University of Southern California Dissertations and Theses

The balanced scorecard and long-term financial performance: evidence from publicly traded companies

(USC Thesis Other)

The balanced scorecard and long-term financial performance: evidence from publicly traded companies

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content ! ! "#$!%&'&()$*!+),-$)&-*!&(*!',(./"$-0!12(&()2&'!3$-1,-0&()$4! ! $52*$()$!1-,0!36%'2)'7!"-&*$*!),03&(2$+! ! ! ! ! 89! ! ! ! ! &:;<;=!&>>!7=><9! ! ! ! ! ! ! ! ! ! ! &!*;??@AB=B;C>!3A@?@>B@D!BC!BE@!! 1&)6'"7!,1!"#$!.-&*6&"$!+)#,,'! 6(25$-+2"7!,1!+,6"#$-(!)&'21,-(2&! ;>!3=AB;=:!1F:G;::H@>B!CG!BE@! -@IF;A@H@>B?!GCA!BE@!*@JA@@! *,)",-!,1!3#2',+,3#7! K%6+2($++!&*02(2+"-&"2,(L! ! ! ! ! ! 0=9!MNOP! ! ! ! ! ! ! )CQ9A;JEB!MNOP! ! ! ! ! ! ! ! &:;<;=!&>>!7=><9! ! i! CONTENTS 1. Introduction..........................................................................................................................1 1.1 Background and Study Purpose..................................................................................1 1.2 Outline of the Study....................................................................................................7 2. Literature Review.................................................................................................................9 2.1 Overview.....................................................................................................................9 2.2 Psychology-based studies that investigate the impact of BSC on managerial judgment and decision-making.................................................................................11 2.3 Case studies or field studies describing the implementation of BSC in a single firm............................................................................................................................17 2.4 Studies that investigate how and for what purpose BSC is used within organizations.............................................................................................................19 2.5 Commentary or studies testing the cause-effect relationship in BSC models ..........21 2.6 Studies investigating the use of BSC on organizational performance......................27 3. Hypothesis Development...................................................................................................36 3.1 Characteristics of Firms that Use BSC .....................................................................36 3.1.1 Intangible Assets............................................................................................36 3.1.2 Strategy ..........................................................................................................37 3.1.3 Recent Financial Performance .......................................................................40 3.1.4 Size.................................................................................................................41 3.1.5 Decentralization .............................................................................................42 3.2 Financial Performance Impact of BSC Use..............................................................44 4. Sample Selection and Research Design.............................................................................49 4.1 Sample Selection.......................................................................................................49 4.1.1 Balanced Scorecard Users..............................................................................49 4.1.2 Control Sample ..............................................................................................51 4.2 Research Design .......................................................................................................53 4.2.1 Prediction Model for Likelihood to Adopt BSC ...........................................53 4.2.2 Univariate Tests of Performance Differences Between BSC-User and Non-User Firms ............................................................................................54 4.2.3 Difference-in-Difference Model to Test the Performance Impact of Use ................................................................................................................55 4.2.4 OLS Model to Test the Effect of BSC-Mismatch on Financial Performance ...................................................................................................56 5. Results................................................................................................................................58 5.1 Prediction Model for Likelihood to Adopt BSC ......................................................58 5.2 Descriptive Statistics – BSC-User vs. Non-User Firms ...........................................60 5.3 Effect of BSC Use on Financial Performance .........................................................61 5.3.1 Univariate Tests ............................................................................................61 5.3.2 Impact of BSC Adoption on Financial Performance .....................................62 5.3.3 Impact of BSC Mismatch on Financial Performance ...................................63 ! ii! 5.4 Additional Tests 5.4.1 Matching on Variables from Prior Literature ...............................................65 5.4.2 Hall of Fame Matched Firms - Only .............................................................67 6. Conclusion ........................................................................................................................69 6.1 Contributions and Limitations .................................................................................71 6.2 Future Directions for Research .................................................................................72 References ......................................................................................................................................73 ! iii! LIST OF TABLES Table 1. Distribution of BSC-User Firms by Identification Source............................................81 Table 2. Logit Models Predicting the Adoption of BSC .............................................................82 Table 3. Covariate Balance – Propensity Score Match ...............................................................83 Table 4. Descriptive Statistics .....................................................................................................84 Table 5. Univariate Tests.............................................................................................................85 Table 6. Effect of BSC Use on Financial Performance – All Firms Dataset ..............................86 Table 7. Effect of BSC Use on Financial Performance – w/o Financials Dataset ......................87 Table 8. Effect of BSC Use on Financial Performance – SEC Identified Dataset......................88 Table 9. Effect of BSC Use on Financial Performance – w/o Hall of Fame Dataset..................89 Table 10. Effect of BSC Mismatch on Financial Performance – All Firms Dataset.....................90 Table 11. Effect of BSC Mismatch on Financial Performance – w/o Financials Dataset.............91 Table 12. Effect of BSC Mismatch on Financial Performance – SEC Identified Dataset ............92 Table 13. Effect of BSC Mismatch on Financial Performance – w/o Hall of Fame Dataset........93 Table 14. Covariate Balance – Single Variable Match .................................................................94 Table 15. Effect of BSC Use on Financial Performance – ROA Matched Dataset ......................95 Table 16. Effect of BSC Use on Financial Performance – Book-to-Market Matched Dataset.....96 Table 17. Effect of BSC Use on Financial Performance – Market Value of Equity Matched Dataset...........................................................................................................................97 Table 18. Effect of BSC Use on Financial Performance – Total Assets Matched Dataset...........98 Table 19. Effect of BSC Mismatch on Financial Performance – ROA Matched Dataset.............99 Table 20. Effect of BSC Mismatch on Financial Performance – Book-to-Market Matched Dataset.........................................................................................................................100 Table 21. Effect of BSC Mismatch on Financial Performance – Market Value of Equity Matched Dataset ..........................................................................................................101 Table 22. Effect of BSC Mismatch on Financial Performance – Total Assets Matched Dataset.........................................................................................................................102 Table 23. Logit Model Predicting the Adoption of BSC – Hall of Fame Firms .........................103 Table 24. Effect of BSC Use on Financial Performance – Hall of Fame Dataset.......................104 ! iv! ABSTRACT This study examines the effect of Balanced Scorecard (BSC) usage on financial performance using accounting-based and market-based performance measures. A logit model predicting the likelihood that a firm will adopt or use BSC is proposed. Firms that have been identified as BSC users are propensity-score matched to a set of control firms using the fitted- values from the logit model. Univariate tests and multiple regression are used to examine the relationship between BSC use and ROA, ROE and one-year buy-and-hold stock returns. The findings of this study suggest that the use of BSC is not positively associated with financial performance. Findings also suggest that deviation from investing in BSC up to an “optimal” level, may be associated with significantly lower financial performance. However, subsequent analysis using a subset of BSC-User firms (Hall of Fame firms) that have received an award for successful implementation of BSC provides strong evidence in support of the hypothesis that the use of BSC is positively associated with future financial performance. An important finding from this subset of data is that performance effects may not be seen for 3 or 4 years after BSC adoption. The results demonstrate a need for managers to ensure that they are investing the necessary amounts of financial and human capital resources to ensure successful implementation of BSC. The results also indicate a need for managers and researchers to allow sufficient time to elapse before assessing or evaluating the success or failure of a BSC implementation. ! 1! CHAPTER 1 INTRODUCTION 1.1 Background and Study Purpose Originally introduced by Harvard Business School Professor Robert Kaplan and business consultant David Norton, the Balanced Scorecard, hereafter BSC, is an excellent example of how business practices can inform academic research and vice versa. Developed during the course of a yearlong research project involving 12 firms considered to be on the cutting-edge of performance measurement, BSC was initially conceived of as a single performance measurement tool that has since evolved in such a way that it is now considered to be a major component of a firm’s overall strategic management system, in which management processes and systems are aligned with organizational strategy (Kaplan and Norton, 1996, 2001a, 2001b). Traditionally, organizational performance was measured using financial metrics such as return-on-assets, return-on-capital-employed, or return-on-equity. As technology advanced and society moved into the “Information Age,” many organizations’ most valuable assets, including human capital and investments in research and development were not captured on the balance sheet. As a result, the traditional financial measures have become less meaningful. Something more is needed to provide managers with indicators or signals of what performance will be prospectively. Financial measures are always backwards-looking and do not provide any indication as to what a firm should do to sustain or improve performance going forward. The notion that financial performance measures alone are not sufficient to evaluate organizational effectiveness is not new. Using non-financial measures to supplement financial ! 2! performance measures was recommended as early as 1971, in the American Accounting Association’s (AAA) “Report of the Committee on Non-Financial Measures of Effectiveness.” The Committee states, “While long-run profits are probably the primary objective of most business firms, the contribution of the firm’s current activities to long-run profits may be inadequately measured by traditional financial statements.” Suggested non-financial performance measures include measures of productivity, employee and customer satisfaction, marketing effectiveness, and the “effectiveness of fulfilling the firm’s social role” (i.e. corporate social responsibility.) 1 Financial performance measures are always lagging indicators of performance outcomes related to past decisions and choices. The basic premise of BSC is that managers should consider non-financial performance measures in addition to financial measures in order to evaluate, predict, and improve future organizational performance. However, these performance measures should not be an ad hoc collection of unrelated performance metrics. The non-financial performance measures included in the BSC should include some indication of past performance as well as some forward-looking, measures i.e. leading indicators of future financial performance that capture the strategic direction and initiatives of the organization. For example, results from a customer satisfaction survey measure an organization’s historical performance; whereas a measure of the time it takes a company to introduce a new product into the marketplace can be predictive of future financial performance. Performance measures in the BSC are grouped into four categories, called “perspectives”: 1. Customer perspective – reflects how the customer views the organization and includes such measures as customer satisfaction and market share !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! " !The Committee recommends non-financial performance measures that would be of interest to investors, employees and labor unions, suppliers, and consumers in addition to measures that would be of interest to internal management (and likely included in a BSC.) ! 3! 2. Innovation and Learning perspective – reflects how employees view the organization and includes measures such as employee turnover/employee retention and employee training 3. Internal Business Processes perspective – reflects what the organization must improve upon to be successful and includes measures such as cost reduction, reduction in manufacturing time, and other efficiency improvements 4. Financial perspective – reflects changes in financial performance and may include such measures as growth in sales, growth in net income, and return on assets According to the BSC creators, performance measures in each perspective should be derived from the organization’s strategy and theoretically should be related to each other in a cause-and-effect manner. The cause-effect relationships should pervade across all four BSC perspectives. Thus, positive performance on learning and growth perspective measures should lead to positive performance on internal business process measures, which should then lead to positive performance on customer perspective measures, ultimately leading to improved financial outcomes. It is this required link to strategy and the assumed cause-effect relationship that distinguishes BSC from a simple collection of non-financial and financial performance measures and from other performance measurement/management tools such as the French Tableau de Bord. 2 The Balanced Scorecard is said to be one of management accounting’s most important innovations, in addition to other innovations, including activity-based costing (ABC), activity- based management (ABM), life-cycle costing (LCC), target costing (TC), and total quality !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 2 Kaplan & Norton (1996) acknowledge that initially this cause-effect relationship must be established subjectively and qualitatively. After a period of time, when enough data and evidence are accumulated, empirical testing can be done to validate the cause-effect relationships. However, Huelsbeck, et al. (2011) #$%&!'()'!*+*%!$%!)!,$%-.*! /0,$%*,,!0%$'!#$12!3$'(!)%!*,')/.$,(*&!,'1)'*-4!)%&!.5%-!6*1#512)%7*!($,'5148!'(*!)/$.$'4!'5!90)%'$#4!)%&! +).$&)'*!'(*!7)0,).!.$%:)-*,!25&*.!1*2)$%,!*.0,$+*; ! 4! management (TQM), (Ittner and Larcker, 1998a, Bjoornenak and Olson, 1999). The BSC framework is taught in nearly every business school curriculum at both the undergraduate and graduate levels, and anecdotal evidence suggests that since it’s introduction in 1992, nearly 60% of Fortune 1000 and global firms including publicly traded for-profit corporations and privately held companies, as well as not-for-profit companies, government agencies, educational institutions, and municipalities, have utilized balanced scorecard methodology incorporating strategy-derived financial and non-financial performance measures into their performance evaluation systems (Silk, 1998; Beimann and Johnson, 2007; Rigby and Bilodeau, 2009). Over the last two decades, Kaplan and Norton and others have published multiple books describing the evolution of BSC from simple performance measurement tool to comprehensive strategy management system and outlining best-practices for successful BSC design and implementation (Olve, Roy and Wetter, 1999; Niven, 2002; Lawson, Desroches and Hatch, 2008). In addition, the BSC continues to be a major topic of interest in both the academic and practitioner literature. A keyword search of the subject “balanced scorecard” returns over 1,000 articles in scholarly journals and nearly twice as many articles when magazines, trade publications, newspapers, working papers and dissertations are included. While plenty has been written about the Balanced Scorecard, the relevance of these publications for the advancement of management accounting practice and scholarship is called to question. In fact, Salterio (2012) argues that behavioral management accounting researchers’ focus on the “bad news” of BSC – the finding that consistent with prior research in psychology, BSC users will rely more on performance measures that are common across evaluation units than on measures that are unique to the unit 3 - presents a missed opportunity for accounting researchers to demonstrate “the advantages of using well-grounded psychology research in management accounting (both in !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! < !This finding is commonly referred to as the “common measures” bias. ! 5! research on management accounting as well as in the design of management accounting practices)...” Individual BSC success stories abound. The academic literature and popular press has documented several successful BSC implementations, most notably early adopters CIGNA Property & Casualty and Mobil U.S. Marketing and Refining. In addition, the Palladium Group inducts several organizations each year into the Balanced Scorecard Hall of Fame 4 . Inducted organizations include public and private corporations as well as government and non-profit entities. Furthermore, over the past two decades several BSC consulting firms have emerged, including the Balanced Scorecard Institute and Palladium Group, which was co-founded by BSC creators Drs. Kaplan and Norton. All of this evidence suggests that organizations do derive some value from the adoption and use of BSC. However, there are also indications that BSC may not be as effective as its’ proponents suggest. Some question the effectiveness of BSC adoption and argue that its’ usage may be obsolete or out-of-date (Udpa, 1997; Mooraj, Oyon and Hostettler, 1999; Schneiderman, 1999; Davis and Albright, 2004; Neely, 2007; DeGeuser, Mooraj and Oyon, 2009; Richardson, 2011). While behavioral management accounting researchers have focused primarily on the common measures bias and how to “correct” it with increased motivation, incentives, and additional knowledge about BSC (Banker, Chang and Pizzini, 2004; Libby, Salterio and Webb, !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! = !Organizations selected for the BSC Hall of Fame must meet six criteria for selection (BSC Hall of Fame Report, 2004, p. 4): 1. Implement BSC as defined by the Kaplan/Norton methodology 2. Exemplify the 5 principles of the Strategy-Focused Organization 3. Earn media recognition their BSC implementation 4. Present their case at a public conference 5. Achieve significant financial or market share gains (private and publicly traded companies) or demonstrate the measurable achievement of mission or customer objectives (public, nonprofit, or shared services organizations) 6. Provide a testimonial from a senior executive linking the BSC to the organization’s results ! 6! 2004; Roberts, Albright and Hibbets, 2004, Dilla and Steinbart, 2005), very little empirical research has investigated the financial performance impact of BSC use (Davis and Albright, 2004; Neely, 2007; De Geueser, Mooraj and Oyon, 2009). The few empirical studies that have been conducted provide limited and inconclusive evidence about the effectiveness of BSC at achieving the ultimate goal of improving financial performance (Lingle and Schiemann, 1996; Hoque and James, 2000; Ittner, Larcker and Randall, 2003b; DeBusk and Crabtree, 2006). Furthermore, we know almost nothing about the characteristics of firms that are likely to use BSC. A key premise of the Balanced Scorecard methodology is that performance measures should be derived from the organization’s strategy, and prior research suggests that a firm’s strategy impacts its’ choice of performance measurement systems and also impacts organizational performance (Snow and Hrebiniak, 1980; Govindarajan and Gupta, 1985). And although organizational strategy is a critical element of the BSC, the question of what type of organizational strategy is best suited for or may benefit most from BSC use remains largely unexplored. Furthermore, other firm-specific characteristics in addition to organizational strategy may determine if a firm will or will not use BSC. These factors include how much of the firm’s value creating assets are not included on the balance sheet, or the amount of the firm’s intangible assets, firm size and the amount of divisionalization or decentralization. A firm’s recent history of poor financial performance may also precipitate the adoption of a new performance measurement system, such as BSC. Twenty years later, it appears that some form of BSC is still used by many organizations, and according to the business press and BSC consultancies, new balanced scorecard-using organizations continue to emerge. However, the significance of the BSC and its’ effectiveness at ! 