Activity-based costing, total quality management and business process reengineering: their separate and concurrent association with improvement in financial performance.

Cagwin, Douglass ; Barker, Katherine J.

ABSTRACT

This study examines whether use of the strategic business initiatives activity-based costing (ABC), total quality management (TQM), and business process reengineering (BPR), are associated with improvement in financial performance. Top executives of 305 firms operating in the motor carrier industry furnished information regarding use of the initiatives. Dependent variable information is obtained from financial statement data filed with the U.S. government. Multiple regression analysis is used to identify the improvement in ROA associated with the use of each initiative, and concurrent use of two initiatives.

A simple effect for use of TQM and BPR is confirmed. Context-specific benefits obtained from concurrent use of ABC with BPR and TQM are identified. It appears that ABC functions as an enabler of other improvement initiatives since its use provides the information necessary to optimize the effectiveness of TQM and BPR. The positive findings regarding ABC are of particular interest to practicing and academic accountants because they are often the primary proponents and administrators of ABC, and there has been little empirical evidence of ABC efficacy.

INTRODUCTION

The focus on cost, quality and time has generated many management changes with significant accounting implications (Smith, 1998). These changes increasingly include the implementation of strategic business initiatives such as activity-based costing (ABC), total quality management (TQM), and business process reengineering (BPR). Profit-maximizing firms would not implement strategic business initiatives if they did not expect a net financial benefit from their use; however there has been little empirical evidence that demonstrates that ABC, TQM, or BPR improves financial performance in any industry.

Researchers have often suggested that ABC and other strategic business initiatives complement and enhance each other, rather than being individually necessary and sufficient conditions for improvement (Cooper and Kaplan, 1991; Anderson, 1995; Evans and Ashworth, 1995; Player and Keys, 1995; Swenson, 1998). There has been no empirical investigation of context-specific benefits obtained from ABC or from the concurrent use of TQM and BPR. In this study the contexts investigated include use of ABC to enhance the benefits of other initiatives, concurrent use of BPR and TQM, and non-concurrent use of ABC, TQM, and BPR.

The purpose of this study is to investigate the improvement in financial performance associated with the single and concurrent use of the strategic business initiatives ABC, TQM, and BPR. Data is obtained through a cross-sectional mail survey of 305 motor carrier industry top executives and from a database containing financial statement information reported to the U.S. government. Multiple regression analysis is used to investigate the association between use of initiatives and improvement in financial performance (proxied by ROA) and to identify positive context-specific effects from the use of TQM with BPR and of ABC with TQM and BPR.

This research adds to the limited body of empirical strategic business initiative research in four ways. The first contribution is to provide empirical evidence that the benefits claimed by initiative advocates are net benefits. Second, this research confirms the existence of a context-specific benefit from concurrent use of ABC with TQM and BPR. Third, the study focuses on the motor carrier industry, an important member of the service sector, which has become the dominant sector of the U.S. economy. Researchers have often postulated, but not tested, the efficacy of initiatives in a service setting. Finally, limitations of previous research are addressed (i.e., the lack of control for simultaneous use of multiple initiatives, and prior level of performance).

The remainder of the paper is organized as follows: Section II defines and describes strategic business initiatives, situates this study in the context of past research, and provides hypothesis development. Section III describes sample selection and the survey instrument. Section IV describes the methodology used, including variable selection and specification. Results are presented in Section V, sensitivity tests in Section VI, and the summary and concluding remarks are contained in Section VII.

BACKGROUND AND HYPOTHESIS DEVELOPMENT

A strategic business initiative is an innovative business technique, strategy or technology that is purported to increase business success. All initiatives broadly advocate change through continuous improvement, but each accomplishes continuous improvement somewhat differently. For example, TQM emphasizes "doing the right thing the first time" and ABC advocates using activity-based cost and performance measures to reveal non-value added activities (Gupta et al., 1997). In the last decade, two main developmental models have dominated the organizational world, TQM and BPR (O'Neill and Sohal, 1999), while ABC has been a subject of intense interest by practicing accountants, consultants, and academicians. The recent attention that ABC, TQM, and BPR have received motivated their selection for the current study. Each initiative is discussed below.

Total Quality Management (TQM)

Total quality management (TQM) has been one of the most popular business models of the last twenty years, widely embraced by many organizations (Forza and Filippini, 1998; Hendricks and Singhal, 1999). TQM is a concept based on continuous improvement in the performance of all processes in an organization and in the quality of the products and services that are the outputs of those processes. It is defined by ISO 8402 as a "management approach of an organization centered on quality, based on the participation of all its members and aiming at long-term success through customer satisfaction and benefits to all members of an organization and to society." Its overriding consideration is building, rather than inspecting, quality into output (Shank and Govindarajan, 1994).

In the literature TQM is described as a collective, interlinked system of quality practices that is associated with organizational performance (Choi and Ebock, 1998). Several quality experts have suggested that a commitment to total quality will result in improved performance in profitability measures (Juran and Gryna, 1970; Crosby, 1979; Feigenbaum, 1986; Deming, 1988). Others have raised concerns about whether TQM programs have actually generated these improvements (Hendricks and Singhal, 1997). As noted by Samson and Terziovski (1999) and Hendricks and Singhal (1997), there is a lack of empirical research confirming the effectiveness of TQM.

Business Process Reengineering (BPR)

Business process reengineering (BPR) has achieved popularity among businesses in a very short period of time (O'Neill and Sohal, 1999). First introduced by Hammer (1990) and Davenport and Short (1990), BPR is a discontinuous improvement practice used by companies to streamline operations and provide enhanced value to customers (Fliedner and Vokurka, 1997). It is defined as the thorough analysis, fundamental rethinking, and radical redesign of business processes to achieve dramatic improvements in critical measures of performance (Hammer and Champy, 1993). BPR is essentially value engineering applied to the system to bring forth, sustain, and retire a product or service, with an emphasis on information flow (Hammer and Champy, 1993).

Both BPR and TQM share certain principles and adopt a process perspective (Jaworski and Kohli, 1993). Klein (1993) suggests that BPR is much more radical than TQM. It is differentiated from TQM in that rather than promoting continuous improvement, it entails fundamental rethinking and radical redesign of business processes. Instead of gradual continuous improvement, BPR is abrupt and discontinuous change (Earl and Khan, 1994; Altinkemer et al., 1998).

BPR is purported to produce positive results for firms including improvements in critical, contemporary measures of performance, such as cost, productivity, service, customer satisfaction, and speed (Fliedner and Vokurka, 1997; Raymond et al., 1998). It can be used to bring about major internal and external quality increases, thus increasing value for both the employee and the customer (Dean, 1996).

In the past decade BPR has received considerable attention in the literature; however empirical research, including the relationship between the adoption of BPR and financial performance, has lagged its use (O'Neill and Sohal, 1999).

Activity-Based Costing (ABC)

The arguments in support of activity-based costing (ABC) are generally based on the superiority of information that can be generated in comparison with that generated by traditional cost management systems (The terms activity-based costing (ABC) and activity-based management (ABM) are sometimes used interchangeably. Strictly speaking, ABC refers only to the actual techniques for determining the costs of activities and outputs that those activities produce. Some researchers and practitioners prefer to use the term activity-based management (ABM) when they describe how the activity information is used to support operating decisions. As in Swenson (1995) and Krumwiede (1998), this study defines ABC very broadly to include activity-based costing and activity-based management).

According to the theory of information economics, better information leads to better decision-making, and better decision-making enhances firm value. For example, Drucker (1995) states that history has shown time and again that a company that enjoys a cost advantage, by correctly identifying and managing the costs of the entire value chain, overtakes the established leaders in a market segment.

Many authors also recommend using ABC to support process improvement (Turney, 1991b) and to develop cost-effective product designs (Cooper and Turney, 1989); however several reservations have been expressed regarding the efficacy of ABC (Innes et al., 2000). Recent research has been successful in detecting a link between the use of ABC and financial improvement in specific business environments. Kennedy and Afleck-Graves (2001) were successful in linking the implementation of ABC with net improvement in financial performance for manufacturers. Ittner et al. (2002), and Cagwin and Bouwman (2002) found that ABC's contribution was an indirect rather than a direct effect on improvement in financial performance.

