates to be reason- able, the authors conducted a series of field interviews with senior managers of seven steel mills. All the managers interviewed acknowledged that they were aware of the substantial hidden environmental costs. Some even stated that these costs could be in the magnitude of 400% of the reported costs, on average. The managers offered several reasons for not attempting to isolate the hidden costs, including the technical difficulty in isolating these costs from other accounts, and the focus of the accounting system on accurately recording aggregate costs rather than capturing disaggregate costs with some noise. The controller of a Fortune 500 company had the following comment regarding identifying hidden environmental cost: There has been so much turnover in our Accounting and Environmental Affairs Depart- ments that the people who would have prepared the cost reports are no longer with the company. So I can’t tell you much about how that information was gathered and compiled. Often the information is requested in a form different from what we use to report under GAAP. This means that we provide what we think is correct but it may not have the con- sistency or reliability of other numbers, just because we don’t have a system in place to routinely define and gather the data. Our intent is to report the information at the lowest cost with the least amount of effort. We would identify direct costs that we believe are environmental but would not spend a lot of time trying to identify overhead costs or opportunity costs or really exploring the economics of the cost. This company declared bankruptcy within the next two years! Other comments from the field interviews included the following: Obtaining more detailed estimates of environmental costs would require a major change in the cost accounting systems. Collecting such level of detail on all regulatory costs does not pass the cost-benefit test. In any case, these costs are eventually captured under other cost categories, so total cost of production is not misstated. Moreover, the auditors have signed off on these estimates. The Joshi et al. (2001) study indicates that there are likely to be substantial errors in the accounting reports of firms that face significant environmental regulations such as steel, chemicals, paper, and oil refineries. These hidden costs arise because of reliance on traditional accounting systems whose main focus is the accurate reporting of aggre- gate costs. Omission of important cost drivers (such as environmental costs), will not only lead to less accurate decisions related to the management of these costs but also reduce the effectiveness of control systems by providing misleading information for use in flexible budgeting systems, variance analysis, and responsibility accounting systems. Hidden costs have implications for important decisions such as pricing, performance China Journal of Accounting Studies 5 measurement, product profitability analysis, plant closure, and investment decisions. Additionally hidden costs reduce the ability of the industry to engage in a debate and dialogue with regulators and policy makers regarding the consequences of regulation on productivity and performance. To understand and manage environmental and social costs, firms could explore refined accounting systems such as Activity-Based Costing, Resource Efficiency Accounting (REA), or Full Cost Accounting (FCA). 2.2. Excess capacity costs The Joshi et al. (2001) study indicated that traditional accounting systems that focus on the accurate reporting of aggregate costs can lead to substantial hidden costs with adverse consequences for a variety of important decisions. Another source of error in traditional accounting systems is the treatment of capacity costs. For decades, manage- ment accountants have argued that traditional, absorption-based accounting systems that allocate all costs, including fixed overhead costs to current production can lead manag- ers to produce in excess of demand and hide overhead costs in inventory (Balakrishnan & Sivaramakrishnan, 2002; Dickhaut & Lere, 1983; Turner & Hilton, 1989). This ‘real earnings management’ behavior has been empirically observed in a variety of industries (Gupta, Pevzner, & Seethamraju, 2008; Roychowdhury, 2006). However, in spite of the economic losses to multiple stakeholders from real earnings management, financial accounting systems continue to ignore the costs of excess capacity. Management accountants have long argued that the tendency for managers to produce in excess of demand will be curtailed if the cost of excess capacity is separated from current pro- duction (Cooper & Kaplan, 1992). The responsibility for excess capacity should not only be reported separately, but also the responsibility for excess capacity should be assigned to the individuals that caused such excess capacity to occur in the first place, and have control over such excess capacity (typically, the top management). In most firms, not only is excess capacity cost not identified separately, but also high powered incentives are attached to the inaccurate accounting numbers that result from these flawed capacity costing systems. Consider the