An examination of the relationships between JIT and financial performance

Journal of Operations Management 21 (2003) 383–404

An examination of the relationships between JIT and financial performance Rosemary R. Fullerton a,∗ , Cheryl S. McWatters b , Chris Fawson c a

b c

School of Accountancy, Utah State University, Logan, UT 84322-3540, USA Faculty of Management, McGill University, Montreal, Que., Canada H3A 1G5 Department of Economics, Utah State University, Logan, UT 84322-9500, USA Received 20 September 2001; accepted 21 November 2002

Abstract Despite abundant information explaining the expected benefits from successful just-in-time (JIT) implementation, only tenuous validation of the linkage between financial performance and JIT exists. Managers act rationally in implementing JIT if they are convinced that JIT enhances firm performance. From both a cross-sectional and longitudinal perspective, this survey study of 253 US manufacturing firms finds significant statistical relationships between measures of profitability and the degree of specific JIT practices used. The evidence provides empirical support to the premise that firms that implement and maintain JIT manufacturing systems will reap sustainable rewards as measured by improved financial performance. © 2003 Elsevier Science B.V. All rights reserved. Keywords: Just-in-time/kanban; Empirical research methods; Accounting/operations interface

1. Introduction Over the last two decades, just-in-time (JIT) and other world-class manufacturing (WCM) practices have been scrutinized and championed around the globe, as firms seek to attain and sustain competitive advantage. The economic benefits of these techniques must be real and long lasting to warrant their application, given the costs and challenges in their implementation. Lower production costs, higher and faster throughput, better product quality, and on-time delivery of finished goods are benefits from successful implementation of a JIT system that are documented in the literature (e.g. Goyal and Deshmukh, 1992; Nakamura et al., 1998; Norris, 1992; Orth et al., ∗ Corresponding author. Tel.: +1-435-797-2332; fax: +1-435-797-1475. E-mail address: [email protected] (R.R. Fullerton).

1990). Increased profitability is often assumed as an outcome of JIT, yet Johnson and Kaplan (1989) note a frequent disparity between improved operations and financial performance. Empirical studies that examine the direct relationship between JIT implementation and financial performance have reported mixed results (Balakrishnan et al., 1996; Callen et al., 2000; Huson and Nanda, 1995; Inman and Mehra, 1993; Kinney and Wempe, 2002). Moreover, Japanese transplant manufacturing firms that employ JIT methods consistently have shown lower profitability in the short term than their counterpart domestic US firms (Nakamura et al., 1998). Cooper (1995) argues that this difference results from the Japanese preference for stability, longterm reliability, and growth. Supporting this view, Johnson and Bröms (2000) demonstrate how Toyota’s stable performance relates to its manufacturing strategies that foster growth and stability over the long run,

0272-6963/03/$ – see front matter © 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0272-6963(03)00002-0

384

R.R. Fullerton et al. / Journal of Operations Management 21 (2003) 383–404

as opposed to the achievement of short-run financial targets. Investment returns from JIT adoption are not immediately observable, due to the long-run nature of its implementation process. This phenomenon may provide a partial explanation of limited empirical validation for a direct association between financial performance and JIT adoption. The purpose of this study is to evaluate empirically whether the degree to which a firm implements a combination of JIT practices systematically affects that firm’s financial performance. The focus on the degree of JIT implementation underscores the notion of complementarity, which “suggests that a firm adapting to environmental change will be most likely to find profitable new activities in areas that are complementary to the newly increased activities” (Milgrom and Roberts, 1995, p. 186). In response to a changing environment, a firm’s coherent set of modifications to strategy, structure, and process is “plausibly associated with increasing income levels” (Milgrom and Roberts, 1995, p. 192). The empirical evaluation is done using a twopronged approach. First, data from 253 US manufacturing firms are examined to determine the static relationship between firm profitability and the degree of JIT implementation in terms of various JIT practices. These practices represent a measurable set of JIT elements suggested in prior research as indicative of JIT. Second, the sample is stratified to focus on the subset of firms that self-identify as JIT adopters and that have a sufficient pre-JIT and post-JIT implementation history to evaluate the time-dependent effects of JIT adoption levels on firm profitability. This longitudinal perspective takes into consideration the traditional annual performance measures that may be askew in the first year or two after implementing a major business-level strategic change such as JIT. This study contributes to the literature in four areas. First, the key contribution is its additional insight into the uncertain relationship between a firm’s financial performance and its adoption of a comprehensive JIT system. The study augments the growing body of empirical evidence by its careful documentation of the various JIT-implementation schemes used by firms and their link to financial performance. This contribution is derived from both a static cross-sectional comparison of firms that have adopted different degrees of world-class JIT manufacturing practices, and a longi-

tudinal setting that allows time for conventional profitability indicators to reflect more fully adjustments by firms that formally have implemented a JIT strategy. Second, the current study further resolves the differing results of Balakrishnan et al. (1996) and Kinney and Wempe (2002). Both of these studies classified their sample firms as either JIT or non-JIT and focused on the impact of JIT adoption on return on assets (ROA). The current study presents additional evidence of JIT’s positive influence on ROA. In contrast to these prior studies, this research examines the degree of JIT implementation by capturing the extent to which sample firms have adopted a combination of JIT elements. These data allow for a more comprehensive assessment of JIT implementation and its effect on financial performance. Third, the study uses publicly reported financial data to test the association between financial performance and the degree of JIT implementation. This approach extends the work of Inman and Mehra (1993), which relied upon survey respondents’ self-evaluation of financial success. Finally, firm-specific responses are collated with their publicly available financial information from the COMPUSTAT database. Balakrishnan et al. (1996) and Kinney and Wempe (2002) relied solely on publicly available data. Callen et al. (2000), which classified its sample as either JIT or non-JIT adopters, is the only other known study to combine both public and private data to assess the financial benefits achieved from implementing JIT. 2. JIT adoption and financial performance 2.1. Definition and benefits of JIT Manufacturing capabilities can be used as a strategic, competitive weapon (Hayes and Wheelwright, 1984). Voss (1995) discussed three major manufacturing strategic paradigms, one of which is “best practices.” This paradigm is supported by the concept of WCM. Embodied within WCM is the JIT manufacturing philosophy, which emphasizes excellence through the continuous elimination of waste and improvement in productivity. Much more than the narrow notion of reduced inventory and optimal batch size (Blackburn, 1991; White and Prybutok, 2001; Yasin et al., 1997), JIT is the genesis of time-based

R.R. Fullerton et al. / Journal of Operations Management 21 (2003) 383–404

competition that “provides manufacturing with flexibility and speed essential to meet global competition” (Blackburn, 1991). For JIT to be most beneficial, it must be accepted as an “organizational philosophy” (Yasin et al., 1997, p. 462), and be aligned with a firm’s key success factors. The lack of a universal definition of JIT reflects some remaining confusion over what exactly it comprises (Mia, 2000). This inability to explain systematically and theoretically JIT manufacturing methods may be due to JIT’s emphasis on practice and implementation (Monden, 1998, p. 458). Descriptions in the literature generally include a broad-based production system that incorporates the manufacturing practices of “efficient material flow, improved quality, and increased employment involvement” (White and Prybutok, 2001, p. 113). Mehra and Inman (1992, p. 172) proposed that JIT was both a vendor strategy and a production strategy “ . . . that strives to achieve excellence in manufacturing by reducing setup times . . . through the use of group technology, cross-training of employees, and sound preventive maintenance.” Most published research over the past two decades has been field studies or anecdotal evidence gleaned from surveys with small samples that attempt to validate empirically the benefits of JIT adoption. The most consistent benefit from JIT adoption found in the empirical studies is a reduction in inventory levels and/or an increase in inventory turns (Balakrishnan et al., 1996; Billesbach, 1991; Billesbach and Hayen, 1994; Crawford and Cox, 1990; Droge and Germain, 1998; Fullerton and McWatters, 2001; Gilbert, 1990; Huson and Nanda, 1995; Im and Lee, 1989; Norris et al., 1994). Other evidence purports improvements in productivity, customer response time, and product quality, along with decreases in scrap and rework, production costs, lead times, setup times, and space requirements. Monden (1998, p. 13) stated that the main purpose for the continuous improvement efforts of JIT was to “increase profits by reducing costs through completely eliminating waste.” Thus, implementing the lean principles of JIT is an increasingly important competitive tool (Conner, 2001). 2.2. JIT and profitability “The result by which any business in a market economy must be measured is the ability to make

385

enough profit to renew itself” (Womack and Jones, 1996, p. 121). The general assumption in the literature has been that JIT implementation translates into higher profitability. JIT is expected to improve firm performance and competitiveness through an even production flow of small lot sizes integrating schedule stability, product quality, short setup times, preventive maintenance, and efficient process layout (Chapman and Carter, 1990; Foster and Horngren, 1987; Hall and Jackson, 1992). Moreover, these production improvements are assumed to bring both indirect and direct financial savings (Anderson et al., 1989; Kaplan and Atkinson, 1989). In theory, JIT improves profitability due to its impact on the two interdependent components of ROA: asset turnover, which measures sales relative to investment; and return on sales (ROS), which measures income relative to sales (Kinney and Wempe, 2002). JIT is expected to improve ROA in a number of ways. First, asset turnover should increase, as JIT frees up assets and capital. A smaller asset base increases ROA. Second, lower inventory levels reduce the asset base, improving asset turnover in the short term. Third, fewer buffer inventories motivate the elimination of non-value-added activities (such as dealing with defects and stock-outs) that have a negative impact on the profit margin (Alles et al., 1995). As emphasized by Balakrishnan et al. (1996, pp. 185–186), these effects are not necessarily automatic, and can be offsetting, especially in the short term. For example, firms may be required to invest in additional training and capital expenditures to accommodate a JIT environment. Training costs initially reduce the profit margin, but are expected to improve long-run productivity. Capital expenditures increase the asset base and depreciation expense in the short run, thus affecting both components of ROA. Despite the expected impacts, few studies have adequately tested the presumed relationship between JIT and improved financial performance. In the first survey study to directly link JIT adoption to enhanced firm profitability, Inman and Mehra (1993) found a significant correlation between self-reported improvements in profitability and the adoption of JIT practices. Balakrishnan et al. (1996) compared the profitability of JIT and non-JIT firms. When using the broad categorization of JIT and non-JIT firms to make profitability comparisons, their study found no differences in ROA. However, in sub-sample stratifications where