7! improving future financial performance are still questioned (Richardson, 2011). Chenhall (2006) states, “…as [the] practices become more widely adopted, evidence becomes available to test for universal effects.” Thus, if the balanced scorecard is in fact the strategic management tool that it was designed to be – aligning people, practices and business activities to sustain competitive advantage and improve stakeholder value – and not simply a novel innovation or management fad that has allowed numerous consultants to generate substantial revenues from design and implementation engagements, then the impact of its adoption should stand up to rigorous empirical testing. “Successful” implementation of BSC should result in measurable improvement in an organization’s financial performance over time. In this study I address two primary research questions: First, what are the characteristics, including realized organizational strategy, of firms that are likely to use BSC? Second, what are the long-term financial performance consequences of using BSC? I develop a prediction model of a firm’s likelihood to use BSC that includes measures of firms’ intangible assets, strategy, strained financial condition, size and amount of decentralization as determinants. The fitted values from the prediction model are used to match BSC-user firms to a set of non-user (control) firms. I examine the impact of BSC use on future financial performance using both accounting and market measures of financial performance and compare the performance of BSC users to non-users. 1.2 Outline of the Study The remainder of the study proceeds as follows. Chapter 2 provides a review of relevant BSC literature and develops the research hypotheses. Chapters 3 and 4 discuss hypothesis development and research design. Results, including the discussion of additional tests and ! 8! sensitivity analyses are presented in Chapter 5. Chapter 6 concludes with contributions, limitations, and possible extensions for future research. 9! CHAPTER 2 LITERATURE REVIEW 2.1 Overview The seminal article on the Balanced Scorecard (BSC) is Kaplan and Norton’s (1992) Harvard Business Review article entitled “The Balanced Scorecard – Measures that Drive Performance.” The original intent of BSC was to complement traditional financial performance measures with non-financial operational measures when evaluating organizational performance. The non-financial measures would be leading indicators of future financial performance, whereas financial measures are always backward-looking indicators reflecting the outcomes of past decisions. Performance measures in a BSC are grouped into four categories or “perspectives”: 1. Customer perspective – reflects how the customer views the organization and includes such measures as customer satisfaction and market share 2. Innovation and Learning perspective – reflects how employees view the organization and includes measures such as employee turnover/employee retention and employee training 3. Internal Business Processes perspective – reflects what the organization must improve upon to be successful and includes measures such as cost reduction, reduction in manufacturing time, and other efficiency improvements 4. Financial perspective – reflects changes in financial performance and may include such measures as growth in sales, growth in net income, and return on assets 10! Originally conceived of as a stand-alone performance measurement tool, the scorecard was positioned in 1996 as an integral part of a larger strategic management system (Kaplan and Norton, 1996) where performance measures in each perspective are derived from the organization’s strategy. Over time, the process of developing and implementing BSC further evolved, and Kaplan and Norton (2001a, 2001b) subsequently show how organizations should align management processes and systems to strategy. An organization using this third-generation BSC should start by articulating or defining its’ strategy, then translate the strategy into specific objectives through the use of a strategy map. 5 The strategy map provides a visual representation of the critical components and cause-effect linkages required for an organization to achieve its’ strategic goals and create long-term value. To achieve “breakthrough success” organizations must focus on five key principles 6 : 1. Translate the strategy to operational terms – present the strategy to employees in common language that everyone will understand 2. Align the organization to the strategy – eliminate functional silos and take advantage of organizational synergies 3. Make strategy everyone’s everyday job – provide top-down communication about strategy but allow lower level employees to develop their own performance objectives derived from the strategy; incentive compensation should be linked to scorecard performance !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! > !A first-generation BSC is a collection of financial and non-financial performance measures based on Kaplan and Norton’s (1992) introduction of BSC as a performance measurement tool. Second-generations scorecards derive performance measures from strategy in a hypothesized cause-effect manner based on Kaplan and Norton’s (1996) positioning of BSC as a strategy management system. Third-generation scorecards incorporate the use of strategy maps (Kaplan and Norton, 2001). ? !Kaplan and Norton (2001b) refer to these as the “Five Principles of the Strategy-Focused Organization.” 11! 4. Make strategy a continual process – provide regular opportunities to evaluate performance and to give and receive feedback, including adapting/revising the strategy 5. Mobilize leadership for change – actively engage the executive team to communicate to the entire organization the need for change and embed the new strategic management system into the organization’s culture. A keyword search of “balanced scorecard” returns over 1,000 articles in scholarly journals and nearly twice as many articles when magazines, trade publications, newspapers, working papers and dissertations are included. Subject areas range from engineering to information systems to medicine and more. The BSC literature that is relevant to management accounting research can be grouped into five broad categories: 1. Psychology-based studies that investigate the impact of BSC use on managerial judgment and decision-making 2. Case studies or field studies describing the implementation of BSC in a single firm 3. Studies that investigate how and for what purpose BSC is used within organizations 4. Commentary or studies testing the cause-effect relationship in BSC models 5. Studies investigating the use of BSC on organizational performance 2.2 Psychology-based studies that investigate the impact of BSC on managerial judgment and decision-making 12! Studies investigating the impact of BSC use on judgment and decision-making use several well-documented theories from cognitive and organizational psychology. Cognitive psychology theory posits that human information processing capacity is limited to seven, plus or minus two, chunks of information (Miller, 1956). Organizational psychology theories posit that when making comparative judgments, evaluators will place more weight on common information than unique information (Slovic and MacPhillamy, 1974) and that objective, outcome-based information is relied upon more heavily than subjective information in evaluation decisions (Baron and Hershey, 1988). Lipe and Salterio (2000) ask participants to evaluate the performance of two divisions (RadWear and PlusWear) of a fictitious women’s clothing store using a scorecard with 16 performance measures. Some performance measures are unique to each division and others are common to both. Results of the experiment show that consistent with psychology theories, common measures affect performance evaluation, but unique measures do not, a result that has come to be known in the management accounting literature as the “common measures bias.” Using the same case involving two divisions of a women’s clothing store, Lipe and Salterio (2002) conduct two experiments to investigate whether the organization of performance measures in the balanced scorecard affects judgments about the division’s performance. In the first experiment, some participants are given performance measures in a BSC format, while other participants are provided an ungrouped list of the same performance measures. Performance in the financial perspective is slightly above target and is on target in the internal business processes and learning and growth perspectives. Performance in the customer perspective is manipulated to be significantly above or below target. With the information provided, participants are asked to evaluate the performance of each division. When performance measures 13! are grouped into four BSC categories, the difference in evaluation between the two division managers is considerably less extreme than when the measures are presented in the ungrouped list. In contrast to the first experiment’s results, when significantly positive or negative performance measures are spread across multiple perspectives, instead of being grouped in one (customer) perspective, the format or presentation of performance measures (in BSC categories or an ungrouped list) does not lead to significant differences in performance evaluation. Ittner, Larcker and Meyer (2003a) examine the use of a BSC-based incentive compensation plan at a North American retail bank. The bank’s previous compensation plan was based primarily on financial performance. After a significant downturn in performance, management switched to the BSC-based plan that included measures for quality, risk management, customer satisfaction, and other non-financial measures of performance, in addition to financial measures. The bank’s scorecard includes six perspectives: financial, strategy implementation, customer, control, people and standards. Individual performance ratings and bonuses are based on the measures included in the BSC. Bank managers were allowed to adjust the weighting of scorecard measures and to adjust bonuses based on factors other than agreed upon scorecard measures. Ittner, et al. (2003a) finds that when managers are allowed to subjectively evaluate employee performance, non-financial performance measures are largely ignored and bonuses were based almost entirely on objectively measured financial performance. Several studies attempt to reduce or eliminate the common measures bias first reported by Lipe and Salterio (2000). Banker, Chang and Pizzini (2004) first replicate the Lipe and Salterio (2000) experiment and confirm the result. They then extend the study by providing some participants detailed information about the organization’s strategy and by categorizing the performance measures into one of four groups: performance measures linked to strategy that are 14! common to both divisions, measures linked to strategy that are unique to each division, measures not linked to strategy that are common to both divisions, and measures not linked to strategy that are unique to each division. Consistent with prior research, they find that common measures are more heavily weighted than unique measures in performance evaluations. However, when detailed information about organizational strategy is provided, unique strategy-linked performance measures are more heavily weighted than unlinked common measures. Libby, Salterio and Webb (2004) use the women’s clothing store case in an experiment that attempts to reduce common measures bias by introducing the accountability of having to justify the performance evaluation to a superior and by improving the perceived quality of the BSC measures through third-party assurance of the scorecard in the form of an audit report on the relevance and reliability of scorecard performance measures and results. After controlling for study participants’ work experience, academic background and major, they find that under conditions of accountability and third-party assurance, participants increase their use of unique measures in evaluating performance. Roberts, Albright and Hibbets (2004) also use a version of the women’s clothing store case in their experiment. Participants are presented both common and unique BSC measures in a disaggregated format (i.e., not grouped into the traditional four BSC perspectives) and are asked to rate the two division managers on each separate performance measure. The ratings of the performance measures are then mechanically aggregated into a scorecard using predetermined weights for each measure. In contrast with the Lipe and Salterio (2002) finding that performance evaluations between the two division managers do not significantly differ when above-target and below-target performance measures are distributed throughout an ungrouped list of BSC measures, Roberts, et al. (2004) find that unique performance measures accounted for a 15! statistically significant difference in the performance evaluation of the two division managers when the measures are presented in a disaggregated format. The authors argue that the finding suggests that eliminating the balanced scorecard four-perspective format will reduce the common measures bias. Dilla and Steinbart (2005) investigate whether more knowledge about the balanced scorecard will mitigate the effect of the common measures bias. Study participants read two articles about the balanced scorecard written by Robert Kaplan and David Norton and a white paper describing and illustrating “an oil’s company’s experience of designing and implementing a BSC.” The researchers find that although unique measures are used in performance evaluation and are not completely ignored, more weight is placed on common measures than on unique ones. Cardinaels and van Veen-Dirks (2010) conduct two experiments to investigate the impact of the organization of BSC measures – unformatted or categorized in multiple perspectives – on the performance evaluations of two divisions of a retail clothing store (Streetware and Family Fashions) and to assess the impact of using markers (plus or minus signs) on BSC performance measures. Consistent with prior research, the authors find that the organization of performance measures in a BSC-format versus an unstructured list, affects management judgment. More weight is place on financial measures than on non-financial measures when presented in the BSC-format, when markers are presented with performance measures, the measures are more heavily weighted regardless of whether presented in a BSC-format or an unformatted list. Unlike many of the prior experimental studies use that focus on mitigating the “bad effects” of BSC use (i.e. reducing the common measures bias that is introduced into evaluations when performance measures are presented in a BSC format), Tayler (2010) conducts a study to 16! investigate how the use of a BSC can have a positive effect on management behavior. Tayler (2010) manipulates the presentation of performance measures in a BSC causal-chain format or as four balanced groups, and manipulates participant involvement in the selection of one of two strategic initiatives to pursue as well as the selection of performance measures against which the initiative will be measured. Psychology theories of motivated reasoning suggest that participants are likely to more favorably evaluate initiatives that they were involved in selecting and less favorably evaluate those that they did not choose, even when evidence indicates that their chosen initiative is not performing well. Consistent with that prediction, Tayler (2010) finds that managers rate the performance of their chosen initiative more favorably than that of the initiative that they were not involved in selecting. In addition, he finds that the presentation format of BSC measures matters, but only when managers are involved in the selection of performance measures. When managers select the initiative, but not the performance measures, motivated reasoning effects persist regardless of whether the measures are presented as four balanced groups or in a BSC casual linkages format. However, motivated reasoning is completely mitigated when performance measures are selected by the manager and presented as a strategy map (i.e. in the causal linkages format). Humphreys and Trotman (2011) modify and extend Banker et al. (2004) by including in their experiment a treatment group that provides evaluations based on a scorecard in which all performance measures are linked to the organization’s strategy. The common measures bias is present when only some of the performance measures are linked to the strategy and disappears when all measures are linked to the organizational strategy. In their second experiment, some participants are provided detailed strategy information and other are provided no information about the strategy, however all performance measures are still linked to the strategy. In the 17! absence of detailed strategy information, the common measures bias persists, whereas it is not present when detailed strategy information is provided. 2.3 Case studies or field studies describing the implementation of BSC in a single firm Much of the academic research on the balanced scorecard has involved case studies or field studies of individual organizations’ BSC implementations. These studies typically describe the steps involved in selecting performance measures used in the scorecard and document the process of implementing the scorecard across the organization. They also provide management’s assessment of the success or failure of the overall BSC implementation process as well as opinions on whether the adoption or use of BSC led to the desired results. Malina and Selto (2001), study a Fortune 500 manufacturing company’s use of a balanced scorecard with its’ distributors. The company had recently implemented BSC for internal performance management and had developed a distributor scorecard that was used as the basis for contract negotiation and for relative performance evaluation of its’ distributors. The distributors were aware of the BSC, but were not involved in selecting any of the performance measures that were included in the scorecard. With the use of data collected in interviews with individuals directly involved with the scorecard, the researchers find that this one-sided communication about the BSC results in conflict and tension between the firm and its’ distributors and conclude that some of the most important factors in the successful implementation of BSC are the use of relevant and reliable measures, attainable performance goals, and appropriate benchmarks for evaluation. Kasurinen (2002) studies the implementation of BSC in a European metal manufacturing company. Over the course of nearly two years, the researcher, engaged as a member of the 18! scorecard project team, documents that an organization’s inability to select performance metrics that clearly measured organizational goals and objectives and metrics that are tied to a clear-cut strategy is a significant barrier to effective BSC implementation. Kasurinen’s experience supports Kaplan & Norton’s emphasis on following the prescribed process for successful implementation of BSC and suggests that poor performance may be related to how the scorecard was designed and implemented, not whether or not the system works. Ittner, Larcker and Meyer (2003a) find that bank managers ignored non-financial measures in performance evaluation when not strictly required to use them (i.e. when allowed to use subjectivity in evaluating performance). By examining the relationship between non- financial performance measures and future financial performance (e.g. changes in revenues, expenses, and margin) they find that managers exclude value-relevant information when conducting their performance evaluations. The authors find that managers consistently place more weight on objective, outcome-oriented, external or quantitative measures of performance than on subjective, input or driver-oriented, internal or qualitative measures. Ittner, et al. (2003a) also documents that the switch to the BSC-based compensation plan does not improve branch managers’ understanding of strategy and business objectives, nor do they feel that their efforts are recognized and rewarded under the new system. In fact, most employees surveyed are not satisfied with the scorecard process and do not believe that it had achieved its’ objectives. The bank subsequently abandoned the use of BSC. Tuomela (2005) documents the implementation of the 3K Scorecard (i.e. BSC) at FinABB over the course of four years, using Simons’ four levers of control as framework for analysis. 7 The implementation and use of the scorecard is examined from the perspective of it !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! @ !Simons’ four levers of control include: belief systems that communicate the core values of the organization; boundary systems which define the boundaries within which the organization and its’ employees work; diagnostic 19! being an interactive control system used to respond to the changing business environment. Tuomela follows the progression of the BSC from the evaluation stage until two years after its’ initial implementation. In addition to documenting that the BSC was used as a means of communicating important information to employees, that performance measures were initially established from the bottom-up but were subsequently modified to conform to company targets, and that incentive compensation was gradually shifted to include non-financial performance measures, the study also reported some resistance to the scorecard as some lower level managers felt their workloads had increased because they were asked to report some of the performance measures instead of the number being provide by the finance departments. The author concludes that how a performance measurement system (e.g. BSC) is as important as which performance measurement system is used in determining implementation success. 2.4 Studies that investigate how and for what purpose BSC is used within organizations Rather than examine the impact of BSC use on judgment and decision-making or organizational performance, some studies investigate the mechanisms by which and the purposes for which BSC is used. Malmi (2001) interviewed “the person most knowledgeable about the development and use of the BSC application” at 17 Finnish companies who were early adopters of BSC. The companies span multiple industries and the interviewed personnel range from business planning and accounting/finance employees to the company CEO. Interview data indicate that most !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! control systems which define performance measures against which the organization is evaluated; and interactive control systems which allow for feedback and revision of strategic plans and goals. 20! companies’ scorecards include the four perspectives 8 suggested by Kaplan and Norton (1996, 2001) and are implemented at the business unit level. Data for scorecard measures is often collected manually, and the scorecard is seen as tool to gather information and as a means of implementing strategy. The study also documents some of the reasons why firms chose to implement BSC, including as a means of “translating strategy into action,” to support other change initiatives, to abandon budgeting, or simply because it was the latest management fad. The authors suggest that the popularity of the BSC in Finland may be due in part to the abundance of BSC consultants in addition to the appeal of the logic of the BSC model. Speckbacher, Bischof and Pfeiffer (2003) develop a typology of balanced scorecards based on the evolution of the BSC literature. Type I scorecards include financial and non- financial performance measures tied to organizational strategy. Type II scorecards include hypothesized cause-effect relationships. Type III scorecards, in addition to causal linkages, are tied to incentive compensation. The authors survey 200 publicly traded companies in German- speaking companies to assess the state of their BSC implementations. 