«Jan Bouwens*, Laurence van Lent* Abstract Using a sample of 140 managers, we investigate the use of various performance metrics in determining the ...»
Assessing the performance of business unit managers
Jan Bouwens*, Laurence van Lent*
Using a sample of 140 managers, we investigate the use of various performance metrics in
determining the periodic assessment, bonus decisions, and career paths of business unit
managers. We show that the weight on accounting return measures is associated with the
authority of these managers, and we document that both disaggregated measures (expenses
and revenues), and non-financial measures play a greater role as interdependencies between business units increase. The results suggest separate and distinct roles for different types of performance measures. Accounting return measures are used to create the proper incentives for managers with greater authority, while disaggregated and non-financial measures are employed in response to interdependencies.
JEL classification: M41 Keywords: performance measures, business units, managerial performance * Tilburg University. We appreciate comments from Margaret Abernethy, Eddy Cardinaels, Henri Dekker, Peter Easton, Chris Ittner, Ken Merchant, Valeri Nikolaev, Mark Penno, Frank Selto, Steve Salterio, and Ann Vanstraelen, and also from workshop participants at the Vrije Universiteit Amsterdam, University of Antwerpen, University of Melbourne, Tilburg University, the 2004 Global Management Accounting Research Symposium, the 2005 Management Accounting Section mid-year meeting in Scottsdale, Arizona, and the 2005 European Accounting Association annual congress in Göteborg. We thank Merle Erickson (Editor) and an anonymous reviewer for their constructive feedback.
1. Introduction To encourage rapid and relevant decision making, firms move towards organizational designs in which authority over operations and strategy is assigned to business unit managers (Aghion and Tirole 1997; Roberts 2004). However, empowering business unit managers is also likely to undermine cooperation among managers. We conjecture that different performance measures play different and distinct roles in evaluating the performance of business unit managers according to the organizational design of the firm. More specifically, we hypothesize that the use of accounting return measures provides incentives to managers to use their authority appropriately and that the use of disaggregated accounting measures, such as costs and revenues, as well as various non-financial measures, mitigates the individual manager’s tendency to attach too little value to the impact of his or her decisions on other parts of the firm (Baiman and Baldenius 2006). To test these predictions, we use survey data on 140 business unit managers and assess the use of a range of measures for evaluating the performance of business unit managers.
Accounting return measures are more aggregated than profit measures, as they relate profit to resources employed (e.g., return on investment, residual income, return on capital).
Profits, in turn, are more aggregated than expenses and revenues. Profits are thus an intermediate category between accounting return measures and fully disaggregated measures such as expenses and revenues. Following previous theoretical research, we assume that accounting return measures are the best choice for assessing the performance of a business unit manager and are more likely to be used when more authority is vested in the business unit manager. To the extent that return and profit measures substitute for each other, however, we will see shifts from returns to profits as the level of authority declines, as well as from more disaggregated measures to profits as the level of authority increases. We suggest that the primary function of disaggregated and non-financial measures is to reduce the noise in aggregated financial measures (accounting returns, profits) or to provide a signal about a manager’s actions. We therefore posit that these measures are used when aggregated
efforts of managers.
We document that accounting return and profit measures are associated with the level of authority of the business unit manager (relative to that of his or her superior) with regard to key decisions. In contrast, the use of disaggregated and non-financial measures, is associated with interdependencies among business units within the firm, i.e., the extent to which the decisions of one business unit manager affect other managers in the firm and vice versa.
This study builds on and extends early work by Scapens and Sale (1985) who seek but fail to establish an association between delegated authority and the use of accounting return metrics. Since this early work, relatively little progress has been made in identifying the circumstances under which firms use various types of performance measures, 1 perhaps because the classifications that researchers use are often too general (e.g., financial vs. nonfinancial or market vs. accounting measures). Thus, by analyzing four different types of performance measures — accounting return, profit, disaggregated, and non-financial measures — we contribute to the literature by distinguishing the differential effects of organizational design on the use of each metric type.
Ittner and Larcker (2001) warn that the use of performance measures is likely to vary with the decision context in which they are applied. To accommodate this possibility, we provide exploratory evidence on the use of different types of performance measures in three decision contexts: (1) the periodic assessment of business unit managers, (2) the determination of managers’ bonuses, and (3) the determination of managers’ long-term career paths. The literature is somewhat critical about the use of survey data in accounting research (Young 1996; Ittner and Larcker 2001; Zimmerman 2001). In performing our tests, we heed the warnings in the literature about potential measurement error, “leading” or “soft” questions, inadequate sampling, and inappropriate econometric techniques.