386

R.R. Fullerton et al. / Journal of Operations Management 21 (2003) 383–404

firms were characterized by high or low customer concentrations and different cost structures, JIT firms that had low customer concentrations exhibited significantly higher ROA than non-JIT firms. In an extension of Balakrishnan et al., Kinney and Wempe (2002) compared the profitability of JIT to non-JIT firms in a similar matched-pairs research design. In contrast, their results showed that in the three post-JIT adoption years tested, ROA fell significantly less for the JIT firms than for the non-JIT firms. They concluded that JIT positively affected adopters’ profitability, and that their conflicting evidence perhaps resulted from a larger sample, whose sample firms were also proportionately larger in size. Both studies relied upon publicly-available sources for JIT classification and adoption year. Callen et al. (2000) surveyed 100 Canadian plants from the auto parts and electronic industries, classifying the plants as either JIT or non-JIT. Their cross-sectional, multivariate analyses showed that JIT implementation led to higher profit and contribution margins and lower variable costs. In their 1995 longitudinal study of 55 JIT firms, Huson and Nanda reported that JIT implementation increased firm performance. Examining a period of 4 years before and after JIT adoption, they found that despite a decrease in post-JIT adoption earnings per share and gross profit margins, the decrease appeared to be less than if the firms had not adopted JIT. 2.3. Research hypotheses Milgrom and Roberts (1995, p. 191) demonstrated how modern manufacturing features similar to those measured in this study are “mutually complementary” and “the move towards adopting them is a profit-maximizing response” for firms. They further pointed out that a firm’s strategy and structure must move in a coherent pattern to meet the needs of a changing environment. If JIT is not implemented as part of a systematic and comprehensive transformation of production and operation procedures, and only a few of the new, advanced manufacturing elements reach optimal levels, the full benefits of change might be diminished. Negative results also might ensue. Many organizations have chosen not to embrace JIT processes as part of their management and production strategies. While the JIT business model is not

appropriate in all firms, its non-adoption by others also suggests that the costs of JIT may outweigh the perceived realizable benefits. Even if management considers these advanced manufacturing practices to be cost beneficial, the organizational-change process to adapt to new environmental conditions is “slow, painful, and uncertain” (Milgrom and Roberts, 1995, p. 180). The reported results from prior research are inconclusive, but suggest cautious optimism about the relationship between JIT implementation and enhanced financial performance. With careful attention to data construct and model structure, this study provides evidence related to the tenuous relationship of JIT implementation and financial performance through two testable hypotheses. Hypothesis H1 is evaluated using a cross-sectional comparison of firms that have identified themselves as being engaged in various stages of JIT process implementation, from no intention to implement certain JIT practices to a full JIT adoption. H1 . Firms that have implemented a higher degree of JIT practices will experience higher profitability. For the sample respondents that identified themselves as formally adopting JIT, it is assumed that integration of JIT methods into a company’s operations is a gradual, long-term and time-consuming process. It is also assumed that JIT firms adopting a more comprehensive set of JIT practices will experience better performance over time. As such, hypothesis H2 is tested using a pooled, cross-sectional regression model that acknowledges the time-dependent adjustment process inherent in the adoption of JIT management practices. H2 . JIT firms that have implemented a higher degree of JIT practices will experience higher profitability over time than JIT firms that have implemented a lower degree of JIT practices.

3. Research method 3.1. Survey and sample design To test the research hypotheses, a five-page survey instrument was designed to collect specific information about the manufacturing operations, product

R.R. Fullerton et al. / Journal of Operations Management 21 (2003) 383–404

costing methods, information and incentive systems, JIT practices employed, perceived results from JIT implementation, and characteristics of the respondent firms. The majority of questions were either categorical or interval Likert scales. To evaluate the survey instrument for readability, completeness, and clarity, a limited pretest was conducted by soliciting feedback from several business professors and managers of five manufacturing firms who were familiar with JIT practices. Appropriate changes were made to reflect their comments and suggestions. The sample of firms used to collect data was identified through a two-stage process. First, to ascertain firms that were practicing JIT, an extensive search of industry literature identified a sample of 169 US manufacturing firms that had implemented JIT. This sample was combined with a random sample of 600 firms obtained from the Compaq Disclosure database that included both JIT and non-JIT manufacturing firms using various degrees of JIT practices. All sample firms were required to pass the following screen: a primary two-digit SIC code within the manufacturing ranges of 20–39, sales between US$ 2 million and US$ 2 billion, significant manufacturing operations in the US, and inclusion in the COMPUSTAT database. After accounting for duplications arising from firms that were common to both the JIT-specific sample and the random sample and eliminating firms that did not meet the preliminary screening criteria, 447 firms were included in the final sample.

387

During the first 4 months of 1997, initial telephone contacts explaining the purpose of the research were made with the sample firms. Receptionists were asked either to provide a name or to direct the call to the appropriate executive. The sample firm executives were contacted personally, then faxed or mailed the pre-tested survey instrument. To improve the response rate, each sample firm was contacted a maximum of three times over a period of 6 weeks. Out of the 447 firms surveyed, 254 returned completed survey instruments, for an overall response rate of 56.8%. One of the returned surveys was unusable. The differences in the means for responding (US$ 404 million) and non-responding (US$ 380 million) firm sales were not statistically significant. In addition, an ANOVA comparing the means of the industry SIC codes (represented as dummy variables) for the non-responding and responding firms showed no statistical differences. Thus, a response bias related to either firm size or industry is not evident. Respondents had titles equivalent to the Vice President of Operations, the Director of Manufacturing, or the Plant Manager. The respondents averaged 17 years of management experience, including 9 years in management with their current firm. Data from 253 valid survey responses were collated with firm-level financial data obtained from the COMPUSTAT database. A distribution of survey respondents by SIC classification is presented in Table 1. Firms that responded to the survey are represented predominantly (72%) by four

Table 1 Distribution of two-digit SIC codes for sample firms Industry

JIT firms frequency

20: 22: 25: 26: 27: 28: 30: 33: 34: 35: 36: 37: 38: 39:

1 2 5 1 1 4 3 3 7 17 24 6 20 1

6 3 1 1 0 20 2 12 7 24 37 5 35 5

7 5 6 2 1 24 5 15 14 41 61 11 55 6

2.8 2.0 2.4 0.8 0.4 9.5 2.0 5.9 5.5 16.2 24.1 4.3 21.7 2.3

95

158

253

100.0

food textiles furniture & fixtures paper & allied products printing/publishing chemicals & allied products rubber products primary metals fabricated metals industrial machinery electronics motor vehicles & accessories instrumentation other manufacturing

Total

Non-JIT firms frequency

Sample frequency

Sample percent

388

R.R. Fullerton et al. / Journal of Operations Management 21 (2003) 383–404

industries: chemicals and allied products (SIC-28), industrial machinery (SIC-35), electronics (SIC-36), and instrumentation (SIC-38). This distribution was similar to the total sample distribution. Ninety-five out of the 253 firms in the cross-sectional sample identified themselves as JIT firms and supplied their year of JIT adoption. To be included in the JIT sample used for the pooled data analysis required to evaluate hypothesis H2 , a firm had to first self-identify itself as having formally implemented JIT, and then have financial (COMPUSTAT) data available for 2 years prior to the JIT adoption year and 3 years post-adoption. This time frame is consistent with the prior research of Balakrishnan et al. (1996) and Kinney and Wempe (2002). Their work suggested that the majority of any direct effects from JIT implementation would be captured in the 3 years following a firm’s adoption year. Therefore, only firms that had implemented JIT before 1995 could be included in this sample, reducing the number of firms in the pooled analysis to 54, for a total of 324 observations (54 firms × 6 years). 3.2. Measures of firm profitability Three variants of profitability measures are obtained from the COMPUSTAT database and used as the dependent variable for hypotheses testing: return on assets, return on sales, and cash flow margin (CFL), which is measured as the ratio of income net of extraordinary items, depreciation, and amortization, to sales. 3.3. Measures of the degree of JIT implementation In order to determine if JIT processes affect firm profitability, the JIT independent variables are constructed to measure the extent to which a firm adopted specific JIT practices. The prior literature suggests that unsatisfactory results from JIT are associated with incomplete and ineffective implementations (Clode, 1993; Daniel and Reitsperger, 1991; Gilbert, 1990; Goyal and Deshmukh, 1992; Milgrom and Roberts, 1995). “The potential synergic benefits are not fully realized until all elements of a JIT system are integrated” (White and Prybutok, 2001, p. 114). Often companies identify themselves as “JIT firms” without a full un-

derstanding of what constitutes JIT practice, and with only minimal efforts at changing their operations (see Ransom, 2001). One objective of this study is to determine what specific JIT elements are related to financial performance, rather than to examine JIT implementation as an either/or proposition. Thus, it was necessary to delineate a set of measurable manufacturing practices that reflect JIT. Although a universal set of JIT elements remains to be specified within the research literature (Davy et al., 1992; White and Ruch, 1990), different practices deemed important to successful JIT adoption are suggested in several studies (Koufteros et al., 1998; Mehra and Inman, 1992; Moshavi, 1990; Sakakibara et al., 1993; Spencer and Guide, 1995; Yasin et al., 1997). White and Ruch (1990) found 10 consensus JIT elements identified in the work of established JIT authors (e.g. Hall, Hay, Monden, Schonberger, Shingo, and Suzaki). Described in previous research as encompassing JIT practices, these consensus elements are used by White (1993), White et al. (1999), White and Prybutok (2001), and Fullerton and McWatters (2001, 2002) as JIT indicators. Although not all inclusive, these 10 practices are considered representative of a comprehensive JIT implementation for the purposes of this study. They are designated as follows: focused factory, group technology, reduced setup times, total productive maintenance, multi-function employees, uniform workload, kanban, JIT purchasing, total quality control, and quality circles. Definitions for each of these practices, which were supplied to survey respondents, along with post-1990 research literature that supports their inclusion as part of a JIT manufacturing environment are presented in Table 2. As explained in the survey instrument, the term focused factory is consistent with the APICS definition (1995) of a factory that focuses on simplifying its organizational structure and operations to coincide with the firm’s competitive strategy. It is not restricted to Skinner’s (1974) more narrow definition of a plant that focuses explicitly on one task. The term uniform workload refers to balancing the daily flow of work through manufacturing operations. Using the above-noted JIT indicators, 11 survey questions asked respondents to identify their firm’s degree of JIT implementation on the basis of a six-point Likert scale, ranging from “no intention” of implementing the identified JIT practice to its being

R.R. Fullerton et al. / Journal of Operations Management 21 (2003) 383–404

389

Table 2 JIT practices defined with literature support JIT element

Definition

Referencesa

Focused factory

A production strategy that is based on corporate strategy. It centers around simplifying the organizational structure, reducing the numbers of products or processes, and minimizing the complexities of physical constraints.

