Journal of Business Inquiry, Volume 12, 2013

              THE
JOURNAL OF
BUSINESS INQUIRY
Volume 12 2013
The Impact of Product Differentiation on the Collapse of Arthur Andersen
Guy McClain, p. 1
Measuring Energy Efficiency in GCC Countries Using Data Envelopment Analysis
Mohammed A. Alsahlawi, p. 15
An Examination of Great by Choice: Redux
Eric Olsen and James Sena, p. 31
Empirically Revisiting the Learning-by-Exporting Theory
Using Data on Chilean Manufacturing Plants
Ruohan Wu, p. 65
Banking and Monetary Crises: Impacts on Exports of MENA Countries
Mohamed Ben Abdallah and Zouheir Bouchaddakh, p. 81
The Impact of Religion on Corruption
Leila Shadabi, p. 102
Utah Valley University Woodbury School of Business
ISSN 2155-4072 (online)i
Report of the Editor of The Journal of Business Inquiry
For the Year 2013, Volume 12
The Year 2013 was another good year for The Journal of Business Inquiry (JBI). Volume 12 published seven articles. We received many high-quality papers with 24.13 percent of acceptance rate. The articles were written by authors, whose primary affiliations include 30 institutions from 11 countries, including Australia, Bahrain, Bangladesh, India, Iran, Kuwait, Puerto Rico, Qatar, Saudi Arabia, Tunisia and the US. Turnaround time took, with almost 27.58 percent of the editorial decisions, less than or 30 days, 48.27 percent between 31 and 180 days, 24.13 percent, between 181 and 272 days.
On behalf of The Journal of Business Inquiry, I would to thank Professor Diane Tardif of the University of Ottawa for copy editing the articles in this issue and Ann Mecham for her excellent administrative assistance and for formatting the articles.
The Journal of Business Inquiry would also like to thank the following individuals, who served as referees over the past year. Some of these individuals refereed more than one paper. Without their help, JBI could not fulfill its mission.
Dr. Adel Al-Shayea: King Saud University
Dr. Maria Arbatskaya: Economics Department, Emory University
Dr. Vaughn Armstrong: Finance and Economics Department, Utah Valley University
Dr. Muhammad Atmeh: German Jordanian University
Dr. Sanjoy Bose: Abu Dhabi University
Dr. Ahmed El-Galfy: Gulf University for Science and Technology Dr. Karim Eslamloueyan: Department of Economics, Shiraz University
Dr. Mamoon Hammad: Department of Civil Engineering, Abu Dhabi University
Dr. James Matthew Hampton: Economics Department, Emory University
Dr. Stan Jenne: Accounting Department, Accounting Department, Utah Valley University
Dr. Michael Jetter: Economics and Finance Department, Universidad EAFIT in Medellin
Dr. Susan Madison: Management Department, Utah Valley University
Dr. Milton Mayfield: Texas A&M International University
Dr. Guy McClain: School of Accountancy, Auburn University
Dr. Gary Mercado: Finance and Economics Department, Utah Valley University
Dr. Kaz Miyagiwa: Florida International University
Dr. Kaushik Mukhopadhaya: Economics Department, Emory University
Dr. Rashid Nikzad: Economics Department, Ottawa University
Dr. Haitham Nobanee: College of Business Administration, Abu Dhabi University
Dr. Andre Oliveira: Finance and Economics Department, Utah Valley University
Dr. Geoffrey Poitras: Faculty of Business Administration, Simon Fraser University
Dr. Manuel Salas Velasco: Department of Applied Economics Campus, University of Granada
Dr. Abdus Samad: Finance and Economics Department, Utah Valley University
Dr. Scott E Seavey: University of Nebraska-Lincoln
Dr. Sheldon Smith: Accounting Department, Utah Valley University ii
Dr. Maritza Sotomayor: Finance and Economics Department, Utah Valley University
Dr. Ramendra Thakur: University of Louisiana at Lafayette
Dr. Hamid Yeganeh: Winona State University
Dr. Chris Westley: Jacksonville State University
Amir Kia
iii
Editorial Board
Publisher: Woodbury School of Business, Utah Valley University
Editor-in-Chief: Amir Kia, Woodbury School of Business, Utah Valley University
Editors:
Maria Arbatskaya, Emory University, USA
Dena Hale, Nova Southeastern University, USA
Karim Eslamloueyan, Shiraz University, Iran
Milton Mayfield, Texas A&M International University, USA
Gary Mercado, Utah Valley University, USA
Kaz Miyagiwa, Florida International University, USA
Haitham Nobanee, Abu Dhabi University, UAE
Geoffrey Poitras, Simon Fraser University. Canada
Ramendra Thakur, University of Louisiana at LafayetteThe Journal of Business Inquiry 2013, 12, 1−14
http:www.uvu.edu/woodbury/jbi/volume12
ISSN 2155-4072
The Impact of Product Differentiation on the Collapse of Arthur Andersen
By GUY MCCLAIN
This study investigates industry specialization during the demise of Arthur Andersen after the Enron scandal. While the literature on industry specialization has centered mainly on the auditing firm’s ability to generate fee premiums and increase audit quality, this paper argues that industry specialization is a method of product differentiation that allows firms to be less imitable and therefore, replaceable. Using multinomial logistic regression from a sample of Fortune 500 firms, I find that Arthur Andersen was not fully differentiated with respect to industry specialization when compared with two of their competitors: Deloitte & Touche and KPMG.
Keywords: Audit Markets, Auditor Replacement, Product Differentiation
JEL Classification: M420
I. Introduction
The demise of Arthur Andersen was an event that rocked the global financial accounting community. So significant was the impact of Andersen’s decline that legislation was enacted through the Sarbanes-Oxley Act of 2002 to restore public confidence in the U.S. financial markets. Most notably, the Sarbanes-Oxley Act restricts the types of services (referred to as management advisory services or non-audit fees) auditors can provide to their public audit clients. It has been argued that auditor dependence on fees (both audit fees and fees from management advisory services) clouds the auditor’s independence and ultimately the judgment auditors use when applying generally accepted accounting principles (Frankel et al., 2002). Because Andersen was so dependent on the fees received from Enron, the fear of losing those fees led them to make suspect judgments or to acquiesce to client demands. While this finding has been debated (Ashbaugh et al., 2003), what remains true is that Andersen is no longer performing audits, and its former competitors have filled the vacancy.
Product differentiation is an important key to having a successful business. Within the context of public accounting, differentiation can be hard to operationalize given the homogenous nature of financial statement audits. Therefore, auditing firms specialize by industry (both nation-wide and city-specific) in order to demand fee premiums (Francis et al., 2005; Ferguson et al., 2006). Product differentiation also acts as protection against competitors as the firm’s product becomes less imitable (Matraves and Rondi, 2007). The reasons can explain why Andersen’s clients dropped the auditing firm prior to the obstruction of justice indictment. For example, Federal Express, Delta Airlines, and Freddie Mac all dropped Andersen prior to the indictment (Day, 2002)
Else School of Management, Millsaps College. E-mail: mcclagm@millsaps.edu. Phone: (601) 974-1270, Fax: (601) 974-1260.
Acknowledgements: I gratefully acknowledge the helpful comments of two anonymous reviewers, Bruce Dixon, participants at the 2012 Southwest Regional Meeting of the American Accounting Association and participants at the 2009 Southeast Regional Meeting of the American Accounting Association. 2 JOURNAL OF BUSINESS INQUIRY 2013
which signaled serious problems for the firm. As such, improved knowledge about Andersen’s specialization within the audit market provides new insights into the surprising and seemingly overnight disappearance of the firm. This paper investigates whether Andersen properly differentiated itself from its competitors by comparing its client characteristics to the client characteristics of its Big 5 competitors: Deloitte and Touche, Ernst and Young, KPMG, and Pricewaterhouse Coopers.
II. Prior Literature
A. The Homogenous Audit Market
Managers of public firms demand audit services in order to provide credibility to their financial statements. This demand is called audit quality, which refers to the likelihood that auditors detect and report financial statement errors or omissions (DeAngelo, 1981). Higher audit quality protects shareholders and managers by decreasing the chance that errors or omissions will exist in the audited financial statements. Previous research has defined audit quality by firm size—Big 5 versus non Big 5 accounting firms (for example, Simunic and Stein, 1987; Francis and Wilson, 1988; Firth and Smith, 1992). Selection of a Big 5 auditor, therefore, is viewed as a signal of financial statement quality. Thus, audit clients are faced with a dilemma: if the clients want a quality audit, must they choose a Big 5 firm? If so, which Big 5 firm should the clients choose?
First, audit clients demand higher audit quality because of agency frictions in the relationship between shareholders and managers. These frictions are the costs incurred to structure, to implement and to monitor contracts related to the performance of managers. As clients become larger (Francis and Wilson, 1988; DeFond, 1992) and more complex (Dopuch and Simunic, 1982), agency costs increase resulting in a need for a higher quality audit. Auditing provides credibility to financial statements and reliability in their role as a monitoring device; thus, auditing helps to mitigate agency costs (Jensen and Meckling, 1976).
Next, audit clients are keenly aware of the risks equity investors incur when investing in the stock of their company: business risk and information risk. Business risk is a measure of general business success or failure. Investors can mitigate this risk by investing in a diversified portfolio of firms. However, diversification can be impaired since it depends on the reliability and adequacy of the financial statement information available to the investor. This risk is known as information risk, the risk or the probability that the financial statements used by the investor are inadequate and unreliable (DeJong and Smith, 1984).
In addition, auditors can be seen as a “deep pocket” (think insurance policy) for any potential lawsuits arising from the performance of the audit and the client’s possible financial failure. Audit firms have three mechanisms to mitigate their litigation: audit design, audit pricing and client selection. Audit design refers to creating and executing a sufficient audit that accurately detects and reports the financial statements. Audit pricing refers to pricing the audit higher than normal to compensate for the increased risk. Simunic and Stein (1987) find that firms do not increase price; they increase the amount of audit work. Client selection refers to selecting clients with as little business risk (potential for failure) as possible.
Audit firms competing within the Big 5 have clients with a similar profile regarding the aforementioned characteristics. In fact, Francis et al., (2013) report the Big 4 (the firms remaining after Andersen’s failure) have a global market share for publicly listed firms of 55 percent and 61 VOL. 12 MCCLAIN: THE IMPACT OF PRODUCT DIFFERERENTIATION 3
ON THE COLLAPSE OF ARTHUR ANDERSEN
percent for firms in the United States. This demand for auditing allows multiple types of audit quality to exist and leads to product differentiation within the Big 5.
B. Industry Specialization
One method used in differentiation is industry specialization. KPMG led the advancement to an industry based audit strategy with its restructuring along industry service lines in 1993 (Hogan and Jeter, 1999). Professional audit standards also stress the importance of understanding a client’s business. Both U.S. standards (the Public Company Accounting Oversight Board’s Audit Standard No. 9, Audit Planning) and the International Standard on Auditing 315, Identifying and Assessing Risks of Material Misstatement through Understanding the Entity and its Environment require the auditor to have a thorough understanding of matters affecting the industry in which the client operates.
Prior research has attempted to identify the reasons why auditors specialize. Early work by Eichenseher and Danos (1981) and Danos and Eichenseher (1986) argues that auditors specialize in industries where the auditor can use economies of scale to reduce the cost of production. They argue these specializations are more common in regulated industries where auditors have to make large investments in industry-specific knowledge. For example, certain types of transactions such as interest rate swaps, long-term leases, and joint ventures are more common in some industries. Other industries such as banking and healthcare have specialized accounting procedures, reporting requirements and internal control systems. In this regard, the auditor’s investments necessary to achieve specialization are similar to those regarding quality in repeat purchase settings (DeAngelo, 1981; Klein and Leffler, 1981).
More recent evidence suggests that auditor specialization has increased in non-regulated industries. Hogan and Jeter (1999) find that auditor specialization in unregulated industries has increased over time during their sample period from 1976-1993. Extending the discussion beyond regulated versus non-regulated industries, Cairney and Young (2006) find auditor specialization in industries where firms have similar operational cost structures. Similarly, Cahan et al., (2008) find that a homogenous investment opportunity set within an industry is positively related to auditor specialization. They refer to this as concentration, not dominance, as firms within a homogenous industry would be reluctant to share an auditor because of concerns regarding the transfer of proprietary information.
O’Keefe et al., (1994) suggest that the provision of audit services to a client includes general, industry-specific, and client-specific knowledge. General knowledge (e.g., knowledge of GAAP and GAAS) and client-specific knowledge are not dimensions of industry specialization. The former is required of all auditors and the latter is acquired as part of a particular engagement and is not transferrable. Gaining industry-specific knowledge and expertise requires considerable investment on the part of auditing firms. Kend (2008) documents Big 5 audit partners’ responses by indicating specialist’s knowledge requires an understanding of the client’s operating environment, key accounting policy issues, business practices, key performance indicators, history, current issues and future direction. This investment, however, allows firms to maximize economies of scales through reductions in costs for technologies, personnel and training (Craswell et al., 1995) as the costs are spread over a larger number of clients.
Industry specialization also produces a higher quality audit which allows industry specialist auditors to charge a fee premium. The evidence on the benefits of auditor specialization is straight forward. First, specialist auditors detect more errors (Owhoso et al., 2002) and make better 4 JOURNAL OF BUSINESS INQUIRY 2013
assessments of risk (Low, 2004). Second, firms employing a specialist auditor are associated with better cash flow predictability (Gramling et al., 1999), higher earnings response coefficients (Balsam et al., 2003), higher analyst evaluations of disclosure quality (Dunn and Mayhew, 2004) and lower instances of financial fraud (Carcello and Nagy, 2004).
The evidence regarding fee premiums, however, is mixed. Early work by Craswell et al., (1995) suggests a fee premium for industry specialists when measured at the national level. Their results, however, are sensitive to the cut-off percentage used to define specialization. In contrast, Ferguson and Stokes (2002) also using Australian data, find no evidence of a premium at the national level. Using city-level industry specialization only adds to the inconsistency. Ferguson et al., (2003) find evidence of a fee premium for auditors identified as specialists at both the city and national level. Thus concluding, both levels of specialization are required to achieve a premium. Francis et al., (2005) find evidence of a city-level fee premium, but the relationship fails in two of their robustness tests. The authors conclude their results are mixed and inconclusive.