45 of 174 responding firms had implemented BSC at the business-unit or corporate level, and the authors classify the firms according to the newly developed typology. Fifty percent of BSC-user firms implemented a Type I BSC and twenty-one percent of firms had implemented a Type II BSC. The remaining twenty-nine percent had implemented a Type III BSC where performance against the scorecard was tied to compensation. Several factors that the researchers believe may affect the type of BSC used are also examined. BSC user firms on average are larger than non-user firms, but the type of BSC used is not related to the organization’s size. Overall, industry is not associated with BSC use of any type with the exception that BSC use by firms in the consumer & retail was !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! A !Kaplan and Norton (1996, 2001) suggest four perspectives, but acknowledge that four perspectives is neither a necessary or sufficient condition to ensure BSC success. They indicate that in their experiences with firms implementing BSC, they have not seen a scorecard with less than four perspectives. 21! significantly lower than other industries. In addition, Type III BSC users regard the scorecard as more important to the organization than Type I and Type II users. Wiersma (2009) surveys individual managers from 19 Dutch BSC user companies to investigate the purposes for which BSC is used. Using exploratory factory analysis to analyze responses to survey questions, Wiersma (2009) identifies three dimensions of BSC use: decision making and decision-rationalizing; coordination of activities; and, self-monitoring. Additional analysis indicates that the more flexible the evaluation system is, the less BSC is used for coordinating activities and self-monitoring, and when rules and procedures are more strictly followed, BSC is used more for decision-making and decision-rationalizing. When managers are more receptive to new types of information and new performance measurement systems, BSC is used more for decision-making, decision-rationalizing, and coordination of organizational activities. 2.5 Commentary or studies testing the cause-effect relationship in BSC models A key premise of the Balanced Scorecard is that performance measures in each perspective should be associated in a cause-effect manner (Kaplan and Norton, 1996). 9 Positive performance on learning and growth perspective measures should lead to positive performance on internal/business process measures, which should then lead to positive performance on customer perspective measures, ultimately leading to positive or improved financial performance. This hypothesized relationship between performance measures in the BSC is often referred to as the “causal linkages hypothesis.” It is the assumed cause-effect relationship between non-financial and financial performance measures that distinguishes the BSC from other !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! B !Ittner, et al. (2003b) find that 76.9% of firms in their study that claim to use a balanced scorecard “place little or no reliance on business models.” 22! performance measurement and management systems. Establishing causality requires three criteria: temporal precedence (i.e. the cause must occur before the effect); covariation between the cause and effect variables (i.e. if the cause is present then the effect is present, and if the cause is not present, the effect is not present); and no alternative explanation is available (i.e. no other factor or influence should explain the obtained relationship or results). Some accounting literature has simply commented on the “causal linkages hypothesis,” while other literature has presented the results of studies that test the hypothesis using analytical models or empirical tests of a specific organization’s strategy map (scorecard) or business model. Though their studies do not specifically test a balanced scorecard model, Ittner and Larcker (1998b) and Banker, Potter and Srinivasan (2000) provide evidence about whether non- financial performance measures (e.g. customer satisfaction) are leading indicators of future financial performance. Using data from a telecommunications firm, Ittner and Larcker (1998b) find a positive association between individual customer satisfaction measures and revenues in the following year, and between business unit customer satisfaction measures and revenues, expenses, and margins. With data for firms included in the American Customer Satisfaction Index (ACSI), they investigate the impact of firm-level customer satisfaction on the market value of equity (MVE) after controlling for information contained in the book value of assets and liabilities. The authors find a positive association between the ACSI index and MVE that is driven largely by firms in the transportation, utilities and communications industries. They also find evidence that the stock market responds positively to information contained in measures of customer satisfaction. Banker, et al. (2000) estimate the impact of two customer satisfaction measures on the revenues, costs, and profits of a hotel chain and find a positive association between one measure, the likelihood that a customer will return to the hotel, and 23! contemporaneous revenues and profits. The other customer satisfaction measure, the average number of complaints received is not statistically associated with any of the financial performance measures, however an increase in the number of complaints is associated with lower revenues. Norreklit (2000) discusses the causal linkages hypothesis in the context of causal and logical relationships. A relationship is causal if one factor can be said to cause the other, e.g. smoking causes lung cancer. However, 1 + 2 = 3 is a logical relationship, not a causal one because it cannot be said that 1 or 2 causes 3. Norreklit argues that whereas causal relationships can be empirically tested, logical relationships cannot be verified and she offers several explanations for why the presumed cause-effect relationship between measures in the balanced scorecard is problematic. Norreklit contends that accounting models, such as BSC, are logical models that cannot be empirically proved or rejected, and that “the relationship between two phenomena cannot be both logical and causal.” Failure to incorporate time (lags) into the BSC model, which is central to the establishment of causality, and the circular interdependencies among BSC measures are presented as further evidence of the challenge of validating the cause- effect relationship between non-financial and financial performance measures in the BSC. Norreklit (2000) acknowledges the counterargument that the creators of the balanced scorecard may have a conceptualization of the cause-effect relationship that differs from that which she describes. Bryant, Jones and Widener (2004) collect data for 125 firms included in the American Society for Quality (ASQ) customer satisfaction index (ACSI) and construct seven proxies of generic BSC measures. 10 Using structural equation models (SEM) they test the relationship !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! "C !According to Kaplan and Norton (1996) scorecard measures should be derived from the organization’s specific strategy. However, they acknowledge the existence of certain generic performance measures that are likely to be 24! between two customer perspective performance measures (market share and customer satisfaction), one internal business process measure (new product introductions) and three measures of financial performance (revenue, operating costs, and profitability). The causal linkages model explicated by Kaplan and Norton (1996) whereby positive performance on learning and growth measures leads to positive performance on internal business process measures, leading to positive performance on customer-focused measures, ultimately resulting in improved financial performance, represents the fully mediated model. This model is deemed to be too simplistic and although each path is significant, the model fit is poor. A partially mediated model is developed whereby it is hypothesized that all of the non-financial measures can have both a direct and indirect effect on financial outcomes and on other measures in the model. In the full model, several positive associations are found between the non-financial measures and different measures of financial performance. The learning and growth, internal business processes and customer satisfaction measures have an indirect, but not a direct effect on revenue and operating cost, whereas market share has a direct effect, but no indirect effect. Internal business processes and customer satisfaction measures have a direct effect on profit, but there are no relationships between the learning and growth and market share measures and profit. Some non-financial performance measures were also shown to have an effect on other non- financial measures. The learning and growth measure (employee skills) affects internal business processes and both customer related measures either directly or indirectly. Customer satisfaction has a positive impact on market share and internal business processes has a positive association with customer satisfaction. When the sample is partitioned into firms that base executive compensation on both financial and non-financial measures or financial measures only, a !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! included in many organizations’ BSC, including: return on investment, economic value-added, customer satisfaction and retention, market share, quality, response time, cost, new product introductions, employee satisfaction and information system availability.) 25! different pattern of results is found. In particular, customer satisfaction has a statistically significant negative effect on revenue and costs and no effect on profit, and internal business processes (new product introductions) has a significantly negative effect on market share. Overall, Bryant, et al.’s findings do not support the model proposed by Kaplan and Norton. However, the firms in the sample were not actual BSC users and results based on a sample of true BSC users may lead to different conclusions. Malina, Norreklit and Selto (2007) test the cause-effect relationship of measures included in the strategy map of a scorecard used by a Fortune 500 company. The company’s distributor balanced scorecard (DBSC) was used to measure the performance of distributors in its North American channel. Initially, the company did not have a strategy map for its DBSC. The strategy map created by the researchers was based on data from interviews conducted with company management and with distributors. It did not conform to the model proposed by Kaplan and Norton (2001), but rather reflected the relationships that company managers and distributors expected. Tests for Granger causality, including lagged dependent and independent variables in the models, provide little support for cause-effect relationships in the DBSC (i.e. measures included in the DBSC are not predictive of future financial performance). The authors argue that performance measurement models (PMM) can be based on logical and finality relationships in addition to the assumed cause-effect relationship, and failure to validate cause- and-effect should not cause one to abandon the PMM. Campbell, Datar, Kulp and Narayanan (2008) test the linkages between performance measures in Store 24’s balanced scorecard. 11 Store 24 had recently adopted a differentiation strategy and used BSC as its’ performance measurement system. Store 24’s BSC performance !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! "" !Store 24 is a privately-held chain of convenience stores in New England. The company provided the researchers detailed information regarding corporate strategy, employee compensation, financial performance and other information to conduct their study. 26! measures were captured in the four traditional perspectives and were presented in a strategy map that depicted the underlying hypothesized relationships between the measures in each perspective. Campbell, et al. test the relationships explicitly stated by Kaplan and Norton (1996) that measures in the learning and growth perspective will positively affect measures in the internal business processes perspective, which will positively affect customer perspective measures, ultimately leading to improved financial performance. In addition, they test whether measures in the internal business processes perspective, which they call strategy inputs, directly affect financial performance and whether the impact of strategy inputs on customer outcomes and the impact of customer outcomes on financial performance are moderated by employee capabilities (measures from the learning and growth perspective). They find a positive association between strategy inputs and customer outcomes, and mixed results or no results for the other hypothesized relationships. Store 24 ultimately abandoned the differentiation strategy, and although the researchers were not able to provide evidence in support of the causal linkages hypothesis, they did conclude that if Store 24 had subjected its’ BSC and strategy map to statistical analysis sooner, they would have known much earlier that the differentiation strategy was not working. Huelsbeck, Merchant and Sandino (2011) tests the business model of a medium-sized medical equipment manufacturer. 12 The company operated as a single business unit, had a stable management team and followed the same business model for more than a decade. Researchers had access to thirty-four quarters worth of data provided by the company and available through Compustat and CRSP. The company seemed to provide the ideal setting to test the hypothesized relationships in the business model. Kaplan and Norton (1992, 1996, 2001) do not discuss how !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! "D !The business model is not presented in the form of a Balanced Scorecard. However, the hypothesized “paths to value creation” represent the same cause-and-effect relationships depicted in the BSC and strategy maps. 27! much lead-time is required for non-financial performance measures included in a balanced scorecard to translate into improved financial performance. They simply state that some of the non-financial measures are leading indicators of future financial performance. Huelsbeck, et al., however, are able to test a model that includes the hypothesized lags 13 between four leading indicators (R&D spending, instrument placements, reagents released and change in gross margin percentage) and two measures of financial performance (accounting-based operating income and market-based stock returns). For example, changes in gross margin percentages are expected to impact financial performance in the subsequent quarter and research and development spending is not expected to impact financial performance for twelve to fifteen quarters after expenditure. In univariate tests, the study finds a positive association between lagged R&D spending and operating income and between lagged instrument placement and operating income. No relationship is found between stock returns and any of the leading indicators in univariate tests. Multivariate tests indicate a positive association between operating income and lagged R&D spending, lagged instrument placement and changes in gross margin. Only instrument placements and reagents released are significant in the model testing the association between leading indicators and stock returns. The inability to validate the company’s business model caused the authors to reflect on the “value of testing business models,” and they assert that there may be rational and valid reasons that a firm may choose not to test its’ business model, that tests of the business model may fail, and that managers may choose to continue using an unvalidated model. 2.6 Studies investigating the use of BSC on organizational performance !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! "< !The hypothesized lags represent management’s beliefs about how long it takes for non-financial performance measures to translate to improved financial performance. 28! One of the strongest criticisms of BSC is that it is not grounded in theory that can be used to predict performance outcomes. 14 Related to this criticism is the fact that extant research provides mixed evidence about whether balanced scorecard measures that are causally linked to strategy will result in improved financial performance. Successful implementation of BSC should result in a significant improvement in financial performance over time. Several articles and books have outlined process measures of success for an effective BSC implementation, including ensuring the BSC measures are aligned with the organization’s strategy, involving all relevant parties in the process of designing and implementing the BSC, and tying the BSC to employee incentives and compensation (Kaplan and Norton, 1996, 2001; Niven, 2002; Albright, Burgess, Davis, and Juras, 2007), however, there is no consensus as to what defines the outcome measure of success. Kaplan and Norton (1996) propose doubling the stock price or return on invested capital or increasing sales by 150% over the course of five years as suggested financial targets, but those are very ambitious goals, and existing studies rarely extend beyond a one- or two-year performance window. In addition, many studies do not examine actual financial performance, but instead report management’s assessment or opinion of how/whether BSC use affects qualitative/subjective or self-reported measures of organizational performance. Hoque and James (2000) survey 188 chief financial controllers of Australian manufacturing firms to investigate the factors associated with the use of BSC as well as the impact of BSC use on organizational performance. BSC use is measured using a 20-item scale, and organizational performance is captured by five self-reported measures of operating performance – return on investment, margin on sales, capacity utilization, customer satisfaction, !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 14 Kerlinger (1973) defines a theory as “a set of interrelated constructs (concepts), definitions, and propositions that present a systematic view of phenomena specifying relations among variables, with the purpose of explaining and predicting the phenomena.” 29! and product quality. 15 OLS regression results indicate BSC usage is greater for large firms and for firms in the emerging/growth life-cycle stage, and is not associated with market position relative to competitors. Two-way analysis of variance (ANOVA) indicates a main effect for the impact of BSC use on organizational performance. However, in contrast with hypothesized predictions, there is no interaction effect between BSC use and organizational size, product life- cycle stage, or market position, indicating that these factors do not moderate the relationship between BSC use and performance. In addition to investigating the impact of subjectivity on performance evaluations, Ittner, Larcker and Meyer (2003a) also examine the impact of the use of non-financial performance measures on revenue, expenses and margin. The research site is a North American retail bank whose scorecard consists of measures in six categories: financial, strategy, customer, control, people and standards. Tests of contemporaneous associations indicate positive (negative) relationship between revenue and margins (expenses). An increase in the number of retail customers is also associated with increased revenue and margins. One quarter lagged customer satisfaction and people measures are associated with higher expenses. Quality and people measures are associated with lower expenses and lower margin, respectively, and increases in the number of retail and premier customers are associated with higher expenses. Four quarter lagged tests provide more mixed results. An increase in the number of business and professional customers is associated with increased revenue and people measures are associated with increased expenses. An increased number or retail customers and quality measures are associated with higher margins, whereas people-focused performance measures are associated with lower margins. Overall, findings regarding the impact of non-financial performance non- financial performance measures on financial performance are inconsistent and inconclusive. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! "> Managers rated their organizations on a 5-point scale (1 = below average to 5 = above average). 30! Ittner, Larcker and Randall (2003b) use data from a survey of 140 U.S. financial institutions to investigate the impact of the use of different performance measurement systems (PMS) on organizational performance. The performance impact of economic value measures and the development of formal causal business models are investigated in addition to the performance impact of BSC use. Organizational performance is measured subjectively through assessment of management’s satisfaction with the PMS, and objectively via accounting and stock return data. Across all PMS, greater measurement diversity is associated with higher levels of management satisfaction and higher one-year stock returns. This result is driven by a focus on non-financial measurement. When the impact of the different PMS’ are tested individually, all have a positive effect on management satisfaction, but there is no relationship between BSC use and market-based measures of financial performance (one-year and three-year stock returns). Evidence for accounting-based performance measures is mixed. There is no association between BSC use and sales growth, and contrary to hypothesized predictions, there is a significantly negative association between BSC use and return on assets. Maiga and Jacobs (2003) survey a sample of firms that use BSC and activity-based costing (ABC) identified from the popular press. They assess the impact of the two management accounting innovations on three subjective measures of organizational performance: product quality, customer satisfaction, and margin on sales. 16 The study finds that BSC positively impacts organizational performance only when interacted with ABC. No direct effect of BSC on organizational performance is found. One of five principles of the Strategy-Focused Organization (SFO) and a key element for successful BSC implementation is a link to incentive compensation (Kaplan and Norton, 2001). !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! "? !Survey respondents were asked to rate their organization’s performance relative to their competitors on a scale from 1 (= below average) to 7 (= above average). 31! Said, HassabElnaby and Wier (2003) examine the performance impact of non-financial measures for firms that disclose the use of both financial and non-financial performance measures for management compensation. 