2–5, 10, 13–16

Group technology

Collecting and organizing common concepts, principles, problems, and tasks. It avoids unnecessary duplication through standardization. It includes sequencing similar parts through the same machine and creating manufacturing cells for processing.

1–8, 10–18

Reduced setup times

Reduction of the time and costs involved in changing tooling and other aspects required in moving from producing one product to another. This reduces lot sizes and the need for buffer inventories.

1–5, 7–18

Total productive maintenance

Rigorous, regularly scheduled preventive maintenance and machine replacement programs. Operators are actively responsible for the maintenance of their machines.

1–7, 9, 11–16, 18

Multi-function employees

Extended training of employees on several different machines and in various tasks.

2–4, 7, 8, 10, 12–16

Uniform workload

Reduction of the fluctuations of the daily workload through line balancing, level schedules, stable cycle rates, and market-paced final assembly rates.

1–5, 8–10, 13–17

Kanban

A card or information system that is used to “pull” the necessary parts into each operation as they are needed.

1–6, 8–11, 13–17

JIT purchasing

A supplier participation and partnership program. Receiving just the right parts just when they are needed. Suppliers, lot sizes, and paperwork are reduced.

2, 3, 5, 6, 9, 10, 12–15, 17

Total quality control

Quality is established as the top priority of the production system. Involvement in quality effort required by all aspects of the organization. Implementation of statistical quality control methods is an integral part of establishing both process and product quality.

1–4, 6, 8, 12–17

Quality circles

Small groups are formed from employees doing similar tasks. The groups are created to encourage employee participation in problem solving and decision making.

2, 7, 8, 13–16

a

The following references are examples of literature that identify the specified element as part of a JIT production system: 1. Conner, 2001; 2. Davy et al., 1992; 3. Fullerton and McWatters, 2001; 4. Fullerton and McWatters, 2002; 5. Jusko, 1999; 6. Koufteros et al., 1998; 7. Mehra and Inman, 1992; 8. Monden, 1998; 9. Moshavi, 1990; 10. Sahin, 2000; 11. Sakakibara et al., 1993; 12. Spencer and Guide, 1995; 13. White and Ruch, 1990; 14. White, 1993; 15. White et al., 1999; 16. White and Prybutok, 2001; 17. Womack and Jones, 1996; 18. Yasin et al., 1997.

“fully implemented”. The questionnaire had two questions representing total quality control: one for process quality and one for product quality. In order to investigate the possibility of a more parsimonious representation for JIT adoption in the empirical model, a principal-components-based exploratory factor analysis was conducted across the 11 JIT categories to identify common structural characteristics. Three

components of JIT implementation, with eigenvalues greater than 1.0, were extracted from the orthogonally rotated factor matrix, representing 63% of the total variance in the data (refer to Table 3 for results of the factor analysis). All of the 11 elements loaded greater than 0.50 onto one of the three constructs, except for quality circles, which was eliminated from further testing.

390

R.R. Fullerton et al. / Journal of Operations Management 21 (2003) 383–404

Table 3 Factor analysis (VARIMAX rotation), factor loadings for JIT variablesa Cronbach’s alpha correlation coefficients

Factor 1 (M) JIT manufacturing 0.831

Focused factory Group technology Reduced setup times Productive maintenance Multi-function employees Uniform work load Product quality improvement Process quality improvement Kanban system JIT purchasing

0.740 0.770 0.706 0.668 0.501 0.731

Factor 2 (Q) JIT quality 0.898b

Factor 3 (U) JIT unique 0.524b

0.917 0.902 0.820 0.825

Note: n = 253. a All loadings in excess of 0.300 are shown. b Significant at P < 0.001.

The three resulting JIT factors are similar to those identified in Davy et al. (1992), which were defined as operating structure and control, quality implementation, and product scheduling. The first factor, a manufacturing dimension, is comprised primarily of indicators that explain the extent to which companies have implemented advanced manufacturing techniques associated with JIT: focused factory, group technology, reduced setup times, productive maintenance, multi-function employees, and uniform work loads. This factor demonstrates the interrelatedness of the streamlining components of an advanced manufacturing system. Standardization of tasks in a cellular type of balanced, continuous flow, where setup times are minimal, works well in a focused factory in which employees are cross-trained and responsible for the maintenance of their machines (Sahin, 2000). The second JIT factor represents a quality dimension comprised of indicators that explain the extent to which companies have implemented procedures for improving product and process quality. Good quality management and productive maintenance are keys to JIT survival, with quality often referred to as the cornerstone of JIT (Banker et al., 1993; Imai, 1998; Sim and Killough, 1998; Swanson and Lankford, 1998; Young et al., 1988). The third factor is a uniquely JIT dimension. It identifies JIT practices that describe the extent to which companies have implemented JIT purchasing and kanban. These two elements are representative of the unique JIT practices related to

product scheduling, purchasing and delivering the right quantities at the right time. The likelihood is low that companies that are not fully committed to a JIT program would adopt these practices. A description of the specific survey questions1 that support these JIT factors is found in Appendix A. The factor solutions for the defined constructs support the construct validity of the survey instrument. Multiple-question loadings for each factor in excess of 0.50 demonstrate convergent validity (see Bagozzi and Yi, 1988). In addition, discriminant validity is supported, since none of the questions in the factor analyses have loadings in excess of 0.30 on more than one factor. In order to further test the construct validity of the resulting constructs, the factor structures were cross validated through the use of split samples (firms either declared as having implemented JIT or those that had not). Similar loadings in the cross-validation samples verified the initial underlying patterns. Cronbach’s alpha (1951) (or correlation coefficients where the construct has only two variables) is used as the coefficient of reliability for testing the internal consistency of the constructs validated by the factor analysis. The correlation coefficients are significant, and the alpha coefficient exceeds the standard of 0.70 for established constructs (Nunnally, 1978), as shown in Table 3. 1 Note: a copy of the questionnaire can be obtained from the leading author.

R.R. Fullerton et al. / Journal of Operations Management 21 (2003) 383–404

3.4. Control variables In addition to the JIT-specific dimensions identified above, the model includes four variables (inventory margin, product life cycle, innovation, and organizational structure) to control for non-JIT factors that are most likely to influence profitability. The literature discusses the strong relationship between inventory and JIT. The general inference is that lower inventory levels are a consequence, rather than a determinant of JIT (Chapman and Carter, 1990; Davy et al., 1992; Green et al., 1992; Hall, 1987; White and Ruch, 1990). Droge and Germain (1998) determined that the levels of JIT and inventory consistently have an inverse relationship, regardless of the contextual environment. Moreover, as demonstrated by Cachon and Fisher (1997), reducing inventories should have a positive effect on profitability. Minimizing inventory frees up working capital and reduces warehousing costs. These lower costs increase cash flows and profitability ratios, such as ROA, which measure profit margins against assets. To validate the relationship between JIT practices and inventory levels, a separate regression was run with inventory margin as the dependent variable and JIT practices as the independent variables. Results demonstrated a strongly significant inverse relationship as expected (not shown). Inventory margin (IY) is included in the regression models to control for reduced inventory effects on profitability created from the implementation of JIT. This measure representing total inventory divided by net sales is obtained from the COMPUSTAT database. Hofer and Schendel (1978, p. 98) suggested that “the most fundamental variable in determining an appropriate business strategy is the stage of the product life cycle (PLC),” which identifies four stages through which products travel: introduction, growth, maturity, and decline. A firm’s strategic responses are related to the stage of the product life cycle in which the majority of its production is occurring (Williams et al., 1995). Early in the growth cycle, new technology and growth take precedence over profit (Nevens et al., 1998). Firms with well-established main products have more stable profits than those firms whose products are in the initial production stages. Thus, mature products are expected to be at the peak of their earning power (Comiskey and Mulford, 1993; Nevens