ASSOCIATION BETWEEN INITIATIVE USE AND IMPROVEMENT IN FINANCIAL PERFORMANCE

The theories of diffusion of innovations (Kwon and Zmud, 1987), transaction cost economics (Roberts and Sylvester, 1996), and information technology (Dixon, 1996) suggest that organizations adopt an innovation to obtain benefits that directly or indirectly affect financial performance measures. There have been numerous claims and counterclaims, rarely supported by objective and rigorous empirical evidence, regarding whether programs have yielded net financial gains. Evidence of the benefits of these systems is largely restricted to theoretical models and anecdotal information obtained from case studies that are dependent on anecdotal information related by practitioners (for examples, see Goyal and Deshmukh (1992), and Golhar and Stamm (1991) for JIT; Bruns and Kaplan (1987), Barnes (1991), Brimson (1991), and Harris (1990) for ABC; Romney (1995), and Dean (1996) for BPR; and Sankar (1995) for TQM).

Results from the limited empirical research examining the link between TQM, BPR, and financial performance are mixed (Wouters et al., 1999). There is weak evidence of an association between TQM and financial performance. To date, other than an exploratory study by Altinkemer et al. (1998) that was hampered by a small sample size of 35 firms, no studies have empirically measured financial performance benefits obtained from using BPR.

Kaynak (1996) found that self-reported "financial and market" performance was enhanced for firms using a combination of TQM and just-in-time purchasing. Easton and Jarrell (1998) found evidence that a very broadly defined TQM is associated with the variance between actual financial performance and that forecasted by Value Line analysts. Finally, Hendricks and Singhal (1997) found a link between change in ROA and implementation of TQM for a sample of quality award winners--evidence that firms that possess a "higher level of seriousness and commitment than other firms" and that think enough of their quality programs to apply for awards have seen financial improvement. Recently, researchers have been successful in detecting a link between use of ABC and improvement in financial performance in specific business environments. Kennedy and Afleck-Graves (2001) were successful in linking the implementation of ABC with a net improvement in financial performance in manufacturers; however, Ittner et al. (2002), and Cagwin and Bouwman (2002) found that ABC's contribution was an indirect, rather than a direct effect on improvement in financial performance.

Firms adopt initiatives in attempts to gain or maintain cost and market advantages (Kinney and Wempe, 1998). These advantages should in turn lead to improvement (or to maintenance of favorable values) in composite financial indicators, in the face of competitive pressures. The first hypothesis is in three parts and is consistent with hypotheses contained in prior research, suggesting that initiatives individually contribute toward an improvement in financial performance.

H1a: There is a positive association between use of ABC and
improvement in financial performance relative to he improvement in
financial performance of non-ABC users.

H1b: There is a positive association between use of TQM and
improvement in financial performance relative to the improvement in
financial performance of non-TQM users.

H1c: There is a positive association between use of BPR and
improvement in financial performance relative to the improvement in
financial performance of non-BPR users.

CONTEXT-SPECIFIC BENEFITS OBTAINED FROM CONCURRENT USE OF INITIATIVES

There may be context-specific benefits (positive or negative) leading to various optimal combinations of factor inputs (e.g., initiatives and management systems (Capon et al., 1988). If firms are rationally maximizing value they would choose initiative combinations that they believe lead to this objective.

There has been considerable research interest in investigating possible context-specific benefits of initiatives with management information systems and management techniques. Recently, researchers have had mixed results in investigating the relationship between TQM and information and reward systems (Ittner and Larcker, 1995; Sim and Killough, 1998), TQM and manufacturing performance measures (Chenhall, 1997), TQM and human resource variables (Snell and Dean, 1992; Youndt et al., 1996) and TQM and strategic priorities, management techniques, and management accounting practices (Chenhall and Langfield-Smith, 1998). There has been no empirical research investigating the combined effects of TQM and BPR, or of ABC and other strategic business initiatives.

Context-Specific Benefits Obtained From Concurrent Use Of BPR With TQM

An increasing number of authors have suggested a need for both continuous (i.e., TQM) and discontinuous (i.e., BPR) improvement (O'Neill and Sohal, 1999). For example, Harrison and Pratt (1992) suggest that TQM and BPR can and should form an integrated strategic management system within organizations and authors such as Furey (1993), Taylor (1993), and Chang (1994) describe programs that integrate TQM and BPR. Most authors agree that if BPR helps focus attention on transformational change, without damaging core competencies and TQM based continuous improvement, it could effectively benefit an organization (O'Neill and Sohal, 1999). These arguments provide additional support for the hypothesis that both TQM and BPR contribute separate positive direct effects on the financial performance of firms (H1).

There are, however, three conflicting sets of arguments regarding a positive context-specific effect obtained from concurrent use of the two initiatives. Some arguments support a hypothesis for a positive interaction between the two initiatives. O'Neill and Sohal (1999) speculate that BPR is less likely to succeed outside TQM, since it uses the methods, processes, and customer orientation of TQM to deliver changes. Cole (1994) concludes that the two initiatives complement each other and that each is a "building block" for the other. In research similar to Cole (1994), Zairi and Sinclair (1995) and Gadd and Oakland (1996) find that TQM and BPR can be considered as two distinct approaches capable of coexisting in the same organization, but used at different times to achieve different levels of performance improvement (i.e., simple but not context-specific benefit). Jarrar and Aspinwall (1999b) feel that there is a clear opportunity to "unite them to fill each other's gaps." Wright (1995) argues that reengineering of existing processes (BPR) can be leveraged through existing TQM processes.

Finally, some authors argue that TQM might even be a barrier to the change required by BPR, a negative context-specific benefit (Redman and Grieves, 1999). For example, Hammer (1991) described sequential performance improvements using the two techniques and warned against using the two concurrently.

Because of the conflicting arguments regarding a context-specific effect obtained from concurrent use of TQM and BPR, the second hypothesis is not directional:

H2: The improvement in financial performance of firms that use TQM
and BPR concurrently is different than that associated with each
initiative used alone.

CONCURRENT USE OF ABC WITH TQM AND BPR

A number of academics, consultants and practitioners have championed the need to link management accounting and strategy (Rangone, 1997). Armitage and Russell (1993), and Shepard (1995), argue that many organizations that are implementing quality systems will see little return on assets in terms of performance improvement, primarily because managers are unable to identify specific opportunities for improvement.

The literature contains many references to the capability of ABC in helping to rectify this lack of information. McConville (1993) stated that ABC complements TQM by providing quantitative data on improvements made as a part of a TQM program. Beheiry (1991) commented that ABC quantifies additional costs regarding non-conformance of requirements by focusing attention on the activities that consume the majority of costs. Steimer (1990) examines ABC in the context of the services performed by the home office for multiple business units. He argues that ABC is perfectly suited to TQM because it encourages management to analyze activities and determine their value to the customer. Shepard (1995) believes that an economics-of-quality approach can be integrated with ABC in order to maximize return-on-investment and begin long-term continuous improvement. Anderson and Sedatole (1998) state that many companies have found that ABC fits well with their cost of quality framework and recommend extending the union of TQM and ABC to the design process. Evidence of the context-specific benefit of TQM and ABC was found in case studies performed by Cooper et al. (1992) where all five manufacturing companies studied found ABC and TQM to be highly compatible and mutually supporting.

Cokins (1996) examines ABC's role in BPR. He states that ABC identifies business processes that are non-value-added and can either be reduced or eliminated, which is the heart of BPR. Hammer (1990), Borthick and Roth (1993), Dean (1996), Altinkemer et al. (1998) have also commented on the value of combining ABC with BPR.

Theory and anecdotal reports support the proposition that the improved costing information and intensive analysis of business activities provided by ABC lead to improved decision-making, and therefore should be associated with improved performance; however Ittner et al. (2002) and Shields et al. (2000) argued that ABC has an indirect effect on financial performance by enhancing improvements contributed by other process improvement initiatives. Krumwiede (1998) provided additional weight to this argument by reporting that all fifteen "best practice" firms had linked ABC to another improvement initiative. Further, in an experimental setting, Drake et al. (1999) found that combining ABC with an innovative activity can produce a higher or lower level of firm profit, depending on whether workers had incentives to cooperate or could use the improved ABC information for their individual benefit (i.e., increase personal as opposed to group output).