Additional work has tested the association between industry specialization and fees in the context of Porter’s (1985) competitive strategy framework. Here, specialization is viewed as a form of differentiation which leads to greater efficiencies and therefore lower production costs. Once again, the empirical results are mixed. Mayhew and Wilkins (2003) find that auditors who are successful in differentiating themselves from their competitors can charge a fee premium. When they are not successful at differentiation they must offer discounts to attract clients. Casterella et al., (2004) using Porter’s (1985) framework find that smaller clients are charged a premium, while larger clients with bargaining power are not charged. Their findings are in contrast with earlier work by Craswell et al., (1995) who find a fee premium for larger clients but not for small companies. They attribute this to large clients having greater agency costs and hence more need for a specialist auditor.
Auditor Switching
While auditing is believed to be a means of reducing agency costs, there is no unifying theory on how companies select a new auditor or weigh the cost/benefit of switching auditors. Prior research suggests three potential costs involved in changing auditors1: switching costs, agency costs and implicit insurance costs (Schwartz and Menon, 1985; Francis and Wilson, 1988; DeFond, 1992; Shu, 2000). I hold the latter constant by only comparing Andersen to other Big 5 auditors. This assumption is consistent with Menon and Williams (1994) who find that the implicit insurance provided by Big N audit firms is relatively equal.
Thus, a client must weigh the cost/benefit of switching costs and agency costs when deciding on changing auditors. Switching costs are the costs incurred by the client for a new audit engagement. These costs typically include the following: cost incurred by the client to educate the auditor about the company’s operations, systems, financial reporting practices and accounting issues, the costs incurred selecting a new auditor and increased risk of audit failure (Blouin et al., 2007).
Consistent with Jensen and Meckling (1976) agency costs are the costs associated with monitoring by the principal, bonding with the agent, and a loss in welfare experienced by the principal because the agent does not always act in the principal’s best interest. Agency problems
1 This is different from audit mergers that brought the auditing profession from the Big 8 to the Big 5. Those mergers were done to increase the ability of the two new firms to compete for large clients (Sullivan, 2002) in an increasingly competitive market along industry lines (Wootton et al., 2003). VOL. 12 MCCLAIN: THE IMPACT OF PRODUCT DIFFERERENTIATION 5
ON THE COLLAPSE OF ARTHUR ANDERSEN
arise when managers have incentives to misallocate or expropriate investor’s funds. An independent audit can weaken these incentives by assuring investors that management is properly reporting in accordance with generally accepted accounting principles (Dopuch and Simunic, 1982; Watts and Zimmerman, 1983). The infrequency of auditor switching (the special case of Andersen with-standing) suggests that the marginal agency benefit gained is significantly less than the cost of switching.
Agency costs typically manifest themselves as changes in client characteristics, which are outside the auditor’s control (Johnson and Lys, 1990; Krishnan and Krishnan, 1997; Hackenbrack and Hogan, 2002). On the other hand, switching costs, relate to the industry and client specific knowledge of the auditor. Prior research on auditor specialization has typically identified industry specialist auditors by using a market share-based approach where a significant share of the industry’s audit fees (or the auditor’s share of total assets or total revenues in an industry) is used to designate a specialist auditor (Hay et al., 2006). This measure can be problematic since an auditor can become an industry specialist in two different ways. That is, the auditor can audit a few large clients within an industry or, alternatively, audit many relatively small clients. Thus, the current market share-based definition embraces two strategies.
Based on Porter (1985) there are two basic competitive strategies: product differentiation and cost minimization. To have the largest firms in an industry as clients, an auditor must develop a high level of technical expertise to deal with the scope and complexity of accounting issues that arise in those types of clients. These auditors differentiate themselves as product specialists. On the other hand, auditors who gain market share by auditing a large number of small clients are more likely to be low-cost specialists. Using companies in the Fortune 500 eliminates the cost minimization strategy to focus on product differentiation.
Industry specialization provides a differentiated service by providing a greater value proposition to the auditor’s clients. The audit should be viewed not as a standardized report, but rather as a process. This process requires efforts by both the auditor and the client. An industry specialist auditor should reduce the client’s effort by reducing the time the client spends explaining industry-specific practices, procedures and trends. Behn et al., (1997) find that industry specialization is a key component of client satisfaction. Furthermore, research provides evidence that the audit process does not simply produce an audit opinion; it produces audited financial statements to which auditors have substantial impact (Kinney and Martin, 1994; Nelson et al., 2002).
Thus, if auditors are specialists, their clients can incur significant costs to change auditors. In the context of Andersen’s decline, previous research investigated the costs associated with changing auditors; specifically, the trade-off between switching costs and agency costs. Chaney and Philipich (2002) find negative market returns for Andersen clients in the three days after Andersen’s admission that documents had been shredded. This result implies that investors had downgraded the quality of an Andersen audit thereby negatively impacting the client’s agency costs. In a direct test of switching versus agency costs, Blouin et al., (2007) find that clients followed their former Andersen team to a new auditing firm when Andersen was an industry specialist, thus reducing switching costs. On the other hand, they also find that clients with greater agency costs were more likely to sever ties with Andersen. Their results are helpful in understanding the costs and benefits weighed by clients when switching auditors. In a similar study, Barton (2005) investigates the timing of client defections from Andersen. Barton (2005) finds that clients defected prior to Andersen’s indictment for criminal misconduct if they were more visible in the capital markets. Measures of agency conflict were not associated with early defections. 6 JOURNAL OF BUSINESS INQUIRY 2013
To summarize the literature on auditor switching relative to Andersen’s criminal indictment implies that once reputational concerns for Andersen arose, the credibility of the financial statements was in doubt. Clients defected because the agency costs simply became too great and switching costs were no longer relevant. Defections prior to the indictment, however, have a different implication. In those situations, clients weighed the agency costs against switching costs and concluded the agency benefit outweighed the switching costs. Perhaps, Andersen had not differentiated its product from its competitors; consequently, there was another auditing firm with comparable technical industry expertise.
This paper’s hypothesis tests the differentiation (industry specialization) of Big 5 auditors prior to the bankruptcy of Andersen. Since Big 4 auditors operate almost exclusively in the market for large publically traded firms (Francis et al., 2013), this theory implies that differentiation takes place on a variable other than client size, risk, or complexity. Rather, differentiation occurs through industry specialization. As it relates to Andersen, lack of industry differentiation from their competitors allowed many of their clients to switch auditors prior to any legal indictment. These notions form the basis for H1:
H1: Andersen is not differentiated by industry from its Big 5 competitors.
III. Sample and Research Design
To explore the issue of auditor differentiation, a sample of 183 firms from the April 15, 2002 Fortune magazine Fortune 500 list are used. The sample firms were selected from Fortune’s listing of selected industries. The industries were judgmentally selected to ensure diversity in services and/or products offered. Within each industry, the largest companies (based on revenues) were selected. Financial and auditor data was collected from Research Insight, and the sample was restricted to a December 31, 2001 year-end, the last year-end in which Andersen performed financial statement audits.
When using a multinomial regression model, one firm must be the reference firm (i.e., the firm with value 0). All other firms are then compared to this base firm. In this study the reference firm is Andersen because Andersen is the firm that went bankrupt and is no longer providing audit services. In addition, the model was also run with Pricewaterhouse Coopers (PWC) as the base firm. PWC was randomly chosen as an opposite to Andersen (i.e., a firm with no regulatory or legal troubles). I compared the Andersen model with the PWC model by comparing the frequencies of actual and predicted outcomes. The Andersen model correctly predicted the same percentage of outcomes as the PWC model, (70 out of 183 or 38.25 percent)
Based on the discussion above, the paper tests whether Andersen is differentiated with respect to its Big 5 competitors using the following multinomial logistic regression model. Multinomial logistic regression is used to predict a dependent variable on the basis of continuous and categorical independent variables. The model is as follows:
AUDITOR = 0 + α1ASSET + α2REV + α3DE +α4OPIN + α5FSUB + α6DSUB + α7D1 + α8D2 + α9D3 + α10D4 +α11D5 +α12D6 + ,
where:
1. AUDITOR= a multinomial variable used to identify each of the Big 5 firms
2. ASSET= the natural log of total assets reported for the year-ended Dec. 31, 2001
3. REV= the natural log of total revenues reported for the year ended Dec. 31, 2001 VOL. 12 MCCLAIN: THE IMPACT OF PRODUCT DIFFERERENTIATION 7
ON THE COLLAPSE OF ARTHUR ANDERSEN
4. DE= the debt-to-equity ratio as reported on Dec. 31, 2001
5. OPIN= a dummy variable for the type of opinion received on the Dec. 31, 2001 financial statements. A 0 is used for an unqualified opinion, a 1 for all other opinion types
6. FSUB= the number of foreign subsidiaries that are consolidated into the financial statements as of Dec. 31, 2001
7. DSUB = the number of domestic subsidiaries that are consolidated into the financial statements as of Dec. 31, 2001
8. D1-D6 = a series of dummy variables used to represent the industry sector in which a company operates. The dummy variables are as follows: D1- the aerospace/defense industry; D2- equipment manufacturing; D3- chemical and petroleum; D4- health care and health related; D5- energy; and D6- telecommunications
IV. Results and Discussion
A. Univariate Results
Table 1 presents the variable means for the sample by industry. Table 2, panel A presents the variable means by auditor. Panel B presents t-tests for mean differences between Andersen and the other Big 5 auditors for the entire sample. Panel B shows that KPMG clients are significantly larger than Andersen clients (p-value 0.004); all other results are not significant. These results suggest that Andersen is not significantly different from its competitors on variables that measure size, agency costs and complexity. Therefore, it is expected that Andersen would differentiate itself through industry specialization which is the purpose of the multinomial logistic regression.
B. Multivariate Results
Table 3 contains the results of the multinomial logistic regression. The model has a chi-square of 28.732 (significance 0.032) suggesting that the final model (with independent variables) is significantly different from the intercept-only model. The McFadden pseudo R2 of 0.230 measures the amount of explained variance in the outcome variable. A McFadden value from 0.2 to 0.4 is considered highly satisfactory (Hensher and Johnson, 1981). The table presents each firm’s results as compared to Andersen. When comparing PWC to Andersen, it is noted that D5 and D6 are significant, p-values .0234 and .0553, respectively. The coefficients on these variables are negative. Thus, one concludes the following: in comparison to Andersen, the probability of PWC auditing a company in the energy or telecommunications industries is less likely than Andersen. Additionally, since the financial services variable is coded 0, it is represented in the constant term. The constant term is negative and significant (p-value .0175). Thus, PWC is less likely to have a financial services client than Andersen. The REV variable is also positive and significant (p-value of .0216). From that p-value, one concludes that PWC clients are more profitable than Andersen’s, but this is more a function of industry specialization than a systematic auditor-client characteristic.
8 JOURNAL OF BUSINESS INQUIRY 2013
Table 1: Descriptive Statistics of Variables by Industry (n=183)
When comparing Andersen to Ernst & Young (EY), the results indicate that D1, D2, and D3 are all positive and significant (p-values of .0275, .0079, and .0430) suggesting that EY is more likely to have clients in the aerospace, equipment manufacturing, and petroleum industries. The ASSET variable is also positive and significant (p-value .0266). This difference may be the result of larger firm sizes for EY specializations than Andersen specializations. Additionally, the FSUB variable is negative and significant (p-value of .0128). It seems counterintuitive that the industries that EY has specialized in, especially petroleum and equipment manufacturing would have fewer foreign subsidiaries than other industries. Thus, this result may be the sign of an auditor characteristic in firm selection.
Mean (standard deviation) Industry Assets (in millions $) Revenue (in millions $) Debt to Equity Foreign subsidiaries Domestic subsidiaries Financial Services 56,914.50 (99,653.49) 10,319.74 (10,068.40) 128.03 (121.28) 13 (45) 81 (143) Aerospace
19,011.64
(11,370.57)
17,261.50
(12,797.83)
57.24
(28.86)
31
(37)
95
(96) Manufacturing 28,798.01 (75,182.99) 20,660.68 (42,623.73) 76.48 (38.50) 107 (98) 153 (140) Chemical
17,370.38
(30,321.34)
23,055.75
(41,279.55)
46.11
(16.37)
70
(101)
115
(143) Health Care 15,978.08 (20,500.44) 14,517.18 (10,980.19) 52.92 (106.20) 44 (84) 223 (430) Energy
18,516.03
(17,814.16)
24,510.40
(22,361.37)
68.97
(21.88)
24
(54)
85
(130) Telecommunications 54,567.66 (51,970.93) 20,364.63 (21,097.22) 63.55 (23.76) 26 (54) 113 (129) VOL. 12 MCCLAIN: THE IMPACT OF PRODUCT DIFFERERENTIATION 9
ON THE COLLAPSE OF ARTHUR ANDERSEN
Table 2: Descriptive Statistics of Variables by Auditor (n=183)
Consistent with my predictions are the non-significant differences when comparing Andersen to Deloitte & Touche (DT) and KPMG. Thus, when the legal troubles ensued for Andersen, their clients and the market, in general, likely viewed them as easily imitable and thus replaceable. The assumption is that Andersen did not adequately protect its product from its competitors. This result and prediction are consistent with the data in Hoitash et al., (2007) and Barton (2005). Hoitash et al., (2007) indicate that as of February 2007, 280 of the 540 (51.8 percent) Andersen clients who had switched firms had switched to either DT or KPMG, while Barton (2005) indicates 12.3 percent of former Andersen clients chose a non-Big 5 auditor.