following example: a plant has the practical capacity to produce 100,000 units. When the capacity was installed, it was expected that actual production would be less than the 100,000 units. The installed capacity decision was based on: (a) the lumpiness of capacity acquisitions; (b) the long-term horizon of investment decisions; and (c) to incorporate a buffer for future growth in demand. Further, assume that the fixed cost of maintaining a facility that can produce 100,000 units is $10 million. For simplicity, assume that the fixed cost of $10 million is comprised of costs such as depreciation, interest, and other long term commitments. Suppose during the year 2012, the expected demand is 50,000 units. Because expected demand can rarely be a point estimate, say, the marketing department estimation of expected demand is between 40,000 and 60,000 units. If the plant man- ager produces 50,000 units, the fixed cost per unit is $200. However, if the manager is provided with a bonus contract based on absorption costing using expected production as the denominator, the manager has the incentive to produce in excess of demand and ‘hide’ some of the fixed costs in ending inventory. For example, producing 60,000 units reduces the unit cost to $167 per unit, albeit at the cost of excess inventory of 10,000 units. These excess inventory units absorb $1.67 million of fixed costs, and thereby increase current reported income at the cost of future income. A study by Brüggen, Krishnan, and Sedatole (2011) examines the effect of absorp- tion costing systems on excess production in the auto industry. Additionally, their study 6 Krishnan investigates the association between excess production and tangible costs such as adver- tisement, inventory carrying costs, and discounts, as well as intangible costs such as brand image. The authors conduct archival as well as field analysis of the drivers and consequences of inappropriate allocation of excess capacity costs. Their field and archi- val analysis indicates that the combination of improper accounting for the fixed over- head cost of excess capacity and a performance measurement system that focuses on short-term costs and profits results in excess production. They find that for each per- centage point of excess capacity there is a 0.49-percentage-point increase in excess pro- duction. They also find a positive association between excess production and customer incentives, indicating that to sell the excess production, firms are required to spend on advertising as well as use discounts and other incentives. Importantly, their study indi- cates that excess production has implications for intangible costs such as brand image: for every additional 1% of rebate, there is a two-point decline in the JD Power APEAL index – a popular measure of brand image. Brüggen et al. (2011) conduct field interviews at a big-three auto maker. The fol- lowing remarks by senior managers reveal the extent of the problem caused by absorp- tion costing systems: ‘Excess capacity costs are not separated.’ ‘We build more to reduce unit costs.’ ‘And then a little more – to fill capacity.’ (Brüggen et al., 2011, p. 92) ‘The issue is that when the executive committee approves those volumes, they have been overly optimistic — extremely overly optimistic. And this is where the truth comes out. And again, this is the crux of the problem – in order to make the money – the profit tar- gets – you have to build more units. So, even though the [marketing department managers] come back and say, “Listen, we really can’t sell that many units”, they are told: “You have to sell more units, because otherwise we can’t hit the profit number.” And so we find a way to sell more units.’ (Brüggen et al., 2011, p. 93) ‘You don’t necessarily want to fall on your sword for the sake of a long-term profit down the road, because you may not be the one that’s in the chair when those long term profits come to roost. So, we get into this short-term cycle.’ (Bruggen et al., 2011, p. 95) In sum, Brüggen et al. (2011) provide systematic field and archival evidence that three characteristics of traditional accounting systems have implications for error in accounting systems. First, absorption costing systems force the inclusion of excess capacity costs into current production. This provides a convenient opportunity for man- agers to increase reported profits by hiding fixed costs into the ending inventory. Sec- ond, traditional costing systems have the goal of reporting transaction-oriented costs accurately and therefore ignore intangible costs such as brand image. When these two factors are added to performance measurement systems that attach high-powered incen- tives to short-term accounting profits, a dangerous cocktail of incorrect decisions emerge. These incorrect decisions can perpetuate for a long time, because managers do not want to ‘fall on their sword’ for the sake of long term profit. In addition to the need for appropriate identification and allocation of excess capac- ity, the Brüggen et al. (2011) study also demonstrates that it is possible to calibrate the value of an intangible asset such as brand image. Accounting research (e.g. Lev, 2001) China Journal