Although, as Shaw (1998) notes, ABC is now recognized as a fundamental business methodology for enabling business improvement, no empirical research has specifically targeted the combination of ABC with either TQM or BPR and their combined association with improvement in financial performance, leading to the following hypothesis:

H3: The financial performance of firms that use ABC concurrently
with TQM or BPR has improved more than the improvement associated
with each initiative used alone.

SAMPLE SELECTION AND SURVEY INSTRUMENT

Most research regarding strategic business initiatives has focused on the manufacturing segment of the economy, however, the major changes that manufacturing companies have experienced in recent years (e.g., technology improvements and globalization) have also occurred in virtually all types of service organizations (Atkinson et al., 1995), and researchers have found that strategic business initiatives can be applied in all types of organizations (Rotch, 1990; Tanju and Helmi, 1991; Jarrar and Aspinwall, 1999a). Since non-manufacturing activities represent the majority of the North American economy, there clearly is opportunity for research to focus on non-manufacturing settings, including transportation (Shields, 1997).

This study focuses on a single service industry: the motor carrier industry (SIC 4213). Restricting to a single industry reduces noise, thereby increasing statistical power, and consequently provides a higher likelihood of identifying valid relationships. The motor carrier industry is selected because of 1) the importance of the motor carrier industry to the nation's economy, 2) interest of the members of the industry in use of business initiatives that can potentially improve their competitive positions, and 3) the availability of detailed financial statement data for members of the industry.

Although it can be argued that the focus on a single industry tends to make results less generalizable than a study that crosses industries, the findings of this study have a wide appeal. The motor carrier industry generates about five percent of the gross domestic product and hauls approximately 55 percent of all domestic freight volume. It affects virtually every organization in every industry and governmental agency in the U.S. economy. Furthermore, transportation is a major component of business logistics and is usually the single largest cost element in the logistics function. Companies not only contract with for-hire carriers but very often maintain private fleets of long-haul vehicles.

The independent variable data used in this study (other than LEVEL) are extracted from an instrument that was used to collect data intended for use both in this study and also for other in-depth analyses of the trucking industry. The instrument is based on a thorough review of prescriptive, conceptual, practitioner, and empirical motor carrier industry literature. Content validity is addressed by asking representatives of the trucking industry, industry experts, and a group of faculty experienced in management innovation and survey research to review the instrument for clarity and meaning. Modifications were made as appropriate.

Most of the questions are closed-ended and ask the respondent to rate or assess the item on a seven-point Likert balanced scale, anchored by 1 = "Almost Always Avoid," and 7 = "Almost Always Use." Some items ask for specific numerical information (e.g., "truckload percent of total freight revenue").

Procedures prescribed by Dillman (1999) are followed to maximize response rates. Specific steps taken to strengthen this study include 1) pre-calling to obtain name of the CEO and to verify the mailing address, 2) sending a preliminary letter and brief summary of the project, 3) pre-calling to ask if the CEO had any questions, 4) including a personalized cover letter, 5) promising to send a summary of results and a technical report, 6) promising confidentiality, 7) including a stamped, self-addressed envelope for reply, 8) mailing a reminder letter at three weeks past initial mailing, and 9) mailing a reminder post card after seven weeks.

The initial population for this study consisted of the 2,002 firms that reported to the Interstate Commerce Commission and were included in the 1999 TTS Blue Book of Trucking Companies. In order to focus on companies of sufficient size to have an established set of practices for conducting business, the population is limited to those companies that have at least thirty employees or $5 million in gross revenues. This constraint reduced the population by 383. From the remaining 1,619 companies, 1,100 were randomly selected for inclusion in the study. Of these, six were eliminated because they were Canadian companies, two were unable to be contacted by telephone or letter, nine had gone out of business, and 14 withdrew or refused to cooperate upon initial contact. The remaining 1,069 firms comprise the final sample. A total of 332 responses were received, a response rate of 31.1 percent. Because of their larger size, the 332 sample firms represent 16.5 percent of the firms in the TTS database but contribute 23.1 percent (equity) to 41 percent (ton-miles) of the aggregated totals.

Financial data are available for 305 of the responses for 2001 & 1998. Sample selection and response are summarized in Table 1.

The median industry, company, and position experience of the respondents is 25, 17, and nine years, respectively and 96 percent are of the rank of controller or officer (70 percent are president, owner, or CEO). The extensive experience and high rank of the respondents lend considerable credibility to the survey responses.

METHODOLOGY

The impact of strategic business initiatives on a firm's improvement in financial performance is examined using the following model:

PERFORMANCE CHANGE= f (Initiative Use, Initiative Use Interactions, Control Variables)

where PERFORMANCE CHANGE is the change in ROA, measured for year t +3 minus year t. The initial variables are the set of binary measures of use of TQM, BPR and ABC, and are used to identify simple effects (H1). Interaction terms are created for concurrent use of TQM with BPR (H2) and of ABC with TQM and with BPR (H3). Variables of interest are discussed below.

Change in Return on Assets (ROA CHANGE)

ROA, defined as after-tax net income scaled by total assets is generally accepted as a composite financial performance variable in empirical research. Seven studies that recently attempted to measure improvement in financial performance resulting from the implementation of JIT (Husan and Nanda, 1995; Balakrishnan et al., 1996; Boyd, 1996; Biggart, 1997) and TQM (Engelkeyer, 1991; Dixon, 1996; Easton and Jarrell, 1998) have operationalized financial performance through the use of ROA as defined above. In addition, Ittner and Larcker (1995), Hendricks and Singhal (1997), and Ittner et al. (1999) used ROA as a dependent variable in their studies of TQM and TQM and supplier strategies. Furthermore, previous research shows a high correlation between ROA and other profitability measures (Prescott et al., 1986) and suggests that ROA can be more readily available in business units than other measures (Jacobson, 1987). For these reasons ROA is selected as the primary dependent variable.

Testing improvement in financial performance poses significant measurement problems. As Roberts and Silvester (1996) observe, numerous complications arise, including:

1. Modeling a company's "expected" profitability against which to
compare realized profitability achieved after use of an initiative

2. Controlling for concurrent changes in the organization

3. Controlling the breadth of implementation and integration of
initiatives throughout the firm.

In general, comparison of "expected profitability" requires either specification of control variables which describe the industry in which the firm operates or the use of "industry mean-adjusted" measures. In the current study, expected profitability is addressed through restricting the study to a single industry, using a fixed period of time (the change from 1998 to 2001), which provides control for macroeconomic and industry-specific factors that affect all firms equally, and controlling for differences in the three segments of the industry. These restrictions allow comparison of the profitability of initiative users against that expected without use, proxied by the performance of equivalent non-users.

Concurrent changes in the organization are addressed through identifying and controlling for use of other initiatives and for prior performance. Control for use of other initiatives separates the effects of individual initiatives and allows comparison of users of an individual initiative to nonusers of that initiative. Controlling for the moderating effects of length and breadth of implementation is addressed by inclusion of variables measuring extent of use derived from survey responses.

Archival dependent variable information is obtained from the TTS database. The TTS Blue Book of Trucking Companies is published by Transportation Technical Services, Inc., New York (TTS). The majority of Blue Book data is extracted from annual reports (Form M) that carriers file with the Interstate Commerce Commission. Form M requires use of standardized accounts defined in the Uniform System of Accounts for Motor Carriers of Property published by the American Trucking Associations, Inc.

Variables of Interest (ABC, TQM, BPR)

The simple variables of interest measure use of the initiatives TQM, BPR, and ABC. These initiatives are established initiatives of significant interest to the motor carrier industry. ABC is of particular interest to the accounting profession.