Panel A
Mean (standard deviation) Auditor Assets Revenues Debt to Equity Foreign Subsidiaries Domestic Subsidiaries Arthur Andersen $20,538.52 (24,356.57) $11,661.67 (10,308.37) 82.44 (90.24) 45.47 (81.56) 110.31 (121.69) Ernst & Young
29,675.60
(36,632.39)
11,875.87
(12,548.51)
79.45
(77.34)
21.00
(41.58)
134.11
(364.25) Deloitte & Touche 32,476.67 (53,686.49) 19,231.77 (30,341.34) 89.01 (74.69) 34.69 (86.46) 88.85 (132.24) KPMG
71,516.29 (102,114.44)
12,511.07 (10,476.09)
131.11
(179.46)
42.14
(96.51)
121.71
(159.02) Pricewaterhouse Coopers 41,836.96 (101,590.04) 24,758.48 (37,409.33) 63.18 (45.28) 57.64 (76.88) 120.94 (155.05) Panel B
t-test of Differences (p-value) Andersen–EY -1.325 (.189) -.084 (.933) .158 (.874) 1.646 (.104) -.399 (.691) Andersen- Deloitte -1.302 (.196) -1.526 (.131) -.355 (.723) .578 (.565) .761 (.449) Andersen-KPMG -3.046 (.004) -.273 (.786) -1.355 (.181) .126 (.900) -.281 (.780) Andersen-PWC -1.325 (.189) -2.196 (.031) 1.293 (.199) .565 (.468) -.358 (.721) 10 JOURNAL OF BUSINESS INQUIRY 2013
Table 3. Multinomial Logistic Regression Results (n=183) Coefficient (p-value)
V. CONCLUSION
In this study, a sample of Fortune 500 companies is examined regarding the differentiation of Andersen relative to its Big 5 competitors prior to their indictment for obstruction of justice. Results indicate little difference between Andersen and its competitors when comparing along measures of size, agency cost and complexity. Differences are discovered when testing along industry specialization. Specifically, Andersen is more likely than PWC to audit a firm in the
Variables
PWC
EY
DT
KPMG Constant -6.8299 (.0175)* -4.0617 (.2054) 2.6398 (.3000) -.6844 (.8845)
ASSET
.0299
(.8799)
.8275
(.0266)*
.1925
(.4281)
-.4200
(.8648) REV .8424 (.0216)* -.5052 (.3142) .5009 (.1905) -.6246 (.9086)
DE
-.0065
(.2084)
-.0013
(.6830)
.0004
(.9052)
.0036
(.1791) OPIN .0644 (.9275) .2866 (.7117) .0836 (.9075) 1.3394 (.1120)
FSUB
-.3363
(.2926)
-.0163
(.0128)*
-.0051
(.1976)
.0005
(.9096) DSUB -.0007 (.7453) .0011 (.4565) -.0005 (.8369) .0011 (.5581)
D1
.1523
(.9062)
3.0256
(.0275)*
.5599
(.6806)
.6857
(.6666) D2 .1327 (.8707) 2.9219 (.0079)* .5988 (.5050) -.7496 (.5964)
D3
.2486
(.7817)
2.3162
(.0430)*
.5237
(.5928)
.5990
(.6107) D4 -.4163 (.6056) 1.54440 (.1447) -.9465 (.3500) -.2578 (.8158)
D5
-2.8244
(.0234)*
-28.0608
(1.0000)
-.4028
(.8670)
-1.0543
(.4213) D6 -1.9558 (.0553)* -.0744 (.9366) -.8829 (.3304) -.70892 (.5815)
McFadden R2 = .230
Chi-square = 28.732 (p-value of 0.032)
*significant at the .05 level VOL. 12 MCCLAIN: THE IMPACT OF PRODUCT DIFFERERENTIATION 11
ON THE COLLAPSE OF ARTHUR ANDERSEN
energy and telecommunications industry, while EY is more likely than Andersen to audit a firm in the aerospace, equipment manufacturing and petroleum industries. There were no differences between Andersen and either DT or KPMG. This result is consistent with the auditor switching results of Hoitash et al., (2007) who find that over half of Andersen’s former clients switched to DT and KPMG and Barton (2005) who finds that 12.3 percent of former Andersen clients switched to a non-Big 5 firm.
Audit firms have incentives to perform audit services for clients that are not easily replicated (Porter, 1985). The results from this study provide modest evidence to suggest that industry specialization (or lack thereof in Andersen’s case) allowed competitors to easily court Andersen’s clients prior to Andersen’s obstruction of justice conviction, as reported by Day (2002) in which FedEx, Delta Airlines and many other larger clients switched firms. Generalizing these results to the audit market as a whole is problematic since the data came from the largest publicly traded companies in the U.S., where the Big 5 (now Big 4) overwhelmingly dominate the market. Perhaps industry specialization at the lower end of the size continuum will yield different results given the impact size plays on audit efficiency and effectiveness.
The results from the present study warrant the following conclusion: in the large audit market, Arthur Andersen did not sufficiently differentiate itself from its competitors. This is consistent with the theory that product differentiation acts a protection against competitors.
References
Ashbaugh, Hollis, Ryan LaFond, and Brian W. Mayhew. 2003. “Do Nonaudit Services Compromise Auditor Independence? Further Evidence.” The Accounting Review, 78(3): 611-39.
Balsam, Steven, Jagan Krishnan, and Joon S. Yang. 2003. “Auditor Industry Specialization and Earnings Quality.” Auditing: A Journal of Practice and Theory, 22(2): 71-97.
Barton, Jan. 2005. “Who Cares about Auditor Reputation?” Contemporary Accounting Research, 22(3): 549-86.
Behn, Bruce, Joseph Carcello, Dana Hermanson, and Roger Hermanson. 1997. “The Determinants of Audit Client Satisfaction among Clients of Big 6 Firms.” Accounting Horizons, 11(1): 7-24.
Blouin, Jennifer, Barbara Murray Grein, and Brian R. Rountree. 2007. “An Analysis of Forced Auditor Change: The Case of Former Arthur Andersen Clients.” The Accounting Review, 82(3): 621-50.
Cahan, Steven, Jayne Godfrey, Jane Hamilton, and Debra C. Jeter. 2008. “Auditor Specialization, Auditor Dominance, and Audit Fees: The Role of Investment Opportunities.” The Accounting Review, 83(6): 1393-1423.
Cairney, Timothy D., and George R. Young. 2006. “Homogenous Industries and Auditor Specialization: An Indication of Production Economies.” Auditing: A Journal of Practice and Theory, 25(1): 49-67.
Carcello, Joseph V., and Albert L. Nagy. 2004. “Client Size, Auditor Specialization and Fraudulent Financial Reporting.” Managerial Auditing Journal, 19(5): 651-68.
Casterella, Jeffrey R., Jere R. Francis, Barry L. Lewis, and Paul L. Walker. 2004. “Auditor Industry Specialization, Client Bargaining Power and Audit Pricing.” Auditing: A Journal of Practice and Theory, 23(1): 49-66.
Chaney, Paul K., and Kirk L. Philipich. 2002. “Shredded Reputation: The Cost of Audit Failure.” Journal of Accounting Research, 40(4): 1221-245. 12 JOURNAL OF BUSINESS INQUIRY 2013
Craswell, Allen T., Jere R. Francis, and Stephen L. Taylor. 1995. “Auditor Brand Name Reputation and Industry Specializations.” Journal of Accounting and Economics, 20(3): 297- 322.
Danos, Paul, and John W. Eichenseher. 1986. “Long-term Trends toward Seller Concentration in the U.S. Audit Market.” The Accounting Review, 61(4): 633-50.
Day, Kathleen. 2002. “Riggs National, FedEx Drop Andersen as Auditor.” The Washington Post, March 12.
DeAngelo, Linda Elizabeth. 1981. “Auditor Independence, ‘Low Balling’, and Disclosure Regulation.” Journal of Accounting and Economics, 3(2): 113-27.
DeFond, Mark L. 1992. “The Association between Changes in Client Firm Agency Costs and Auditor Switching.” Auditing: A Journal of Practice and Theory, 11(1): 16-31.
DeJong, Douglas V., and John H. Smith. 1984. “The Determination of Audit Responsibilities: an Application of Agency Theory.” Auditing: A Journal of Practice and Theory, 4(1): 20- 34.
Dopuch, Nicholas, and Dan A. Simunic. 1982. “Competition in Auditing: An Assessment.” In Fourth Symposium on Auditing Research, 401- 50. University of Illinois.
Dunn, Kimberly A., and Brian W. Mayhew. 2004. “Audit Firm Industry Specialization and Client Disclosure Quality.” Review of Accounting Studies, 9(1): 35-58.
Eichenseher, John W., and Paul Danos. 1981. “The Analysis of Industry-Specific Auditor Concentration: towards an Explanatory Model.” The Accounting Review, 56(3): 479-92.
Ferguson, Andrew, and Donald Stokes. 2002. “Brand Name Audit Pricing, Industry Specialization and Leadership Premiums Post-Big 8 and Big 6 Mergers.” Contemporary Accounting Research, 19(1): 77-110.
Ferguson, Andrew, Jere R. Francis, and Donald J. Stokes. 2003. “The Effects of Firm-Wide and Office-Level Industry Expertise on Audit Pricing.” The Accounting Review, 78(2): 429-48.
Ferguson, Andrew, Jere R. Francis, and Donald J. Stokes. 2006. “What Matters in Audit Pricing: Industry Specialization or Overall Market Leadership?” Accounting and Finance, 46(1): 97-106.
Firth, Michael, and Andrew Smith. 1992. “Selection of Auditor Firms by Companies in the New Issue Market.” Applied Economics, 24(2): 247- 55.
Francis, Jere R., and Earl R. Wilson. 1988. “Auditor Changes: A Joint Test of Theories Relating to Agency Costs and Auditor Differentiation.” The Accounting Review, 63(4): 663-83.
Francis, Jere R., Paul N. Michas, and Scott E. Seavey. 2013. “Does Audit Market Concentration Harm the Quality of Audited Earnings? Evidence from Audit Markets in 42 Countries.” Contemporary Accounting Research, 30(1): 325-55.
Francis, Jere R., Kenneth Reichelt, and Dechun Wang. 2005. “The Pricing of National and City-Specific Reputations for Industry Expertise in the U.S. Audit Market.” The Accounting Review, 80(1): 113-36.
Frankel, Richard M., Marilyn F. Johnson, and Karen K. Nelson. 2002. “The Relation between Auditors’ Fees for Nonaudit Services and Earnings Management.” The Accounting Review, 77(S-1): 71-105.
Gramling, Audrey A., Van E., Johnson, and Inder K. Khurana. 1999. “The Association between Audit Firm Industry Experience and Financial Reporting Quality.” Georgia State University Working Paper.
Hackenbrack, Karl E., and Chris E. Hogan. 2002. “Market Response to Earnings Surprises Conditional on Reasons for an Auditor Change.” Contemporary Accounting Research. 19(2): 195-223. VOL. 12 MCCLAIN: THE IMPACT OF PRODUCT DIFFERERENTIATION 13
ON THE COLLAPSE OF ARTHUR ANDERSEN
Hay, David C., W. Robert Knechel, and Norman Wong. 2006. “Audit Fees: A Meta-Analysis of Supply and Demand Attributes.” Contemporary Accounting Research, 23(1): 141-91.
Hensher, David, and Lester Johnson. 1981. Applied Discrete Modeling. London: Croom Helm.
Hogan, Chris E., and Debra C. Jeter. 1999. “Industry Specialization by Auditors.” Auditing: A Journal of Practice and Theory, 18(1): 1-17.
Hoitash, Rani, Ariel Markelevich, and Charles A. Barragato. 2007. “Auditor Fees and Audit Quality.” Managerial Auditing Journal, 22(8): 761-86.
Jensen, Michael C., and William H. Meckling. 1976. “Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure.” Journal of Financial Economics, 3(4): 305-60.
Johnson, W. Bruce, and Thomas Lys. 1990. “The Market for Audit Services: Evidence from Voluntary Auditor Changes.” Journal of Accounting and Economics, 12(1-3): 281-308.
Kend, Michael. 2008. “Client Industry Audit Expertise: Towards a Better Understanding. Pacific Accounting Review, 20(1): 49-62.
Kinney, William R., and Roger D. Martin. 1994. “Does Auditing Reduce Bias in Financial Reporting? A Review of Audit-Related Adjustment Studies.” Auditing: A Journal of Theory and Practice, 13(1): 149-56.
Klein, Benjamin, and Keith B. Leffler. 1981. “The Role of Market Forces in Assuring Contractual Performance.” Journal of Political Economy, 89(4): 615-41.
Krishnan, Jagan, and Jayanthi Krishnan. 1997. “Litigation Risk and Auditor Resignations.” The Accounting Review, 72(4): 539-60.
Low, Kin-Yew. 2004. “The Effects of Industry Specialization on Audit Risk Assessments and Audit-Planning Decisions.” The Accounting Review, 79(1): 201-19.
Matraves. Catherine, and Laura Rondi. 2007. “Product Differentiation, Industry Concentration and Market Share Turbulence.” International Journal of the Economics of Business, 14(1): 37-57.
Mayhew, Brian W., and Michael S. Wilkins. 2003. “Audit Firm Industry Specialization as a Differentiation Strategy: Evidence From Fees Charged to Firms Going Public.” Auditing: A Journal of Practice and Theory, 22(2): 33-52.
Menon, Krishnagopal, and David Daou Williams. 1994. “The Insurance Hypothesis and Market Prices.” The Accounting Review, 69(2): 327-42.
Nelson, Mark N., John A. Elliott, and Robin L. Tarpley. 2002. “Evidence from Auditors about Managers’ and Auditors’ Earnings Management Decisions.” The Accounting Review, 77(S-1): 175-202.
O’Keefe, Terrence B., Dan A. Simunic, and Michael T. Stein. 1994. “The Production of Audit Services: Evidence from a Major Public Accounting Firm.” Journal of Accounting Research, 32(2): 241-61.
Owhoso, Vincent E., William F. Messier Jr., and John G. Lynch Jr. 2002. “Error Detection by Industry Specialized-Teams During Sequential Audit Review.” Journal of Accounting Research, 40(3): 883-900.
Porter, Michael E. 1985. Competitive Advantage: Creating and Sustaining Superior Performance. New York: Free Press.
Schwartz, Kenneth B., and Krishnagopal Menon. 1985. “Auditor Switches by Failing Firms.” The Accounting Review, 60(2): 248-61.
Shu, Susan Zhan. 2000. “Auditor Resignations: Clientele Effects and Legal Liability. Journal of Accounting and Economics, 29(2): 173-205. 14 JOURNAL OF BUSINESS INQUIRY 2013
Simunic, Dan A., and Michael Stein. 1987. “On the Economics of Product Differentiation in Auditing.” In Proceedings of the 1986 Touche Ross/University of Kansas Symposium on Auditing Problems: 69-99, University of Kansas.