of Accounting Studies 7 has argued that intangible assets such as patents, brands, and a unique organizational structure, are drivers of future firm value. These intangible drivers are not completely reflected in the physical and financial assets reported in the balance sheet, nor are they captured in the income statement where only the costs of depleting physical and finan- cial assets are reported. It is important for firms to understand how to measure the value of intangible assets because these assets are often the first to reflect the implica- tions of a company’s strategy (such as customer relationship management, which is reflected first in customer satisfaction and then only makes its way into the financial accounting system). A strategy of building to fill capacity to meet short term goals leads to overproduction and can erode brand image, which is an important intangible asset (Ashton, 2005). 2.3. Judgment errors in accounting Research in accounting provides ample guidance for the design of incentive contracts in firms to mitigate agency problems (Banker & Datar, 1989; Feltham & Xie, 1994). For example, research recommends that the weights placed on performance measures used for managerial incentive contracts should be in proportion to their precision (i.e., lack of noise), sensitivity (the extent to which the performance measure responds to managerial effort), and congruence (the association between the performance measure and the outcome desired by the firm, such as market value). When managers perform more than one task, they have to be evaluated based on multiple performance measures. In the presence of multiple performance measures, the optimal bonus contract needs to not only induce managerial effort (i.e., reduce moral hazard), but also ensure the appro- priate allocation of effort amongst the multiple tasks (Feltham & Xie, 1994; Holmström & Milgrom, 1991). Although analytical models provide guidance for the design of bonus contracts in the presence of multiple performance measures, the task of assigning weights to each of the performance measures is challenging because of the complexity of these models. Krishnan, Luft, & Shields (2005) experimentally study how individuals cope with the task of assigning weights to performance measures in multi-task situations. They study two properties of performance measures, i.e., precision and error covariance. In their setting, the manager is responsible for both cost and quality performance and therefore needs to be evaluated based on two performance measures, one for cost per- formance and the other for quality performance. As per the analytical model of Feltham and Xie (1994), if the precision of one of the measures (say, the cost measure) changes, then not only does the weight on the cost measure need to be changed, but the weight on the quality measure also needs to be changed. This is because if the weights on both measures are not changed, it will cause a sub-optimal allocation of effort. For example, if the precision of the cost measure reduces, then the firm has to reduce the weight on the cost measure as well as the quality measure. But if the firm does not reduce the weight on the quality measure, then the manager will shift effort towards the quality measure (that is more precisely measured) to an extent that is not optimal. Experimental results in Krishnan et al. (2005) indicate the following: first, many individuals do pay attention to the decrease in the cost measure precision. However, a significant number of individuals place a higher weight on the lower precision cost measure, contrary to the analytical model. The reason for this judgment error is that individuals do not have analytical ability to compute complex models. When con- fronted with complexity, individuals use qualitative and incomplete models, and 8 Krishnan substitute familiar attributes for formal models. When the precision of the cost measure decreases (i.e., it becomes more noisy/risky), instead of decreasing the weight on the cost measure, individuals use incomplete rule of thumb principles such as ‘risk and return should be positively associated’ and therefore increase the weight on the more risky measure. Many individuals also do not adjust the quality measure because they are unable to see the association between the two measures. The Krishnan et al. (2005) study indicates that individuals’ decisions systematically depart from model predictions because of errors of judgment and decision making. Research in accounting has demonstrated the presence of other types of judgment errors. For example Vera-Munoz (1998) finds that accounting knowledge is positively associated with the neglect of opportunity costs in business decisions, i.e., individuals with higher accounting knowledge are more likely to ignore opportunity costs. Vera- Munoz (1998) attributes this to the GAAP orientation of accounting graduates. When individuals with a high accounting knowledge are confronted with a resource allocation problem, they use their knowledge of GAAP-based rules to solve the problem. Because GAAP rules do not allow for the inclusion of opportunity costs in most instances, they tend to ignore opportunity costs while making business decisions. In sum, judgment errors can cause systematic departures of actual managerial reporting and decision making from optimal. There is a significant body of research in accounting that provides evidence of judgment errors and provides guidance on how to mitigate these errors. To the extent this research is not communicated in accounting undergraduate and graduate programs, these errors will continue to occur, causing the reported accounting performance to depart from the economic optimal. 