Cross-sectional survey data are collected regarding the extent of use (diffusion) of initiatives at the survey date (mid-1999). The variables of interest are developed from 7-point Likert balanced scale (Dillman, 1999) responses to a survey item introduced as "How much do you avoid or use the following competitive tactics to realize your competitive strategies?" Possible responses are:

 1 = "Almost Always Avoid,
 2 = Mostly Avoid,
 3 = Sometimes Avoid,
 4 = Neither Avoid or Use,
 5 = Sometimes Use,
 6 = Mostly Use, and
 7 = Almost Always Use."

In addition, respondents furnished the year that they began use of each initiative. Responding firms are classified as significant users if their response was a 6 or 7 to the question regarding ABC, TQM, or BPR, and, because strategic initiatives are inherently multi-year projects, their year of beginning use was not 1999. Binary variables (BPR, TQM and ABC) differentiate significant users from the remainder of the sample. These variables are the variables of interest for testing Hypothesis 1, which tests for a positive simple effect from use of the individual initiatives.

CONTROL VARIABLES

The implications of three control variables--two variables denoting type of company, TL and LTL (with specialized carrier the default), and beginning mean-adjusted LEVEL of performance are discussed in the following section.

Type of Company (TL, LTL)

The motor carrier industry is not entirely homogenous, but can be partitioned into three segments. One important distinction is between less-than-truckload (LTL) and truckload (TL) carriers. LTL carriers provide service to shippers who tender shipments lower than the minimum truckload quantities (i.e., 500 to 15,000 pounds). Consequently, the LTL carrier must consolidate the numerous smaller shipments into truckload quantities for inter-city movement and break down full truckloads at the destination city for delivery in smaller quantities. In contrast, the truckload carrier picks up a truckload and delivers the same truckload at destination.

Carriers may also be classified by the type of commodity they haul, general or specialized commodities. Specialized equipment carriers are carriers of goods requiring special handling (e.g., liquefied gases, frozen products, automobiles, or household goods). A specialized carrier is not permitted to transport other specialized commodities, or general commodities, in the same equipment.

The LTL segment of the industry requires significant capital assets, including terminal facilities and complex computer and communications systems, a skilled work force, and a large sales organization to operate a network of terminals and freight handling equipment to consolidate and distribute freight (Harmatuck, 1990). This network is generally not needed by the TL carrier. Specialized equipment carriers usually have larger investments in equipment and terminals than those transporting general freight.

Industry type has been demonstrated as important in previous work (e.g., Zmijewski and Hagerman, 1981; Healy, 1985; Watts and Zimmerman, 1986; Capon et al., 1988) explaining cross-sectional variation in financial performance. In effect, the characteristics of the three types of service offered by carriers (TL, LTL, and specialized) reflect three mini-industries. The impact of industry type is appropriately addressed through use of control variables. Because firms often offer more than one type of service, participating in more than a single mini-industry, self-reported continuous variables measuring the percentage of total freight revenues attributable to each classification (TL and LTL, with specialized carrier the default) are created. These variables provide control for differences in competitive environments, accounting practices, and other classification-specific attributes that may affect performance. It is expected that LTL will be negatively signed because that segment of the industry has been underperforming the other segments during this decade.

Level of Performance (LEVEL)

As Balakrishnan et al. (1996) noted in their discussion of JIT, a firm's pre-adoption operating efficiency will influence its ROA response to the increased efficiency of initiative use. Because it appears that there are continuing pressures that tend to pull the performance of firms towards the average (Bernard, 1994), higher performing companies may implement business initiatives to retain their comparative advantage, rather than to show improvement. This condition causes problems in detecting the association of the initiatives with improved financial performance (Huson and Nanda, 1995). In addition, firms are generally unable to sustain extremely poor performance for an extended period of time. They must either improve their performance towards the mean, or go out of business and thus would be not included in a cross-sectional study. These conditions may effectively create a "collar" around the performance of a sample firm, a ceiling limiting the improvement of the top performers and a floor limiting the deterioration of the already poor performers, resulting in a phenomenon with the statistical characteristics of mean reversion.

Significance of the variable of interest could result from lack of control for the effects of this "collar." If below average performers tend to implement initiatives more than successful firms, an upward change in performance may be due to the pressures noted above that tend to pull the performance of firms towards the average rather than efficacy of the initiatives. To control for the effects of mean reversion, beginning of test period (t) mean-adjusted level of performance (ROA) is included as an independent variable. It is expected that the sign of the regression coefficient associated with this variable will be negative (i.e., performance will be drawn toward the mean).

Regression Model

Testing of the three hypotheses is accomplished through estimation of the following OLS multiple regression as follows:

ROA CHANGE = A + [B.sub.1]TQM + [B.sub.2]BPR + [B.sub.3]ABC + [B.sub.4]TL+ [B.sub.5]LTL + [B.sub.6]ROA + [B.sub.7]TQM * BPR + [B.sub.8]TQM * ABC + [B.sub.9]BPR * ABC

The expected signs of the coefficients are: [B.sub.1] through [B.sub.3], and [B.sub.8] through [B.sub.9], positive; [B.sub.5] and [B.sub.6], negative; [B.sub.4] and [B.sub.7] are not predicted.

DESCRIPTIVE STATISTICS

Statistics relating to the use of TQM, BPR and ABC are reported in panel A of Table 2. Over thirty percent of the respondents indicated that their firm "mostly" or "almost always" uses TQM or ABC, with 49 (16%) making heavy use of both. As might be expected given the magnitude of firm disruption, fewer respondents (13.1%) use BPR heavily, and approximately nine and seven percent use BPR with TQM or ABC. There appears to be an adequate balance of users and non-users (control firms) to provide the contrast necessary to obtain adequate statistical testing power.

Descriptive statistics relating to the dependent and control variables used in statistical testing are presented in Table 3. The median (mean) change in ROA is slightly negative 0.6 percent (positive 0.1 percent) from 1998 to 2001, reflecting the recent decline in profitability of the industry. The median level of performance for 1998 was a 3.6 percent ROA. Because the sample includes somewhat larger and less TL oriented firms than the industry population, this performance could indicate reduced profitability for the LTL segment of the industry.

The correlation matrix of the simple effect and control variables is shown in Table 4. As expected, use of initiatives is moderately positively correlated, with individual correlations ranging from 0.19 for ABC with BPR to 0.32 for BPR with TQM. Initiative users also are more likely to be less-than-truckload (LTL) companies than truckload carriers (TL). Consistent with these pair-wise correlations, regressions of individual variables on the remaining independent variables show that the initiative variables have a moderate multivariate relationship with significance levels in the alpha equals 0.10 range. ABC and BPR are also moderately correlated with LTL. In no case does R2 exceed 0.19 for these regressions. The extent of these correlations does not suggest that correlation among variables is a serious econometric issue.

There are statistically significant negative correlations between prior level of performance and BPR (-0.22) and ABC (-0.12), an indication of possible endogeneity. Lower performing firms tend to use BPR and ABC more often than high performers. A regression of initiative use on year t level of ROA (not presented) confirms that heavy users of initiatives tended to be slightly below mean in level of performance. If the form of the LEVEL variable does not adequately model mean-reversion (e.g., due to non-linearity), then performance improvement from lower performing BPR or ABC users cannot be specifically attributed to the initiative. This potential problem is addressed by performing an alternate test where the sample is partitioned into two groups based on prior performance, as shown in panel B of Table 2. As discussed later in the paper, consistent results for both groups indicate that results for the lower performing firms are not biased by improper modeling of mean reversion.

TESTS OF ASSOCIATION BETWEEN INITIATIVE USE AND FINANCIAL PERFORMANCE

Results of the formal hypothesis tests are reported in Table 5. The model is highly significant with an F-statistic of 22.90 and an adjusted R-square of .5212. BPR and TQM have positive simple effects at the [alpha]=0.05 level. Hypothesis 1 is confirmed for TQM (H1b) and BPR (H1c). Although ABC (H1a) is positively signed, it does not attain statistical significance at conventional levels (p is less than 0.157).

The interactions of ABC with TQM and with BPR are also positively significant, with the TQM-ABC combination significant at [alpha]=0.05. Significance of a positively signed interaction term confirms that there is a positive effect created from concurrent use of the two tested initiatives (i.e., there is an association with improvement in financial performance over and above that of the sum of the effects of the initiatives used in isolation). Therefore, it appears there is a positive context-specific benefit created from concurrent use of these pairs of initiatives. Hypothesis 3 is therefore confirmed.