Sullivan, Mary W. 2002. “The Effect of the Big Eight Accounting Firm Mergers on the Market for Audit Services.” Journal of Law and Economics, 45(2): 375-99.
Watts, Ross L., and Jerold L. Zimmerman. 1983. “Agency Problems, Auditing and the Theory of the Firm: Some Evidence.” Journal of Law and Economics, 26(3): 613-33.
Wootton, Charles W., Carel M. Wolk, and Carol Normand. 2003. “An Historical Perspective on Mergers and Acquisitions by Major U.S. Accounting Firms.” Accounting History, 8(1): 25-60.
The Journal of Business Inquiry 2013, 12, 15−30
http:www.uvu.edu/woodbury/jbi/volume12
ISSN 2155-4072
Measuring Energy Efficiency in GCC Countries
Using Data Envelopment Analysis
By MOHAMMED A. ALSAHLAWI
Energy efficiency is a key issue in determining the direction of global concern about preserving the environment. The Gulf Cooperation Council (GCC) is an emerging economic block of six countries with abundant oil reserves and high energy consumption. However, satisfying high levels of energy efficiency in GCC countries might hinder their fast growing economies. Therefore, it is important to measure energy efficiency of GCC countries in order to set the appropriate policies without adverse effects on their economic development strategies. This paper is the first attempt to measure energy efficiency in GCC countries using two models of Data Envelopment Analysis (DEA). As a complementary step, energy intensity for GCC countries and causality of the relationship between energy consumption and economic growth have been tested. The results indicate several policy implications with regard to energy conservation and efficient use of energy.
Keywords: GCC, Energy Efficiency, Energy Intensity
JEL Classification: C54, C81, Q48
I. Introduction
International concern over environmental issues such as global warming and climate change has put severe political and economic pressures on governments of both developed and less developed countries. Energy efficiency is one of the relevant targets to be met by international environmental standards. Therefore, improvement of energy efficiency is one of the most important objectives for any energy policy, especially for countries with high dependency on imported energy. It also exerts political pressures to deal with the climate change challenge (Al-Mansour, 2011).
The objective of improving energy efficiency is not only for environmental benefits, such as reducing CO2 emissions, but also for attaining commercial, industrial competitiveness and energy security. However, the issue in measuring energy efficiency performance is to define the term “energy efficiency” (Patterson, 1996). Different definitions of energy efficiency would lead to different indicators being used to monitor changes in energy efficiency. In this paper we employ two models of DEA to measure and compare energy efficiency in GCC countries.
The problem of increasing energy efficiency in GCC countries by reducing energy consumption might slow down their economic growth where most GCC countries depend on fossil fuels, in particular oil. Thus, analyzing the relationship between the economic growth as presented by gross domestic product (GDP) and energy consumption is very important in setting energy policies.
Professor of Energy Economics, Department of Finance and Economics, College of Industrial Management, King Fahd University of Petroleum and Minerals, P.O. Box 1617, Dhahran 31261, Saudi Arabia. E-mail: msahlawi@kfupm.edu.sa. Phone: 966-3-860-2668, Fax: 966-3-860-2772. 16 JOURNAL OF BUSINESS INQUIRY 2013
The relationship between GDP and energy consumption has been of interest to many researchers in energy economics (Hannesson, 2009). This interest has been stimulated by the unprecedented oil price increase of the early 1970s, which substantially increased the energy bill in oil-importing countries (Al-Iriani, 2006). Different studies have focused on different countries over different time periods to examine the relationships between energy and other macroeconomic variables (Haji and Said, 2011). To assess the effect of energy conservation policies on economic growth, the direction of the causality relationship between the GDP and energy consumption is usually tested (Sa’ad, 1010). In this paper, we will test the causality between these two variables for GCC countries to draw some policy implications.
As far as energy efficiency is concerned, this paper attempts to measure the energy efficiency in GCC countries using two different models of DEA and utilizes the results of the causality test between GDP and the energy consumption for each GCC country and for the GCC as a panel.
II. General Review of GCC Economies
GCC countries share several homogeneous aspects. They have the same language, culture, history and similar economic challenges which emphasize the need for economic integration and diversification. The GDP for the total GCC has increased from US$245 billion in 1980 to US$1,384 billion in 2011 as presented in Table 1.
Table 1: Major Development in GCC Countries, Economic Indicators (Billion US dollar)
Economic Indicator
UAE
Bahrain
Saudi Arabia
Oman
Qatar
Kuwait
2008
2011
2008
2011
2008
2011
2008
2011
2008
2011
2008
2011
Nominal GDP
254.4
366.2
21.9
25.8
472.3
592.5
60.3
67.9
100.4
173.5
148.2
158.0
Real GDP growth %
7.4
3.5
6.3
2.2
4.2
7.1
12.8
5.0
13.4
14.1
8.5
4.8
Interest rates
on US Dollar
(3 months)
2.2
0.3
2.4
0.3
2.2
0.3
2.9
0.7
2.2
0.3
2.9
0.7
Inflation rate
12.3
0.9
3.5
-0.4
9.9
5.0
11.8
4.0
15.1
1.9
10.6
4.7
Money supply*
29.1
1.6
20.8
5.3
17.6
13.3
23.1
12.2
21.9
27.5
15.9
8.5
Imports (FOB)
176.3
198.5
14.2
12.1
100.6
119.1
20.7
21.3
25.1
22.2
22.9
20.7
Exports (FOB)
239.2
295.5
17.3
19.9
313.4
364.6
37.7
47.1
54.9
111.4
78.0
95.1
Current account
22.3
52.3
2.3
3.3
132.3
158.5
5.5
9.6
14.2
51.4
64.8
62.7
Ratio of current account balance in GDP**
8.8
14.3
10.3
12.6
27.9
26.8
9.1
14.2
14.1
29.6
43.7
39.7
Government balance (% of GDP)**
136.0
7.4
7.5
-2.3
32.6
13.1
6.0
3.7
10.5
7.1
28.5
33.6
Population (million)
5.6
4.8
1.1
1.4
24.8
28.1
2.9
2.8
1.4
1.8
3.4
2.8
*M2 in Oman represents broad money supply, while M3 represents broad money supply in the remaining GCC countries.
**Ratio of surplus/deficit to GDP (at current prices)
Source: Saudi Arabian Monetary Agency, 2010 and 2012.
GCC countries excluding Bahrain are major producers and exporters of oil. They depend on oil revenue for their economic development as shown in Table 2. VOL. 12 ALSAHLAWI: MEASURING ENERGY EFFICIENCY IN GCC COUNTRIES 17
USING DATA ENVELOPMENT ANALYSIS
Table 2: World Oil and Gas Reserves, Year End 2010
Region
Oil Reserves
Gas Reserves
Thousand Million bbls
%
Trillion CM
%
Americas
313.8
22.7
17.3
9.2
Europe and Eurasia
139.7
10.1
63.1
33.7
Middle East–of which:
752.5
54.4
75.8
40.5
(Gulf Region)
749.2
54.2
74.8
40.0
Africa
132.1
9.6
14.7
7.9
Asia Pacific
45.2
3.3
16.2
8.7
Total World
1383.2
100
187.1
100
Source: British Petroleum (BP) Statistical Review, 2011.
Recently, GCC countries have embarked on plans for developing their non-energy sectors such as manufacturing and developing renewable energy sources. Such a policy has been driven by the international search for energy alternatives to oil. A fast demographic change is another factor that puts pressure on GCC countries to diversify their economies from oil as the main source for economic development to meet employment aspirations.
The industrial development in the GCC is concentrating on industries, such as basic petrochemicals, fertilizers, and steel, as well as aluminum and non-durable consumer goods. Most of these industrial products are for export and are primarily energy intensive. However, to reach an effective diversification, GCC countries need to integrate economically and politically.
Table 2 shows that GCC countries collectively account for 54.2 percent of world proven oil reserves, and produce more than 21 percent of world crude oil production. As shown in Table 2, GCC countries have around 40 percent of world natural gas reserves and 9.1 percent of the world natural gas production. The GCC countries enjoy relatively cheap extraction costs of oil and gas, which puts them in an advantageous position relative to other oil and gas producers. The wealth generated by the GCC countries from oil and gas exports allows their economies to reach unprecedented development, with higher standards of living and modern physical infrastructure. However, subsidized domestic oil prices in GCC countries are contributing to high consumption of oil which will lead to high CO2 emissions and reduce their energy efficiency. Therefore, GCC countries have to adopt an energy policy in line with the diversification and integration strategy which improves energy efficiency and economic growth.
III. Energy Efficiency
Energy efficiency generally refers to using less energy to produce maximum output. Patterson (1996) has introduced four types of indicators which can be used in measuring energy efficiency: thermodynamic, physical-thermodynamic, economic-thermodynamic and economic indicators. Each indicator differs in terms of the measurement unit of input and output. In this paper, we used the economic-thermodynamic indicators to compare the level of efficiency of each GCC country relative to one another. The most commonly used indicator to measure aggregate of a nation’s energy efficiency is the energy GDP ratio, or usually called “energy intensity”. Other methods exclude the extraneous factors, such as changes in energy input mix, energy for labor 18 JOURNAL OF BUSINESS INQUIRY 2013
substitution, and changes in structure of economy, from the energy GDP ratio, in order to isolate the underlying technical energy efficiency.
Zhou and Ang (2008) use production framework to measure energy efficiency performance. By utilizing DEA, energy consumption is treated as one of the inputs within the production framework with labor and capital as other established economic inputs. Different energy sources are treated as different inputs so that changes in energy mix could be accounted for in evaluating energy efficiency. The undesirable output of energy input which is CO2 emission is also included to measure the efficiency performance of 21 Organization for Economic Co-operation and Development (OECD) countries.
On the other hand, energy intensity as one of the indicators is used to measure energy efficiency. High energy intensities indicate a high cost of converting energy into GDP. Figures 1 and 2 show the trend of energy efficiency in GCC countries based on their energy intensity as one bloc and as individual countries, respectively.
Figure 1: Energy Intensity of GCC: As a Ratio of Energy/GDP
Figure 2: Energy Intensity of Individual GCC Countries: As a Ratio of Energy/GDP
VOL. 12 ALSAHLAWI: MEASURING ENERGY EFFICIENCY IN GCC COUNTRIES 19
USING DATA ENVELOPMENT ANALYSIS
Figure 2 depicts Bahrain as the most inefficient GCC country in terms of energy intensity as a ratio of energy divided by GDP. Oman was the most efficient in 2001, 2003 and 2005, and it indicated some performance decreases in this efficiency level since its energy intensity was increased.
In 2007 and 2008, Kuwait became the most efficient GCC country with the lowest energy intensity. The UAE and Kuwait had shown good improvements in their energy efficiency in the last 8 years. Qatar had shown the highest level of improvement compared to other GCC countries and reached the most efficient country, Kuwait, in 2008. For Saudi Arabia, there was a moderate level of improvement in the energy efficiency over these time periods.
Many factors influence an economy's overall energy intensity. Such factors are the requirements for general standards of living and weather conditions. It is not typical for particularly cold or hot climates to require greater energy consumption in homes and workplaces for heating or cooling purposes, given the differences in standard of living.
Testing the causality between GDP and energy consumption is very essential for determining energy intensity. Empirical results in oil importing countries have been mixed (Chontanawat et al., 2006 and Hertog and Luciani, 2009). This led to some confusion about the effects that energy conservation policies have on economic growth in both developed and developing countries. The disparity in results has most likely been a product of methodological and data differences.
Al-Iriani (2006) investigated the causality relationship between GDP and energy consumption in the GCC. He utilized the recently developed panel cointegration and causality techniques to test the direction of energy-GDP causality in the GCC. Results indicated a unidirectional causality running from GDP to energy consumption. Evidence shows no support for the hypothesis that energy consumption is the source of GDP growth in the GCC countries. Such results suggest that energy conservation policies have been adopted without much concern about their adverse effects on the economic growth of GCC countries.
Mehrara (2007) examined the causality issue between energy consumption and economic growth for three typical oil-exporting countries: Iran, Kuwait and Saudi Arabia. By using two different test methods, a unidirectional long-run causality from economic growth to energy consumption for Iran and Kuwait and a unidirectional strong causality from energy consumption to economic growth for Saudi Arabia were consistently shown.
Chontanawat et al. (2008) tested causality between energy to GDP for over 100 countries. He found that causality from energy to GDP more prevalent in the developed OECD countries, compared to the developing non-OECD countries. Its implication on policy is that reducing energy consumption aimed at reducing emissions is likely to have greater impact on the GDP of the developed rather than developing world.
IV. Measuring Energy Efficiency
In order to measure energy efficiency, we used two different models. The first model emphasizes more energy intensity, and the second model incorporates other economic factors in the measurement.
20 JOURNAL OF BUSINESS INQUIRY 2013
,
.
A. Model I
The objective is to maximize the DEA efficiency rating for each country as the relative
efficiency of unit j using only energy consumption as input is given by
vxj
wyj
relativeinput value
relative output value

(1)
We define the variables as follows:
w = relative output weight applied to country’s GDP
y = amount of output (GDP) from unit j
v = relative input weight applied to country’s energy consumption
x = amount of input (energy consumption)
j = unit/ country being measured
Looking at the model, we can see that it is a non-linear function in the decision variables.
However, since all quantities are relative to some value, we shall choose to make the denominator
equal to 1. So vxj=1 will be a constraint in the model. Given this, the objective function is modified
to .
The second constraint is that all DEA efficiency ratings must not exceed 1 for all countries,
All weights must be non-negative, shown as resulting in w,v  0.