17 The authors match the sample of firms that use non-financial performance measures (NFM) for compensation to a sample of firms that only use financial performance measures and compare performance using accounting-based (return-on-assets) and market-based (stock returns) measures. Contemporaneous and one-year-ahead relationships are assessed. Univariate tests indicate that contemporaneous and one-year-ahead return-on-assets (ROA) and stock returns are higher for firms that use both financial and non-financial performance measures to determine incentive compensation. After controlling for the debt-to- equity ratio, firm size, growth opportunities and other factors thought to affect financial performance, the use of NFM was associated with contemporaneous stock returns and both one- year-ahead ROA and stock returns, but not with contemporaneous ROA. To address endogeneity concerns, Said, et al. develop a model to predict the likelihood of including NFM for compensation. Over-investment in non-financial performance measures (i.e. large positive residual from the prediction model) is associated with lower contemporaneous stock returns and lower one-year-ahead ROA and stock returns. Firms that do not use NFM when the model predicts they should also experience lower contemporaneous stock returns and lower returns and ROA one year ahead. No association is found between contemporaneous accounting performance (ROA) and the mismatch in NFM use. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! "@ !Said et al. (2003) does not explicitly state that their sample firms are BSC users, however, the sample firms are sourced from information contained in the proxy statements. The current study identifies several BSC-user firms based on disclosures in the proxy statement along with information about the use of non-financial and financial performance measures for incentive compensation. 32! In a quasi-experimental 18 study of BSC implementation at a regional U.S. bank, Davis and Albright (2004) directly test the impact of adoption/implementation of BSC on future financial performance. Four branches in the bank’s southern division comprise the BSC- adopting treatment group and five northern division branches comprise the control group. Using a composite measure of financial performance encompassing nine financial measures that bank management believe are key to branch success, the authors find statistically significant performance differences between BSC-adopting and non-adopting branches. Davis and Albright (2004) attribute the ability to find a positive association between BSC usage and financial performance to including measures in the scorecard that are logically derived and causally linked to branch financial performance and argue that prior research’s mixed results regarding the association between non-financial measures and financial performance may be due to the “lack of a coherent linkage between the measures chosen for the performance system and the targeted financial measure of interest.” Van der Stede, Chow and Lin (2006) investigate the performance effect of performance measurement diversity and the joint effects of strategy and performance measurement diversity on organizational performance using survey data from 128 European manufacturing firms and manufacturing firms in Southern California. Diversity of performance measurement is captured by two measures. The first measure is a count of financial and non-financial measures. The second measure is a categorical variable that captures whether a firm uses: financial, objective nonfinancial and subjective nonfinancial measures; financial and objective nonfinancial measures; financial and subjective nonfinancial measures; or financial measures only. The strategy measure captures the firm’s focus on quality-based manufacturing and performance is measured across four dimensions: financial, operating, employee-oriented, and customer- !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 18 BSC-adopting (treatment) and non-adopting (control) bank branches were not randomly assigned. 33! oriented. 19 The authors find a positive association between the measurement diversity and organizational performance. Higher performance is also associated with firms that emphasize a quality manufacturing strategy and extensively use subjective nonfinancial performances measures. In addition, the authors find that performance measurement misalignment (i.e. not using enough performance measures) is associated with lower organizational performance. Crabtree and DeBusk (2008) use data collected from firms that participated in an Institute of Management Accountants (IMA) survey of strategic management techniques to investigate the impact of balanced scorecard use on shareholder returns. They match fifty-seven publicly traded BSC user firms to non-user firms and find that three-year buy-and-hold returns are significantly higher for firms that use BSC. Comparisons of BSC-user firms to control firms on accounting-based performance measures (operating margin, return on assets, and return on equity) provide mixed results. Ittner (2008) reviews the existing literature to address the question “Does measuring intangibles for management purposes improve performance?” The literature review does not exclusively examine BSC studies, however, BSC is cited as an example of the type of performance measurement system for which measurement of intangible assets is important. Ittner classifies the literature into four groups: 1) studies that use perceptual measures of performance outcomes; 2) studies that use actual economic outcomes; 3) quasi-experimental studies; and 4) causal business model studies. Most studies using perceptual performance outcomes find a positive relationship with intangible asset measurement, while results for those studies using actual financial performance outcomes provide mixed evidence in support of a positive association. Quasi-experimental studies that have a control group against which to compare and contrast performance outcomes provide the strongest tests, but the evidence is !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 19 These dimensions represent the four Balanced Scorecard perspectives. 34! mixed. Causal business model studies also provide a wide range of mixed results. Most studies only test a relationship between a single variable/measure and financial performance, instead of a model with multiple causal links. Results vary from finding a positive association between the variable of interest and accounting and stock-based performance outcomes, to finding an association between the variable and accounting but not stock-based performance, or vice versa, to finding no association between the intangible asset measure and any measure of financial performance. One study, Ittner, et al. (2003b) even finds a significant negative association between intangible assets, proxied by the use of BSC as the organization’s performance measurement system, and accounting-based financial performance (ROA). De Geuser, Mooraj and Oyon (2009) survey 164 European BSC adopter firms to evaluate the impact of BSC use on five measures of organizational performance. The measures capture subjective/qualitative assessment of organizational performance in addition to financial performance. Survey questions capture management’s evaluation of the success of the BSC, opinions about the integration of key process in the BSC model and whether BSC development leads to greater business unit autonomy. In addition, the study evaluates the costs and benefits of BSC adoption. Survey results suggest that the use of BSC has a positive effect on self-reported, subjective measures of organizational performance. Furthermore, De Geuser, et al. (2009) investigates the source of organizational performance in accordance with Kaplan and Norton’s (2001) five criteria for the Strategy-Focused-Organization. Translating the strategy into operational terms, aligning the organization to strategy and making strategy a continual process are associated with some, but not all, measures of organizational performance. However, in contrast to Kaplan and Norton’s (2001) assertions, making strategy everyone’s everyday job and 35! mobilizing change through executive leadership are not associated with any of the measures of organizational performance. ! 36! CHAPTER 3 HYPOTHESIS DEVELOPMENT 3.1 Characteristics of Firms that Use BSC The primary research question that this study answers is: Does the use of Balanced Scorecard improve financial performance? However, any study involving a firm’s choice to undertake or not undertake a particular activity (e.g. to adopt or not to adopt BSC) must first consider the endogenous nature of such choice. Several factors may influence a firm’s decision to adopt a Balanced Scorecard. Those factors may also be correlated with the firm’s financial performance. Chenhall (2006) argues, “enhanced performance outcomes will depend on how different types of measurement systems best suit, or fit, with an organization’s specific context.” Based on theory and following prior literature, I develop a model of contextual factors to predict a firm’s likelihood to adopt BSC as its’ performance measurement system. I posit that a firm’s level of intangible assets, strategy, recent financial performance, size and amount of divisionalization are associated with its’ propensity to adopt BSC. 20 3.1.1 Intangible Assets The Balanced Scorecard was initially developed from a study involving firms with large amounts of intangible assets (Kaplan, 2010). The International Accounting Standards Board (IASB) defines an asset as “a resource controlled by the enterprise as a result of past events and from which future economic benefits are expected to flow to the enterprise.” Intangible assets are long-term assets that do not have physical substance, yet provide substantial economic !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! DC !A few studies explore determinants of BSC usage using survey data. ! 37! benefit to the organization, e.g. patents, copyrights, trademarks, and computer software. Many knowledge-based, technology and service firms cannot record their most valuable assets, such as human capital, internally developed IT systems and expertise, and internally developed goodwill, on their balance sheets. Because these assets are not easily imitable they can provide a source of sustainable competitive advantage for a firm. However, measuring the value of intangible assets and demonstrating how they ultimately impact the bottom-line are difficult tasks. The Balanced Scorecard and its’ corresponding strategy maps provide a framework for illustrating how intangible assets (e.g. a highly skilled/trained customer support staff or sales force) affect future financial performance (Kaplan and Norton, 2004). Thus, firms with large amounts of unrecorded intangible assets are conjectured to benefit most from the use of BSC and to be more likely to adopt it. 21 H 1a : Firms with more intangible assets are more likely to adopt BSC. 3.1.2 Strategy There is no single definition of strategy. Porter (1996) asserts that while operational effectiveness is necessary for strategy, it is not in and of itself “strategy.” Strategy requires defining a company’s competitive position, including core competencies and areas of competitive advantage, making trade-offs or choices to sustain competitive advantage and aligning organizational activities to achieve strategic goals. In addition, a firm’s chosen strategy !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! D" !Hendricks, Menor and Wiedman (2004) measure intangible assets as the mean ratio of intangible assets (undefined) to total assets and find no relationship with BSC adoption for a sample of Canadian firms. Using three alternative specifications for intangible assets: the sum of the rank of the firm’s ratio of intangible assets to total assets adjusted for the industry median and the rank of the ratio of investment in R&D and innovation (undefined) to industry-median-adjusted sales (RANK_INTANG); ratio of intangible assets to total assets adjusted for the industry mean in the year prior to BSC adoption (INV_INTANG); and an indicator variable set to 1 if the firms is a member of one of the above-median NAICS industries for investment in intangible assets (H_INTANG), Hendricks, Hora, Menor and Wiedman (2012) find no association between intangible assets and the likelihood to adopt BSC. In contrast to these prior studies, I measure intangible assets as the difference between the market value of equity and book value of equity scaled by total assets. ! 38! or strategic positioning should span several years, not just one operating cycle. Kaplan and Norton (1996) define strategy as “a set of hypotheses about cause-and-effect than can be expressed by a sequence of if-then statements.” Miles and Snow (1978) develop a typology of business strategy in which firms are classified as Defenders, Analyzers, Prospectors, or Reactors. 22 Defender firms often follow a cost-leadership orientation and may be characterized by more organizational stability. They tend to be more vertically integrated, have more formal processes and procedures in place, are slow to change, and are more conservative. Prospector firms seek to develop new products and markets, are more innovative, and are characterized by a more dynamic or uncertain operating environment. They tend to be more decentralized, with few layers of management, and are more entrepreneurial. Defender and Prospector firms are at opposite end of the strategy continuum. Reactor firms have no clearly defined strategy. The strategic goals and priorities of these firms change quickly and often, in response to changes in the business environment. Analyzer firms have characteristics of both Defender firms and Prospector firms. Analyzer firms try to maintain a balance between new product innovation and cost containment. Using the Miles and Snow strategy classifications, Simons (1987) finds that the characteristics of accounting control systems differ between Prospector and Defender firms. In particular, the extent to which control systems are tailored to departmental needs, the frequency of change in control systems, and the importance of using informal communication to convey control system information are associated with pursuing a Prospector strategy. Baines and Langfield-Smith (2003) conduct a structural equation model analysis of the antecedents of management accounting change. They find that changes towards a differentiation strategy, as !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! DD !Other strategy typologies exist. Porter (1980) creates two generic categories of strategy classification: cost leadership and product differentiation. Gupta and Govindarajan (1984) classify organizational strategy as build, hold, harvest, or divest. ! 39! are characteristic of Prospector firms, are associated with greater reliance on advanced management accounting practices, e.g. BSC. Said, HassabElnabby and Weir (2003) first investigate the performance consequences of non-financial performance measures (NFM) and subsequently investigate factors associated with greater NFM use. Following, Ittner, Larcker and Rajan (1997), the authors create a composite measure of organizational strategy where higher values reflect a Prospector strategy and lower values reflect a Defender strategy. Findings indicate that NFM are used more by Prospector firms than by Defender firms. Hendricks, Menor, and Wiedman (2004) surveys senior executives from Canadian firms to examine the association between several contingency variables, including strategy, and BSC adoption. The measure of strategy is an indicator variable set to 1 if the firm strategy is classified as Prospector or Analyzer based on key informant response to a survey question, and 0 if the firm is classified as pursuing a Defender strategy. Reactor firms are excluded from the analysis. Results indicate that Prospector and Analyzer firms are more likely to adopt BSC than Defender firms. Naranjo- Gil, Maas and Hartmann (2009) uses survey and archival data from a public hospital in Spain to understand how CFOs determine management accounting innovation. Using a self-reported measure of organizational strategy, hospitals are classified as pursuing a Prospector or Defender strategy. Results show a positive and significant path between strategy and use of innovative management accounting systems (MAS), such as BSC, indicating that Prospectors use more innovative and sophisticated MAS. Gosselin (2011) surveys Canadian manufacturing firms and finds that firms pursuing a Prospector strategy use more non-financial performance measures and adopt more innovative performance measurement approaches such as BSC. As prior research finds that strategy can be an antecedent of management accounting system (MAS) innovation, including the use of BSC, and that Prospector firms are more likely to ! 40! adopt innovative MAS, I conjecture that following a Prospector strategy versus a Defender strategy will positively influence the likelihood of adopting a Balanced Scorecard. H 1b : Firms pursuing a Prospector strategy are more likely to adopt BSC than those pursuing a Defender strategy. 3.1.3 Recent Financial Performance While strategy and strategic orientation tend to remain constant over time i.e. to persist, poor financial performance has been shown in prior literature to precipitate innovation and strategic change. In the Management literature, Lant, Milliken, and Batra (1992) study the process of strategic reorientation in a sample of firms from the furniture and computers software industries. They find a significantly negative association between poor financial performance measured by return-on-assets and strategic reorientation, i.e., after a period of poor financial performance, firms are more likely to change strategies. Said, et al. (2003) investigates factors associated with the use of NFM. In addition to strategy and other factors, they examine the impact of financial distress on NFM use. Using a composite measure of financial distress that includes the probability of bankruptcy from Ohlson (1980), the leverage ratio, and the leverage ratio scaled by R&D, the authors find that distressed firms are more likely to use NFM. 23 Hendricks, et al. (2004) examines return-on-assets (ROA) and return-on-sales (ROS) for the three-year period prior to BSC adoption for a sample of Canadian firms. They find weak evidence in support of the hypothesis that poor financial performance is associated with the decision to adopt BSC. Naranjo-Gil, et al. (2009) investigates how CFOs determine management accounting innovation. Innovative management accounting system (MAS) system !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! D< !The leverage ratios do not reflect recent poor financial performance. The Ohlson (1980) bankruptcy prediction model includes a variable to indicate if net income was negative for the past two years (INTWO), however, the variable is not significant in empirical tests. ! 41! use is indicated by whether the organization uses benchmarking, activity-based costing (ABC) and BSC. In addition to examining characteristics of the CFO (e.g. age, tenure and educational background), they also examine organizational factors that may impact management accounting innovation (e.g. organizational strategy and historical financial performance). Results from partial least squares (PLS) analysis show a significant negative path from historical financial performance to innovative MAS use, indicating that poor financial performance leads to MAS innovation and use of tools such as BSC. Hendricks, et al. (2012) examine the likelihood of Canadian firms to adopt BSC and find that a firm’s ROA in the year prior to adoption is negatively associated with the propensity to adopt, i.e. firms with low or negative ROA are more likely to adopt BSC than firms with high ROA. Firms that have recently experienced losses are more likely to change strategic direction and have a need to find new ways to measure and improve performance to achieve desired outcomes (Westphal and Fredrickson, 2001). Thus, these loss firms are thought to be more likely to adopt BSC. H 1c : Firms with poor prior financial performance are more likely to adopt BSC. 3.1.4 Size Merchant (1981) studies the characteristics of control systems in electronics industry firms and finds that size is positively associated with control system sophistication. In a study of Finnish companies, Malmi (1999) finds a significant relationship between size (measured by the number of personnel in each firm) and the likelihood of adopting a new management accounting innovation, specifically activity-based costing. Hoque and James (2000) survey Australian manufacturing firms and investigate the association between BSC usage and size, measured by ! 42! the number of employees, sales turnover and total assets. BSC usage is measured using a 20- item instrument that includes financial and non-financial performance measures that would be included in a scorecard. Items were rated on a Likert-scale from 1 to 5. Results show that firm size is positively related to BSC usage. In Speckbacher, et al.’s (2003) study of firms in German-speaking countries, they find that larger firms are more likely to adopt BSC, but do not find that larger firms are more likely to adopt a specific type of BSC, according their typology of Type I, Type II, and Type III BSC users. Abdel-Kader and Luther (2008) survey United Kingdom food and beverage sector firms and find that large firms adopt more sophisticated management accounting practices. Duh, Xiao and Chow (2009) studies Chinese listed firms and finds that firm size, measured as the natural log of total assets, is positively associated with the extent of use of management accounting and controls (MAC), including the use of a mix of leading and lagging performance indicators explicitly tied to the organization’s strategy. Joshi (2011) measures size by the firm’s total assets and finds that the use of both financial and non- financial performance measures is associated with larger firms. Based on prior literature, I expect an organization’s size to positively impact the likelihood of BSC adoption. H 1d : Large firms are more likely to adopt BSC. 3.1.5 Decentralization In decentralized organizations, decision-making authority is pushed down to lower levels of management rather than being concentrated among a small number of key senior level executives. Decentralized firms require highly developed management accounting systems that provide managers with the information needed for decision-making and management control. Merchant (1981) finds that decentralization is associated with more formal control systems. ! 43! Abernethy, Bouwens and van Lent (2004) survey divisional managers of Dutch firms and find that the use of divisional summary performance measures instead of firm-wide measures is positively associated with the degree of decentralization. Abernethy and Bouwens (2005) find that sub-unit managers who are involved in management accounting system design and who have to ability to redeploy human and financial resources are more accepting of management accounting system innovation. In addition to finding a positive association between firm size and management accounting practice sophistication, Abdel-Kader and Luther (2008) find that decentralized firms that delegate responsibility for planning and control activities to business unit managers have more sophisticated management accounting practices than those of centralized firms. Lee and Yang (2011) firms listed on the Taiwan Stock Exchange and their adoption of “Western” Management Control Systems (MCS). The study finds that organic organizations, characterized by greater decentralization, fewer formal rules and a wider control range, are associated with greater use of integrated performance measures, and greater use of measures in each of the four BSC perspectives (financial, customer, internal process and learning & growth) than mechanistic organizations. The study also finds that organic organizations have performance measurement systems (PMS) that are at a higher developmental stage (i.e. PMS reflect cause-effect relationships) than mechanistic organizations. Gosselin (2011) surveys Canadian manufacturing firms and finds that decentralization is associated with greater use of non-financial performance measures, but not with the adoption of innovative management accounting practices. Based on these findings, decentralization is expected to have a positive influence on the adoption of BSC. H 1e : Decentralized firms are more likely to adopt BSC than centralized firms ! 