391

et al., 1998; Yasukata and Kobayashi, 2001). To determine the PLC of the respondent firms, the survey instrument asked the respondents to select in which stage of the PLC, based on their main products, would they classify their firm: introductory, growth, maturity, or other. The responses were coded as “1” if the respondent firm answered maturity stage, and “0” if otherwise (introductory, growth, or other). This dummy variable was used as a control variable for PLC. Innovation is critical to the strategy of successful organizations. Entrepreneurial actions, if supported by top management and implemented effectively, help firms to create wealth by staying ahead of the competition (Ireland et al., 2001). By effectively evaluating and understanding their market place, technologically adept and innovative firms have increased sales and profits (Judd, 2000; Sivadas and Dwyer, 2000). Ittner and Larcker (1997) reported that innovation had a significant effect on profitability. However, the impact was industry specific. Innovation (IN) is measured by a firm’s response to the five-point Likert-scaled question on the survey instrument as to whether the firm was a leader or a follower in product technology, product design, and process design (Ittner and Larcker, 1997). Organizational structure also can influence a firm’s ability to be flexible and make major operational changes. Decentralization refers to the level of decision-making authority that is found in a firm (Aiken and Hage, 1968). If a firm is highly centralized, employees will be much less involved in decision making and organizational changes than if it is more decentralized. Decentralization allows firms to take better advantage of and respond more quickly to opportunities and events through decision making at the level of day-to-day activities (Sabath et al., 2001). Woodward’s (1980) seminal work demonstrated that organizational success was contingent upon the right combination of organizational structure and manufacturing technology. More recent studies have reported inconclusive results as to the relationships among JIT, decentralization, and firm performance (Claycomb et al., 1999; Germain et al., 1994). Kalagnanam and Lindsay (1998) showed how adapting more organic (decentralized) organizational structures led to greater performance benefits from JIT adoption. Additionally, Alles et al. (1995) displayed how JIT adoption and changes in the organizational structure have complementary effects, especially as JIT makes

392

R.R. Fullerton et al. / Journal of Operations Management 21 (2003) 383–404

the production process more transparent. The firm’s organizational structure is identified by responses to questions on decentralization, measured by five-point Likert-scales. A description of the specific survey questions that support the control variables of innovation and organizational structure is found in Appendix A. In addition, the six survey questions related to firm innovation and organizational structure were reduced and summarized using factor analysis. These converged into the two anticipated distinct factors (IN and S) with eigenvalues in excess of 1.0, accounting for 66% of the total variance in the data. Refer to Table 4 for the rotated factor solution. While some firms may have implemented certain JIT practices, they may not identify themselves specifically as a JIT firm. One contribution of this study is that it tests the financial effects from implementing specific JIT elements, rather than assuming each firm is either completely representative of a JIT firm or not. It is important, therefore, that these individual measures capture the JIT concept as understood in the manufacturing environment. In addition to identifying the degree of implementation of the individual JIT practices, respondents were asked to indicate whether or not (yes

Table 4 Factor analysis (VARIMAX rotation) factor loadings for control variablesa Cronbach’s alpha

Organizational structure Overall company Individual operations Individual departments

Factor 1 (S) organizational structure 0.793

Factor 2 (IN) innovation 0.677

0.703 0.877 0.844

Firm’s innovation strategy in Product technology Process design Product design

0.763 0.604 0.815

Note: n = 253. a All loadings in excess of 0.300 are shown.

or no) they had formally implemented JIT. Descriptive statistics depicting the means for the variables tested in the model for the total sample, self-identified JIT, and non-JIT firms are shown in Table 5. Statistical differences between the means of the JIT and non-JIT firms are also indicated. The ANOVA comparison of the means between the JIT and non-JIT firms consistently shows highly significant differences for the

Table 5 Descriptive statistics for model variables and comparison of means between JIT firms and non-JIT firms Full sample means

JIT firms means

Non-JIT firms means

ANOVA F-value

Significant F

measuresa

Profitability Return on sales (ROS) Return on assets (ROA) Cash flow margin (CFL) JIT variablesb JIT manufacturing (M) JIT quality (Q) JIT unique (U) Control variables Organizational structure (S)c Innovation (IN)d Product life cycle stage (P)e Inventory margin (IY)a Firm size (net sales)a

−5.071 1.490 −0.801

1.294 6.212 5.051

−8.907 −1.357 −4.351

3.287 4.310 3.264

0.071 0.039 0.072

3.412 4.665 3.233

4.055 5.026 4.263

3.020 4.446 2.606

54.975 17.809 114.106

0.000 0.000 0.000

2.860 3.694 0.494 17.649

3.075 3.940 0.473 14.483

2.731 3.543 0.506 19.570

7.391 14.467 0.257 11.115

0.007 0.000 0.613 0.001

465.276

800.109

264.489

19.801

0.000

Notes: n = 253. a Information provided from COMPUSTAT database. b Possible responses: no intention = 1; considering = 2; beginning = 3; partially = 4; substantially = 5; fully = 6. c Possible responses: highly centralized = 1, . . . , 2, . . . , 3, . . . , 4, . . . , 5 = highly decentralized. d Possible responses: follower = 1, . . . , 2, . . . , 3, . . . , 4, . . . , 5 = leader. e Possible responses: 1 = mature; 0 = introductory, growth, and other.

R.R. Fullerton et al. / Journal of Operations Management 21 (2003) 383–404

JIT variables, suggesting that the JIT measures are representative of JIT practices. Significance also exists in the profitability measures when examining the firms as either JIT or non-JIT firms. This result is in agreement with the findings of Kinney and Wempe (2002).

4. Research results and discussion 4.1. Cross-sectional model tests and results The empirical model of the relationship between the degree of JIT implementation and firm profitability in this study draws upon the relationships theorized and described by Milgrom and Roberts (1995). Their framework analyzes the relationships and impact on firm performance of the various parts of the manufacturing system. “They suggest that organizations often experience a simultaneous shift in competitive strategy along with various elements of organizational design when they move from mass production to JIT/TQM manufacturing. In addition, synergies, or complementarities, often arise with clusters of these elements that improve overall performance. . . ” (Sim and Killough, 1998, pp. 328–329). Tests of hypotheses are formulated in the context of restrictions placed on coefficients in multiple linear regression models (MLR). The cross-sectional model used to evaluate hypothesis H1 is presented as follows: JIT JIT JIT C Πi,j = β0,j + β1,j Mi + β2,j Qi + β3,j Ui + β1,j Si C C C + β2,j INi + β3,j Pi + β4,j IYi + εi

where Πi,j is the jth measure of profitability (ROA, ROS, CFL) for the ith firm, Mi the JIT manufacturing measure for the ith firm, Qi the JIT quality measure for the ith firm, Ui the JIT unique measure for the ith firm, Si the organizational structure measure for the ith firm, INi the innovation measure for the ith firm, Pi the product life cycle dummy variable for the ith firm, IYi the inventory margin measure for the ith firm. The external data of profitability and inventory measures for the cross-sectional model corresponded to the 1997 internal data collected. To partition the control variables from the explanatory variables, hierarchical multiple linear regressions were run independently for each of the profitability measures. In each

393

regression, the four control variables were first entered into the equation. Step two added the JIT independent variables and contributes to the understanding of the discrete impact of their explained variances. Before estimating the model, tests of potential multicollinearity among the set of independent variables were performed. The variance inflation factor (VIF) was used to evaluate evidence of collinearity. None of the VIFs exceeded 2.0, which is well below the conventional critical value of 10 at which point multicollinearity becomes problematic (Neter et al., 1983). An examination of the tolerance of the variables and the condition indices associated with the eigenvalues also support the lack of collinearity. Probability plots and the Kolmogorov–Smirnov (Smirnov, 1948) test demonstrated that the data were normally distributed. In addition, the models had similar results when outliers were excluded from the testing. Based on the F-statistic, each of the MLR equations is significant at a 0.01 level of significance. The adjusted R2 for the three performance measures ranges from 10.1% for ROA to 13.6% for CFL, indicating that other significant measures affecting firm performance have not been captured in the equations. Both the control variables and the JIT variables make a significant separate contribution to the model, although the JIT variables contribute less than one-third to the explained variances (refer to Table 6). 4.1.1. JIT explanatory variables The empirical results presented in Table 6 are generally consistent across all three profitability measures and demonstrate support for hypothesis H1 . Although the majority of the explained variances is from the control variables, JIT variables make a significant contribution to each of the models, suggesting that the relationship between JIT implementation and financial performance is robust across different indicators of firm profitability. The JIT manufacturing variable makes a significant contribution to the profitability measures. These results substantiate the conclusions of Inman and Mehra (1993) and Kinney and Wempe (2002), but contradict some of the findings of Balakrishnan et al. (1996), which found no differences between the financial performances of JIT and non-JIT firms unless controlled by other variables. The unexpected inverse relationships found between quality and profitability provide additional

394

R.R. Fullerton et al. / Journal of Operations Management 21 (2003) 383–404

Table 6 Regression results for the relationship between JIT practices and firm profitability ROS (j = 1)

ROA (j = 2)

CFL (j = 3)

Constant: t

0.449

0.266

0.493

Step 1 R2 F

0.118 7.659∗∗∗

0.090 5.650∗∗∗

0.125 8.164∗∗∗

4.290 2.312∗∗

6.799 3.082∗∗∗

Organizational structure C 7.089 S: β1,j S: t 2.960∗∗∗ Innovation strategy C −5.873 IN: β2,j IN: t −2.004∗∗

−2.009 −0.885

−5.119 −1.902∗

Product life cycle C P: β3,j P: t

10.383 2.179∗∗

8.579 2.324∗∗

9.790 2.232∗∗

Inventory margin C IY: β4,j IY: t

−0.470 −2.718∗∗∗

−0.354 −2.642∗∗∗

−0.463 −2.913∗∗∗

Step 2 R2 Change in R2 F

0.157 0.039 3.480∗∗

0.128 0.038 3.325∗∗

0.162 0.037 3.308∗∗

JIT manufacturing practices JIT M: β1,j 6.183 M: t 2.542∗∗

4.545 2.409∗∗

5.943 2.661∗∗

JIT quality practices JIT Q: β2,j −5.446 Q: t −2.375∗∗

−4.018 −2.260∗∗

−4.454 −2.088∗∗

1.358 0.992

1.029 0.634

0.101 4.752∗∗∗

0.136 6.224∗∗∗

JIT unique practices JIT U: β3,j U: t

1.509 0.855

Overall model statistics Adjusted R2 0.131 F value 6.010∗∗∗ Note: n = 253. ∗ P < 0.10. ∗∗ P < 0.05. ∗∗∗ P < 0.01.