Then Model I:
Max output = wyj (2)
Subject to vxj = 1
wyl – vxl  0 1=1,2,…..9
w,v  0
l = units (countries) included in the measurement
B. Model II
The objective is to maximize the DEA efficiency rating for each country. The relative
efficiency of unit j using only economic factors as input is given as
(3)
1 1 2 2 3 3
v x j v x j v x j
wyj
Therelative efficiency of unit j
 

where
w = relative output weight applied to country’s GDP
y = amount of output (GDP) from unit j
v1 = relative input weight applied to country’s labor
x1j = amount of input 1 (labor) from unit j
v2 = relative input weight applied to country’s energy consumption
x2j = amount of input 2 (energy consumption) from unit j
v3 = relative input weight applied to country’s capital
x3j = amount of input 3 (capital) from unit j
VOL. 12 ALSAHLAWI: MEASURING ENERGY EFFICIENCY IN GCC COUNTRIES 21
USING DATA ENVELOPMENT ANALYSIS
Following the same steps we did for Model I in establishing the constraints, we came up with
Equation (4) in Model II:
Max output = wyj (4)
 
2
1
i
Subject to vt xtj
  
3
0 1,2, .....9
i
wyl vi xil i
2,v1,v2,v3  0
1=units (countries) included in the measurement
In measuring energy efficiency, we used total energy consumption for each GCC country
without differentiating energy source types because all of these countries depend on fossil fuels,
mainly oil. Based on World Bank data (2012), all GCC countries derive 100 percent of their energy
consumption from fossil fuels, United Arab Emirates, Saudi Arabia and Bahrain use less than
1 percent of alternative sources of energy.
V. Data and Methodology
A. The Data
The data sources which are used in the modeling are secondary. In performing the unit root,
cointegration and causality tests between the GDP and energy consumption, we used time series
data from 1980 to 2009. The GDP data obtained from the IMF (2011) and energy consumption
data was obtained from the World Bank (2012).
In performing the DEA to measure the energy efficiency, the data of 2001, 2003, 2005, 2007
and 2008 is used for three inputs (capital, labor force, and energy consumption) and one output
(GDP). For all input, World Bank (2012) data was used and IMF (2011) data was used for the
output. The description of data is presented in tables 3, 4, and 5.
Table 3: Data Used for Causality Test
Variables Descriptions
Energy consumption
Energy use in (tons of oil equivalent), World
Bank (2012)
GDP
Gross domestic product based on local constant price
(in billions of national currency units), IMF (2011)
World Economic Outlook
22 JOURNAL OF BUSINESS INQUIRY 2013
Table 4: Data Used for DEA Measurement
Variables
Descriptions
Energy consumption
Energy use (ton of oil equivalent), World Bank (2012)
Capital
Gross capital formation (current US$), World Bank (2012)
Labor force
total labor force, World Bank data
GDP
Gross domestic product based on purchasing-power-parity (PPP) valuation of country GDP (billions in current international dollar), IMF (2011) World Economic Outlook
Table 5: Summary of Inputs and Output Variables for Six GCC Countries
and Three Developed Countries, 1980-2009
Variables
2001
2003
2005
2007
2008
Mean Input 1- Capital (billions in current US$)
27.8
35.7
51.8
72.8
89.7
Stdev Input 1- Capital (billions in current US$)
28.1
36.5
55.1
69.1
88.4
Mean Input 2- Energy consumption in (tons of oil equivalent)
38984.6
42166.6
46966.8
49625.1
52784.4
Stdev Input 2- Energy consumption in (tons of oil equivalent)
37385.6
41295.2
47192.8
46920.5
49539.7
Mean Input 3- Labor (in thousands)
3059.8
3251.0
3491.5
3823.9
4004.7
Stdev Input 3- Labor (in thousands)
3119.9
3312.0
3507.8
3687.4
3766.6
Mean Output-GDP (billions in current US$)
179.8
199.8
232.9
269.4
286.3
Stdev Output-GDP (billions in current US$)
169.5
187.6
212.8
237.8
248.4
B. The Methodology
The Augmented Dickey Fuller (ADF) (1979) and Phillips-Perron (1988) tests were implemented to assess the unit root. While performing the ADF test, the Schwarz (1987) Information Criterion was used for the lag length including the trend and intercept in the equation. The Phillips-Perron test was performed using automatic bandwidth selection of Newey-West (1987). This was implemented by including trend and intercept as well. For the cointegration test, the Johansen maximum likelihood cointegration test was employed (Masih and Masih, 1996).
After conducting the unit root and cointegration tests, we continued by testing the causality between energy consumption of the GDP by using Granger Causality Test (Granger, 1969). In order to test the causality from energy consumption to GDP, the following log-linear equation is estimated for GCC countries over the period from 1990-2009: VOL. 12 ALSAHLAWI: MEASURING ENERGY EFFICIENCY IN GCC COUNTRIES 23
USING DATA ENVELOPMENT ANALYSIS
Lnyt= λ0+ λ1 7Lnyt-1 +λ3 Lnxt-1+ μ, (5)
where
yt = real per capita GDP in period t
yt-1 = real per capita GDP in period t-1
yt-1 = layed value for per capita energy consumption in period 5-1
μ = error term
The presence of Granger-causality depends on the significance of Xx-I term in Equation (5); energy consumption causes GDP if the current value of GDP is predicted better by including lagged value of energy consumption.
In measuring the performance of energy efficiency among GCC countries, DEA methodology was used. Various indices have been used in comparing performances of countries across the world, e.g., the human development index and the global competitiveness index, by considering some relevant attributes in developing them (Savic and Martic, 2001 and Thore, 2008). We can classify them as fixed weight schemes since they combine performances in terms of various attributes using pre-fixed weights, which may be subjectively chosen. The advantage of DEA vs. fixed weight schemes is that the weights are not subjective but determined using linear programming (Ramanathan, 2006). The DEA approach computes these weights that maximize the efficiency score of a country subject to the efficiencies of other countries (calculated using the same set of weights) falling between 0 and 1.
Despotis (2005) considers DEA as a mathematical programming methodology based on the Frontier approach. It has been successfully employed to study the comparative performance of units that consume similar inputs and produce similar outputs. The units are generally referred to as Decision Making Units (DMUs). When we are assessing the performance of nations, DEA combines performances of countries in terms of several desirable and undesirable attributes into a single scalar measure, called the efficiency score. Countries that have unit efficiency scores of 1 are considered efficient, with the highest value of desirable attributes and the lowest values of undesirable attributes. Countries with efficiency scores of less than one are considered to operate sub-optimally for a given set of variables.
There are two possible assumptions that could be made while computing efficiency scores using DEA, namely constant returns to scale (CRS) and variable returns to scale (VRS). The assumption of CRS occurs when an increase in all inputs (i.e., increase in terms of undesirable attributes) by 1 percent leads to an increase in all outputs (i.e., increase in terms of desirable attributes) by 1 percent, while the assumption of VRS is for situations when the CRS assumption is not satisfied. The VRS efficiency of a DMU measures only technical efficiency, while CRS efficiency accounts for both technical efficiency and efficiency loss, when the DMU does not operate in its most productive scale size. The scale efficiency is the ratio of CRS to VRS scores. The scale efficiency of 1 would be for DMU which operates in its most productive size (Savic, 2001).
Taqi and Shah (2006) considered DEA as a non-parametric linear programming based methodology originally introduced by Charnes et al. (1978). Choosing the DMUs as the entities responsible for converting inputs into outputs is the key element in DEA. It compares each DMU based on its input and output factors with all other similar DMUs taken into consideration. Consequently, choosing outputs and inputs is a very important activity in the DEA process. 24 JOURNAL OF BUSINESS INQUIRY 2013
DEA can be used to obtain an overall measure of efficiency for each DMU, given the choices and the observed values of the inputs used and the outputs produced by each DMU. By using mathematical models, we obtained efficiency results depending on the performance attained by all of the DMUs. Thus, we can say that the efficiency measures through DEA are obtained by comparing them relatively with other DMUs. For example, one DMU, designated as DMU1, is being evaluated relative to some other DMUs which produced the same amount of output as DMU1 but used smaller levels of input. Then, DMU1 would be rated as inefficient relative to the other DMUs, where differences in their inputs would represent sources and amounts of inefficiency in DMU (Nordin, 2007).
VI. Results and Analysis
A. Causality Test
For policy purposes, the causality relationship between economic growth and energy consumption will be tested using a unit root test to find whether or not there is a unit root effect in the series. The result is presented in Table 6.
Table 6: Unit Root Test Results
Series
Augmented Dickey Fuller Test
Phillips-Perron Test
Level
1st difference
2nd difference
Level
1st difference
Saudi Arabia-Energy consumption
non stationary
stationary
non stationary
stationary
Saudi Arabia GDP
non stationary
stationary
non stationary
stationary
Bahrain-Energy consumption
non stationary
non stationary
stationary
non stationary
stationary
Bahrain-GDP
non stationary
stationary
stationary
non stationary
stationary
Kuwait- Energy consumption
non stationary
stationary
non stationary
stationary
Kuwait-GDP
non stationary
stationary
non stationary
stationary
Qatar- Energy consumption
non stationary
non stationary
stationary
non stationary
stationary
Qatar-GDP
non stationary
non stationary
stationary
non stationary
stationary
(at 10%)
Oman- Energy consumption
non stationary
stationary
non stationary
stationary
(at 10%)
Oman-GDP
non stationary
stationary
non stationary
stationary
UAE- Energy consumption
non stationary
stationary
non stationary
stationary
UAE-GDP
non stationary
stationary
non stationary
stationary
Panel- Energy consumption
non stationary
stationary
non stationary
stationary
Panel-GDP
non stationary
stationary
non stationary
stationary
VOL. 12 ALSAHLAWI: MEASURING ENERGY EFFICIENCY IN GCC COUNTRIES 25
USING DATA ENVELOPMENT ANALYSIS
Based on the unit root test, all the series are non-stationary at the level, since they contain trends. Taking the first difference for all of them, they become stationary at first difference. Since the stationarity of the series at the same level is a necessary condition in order to get valid result of causality test, then the causality test can be continued.
However, the co-integration test between the GDP and energy consumption for each country is shown in Table 7.
Table 7: Co-integration Test Results
Countries
Cointegration between energy and GDP
UAE
Not cointegrated
SAUDI ARABIA
Cointegrated
QATAR
Cointegrated
OMAN
Cointegrated
KUWAIT
Not cointegrated
BAHRAIN
Cointegrated
PANEL
Cointegrated
Although the results for the UAE and Kuwait showed that energy consumption and GDP are not cointegrated at a 5 percent confidence level, further testing of the causality between the two series (energy consumption and GDP) is not a necessary condition for the causality test. However, correlation does not necessarily imply causation.
Based on Granger (1988), whether x causes y in Equation (5) depends on how much of the current y can be explained by past values of y and then to see whether adding lagged values of x can improve the explanation. It is said that y is Granger-caused by x if x helps in the prediction of y, or equivalently if the coefficients on the lagged x’s are statistically significant. The result of applying the Granger Causality test on the relationship between GDP and energy consumption is presented in Table 8. The results are confirmed by doing F-tests for the coefficients of the independent results including error terms at a 5 percent significance level, while R-squares range from 80 percent to 95 percent.
Table 8: Granger Causality Test Results
Granger Causality
Saudi Arabia
Bahrain
Kuwait
Qatar
Oman
UAE
Panel
GDP cause energy consumption
no
yes
yes
no
no
no
yes
Energy consumption
cause GDP
yes
no
no
yes
yes
yes
no
It is noted that there is a strong causality from energy consumption to GDP for Saudi Arabia. Results for Oman are similar to Saudi Arabia, while Qatar and the UAE show a moderate causality from energy consumption to GDP. The result for Bahrain shows the opposite direction that GDP 26 JOURNAL OF BUSINESS INQUIRY 2013
causes energy consumption, while Kuwait somehow tends to have stronger causality running from GDP to energy consumption. For the panel of GCC, causality runs from GDP to energy consumption, which is consistent with previous research by Al-Iriani (2006) and Mehrara (2007).
Bahrain and Kuwait tend to be in line with developing countries when causality runs from GDP to energy consumption, while the other countries showed that they are close to what was found in the developed countries by Chontanawat et al. (2006).
B. Energy Efficiency
After performing the causality tests, the measurement of energy efficiency is performed using the input oriented DEA with CRS envelopment analysis.
In order to have benchmark for comparison, three other OECD countries were added in the measurement. We selected efficient countries based on the results of measurements conducted by Zhou and Ang (2008) for 21 OECD countries. The result of energy efficiency measurement using Model I is presented in Table 9.
Table 9: Energy Efficiency Measure Using Model I
No.
DMU Name
2001
2003
2005
2007
2008
1
UAE
0.41
0.45
0.48
0.40
0.35
2
SAUDI ARABIA
0.49
0.46
0.43
0.42
0.32
3
QATAR
0.55
0.53
0.64
0.65
0.41
4
OMAN
0.54
0.48
0.46
0.45
0.46
5
KUWAIT
0.45
0.55
0.51
0.45
0.52
6
BAHRAIN
0.47
0.46
0.46
0.44
0.45
7
AUSTRALIA
0.60
0.59
0.57
0.54
0.54
8
NORWAY
0.79
0.78
0.80
0.76
0.72
9
SWITZERLAND
1.00
1.00
1.00
1.00
1.00
Saudi Arabia and Bahrain did not show significant improvement and were still in the inefficient rankings compared to others countries as measured in Table 9. Qatar showed significant improvement, while the UAE’s improvement was slightly slow. Oman’s performance has declined since 2007, and Kuwait exhibited significant improvement to become the most efficient country in the GCC surpassing Oman.
By comparing the energy efficiency of GCC countries with energy efficiency in developed countries, it seems that the GCC is still far behind but has high potential for energy preservation.
Comparing the results of the energy efficiency measurement in Model I with the efficiency findings using the energy intensity in Figure 2, the results are consistent.
The advantage of using the energy intensity of Figure 2, is that improvement in country efficiency can be shown for recorded years while Model I, gives efficiency improvement of GCC countries relative to other depicted countries at one time period.
In Model II, instead of using only energy consumption as the only input, labor and capital as other economic features were added in order to measure energy efficiency within an economic perspective. VOL. 12 ALSAHLAWI: MEASURING ENERGY EFFICIENCY IN GCC COUNTRIES 27
USING DATA ENVELOPMENT ANALYSIS
Measurement results using Model II are shown in Table 10. Even classified as inefficient in terms of energy consumption, Saudi Arabia appears in the efficient frontier in economy-wide performance since 2005. Qatar and Kuwait have maintained their efficient economy-wide relative to others as in Model I, while Oman has been inefficient since 2007. The UAE and Bahrain have shown more room for improvement to attain economic efficiency.
Table 10: Energy Efficiency Measure Using Model II
No.