44! In summary, I formally hypothesize the following: H 1a : Firms with more intangible assets are more likely to adopt BSC. H 1b : Firms pursuing a Prospector strategy are more likely to adopt BSC relative to those pursuing a Defender strategy. H 1c : Firms with poor prior financial performance are more likely to adopt BSC. H 1d : Large firms are more likely to adopt BSC. H 1e : Decentralized firms are more likely to adopt BSC than centralized firms. 3.2 Financial Performance Impact of BSC Use All of the performance measures in each perspective of a Balanced Scorecard should be causally linked to financial objectives as the ultimate outcome, with improvements in financial performance usually seen two to three years after adoption (Kaplan and Norton, 1996, p. 34; 2001). Investment in technological innovation should result in a measurable comparative advantage for adopting firms compared to non-adopting firms. Successful implementation of BSC requires a substantial amount of time and human capital, and firms that make those investments and adopt BSC should realize returns on their investments by outperforming firms that do not use BSC. However, prior research investigating the impact of BSC usage on financial performance has provided mixed results. Lingle and Schiemann (1996) does not specifically mention BSC, however the authors identify six strategic performance areas – financial performance, operating efficiency, customer satisfaction, employee performance, innovation/change, and community/environmental issues - and partition their sample into measurement-managed organizations (non-measurement-managed organizations) based on whether senior management used 3 or more (less than 3) primary ! 45! performance areas in semi-annual performance reviews. They find that measurement-managed organizations (i.e. BSC adopting) outperform non-measurement-managed organizations on subjective (i.e. self-reported) financial performance measures. In addition to finding that firm size is positively associated with BSC usage, Hoque and James’ (2000) survey of chief financial controllers of Australian manufacturing firms finds a positive association between BSC usage and managers’ self-reported assessment of five financial and non-financial performance measures – return on investment, margin on sales, capacity utilization, customer satisfaction, and product quality. Ittner, Larcker and Meyer (2003a) examine the impact of the use of non-financial performance measures (NFM) on revenue, expenses and margin. Tests of contemporaneous associations between NFM and financial performance indicate positive (negative) relationships associations with revenue and margins (expenses). An increase in the number of retail customers is also associated with increased revenue and margins. One quarter lagged customer satisfaction and people measures are associated with higher expenses. Quality and people measures are associated with lower expenses and lower margin, respectively, and increases in the number of retail and premier customers are associated with higher expenses. Tests using four-quarter lagged values provide more mixed results. An increase in the number of business and professional customers is associated with increased revenue and performance measures related to people are associated with increased expenses. An increased number or retail customers and performance measures related to quality are associated with higher margins, whereas people- focused performance measures are associated with lower margins. Overall, findings regarding the impact of non-financial performance non-financial performance measures on financial performance are inconsistent and inconclusive. ! 46! Ittner, Larcker and Randall (2003b) objectively measure organizational performance using accounting and stock return data. Greater measurement diversity is associated with higher one-year stock returns. This result is driven by a focus on non-financial measurement. When the impact of the different PMS’ (Economic Value Added, Business modeling and BSC) are tested individually, there is no relationship between BSC use and market-based measures of financial performance (one-year and three-year stock returns). Evidence for accounting-based performance measures is mixed. There is no association between BSC use and sales growth, and contrary to hypothesized predictions, there is a significantly negative association between BSC use and return on assets. Said, HassabElnaby and Wier (2003) examine the performance impact of non-financial measures for firms that disclose the use of both financial and non-financial performance measures for management compensation. Contemporaneous and one-year-ahead relationships are assessed for accounting-based (return-on-assets) and market-based (stock returns) measures. Univariate tests indicate that contemporaneous and one-year-ahead return-on-assets (ROA) and stock returns are higher for firms that use both financial and non-financial performance measures to determine incentive compensation. OLS regression indicates that the use of NFM is associated with contemporaneous stock returns and both one-year-ahead ROA and stock returns, but not with contemporaneous ROA. Davis and Albright (2004) directly test the impact of adoption/implementation of BSC on future financial performance for a sample of bank branches. Four branches in the bank’s southern division comprise the BSC-adopting treatment group and five northern division branches comprise the control group. Using a composite measure of financial performance ! 47! encompassing nine financial measures that bank management believe are key to branch success, the authors find significantly better financial performance for BSC-adopting branches. Van der Stede, et al. (2006) investigates the performance impact of measurement diversity (i.e. the extensive use of objective and subjective nonfinancial performance measures in addition to financial performance measures) for a sample of European and U.S. manufacturing firms. The authors find a positive relationship between organizational performance and the extent of measurement diversity using both continuous and categorical measures of measurement diversity. Crabtree and DeBusk (2008) investigate the impact of balanced scorecard use on shareholder returns. They match fifty-seven publicly traded BSC user firms to non-user firms and find that three-year buy-and-hold returns are significantly higher for firms that use BSC. Comparisons of BSC-user firms to control firms on accounting-based performance measures (operating margin, return on assets, and return on equity) provide mixed results. Xiong, Su and Lin (2008) survey senior executives and financial officers of mid-size and large firms in China to investigate the impact of the use of financial and nonfinancial performance measures on organizational performance. The study finds that Chinese firms extensively use both financial and nonfinancial performance measures and consider the cause- effect relationship between performance measures. The study also finds that greater use of nonfinancial performance measures and higher levels of cause-effect relationships among performance measures are associated with higher organizational performance. Duh, et al. (2009) includes objective performance measures, subjective performance measures and an aggregate measure of performance in the analysis of Chinese listed firms. The ! 48! authors find a positive relationship between the use of MAC and each type of performance, including objectively measured cost efficiency. Lee and Yang (2011) include five self-reported measures of organizational performance in their survey of Taiwanese listed firms. Managers rate their respective organizations on gross profit, return on investment, customer satisfaction, product/service quality and employee productivity on a five-point Likert scale. The study finds that the use of integrated PMS and having a more highly developed PMS are both associated with higher organizational performance. Zhang, Pan and Lin (2013) also find that the extent to which firms use comprehensive performance measures and nonfinancial performance measures is associated with better performance in a study of Chinese firms. The extant evidence is clearly mixed and inconclusive. Nevertheless, if Kaplan and Norton’s assertions are correct, and if the use of BSC is congruent with the objective of maximizing shareholder wealth, then the adoption of a BSC approach to performance measurement/ management should lead to improvements in financial performance over time. Thus, I formulate the following prediction: H 2 : The use of BSC improves financial performance. ! 49! CHAPTER 4 SAMPLE SELECTION AND RESEARCH DESIGN 4.1 Sample Selection 4.1.1 Balanced Scorecard Users The treatment group consists of firms that claim to have implemented a Balanced Scorecard. 24 I use the Mergent Online database to identify firms that use either of the phrases “balanced scorecard” or “corporate scorecard” or the word “scorecard” in any of their government (i.e. SEC) filings (e.g. 6-K, 10-K, 8-K, DEF 14) from 1993 to 2010. Foreign corporations use the 6-K to disclose information, while U.S. domestic corporations use the 8-K (current report), 10-K/Q (annual/quarterly report), or DEF 14 (proxy statement). Mention of the use of BSC in the proxy statement is most often related to the firm’s use of the BSC for compensation in addition to using it to measure organizational performance. I read the filing documents to confirm that the word or phrase is used in the proper context. If a firm discloses in its’ SEC filings the use of a balanced scorecard for performance measurement and/or for compensation of the CEO or other executives, it is deemed for the purposes of this study to have implemented a Balanced Scorecard. For example, in the biography of Executive Vice President and Chief Financial Officer Abiola Lawal included in its’ July 29, 2010 8-K, CAMAC Energy states that “he (Mr. Lawal) worked as a Financial Analyst on the Balanced Scorecard Project in the Financial Planning Department for Walt Disney Company…” CAMAC Energy is not !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 24 This study is not an experiment, however, firms that use BSC are considered the treatment group and the matched group of non-using firms is considered the control group. ! 50! considered a BSC-using firm for the purposes of this study. However, manufacturing company KLA-Tencor Corporation states the following in its’ proxy statement: “Though the balanced scorecard has been used with the Company for many years as a tool for assessing the Company’s performance across a broad range of key areas, fiscal year 2010 represented the first time that the scorecard was formally incorporated into our executive compensation program. The balanced scorecard takes into account our strategic objectives of growth, customer focus, operational excellence and talent…and applies scores for the Company’s performance against a variety of specific goals within each of those variables. The scorecard is tracked throughout the year and is reviewed every quarter, then formally presented to the Compensation Committee and the Independent Board Members following the conclusion of the fiscal years for assessment as to the Company’s success in achieving the pre-established annual goals.” KLA-Tencor is classified in this study as a BSC-user. Additionally, publicly traded companies that have been inducted into the Balanced Scorecard Hall of Fame, or which are listed on Palladium Group’s and the Balanced Scorecard Institute’s websites are included in the sample of BSC-adopter firms. 25 If a company is found in multiple sources, its’ source is designated in the following order – Hall of Fame, Palladium Group, BSC Institute. The company is then considered a duplicate for the other source groups. Firms such as COGNOS Inc., Gentia Software, Veritas Software, and Xformity Technologies, that provide Balanced Scorecard software and/or consulting services are excluded. In addition, some firms identified as BSC-User firms in SEC filings do not have a gvkey, the unique code used to identify companies in Compustat, and others did not have the minimum required five years of data to create a strategy score or were missing data for variables needed for analysis. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! D> !In the context of BSC usage, the words “adopt”, “implement”, and “use,” as well as variants thereof, are used interchangeably. ! 51! I identify a pool of 314 publicly traded potential BSC-adopting firms. After excluding firms based on the criteria discussed above, the remaining sample includes 155 unique firms, hereafter, BSC-User firms. Table 1 presents the distribution of BSC-User firms by source. 4.1.2 Control Sample Because it is impossible to know what the financial performance of the BSC-adopting firm would have been had the firm not chosen to adopt BSC, I select a control group as the counterfactual against which to compare performance. Matched-pair designs have been used in prior studies involving the comparison of innovation-adopting and non-adopting firms (Easton and Jarrell, 1998; Kennedy and Affleck-Graves, 2002; Kinney and Wempe, 2002; Crabtree and DeBusk, 2008). In addition, there is the potential for selection bias in any research involving managerial choice. Management’s decision to adopt a balanced scorecard is not random. Therefore its’ effect on firm financial performance as compared with that of non-adopting firms’ financial performance may be biased. Prior research matches BSC-adopter firms to control firms on a single variable, e.g. market value of equity, book-to-market ratio and total assets (Said, et al., 2003; Crabtree and DeBusk, 2008). Matching on multiple criteria increases the likelihood that any differences found between the treatment and control groups are due to the treatment, i.e. the use of BSC, and not due to other correlated variables or factors (Tucker, 2010). Multi- criteria matching becomes exceedingly complex as the number of covariates increases. To address this issue and to control for selection bias, I select a group of control firms by matching BSC adopter firms to non-adopter firms using the propensity-score calculated by my prediction model of a firm’s likelihood to adopt a balanced scorecard. I hypothesize that a firm’s size, level ! 52! of intangible assets, realized strategy, recent financial performance, and amount of decentralization are related to its’ propensity to adopt BSC. 26 It is difficult to precisely determine the adoption year for many BSC-user firms. The BSC Hall of Fame Reports identify the year of adoption for some, but not all, Hall of Fame firms. For a very small number of other firms, the adoption year is disclosed either in a required SEC filing or press release. The year before the first year in which the use of Balanced Scorecard is disclosed either in an SEC filing or in a press release is used as the adoption year for the remaining firms. Kaplan and Norton (1996) state that the scorecard is disseminated to the entire organization “at the end of one year, when the management teams are comfortable with the strategic approach…” Therefore, I assume that adoption occurs in the year prior to any public disclosure. Since the actual month and day are indeterminable, I use January 1st of the adoption year as the BSC Adoption Date following prior research (Haka, et al., 1985; Kennedy and Affleck-Graves, 2001; Crabtree and DeBusk, 2008). All firms in Compustat that are not missing data to compute the variables required for analysis are potential matches for the BSC adopter firms. Adopter firms are matched one-to-one, without replacement to the nearest-neighbor non-adopter firm within the same industry code and fiscal year. Matching is performed once, in the year prior to BSC adoption. Thus, an individual BSC adopter firm is paired with only one non-adopter firm in the dataset. In addition to matching all 155 BSC-User firms using propensity-score matching to address self-selection concerns (All Obs dataset), I also create datasets excluding the most “problematic firms.” The “Minus Financials” dataset excludes firms in Fama-French industry !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 26 While the use of propensity score matching controls for bias due to observable firm characteristics that may affect a manager’s decision to adopt BSC, it does not control for selection bias that may arise from unobservable factors that may also impact managers’ decisions (Lennox, et al. 2012). ! 53! group 16 – Banks, Insurance & Other Financial firms. 27 The “SEC Only” dataset includes those firms identified as a BSC-User through mention of the use of BSC in SEC filings and not by any other source and the “Minus HOF” dataset excludes the 18 Hall of Fame firms. The matched sample datasets are used to test H2. 4.2 Research Design 4.2.1 Prediction Model for Likelihood to Adopt Balanced Scorecard In order to test hypotheses 1a – 1e, I use the following probit (prediction) model: Pr(BSC_User) it = f (Intangible_Assets it , Strategy_Score it , Financial_Distress it , Size it , Decentralization it ,) (1) Variables are measured as follows: BSC User is an indicator variable set to 1 if the firm is one of the 155 publicly traded BSC using sample firms, and 0 otherwise. Intangible_Assets are resources that generate future economic benefits that are not captured on the balance sheet and are measured as the difference between market value of equity and book value of equity scaled by total assets. Strategy is measured following Higgins, Omer and Phillips (2010) who expand upon Ittner, et al.’s (1997) operationalization of the strategy construct. Five variables are used to create a Strategy_Score: (1) the ratio of research-and-development expenses to sales, (2) the ratio of employees to sales, (3) the market-to-book ratio (MTB), (4) the ratio of advertising expenses to sales, and (5) the PPE intensity ratio (the ratio of property, plant, and equipment (PPE) to lagged assets). Consistent with prior literature, each ratio is calculated using a rolling average of the previous !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! D@ !Industries are based on Fama-French 17 industry classification from Kenneth French’s website: http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/Data_Library/det_17_ind_port.html ! 54! five years. I use the inverse of the PPE intensity ratio so that for all variables higher scores are associated with a Prospector strategy and lower scores are associated with a Defender strategy. Each ratio used to compute the Strategy Score is intended to capture a different element of firm strategy. The R&D/Sales ratio captures a firm’s innovation (or its’ “propensity to search for new products”) and the Employee/Sales ratio captures efficiency. MTB proxies for the firm’s growth opportunities, while the Advertising Expense/Sales ratio and PPE intensity ratio are intended to capture the firm’s emphasis on sales and marketing and focus on capital assets, respectively. I subsequently rank each of the variables by forming quintiles within each 2-digit SIC industry- year. Firms in the highest quintile receive a score of 5; firms in the second highest quintile receive a score of 4, etc. For each firm-year, the Strategy Score is the sum of the rankings across all five variables such that the highest (lowest) score is 25 (5). Firms with strategy scores at the low end of the continuum are classified as Prospector firms and those at the high end of the continuum are classified as Defender firms. Financial_Distress is an indicator variable set to 1 if a firm has experienced two consecutive years of negative earnings and 0 otherwise. Size is measured as the number of firm employees reported in Compustat. Decentralization is measured as the number of business segments reported in Compustat. 4.2.2 Univariate Tests of Performance Differences Between BSC-User and Non-User Firms To test the overall difference in financial performance between the BSC-User firms and Non-User firms, I perform a t-test of the difference in the means between the treatment and control groups. This test does not control for additional factors that may affect performance, such as firm size and fiscal year, which are subsequently controlled for in the OLS regressions. ! 55! 4.2.3 Difference-in-Difference Model to Test the Performance Impact of BSC Use To test the performance consequences of BSC use (hypothesis 2), I use a difference-in- difference design to compare the change in financial performance before and after BSC adoption for both BSC-user and Non-user firms. The model is as follows: Performance it = ! + " 1 BSC_User it + " 2 BSC_User it * Post_Adoption it + " 3 Size it + " 4 Strategy_Score it + Year_Fixed_Effects (2) In this model, BSC_User is an indicator variable set to 1 for BSC-user firms and 0 for the non-user firms (1), and Size is measured as the log of total assets, instead of the number of firm employees. While it is hypothesized the that size, measured as the number of firm employees, will positively impact its’ likelihood to adopt BSC, size, measured as a firm’s total assets, is used as a control variable in assessing the effect of BSC usage on firm financial performance as firm size has been shown in prior research to be positively impact associated with firm performance. Fiscal year is included as a control variable. All other independent variables are measured as previously defined. The dependent variable, financial Performance is captured using both accounting-based and market-based measures. Return-on-assets (ROA) and return-on-equity (ROE) are the accounting-based measures of performance and are calculated from Compustat data as net income (loss) divided by total assets, and net income (loss) divided by common equity, respectively. Buy-and-hold stock returns calculated from CRSP monthly returns are the market- based measure of firm performance. Following prior research, I examine current-year (contemporaneous) and one-year-ahead accounting and stock performance (Said, et al. 2003). ! 