insight into the on-going and widespread debate on the costs of quality. The JIT variable measuring the implementation of quality practices is significant and negative at the 0.05 level for all three profit equations. While some early references espoused that “quality is free” (Crosby, 1979), more recent studies suggest otherwise. Nakamura et al. (1998) discussed anecdotal

evidence that some firms have focused excessively on quality programs without adequate consideration of costs. Cammarano (1996) explained that companies frequently lack the vision of the total manufacturing process and concentrate on improving only one aspect, such as quality. They may achieve their goal of improving quality at the expense of higher costs. Safizadeh et al. (2000) examined specific trade-off concepts in manufacturing strategies. Their findings indicated a significant inverse relationship between cost and quality. The third explanatory JIT variable, JIT unique (U), is most analogous to firms that have formally adopted JIT. Its informational contribution to the model is inconsequential, showing no statistically significant association with measures of firm profitability. The differing results of the three JIT variables suggest that certain JIT practices will have a greater influence on profitability than will others. 4.1.2. Control variables The results outlined in Table 6 show that the four control variables account for the majority of the explained variances, and each of them makes a significant contribution to the regression equations. As expected, inventory level has a significant, negative relationship with all of the profitability measures. One of the strongest motivators for implementing JIT is to reduce inventory and correspondingly to reduce costs. The results in this study provide direct evidence that lower inventory margins are associated with increased financial performance. As the regression results of Table 6 demonstrate, both organizational structure and PLC have significant relationships to firm financial performance for all three of the profitability measures. The firms that are more decentralized and consider themselves to be in the mature stage of the PLC perform better. As explained by Nevens et al. (1998), the profit model for a mature firm works better in a more decentralized organizational structure where decision making is pushed down the line to those managing operations. An intriguing result is the negative association of innovation with the profit measures of ROS and CFL. This negative relationship could indicate that the sample firms that are more stable and less willing to lead out with risky and expensive innovations perform better. The lack of top management support and/or ineffective

R.R. Fullerton et al. / Journal of Operations Management 21 (2003) 383–404

implementation of innovations also might lead to failed efforts (Ireland et al., 2001). Alternatively, as Schrage (2000) proposed, a glut of innovation currently exists, much of which hurts rather than boosts a firm’s profit margin. 4.2. Time-series model tests and results In order to analyze the time-dependent structure of firm profitability as it relates to the implementation of JIT practices, a pooled cross-sectional, time-series regression model is used to evaluate hypothesis H2 . Time-series data are rich and unique, allowing for firm comparison to itself over time (Bernhardt et al., 2000). To address concerns that the data are likely to exhibit violations to the classical linear regression model assumptions of homoscedasticity and autocorrelation, a generalized least squares (GLS) cross-sectionally heteroscedastic and timewise autoregressive model (Kmenta, 1986) is estimated. This approach, which is frequently used in time-series studies (Droms and Walker, 1994; Hill and Phan, 1991; Kashlak et al., 1998; Krishnan and Largay, 2000), combines the assumptions made about cross-sectional observations (regression disturbances are mutually independent, but heteroscedastic) with those made about time-series (disturbances are autoregressive, but not heteroscedastic). The observations are subjected to a double transformation for removing both autocorrelation and heteroscedasticity, before applying the conventional ordinary least squares methodology (Kmenta, 1986, p. 509). To capture the linear relationships, as well as the property of increasing (or decreasing) returns to the degree of JIT implementation, hypothesis H2 is formulated in the context of a quadratic panel model specification as follows: Πi,j,t =

5


β0,k,j × Dk +

k=0

+

5


5


JIT β1,k,j × Mi × Dk

k=0 JIT β2,k,j × Qi × Dk +

k=0

× Dk +

5


JIT β3,k,j × Ui

k=0 5
k=0

JIT β4,k,j × Mi2 × Dk +

5
k=0

JIT β5,k,j

× Q2i × Dk + +

C β2,j INi

5


395

JIT C β6,k,j × Ui2 × Dk + β1,j Si

k=0 C + β3,j Pi

C C + β4,j IYi,t + β5,j Ci + εi,t

where Πi,j,t is the jth measure of profitability (ROA, ROS, CFL)  for the ith firm in period t (t = 1, . . . , 6), 1 when k = 0 or when k = t, t = 1, . . . , 5 Dk = 0 otherwise and Ci is the Top Management Commitment index for the ith firm. All other variables are as specified in hypothesis H1 . JIT implementation levels are not observed in every sample period; therefore, the time-dependent structure of JIT implementation processes is captured through the specification of dummy variables D1 , D2 , D3 , D4 , and D5 . The dummy variables are used as indicators of elapsed time surrounding the JIT adoption year. D1 , and D2 are indicators for the 2 years preceding adoption, D3 is an indicator of the year of adoption, and D4 , and D5 reflect the 2 years following JIT implementation. Year 6, which is the third year of post-JIT adoption, is the omitted variable benchmark, D0 , which makes possible the dummy variable comparisons of change in profitability levels over time. The quadratic specification included in the model allows for identification of the turning point at which the marginal return to implementing JIT is zero. If the quadratic coefficients are significant, firms should expect that marginal returns will not remain constant as the degree of JIT implementation changes. In this vein, the quadratic model specification is critical in addressing the question of whether, and how the degree of implementation of JIT practice affects firm profitability. Successful implementation of JIT requires complete support of the firm’s top executives (Ahmed et al., 1991). According to Willis and Suter (1989), the project is “doomed” if top management does not support the JIT philosophy. In any change effort, management must be willing to devote the necessary resources and provide the proper training and motivation for it to reap its potential benefits. The level of top management commitment to JIT implementation (C), as measured by a five-point Likert scale, is added to the pooled model as a control variable. Only those respondents who indicated that they formally had implemented JIT were asked to assess top-management commitment; therefore, the latter could not be used

396

R.R. Fullerton et al. / Journal of Operations Management 21 (2003) 383–404

as a control variable in the full-sample cross-sectional analysis. In addition, the four control variables used in the cross-sectional model are included in the longitudinal analysis. Results of the pooled cross-sectional, time-series regression are shown in Tables 7–9. To facilitate interpretation of the regression results, two conventions are used in reporting the time-related effects. First, with the exception of the constant terms, for each variable that interacts with a time dummy variable, JIT + βJIT , m = the combined coefficients (i.e. βm,0,j m,k,j 1, . . . , 6; k = 1, . . . , 5; and j = 1–3) are reported, along with their respective calculated t-statistics (based on calculated standard errors from the variance/covariance matrix). This convention facilitates the interpretation of the coefficients’ direct effect on profitability, rather than their effect on profitability relative to the omitted variable benchmark. Second, the turning-point statistics are reported for the curvilinear effects. The turning-point statistics are calculated by taking the derivative of the regression equation with respect to the quadratic variable, setting the derivative equal to zero and solving for the value of the quadratic variable as shown in the following JIT + βJIT )/2(βJIT + βJIT ), m = equation: −(βm,0,j n,0,j m,k,j n,k,j 1–3; n = m + 3; k = 1, . . . , 5; and j = 1–3. This provides the value of the variable where the slope is zero (turning point). Investments in JIT practices beyond this point will have increasing (decreasing) returns if the coefficient is positive (negative). The Buse Raw Moment R2 (Buse, 1973) is 0.499 for dependent variable ROS, 0.602 for dependent variable ROA, and 0.710 for dependent variable CFL. The Durbin–Watson D-statistic, which is a standard measure for evaluating serial correlation, indicates that the GLS estimation technique adequately corrected for potential serial correlation. 4.2.1. JIT explanatory variables Results for the three time-series regression models show interesting and fairly consistent patterns among the JIT explanatory variables. The JIT quality and JIT unique indicators offer more statistical validity to the model than the JIT manufacturing indicator. This result is in contrast to the cross-sectional results in which JIT manufacturing is generally significant. Coefficients on the quadratic JIT unique indicator provide the strongest argument for increasing marginal

returns to investment in JIT processes. Turning-point values are generally smallest in the post-adoption year 3 for the JIT quality and JIT unique indicators. None of the turning points in the post-adoption years for investments in unique JIT practices exceed 4.0 on the 6.0 scale. These declining turning-point values suggest that firms are able to extract the benefits of even modest implementations, as they gain experience with the JIT system. Thus, the trend results imply that the degree to which some JIT practices are implemented have long-term consequences for firm profitability. Table 8 indicates that in the first year prior to JIT adoption, those firms that had implemented a higher degree of JIT manufacturing practices had significantly lower ROA than in the post-JIT adoption years. Beyond that, no evidence exists that the degree to which JIT firms have implemented JIT manufacturing practices affects long-term financial returns. The JIT quality variable exhibits the most significant linear relationship with firm profitability. Interestingly, the statistically significant coefficients demonstrate that higher investments in quality contribute to stronger profitability in the pre-implementation years. This divergent relationship ceases to exist in post-adoption years, as indicated by insignificant Q coefficients across all of the models. Specifically, Table 8 shows that firms with higher implementation degrees of quality practices had more positive ROAs in the JIT pre-adoption and adoption years than in the JIT post-adoption period. For the quadratic terms, a significant relationship also exists between the degree of JIT implementation and ROA in the years before JIT adoption, but not after. As expected, the signs of the quadratic term coefficients are the opposite of the signs for the linear coefficients. Given the signs for both the linear and quadratic Q coefficients, returns from JIT quality investments are positive, but declining in both the degree of implementation and over time. These results suggest that the costs incurred in the first few years of implementing quality improvements may be such that they offset any potential financial gains associated with JIT implementation. In the cross-sectional, full sample results of hypothesis H1 , JIT unique demonstrated no significance. However, the pooled time-series analysis shows some financial improvement for those JIT firms that have implemented higher degrees of kanban and

R.R. Fullerton et al. / Journal of Operations Management 21 (2003) 383–404

397

Table 7 Pooled cross-section time-series regression results for the relationship between JIT practices and return on sales Control variables Constant Constant: β0,k,1 Constant: t JIT manufacturing practices JIT JIT Ma : β1,0,1 + β1,k,1 M: t JIT JIT M2 a : β4,0,1 + β4,k,1 M2 : t