DMU Name
2001
2003
2005
2007
2008
1
UAE
0.45
0.85
0.86
0.77
0.75
2
SAUDI ARABIA
0.42
0.84
1.00
1.00
1.00
3
QATAR
0.51
1.00
1.00
1.00
1.00
4
OMAN
0.56
0.99
1.00
0.86
0.82
5
KUWAIT
0.42
1.00
1.00
1.00
1.00
6
BAHRAIN
0.45
0.71
0.74
0.81
0.70
7
AUSTRALIA
0.54
0.90
0.87
0.86
0.84
8
NORWAY
0.72
1.00
1.00
1.00
1.00
9
SWITZERLAND
1.00
1.00
1.00
1.00
1.00
In comparison with efficient developed countries, Model II also shows that GCC countries have to improve their economic energy efficiency. It is noted that results from Model I and Model II differ substantially. This is due to the structural difference between Model I and Model II, where the latter explains economic efficiency more than energy efficiency. Including labor, and capital as a factor of production in Model II, has also changed the model specifications and, ultimately, the results. Furthermore, the structural and policy differences among GCC countries are clearer in Model II which affected its results and caused some discrepancies between the results from Model I and Model II.
VII. Conclusion and Policy Implications
With increased pressure on scarce water resources and high energy consumption, the GCC will be adversely affected by climate change and rising air pollution vulnerabilities. Therefore, GCC countries have directed their energy policy to energy efficiency policies. Recently, research on alternative energy has been encouraged with more strategic movement toward energy conservations. Table 11, shows the policy implications of the results.
28 JOURNAL OF BUSINESS INQUIRY 2013
Table 11: Energy Efficiency—Results and Policy Implications
Countries
Model II
Model I
Causality
Policy Implication
Saudi Arabia
Efficient
Inefficient
Energy consumption to GDP
It seems that Saudi Arabia depends significantly on oil consumption in its economy. Their economy is efficient, but their energy consumption is not efficient compared to other GCC countries. Energy consumption in Saudi Arabia is very high since they have very cheap energy. Increasing oil prices (reduce subsidy) to reduce the oil consumption will obviously reduce CO2 emissions. Since the causality strongly runs from EC to GDP for Saudi Arabia, then it needs to complement the strategy by providing alternative power environment-friendly sources of energy like solar to substitute oil, so that the effect on the GDP is controlled.
Bahrain
Inefficient
Inefficient
GDP to energy consumption
Bahrain’s economy is inefficient due to its inefficient use of energy. Thus, energy conservation policy would be appropriate for Bahrain, since the causality runs from GDP to energy consumption. In fact, the more efficient their energy use, the more efficient the economy will be.
Kuwait
Efficient
Efficient
GDP to energy consumption
For Kuwait, it is efficient for both energy and economic efficiency compared to other GCC countries. Since the strong causality runs from GDP to energy consumption in this country, it is suggested for the government to implement energy conservation policy to become more energy efficient which eventually will become more efficient in the economy as well as toward that of more developed countries that were used as benchmark.
Qatar
Efficient
Inefficient
Energy consumption to GDP
The economy of Qatar is efficient compared to other GCC countries, and the trend for its energy efficiency is improving. Based on their causality test, they have weak causality which runs from EC to GDP that will not have a great negative effect on their economy if they implement energy conservation policy.
Oman
Inefficient
Inefficient
Energy consumption to GDP
Unlike in other GCC countries, the trend of energy efficiency is decreasing for Oman. It started to become inefficient in using energy and the economy has tended to become inefficient, as well since 2007. By implementing energy conservation policy, Oman needs to also adopt an alternative energy policy to replace the oil dependency in its economy, since we found strong causality running from energy consumption to GDP.
UAE
Inefficient
Inefficient
Energy consumption to GDP
Both the economy and energy use of the UAE is classified as inefficient compared to those of other GCC countries. They can improve their efficiency by implementing energy conservation policy since they have only weak causality running from energy consumption to GDP. VOL. 12 ALSAHLAWI: MEASURING ENERGY EFFICIENCY IN GCC COUNTRIES 29
USING DATA ENVELOPMENT ANALYSIS
Increasing international concerns over global warming and climate change have brought a significant dilemma for some countries, especially developing countries. It becomes important for GCC policy makers to realize the importance of energy efficiency. Some indicators were developed to measure energy efficiency, such as energy intensity and economic-thermodynamics to assist in quantifying such measures. In this paper, we applied linear programming techniques to measure and compare energy efficiency of the GCC countries.
In order to complement the results, the causality test between energy consumption and GDP for each GCC country and for all GCC countries as a panel has been conducted. This type of test will be useful for determining the appropriate policy to adopt to achieve energy efficiency without crippling economic growth.
For further research, a sectoral energy analysis related to the GDP of GCC countries should be performed to develop sound policies for each sector without harming the economic growth of the country.
References
Al-Iriani, Mahmoud A. 2006. “Energy-GDP Relationship Revisited: An Example from GCC Countries Using Panel Causality.” Energy Policy, 34(17): 3342-50.
Al-Mansour, Fouad 2011. “Energy Efficiency Trends and Policy in Slovenia.” Energy, 36(4): 1868-77.
British Petroleum (BP). 2011 Statistical Review of World Energy Report. British Petroleum. http://www.bp.com/sectiongenericarticle800.do?categoryId=9037161&contentId=7068605 (accessed May 14, 2012).
Charnes, A., W. W. Cooper, and E. Rhodes. 1978. “Measuring the Efficiency of Decision Making Units.” European Journal of Operational Research, 2(6): 429-44.
Chontanawat, Jaruwan, Lester Charles Hunt, and Richard G. Pierse. 2006. “Evidence from 30 OECD and 78 non-OECD Countries.” Surrey Energy Economics Discussion Paper, (SEEDS) 113. Surrey Energy Economics Centre (SEEC), University of Surrey,
www.seec.surrey.ac.uklResearchlSEEDS.htm.
Chontanawat, Jaruwan, Lester Charles Hunt, and Richard G. Pierse. 2008. “Does Energy Consumption Cause Economic Growth? Evidence from Systematic Study of over 100 Countries.” Journal of Policy Modeling, 30(2): 209-20.
Despotis, D. K. 2005. “Measuring Human Development via Data Envelopment Analysis: The Case of Asia and the Pacific.” The International Journal of Management Science, 33(5): 385-90.
Dickey, David A., and Wayne A. Fuller. 1979. “Distribution of the Estimators for Autoregressive Time Series with a Unit Root.” Journal of the American Statistical Association, 74(366): 427-31.
Granger, C. W. J. 1969. “Investigating Causal Relations by Econometric Models and Cross-Spectral Methods.” Journal of Econometrics, 36: 424-38.
Granger, C. W. J. 1988. “Some recent developments in a Concept of Causality.” Journal of Econometrics, 39(1-2): 199-211.
Haji, Mohamad Nordin, and Fatimah Binti Said. 2011. “Comparing Macroeconomics Performance of OIC Member Countries.” International Journal of Economics and Management Sciences, 1(3): 90-104.
Hannesson, Rögnvaldur. 2009. “Energy and GDP Growth.” International Journal of Energy Sector, 3(2): 157-70. 30 JOURNAL OF BUSINESS INQUIRY 2013
Hertog, Steffen, and Giacomo Luciani. 2009. Energy and Sustainability Policies in the GCC. Kuwait: The Center for the Study of Global Governance.
International Monetary Fund. 2011. “World Economic Outlook Database.” September.
Masih, Abdul M. M., and Rumi Masih. 1996. “Energy Consumption, Real Income and Temporal Causality: Results from a Multi-Country Study Based on Cointegration and Error-Correction Modelling Techniques.” Energy Economics, 18(3): 165-83.
Mehrara, Mohsen. 2007. “Energy-GDP Relationship for Oil-Exporting Countries: Iran, Kuwait and Saudi Arabia.” OPEC Review, 31(1): 1-16.
Newey, Whitney K., and Kenneth D. West. 1987. “A Simple, Positive Semi-Definite Hetroskedasticity and Autocorrelation Consistent Covariance Matrix.” Econometrica, 55(3): 703-08.
Nordin, Mohamad. 2007. “A Linear Programming Formulation of Macroeconomic Performance: the Case of Asia Pacific.” Matematika, 23(1): 29-40.
Patterson, Murray G. 1996. “What is Energy Efficiency? Concepts, Indicators and Methodological Issues.” Energy Policy, 25(1): 377-90.
Phillips, Peter C. B., and Pierre Perron. 1988. “Testing for a Unit Root in Time Series Regression.” Biometrica, 75(2): 335-46.
Ramanathan, Ramakrishnan. 2006. “Evaluating the Comparative Performance of Countries of the Middle East and North America: A DEA application.” Socio-Economic Planning Sciences, 40(2): 156-67.
Sa'ad, Suleiman. 2010. “Energy Consumption and Economic Growth: Causality Relationship for Nigeria.” OPEC Energy Review, 34(1): 15-24.
Saudi Arabian Monetary Agency. 2010. Forty-Sixth Annual Report. Riyadh: SAMA, 18.
Saudi Arabian Monetary Agency. 2012. Forty-Eighth Annual Report. Riyadh: SAMA.
Savić, Gordana, and Milan Martić. 2001. “An Application of DEA for Comparative Analysis and Ranking of Regions in Serbia with Regards to Social-Economic Development.” European Journal of Operational Research, 132(2): 343-56.
Schwarz, Gideon. 1987. “Estimating the Dimension of a Model.” The Annual of Statistics, 6(2): 461-64.
Taqi, N. A., and M. S. Shah. 2006. “Productivity Assessment of The GCC Countries By Means of Data Envelopment Analysis.” International Journal of Business Research, 122-26.
Thore, Sten, and Ruzanna Tarverdyan. 2008. “Using Data Envelopment Analysis to Screen the Possibility of a Fair Globalization.” Portuguese Journal of Management Studies, 13(1): 129-51.
World Bank. 2012. “World Development Indicators.”
Zhou Peng, and Beng Wah Ang. 2008. “Linear Programming Models for Measuring Economy-Wide Energy Efficiency Performance.” Energy Policy, 36(8): 2911-16The Journal of Business Inquiry 2013, 12, 31−64
http:www.uvu.edu/woodbury/jbi/volume12
ISSN 2155-4072
An Examination of Great by Choice: Redux
By ERIC OLSEN AND JAMES SENA
Given the works by Collins, and later Collins with Hansen, we examined the performance and practices of the fourteen companies in Great by Choice [GBC]. We looked at the financial performance for the 11 years ending the GBC comparison period (1991-2001) and the 11 years of our research period (2002-2012). We used financial analysis to develop and examine research propositions as to whether the companies continued, discontinued, or started employing the GBC practices. Overall, we concluded that GBC has good advice for companies.
Keywords: Leadership, Management Best Practices, Practice Versus Performance, Comparison Case Studies, Good to Great, Great by Choice
JEL Classification: Y10, O34, O31, M11, K42, L25, M21
I. Introduction
In a series of books, starting with Built to Last (Collins, 2001), Jim Collins and a team of researchers sought to discover the underlying principles of greatness. Their approach was to identify companies that differentiated themselves through outstanding marketplace performance, comparing them to other, equally promising companies that were less successful. These comparisons covered a specific period where the firms had equal potential for success. It is to a recent edition of the Collins series that we address this paper – Great by Choice (Collins and Hansen, 2011).
Collins and Hansen sought to identify principles and practices that were unique to successful companies across sample sets. This approach appears to be scientifically rigorous. We decided to test their conclusions beyond their sample sets’ dynastic period. The principles and practices apply to companies within the period of analysis, but what about beyond? There are certainly examples of once great companies that have fallen off the pedestal in Collins’ own body of work (Collins, 2009). Collins made the argument that those that fall from greatness did not invalidate his conclusions because during the dynastic period the companies were engaging in those practices identified as being great. His assumption was that the companies are no longer great because they are no longer modeling the practices. In this paper, we test to see if continued great performance is explained by the application of GBC practices or if a reduction in performance is explained by companies discontinuing the use of the practices that ostensibly made them great. Similarly, we look to see if the comparison companies identified in GBC improve their financial performance if they adopt GBC practices. We begin with a review of Collins’ previous works to determine how their conclusions and practices were derived (Table 1).
James Sena: Orfalea College of Business, California Polytechnic State University. E-mail: jsena@calpoly.edu. Eric Olsen: Orfalea College of Business, California Polytechnic State University. E-mail: eolsen@calpoly.edu. 32 JOURNAL OF BUSINESS INQUIRY 2013
Table 1: Quick Reference to Collins and Team Series of Books
Title
Reference
Objective
Built to Last
Collins and Porras (1994 and 2001)
Identify practices that enable the transformation from a mediocre (good) company to a great company.
Good to Great
Collins (2001)
Identify practices of great companies.
Good to Great and the Social Sectors
Collins (2011)
Identify practices of great companies in the social sector.
How the Mighty Fall
Collins (2009)
Identify mechanisms that cause once great companies to fail.
Great by Choice
Collins and Hansen (2011)
Uncertainty, chaos luck—why some thrive despite them all.
In Good to Great Collins showed how great companies triumph over time and how long-term sustained performance could be engineered into the enterprise. He identified a set of elite companies that made the transition from mediocre to extraordinary results and sustained those results for at least fifteen years. After the transition, the good to great companies generated cumulative stock returns that beat the general stock market by an average of seven times in fifteen years, better than twice the results delivered by a composite index of the world's greatest companies. They noted that the transformation was a process of build-up, followed by breakthroughs, broken into three broad stages, with a few key differentiators in each as described in Figure 1.
Figure 1: Good to Great Stages and Key Differentiators
Stage I
Disciplined People
Level 5 Leadership
A surprising style, required for greatness. These executives build "enduring greatness through a paradoxical blend of personal humility and professional will." They are modest and freely give others credit, act with calm determination, and channel ambition into the company (not to themselves).
First Who, Then What
The key distinguishing point here resulting from Collins team's research was not just about the importance of assembling the right team, but about, first, getting the right people on the bus, the wrong people off the bus, and the right people in the right seat. And then, determining the proper strategic direction that should be followed. Often we think of accomplishing these two steps in the opposite order.