1 Several authors (Gordon and Narayanan 1984; Chenhall and Morris 1986; Gul and Chia 1994; Scott and Tiessen 1999) investigate the association between organizational design and “broad control system” issues (Ittner and Larcker 2001).
business unit managers enjoy more decision-making authority is important because some authors stress that the use of performance measures is independent of a manager’s authority (Solomons 1965), whereas others maintain that profit and return measures are only meaningful if a manager has significant decision-making power (Vancil 1978). These views are often expressed when discussing the use of responsibility centers (emphasis added in the
“... A common misconception is that the term profit center (and in some cases investment center) is a synonym for a decentralized subunit... [M]anagers in a division organized as a profit center may have little leeway in making decisions …” (Horngren, Foster and Datar 1994, pg.863) “Responsibility accounting is a term used to describe the use of accounting data in managerial evaluation. … Authority and responsibility are distinct. For example, the manager of a fast food facility is usually evaluated as a profit center. Yet the menu and prices, not to mention cooking procedures and ingredients, are determined by central management. The manager has little authority but considerable responsibility.” (Demski 1994, pg. 538) and, in contrast, “... one of the major purposes behind the use of a profit center [is] to encourage local decision making and initiative. … A profit center is a unit for which the manager has the authority to make decisions on sources of supply and choice of markets...” (Kaplan and Atkinson 1989, pg. 590) Our evidence takes issue with both sides of the debate. First, we show that the use of performance measures (and thus the designation of responsibility centers) is strongly correlated with the authority of business unit managers. At the same time, we believe that requiring investment or profit centers to be decentralized units by definition obfuscates the substantial observed variety in the authority of managers who are evaluated on accounting returns and/or profits. It also prevents researchers from investigating the mapping between performance measure and authority.
We find no evidence that the use of accounting return measures is affected by the presence of interdependencies, nor is the use of disaggregated and non-financial measures associated with the level of authority. Instead, firms use aggregated measures relatively more than disaggregated measures when more authority is delegated to business unit managers, and
short, performance measures have very distinct roles: whereas the use of accounting return measures is correlated with the authority of business unit managers, disaggregated and nonfinancial measures are employed to provide incentives for managers to cooperate in the presence of interdependencies between business units.
2. Hypothesis development
2.1. Authority of business unit managers Performance evaluation measures should fulfill two purposes: (1) to give managers incentives to use their authority optimally, and (2) to “disaggregate” the firm’s total economic performance into a summary estimate of each manager’s contribution to firm value (Zimmerman 1997; Raith 2005). Accounting return measures, designed to capture the economic value generated from specified resources (Scapens 1979; Anthony and Govindarajan 2004), should accomplish both purposes. However, the normative literature warns against using accounting return measures when the authority of managers does not extend to decisions about the acquisition of new resources or the disposal of existing resources: 2 “Both the measure of performance and the standard against which it is compared must reflect the degree of control that the division manager can exert on profit and investment.” (Shillinglaw 1959, page 46) Notwithstanding this warning, recent theoretical work concludes that accounting return measures provide not only information about how well resources are used, but also incentives that encourage optimal resource use (Reichelstein 1997; Rogerson 1997;
Zimmerman 1997; Dutta and Reichelstein 1999). Even if managers have little authority over investment decisions, theory suggests that accounting return measures are informative (Holmstrom 1979; Indjejikian 1999) about the way in which managers wield their authority.
2 The normative literature traditionally invokes the “controllability principle” to argue that managers should only be charged for the resources over which they have control. The controllability principle has been somewhat discredited in more recent theoretical (Indjejikian 1999) and empirical (Merchant
they substantially affect the use of current resources. Accounting return measures should therefore provide valuable information about their performance.
Extant empirical research supports textbook prescriptions calling for performance measures that capture the decisions and actions of managers (Chenhall and Morris 1986).
Whereas earlier empirical work does not specify the type of metric that fulfils this purpose, we propose that accounting return measures summarize the performance of managers whose authority impacts the value of the resources they use (Scapens and Sale 1985, pg. 240), and we therefore expect a positive association between the weight on accounting return measures and the authority of business unit managers.
H1: The weight on accounting return measures in the evaluation of business unit managers is positively associated with their decision-making authority.
We also want to know whether an increased weight on accounting returns implies a reduction in the weight on disaggregated and non-financial measures. Agency models suggest that the relative importance of a performance measure depends on the “signal” provided by the measure about an employee’s action and the “noise” with which the actions are captured (Lambert and Larcker 1987; Banker and Datar 1989; Indjejikian 1999). Several authors have noted the lack of causal relation between aggregated financial measures and the actions of managers with low authority, and advocate the use of disaggregated and non-financial measures as being more directly under these managers’ control (Wruck and Jensen 1994;
Ittner and Larcker 1998b; Ittner and Larcker 2003). Disaggregated and non-financial measures are less subject to exogenous events and therefore contain less noise. In contrast, accounting return measures are expected to be too coarse to provide desirable signals and lack precision when managers have little authority (Horngren 2004). This “signal-to-noise” argument suggests a negative association between authority and the use of disaggregated and non-financial measures.
3 Managers are said to have full investment authority if they can acquire or dispose fixed assets without prior approval from their superior.