3. Bias in accounting reports The previous section discussed the causes and consequences of errors in accounting reports. This section briefly summarizes the causes and consequences of bias in accounting reports. One of the primary drivers of bias is faulty accounting systems combined with managerial incentive contracts that assign reward based on flawed per- formance measures that result from these faulty systems. There is plentiful research in accounting that examines earnings management behaviors and calibrates the types of earnings management and magnitudes thereof. There is also general agreement that bonus contracts can encourage earnings management and slack-building behaviors. Considerable research in management accounting has studied the design of accounting and control systems to minimize slack. Prominent examples of such control systems include: participative budgeting (Baiman & Evans, 1983), and hurdle contracts (Antle & Eppen, 1985; Antle & Fellingham, 1995). If incentive compensation is a driver of bias, then it raises the question – why do firms use incentive contracts with managers? An important purpose of incentive com- pensation is to mitigate agency problems arising from motivational and informational problems that exist in decentralized organizations. These problems arise because the principal (shareholder or owner) is not able to observe the actions of the risk-averse and effort-averse agent (manager). Because of unobservability of effort, the agent may shirk instead of exerting effort (the moral hazard problem). Accounting literature has explored the design of compensation contracts such that the agent’s actions are in the best interest of the principal thereby achieving goal congruence. This is a major area of research in accounting because the design of contracts and the choice of performance measures require information from the management accounting system. Noisy, China Journal of Accounting Studies 9 incongruent, or insensitive performance measures can make incentive compensation ineffective. Noisy performance measures impose risk on the agent without commensu- rate return; therefore, the organization has to compensate the agent for the additional risk, which increases the cost of incentive compensation. Performance measures that are not sensitive to the agent’s effort are less efficacious at motivating effort. Incongruent performance measures allow the agent to improve the performance measure but not the underlying economic performance (Banker & Datar, 1989; Feltham &Xie, 1994). Although incentive compensation can be used as a tool to align the interests of the manager and the owner, considerable accounting literature has found that incentive compensation can cause dysfunctional behaviors including gaming, budget biasing, and earnings management. For example, Healy (1985) finds evidence that earnings-based incentive compensation is associated with earnings management. His results indicate that managers choose income-increasing accruals when their bonus plans do not contain an upper bound on earnings. Similarly, Holthausen, Larcker, & Sloan (1995) find that managers perform earnings decreasing manipulations after they achieve the maximum limit on their bonuses. A large body of literature in management accounting has found that incentive com- pensation based on a budgetary target can lead to budgetary biasing and budgetary slack (Dunk, 1993; Schiff & Lewin, 1970; Young, 1985). Often, the primary assump- tion in the economics based literature that influences the design of budgeting systems is that employees will try to create budgetary slack and misreport to the extent possible to maximize their earnings. Considerable research in management accounting has explored the design and implications of budgeting practices, using a variety of theoreti- cal perspectives. Research has examined the role of budgets in performance measure- ment, budgetary target setting, budget biasing and gaming, and the causes and consequences of budgetary slack (see Covaleski, Evans, Luft, & Shields, 2003, and Luft & Shields, 2009, for reviews). At the same time, research in accounting also shows that individuals often do not misreport, even when they have the opportunity to do so. The pioneering research in behavioral decision theory (BDT) by Daniel Kahneman, which won the 2002 Nobel Prize in economics, challenges the basic assumptions of neo-classical economic theory. Research in accounting and finance has begun to explore the implications of BDT on the behavior of financial markets and design of accounting systems. In the design of control systems, it is important to take into consideration individuals’ inherent prefer- ences for