56! Schneiderman (1999) cautions that the lead-time required for BSC usage to show any measurable improvement in firm performance may far exceed the short window within which most research is conducted. More importantly, Kaplan and Norton (1996) recommend that managers establish targets for scorecard measures that will be achieved three to five years out. Not allowing enough time to elapse before measuring performance likely accounts for the failure to find significant relationships between BSC use and financial performance in some prior studies. To address the concern that one year may not be enough time for positive financial performance effects to be shown and to be consistent with the BSC creators intentions regarding when to expect performance improvements, in addition to contemporaneous and one-year ahead financial performance, I also examine accounting performance two to five years into the future and two-year-ahead thru five-year-ahead stock returns. 4.2.4 OLS Model to Test the Effect of BSC-Mismatch on Financial Performance Ittner, Lanen and Larcker (2002) and Said, et al. (2003) assert that there should be no performance differences between firms if all firms invest in performance measurement systems (and other technologies) at their respective “optimal” levels. Thus, performance differences should be associated with the degree to which firms deviate from “optimal practice.” I also examine the impact of “BSC-mismatch” or “fit” using the following model: Performance it = ! + " 1 Pos_BSC_Residual it + " 2 Neg_BSC_Residual it + " 3 Size it + " 4 Strategy_Score it + Year_Fixed_Effects (3) ! 57! Instead of using an indicator variable (Adopter) to identify firms that use BSC, I include two variables to capture the amount of over- or under-investment in BSC. Pos_BSC_Residual represents over-investment and is equal to the residual from the logit model if the residual is positive (i.e. the firm uses BSC but the predicted probability is less than one), and 0 otherwise. Neg_BSC_Residual represents under-investment and is equal to the residual from the logit model if the residual is negative (i.e. the firm does not use BSC but the predicted probability of use is greater than zero), and 0 otherwise. I expect both over-investment and under-investment in BSC to be associated with lower performance. Thus, I predict the coefficient of Pos_BSC_Residual to be negative and the coefficient of Neg_BSC_Residual to be positive, respectively. ! 58! CHAPTER 5 RESULTS 5.1 Prediction Model for Likelihood to Adopt BSC The results of the logit regressions predicting a firm’s use of Balanced Scorecard (equation 1) are presented in Table 3. The logit regressions are run with all BSC-User identified firms (All Firms), excluding Hall of Fame firms (w/o HOF), with SEC filing identified firms only (SEC Identified), and excluding financial firms (w/o Financials). The models appear well specified as indicated by the Chi-square and ROC values. 28 I find that firms with a higher level of unrecorded value-adding assets (i.e. with large amounts of intangible assets) are more likely to adopt BSC, in the All Firms model and the w/o Financials model. The w/o HOF and SEC Identified models show no association between a firms’ level of intangibles assets and likelihood to adopt BSC. Consistent with prior research, I find that firms with higher strategy scores (i.e. prospector firms) are significantly more likely to adopt BSC than those with low strategy scores (i.e. defender firms) across all models (Said, et al., 2003; Hendricks, et al., 2004; Naranjo-Gil, et al., 2009; Gosselin, 2011). 29 Prior research also indicates that recent poor financial performance is positively associated with the use of BSC (Hendricks, et al., 2004, 2011; Naranjo-Gil, et al., 2009). In contrast with these prior results, across all four models, I find that firms that have experienced recent financial distress are significantly less likely to adopt BSC than other firms. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 28 The ROC (Receiver Operating Characteristic) calculates the percentage of true-positives vs. false-positives determined by the binary logit model. Minimally, the models should perform at 0.5. Higher values indicate better model specification. DB !I run the model using two indicator variables instead of the continuous Strategy_Score variable. Prospector equals 1 if the strategy score is in the range 20-25, and 0 otherwise. Defender equals 1 if the strategy score is in the range 1-10, and 0 otherwise. Results are qualitatively the same. Defender firms are less likely to adopt BSC. ! 59! This may be explained by the fact that firms that have recently experienced losses may be financially constrained and unable to make the financial and human capital investments required to effectively adopt and implement a balanced scorecard performance measurement / strategy management system. Consistent with prior research investigating the effect of size on BSC usage (Hoque and James, 2000; Speckbacher, et al., 2003; Abdel-Kader and Luther, 2008; Joshi, 2001), I find that BSC usage is associated with larger firms. 30 I also find across all models that firms with more business segments (i.e. decentralized firms) are more likely to use BSC. I include industry indicators based on Fama-French 17 industry portfolios in the analysis. I generally find that firms in the following industries are more likely to adopt BSC: Mining & Minerals; Chemicals; Drugs, Soaps, Perfumes & Tobacco; Steel Works, etc.; Transportation; and Utilities. On average, firms in these industries have larger amounts of intangible assets and more business segments than overall averages. In contrast with the proportion of Banking-industry firms in the BSC-User sample, I find that Banks, Insurance & Other Financials are less likely to adopt BSC. Consumer Durables firms, firms in the Automobile and Retail Stores industries are also significantly less likely to adopt BSC than other firms. On average firms in these industries have smaller amount of intangible assets and fewer business segments than overall averages. Differences in the amount of intangible assets and business segments may provide some explanation for the differences in likelihood to adopt BSC for various industries. Predicted values from the logit regressions are used to match BSC-User firms to Non- User firms. Table 3 shows the balance of the five matching covariates (intangible assets, strategy score, financial distress, size and decentralization) between the treatment (BSC-User) group and control (Non-User) group for all four models. The analysis is run for the matching !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! <C !I run the model using the log of total assets as the measure of firm Size instead of the number of firm employees. Results are qualitatively the same. ! 60! year (i.e. the year before the BSC-User firm adopts BSC). Results indicate a good match, with no significant differences between the treatment and control groups for most matching variables irrespective of which model is used to create the predicted value used to perform the matching. 5.2 Descriptive Statistics – BSC-User vs. Non-User Firms Table 4 shows descriptive statistics for BSC-User and Non-User firms for all four matched datasets. On average, BSC-User firms are significantly larger than Non-User firms, with significantly higher total assets and market value of equity, and more employees across all matched datasets. A significant difference in the amount of decentralization between BSC-Users and Non-Users is found only in the analysis using the dataset that includes BSC-User firms identified through SEC Filings (Panel C). BSC-User firms are more decentralized than Non- Users, with mean number of reporting segments of 3.5 and 3.2, respectively. Other datasets, shown in Panels A, B, and D, show no difference in the amount of decentralization between BSC-User and Non-User firms. For 3 of 4 datasets (Panels A, C, and D), there is no significant difference in strategy score between BSC Users and Non-Users. This is primarily due to Strategy Score being used as a matching variable and because firm strategy and the Strategy Score tend to remain stable over time. The difference is statistically significant for the dataset that excludes financial firms. However the average strategy scores are in the middle of the range at 12.8 and 12.9 for BSC-Users and Non-Users, respectively putting the average firm in the Analyzer category (having characteristics of both Prospector and Defender firms) in the Miles and Snow (1978) strategy continuum. Financial distress is significantly higher for BSC-User firms compared to Non-User firms only in the dataset that excludes financial companies (Panel B), suggesting that on average firms that use BSC have experienced more recent net losses than Non-User firms. Though this difference is statistically significant, it is driven by a very small ! 61! number of observations. Differences for the financial distress variable are not significant in the other datasets. On average, intangible assets are higher for BSC-User firms than for Non-User firms. The differences are significantly higher for the datasets including All Firms (Panel A) and excluding Hall of Fame firms (Panel D) and significantly lower for the dataset that excludes financials (Panel B). No significant difference is found for the level of intangible assets in the dataset that includes only BSC-User firms identified in SEC filings (Panel C). 5.3 Effect of BSC Use on Financial Performance 5.3.1 Univariate Tests Table 5 reports the results of univariate tests of the difference in financial performance between BSC-User and Non-User firms. Results are presented for five windows around the year of BSC adoption. The ‘t-1 to t+1’ window includes the year before and the year after BSC adoption. The ‘t-2 to t+2’ window includes two years before and two years following BSC adoption. The same procedure is used to create windows through ‘t-5 to t+5.’ The year of adoption (t=0) is excluded from each window. ROA and ROE are the accounting performance measures and market-based performance is measured using 1-year buy-and-hold stock returns. Panel A presents results for the dataset including all BSC-User firms and shows a significant difference between BSC-Users and Non-Users for ROE in the 5-year window (t-5 to t+5) (p- value=0.010). ROE is also significantly higher for BSC-User firms in the 5-year window for the dataset that excludes Hall of Fame firms (Panel D). No other significant performance differences are shown for BSC-User and Non-Users in Panels A and D. Univariate tests do not reveal any significant differences between BSC-Users and Non-Users for ROA or stock returns in any window using any of the 4 matched datasets. ! 62! Overall, the univariate tests provide very weak support of the assertion that it may take several years for the adoption or implementation of BSC to manifest in improved financial outcomes, H2. Results of regression analyses that include an indicator for fiscal years before and after the adoption of BSC (Post) and additional control variables are presented in the next section. 5.3.2 Impact of BSC Adoption on Financial Performance I use equation 2 to test the hypothesis that use of the Balanced Scorecard improves financial performance. Tables 6 thru 9 report the results of the OLS regressions of BSC use on accounting and market-based financial performance for each dataset. In these regressions, an indicator variable is used as the measure of BSC use. Panels A and B report results using ROA and ROE as the dependent measures. Panel C reports results for the market-based measure, using 1-year buy-and-hold stock returns. As reported in Panel A of Table 6, I find no relationship on average between the use of BSC and performance using ROA as the measure of accounting-based financial performance in the All Firms dataset. Coefficients on the Adopter indicator variable are insignificant in each window. The significantly lower interaction term (Adopter x Post) suggests that ROA is lower for BSC-User firms in subsequent years after BSC adoption. Panel B of Table 6 reports the results using ROE as the measure of financial performance. No main effect is found between the use of BSC and ROE in any of the performance windows. Consistent with the results for ROA in Panel A, I find a significantly negative interacting term (Adopter x Post) suggesting that ROE is lower for BSC-User firms in the years after BSC adoption. Results for regressions using stock ! 63! returns as the dependent variable are reported in Panel C. No significant main effect is found between the use of BSC and stock returns in any of the performance windows. As reported in Table 7, I find no relationship on average between the use of BSC and financial performance using the dataset that excludes financial firms. Coefficients on the Adopter indicator variable are insignificant in each window. I find that BSC-User firms on average have lower accounting-based financial performance (Panels A and B) in the years following BSC adoption than Non-User firms as indicated by the significant negative coefficient on the interaction term (Adopter x Post). Results for the dataset of SEC-filing-identified firms (Table 8) show no relationship on average between the use of BSC and accounting-based or market-based financial performance. Post-BSC-adoption performance is lower for BSC-User firms in the t-2 to t+2 thru t-4 to t+4 windows for ROE (Panel B) and for the t-4 to t+4 window for ROA (Panel A). Table 9 reports results using the dataset of matched firms that excludes the Hall of Fame Firms. No significant relationship is found between BSC use and any measure of financial performance. However, on average, ROA is lower in the post-adoption period for all firms - both BSC-User and Non-User firms – as indicated by the negative and significant coefficient on Post for all regression windows in Panel A. 5.3.3 - Impact of BSC Mismatch on Financial Performance Results for the tests of the impact of BSC mismatch on financial performance are presented in Tables 10 - 13. Analyses for the All Firms, Minus Financials, SEC Only, and Minus HOF datasets are shown in Tables 10 to 13, respectively. As shown in Panel A of Table 10, consistent with predictions, the signs of the coefficients are negative and significant for ! 64! Pos_BSC_Residual in the 2-year thru 5-year regression windows and significant and positive for Neg_BSC_Residual in the 2-year, 3-year and 5-year windows. This suggests that both over- investment and under-investment in the use of BSC will have a negative impact on financial performance. When ROE is used as the measure of accounting performance (Panel B), the coefficients on Pos_BSC_Residual are significant in the 2-year thru 4-year regression windows, however, none of the coefficients on Neg_BSC_Residual are significant. No inferences can be made from the results presented in Panel C, where stock returns are the measure of financial performance, as none of the coefficients on Pos_BSC_Residual or Neg_BSC_Residual are significant. Panel A of Table 11 shows that the coefficient of Pos_BSC_Residual is consistently significant and negative for all five windows when performance is measured as ROA, suggesting that over-investment in the use of BSC is associated with lower accounting performance. The coefficient on Neg_BSC_Residual is consistently positive, but is only significant in the 3-year and 5-year windows. Panel B of Table 11 shows results using ROE as the measures of performance. The coefficient of Pos_BSC_Residual is only significant for the 2-year, 3-year and 4-year windows. Coefficients for Pos_BSC_Residual in regressions for the 1-year and 5-year windows and all coefficients for Neg_BSC_Residual are insignificant. Panel C shows results using stock returns as the measure of financial performance. None of the coefficients for Pos_BSC_Residual or Neg_BSC_Residual are significant in any of the regression windows. Panels A and B of Table 12 indicate that over-investment in the use of BSC is associated with significantly lower accounting performance as shown by the negative and significant coefficients on Pos_BSC_Residual in the 2-year thru 5-year windows. Under-investment in the use of BSC is also associated with significantly lower financial performance in the 3-year to 5- ! 65! year windows for ROA and the 4-year and 5-year windows for ROE as the measures of performance, as indicated by the positive and significant coefficient on Neg_BSC_Residual. Neither over-investment nor under-investment in the use of BSC are associated with financial performance when stock returns are examined, as shown in Panel C. Table 13 presents results using the dataset that excludes Hall of Fame firms. No inferences can be made from results presented in Table 13. When ROA (Panel A) and stock returns (Panel C) are used as the measures of financial performance, none of the coefficients on Pos_BSC_Residual or Neg_BSC_Residual are significant. When ROE (Panel B) is used as the measure of accounting performance, the coefficients on Pos_BSC_Residual consistently have the correct sign, but only the coefficient of the 2-year window is significant. 5.4 Additional Tests 5.4.1 Matching on Variables from Prior Literature Prior research matches BSC-adopter firms to control firms on a single variable, e.g. market value of equity, book-to-market ratio, ROA or total assets (Said, et al., 2003; Crabtree and DeBusk, 2008). To demonstrate that my analysis is robust to different matching specifications, I create four additional datasets matching within +/- 25% of ROA, book-to-market ration, market value of equity and total assets within the same fiscal year and 4-digit SIC code. When a match within +/- 25 cannot be found in the 4-digit SIC code, I attempt a match within the same 3-digit SIC code, then 2-digit SIC code, and finally within 1-digit SIC code. Using this methodology, I am able to match all BSC-User firms. Differences between BSC-User and Non-User firms on the five variables hypothesized to influence a firm’s likelihood of adopting BSC are shown in Table 14. Across all four matching ! 66! specifications, BSC-User firms are more decentralized (i.e. have more reporting segments) than their matched pairs. 31 BSC-User firms are larger in the ROA-matched and Book-to-Market- matched pairs, and Strategy Scores are higher (indicating a Prospector orientation) for the ROA- matched and Total Assets-matched pairs. Results for tests of the impact of BSC use on financial performance for the datasets matched on ROA, book-to-market, market value of equity and total assets are presented in Tables 15-18, respectively. For the ROA-matched dataset, the coefficient on Adopter is only positive and significant in the 3-year window for stock returns (Panel C of Table 15). The coefficients on Adopter are significant for the 2-year thru 5-year windows for ROA (Panel A of Table 15), however the sign of the coefficient is opposite of the expected result (i.e. negative). The coefficients on Adopter are also significantly negative for the 2-year to 5-year windows for ROA (Panel A of Table 16), and the 2-year and 3-year windows for ROE (Panel B of Table 16) in the Book-to-Market-matched dataset. The coefficients on Adopter are not significant in any of the other regressions for the Book-to-Market matched dataset. None of the coefficients on Adopter are positive and significant for the datasets matched on Market Value of Equity or Total Assets (Tables 17 and 18). Results for tests of the impact of BSC mismatch on financial performance are presented in Tables 19-22, respectively. Only the results for the dataset matched on ROA (Table 19) are internally consistent. The coefficients on Pos_BSC_Residual are consistently negative for ROA and ROE and are statistically significant in the 2-year thru 5-year windows (Panels A and B). The coefficients on Neg_BSC_Residual are mostly positive, but it is only significant in the 5- year window when ROA is the measure of financial performance. The results suggest that over- !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! <" !Although the Non-User firm meets the matching requirement of being within +/- 25, the difference between the means of the two groups may be statistically significant. ! 67! investment in the use of BSC is associated with lower financial performance. Coefficients on both variables of interest in the stock return regressions have the incorrect sign and are mostly insignificant. Results using the datasets matched on Book-to-Market (Table 20) are consistent with predictions, negative and significant coefficients on Pos_BSC_Residual in the 2-year thru 5- year windows and 2-year thru 4-year windows for ROA and ROE, respectively (Panels A and B). Results are only consistent with predictions for Pos_BSC_Residual in the 2-year thru 4-year windows for ROE (Panel B of Table 21) and the 5-year window for stock returns (Panel C of Table 21) for the dataset matched on Market Value of Equity. The coefficients on Neg_BSC_Residual are insignificant in all regression windows. The coefficient of Pos_BSC_Residual is only significant in the 5-year window for ROA in the dataset matched on total assets (Panel A of Table 22). All other coefficients on both of the variables of interest are insignificant. This mixed and mostly inconclusive evidence makes it impossible to draw inferences based on these samples. 5.4.2 Hall of Fame Matched Firms - Only Results from the previous tests provide at best mixed and mostly inconclusive evidence regarding the relationship between the use of the Balanced Scorecard and future financial performance. Initially, the Hall of Fame firms were excluded from the analysis to ensure that these firms did not drive the study results. However, if a positive performance effect from the use of BSC should be found anywhere, it should be with the firms that have successfully implemented BSC according to the guidelines set forth by it’s creators and the organization that makes the award. Therefore, I run the same analysis using a much smaller subset of identified BSC-User firms – Hall of Fame firms only. I run a logit model to predict the likelihood BSC ! 68! adoption. Results from the logit regression are presented in Table 23. Consistent with previous results, I find that BSC adoption is positively associated with the amount of intangible assets and decentralization and firm size. The likelihood to adopt BSC is negatively associated with financial distress, likely due to distressed firms being constrained by limited financial and human capital. I do not find that strategy score influences that likelihood that a firm will adopt BSC. The Chi-square (p<0.01) and ROC statistics (0.743) are consistent with the previous models indicating a good model fit. Regression results testing the effect of BSC use on financial performance are presented in Table 24. I find a positive association between BSC use and accounting based performance measures (ROA and ROE) in the 3-year thru 5-year windows (Panels A and B of Table 24) and in the 4-year and 5-year windows for stock returns (Panel C of Table 24). The results from Table 24 provide the strongest evidence in support of the assertions that how BSC is implemented, i.e. with support from top management, tied to compensation, with performance measures linked to specific strategic outcomes, etc., matters and support for the assertion that the short windows in which many change initiatives like BSC are evaluated, may be too narrow for any performance effects to be shown. In the current analysis, positive associations are not manifested until at least three years after adoption of BSC. ! 