There is no evidence, however, that concurrent use of TQM and BPR creates a positive or negative context-specific benefit. This interaction is negatively signed but is not statistically significant (p<0.567). Therefore Hypothesis 2 is not confirmed.

Results are consistent when each initiative's set of three variables (one simple and two interaction terms) is dropped from the model. In all cases, the adjusted R2 decreases. Inclusion of each initiative adds to the explanatory power of the model. Also, the ABC-BPR and ABC-TQM combinations individually contribute a positive adjusted R2. TQM-BPR does not add to explanatory power.

Of the control variables, LTL is negatively signed and significant at a=0.10, and LEVEL is negatively signed and highly significant at 0.001. The negative significance of LTL confirms that, as discussed in the motor carrier and transportation literatures, LTL and large companies did not perform as well as specialized carriers or TL companies during this period. The negative significance of LEVEL confirms the mean-reversion of earnings in the motor carrier industry.

SENSITIVITY TESTS

Regression diagnostics reveal no serious problems with multicollinearity. For a model without initiative interaction terms the condition index is 6, with no variance inflation factors above 2, well within the guidelines established by Belsley (1980). Addition of interaction terms increased the condition index to 24, still within acceptable limits. However, the addition of interaction terms tends to bias against finding simple effects and prevents interpretation of the individual initiative coefficients.

White's (1980) heteroskedasticity adjusted t-statistics are reported. Analysis of the Durbin-Watson statistics indicates no misspecification of variables. Influential data points, generally outliers with extreme values of the dependent variable, are identified through analysis of the R-student residuals. Outliers are expected because extreme observations of ratios (e.g., ROA) occur frequently relative to typical level variables. Influential data points are addressed through an iterative process whereby a regression is run, the observation with the largest r-student residual (exceeding "3") is identified, investigated and eliminated, and the regression repeated. This process results in the elimination of eight observations (2.6 percent), well within normal limits. As discussed later in the paper, sensitivity testing is performed whereby the values of the dependent variables are transformed to eliminate the need for eliminating observations. Results are robust to these specifications.

Several sensitivity tests are performed including alternative modeling of prior level of performance, and a search for missing variables (i.e., controlling for level of equity and firm size using revenues, log of revenues, total assets and log of total assets). Results are robust to these specifications of the model. Four additional categories of sensitivity tests are described in detail below.

ALTERNATE DEPENDENT VARIABLES

ROA and return on equity (ROE) are both important in analytical and empirical research (Auks, et al., 1996). It can be argued that the goal of the firm is to maximize shareholder wealth and that return on equity (ROE) is more closely tied to shareholder wealth than ROA (Brigham et al., 1999; Shapiro and Balbirer, 1999). An alternative test is performed where ROE replaced ROA as the dependent variable. Results (not presented) are robust to this alternative dependent variable definition, except that the simple effect for BPR is significant at the [alpha]=0.10 rather than at [alpha]=0.05. A limitation of the use of a ratio-based dependent variable such as ROA is that TQM and BPR often involve improving efficiencies or restructuring, which reduce a firm's asset base. This could cause ROA of users to grow even though cash flows and firm value may actually decline over the same period. To alleviate concerns over this possible bias, an alternate test is performed where the dependent variable is percentage change in net income from year t to year t +3, defined as (net income after tax for fiscal 2001--net income after tax for fiscal 1998) / (net income after tax for fiscal 1998). Results are robust except that the simple effect for BPR is significant at the alpha = 0.10 rather than at alpha = 0.05.

To eliminate the need for dropping outlier observations from the tests and to alleviate possible concern about non-normality of the continuous variables, two additional tests are performed. In the first test, logarithmic transformations [DELTA]ROA and level of ROA are substituted for the raw values. In the second the variables are winsorized to the 5th and 95th percentile of the raw variables. Other than a weakening of [R.sup.2] of the winsorized model, results are robust to these modifications.

ALTERNATIVE SPECIFICATIONS OF VARIABLES OF INTEREST

The 7-point initiative use data were transformed into binary variables because the meaning of the responses is ambiguous, specifically the interpretation of 3 = "Sometimes Avoid" and 4 = "Neither Avoid or Use" when applied to TQM, BPR or ABC. The transformation allows separation into two classes--those firms that mostly or almost use and those that never or seldom used initiatives. Presumably the users would be expected to receive positive gains from initiative use. Use of binary variables also allows separate identification and interpretation of simple and interaction effects through a factorial design; however transformation means that the inherent properties of the survey information are ignored, potentially affecting results. Therefore, the analysis is run with the original 7-point Likert measures. Results are not materially affected.

Another issue is that there may be significant implementation costs for firms that have implemented but do not heavily use initiatives. If so, categorization of these firms as non-users would bias in favor of finding positive results. To guard against this bias, alternative tests are performed where three categories of initiative use are created: nonuser = response of 1, 2, 3, or 4 for initiative use with no implementation date; these firms would presumably have immaterial gain from using the initiative and zero or insignificant cost from implementation. light user = response of 5, or response of 2, 3, or 4 with an implementation date, or response of 6 or 7 with an implementation date of 1999; these firms presumably have some gains, but gains would likely be at least offset by implementation costs. Heavy user = response of 6 or 7 with implementation date before 1999; these firms would presumably have net gains from use of the initiative. Six binary variables are created, two for each initiative. Interpretation of the results is not affected and no significance, positive or negative, is obtained from of any of the LIGHT variables. The overall benefits and costs of light use may offset each other, and/or implementation costs may not be material. This finding is consistent with Hendricks and Singhal (1997) who found no effect, positive or negative, for recent TQM implementers. Future research could investigate the alternative explanations by testing recent implementers and light users separately.

THREE-WAY INTERACTION OF VARIABLES OF INTEREST

Nine firms are users of all three initiatives (TQM, BPR, and ABC). When a term is introduced into the model whereby the three initiatives are interacted there is no improvement in adjusted [R.sup.2]. The three-way interaction is not significant and there is little change in the significance or coefficients of the variables, possibly because of the limited number firms in this category.

PARTITIONED SAMPLE

To ascertain whether the results reported above have been biased by the higher tendency of low-performing firms to use BPR and ABC, the sample is partitioned at the median level of year-t performance. Results are consistent for both the high and low performing groups. The TQM-ABC combination remains significant at [alpha]=0.05 for both the high and low performers. BPR-ABC increases to [alpha]=0.05 for the low performers. TQM-BPR does not attain significance for either group.

The simple effects also are fairly consistent but are slightly weaker, possibly due to the smaller sample size. BPR remains significant at alpha = 0.05 for the low performers but drops to alpha = 0.10 for the high performers. One could easily expect companies that are performing poorly to benefit more from dramatic, discontinuous change than those that are already performing above the industry median. TQM remains significant at alpha = 0.10 for high performers, but is not significant at conventional levels for the low performers. These results suggest that there are similar positive effects generated from use of strategic business initiatives for both high and low performing firms. It does not appear that endogeneity between beginning level of performance and initiative use is a serious concern in interpreting the results of this study.

SUMMARY AND DISCUSSION

This study investigates the use of TQM, BPR and ABC in the motor carrier industry and the association of those initiatives with improvement in financial performance. Knowledge of the efficacy and context-specific benefit of business initiatives is of significant interest to three communities: 1) the practitioner community that use, promote, instruct in the use of, or contemplate the implementation of ABC, TQM, or BPR (including accountants, managerial decision-makers, potential project leaders, professional associations, and consultants), 2) researchers interested in the theoretical and empirical literature regarding these initiatives, and 3) educators who communicate commonly-believed benefits of these initiatives and instruct in their use.

Archival financial information obtained for 305 motor carriers is used to regress one-year changes in financial performance against initiative use. The first finding is that, consistent with the literature and after control for previous level of performance and for use of other initiatives, use of TQM and BPR are significantly associated with ROA improvement. The second finding is that, although there was not a statistically significant simple effect obtained from use of ABC, there is empirical evidence that, consistent with management accounting, TQM, and BPR literatures, context-specific benefits are obtained from concurrent use of ABC with TQM and BPR. These results are robust to the partitioning the sample into high- and low-performing groups. It is likely that ABC functions as an enabler of other improvement initiatives, providing the information necessary to optimize the effectiveness of TQM and BPR. The positive findings regarding ABC are of particular interest to practicing and academic accountants because they are often the primary proponents and administrators of ABC, and previous evidence of ABC efficacy has been theoretical or anecdotal.