Turning point JIT quality practices JIT JIT Qa : β2,0,1 + β2,k,1 Q: t JIT JIT + β5,k,1 Q2 a : β5,0,1 Q2 : t

Pre 2 (k = 1)

JIT unique practices JIT JIT Ua : β3,0,1 + β3,k,1 U: t JIT JIT U2 a : β6,0,1 + β6,k,1 2 U :t

Turning point Organizational structure C S: β1,1 S: t

0.22 0.47

Innovation strategy C IN: β2,1 IN: t

0.27 0.49

Product life cycle C P: β3,1 P: t

0.57 0.64

Inventory margin C IY: β4,1 IY: t

−0.69 −2.24∗∗

Adopt (k = 3)

Post 1 (k = 4)

Post 2 (k = 5)

Post 3 (k = 0)

−8.79 0.46

7.80 0.43

−6.61 −0.39

3.47 0.23

6.75 0.56

−7.42 −0.47

0.73 0.09

−2.11 −0.28

2.90 0.39

1.10 0.16

3.86 0.60

2.90 0.61

−0.11 −0.18

0.28 0.53

−0.22 −0.42

−0.11 −0.20

−0.48 −0.88

−0.25 −0.43

3.39

3.73

6.50

5.11

4.05

5.73

11.00 1.99∗∗

7.04 1.37

8.30 1.75∗

3.84 0.84

−4.57 −0.97

4.07 0.80

−0.79 −1.42

−0.91 −1.65∗

−0.43 −0.78

0.46 0.83

−0.48 −0.84

−1.11 −1.95∗

Turning point

Pre 1 (k = 2)

4.95

4.46

4.54

4.44

4.95

4.28

−6.87 −1.11

−6.59 −1.09

−6.68 −1.17

−5.01 −0.95

−0.26 −0.06

−5.84 −1.57

0.81 1.86∗

0.76 1.81∗

0.79 1.90∗

0.66 1.58

0.20 0.47

0.79 1.84∗

4.26

4.35

4.23

3.80

0.66

3.68

Top management commitment C 0.91 C: β5,1 C: t 1.64 Durbin–Watson D-statistic 1.65 Buse Raw Moment R2 0.499 Model F statistic 6.17∗∗∗ Note: n = 324 (number of firms = 54; number of years for each firm = 6). a For k = 0, the reported estimate is solely βJIT , m = 1, . . . , 6. m,0,1 ∗ P < 0.1. ∗∗ P < 0.05. ∗∗∗ P < 0.01.

398

R.R. Fullerton et al. / Journal of Operations Management 21 (2003) 383–404

Table 8 Pooled cross-section time-series regression results for the relationship between JIT practices and return on assets Control variables Constant Constant: β0,k,2 Constant: t JIT manufacturing practices JIT JIT + β1,k,2 Ma : β1,0,2 M: t JIT JIT + β4,k,2 M2 a : β4,0,2 M2 : t

Pre 2 (k = 1) Pre 1 (k = 2) Adopt (k = 3) Post 1 (k = 4) Post 2 (k = 5) Post 3 (k = 0)

−58.07 −2.98∗∗∗

0.34 0.02

−36.00 −2.02∗∗∗

−7.05 −0.42

−14.46 −1.04

22.59 1.52

0.50 0.10

−10.6 −2.29∗∗∗

0.13 0.03

−2.40 −0.52

−1.36 −0.29

−3.95 −0.82

0.05 0.08

0.35 0.63

0.19 0.33

0.48 0.82

3.39

3.66

4.08

−0.15 −0.28

Turning point

1.61

JIT quality practices JIT JIT Qa : β2,0,2 + β2,k,2 Q: t JIT JIT + β5,k,2 Q2 a : β5,0,2 Q2 : t

Turning point JIT unique practices JIT JIT Ua : β3,0,2 + β3,k,2 U: t JIT JIT U2 a : β6,0,2 + β6,k,2 2 U :t

Turning point Organizational structure C S: β1,2 S: t

0.96 2.23∗∗

Innovation strategy C IN: β2,2 IN: t

0.09 0.16

Product life cycle C P: β3,2 P: t

−0.34 −0.41

Inventory margin C IY: β4,2 IY: t

−12.79 −3.6∗∗∗

Top management commitment C C: β5,2 C: t Durbin–Watson D-statistic Buse Raw Moment R2 Model F statistic

1.41 2.50∗∗ 3.77

N/A

19.09 4.71∗∗∗

9.45 2.35∗∗

11.38 2.87∗∗∗

−1.53 0.39

−4.70 −1.18

0.22 0.05

−2.00 −4.44∗∗∗

−1.17 −2.65∗∗

−1.32 −3.04∗∗∗

−0.02 −0.04

0.33 0.76

−0.29 −0.64

4.76

4.02

4.30

7.12

0.38

−4.62 −1.25

−11.12 −3.33∗∗∗

−5.85 −1.76∗∗

−2.72 −0.82

2.63 0.79

−5.19 −1.41

0.61 1.43

1.29 3.33∗∗∗

0.70 1.82∗

0.39 1.02

−0.10 −0.26

0.76 1.80∗

3.77

4.30

4.18

3.46

13.09

3.40

0.73 1.79∗ 1.63 0.620 9.64∗∗∗

Note: n = 324 (number of firms = 54; number of years for each firm = 6). a For k = 0, the reported estimate is solely βJIT , m = 1, . . . , 6. m,0,2 ∗ P < 0.1. ∗∗ P < 0.05. ∗∗∗ P < 0.01.

N/A

R.R. Fullerton et al. / Journal of Operations Management 21 (2003) 383–404

399

Table 9 Pooled cross-section time-series regression results for the relationship between JIT practices and cash flow Control variables Constant Constant: β0,k,3 Constant: t JIT manufacturing practices JIT JIT Ma : β1,0,3 + β1,k,3 M: t JIT JIT M2 a : β4,0,3 + β4,k,3 M2 : t

Turning point JIT quality practices JIT JIT Qa : β2,0,3 + β2,k,3 Q: t JIT JIT + β5,k,3 Q2 a : β5,0,3 Q2 : t

Turning point

−19.50 −1.00

−6.74 −0.37

−18.86 −1.08

−4.15 −0.26

1.12 0.87

15.73 1.03

−2.28 0.26

−1.78 −0.21

1.70 0.21

−2.07 −0.27

0.81 0.12

−3.55 −0.69

0.34 0.54

0.32 0.55

−0.09 −0.16

0.31 0.54

−0.07 −0.12

0.60 0.96

3.32

2.78

9.21

3.37

5.73

2.99

11.18 2.47∗∗

6.72 1.60

6.01 1.62

2.58 0.74

−6.15 −1.60

2.40 0.52

−1.21 −2.37∗∗

−0.82 −1.64

−0.70 −1.41

−0.35 −0.71

0.57 1.14

−0.37 −0.74

4.09

4.26

3.63

5.38

3.22

−8.03 −1.50

−7.38 −1.46

−3.67 −0.80

−8.30 −2.26∗∗

4.61

JIT unique practices JIT JIT Ua : β3,0,3 + β3,k,3 U: t

−9.32 −1.66∗

JIT JIT U2 a : β6,0,3 + β6,k,3 2 U :t

Turning point Organizational structure C S: β1,3 S: t Innovation strategy C IN: β2,3 IN: t

Pre 2 (k = 1) Pre 1 (k = 2) Adopt (k = 3) Post 1 (k = 4) Post 2 (k = 5) Post 3 (k = 0)

−10.44 −1.89∗

1.08 2.49∗∗

1.23 2.94∗∗∗

0.98 2.37∗∗

0.96 2.32∗∗

0.64 1.52

1.09 2.52∗∗

4.32

4.23

4.08

3.83

2.89

3.82

0.25 0.63 0.76 1.34

Product life cycle C P: β3,3 P: t

0.46 0.50

Inventory margin C IY: β4,3 IY: t

−1.92 −0.53

Top management commitment C C: β5,3 0.78 C: t 1.48 Durbin–Watson D-statistic 1.68 Buse Raw Moment R2 0.710 Model F statistic 15.13∗∗∗ Note: n = 324 (number of firms = 54; number of years for each firm = 6). a For k = 0, the reported estimate is solelyβJIT , m = 1, . . . , 6. m,0,3 ∗ P < 0.1. ∗∗ P < 0.05. ∗∗∗ P < 0.01.

400

R.R. Fullerton et al. / Journal of Operations Management 21 (2003) 383–404

JIT purchasing. Linear coefficients for variable U in post-adoption year three are all negative, significantly so for the cash-flow model. Although a negative relationship to profit exists for those firms that have implemented higher degrees of JIT unique practices, the results suggest an improvement from a significantly lower pre-adoption profitability level. In addition, the quadratic coefficients are all positive and indicate increasing marginal returns as the level of JIT implementation rises. Tables 7–9 show turning points in the 3.4–3.8 range for post-adoption year 3, indicating rising returns to JIT unique investments over time, even for modest implementation of JIT unique practices. 4.2.2. Control variables Except for inventory margins, the control variables are relatively unimportant contributors to the time-dependent models (refer to Tables 7–9). Lower inventory levels have a significant effect on ROA and ROS, but not CFL. The insignificant results for the cash-flow model may be related to inventory carrying costs. Lower inventory levels would translate into diminished space requirements for inventory storage. These costs generally are expensed through depreciation. Decreasing depreciation expense over time would influence profitability measures of ROA and ROS, but would not affect cash flows directly. The added control variable in hypothesis H2 is the level of commitment to JIT by top management. The coefficient is positive and significant only for the ROA profitability model. Earlier descriptive survey studies have found the lack of support from top management to be a serious problem in the JIT implementation process (Ansari and Modarress, 1986; Celley et al., 1986; Im, 1989; Lee, 1997). An ANOVA examining the relationship between JIT implementation and top management support is significant to P < 0.01 for all of the JIT variables (not shown). However, the degree of commitment to JIT implementation by top management appears to have minimal effect on the three selected measures of firm performance.