69! CHAPTER 6 CONCLUSION In this dissertation I propose and empirically test a model to predict the likelihood that a firm will adopt a Balanced Scorecard. I also empirically test whether the use of BSC improves financial performance using contemporaneous and future measures of accounting-based (ROA and ROE) and market-based (stock return) performance. Ittner, Lanen and Larcker (2002) assert that if all firms are optimizing with respect to a particular choice variable, i.e. the use of BSC, there should be no association between performance and BSC use. To address this concern, in addition to testing the performance impact of BSC usage using an indicator variable, I also examine the impact of BSC mismatch on financial performance. BSC mismatch is the over- investment or under-investment in the use of BSC as determined by the residuals from the prediction model of the likelihood of BSC adoption. Results from the prediction model indicate that larger, more decentralized firms are more likely to adopt BSC. Firms that have recently experienced losses and firms pursuing a Prospector strategy are less likely to use BSC. Some industries (e.g. Chemicals, Utilities, and Banks) are more likely to adopt BSC, and others (e.g. Consumer Durables and Retail Stores) are less likely to adopt BSC. Overall, results from the regressions of the use of BSC on financial performance do not support the hypothesis that the use of BSC improves financial performance. In fact, the coefficient on the use of BSC (variable = Adopter) is insignificant in all 60 regressions using the four primary matched dataset. This might suggest that BSC use is not associated with improved ! 70! or increased financial performance. However, finding no significant performance effect for using BSC suggests that the equilibrium condition holds, i.e. if all firms invest in the “right” amount of BSC, then no performance differences should be found. If this is the case, then firms that do not “optimally” invest in the use of BSC should have lower financial performance than those that do make the optimal investment. Overall, results for tests of the mismatch of BSC use on financial performance indicate that investing too much in BSC usage is associated with lower accounting- based financial performance. Additional analysis using a sample of BSC Hall of Fame and matched pairs indicates a positive association between the use of BSC and financial performance that may not be seen for three to four years after BSC adoption. This study has the following practical implications. First, the study shows that the use of multiple criteria matching (e.g. propensity score matching) provides a more balanced set of matched pairs than matching on a single variable. Results using the propensity score matched datasets are stronger and more consistent than results using datasets matched on ROA, market- to-book ration, market value of equity, and total assets. Second, the simple use of BSC does not necessarily improve financial performance as suggested by proponents of using any type of performance management system. Rather, investing in the appropriate level of BSC appears to be more important for financial performance. For example, if the level of BSC use is classified according to Speckbacher, et al.’s (2003) typology of BSC use, then it may not be appropriate for all firms to have Type III BSC implementations where the scorecard contains causal linkages and is tied to incentive compensation. 32 For some firms, a simple scorecard that include both financial and nonfinancial performance measures derived from the organization’s strategy, but without specified cause-effect relationships may be a sufficient performance management system !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! <D !Type I scorecards include financial and nonfinancial performance measures tied to organizational strategy. Type II scorecards include hypothesized cause-effect relationships. Type III scorecards include a tie to incentive compensation in addition to causal linkages. ! 71! that provides the framework to ensure that strategic goals and objectives are being met and to ultimately drive improved financial performance. Additionally, the findings from the additional analysis using the Hall of Fame firms and matched pairs suggests that firms should make investments to ensure a successful BSC implementation, including adopting the principles of the Strategy-Focused Organization. The additional findings also suggest that both practitioners and researchers need to expand the windows for analysis and evaluation as the performance impact of BSC adoption or the adoption of any other innovation may not be seen for several years. 6.1 Contributions and Limitations There are several contributions of this research. First, this research provides a model predicting the likelihood that a firm will adopt BSC that includes objectively measured determinants and a measure of intangible assets that is not included in prior literature. Second, this study provides support for using propensity scores to determine better matches between treatment and control groups. Third, this study provides an empirical test of the performance impact of BSC using a broad sample of objectively identified BSC-User firms rather than the convenience samples used in other studies. Finally, this study provides evidence that deviation from optimal investment in performance measurement/management technology has a significantly negative effect on performance outcomes. Thus, it is extremely important to determine and invest in the “right” amount of any particular technology. While this research makes several contributions, there are also some limitations to this study. First, many firms were identified as BSC-Users through disclosure in an SEC filing. Though this was intended to be an unbiased measure of determining BSC use, it does not capture those firms that use BSC but for some reason choose not to disclose. Second, the measure of ! 72! intangible assets is not significant in many regressions. A better measure may need to be used in future analyses. Third, once a firm discloses that it has used BSC, it is considered a BSC-User for all years in the analyses. The measure of BSC use does not account for any decisions to abandon the use of BSC, which has been documented for some firms. 6.2 Future Directions for Research There are several possible extensions for this research. First, the study could be enhanced by identifying a true control group of firms that are known not to use or not to have used BSC. For example, some firms only use financial performance measures and some use generic financial and nonfinancial performance measures that are not derived from a particular organizational strategy. Second, an investigation of additional determinants of BSC adoption may provide a more robust prediction model. Kaplan and Norton indicate that the most important factor for successful BSC implementation is CEO leadership. Development of a suitable proxy for CEO leadership would strengthen the model. Third, since findings indicate that some industries are more or less likely to adopt BSC, future research may benefit from focus on specific industries. Finally, since it has been shown that it is not whether or not firm uses BSC, but rather whether the firm uses “the right amount” of BSC that affects financial performance, additional research into determining the “optimal” amount of investment in BSC would be beneficial. ! ! 73! REFERENCES Abdel-Kader, M., Luther, R. (2008). The impact of firm characteristics on management accounting practices: A UK-based empirical analysis. The British Accounting Review, 40, 2- 27. Abernethy, M. A., & Bouwens, J. (2005). Determinants of accounting innovation implementation. Abacus, 41(3), 217-240. Abernethy, M. A., Bouwens, J., & Van Lent, L. (2004). Determinants of control system design in divisionalized firms. The Accounting Review, 79(3), 545-570. Albright, T., Burgess, C. M., Davis, S. B., and Juras, P. E. (2007). Making a Balanced Scorecard Program a Success: Beyond the Measure-Selection Process. The Journal of Corporate Accounting and Finance, 18(4) 65. American Accounting Association Committee on Non-Financial Measures of Effectiveness (1971), “Report”, Accounting Review,Vol. 46 No. 4, Supplement, p. 165. Baines, A., & Langfield-Smith, K. (2003). Antecedents to management accounting change: a structural equation approach. Accounting, organizations and society, 28(7), 675-698. Banker, R. D., Chang, H., and Pizzini, M. J. (2004). The Balanced Scorecard: Judgmental Effects of Performance Measures Linked to Strategy. The Accounting Review, 79(1), 1-23. Banker, R. D., Potter, G., & Srinivasan, D. (2000). An empirical investigation of an incentive plan that includes nonfinancial performance measures. The Accounting Review, 75(1), 65- 92. Barber, B. and Lyon, J. (1996). Detecting Abnormal Operating Performance: The Empirical Power and Specification of Test Statistics. Journal of Financial Econonomics. 41, pp. 359- 99. Baron, J., & Hershey, J. C. (1988). Outcome bias in decision evaluation. Journal of personality and social psychology, 54(4), 569-579. Beimann, I., and Johnson, C. C. (2007). Balanced scorecard in developed and transitional economies. Balanced scorecard for state-owned enterprises (pp. 15-14) Asian Development Bank. Bjøornenak, T., and Olson, O. (1999). Unbundling management accounting innovations. Management Accounting Research, 10(4), 325-338. ! ! 74! Bryant, L., Jones, D. A., and Widener, S. K. (2004). Managing Value Creation within the Firm: An Examination of Multiple Performance Measures. Journal of Management Accounting Research, 16,1. Campbell, D., Datar, S., Kulp, S., and Narayanan, V. G. (2008). Testing strategy with multiple performance measures evidence from a balanced scorecard at Store24. Cardinaels, E., & van Veen-Dirks, P. M. (2010). Financial versus non-financial information: The impact of information organization and presentation in a Balanced Scorecard. Accounting, Organizations and Society, 35(6), 565-578. Chenhall R.H. (2003). Management control systems design within its organizational context: Findings from contingency-based research and directions for the future. Accounting, Organizations and Society, 28 (2-3), 127-168. Chenhall, R. H. (2006). Theorizing contingencies in management control systems research. Handbooks of Management Accounting Research, 1, 163-205. Crabtree, A. D., and DeBusk, G. K. (2008). The effects of adopting Balanced Scorecard on shareholder returns. Advances in Accounting, 24(1), 8-15. Davis, S., and Albright, T. (2004). An investigation of the effect of Balanced Scorecard implementation on financial performance. Management Accounting Research, 15(1), 135- 153. De Geuser, F., Mooraj, S., and Oyon, D. (2009). Does the balanced scorecard add value? empirical evidence on its effect on performance. European Accounting Review, 18(1), 93. DeBusk, G. K., and Crabtree, A. D. (2006). Does the balanced scorecard improve performance? Management Accounting Quarterly, 8(1), 44. Dechow, N. (2012). The balanced scorecard: subjects, concept and objects – commentary. Journal of Accounting & Organizational Change, 8(4), 511-527. Dilla, W. N., and Steinbart, P. J. (2005). Relative weighting of common and unique balanced scorecard measures by knowledgeable decision makers. Behavioral Research in Accounting, 17, 43. Duh, R. R., Xiao, J. Z., & Chow, C. W. (2009). Chinese firms' use of management accounting and controls: facilitators, impediments, and performance effects. Journal of International Accounting Research, 8(1), 1-30. Easton, George S. and Sherry L. Jarrell. (1998). The Effects of Total Quality Management on CorporatePerformance: An Empirical Investigation. The Journal of Business, 71(2): 253- 305. ! ! 75! Frigo, Mark L. (2012). The Balanced Scorecard: 20 Years and Counting. Strategic Finance, 49- 53. Gosselin, M. (2011). Contextual factors affecting the deployment of innovative performance measurement systems. Journal of Applied Accounting Research, 12(3), 260-277. Govindarajan, V., & Gupta, A. K. (1985). Linking control systems to business unit strategy: impact on performance. Accounting, Organizations and Society, 10(1), 51-66 Gupta, A. K., & Govindarajan, V. (1984). Business Unit Strategy, Managerial Characteristics, and Business Unit Effectiveness at Strategy Implementation. Academy of Management journal, 27(1), 25-41 Haka, S., Gordon, L., Pinches, G. (1985). Sophisticated capital budgeting selection techniques and firm performance. The Accounting Review 60, 651–669. Hendricks, K., Hora, M., Menor, L., & Wiedman, C. (2012). Adoption of the Balanced Scorecard: A Contingency Variables Analysis. Canadian Journal of Administrative Sciences/Revue Canadienne des Sciences de l'Administration, 29(2), 124-138. Hendricks, K., Menor, L., & Wiedman, C. (2004). The balanced scorecard: to adopt or not to adopt?. Ivey Business Journal, 69(2), 1-7. Higgins, D., T. C. Omer, and J. D. Phillips. (2011). Does a Firm’s Business Strategy Influence its Level of Tax Avoidance? (February 14, 2011). 2011 American Taxation Association Midyear Meeting: JATA Conference. Hoque, Z., and James, W. (2000). Linking balanced scorecard measures to size and market factors: Impact on organizational performance. Journal of Management Accounting Research, 12, 1. Huelsbeck, D. P., K. A. Merchant and T. Sandino. (2011). On Testing Business Models. The Accounting Review 86(5): 1631–1654. Humphreys, K. A., & Trotman, K. T. (2011). The balanced scorecard: The effect of strategy information on performance evaluation judgments. Journal of Management Accounting Research, 23(1), 81-98. Ittner, C. D. (2008). Does measuring intangibles for management purposes improve performance? A review of the evidence. Accounting and Business Research, 38(3), 261-272. Ittner, C. D., Lanen, W. N., & Larcker, D. F. (2002). The Association Between Activity‐Based Costing and Manufacturing Performance. Journal of Accounting Research, 40(3), 711-726. Ittner, C. D., D. F. Larcker and M. V. Rajan. (1997). The Choice of Performance Measures in Annual Bonus Contracts. The Accounting Review 72 (2): 231-255. ! ! 76! Ittner, C. D., and Larcker, D. F. (1998a). Are Nonfinancial Measures Leading Indicators of Financial Performance? An Analysis of Customer Satisfaction. Journal of Accounting Research, 36, 1-35. Ittner, C. D., and Larcker, D. F. (1998b). Innovations in performance measurement: Trends and research implications. Journal of Management Accounting Research, 10, 205-238. Ittner, C. D., Larcker, D. F., and Meyer, M. W. (2003a). Subjectivity and the Weighting of Performance Measures: Evidence from a Balanced Scorecard. The Accounting Review, 78(3) 725-758. Ittner, C., Larcker, D. and Randall, T. (2003b). Performance implications of strategic performance measurement in financial services firms. Accounting, Organizations and Society, 28: 715–741. Joshi, P. L. (2001). The international diffusion of new management accounting practices: The case of India. Journal of International Accounting, Auditing and Taxation 10 (1): 85–109. Kaplan, R. S. (2010). Conceptual foundations of the balanced scorecard. Harvard Business School. Working Paper 10–074. Retrieved from http://hbswk. hbs. edu/item/6395. html. Kaplan, R. S., and Norton, D. P. (1992). The balanced scorecard - measures that drive performance. Harvard Business Review, 70(1), 71. Kaplan, R. S., and Norton, D. P. (1996). Using the balanced scorecard as a strategic management system. Harvard Business Review, 74(1), 75. Kaplan, R. S., and Norton, D. P. (2001a). Transforming the balanced scorecard from performance measurement to strategic management: Part I. Accounting Horizons, 15(1), 87. Kaplan, R. S., and Norton, D. P. (2001b). Transforming the balanced scorecard from performance measurement to strategic management: Part II. Accounting Horizons, 15(2), 147. Kaplan, R.S. and Norton, D.P. (2004), “The strategy map: guide to aligning intangible assests”, Strategy & Leadership, Vol. 32 No. 5, pp. 10-17. Kasurinen, T. (2002). Exploring management accounting change: The case of balanced scorecard implementation. Management Accounting Research, 13(3) 323-343. Kennedy, T. and J. Affleck-Graves. (2001). The Impact of Activity-Based Costing Techniques on Firm Performance. Journal of Management Accounting Research 13: 19-45. Kerlinger, Fred N. (1973). Foundations of Behavioral Research. 2nd edition. Holt, Rinehart and Winston. ! ! 77! Kinney, M. R., and W. F. Wempe. (2002). Further evidence on the extent and origins of JIT’s profitability effects. The Accounting Review 77 (1): 203–45. Lant, T. K., Milliken, F. J., & Batra, B. (1992). The role of managerial learning and interpretation in strategic persistence and reorientation: An empirical exploration. Strategic Management Journal, 13: 585-608. Lawson, R., Desroches, D., & Hatch, T. (2008). Scorecard best practices: design, implementation, and evaluation. Wiley. Lee, C. L., & Yang, H. J. (2011). Organization structure, competition and performance measurement systems and their joint effects on performance. Management Accounting Research, 22(2), 84-104. Lennox, C. S., Francis, J. R., & Wang, Z. (2011). Selection models in accounting research. The accounting review, 87(2), 589-616. Libby, T., Salterio, S. E., and Webb, A. (2004). The balanced scorecard: The effects of assurance and process accountability on managerial judgment. The Accounting Review, 79(4), 1075. Lingle, J. H., and Schiemann, W. A. (1996). From balanced scorecard to strategic gauges: Is measurement worth it? Management Review, 85(3), 56. Lipe, M. G., and Salterio, S. E. (2000). The balanced scorecard: Judgmental effects of common and unique performance measures. The Accounting Review, 75(3), 283. Lipe, M. G., and Salterio, S. (2002). A note on the judgmental effects of the balanced scorecard's information organization. Accounting, Organizations and Society, 27(6), 531. Maiga, A. S., & Jacobs, F. A. (2003). Balanced scorecard, activity-based costing and company performance: an empirical analysis. Journal of Managerial Issues, 283-301. Malina, M. A., Nørreklit, H. S., & Selto, F. H. (2007). Relations among Measures, Climate of Control, and Performance Measurement Models*. Contemporary Accounting Research, 24(3), 935-982. Malina, M. A., and Selto, F. H. (2001). Communicating and controlling strategy: An empirical study of the effectiveness of the balanced scorecard. Journal of Management Accounting Research, 13, 47. Malmi, T. (1999). Activity-based costing diffusion across organizations: an exploratory empirical analysis of Finnish firms. Accounting, Organizations and Society, 24(8), 649-672. Malmi, T. (2001). Balanced scorecards in Finnish companies: a research note. Management Accounting Research, 12(2), 207-220. ! ! 78! Merchant, K. A. (1981). The design of the corporate budgeting system: influences on managerial behavior and performance. Accounting Review, 813-829. Miles, R. E., & Snow, C. C. (1978). Organization strategy, structure and process. New York: McGraw-Hill. Miller, G. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. The psychological review, 63, 81-97. Mooraj, S., D. Oyon and D. Hostettler. (1999). The Balanced Scorecard—a necessary good or an unnecessary evil? European Management Journal 17(5), 481–491. Naranjo-Gil, D., V. S. Maas, and F. G. H. Hartmann. (2009). How CFOs determine management accounting innovation: An examination of direct and indirect effects. European Accounting Review 18 4: 667–695. Neely, A. (2007). ‘Does the balanced scorecard work: an empirical investigation’. Working paper, Cranfield School of Management. Niven, P. (2002). Balanced Scorecard Step by Step: Maximizing Performance and Maximizing Results, John Wiley and Sons, New York, NY. Norreklit, H. (2000). The balance on the balanced scorecard - A critical analysis of some of its assumptions. Management Accounting Research, 11(1), 65. Ohlson, J. A. (1980). Financial ratios and the probabilistic prediction of bankruptcy. Journal of accounting research, 18(1), 109-131. Olve, N., Roy, J. and Wetter,M. (1999). A Practical Guide to Using the Balanced Scorecard, Wiley, Chichester. Porter, M. (1980). Corporate strategy. The Economist. Porter, M. E. (1996). What is strategy?. Harvard Business Review (November – December). Richardson, Sandy. Is the Balanced Scorecard Out of Date? Retrieved December 13, 2011, from http://www.linkedin.com/groups/Is-Balanced-Scorecard-Out-Date- 3613389.S.74906003?view=andsrchtype=discussedNewsandgid=3613389anditem=74906003an dtype=memberandtrk=eml-anet_dig-b_pd-ttl-cnandut=3THL_N6TmjS501 Rigby, D., and Bilodeau, B. (2009). Management tools and trends 2009. Bain and Co. Roberts, M. L., Albright, T. L., & Hibbets, A. R. (2004). Debiasing balanced scorecard evaluations. Behavioral Research in Accounting, 16(1), 75-88. ! ! 79! Said, A. A., HassabElnaby, H. R., and Wier, B. (2003). An Empirical Investigation of the Performance Consequences of Nonfinancial Measures. Journal of Management Accounting Research, 15(1) 193-223. Salterio, S. (2012). Balancing the scorecard through academic accounting research: opportunity lost? Journal of Accounting and Organizational Change, 8(4) 458-474. Schneiderman, A. (1999), ``Why balanced scorecards fail’’, Journal of Strategic Performance Measurement, special edition, pp. 6-11. Silk, S. (1998). Automating the Balanced Scorecard. Management Accounting, 79(11) 38-44. Simons, R. (1987). Accounting control systems and business strategy: an empirical analysis. Accounting, Organizations and Society, 12(4), 357-374. Slovic, P., & MacPhillamy, D. (1974). Dimensional commensurability and cue utilization in comparative judgment. Organizational Behavior and Human Performance, 11(2), 172-194. Snow, C. C., & Hrebiniak, L. G. (1980). Strategy, distinctive competence, and organizational performance. Administrative Science Quarterly, 317-336. Speckbacher, G., Bischof, J., & Pfeiffer, T. (2003). A descriptive analysis on the implementation of balanced scorecards in German-speaking countries. Management accounting research, 14(4), 361-388. Tayler, W. B. (2010). The balanced scorecard as a strategy-evaluation tool: The effects of implementation involvement and a causal-chain focus. The Accounting Review, 85(3), 1095- 1117. Tucker, J. (2010). Selection bias and econometric remedies in accounting and finance research. Journal of Accounting Literature, 29: 31-57. Tuomela, T.-S. (2005). The interplay of different levers of control: A case study of introducing a new performance measurement system. Management Accounting Research, 16(3) 293-320. Udpa, S. C. (1997). The balanced scorecard. The Accounting Review, 72(1), 178. Van Der Stede, W. A., Chow, C. W., & Lin, T. W. (2006). Strategy, choice of performance measures, and performance. Behavioral Research in Accounting, 18(1), 185–205. Westphal, J. D., & Fredrickson, J. W. (2001). Who directs strategic change? Director experience, the selection of new CEOs, and change in corporate strategy. Strategic Management Journal, 22(12), 1113-1137. Wiersma, E. (2009). For which purposes do managers use Balanced Scorecards?: An empirical study. Management Accounting Research, 20(4), 239-251. ! ! 80! Xiong, Y., Su, W., & Lin, T. W. (2008). The use of financial and nonfinancial performance measures in Chinese firms. Journal of cost management, 22(5), 37-46. Zhang, C., Pan, F., & Lin, T. W. (2013). Market Competition and the Use of Performance Measures in Chinese Firms. Journal of Corporate Accounting & Finance, 24(3), 65-70. ! ! ! 81! ! ! 82! ! ! 83! ! ! 84! ! ! 85! ! ! 86! ! ! 87! ! ! 88! ! ! 89! ! ! 90! ! ! 91! ! ! 92! ! ! 93! ! ! 94! ! ! 95! ! ! 96! ! ! 97! ! ! 98! ! ! 99! ! ! 100! ! ! 101! ! ! 102! ! ! 103! ! ! 104! !