Due to the inconsistency of the foregoing results, more research is needed to explain how this effect occurs. It is possible that improvement in performance results more from the introspection and internal and external communication that occurs whenever the initiative is implemented rather than results achieved from its mechanical application. Research that investigates the conditions under which improvement occurs and that identifies the components of financial performance that are affected by initiative use would be of benefit.

There was no evidence of a context-specific benefit (positive or negative) from concurrent use of TQM and BPR. Therefore, the divergence of opinions in the literature regarding concurrent use of continuous and discontinuous improvement initiatives must continue.

A third significant finding of this study is that there is a pronounced mean reversion of earnings, at least in the motor carrier industry. Since deregulation in the 1970s, the industry has become highly competitive, largely because of 1) low entry costs in the TL and specialized carrier segments, and 2) increased competition with other modes of transport. Overall the industry lacks the capital investment requirements, proprietary processes, technology, and territory and patent protection typical of many other industries. Therefore, trucking firms are not able to maintain their competitive position over extended periods of time without continuing improvements in efficiency and service (Coyle et al., 1994). To maintain their position, the best performing firms must implement solutions to counter the "collar" effect that pulls their performance towards the mean. Although cause cannot be directly inferred from this study, there is evidence that the use of initiatives can help to offset this effect, thereby facilitating top performers in maintaining their relative position.

As with all studies, there are several important limitations to the analyses. It is assumed that respondents know the extent of initiative use and have responded honestly. Although respondents were generally top executives who should be knowledgeable about major initiatives, the possibility exists that the responses do not represent actual company practices. Secondly, although this study is restricted to a single industry, level-of-use may not capture the effectiveness of an individual firm's implementation of an initiative. As argued by previous researchers (e.g., Cooper, 1988; Cooper and Kaplan, 1991), firm-specific factors such as complexity, diversity, and information technology may limit or enhance this effectiveness. Further research testing the arguments of prior researchers would be of value.

Restriction to a single industry yields significant advantages in empirical testing. Although the motor carrier industry affects virtually all firms, there is no assurance that results are generalizable to firms in other industries. Research investigating other industries would complement the findings of this study.

Finally, significant interaction terms preclude interpretation of the individual coefficients of the initiative variables, and prevent the determination of the individual economic effect of TQM, BPR, and ABC. A study that utilizes a different methodology, but maintains control for concurrent use (possibly through control groups) would be welcome.

REFERENCES

Aksu, C., C. Eckstein, W.H. Greene, & J. Ronen (1996). Time-series properties, adjustment processes, and forecasting of financial ratios, Journal of Accounting, Auditing & Finance, 11(1), 1-44.

Altinkemer, K., A. Chaturvedi, & S. Knodareddy (1998). Business process reengineering and organizational performance: an exploration of issues, International Journal of Information Management, 18(6), 381-392.

Anderson, S. W. (1995). A framework for assessing cost management system changes: the case of activity-based costing implementation at General Motors, 1986-1993, Journal of Management Accounting Research, Fall, 1-51.

Anderson, S. W. & K. Sedatole (1998). Designing quality into products: the use of accounting data in new product development, Accounting Horizons, 12(3), September, 213-233.

Armitage, H. & G. Russell (1993). Activity-based management information: TQM's missing link, Cost and Management, 67(2), March, 7.

Atkinson, A.A., R.D. Banker, R.S. Kaplan, & S.M. Young (1995). Management Accounting. Englewood Cliffs, NJ, Prentice-Hall, Inc.

Balakrishnan, R., T.J. Linsmeier, & M. Venkatachalam (1996). Financial benefits from JIT adoption: effects of customer concentration and cost structure, The Accounting Review, April, 183-205.

Barnes, F.C. (1991). IES can improve management decisions using activity-based costing, Industrial Engineering, September, 44-50.

Beheiry, M.F. (1991). New thoughts on an old concept: the cost of quality, CMA Magazine, June, 24-25.

Belsley, D.A., E. Kuh, & R.E. Welsch (1980). Regression Diagnostics. New York: John Wiley & Sons.

Bernard, V.L. (1994). Accounting-based valuation methods, determinants of market-to-book ratios, and implications for financial statement analysis, Working Paper, Michigan Business School.

Biggart, T. (1997). The effects of just-in-time system adoption of firm performance. Unpublished doctoral dissertation, Florida State University.

Borthick, A.F. & H.P. Roth (1993). Accounting for time: reengineering business processes to improve responsiveness, Journal of Cost Management, Fall, 4-14.

Boyd, D. T. III (1996). An Analysis of Publicly Available Performance Measures of Companies Adopting JIT. Unpublished doctoral dissertation, Louisiana Tech University.

Brigham, E.F., L.C. Gapenski, & M.C. Ehrhardt (1999). Financial Management, 9th edition, New York: The Dryden Press.

Brimson, J.A. (1991). Activity Costing, An Activity-Based Costing Approach, New York: John Wiley & Sons, Inc.

Bruns, W. J. & R.S. Kaplan (1987). Accounting & Management Field Study Perspectives, Boston, MA: Harvard Business School Press.

Cagwin, D. & M.J. Bouwman (2002). The association between activity-based costing and improvement in financial performance, Management Accounting Research, 13, 1-39.

Capon, N., J.U. Farley & S. Hoenig (1988). Determinants of financial performance: A meta-analysis, Management Science, 36, 1143-1159.

Chang, R.Y. (1994). Improve processes, reengineer them, or both? Training and Development, 48(3), 54-58.

Chenhall, R.H. (1997). Reliance on manufacturing performance measures, total quality management and organizational performance, Management Accounting Research, 8,187-206.

Chenhall, R,H. & k. Langfield-Smith (1998). The relationship between strategic priorities, management techniques and management accounting: an empirical investigation using a systems approach, Accounting, Organizations and Society, 23(3), 243-264.

Choi, T.Y. & k. Eboch, K (1998). The TQM paradox: relations among TQM practices, plant performance, and customer satisfaction, Journal of Operations Management, 17, 59-75.

Cokins, G. (1996). Activity-Based Cost Management: Making It Work. Chicago, IL: Irwin Professional Publishing.

Cole, R. (1994). Reengineering the corporation: a review essay, Quality Management Journal, 1(4), 77-85.

Cooper, R. (1988). The rise of activity-based costing--part two: when do I need an activity-based system? Journal of Cost Management, Summer, 45-54.

Cooper, R. & R.S. Kaplan (1991). The Design of Cost Management Systems: Text, Cases, and Readings. Englewood Cliffs, NJ: Prentice Hall.

Cooper, R., R.S. Kaplan, L. Maisel & R. Oehm (1992). Implementing Activity-Based Management: Moving from Analysis to Action. Montvale, NJ: Institute of Management Accountants.

Cooper, R. & P.B. Turney (1989). Hewlett-Packard: The Roseville Network Division. Boston, MA: Harvard Business School (Case #9-189-117).

Coyle, J.J., E.J. Bardi & R.A. Novack (1994). Transportation, 4th edition. St. Paul, MN: West Publishing Company.

Crosby, P.B. (1979). Quality Is Free: The Art of Making Quality Certain. New York: McGraw-Hill.

Davenport, T..H. & J.E. Short (1990). The new industrial engineering: information technology and business process redesign, Sloan Management Review, 31(4),11-27.

Dean, E.B. (1996). Business process reengineering from the perspective of competitive advantage, http://garcia1.larc.nasa.gov/dfca/bpre.html (December).

Deming, W. E. (1988). Deming's Point Four: A Study, Quality Progress, December, 31-35.

Dillman, D.A. (1999). Mail and Telephone Surveys: The Total Design Method (Second Edition). New York: John Wiley & Son.