Stage II
Disciplined Thought
The Stockdale Paradox
The need to confront the brutal facts and continually refine the path to greatness, yet never losing faith in the ability to prevail. Understanding your current reality, good and bad, can only fully occur when an open culture is fostered where the truth can be discussed and a climate exists where intense dialogue is welcome and handled professionally.
The Hedgehog Concept
To transition from good to great, it is necessary to gain a deep understanding of three intersecting circles: 1) what you can be the best in the world at, 2) what economic denominator best drives your economic engine, and 3) what you are deeply passionate about. Next, this understanding must be translated into a concept that can be executed upon.
VOL. 12 SENA and OLSEN: AN EXAMINATION OF GREAT BY CHOICE: REDUX 33
Figure 1: Good to Great Stages and Key Differentiators: Continues
Stage III
Disciplined Action
The Culture of Disciipline
The good-to-great companies built a consistent system with clear constraints, but they also gave people freedom and responsibility within the framework of that system.
Technology Accelerators
Good-to-great companies think differently about technology. They do not use technology to transform from good to great, but to accelerate their progress, staying focused on their Hedgehog. Collins asserts that technology by itself is never a primary cause of either greatness or decline.
The Flywheel and the Doom Loop
Why those who do radical restructuring fail to make the leap to greatness, while those that consistently accumulate momentum (turn upon turn, step by step) can transform themselves. Patience, persistence and discipline are critical to achieving desired results.
Collins and Hansen (2011), subsequent to Good to Great, extended this research work in GBC to companies that they refer to as “10x” cases. During the study period, these companies outperformed other companies in their industry by 10 times or more. The final organizations that met their criteria, after considering over 20,000 companies, were Amgen, Biomet, Intel, Microsoft, Progressive Insurance, Southwest Airlines, and Stryker. These companies started from a position of vulnerability, rose to become great by choice with spectacular performances, and did so in unstable environments characterized by big forces that were out of their control, fast moving, uncertain, and potentially harmful. Simultaneously, they compared these companies to matched pairs that failed to become great in the same extreme environments. They used the contrast between winners and the “also-ran” comparison companies to uncover the distinguishing practices that allow some to thrive in uncertainty. Table 2 presents the GBC practice concepts.
Table 2: Great by Choice Practice Concepts
Practice
Analogy
Description
Fanatic Discipline
The 20 Mile March
Consistent execution without overreaching in good times or underachieving in bad times.
Productive Paranoia
Leading above the Death Line
Learning how to effectively manage risk so that the risks your organization takes never put it in mortal danger.
Return on Luck
“The critical question is not whether you’ll have luck, but what you do with the luck that you get.”
Empirical Creativity
Firing Bullets, Then Cannonballs
Testing concepts in small ways and then making adjustments rather than placing big, unproven bets. But then placing big bets when you have figured out exactly where to aim.
Level 5 Ambition
Ambition for the success of the organization rather than self.
34 JOURNAL OF BUSINESS INQUIRY 2013
Given these works by Collins, and later Collins with Hansen, we set out to examine the performance and practices of the fourteen companies in GBC since the study was completed. We used a convention of looking at the financial performance for a specific time periods, namely the 11 years ending the GBC comparison period (1991-2001) and the 11 years of our research period (2002-2012). This approach is depicted in Diagram 1 and Diagram 2.
Diagram 1: Collins Approach During Dynastic Period
We think this is a valid and comparable financial performance indicator that meets the intent of Collins and Hansen to compare matched pair companies under identical economic conditions. Basically in this paper we replicate the methodology presented in Collins and Hansen’s GBC over the end of their period of examination (1991–2001) and extend it into a second period (2002–2012). Our objective is to determine if the practices identified in GBC, and related performance, continued (or increased) or discontinued (or decreased) based on financial and practice analysis comparable to that used in GBC. We used our financial analysis to develop research questions and propositions as to whether the eleven remaining companies in the GBC sample were or were not employing the GBC practices. We used propositions rather than hypotheses because we viewed the practice analysis to be of a qualitative nature. We then tested these propositions by seeing if the companies indeed were or were not using the practices based on a review of the relevant literature. The sections that follow describe our financial analysis, the development of our propositions, our practice analysis, and conclusions as to the validity of Collins and Hansen’s assertion that GBC practices lead to great performance.
Practice Analysis
1
Fanatic DISCIPLINE
2
Productive PARANOIA
3
Empirical CREATIVITY
4
Level 5 AMBITION VOL. 12 SENA and OLSEN: AN EXAMINATION OF GREAT BY CHOICE: REDUX 35
Diagram 2: Our Research Approach 2002-2012
II. Financial Performance Analysis
A. GBC Procedure
Collins and Hansen selected and compared companies based on total price return performance from 1972 to 1992. If the company was not publicly traded in 1972 its performance was held at the “general stock market rate of return.” This convention actually created some interesting anomalies in the “10x’ers” claim. Some companies benefited considerably from riding the general market trend until its initial public offering. For example, the GBC return for $10,000 purchased December 31, 1972, and assumed to be held at the general stock market rate until Amgen went public in 1980, and then held until 2002, would yield $4.5 million. This equates to 24x the market. However, $10,000 invested in 1980 and sold at the end of 2002 would yield only 4420 percent or $442,000. 36 JOURNAL OF BUSINESS INQUIRY 2013
The S&P 500 Total Price Return was 226.6 percent for the same period. Therefore, Amgen outperformed the market by 19.5x since 1980. During the last 10 years of the GBC period from 1992 to 2002, Amgen only outperformed the S&P 500 by 2.1x. While this performance may still be considered “great,” it points to an inherent problem in allowing the results to be influenced by variable public offering dates.
B. Our Process
Table 3 shows the Total Price Return percentage and “times better” of the eleven remaining GBC and comparison companies for the two periods: 1991–2001 and 2002–2012. Table 3a provides data when full 11-year periods were available. Table 3b provides data when less than full periods of date were available.
The first check we performed was to see if the relative financial performance identified by Collins and Hansen was maintained in the last eleven years of their study. To do this, we examined how the GBC companies performed in comparison to the Standard and Poor’s 500 and to their comparison companies. All the GBC companies we examined performed 2.1 to 7.1 times better than the S&P 500. Two of the three comparison companies examined did worse than the S&P 500: Safeco and Apple. We looked at the five remaining comparison pairs in the sample: Amgen-Genentech, Progressive-Safeco, Intel-AMD, Microsoft-Apple, and Stryker-USSC. The Stryker and USSC comparison was only done for 1991-1998 because USSC was purchased by Tyco in 1998. All these pairs showed that the GBC Company outperformed their comparison company by 3.3 to 29.4 times. Although not consistently “10x,” these observations generally support the selection of GBC and comparison companies by Collins and Hansen and for continued use in our research.
Table 3a: Total Price Return Percentage Comparison for Full Periods
Notes: continued on next page
Notes: from Table 3a.
Percentage Change Total ReturnTimes better (times worse) than S&P 500Times better (times worse) than Comparison CompanyPercentage Change Total ReturnTimes better (times worse) than S&P 500Times better (times worse) than Comparison CompanyS&P 500 Index34654Amgen20764.93.9571.0Genentech (1)4601.3Biomet (1)9382.3Kirschner (2)Intel25746.04.1-17(1.9)5.5AMD5511.5-85(10.2)Progressive8212.13.31181.4Safeco (1)180(1.6)Microsoft30697.129.47(1.4)(46.0)Apple8(4.1)480131.9Southwest Airlines15343.7-44(2.7)PSA (2)Stryker14283.41021.3USSC (3)See Table 3bSee Table 3bSee Table 3bGBC (1991-2001)Update (2002-2012)VOL. 12 SENA and OLSEN: AN EXAMINATION OF GREAT BY CHOICE: REDUX 37
1. Updates for Genentech and Safeco are included in Table 3b due to change or discontinuation of business during 2002-2012.
2. PSA dropped from analysis due to acquisition early in the GBC period.
3. USSC GBC partial period included in Table 3b due to acquisition after 1998.
Table 3b: Total Price Return Percentage Comparison for Partial Periods
Company
Comparison
Period
S&P 500 Basis
Percentage
Change
Percentage Change
Times better (times worse) than S&P 500
Times better (times worse) than Comparison Company
Great by Choice Partial Period Stryker
1991 – Nov. 1998
335
449
1.3
4.9
USSC
13
(3.9)
Update Partial Period Amgen
2002 – Apr. 2009
-13
-14
(1.2)
(4.1)
Genentech
250
3.5
Biomet
2002 – Oct. 2007
49
53
1.5
Progressive
2002 – Oct. 2008
-5
28
1.3
(2.0)
Safeco
152
2.5
C. GBC-Redux
We looked at the 11-year update period 2002-2012. Most of the companies did worse due to the 2008 recession, except for Apple, which went from being 4.4 times worse than the general market to being 31.9 times better. To test Collins and Hansen’s implied hypothesis that GBC practices lead to great performance and the lack of these same practices leads to worse performance, we applied the same performance comparisons to the S&P 500 and the remaining comparison companies to develop indicators of great and mediocre financial performance.
Overall, the GBC companies did not fair very well in the update period. In comparison to the S&P 500, only Progressive (1.4x) and Stryker (1.3x) were doing slightly better than the market. Amgen and Biomet performed only as well as the general market (1.0x). The other three GBC companies were doing worse than the general market: Intel (1.9x worse), Microsoft (1.4x worse), and Southwest Airlines (2.7x worse). With respect to the comparison companies, Genentech, Safeco and Apple now outperform the S&P 500 by 4.x, 2.7x, and 31.9x respectively. AMD and USSC are the only remaining comparison companies that continue to do worse than the general market.
Collins and Hansen had seven matched pairs of companies for comparison. In the update period, we had enough data to make comparisons for four remaining matched pairs. For three of the matched pairs, the GBC Company was now performing worse than the comparison company: Amgen-Genentech, Progressive-Safeco, and Microsoft-Apple. Collins and Hansen observe that the true test of a company’s ability to handle a turbulent business environment is best accomplished by comparing like companies operating in the same environment. For the four remaining comparisons that we looked at, only Intel is still doing better than its comparison company, AMD. The relationship for the other three comparisons has reversed, with the comparison companies Genentech, Safeco, and Apple doing better than the previous “great” companies in the update period 2002-2012. From a survival analysis perspective, the fact that more agile firms absorbed Kirschner, PSA, and USSC provides support for the GBC hypothesis. In addition, the two other companies absorbed—Genentech and Safeco—improved their performance during our research period and became targets for acquisition. Biomet actually acquired Kirschner and went private during our period of study. 38 JOURNAL OF BUSINESS INQUIRY 2013
Based on our financial analyses we proposed a set of eleven research questions and propositions related to the GBC practices. These research questions, shown in Table 4, depict our expectations for GBC practice or lack of practice given on our financial analysis of the update period. Next, we looked for evidence of practice use during the update period 2002-2012 and summarized our observations in the table. Our supporting analyses and literature examination is discussed in the sections that follow.
Table 4: GBC Practice Usage During Update Period 2002-2012 Propositions
III. Practice Comments
In GBC 10x leaders were both "disciplined" and "creative," "prudent" and "bold”—they went fast when they must, but slow when they could—they were consistent, but open to change. According to Collins and Hansen, successful companies were not as innovative as the control companies; they were in some cases less innovative. Rather, they managed to “scale innovation,” introducing changes gradually, then moving quickly to capitalize on those that showed promise. The successful companies were not necessarily the most likely to adopt internal changes as a response to a changing environment. “The 10x companies changed less in reaction to their changing world than the comparison cases” (Murray, 2011).
Collins and Hansen began the process of identifying and further explicating the unique factors and variables that differentiate GBC companies. One of the most significant differences is the quality and nature of leadership. They found that many of those classified in this group displayed an unusual mix of intense determination and profound humility, often having a long-term, personal sense of investment in the company and its success, cultivated through a career-spanning climb through the company’s ranks. Personal ego and individual financial gain are not as important as the long-term benefits to the team and the company. Designated as “Level 5 Ambition,” this is the centerpiece and grounding practice of leaders that support the other three GBC practices—“Fanatic Discipline,” “Productive Paranoia,” and “Empirical Creativity.”
“Fanatic Discipline:” In GBC the authors introduced the “20-Mile March” analogy to illustrate the practice of fanatic discipline. It is about having concrete, clear, intelligent, and rigorous performance mechanisms with two types of self-imposed control: (1) the discomfort of unwavering commitment to high performance in difficult conditions, and (2) the discomfort of holding back in good conditions. GBC leaders and companies demonstrate the discipline to make well-reasoned, measured commitments and stick to them.
“Productive Paranoia:” In practicing productive paranoia, GBC leaders and companies are positioned better than their comparison companies to take advantage should the worst (or best)
Research QuestionCompany GBC Practice Proposition (per financial analysis)Fanatic DISCIPLINEProductive PARANOIAEmpirical CREATIVITYLevel 5 AMBITIONSummary of Practice ObservationsGBC Practice Use Supported by Literature?1AmgenStopped using4.06.07.04.05.3Somewhat Agree2GenentechStarted using5.06.05.57.05.9Agree3BiometMaintained6.06.06.05.55.9Agree4Intel Maintained3.34.95.32.03.9Neutral5AMDNever used0.13.03.02.02.0Disagree6Progressive Maintained6.06.06.06.06.0Agree7SafecoStarted Using6.25.95.74.95.7Agree8MicrosoftStopped using4.43.35.33.64.2Neutral9AppleStarted using5.05.45.44.95.2Somewhat Agree10Southwest Airlines Stopped using4.54.15.81.03.9Neutral11Stryker Maintained6.03.07.06.05.5AgreeFinancial ObservationsPractices Observations (according to Literature)VOL. 12 SENA and OLSEN: AN EXAMINATION OF GREAT BY CHOICE: REDUX 39
happen. GBC leaders continuously scan the environment in what Collins and Hansen call “zoom out” mode and then “zoom in” to put specific plans and resources in place to cover even lower probability eventualities if the effect is potentially devastating enough. Another analogy that Collins and Hansen make in support of this practice is “return on luck.” Both GBC and comparison companies have the same good and bad luck opportunities. However, GBC companies are better positioned to take advantage of extraordinary events or situations.