5. Summary According to its proponents, JIT’s global management philosophy of waste elimination and continu-

ous improvement leads to more efficient operations. Managers act rationally in adopting and supporting JIT implementation when these operating efficiencies translate into improved financial performance. The principal objective of this research has been to determine, from both a static and time-dependent perspective, if JIT practices make a positive contribution to firm profitability. The results of hypothesis H1 provide empirical support to the premise that firms that implement higher degrees of JIT manufacturing practices should outperform competitors who do not. A positive relationship exists between firm profitability and the degree to which waste-reducing production practices, such as reduced setup times, preventive maintenance programs, and uniform workloads, are implemented. These findings complement the views of Womack and Jones (1996) that lean enterprises employing JIT manufacturing techniques are consistently more profitable than their counterparts. The research data indicate that the implementation of a greater degree of JIT quality practices decreased sample firm profitability. These results are not conclusive, since they may imply either that the degree of implementation of JIT quality indicators reduces profitability, or that firms with low profitability recognize their strategic disadvantage and increase their focus on quality improvement by implementing JIT quality processes. Schonberger (1986) and many other proponents of TQM claimed that the trade-off between quality and cost was a myth. TQM advocates contended that leaders in cost reduction were also leaders in quality and vice versa. Hendricks and Singhal (1997) found that after winning quality awards, TQM sample firms outperformed non-TQM sample firms. Alternatively, Hill (1993) described the demise of the Wallace Co., after sustaining unrecoverable losses from heavy expenditures on quality that had contributed previously to its winning of the Malcolm Baldrige National Quality award. Rust et al. (1995) outlined an approach to evaluate quality improvement efforts, noting the disparate validity of such investments. The JIT unique measure demonstrates no significant relationship with profitability. The diverse contribution of the three JIT variables to profitability may provide partial explanation for the inconsistent results of earlier studies in which firms were classified as either JIT or non-JIT. The examination of individual

R.R. Fullerton et al. / Journal of Operations Management 21 (2003) 383–404

components of JIT processes and their degree of implementation provides insights into which specific JIT techniques affect firm financial performance. The results of hypothesis H2 indicate increasing marginal returns to long-term JIT investment, especially for unique JIT practices such as kanban and JIT purchasing. In contrast, the insignificant association in the full cross-sectional, static model suggests that the benefits of these JIT practices are realized only over time due to their higher implementation costs. Evidence from hypothesis H2 also suggests caution when implementing product- and process-quality improvements, as potential trade-offs between cost and quality can affect financial performance negatively. Reduced firm profitability can ensue when those responsible for quality costs are not accountable for overall production costs and the costs that they impose on other activities. Specific limitations might reduce the generalizability and applicability of the research findings. First, a necessary assumption in survey research is that the respondents had sufficient knowledge to answer the items, and that they answered the questions conscientiously and truthfully. Respondents might have been unfamiliar with questionnaire terms used to describe JIT methods, and reluctant to take the necessary time to examine the attached glossary explaining the JIT terminology. Second, an important element of this survey instrument is capturing the degree of JIT implementation. The 11 JIT indicators on the survey were supported through a thorough study of JIT literature; yet, they might not have been completely indicative of actual company practices. Finally, while the majority of the sample firms were selected through a random process, as in Balakrishnan et al. (1996) and Kinney and Wempe (2002), the identification of the JIT sample firms precluded random selection. The diversion from completely random sample selection might make the test sample non-representative of other US manufacturing firms. This study explains a small subset of factors affecting financial performance. Future research extending the time-series to include additional years of post-adoption data would provide fresh insights into the sustainability of financial performance in a JIT environment. Research is also necessary to capture more fully the complementarities that exist amongst JIT implementation and other organizational policies

401

and procedures, including the economic and environmental contexts that influence both the choice of different manufacturing strategies and their subsequent impact on financial performance. Acknowledgements This paper is based on a portion of the first author’s doctoral dissertation completed at the University of Utah. The valuable input from graduate committee faculty is gratefully acknowledged, as well as the comments from colleagues, reviewers, discussants, and participants at the Management Accounting Research Conference (January 2000) and the ASAC-ISAM Conference (July 2000). Appendix A. Survey items measuring JIT implementation factors and control variables JIT manufacturinga Indicate the extent to which your firm has implemented the following techniques: Focused factory Group technology Action plan to reduce setup times Total productive maintenance Multi-function employees Uniform work load JIT qualitya Indicate the extent to which your firm has implemented the following techniques: Product quality improvement Process quality improvement JIT uniquea Indicate the extent to which your firm has implemented the following techniques: Kanban system JIT purchasing Innovation strategyb What most closely matches your firm’s strategy related to innovation in: Product technology? Process design? Product design?

402

R.R. Fullerton et al. / Journal of Operations Management 21 (2003) 383–404

Appendix A. (Continued ) structurec

Organizational What is the organizational structure of your firm’s: Overall company? Individual operations? Individual departments? Possible responses: no intention = 1; considering = 2; beginning = 3; partially = 4; substantially = 5; fully = 6. b Possible responses: follower = 1, . . . , 2, . . . , 3, . . . , 4, . . . , 5 = leader. c Possible responses: highly centralized = 1, . . . , 2, . . . , 3, . . . , 4, . . . , 5 = highly decentralized. a

References Ahmed, N.U., Runc, E.A., Montagno, R.V., 1991. A comparative study of US manufacturing firms at various stages of just-in-time implementation. International Journal of Production Research 29 (4), 787–802. Aiken, M., Hage, J., 1968. Organizational independence and intraorganizational structure. American Sociological Review 33, 912–930. Alles, M., Datar, S.M., Lambert, R.A., 1995. Moral hazard and management control in just-in-time settings. Journal of Accounting Research 33 (Suppl.), 177–204. Anderson, R.H., Needles Jr., B.E., Caldwell, J.C.J., 1989. Management Accounting. Houghton Miffin Co. Ansari, A., Modarress, B., 1986. Just-in-time purchasing: problems and solutions. Journal of Purchasing and Materials Management Summer, 11–15. Bagozzi, R.P., Yi, Y., 1988. On the evaluation of structural equation models. Journal of Academy Marketing Science 16 (1), 74–94. Balakrishnan, R., Linsmeier, T.J., Venkatachalam, M., 1996. Financial benefits from JIT adoption: effects of customer concentration and cost structure. The Accounting Review 71 (2), 183–205. Banker, R., Potter, G., Schroeder, R., 1993. Reporting manufacturing performance measures to workers: an empirical study. Journal of Management Accounting Research 5 (Fall), 33–53. Bernhardt, K.L., Donthu, N., Kennett, P.A., 2000. A longitudinal analysis of satisfaction and profitability. Journal of Business Research 47 (2), 161–171. Billesbach, T.J., 1991. A study of the implementation of justin-time in the United States. Production and Inventory Management Journal 3, 1–4. Billesbach, T.J., Hayen, R., 1994. Long-term impact of just-in-time on inventory performance measures. Production and Inventory Management Journal 1, 62–67. Blackburn, J.D., 1991. Time-Based Competition: The Next Battleground in American Manufacturing. Business One Irwin, Homewood, IL, p. 25.

Buse, A., 1973. Goodness of fit in generalized least squares estimation. American Statistician 27, 106–108. Cachon, G., Fisher, M., 1997. Campbell soup’s continuous replenishment program: evaluation and enhanced inventory decision rules. Production and Operations Management 6 (3), 266– 276. Callen, J.L., Fader, C., Krinsky, I., 2000. Just-in-time: a crosssectional plant analysis. International Journal of Production Economics 63, 277–301. Cammarano, J.R., 1996. Is it time to turn out the lights? IIE Solutions 28, 24–33. Celley, A.F., Clegg, W.H., Smith, A.W., Vonderembse, M.A., 1986. Implementation of JIT in the United States. Journal of Purchasing and Materials Management Winter, 9–15. Chapman, S.N., Carter, P.L., 1990. Supplier/customer inventory relationships under just-in-time. Decision Sciences, 35–51. Claycomb, C., Droge, E., Germain, R., 1999. The effect of justin-time with customers on organizational design and performance. International Journal of Logistics Management 10 (1), 37–58. Clode, D.M., 1993. A survey of UK manufacturing control over the past 10 years. Production and Inventory Management Journal 2, 53–56. Comiskey, E.E., Mulford, C.W., 1993. Anticipating trends in operating profits and cash flow. Commercial Lending Review 8 (2), 38–49. Conner, G., 2001. Lean Manufacturing for the Small Shop. Society of Manufacturing Engineers, Dearborn, MI. Cooper, R., 1995. When Lean Enterprises Collide: Competing Through Confrontation. Harvard Business School Press, Boston, MA. Crawford, K.M., Cox, J.F., 1990. Designing performance measurement systems for just-in-time operations. International Journal of Production Research 28 (11), 2025–2036. Cronbach, L.J., 1951. Construct validity in psychological tests. Psychological Bulletin, 297–334. Crosby, P.G., 1979. Quality is Free. McGraw-Hill, New York. Daniel, S.J., Reitsperger, W.D., 1991. Linking quality strategy with management control systems: empirical evidence from Japanese industry. Accounting, Organizations and Society 16 (7), 601– 618. Davy, J., White, R., Merritt, N., Gritzmacher, K., 1992. A derivation of the underlying constructs of just-in-time management systems. Academy of Management Journal 35 (3), 653–670. Droge, C., Germain, R., 1998. The just-in-time inventory effect: does it hold under different contextual, environmental, and organizational conditions? Journal of Business Logistics 19 (2), 53–71. Droms, W.G., Walker, D.A., 1994. Investment performance of international mutual funds. The Journal of Financial Research Spring, 1–14. Foster, G., Horngren, C.T., 1987. JIT: cost accounting and cost management issues. Management Accounting 6, 19–25. Fullerton, R.R., McWatters, C.S., 2001. The production performance benefits from JIT implementation. Journal of Operations Management 19, 81–96.