Dixon, J.M. (1996). Total Quality Management in ISO-9000 Registered Organizations: An Empirical Examination of the Critical Characteristics Associated with Levels of Financial Performance. Unpublished doctoral dissertation, Florida State University.

Drake, A.R., S.F. Haka & S.P. Ravenscroft (1999). Cost system and incentive structure effects on innovation, efficiency and profitability in teams, The Accounting Review, 74(3), July, 323-345.

Drucker, P.F. (1995). The information executives truly need, Harvard Business Review, January-February, 55-63.

Earl M. & B. Khan (1994). How new is business process redesign, European Management Journal, 12(1), 20-30.

Easton, G.S. & S.J. Jarrell (1998). The effects of total quality management on corporate performance: an empirical investigation, Journal of Business, 71(2), April, 253-307.

Engelkemeyer, S.W. (1991). An Empirical Study of Total Quality Management Practices in U.S. Manufacturing Firms. Unpublished doctoral dissertation, Clemson University.

Evans, H. & G. Ashworth (1995). Activity-based management: moving beyond adolescence, Management Accounting-London, December, 26-30.

Feigenbaum, A.V. (1986). Quality: the strategic business imperative, Quality Progress, February, 26-30.

Fliedner, G. & R. Vokurka (1997). Agility: the next competitive weapon, APICS--The Performance Advantage, January, 56-59.

Forza, C. & R. Filippini (1998). TQM impact on quality conformance and customer satisfaction: a causal model, International Journal of Production Economics, 55, 1-20.

Furey, T.R. (1993). A six-step guide to process reengineering, Planning Review, 21(2), 20-23.

Gadd, K. & J. Oakland (1996). Chimera or culture? Business process reengineering for total quality management, Quality Management Journal, 3(3), 20-38.

Golhar, D.Y. & C.L. Stamm (1991). The just-in-time philosophy: a literature review, International Journal of Production Research, 29(4), 657-676.

Goyal, S.K. & S.G. Deshmukh (1992). A critique of the literature on just-in-time manufacturing, International Journal of Operations and Production Management, January, 18-28.

Gupta, M., S. Baxendale & K. McNamara (1997). Integrating TOC and ABCM in a health care company, Journal of Cost Management, July/August, 23-28.

Hammer, M. (1990). Reengineering work. Don't automate, obliterate, Harvard Business Review, July/August, 104-112.

Hammer, M. (1991). Why we need both the continuous and discontinuous improvement, Perspectives on the Management of Information Technology, 8(1), 6-7.

Hammer, M. & J. Champy (1993). Reengineering the Corporation: A Manifesto for Business Revolution. New York: Harper Business.

Harmatuck, D.J. (1990). Motor carrier cost function comparisons, Transportation Journal, Summer, 31-46.

Harris, E. (1990). The impact of JIT production on product costing information systems, Production and Inventory Management Journal, 1st Qtr, 44-48.

Harrison, D.B. & M.D. Pratt (1992). A methodology for reengineering businesses, Planning Review, 21(2), 6-11.

Healy, P.M. (1985). The effect of bonus schemes on accounting decisions, Journal of Accounting and Economics, April, 85-107.

Hendricks, K.B. & V.R. Singhal (1997). Does implementing an effective TQM program actually improve operating performance? Empirical evidence from firms that have won quality awards, Management Science, 43(9), September, 1258-274.

Hendricks, K.B. & V.R. Singhal (1999). Don't count TQM out, Quality Progress, 32(4), April, 35-42.

Huson, M. & D. Nanda (1995). The impact of just-In-time manufacturing on firm performance, Journal of Operations Management, January, 5-14.

Innes, J., F. Mitchell & D. Sinclair (2000). Activity-based costing in the UK's largest companies: a comparison of 1994 and 1999 survey results, Management Accounting Research, 11, 349-362.

Ittner, C.D. & D.F. Larcker (1995). Total quality management and the choice of information and reward systems, Journal of Accounting Research, Supplement, 1-34.

Ittner, C.D. & D.F. Larcker. (1999). Quality strategy, strategic control systems, and organizational performance, Accounting, Organizations and Society, 22(3/4), 293-314.

Jacobson, R. (1987). The validity of ROA as a measure of business performance, The American Economic Review, 77, 470-478.

Jarrar, Y.F. & E.M. Aspinwall (1999a). Business process reengineering: learning from organizational experience, Total Quality Management, 10(2),March, 173-186.

Jarrar, Y.F. & E.M. Aspinwall (1999b). Integrating total quality management and business process reengineering: is it enough? Total Quality Management, 10(4/5), July, S584-S593.

Jaworski, B. & A.K. Kohli (1993). Market orientation: antecedents and consequences, Journal of Marketing, 57, 53-70.

Juran, J.M. & F.M. Gryna Jr. (1970). Quality, Planning and Analysis, New York, McGraw-Hill.

Kaynak, H. (1996). The Relationship between Just-in-Time Purchasing and Total Quality Management and Their Effects on the Performance of Firms Operating in the U.S.: An Empirical Investigation, Unpublished doctoral dissertation, University of North Texas.

Kennedy, T. & J. Affleck-Graves (2001). The impact of activity-based costing techniques on firm performance, Journal of Management Accounting Research, 13, 19-45.

Kinney, M.R. & W.F. Wempe (1998). Further evidence on the extent and origins of JIT's profitability effects, Working Paper, Texas A&M University.

Klein, M. M. 1993. The fill facilitator role in benchmarking operations to improve performance, Industrial Engineering, 25(9), 40-42.

Krumwiede K.R. (1998). ABC: why it's tried and how it succeeds, Management Accounting, April, 38-38.

Kwon, T.H. & R.W. Zmud (1987). Unifying the fragmented models of information systems and implementation, in Critical Issues in Information Systems Research, edited by R. J. Boland and R. Hirscheim. New York: John Wiley & Sons.

McConville, D.J. (1993). Start with ABC, Industry Week, September, 6, 33-36.

O'Neill, P. & A.S. Sohal (1999). Business process reengineering: a review of the current literature, Technovation, 19, 571-581.

Player, R.S. & D.E. Keys (1995). Lessons from the ABM battlefield: moving from pilot to mainstream, Journal of Cost Management, Fall, 31-41.

Prescott, J.E., A.K. Kohli & N. Venkatraman, N. (1986). The market share-profitability relationship: an empirical assessment of major assertions and contradictions, Strategic Management Journal, 7, 377-394.

Rangone, A. (1997). Linking organizational effectiveness, key success factors and performance measures: An analytical framework, Management Accounting Research, 8, 207-219.

Raymond, L., F. Bergeron & S. Rivard (1998). Determinants of business process reengineering success in small and large enterprises: an empirical study in the Canadian context, Journal of Small Business Management, 36, January, 72-85.

Redman, T. & J. Grieves (1999). Managing strategic change through TQM: learning from failure, New Technology, Work and Employment, 14(1), 45-61.

Roberts, M.W. & K.J. Silvester (1996). Why ABC failed and how it may yet succeed, Journal of Cost Management, Winter, 23-35.

Romney, M. (1995). Business process reengineering, Internal Auditor, June, 24-29.

Rotch, W. (1990). Activity-based costing in service industries, Journal of Cost Management, Summer, 4-14.

Samson, D. & M. Terzioski (1999). The relationship between total quality management practices and operational performance, Journal of Operations Management, 17, 393-409.

Sankar, M. (1995). Self-Assessment Using the Baldrige Criteria, Milwaukee, WI, ASQC Professional Technical Development.

Shank, J.K. & V. Govindarajan (1994). Measuring the "cost of quality": a strategic cost management perspective, Journal of Cost Management, Summer, 5-17.

Shapiro, A.C. & S.D. Balbirer (1999). Modern Corporate Finance. Upper Saddle River, NJ: Prentice Hall.

Shaw, R. (1998). ABC and ERP: partners at last? Management Accounting, November, 56-58.

Shepard, N. (1995). The bridge to continuous improvement, CMA Magazine, March, 29-32.

Shields, M.D. (1997). Research in management accounting by North Americans in the 1980s, Journal of Management Accounting Research, 9, 4-61.