honesty, fairness, reciprocity, and other social norms. For example, Evans, Hannan, Krishnan, and Moser (2001) test how preferences for wealth and honesty affect managerial reporting, through a series of experiments. Contrary to predictions from conventional agency theory, which assumes that people obtain utility only from monetary wealth, they find that subjects often sacrifice wealth to make honest or par- tially honest reports, and they generally do not lie more as the payoff to lying increases. Evans et al. (2001) also find less honesty under a contract that provides a smaller share of the total surplus to the manager compared with a contract that provides a larger share, suggesting that the extent of honesty is likely to depend on how the sur- plus is divided between the manager and the firm. Incorporating such preferences for honesty and reciprocity can assist in the design of contract and control systems that lead to improved outcomes compared with those based exclusively on conventional economic assumptions. 10 Krishnan 4. Conclusions This paper examines how accounting systems can include error and bias, resulting in the output from the accounting system departing from true economic performance. It discusses three sources of error in accounting systems, i.e., the use of aggregate, vol- ume-based allocation systems, absorption costing systems that incorrectly include excess capacity costs in current production, and judgment errors by managers, primarily arising from cognitive limitations and a GAAP mentality. Bias arises primarily from an excessive bonus attached to flawed performance measures that result from poorly designed accounting systems. Moreover, compensation contracts are typically based on the assumption that individuals have utility primarily for monetary wealth. I discuss how control systems that disregard the preferences of individuals can lead to the design of accounting and control systems that can crowd out preferences for honesty and reci- procity. A well-designed MCS should provide managers with the ability to anticipate errors and biases, make assessments about tolerance limits around the errors and biases, and ensure that the sub-systems built around the overall MCS (such as budgeting and variance analysis, compensation systems, etc.) can function effectively even when there are errors and biases. There are a number of research opportunities for scholars in the area of MCS. Error in MCS arises from (a) inappropriate identification of costs; (b) inappropriate aggrega- tion; and (c) inaccurate cost prediction models. Future research could explore decompo- sition of errors along these dimensions and analyze the relative contributions and causes. However, the major challenge in error analysis is establishing the counterfactual – that is, what is the error-free baseline? This often means going down to the transac- tion-level analysis and systematically re-analyzing and re-aggregating the costs into appropriate cost pools, which may require field-based studies. Research is also needed to explore the tradeoffs between the costs and benefits of reducing errors. These trade- offs are ultimately endogenous to the design of MCS. Regarding bias, future research could examine how firms contract with managers. The question of whether high-pow- ered incentives crowd out managers’ preferences for honesty has not been studied. Human resource policies that build organizational commitment could weaken the need for using incentive compensation for goal congruence. There is an entire array of behavioral drivers of managerial motives for misreporting such as trust, reciprocity, gift-exchange, social distance, identity, other-regarding behaviors, and so on, which can be incorporated into optimal contract design. The competitive and regulatory landscape in which companies operate is constantly changing, which places pressures on even the best designed MCS. In addition, organi- zations across the world are exploring new types of performance measurement and con- trol systems, information systems, and stretching the boundaries of their firms by collaborative contracting with suppliers and customers. These changes create challenges in designing appropriate contracting and performance measurement systems. Research could explore the implications of these changes using data from public or proprietary sources. This is an important topic because inappropriate design of MCS can have serious implications for firms and managers as well as for policy. Acknowledgements This paper was the based on a talk delivered at the annual conference of the China Journal of Accounting Studies in Dalian in May 2013. I thank Jason Xiao, Satish Joshi, and two anonymous reviewers for their helpful comments. China Journal of Accounting Studies 11 Note 1. See Sprinkle and Williamson (2007) for a comprehensive review of the decision facilitating and decision influencing roles of management accounting information and their implications for experimental research in accounting. References Antle, R., & Eppen, G.D. (1985). Capital rationing and organizational slack in capital Budgeting. Management Science, 31, 163–174. Antle, R., & Fellingham, J. (1995). Information rents