“Empirical Creativity:” A key practice of GBC companies is their unique ability to collect and analyze their own data. GBC companies are data driven. They are not “me, too, companies” that mindlessly follow the crowd. The analogy that Collins and Hansen use to describe a typical application of this practice is “first bullets, and then cannon balls.” GBC companies do many small experiments and tests of the market place, before committing to a huge investment of time and resources.
We used these practice descriptions, and those in GBC, to identify practice usage by the sample companies. To better understand the context and business environment of the update period we also considered a number of additional factors that complemented and might be considered correlated to the GBC practices. These included counts by year of acquisitions and divestitures, joint ventures, infrastructure incidents, significant personnel actions, philanthropic activity, litigation, financial announcements, and recognitions/presentations. These factors were particularly helpful in analyzing and assigning ratings in situations where there was extensive activity. Examples are litigation dealing with the drug companies; the joint ventures of AMD in their attempt to compete with Intel; the acquisition activity of Microsoft; and the personnel changes and leadership ratings of Southwest and Amgen and Stryker.
We performed a detailed practice analysis of the eleven companies depicted in Table 4. To verify our propositions we examined an extensive set of sources and references. Of note, there was neither uniform nor consistent availability of the company data. For example, Wikipedia was somewhat useful for providing a ready supply of current links and sources. For some organizations, such as Intel and Microsoft, company websites overwhelmed us with data while others (e.g. Stryker) had minimal content. We visited all of the company websites and examined their financial declarations for the eleven years of our study. There was much variability among the companies in the form and content of reporting. For a number of the companies, media and press releases were quite useful—in most cases, this involved sifting through two to three hundred references for each of the eleven years. Two other sources we used were Brint.com, a specialized business search engine and Motley Fool, a specialized stock portfolio analysis tool. The Brint.com source allowed us to view academic journals, business magazines and newspapers, as well as industry specific publications, while deploying various search filters.
Overall, we rated each company as shown in Table 5 on the four practices: “Fanatic Discipline,” “Productive Paranoia,” “Empirical Creativity” and “Level 5 Ambition” and noted whether the data supports or does not support our proposition. We scored articles and incidents using GBC discussions and descriptions. The scores were converted into a 7-point scale from “strongly disagree that the practice is being used” (1) to “strongly agree the practice is being used” (7). If the practice rating supports our propositions on practice usage based on financial performance (Table 5), then our analysis supports Collins and Hansen’s work in GBC. For example, in the case of Genentech we proposed that they had started using GBC practices based on their better-than-comparison and market financial performance. An overall practices rating of 5.9 (agree) indicates that this was indeed the case. Collins and Hansen’s work is supported. In another instance, Amgen had worse-than-comparison and industry financial performance. We 40 JOURNAL OF BUSINESS INQUIRY 2013
proposed that they had stopped using GBC practices. However, an overall practices score of 5.3 (somewhat agree) indicates that the GBC practices were still in place. In this case, Collins and Hansen’s work is not supported.
Of the eleven companies we examined from 2002 to 2012, only one case clearly did not support the proposition that GBC practices lead to better financial performance. Amgen seems to have continued to use the GBC practices, but had worse financial performance. Seven of the cases clearly supported the use of GBC practices either by confirming great financial performance or by confirming poor performance through the lack of practice usage. Three of the cases lacked decisive evidence for practice usage either way (neutral).
In the remainder of this paper, we provide a case-by-case description of our analysis and conclusions with respect to GBC practice usage by the study companies in the period from 2002 to 2012. At the conclusion of the paper, we summarize our findings and conclude with some observations about the nature, replication and extension of Collins and Hansen’s work.
A. Proposition 1: Amgen Stopped Using GBC Practices
During our study period, 2002 to 2012, Amgen started applying the scientific method and experimentation to the sales process. Typically, pharmaceutical salespersons represent several products. However, selling new medications entails educating physicians and medical staff making the sales effort very labor intensive. Based on experiments suggested by an Amgen research scientist, salespersons were made specialists in individual products. This has proven to be beneficial in some industries; however, for Amgen it became a distraction and a deviation from the GBC practices of “Fanatic Discipline” and “Productive Paranoia.”
As part of our research on Amgen, we examined and categorized 138 news releases over our study period (Amgen News Releases). Table 5 presents a summary of these categorizations.
Table 5: Amgen News Releases 2002-2012
A. 1. Fanatic Discipline [Neutral]
During the period of our study, Amgen had a stable of drugs under patent protection generating cash flows. It also had several high potential drugs in its R&D pipeline. Amgen is the largest independent biotechnology firm in the world. The market for biotechnology products is expanding significantly, and the market will continue this expansion. Amgen had a record of accomplishment of delivering solid growth earnings, causing it to be selected by Collins and Hansen as a GBC company. However, in 2009, 2008 and 2007, Amgen incurred restructuring charges of $70 million, $148 million and $739 million respectively, related primarily to staff separation costs, asset impairment charges and accelerated depreciation and loss accruals (Flinn, 2012). These developments had an adverse impact on sales and operations. However, Amgen continued its disciplined focus on solving compelling medical problems using good science (Bashe, 2008).
AcquisitionsJt VenturesPersonnelPhilanthropyLitigationFinancialRecognition13132918101837VOL. 12 SENA and OLSEN: AN EXAMINATION OF GREAT BY CHOICE: REDUX 41
A. 2. Productive Paranoia [Agree]
Amgen’s patents protected them from competition. This is an example of adherence to “Productive Paranoia.” However, manufacturing difficulties, disruptions or delays have limited the supply of their products and consequently their product sales. Current economic conditions have magnified certain risks that affect their business (U.S. Security and Exchange Commission, 2010). Another measure for productive paranoia, as used by Collins and Hansen, is to examine the relative current ratio and debt to equity ratios for each firm. The data for all firms are included in Table 6. This data is comparable to the data provided in GBC, which stated that the great companies have current ratios better than comparisons 72 percent of the time and better total debt-to-equity ratios 64 percent of the time. Essentially, great companies have more cash and less debt to guard against uncertainties in the environment. Our analysis concurs with Collins and Hansen in that three of the four great companies had better average performance than the comparison companies in the last eleven years of the GBC period. Amgen was the exception. Amgen also had only a slight improvement in current ratio performance (7 percent) and a decrease in debt to equity ratio performance of 162 percent in the update period.
Table 6: Average Current Ratio and Debt-to-Equity Ratio
GBC (1991-2001)
Update (2002-2012)
Average Current Ratio
Percent Better than Comparison Company
Average Debt/Equity Ratio
Percent Better than Comparison Company
Average Current
Ratio
Percent Better than Comparison Company
Percent Better than GBC period (1991-2001)
Average Debt/Equity Ratio
Percent Better than Comparison Company
Percent Better than GBC period (1991-2001)
Amgen (1)
2.56
-63%
0.13
-31%
2.74
-4%
7%
0.34
-56%
-162% Genentech (1)
4.16
0.09
2.85
-31%
0.15
-67%
Biomet
5.18
0.03
3.19
-38%
0.31
-933% Kirschner (3)
Intel
2.50
28%
0.08
288%
2.70
37%
8%
0.08
2700%
0% AMD
1.80
0.31
1.70
-6%
2.24
-623%
Progressive
na
0.39
18%
na
0.34
-15%
13% Safeco
na
0.46
na
0.29
37%
Microsoft
3.57
32%
0.00
much better number
2.74
13%
-23%
0.06
-83%
slightly worse
number Apple
2.42
0.30
2.39
-1%
0.01
97%
Southwest Airlines
0.91
0.50
1.10
21%
0.44
12% PSA (3)
Stryker
2.46
0.64
3.12
27%
0.09
86% USSC (3)
Notes:
1. Amgen-Genentech update comparison period is 2002-2008.
2. Biomet update period 2002-2007.
3. Company dropped from analysis due to acquisition or discontinued business early in the update period.
Our conclusion is that Amgen is not consistently applying this practice.
A. 3. Empirical Creativity [Strongly Agree]
Amgen expected to encounter increasing competition from bio-similar products and anticipated this impact on their profitability. Regardless they did not hesitate to stop late-stage studies (Berkrot, 2012). At Amgen, research is very creative. In some ways, chaos runs rampant. The most persistent, aggressive, and gifted people make it to the top. Their research function consisted of constantly shifting networks of projects and allegiances that formed, prospered, and 42 JOURNAL OF BUSINESS INQUIRY 2013
faded with little direction from above. A project that looked promising would attract scientists. The mechanism of shepherding researchers was mostly informal. In general, about twenty percent of a scientist’s time was free to spend on their own projects (Berkley, 1992). This creates a hail of “bullets” which then are followed by cannon balls as critical mass is reached through the attraction of scientists to successful projects.
A. 4. Level 5 Ambition [Neutral]
Over the period of our study, it appeared that Amgen believed they could make change happen simply by force of personality, position, or intellect. Under CEO Sharer’s leadership, Amgen overhauled its management team, altered its culture, and fired a couple of cannon balls (George, 2007). The CEO was considered an interventionist when it came to acquisitions and mergers by not allowing the process to follow naturally. There were several situations where Amgen restructured or stepped back from planned expansions (Fubini, 2006). In 2008, Sharer ranked second for CEOs with the highest disapproval ratings and was at risk of being ousted (54 percent disapproved) (Glassdoor Team, 2008). Sharer retired in 2011.
In Research Question 1, we proposed that Amgen stopped using GBC practices based on our financial analysis. However, our examination of the four practices found that Amgen did indeed continue the practices. This finding rejects our proposition and does not support the findings in GBC.
B. Proposition 2: Genentech Started Using GBC Practices
Genentech’s greatest strength lies in its research team and in its capacity to innovate specialized, highly effective drugs. However, a number of other extraneous, and possibly unnecessary, investments had drained Genentech’s resources. This included the development of unspecialized drugs for health problems such as asthma. In addition, one division consists of highly educated and specialized marketing experts that sell drugs to specialists.
In 2009, Roche acquired control of Genentech, making it a wholly owned subsidiary. Consequently, our research on Genentech has focused on the public period from 2002 to 2008 before their acquisition.
As part of our research on Genentech, we examined and categorized 125 media releases over the eight years of available data. Table 7 presents a summary of the categorizations.
Table 7: Genentech Media Releases
B. 1. Fanatic Discipline [Somewhat Agree]
Genentech had an extensive internal pipeline of potential blockbuster drugs. They are noted for their investment in research and development of products while maintaining a talented workforce in research labs and marketing divisions. The company’s defining success lays in the field of oncology, producing the most effective and advanced cancer drugs on the market.
AcquisitionsJt VenturesPersonnelPhilanthropyLitigationFinancialRecognition618267111319VOL. 12 SENA and OLSEN: AN EXAMINATION OF GREAT BY CHOICE: REDUX 43
Most large companies in the pharmaceutical industry purchase small companies to acquire new product pipelines. Genentech preferred to spend its resources strengthening its own pipelines instead. This strategy allowed Genentech to support small companies with promising drugs or technology that were still in an early stage of development licensing. They had only six acquisitions over the ten years of our analysis, excluding their own acquisition by Roche. Overall, they entered into 18 alliances or joint ventures over the period of 2002 to 2008 (Hall, 2012). In addition, Genentech had a strong reputation in the buying and licensing of compounds and technology platforms at all stages of development.
B. 2. Productive Paranoia [Agree]
Genentech had a sense of mission and a role in efforts to improve the lives of patients. Employees regard Genentech to be a great place to work. Each year, the IT part of the organization spends two days off-campus at the "Full Spectrum IT" event, where the group meets face-to-face with patients who have benefitted from the medications developed for people with difficult-to-treat diseases (King, 2010). Sensitivity to employees and customers is a special way that Genentech demonstrated the practice of productive paranoia. They obsess about getting this right and reap the benefits. On the other hand, from a fiscal standpoint, Genentech’s current ratio and debt-to-equity ratio slipped in comparison to the GBC period (Table 6).
B. 3. Empirical Creativity [Somewhat Agree]
Genentech was conscious of the pharmaceutical industry changes over that time, becoming more regulated and more cost-conscious. According to their Marketing VP, “For many years Genentech was immune to some of these things—they were insulated, because they were delivering on the innovation, on breakthrough drugs, and technology—many salespeople took things for granted—that this was a job that was stable, where they could earn a great living and be able to provide for their families. And then all of a sudden that gets shattered” (Looney, 2011). Similarly a senior scientist noted that Genentech was “regarded as the number one company in innovation, and employees felt blessed and pleased to work with Genentech and Roche;” and “They come up with these incredible brands, patient services and communications that just scream innovation” (Iskowitz, 2011). These observations support the proposition that Genentech has become good at systematically investing in, and benefitting from, their creativity.
B. 4. Level 5 Ambition [Strongly Agree]
The past 18 months at Roche have involved rapid change and reorganization. Genentech was fully integrated along with the seamless design of a research and early development organization. A Genentech employee that was part of this process noted that change was not to be feared, but instead to be embraced—change can be exciting and often leads to new and interesting opportunities (Anonymous, 2010). In our study of Genentech, we noted a consistent pattern of recognition internally and externally that resulted in Genentech’s being considered a model employer (King, 2012). In 2008, Genentech’s Art Levinson came in as the number one rated CEO with a 92 percent approval rating (Glassdoor Team, 2008).
In Research Question 2, we proposed that Genentech started using GBC practices based on our financial analysis. In addition, our examination found that Genentech did adopt the GBC 44 JOURNAL OF BUSINESS INQUIRY 2013
practices. This could explain their improved performance relative to their comparison company and industry.
C. Proposition 3: Biomet Maintained the Use of GBC Practices
Established in 1977, Biomet, Inc. and its subsidiaries design, manufacture and market products used primarily by musculoskeletal medical specialists in both surgical and non-surgical therapy. At the end of 2006, our last year of analysis, Biomet as a family of companies enjoyed steady growth at levels exceeding market expansion. Biomet’s performance demonstrates a responsive, customer-oriented approach to the healthcare market. Biomet thrives by strengthening its commitment to innovation and partnership.
Biomet recorded its twenty-eighth consecutive year of record year-over-year sales and earnings growth. 2006 was a milestone year for the company, attaining net sales exceeding $2 billion, an increase of 8 percent over last year’s sales, which approached $1.9 billion. In September 2007 the company was acquired by a consortium of private equity firms and ceased trading on NASDAQ.
C. 1. Fanatic Discipline [Agree]
From its beginn