Shields, M.D., F.J. Deng & Y. Kato (2000). The design and effects of control systems: tests of direct- and indirect- effect models, Accounting Organizations & Society, 25(2), 185-202.

Sim, K.L. & L.N. Killough (1998). The performance effects of complementarities between manufacturing practices and management accounting systems, Journal of Management Accounting Systems, 10, 325-346.

Smith, M. (1998). Innovation and the great ABM trade-off, Management Accounting-London, 76(1), Jan, 24-26.

Snell, S.A. & J.W. Dean, Jr. (1992). Integrated manufacturing and human resource management, Academy of Management Journal, 35(3), August, 467-504.

Steimer, T.E. (1990). Activity-based accounting for total quality, Management Accounting, October, 39-42.

Swenson, D. (1995). The benefits of activity-based cost management to the manufacturing industry, Journal of Management Accounting Research, Fall, 167-180.

Swenson, D. (1998). Managing costs through complexity reduction at Carrier Corporation, Management Accounting, April, 20-28.

Tanju, D.W. & M. Helmi (1991). ABCs for internal auditors, Internal Auditor, December, 33-37.

Taylor, S. (1993). Eastman Chemical strives for better than world class, Industrial Engineering, November, 28-34.

Transportation Technical Services, Inc. (TTS), (1999). TTS Blue Book of Trucking Companies 1998-1999. New York: Transportation Technical Services, Inc.

Turney, P.B.B. (1991b). How activity-based costing helps reduce cost, Journal of Cost Management, Winter, 29-35.

Watts, R. & J. Zimmerman, J. (1986). Positive Accounting Theory. Englewood Cliffs, NJ: Prentice-Hall, Inc.

White, H. (1980). A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity, Econometrica, 48, 817-838.

Wouters, M., K. Kokke, J. Theeuwes & K. van Donselaar (1999). Identification of critical operational performance measures--a research note on a benchmarking study in the transportation and distribution sector, Management Accounting Research, 10, 439-452.

Wright, B. (1995). Leveraging TQM through TQM, The TQM Magazine, 7, 50-53.

Youndt, M.A., S.A. Snell, J.W. Dean, Jr. & D.P. Lepak (1996). Human resource management, manufacturing strategy, and firm performance, Academy of Management Journal, 39(4), August, 836-867.

Zairi, M. & D. Sinclair (1995). Business process reengineering and process management: A survey of current practice and future trends in integrated management, Business Process Reengineering and Management Journal, 1(1), 8-30.

Zmijewski, M.E. & R.L. Hagerman (1981). An income strategy approach to the positive theory of accounting setting/choice, Journal of Accounting and Economics, March, 129-149.

Douglass Cagwin, Zayed University

Katherine J. Barker, SUNY Fredonia

Table 1: Summary of Sample

Initial Population 2,002
Less: Firms with Less than Thirty Employees or
 $5 million in Revenues 383
Population of Interest 1,619
Random Selection 1,100
Less: Canadian Companies 6
 Undeliverable 2
 Out of Business 9
 Withdrew or Refused to Cooperate Upon
 Initial Contact 14 31
Net Responses Possible 1,069
Responses Received 332
Response Rate 31.1%
Less: Data from 1999 Unavailable 27
Final Sample 305

Table 2: Descriptive Statistics

PANEL A
Characteristics of Responding Firms Use of Initiatives
(n=305)

 # Responses

 Nonusers Users

 # % # %

Initiative
Activity-Based Costing (ABC) 211 69.2 94 30.8
Total Quality Management (TQM) 208 68.2 97 31.8
Business Process Reengineering (BPR) 265 86.9 40 13.1
Interactions
TQM*BPR 28 9.2
TQM*ABC 49 16.0
BPR*ABC 21 6.9

Use of ABC, TQM, and BPR = a response of 6 or 7 with implementation
date completed and before 2001.

PANEL B
Sample Partitioned into Low and High Performing Firms
Based on LEVEL of Prior ROA

 # Nonusers # Users

 LEVEL LEVEL

 Low High Low High

Initiative
 Activity-Based Costing (ABC) 102 109 51 43
 Total Quality Management (TQM) 101 106 52 46
 Business Process Reengineering (BPR) 129 136 24 16
Interactions
 TQM*BPR 16 12
 TQM*ABC 26 23
 BPR*ABC 13 8

Firms are partitioned into High and Low performance at the median.

Table 3: Descriptive Statistics

Panel A: Characteristics of Tested Firms

 Mean Median Std. Dev.

Financial Performance
 1998 ROA 0.040 0.036 0.096
 1998 Net Income (000s) 487.7 419.0 3210.0
Type (%)
 TL 48.360 35.0 44.207
 LTL 15.016 0.0 31.499
 Specialized 36.524 0.0 35.907
Size (000s)
 Revenue 53,171 29,087 107,992
 Assets 27,627 12,415 86,486

Panel B: Dependent and Control Variables

 Mean Median Std. Dev.

Performance
 DROA 0.001 (0.006) 0.101
 % DINC (% Change in Income) 0.087 0.046 0.088
 Level (ROA), (t) before mean
 adjustment 0.039 0.037 0.098
Type (%)
 Truckload (TL) 48.360 35.0 44.207
 Less-than-Truckload (LTL) 15.016 0.0 31.499

Table 4: Spearman Correlation Matrix of the Independent Variables
(N = 305)

 ABC TQM BPR

Activity-Based Costing (ABC) 1
Total Quality Management (TQM) .29 1
Business Process Reengineering (BPR) .19 .32 1
Truckload % (TL) -.09 -15 -.11
Less-than-Truckload % (LTL) .20 .13 .15
LEVEL of year t ROA -.12 -.09 -.22

 TL LTL LEVEL

Activity-Based Costing (ABC)
Total Quality Management (TQM)
Business Process Reengineering (BPR)
Truckload % (TL) 1
Less-than-Truckload % (LTL) -.12 1
LEVEL of year t ROA .01 -.02 1

Use of ABC, TQM, and BPR = a response of 6 or 7 with implementation
date completed and before 1999; LTL and LTL = the percentage of
truckload and less-than-truckload carriage; and LEVEL = the
industry-adjusted level of the prior year's ROA.

Table 5
Regression of 1-Year Change in ROA on Initiatives including
Interactions of TQM with BPR and ABC with TQM and BPR

[DELTA]ROA = [alpha] + [[beta].sub.1]TQM + [[beta].sub.2]BPR +
[[beta].sub.3]ABC + [[beta].sub.4]TL + [[beta].sub.5]LTL +
[[beta].sub.6]LEVEL + [[beta].sub.7]TQM*BPR + [[beta].sub.8]TQM*ABC +
[[beta].sub.9]BPR*ABC + [epsilon]

F 22.90
P-Value 0.001
[R.sup.2] .5360
Adjusted [R.sup.2] .5212

 Expected
 Sign Coefficient [tau]-Stat p-value

Intercept -0.007 2.229 0.026
Initiative Simple
 Effect
 Activity-Based
 Costing (ABC) + 0.006 1.007 0.157
 Total Quality
 Management (TQM) + 0.008 1.658 0.049#
 Business Process
 Reengineering
 (BPR) + 0.015 2.167 0.015#

Interactions
 TQM*BPR ? -0.007 -0.574 0.567
 TQM*ABC + 0.013 2.055 0.020#
 BPR*ABC + 0.007 1.567 0.059^

Control Variables
 Truckload (TL) ? 0.000 -0.744 0.458
 Less-than-Truckload
 (LTL) - 0.000 -1.436 0.098^
 LEVEL - -0.116 4.644 0.001#

Where use of ABC, TQM, and BPR = a response of 6 or 7 with
implementation date completed and before 1999; LTL and LTL = the
percentage of truckload and less-than-truckload carriage; and LEVEL =
the industry-adjusted level of the prior year's ROA.

Bold = significant at the [alpha] = 0.05 level; Italicized =
significant at the [alpha] =0.10 level

Tests on the coefficients are one-tailed for variables with an
expected sign; two tailed for remaining variables.

Note: Bold = significant at the [alpha] = 0.05 level indicated with #;
Italicized = significant at the [alpha] =0.10 level indicated with ^.