and preferences among information systems in a model of resource allocation. Journal of Accounting Research, 33,41–58. Ashton, R.H. (2005). Intellectual capital and value creation: A review. Journal of Accounting Literature, 24, 53. Baiman, S. (1990). Agency research in managerial accounting: A second look. Accounting, Organizations and Society, 15(4), 341–372. Baiman, S., & Evans III, J.H. (1983). Pre-decision information and participative management control systems. Journal of Accounting Research, 22, 371–395. Balakrishnan, R., & Sivaramakrishnan, K. (2002). A critical overview of the use of full-cost data for planning and pricing. Journal of Management Accounting Research, 14,3–31. Banker, R., & Datar, S. (1989). Sensitivity, precision and linear aggregation of signals for perfor- mance evaluation. Journal of Accounting Research, 27,21–39. Brüggen, A., Krishnan, R., & Sedatole, K. (2011). Drivers and consequences of short-term production decisions: Evidence from the auto industry. Contemporary Accounting Research, 28,83–123. Cooper, R., & Kaplan, R.S. (1992). Activity-based systems: measuring the costs of resource usage. Accounting Horizons, 6(3), 1–13. Covaleski, M., Evans III, J.H., Luft, J., & Shields, M.D. (2003). Budgeting research: three theo- retical perspectives for selective integration. Journal of Management Accounting Research, 15,3–49. Demski, J., & Feltham, G. (1976). Cost determination: A conceptual approach. Ames, IA: Iowa State Press. Dickhaut, J., & Lere, J.C. (1983). Comparison of accounting systems and heuristics in selecting economic optima. Journal of Accounting Research, 27(2), 495–513. Dunk, A.S. (1993). The effect of budget emphasis and information asymmetry on the relation between budgetary participation and slack. The Accounting Review, 68(2), 400–410. Evans III, J.H., Hannan, R.L., Krishnan, R., & Moser, D. (2001). Honesty in managerial reporting. The Accounting Review, 76, 537–559. Fama, E., & Jensen, M. (1983). Agency problems and residual claims. Journal of Law and Economics, 26, 301–325. Feltham, G.A., & Xie, J. (1994). Performance measure congruity and diversity in multi-task prin- cipal/agent relations. The Accounting Review, 69, 429–453. Gupta, M., Pevzner, M., & Seethamraju, C. (2008). The implications of absorption cost account- ing and production decisions for firms’ future performance and valuation. Working paper. Healy, P. (1985). The effect of bonus schemes on accounting decisions. Journal of Accounting and Economics, 7,85–107. Holmström, B., & Milgrom, P. (1991). Multitask principal-agent analyses: Incentive contracts, asset ownership and job design. Journal of Law, Economics & Organization, 7,24–52. Holthausen, R.W., Larcker, D.F., & Sloan, R.G. (1995). Annual bonus schemes and the manipulation of earnings. Journal of Accounting and Economics, 19,29–74. Jensen, M.C., & Meckling, W.H. (1976). Theory of the firm: Managerial behavior, agency costs, and ownership structure. Journal of Financial Economics, 4, 469–506. Joshi, S., Krishnan, R., & Lave, L. (2001). Estimating the hidden costs of environmental regulations. The Accounting Review, 76, 171–198. Krishnan, R., Luft, J., & Shields, M.D. (2005). Effects of accounting-method choices on subjective performance-measure weighting decisions: experimental evidence on precision and error covariance. The Accounting Review, 80, 1163–1192. 12 Krishnan Lev, B. (2001). Intangibles: Management, measurement, and reporting. Washington, DC: Brookings Institution Press. Luft, J., & Shields, M.D. (2009). Psychology models of management accounting. Foundations and Trends in Accounting, 4, 199–345. Macintosh, N.B., & Quattrone, P. (2010). Management accounting and control systems: An organizational and sociological approach. UK: Wiley. Roychowdhury, S. (2006). Earnings management through real activities manipulation. Journal of Accounting and Economics, 42, 335–370. Schiff, M., & Lewin, A.Y. (1970). The impact of people on budgets. The Accounting Review, 45 (2), 259–268. Sprinkle, G.B., & Williamson, M.G. (2007). Experimental research in managerial accounting. In C. Chapman, A. Hopwood & M. Shields (Eds.) Handbook of management accounting research. New York: Elsevier. Turner, M.J., & Hilton, M.G. (1989). Use of accounting product-costing systems in making production decisions. Journal of Accounting Research, 27, 297–312. Vera-Munoz, S. (1998). The effects of accounting knowledge and context on the omission of opportunity costs in resource allocation decisions. The Accounting Review, 73,47–72. White, A.L., Savage, D.E., Brody, J., Cavander, D., & Lach, L. (1995). Environmental cost accounting for capital budgeting: A benchmark survey of management accountants. US Environmental Protection Agency, Report no. EPA742-R-95-005. Young, S.M. (1985). Participative budgeting: The effects of risk aversion and asymmetric information on budgetary slack. Journal of Accounting Research, 23(2), 829–842. Zimmerman, J. (2009). Accounting for decision making and control. New York: McGraw-Hill

Journal

China Journal of Accounting Studies – Taylor & Francis

Published: Jan 2, 2014

Keywords: management accounting system design; error; bias

There are no references for this article.