R.R. Fullerton et al. / Journal of Operations Management 21 (2003) 383–404 Fullerton, R.R., McWatters, C.S., 2002. The role of performance measures and incentive systems in relation to the degree of JIT implementation. Accounting, Organizations and Society 27 (8), 711–735. Germain, R., Droge, C., Daugherty, P., 1994. The effect of just-in-time selling on organizational structure: an empirical investigation. Journal of Marketing Research 16, 471–483. Gilbert, J.P., 1990. The state of JIT implementation and development in the USA. International Journal of Production Research 28 (6), 1099–1109. Goyal, S.K., Deshmukh, S.G., 1992. A critique of the literature on just-in-time. Journal of Operations and Production Management 12 (1), 18–28. Green, F.B., Amenkhienan, F., Johnson, G., 1992. Performance measures and JIT. Management Accounting 10, 32–36. Hall, R.W., 1987. Attaining Manufacturing Excellence. Business One Irwin, Homewood, IL. Hall, D., Jackson, J., 1992. Speeding up new product development. Management Accounting 10, 32–36. Hayes, R.H., Wheelwright, S.C., 1984. Restoring Our Competitive Edge. Collier Macmillan, New York. Hendricks, K.B., Singhal, V.R., 1997. Does implementing an effective TQM program actually improve operating performance? Empirical evidence from firms that have won quality awards. Management Science 43 (9), 1258–1274. Hill, C.W.L., Phan, P., 1991. CEO tenure as a determinant of CEO pay. Academy of Management Journal 34 (3), 707–717. Hill, R.C., 1993. When the going gets rough: a Baldrige award winner on the line. Academy of Management Executive 7 (8), 75–79. Hofer, C.W., Schendel, D., 1978. Strategy Formulation: Analytical Concepts. West Publishing, St. Paul, MN. Huson, M., Nanda, D., 1995. The impact of just-in-time manufacturing on firm performance in the US. Journal of Operations Management 12, 297–310. Im, J.H., 1989. How does kanban work in American companies? Production and Inventory Management Journal 4, 22–24. Im, J.H., Lee, S.M., 1989. Implementation of just-in-time systems in US manufacturing firms. International Journal of Operations and Production Management 9 (1), 5–14. Imai, M., 1998. Will America’s corporate theme song be ‘Justin-Time’? Journal for Quality and Participation 21 (3/4), 26–28. Inman, R.A., Mehra, S., 1993. Financial justification of JIT implementation. International Journal of Operations and Production Management 13 (4), 32–39. Ireland, R.D., Hitt, M.A., Camp, S.M., 2001. Integrating entrepreneurship and strategic management actions to create firms wealth. The Academy of Management Executive 15 (1), 49–63. Ittner, C.D., Larcker, D.F., 1997. Quality strategy, strategic control systems, and organizational performance. Accounting, Organizations and Society 22 (3/4), 293–314. Johnson, H.T., Bröms, A., 2000. Profit Beyond Measure. The Free Press, New York. Johnson, H.T., Kaplan, R.S., 1989. Management by accounting is not management accounting. The Accounting Systems Journal Fall, 14–17. Judd, J., 2000. Add more quarters. Upside 12 (10), 167–174.

403

Jusko, J., 1999. A look at lean. Industry Week 248 (22), 88–92. Kalagnanam, S.S., Lindsay, R.M., 1998. The use of organic models of control in JIT firms: generalising Woodward’s findings to modern manufacturing practices. Accounting, Organizations and Society 24 (1), 1–30. Kaplan, R.S., Atkinson, A., 1989. Advanced Management Accounting. Prentice-Hall, London. Kashlak, R.J., Chandran, R., Di Benedetto, C.A., 1998. Reciprocity in international business: a study of telecommunications alliances and contracts. Journal of International Business Studies 29 (2), 281–304. Kinney, M.R., Wempe, W.F., 2002. Further evidence on the extent and origins of JIT’s profitability effects. The Accounting Review 77 (1), 203–225. Kmenta, J., 1986. Elements of Econometrics. Macmillan, New York. Koufteros, X.A., Vonderembse, M.A., Doll, W.J., 1998. Developing measures of time-based manufacturing. Journal of Operations Management 16, 21–41. Krishnan, G.V., Largay, J.A., 2000. The predictive ability of direct method cash flow information. Journal of Business, Finance, and Accounting 27 (1/2), 215–245. Lee, C.Y., 1997. JIT adoption by small manufacturing firms in Korea. Journal of Small Business Management 35 (7), 98–107. Mehra, S., Inman, R.A., 1992. Determining the critical elements of just-in-time implementation. Decision Sciences 23, 160–174. Mia, L., 2000. Just-in-time manufacturing. Accounting and Business Research 30 (2), 137–151. Milgrom, P., Roberts, J., 1995. Complementarities and fit: strategy, structure and organizational change in manufacturing. Journal of Accounting and Economics 19, 179–208. Monden, Y., 1998. Toyota Production System: An Integrated Approach to Just-In-Time. Engineering and Management Press, Norcross, GA. Moshavi, S., 1990. Well Made in America: Lessons from Harley-Davidson on Being the Best. McGraw-Hill, New York. Nakamura, M., Sakakibara, S., Schroeder, R., 1998. Adoption of just-in-time manufacturing methods at US- and Japanese-owned plants: some empirical evidence. IEEE Transactions on Engineering Management 45, 230–240. Neter, J., Wasserman, W., Kutner, M., 1983. Applied Regression Models, Irwin, Homewood, IL. Nevens, T.M., Guiniven, J.R., Paulsen, A.B., 1998. The key to success in electronics: shifting the balance between profit and growth. The McKinsey Quarterly 2, 140–147. Norris, D.M., 1992. A study of JIT implementation techniques using the analytic hierarchy process model. Production and Inventory Management Journal 3, 49–53. Norris, D.M., Swanson, R.D., Chu, Y., 1994. Just-in-time production systems: a survey of managers. Production and Inventory Management Journal 2, 63–66. Nunnally, J., 1978. Psychometric Theory. McGraw-Hill, New York. Orth, D., Hybil, R., Korzan, D., 1990. Analysis of a JIT implementation at Dover Corporation. Production and Inventory Management Journal 31 (3), 79–82. Ransom, C., 2001. Lean manufacturing provides big investment fun with solid returns. Manufacturing News 8 (16 April), 5–10.

404

R.R. Fullerton et al. / Journal of Operations Management 21 (2003) 383–404

Rust, R.T., Zahorik, A.J., Keningham, T.L., 1995. Return on quality (ROQ): making service quality financially accountable. Journal of Marketing 59 (2), 58–80. Sabath, R.E., Autry, C.W., Daugherty, P.J., 2001. Automatic replenishment programs: the impact of organizational structure. Journal of Business Logistics 22 (1), 91–105. Safizadeh, M.H., Ritzman, L.P., Mallick, D., 2000. Revisiting alternative theoretical paradigms in manufacturing strategy. Production and Operations Management 9 (2), 111–124. Sahin, F., 2000. Manufacturing competitiveness: different systems to achieve the same results. Production and Inventory Management Journal 41 (1), 56–65. Sakakibara, S., Flynn, B.B., Schroeder, R.G., 1993. A framework and measurement instrument for just-in-time manufacturing. Production and Operations Management 2 (3), 177– 194. Schonberger, R.J., 1986. World-Class Manufacturing: The Lesson of Simplicity Applied. The Free Press, New York. Schrage, M., 2000. Getting beyond the innovation fetish. Fortune 11 (13), 225–232. Sim, K.L., Killough, L.N., 1998. The performance effects of complementarities between manufacturing practices and management accounting systems. Journal of Management Accounting Research 10, 325–346. Sivadas, E., Dwyer, F.R., 2000. An examination of organizational factors influencing new product success in internal and alliance-based processes. Journal of Marketing 64 (1), 31– 49. Skinner, W., 1974. The focused factory. Harvard Business Review 52 (3), 113–122. Smirnov, N.V., 1948. Table for estimating the goodness of fit of empirical distributions. Annals of Mathematical Statistics 19, 279–281. Spencer, M.S., Guide, V.D., 1995. An exploration of the components of JIT. International Journal of Operations and Production Management 5 (5), 72–83.

Swanson, C., Lankford, W.M., 1998. Just-in-time manufacturing. Business Process Management Journal 4 (4), 333–341. Voss, C.A., 1995. Alternative paradigms for manufacturing strategy. International Journal of Production Management 15 (4), 5–16. White, R.E., 1993. An empirical assessment of JIT in US manufacturers. Production and Inventory Management 34 (2), 38–42. White, R.E., Ruch, W.A., 1990. The composition and scope of JIT. Operations Management Review 7 (3/4), 9–18. White, R.E., Prybutok, V., 2001. The relationship between JIT practices and type of production system. Omega 29, 113–124. White, R.E., Pearson, J.N., Wilson, J.R., 1999. JIT manufacturing: a survey of implementations in small and large US manufacturers. Management Science 45, 1–15. Williams, F.P., D’Souza, D.E., Rosenfeldt, M.E., Kassaee, M., 1995. Manufacturing strategy, business strategy, and firm performance in a mature industry. Journal of Operations Management 13, 19–33. Willis, T.H., Suter Jr., W.C., 1989. The five M’s of manufacturing: a JIT conversion life cycle. Production and Inventory Management Journal 1, 53–56. Woodward, J., 1980. Industrial Organization: Theory and Practice. Oxford University Press, Oxford. Womack, J.P., Jones, D.T., 1996. Lean Thinking: Banish Waste and Create Wealth in Your Corporation. Simon & Schuster, New York, p. 121. Yasin, M.M., Small, M., Wafa, M., 1997. An empirical investigation of JIT effectiveness: an organizational perspective. Omega 25 (4), 461–471. Yasukata, K., Kobayashi, T., 2001. Performance measurement and evaluation systems in view of strategic management control: a survey of management accounting practices in Japan. Asia Pacific Management Review 6 (1), 53–72. Young, S.M., Shields, M.D., Wolf, G., 1988. Manufacturing controls and performance: an experiment. Accounting, Organizations and Society 13 (6), 607–618.