Abstract (if available)

Abstract Despite the large number of organizations that claim to have adopted a Balanced Scorecard, few studies examine its impact on financial performance. Using publicly‐traded firms that are identified as Balanced Scorecard users and a propensity‐score matched sample of non‐user firms, this study investigates the financial performance impact of BSC usage. Results indicate that over‐ or under‐investment in BSC leads to weaker financial performance and are strongest for the subset of BSC Hall of Fame firms. The findings suggest that implementation of BSC according to the authors' methodology results in significantly better financial performance.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The effects of accounting performance and professional relationships on promotion, dismissal, and transfer decisions in a conglomerate

PDF

Three essays on distance: examing the role of institutional distance on foreign firm entry, local isomorphism strategy and subsidiary performance

PDF

Slashing liquidity through asset purchases: evidence from collective bargaining

PDF

Do shared auditors improve audit quality? Evidence from banking relationships

PDF

Implied equity duration: international evidence

PDF

Two essays on the mutual fund industry and an application of the optimal risk allocation model in the real estate market

PDF

Essays on information and financial economics

PDF

Two essays on financial crisis and banking

PDF

Workplace organization and asset pricing

PDF

Real asset liquidity and asset impairments

PDF

The effects of curiosity-evoking events on consumption enjoyment

PDF

Essays on the firm and stakeholders relationships: evidence from mergers & acquisitions and labor negotiations

PDF

Addressing a historical mission in a performance driven system: a case study of a public historically Black university engaged in the equity scorecard process

PDF

Incorporating social preferences into incentive system design

PDF

Knowledge, motivation, and organizational influences in aligning learning to business performance results: a study of a promising practice

PDF

Accounting And Reporting For Long-Term Unconsolidated Intercorporate Investments In Domestic Companies

PDF

Shareholder litigation as a disciplining device: evidence from firms' financial policies

PDF

Who manages the firm matters: the incremental effect of individual managers on accounting quality

PDF

Forgiveness: elucidating the underlying psychological processes that foster brand forgiveness and interpersonal forgiveness

PDF

The role of social status and gender in the composition of bully-victim dyads in early adolescence

Asset Metadata

Creator Yancy, Alicia Ann (author)

Core Title The balanced scorecard and long-term financial performance: evidence from publicly traded companies

School Marshall School of Business

Degree Doctor of Philosophy

Degree Program Business Administration

Publication Date 05/07/2014

Defense Date 12/17/2013

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag balanced scorecard,BSC,Hall of Fame,OAI-PMH Harvest,propensity-score,strategy

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Merchant, Kenneth A. (committee chair), Erkens, David H. (committee member), Huey, Stanley J., Jr. (committee member), Lin, Thomas (committee member)

Creator Email alicia_yancy@hotmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-410967

Unique identifier UC11296347

Identifier etd-YancyAlici-2496.pdf (filename),usctheses-c3-410967 (legacy record id)

Legacy Identifier etd-YancyAlici-2496.pdf

Dmrecord 410967

Document Type Dissertation

Format application/pdf (imt)

Rights Yancy, Alicia Ann

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA