Embed Size (px)
Citation preview
LEAN MANUFACTURING MANAGEMENT: THE RELATIONSHIP
BETWEEN PRACTICE AND FIRM LEVEL FINANCIAL PERFORMANCE
DISSERTATION
Condensed / Summary Version for Survey Participants*
Eric Olsen 30Jun04
Presented in Partial Fulfillment of the Requirements for the Degree Doctor of
Philosophy in the Graduate School of The Ohio State University
By
Eric Oscar Olsen, B.S., M.B.A., M.A.
*****
The Ohio State University 2004
* This version was condensed and summarized by the author from the full dissertation document. Sections of primarily academic and less practical managerial significance were removed. Survey participants are referred to a separate document, “Cross Reference of Company Practice and Performance Values,” for a listing of the values calculated for their own company.
All information in this document is copy righted by Eric Olsen. A complete version of this dissertation will eventually be available through UMI Dissertation Publishing at: http://www.umi.com/umi/dissertations/
ii
ABSTRACT
The relationship between lean manufacturing management practices and business financial performance is examined through the use of empirical surveys and archival accounting data from Compustat and stock return data from CRSP. A sample frame of small to medium sized discrete product and process manufacturing companies reporting participation in only one four-digit SIC was identified as the sample frame. The five-year (1998-2002) financial performance for these companies was analyzed at the operations and business levels using a median z-score comparing median firm performance with the median performance of a matched portfolio of firms. Operations measures included asset and employee productivity, gross margin ratio and two measures of aggregate cycle time. Business measures included return on equity (ROE), sales growth, and stock return. A web-based survey was used to collect data on seven lean practices including just-in-time production management, statistical process control, total productive maintenance, group technology, employee involvement, supplier communication, and customer involvement. Forty-two responding firms were classified as being either lean or non-lean based on a cluster analysis of factor scores. The results demonstrated that lean practices act as a synergistic, mutually supportive set rather than linearly additive individual practices in affecting operations financial performance. Lean classification was associated with better total and cash-to-cash cycle times, but was not related to either better or worse asset or employee productivity. Lean firms also tended to have narrower grow margins than non-lean firms. With respect to business level performance, lean firms tend to have better ROE, but no relationship was found with respect to either stock return or sales growth. Of all the lean practices tested only employee involvement demonstrated a significant relationship to business level performance. Firms with high ROE tend to have high employee involvement. A literature review topology is presented to demonstrate the need for studies combining empirical survey data and archival measures of performance. Opportunities for future research are outlined.
1
CHAPTER 1
INTRODUCTION
1.1 Research Background Both operations management literature and common sense contend that implementing best practices on the factory floor inevitably results in improved business level financial performance. However, empirical research studies fall short in consistently substantiating this relationship with respect to the broad set of operations commonly known as “lean manufacturing” practices. Lean manufacturing may be considered as a synergistic set of integrated modern manufacturing management practices commonly classified under subsets of just-in-time (JIT), total quality management (TQM), total productive maintenance (TPM), and a collection of supportive human resource management practices including teamwork and employee empowerment. Lean manufacturing encompasses such practices as employee involvement in problem solving, statistical process control (SPC), reengineering setups, cellular manufacturing, supplier information sharing and partnership, supply base rationalization, pull production, worker teams, integrated product design, in-house designed technology, and customer requirements integration. The Toyota Motor Corporation offers a preeminent example of a successful lean manufacturer (Bremner & Dawson, 2003; Drickhamer, 2004; Womack, Jones, & Roos, 1991). For 50 years, Toyota has continuously improved its version of lean practice under the banner “Toyota Production System (TPS).” TPS performance positions Toyota to rival DaimlerChrysler, Ford, and General Motors for world auto market domination. A recent BusinessWeek cover article points out that although other auto manufacturers may surpass the auto maker’s individual performance measures, “no car company is as strong as Toyota in so many areas….Its operating profit margin of 8% now dwarfs those of Detroit’s Big Three” (Bremner & Dawson., 2003). Not surprisingly, Toyota’s financial and market success relies largely on lean manufacturing management practices. Nevertheless, the proof that lean works for the broad spectrum of manufacturing firms is specious. Even as practitioners attest that proof exists, studies by both operations management (OM) and finance researchers have proven inconsistent in establishing a significant positive relationship between lean practices and archival business financial performance. In all fairness, most research studies find a positive association with at least one or two financial measures. Reductions in some form of inventory consistently occur in lean implementations. Yet measures of return on assets (ROA), return on sales (ROS), return per employee, and profit margin prove inconsistent. The ambiguity of even “good” empirical research on this question results from three problems. The first relates to the managerial perceptions of performance typically used as the dependent variable in operations management empirical studies. Managers assess
2
one or more of the following: performance relative to competitors, change in performance, internal goal, and absolute performance for specific measures. Surveys asking both practice and performance questions of the same manager run a high risk of encouraging biased responses. Most studies of this nature find a positive relationship between lean practices and perceived financial performance. Managers who are responsible for, or involved in, implementing lean practices may tend to notice only those performance indicators that affirm their belief. An independent source of performance data is required to enhance the validity of these research studies. The second problem occurs in accounting and finance studies of the practice-performance relationship. These studies rely on more objective, archival sources of financial performance data along with rigorous statistical analysis of that data. These studies tend to lack clarity in defining what they consider either a lean or a non-lean firm. They depend on public announcements or other forms of self-identification as indicators of implementing specific practices. Not surprisingly, these studies arrive at mixed results. The third problem concerns the practice of measuring lean practices individually, or in limited subsets, to gauge their relationship to performance. This research maintains that lean practices act as a synergistic whole in affecting operations and financial performance. Therefore, it is not unusual that research studying the individual effects of JIT, TQM, or TPM as subsets of practices often fall short of accounting for significant changes in financial performance. The major contribution of this research is to more definitively assess the practice-performance relationship by using both well-validated practice survey questions (for assessing the actual level of lean implementation) and archival performance data. Standard and Poor’s Compustat1 and the Chicago Graduate School of Business’ Center for Research in Security Prices (CRSP) 2 databases provide independent sources of historical accounting, operations, and business financial performance data. Thus, this research study uses the most applicable techniques from both research streams while overcoming major obstacles that each face separately. The managerial relevance for removing ambiguities from causal connection between practice and performance are undeniable. The prescriptive usefulness of operations
1 “Standard & Poor's Compustat, a division of McGraw-Hill, Inc, is the premier supplier of financial information and produces a variety of databases and software products for institutional investors, financial, and corporate clients…Compustat is a database of financial, statistical, and marketing information providing more than 300 annual and 100 quarterly Income Statement, Balance Sheet, Statement of Cash Flows, and supplemental data items on more than 24,000 publicly held companies.” Wharton Research Data Services; http://wrds.wharton.upenn.edu/home/index.shtml; Oct 2003. 2 “The Center for Research in Security Prices (CRSP) maintains the most comprehensive collection of standard and derived security data available for the NYSE, AMEX and Nasdaq Stock Market. CRSP is a research center at the University of Chicago Graduate School of Business and maintains historical data spanning from December 1925 to the present. CRSP's trademark unique issue identifiers tracks a continuous history of securities and provides a seamless time-series examination of the issue's history.” Wharton Research Data Services; http://wrds.wharton.upenn.edu/home/index.shtml; Oct 2003.
3
management research depends on the resultant ability to advise a manager, “Do this, and your results will improve.” Lean practice implementation that can be shown to lead to financial results that operations managers, accountants, and financial officers all appreciate, can pave the way for more companies to “become lean.” An implied connection exists between the focus of lean on reducing non-value adding activities (i.e. waste reduction) to “make value flow” (Womack & Jones, 1996) and the criteria for good investment and management practice held by value investors, such as Warren Buffet (Graham, 1949; Graham & Dodd, 1934; Vick, 2001). Buffet uses a company’s long-term ability to maintain a high, stable return on equity as a primary indicator of the company’s ability to generate value for its shareholders (Vick, 2001). 1.2 Research Problem This study addresses the general question, “Do lean manufacturing management practices improve financial performance?” This question subsumes several preliminary questions concerning what types and implementation levels of practices constitute “lean practice” and financial performance. As mentioned earlier, this study views lean as a synergistic set of mutually supportive and integrated management practices. The research confirms and measures the relative composition of the lean practice set by identifying lean and non-lean archetypes within a well-defined sample frame of manufacturing companies. Specifically, this research examines whether lean practices as a set, or as individual, specific lean practices--as reflected in survey data--are related to two levels of sustained financial performance (the operations and the business level). At the operations level, Compustat data is used to assess performance by measuring five-year median asset and employee productivity, gross profit margin, and total cycle time. At the business level, return on equity (ROE) and sales growth are measured using Compustat data and stock return performance is assessed using CRSP data. The study tests several research propositions to systematically analyze the question of whether lean practices affect financial performance. These propositions are structured by articulating the problem as a need to understand the dyadic relationships between three ideas: lean practice, operations financial performance, and business financial performance (Figure 1.1). The first to be examined is the relationship between lean practice and either operations financial performance (L-O) or business financial performance (L-B). Lean practice is tested in two ways for each practice-performance relationship: both as a mutually supportive set of practices and as individual practices. Next, the relationship between operations and business financial performance is tested (O-B). This scheme results in the following abbreviated research propositions: Proposition 1: Lean archetypes tend to have better operations financial performance than non-lean archetypes. Proposition 2.1: The extent of lean practice implementation as measured in simple linear combinations is not associated with operations financial performance.
4
Proposition 2.2: The implementation levels of all lean practices are positively associated with all measures of operations performance. Proposition 3: Lean archetypes tend to have better business financial performance than non-lean archetypes. Proposition 4.1: The extent of lean practice implementation as measured in simple linear combinations is not associated with business financial performance. Proposition 4.2: Implementation levels of all lean practices are positively associated with all measures of operations performance. Proposition 5: Categorization as a lean archetype based on operations financial performance is positively associated with business financial performance.
Research Frame
LeanPractices
BusinessFinancial ($)Performance
OperationsFinancial ($)Performance
“L-O”
“O-B”
“L-B”
ManufacturingCompanies
Figure 1.1: Research frame and relationship diagram.
1.3 Research Method The approach taken to study the practice-performance problem combines the best aspects of two research streams. From the operations management empirical research stream the study derives its survey methodology and well-defined, validated multi-item constructs to measure the extent of practice implementation in sample companies. From accounting and finance research, it assumes a rigorous statistical approach to analysis and measurement of financial performance. Combining perceptual practice measurement from surveys with archival financial data enables this research to make substantive claims with respect to its findings’ validity.
5
The research method included first adopting a sample frame of manufacturing companies in four broad industry categories. To enhance the generalizability of the conclusions, the categories range from discrete product to process industries. A strict selection criterion required that the companies report participation in only one four-digit standard industry classification. Along with a size restriction of fewer than 9000 employees, this criterion served to ensure that survey responses were applicable to the performance of the firm as a whole. A high-level company executive for each sample firm completed a survey instrument with 36 questions (Appendix B) covering eight lean practices previously validated in the research literature. The practice data were factor analyzed and used to identify lean and non-lean firms through cluster analysis. Archival financial data from 1998 to 2002 was used to measure long-term performance. A robust, relative measure of financial performance was developed for each metric financial measure used in the study. Measures were developed by identifying a comparison portfolio of companies in the sample frame and calculating the difference between the median 5-year performance for the sample firm and that of its matched comparison portfolio. The relationships between lean and non-lean firms were compared with respect to operations financial and business financial performance using non-parametric statistics and logistic regression. As accounting and financial data is prone to skewness and statistical outliers, non-parametrics were employed to avoid violations to the assumption of normality required in parametric inferential statistics. This research approach offers several advantages. The first is its elimination of common methods bias as a confounding factor. Respondents feel free to describe practice implementation without making performance value judgments since financial performance is measured independently from the practice survey. The second advantage is the availability of a wide range of longitudinal financial performance data through Compustat that allows for a comprehensive financial picture of each sample firm. A broad set of financial ratios can be analyzed and divided into more specific measures to better understand underlying mechanisms. The availability of a non-survey source of performance measures also allows the survey to be shorter and ostensibly reduces respondent fatigue and increases the response rate from busy executives. The use of Compustat represents a significant opportunity for future research into the lean practice-performance relationship. Longitudinal data is available on over 24,000 firms. Specifically as regards the current sample companies in this research, follow-up studies can examine future performance based on annually updated archival data without the necessity of a future survey. Implementation timing and sequence issues can be examined in a follow-up study and provide answers to questions such as, “What is the best practice set for sustainability of results?” In addition, the opportunity is presented to expand lean performance research into other manufacturing and service industries using the techniques developed in this study.
6
1.4 Importance and Contribution of Research Aside from establishing a more definitive relationship between lean practices and archival financial performance, several lesser, but nonetheless unique, contributions of this study are notable from both a content and research methods perspective:
1. Previous empirical OM practice survey studies have not looked at business financial performance in the form of return on equity, sales growth, and stock return. This is the first study to make the connection from broadly measured lean practice constructs to these measures of business level financial performance.
2. A contribution to methods in OM research is the use of Compustat’s standard
industry classification code report to identify companies that participate in only one four-digit SIC code. This approach provides an alternative to the plant-level data restriction imposed by previous OM practice studies. Being able to make the claim that practice survey data applies to all the manufacturing operations of a company, the lean practice effects can then be traced through a company’s financial records.
3. The development of a normalized measure based on the median performance of
comparison portfolios of companies is a major contribution to this study. It successfully controlled for both industry and size effects that would have limited the ability to identify practice effects in a relatively small sample size. This technique in combination with the single SIC code identification feature in Compustat provides a vehicle for more research into the company-level effects of OM practices.
4. This study utilizes cycle time as a key metric in evaluating lean performance. It
offers two aggregate measures of cycle time (total and cash-to-cash) that view cycle time beyond the bounds of inventory, the metric traditionally imposed in OM research. Accounts payable, accounts receivable, and inventory are combined to provide a companywide, cross-functional view of cycle time performance.
5. Finally, this study uses an interesting mechanism to improve survey response rate.
Potential respondents were provided with a relative ranking of their position in the sample frame based on a preliminary analysis of the Compustat data. Letters to the respondents informed them of their company’s ranks with respect to several individual and summated operations and business measures. The respondents were encouraged to respond in order to help the researcher “understand how [their] manufacturing and operation practices contribute to [their] performance.” An added incentive allowed respondents to view rankings for all potential respondents in the sample frame by visiting the website and completing a survey. This mechanism created a unique “curiosity hook” that may be helpful in improving the response rate in future OM survey research. Such incentives may
7
also move some of the practitioner value derived from academic research to the front of the process, thereby improving the partnership between researchers and businesses.
1.5 Organization of the Dissertation The remainder of the dissertation follows in six chapters. Chapter 2 reviews recent OM and financial and accounting studies relevant to the practice-performance question. Chapter 3 develops a series of research propositions based on the literature. Chapter 4 describes the research methods used to test and analyze the positions. Chapter 5 discusses the results and potential threats to their validity and generalizability. Chapter 6 presents conclusions and offers directions for future research into the relationship between lean manufacturing management practices and financial performance.
8
CHAPTER 2
LITERATURE REVIEW
2.0 Objectives and Map of Domain This study aims to capture and quantify the relationship between lean manufacturing management practices and financial performance at both the operations and business levels. The literature review examines how past studies addressed the practice-performance connection, and will show that:
1. The practice-performance connection has not yet been sufficiently established, especially regarding financial performance.
2. Recent operations management (OM) studies have developed a reliable and valid set of constructs for measuring lean practices.
3. Analysis methods used in financial and accounting literature, along with the practice constructs developed in the OM studies, offer a means for more accurately describing and testing the practice-performance connection.
The relevant domain for empirical research into the practice-performance relationship can be classified into three general categories, which, broadly considered, comprise the three relevant quadrants of a two-by-two matrix (Figure 2.1). This study uses the matrix axes’ to specify lean practice usage (vertical axis) and the source of dependent performance variable data (horizontal axis). Either studies are categorized as using empirical surveys or some form of public announcement as a determinant of lean practice usage and are further categorized according to the performance variable’s data source (i.e. survey-based manager perceptions or archival data sources). Other aspects of each study, such as unit and method of analysis, correlate with their corresponding matrix categories. The first category (survey-perceptual) includes cross-sectional, survey-based studies that describe, categorize, and measure multiple manufacturing management practices. They include an analysis of the relationship between practices and between practice and operations performance as measured by managers’ perceptions of performance. This is typically measured relative to competitors’ performance or, internally, to the company’s own past performance or established goals. Having evolved with the development of empirical research in OM over the last ten years, these studies use multi-item constructs that have demonstrated high validity and reliability in capturing and describing management practices. However, performance variables in survey-perceptual literature, which lack definitive connection to archival forms of accounting and financial data, comprise the category’s major shortcoming. The second literature category (announcement-archival) includes studies that use archival data to measure operations or business financial performance. Typically, these studies use public announcement data in the form of press releases or annual reports to designate practice implementation dates or a survey that asks respondents to self-classify their
9
company as either practice or non-practice. This category includes accounting and financial studies of companies that have been identified as having implemented some subset of lean management practices, usually just-in-time (JIT) or total quality management (TQM). None of the announcement-archival studies either explicitly or comprehensively examines lean practices. Yet archival data studies can be quite sophisticated in using event studies and robust statistical analysis to identify changes in operations and business performance attributable to practice implementation. The deficiency in Category 2 studies is their often high reliance on public announcements of practice implementation without adequate verification. Single management practices go by many names and vary in definition and degree of implementation as well as exclusivity. Howton, Higgins, and Biggart (2000) alone use six different terms to identify companies implementing JIT. Claiming a practice “has been implemented” is not equivalent to measuring the extent of its practice usage. Descriptive studies in the OM literature (Category 1) testify to this lack of uniformity in mix and extent of practice implementation. The third, and final literature category (survey-archival), hybridizes the first two. These studies used a survey instrument to collect practice information from subject companies along with archival sources of performance data to measure the practice-performance relationship. These studies’ main deficiency results from the limited capability of their survey instruments to reliably capture a full range of lean practices. In addition, none of the Category 3 studies examines the relationship of practices to business level performance such as return on equity (ROE), earnings per share, or stock return. Studies such as Upton (1998), which use perceived performance measures and announcement-based indicators of practice usage, are less persuasive and are considered irrelevant to this literature review. Table 2.1 lists the literature review’s primary studies and the categories into which they fall. The table uses four broad practice areas that are typically included in lean operations management, namely just-in-time (JIT), total quality management (TQM), total productive maintenance (TPM), and infrastructure or common practices. Practices described in the articles cited are placed in appropriate corresponding lean practice areas. For example, setup time reduction is classified as a JIT practice, and customer focus or involvement is classified as being primarily TQM. The practice areas are not necessarily exclusive. A major contention in the literature is that each lean practice, being synergistic and mutually supportive, is difficult to isolate in its resulting impact on performance (Shah, 2002; Ward, Bickford, & Leong, 1996). Flynn, Sakakibara, and Schroeder (1995) used the concept of “infrastructure” or “common” practices for both JIT and TQM to circumvent the problem of practice classification while recognizing the interdependencies of lean practices. Table 2.1 depicts the high coverage of lean practice topics in survey-perceptual studies that do not use corresponding archival data to analyze financial performance. Announcement-archival studies, measuring financial operations and business performance with archival data, prove deficient in the limited, binomial nature of the practice usage designation. To explore the breadth of studies available, Table 2.1 purposefully uses minimal criteria for practice coverage. The studies by
10
Fullerton and McWatters (2001) and Fullerton, McWatters, and Fawson (2003) ask only ten survey questions to capture the extent of implementation across all lean practice areas. In contrast, Cua, McKone, and Schroeder (2001) use 17 factors derived from a survey containing 69 questions. Table 2.1 further indicates whether a study’s analysis set controls for the firm’s industry and size, both of which may significantly influence actual performance levels. Table 2.2 summarizes other relevant features of studies in this literature review including unit of analysis, sample size, industry coverage, primary analysis method, survey type, and time span of the data analysis. The survey type column differentiates studies that examine performance over time (longitudinal) from cross-sectional studies. The use of longitudinal studies coincides with availability of archival data and is restricted to Category 2 and 3 studies. The following sections describe each literature category in turn and focus on areas where this research study leverages or improves upon past research.
Perceptual Archival
Category 1: Survey-Perceptual Category 3: Survey-Archival
Shah & Ward (2003)Shah (2002)
Cua, et al (2001)Fullerton & McWatters (2001) Fullerton, et al (2003)
Claycomb et al (1999) Callen, et al (2000)Samson & Terziovski (1999) MacDuffie (1995)
Koufteros et al (1998)Sakakkibara, et al (1997)
Flynn, et al (1995)Inman & Mehra (1993)
Category 2: Announcement-Archival
Biggart & Gareya (2002)Boyd, et al (2002)
Kinney & Wempe (2002)[Not relevant] Howton, et al (2000)
Easton & Jarrell (1998)Balakrishnan, et al (1996)
Hendricks & Singhal (1996)Chang & Lee (1995)
Hudson & Nanda (1995)Billesbach & Hayen (1994)
Source of Performance (dependent) Variables
Sour
ce o
f Lea
n Pr
actic
e (in
depe
nden
t) Va
riabl
es
Survey
Announcement
Figure 2.1: Practice-performance literature domain typology and categories.
Operations Performance
Business Performance Industry Size JIT TQM TPM
Infra-struc- ture
Shah 2002 X X X X XCua, et al 2001 X X X X X XFullerton & McWatters 2001 X X X XClaycomb et al 1999 X XSamson & Terziovski 1999 X XKoufteros et al 1998 X X X XSakakkibara, et al 1997 X X X XFlynn, et al 1995 X X XInman & Mehra 1993 X
Biggart & Gareya 2002 X X XBoyd, et al 2002 X X XKinney & Wempe 2002 X X X XHowton, et al 2000 X XEaston & Jarrell 1998 X X X X XBalakrishnan, et al 1996 X X X XHendricks & Singhal 1996 X XChang & Lee 1995 X X X XHudson & Nanda 1995 X X XBillesbach & Hayen 1994 X X
Fullerton, et al 2003 X X X X X XCallen, et al 2000 X X X XMacDuffie 1995 X X X
Notes: Cells with X designate that the column topic is addressed at least to a minimal degree.
Category 3: Studies with limited practice scope and archival measures of operations and business financial performance
Category 2: Studies with public or self-identified nominal practice usage and archival measures of financial performance
Article Date
Archival Performance Data Controls
Category 1: Studies with multiple lean practices and non-archival measures of performance
Lean Practice Coverage Areas
Table 2.1: Practice and performance coverage of key studies.
19
Unit of analysis
Sample size Industry coverage Primary analysis method
Survey type (1) Time span (2)
Category 1: Studies with multiple lean practices and non-archival measures of performanceShah 2002 Plant 271 Mfg SIC's 34, 35, 36, 37, 38 Structural equation modeling C 0Cua, et al 2001 Plant 163 Electronics, machinery, transportation parts Dircriminant analysis C 0Fullerton & McWatters 2001 Firm 91 All mfg ANOVA C 0Claycomb et al 1999 Firm 200 All mfg Structural equation modeling C 3 yearsSamson & Terziovski 1999 Mfg site 1024 All mfg Multiple regression C 0Koufteros et al 1998 Plant 244 Mfg SIC's 34, 35, 36, 37 Structural equation modeling C 0Sakakkibara, et al 1997 Plant 41 Electronics, machinery, transportation parts Canonical correlation C 0Flynn, et al 1995 Plant 42 Electronics, machinery, transportation parts Hierarchical regression C 0Inman & Mehra 1993 Plant 114 All mfg Correlation C 0Category 2: Studies with public or self-identified nominal practice usage and archival measures of financial performanceBiggart & Gareya 2002 Firm 74 All mfg ANOVA L -3,-2 vs 2,3Boyd, et al 2002 Firm 11 Mfg Bivariate regression L -3 to 7Kinney & Wempe 2002 Firm 201 Mfg, wholesale, retail Wilcoxon ranked sign test L -3,-2, -1 vs 1,2,3Howton, et al 2000 Firm 97 All mfg Event study - t-test L -200 to -11 vs 0 daysEaston & Jarrell 1998 Firm 108 Mfg and service Wilcoxon ranked sum test C 5 yearsBalakrishnan, et al 1996 Firm 46 All mfg ANOVA L -3,-2, -1 vs 1,2,3Hendricks & Singhal 1996 Firm 91 All industries Event study - Wilcoxon L -200 to -11 vs 0 daysChang & Lee 1995 Firm 44 All mfg MANOVA L -3 vs +3Hudson & Nanda 1995 Firm 55 SIC's 22 - 59 Path model L -4 to +4Billesbach & Hayen 1994 Firm 28 All mfg Wilcoxon ranked sign test L -5,-6 vs 5,6Category 3: Studies with limited practice scope and archival measures of operations and business financial performanceFullerton, et al 2003 Firm 212 All mfg Hierarchical regression C 0Callen, et al 2000 Plant 100 Automotive parts and electronics Multiple regression C 0MacDuffie 1995 Plant 70 Automotive Multiple regression C 0
Note: 1. C = cross-sectional, L = longitudinal2. Stated in years unless otherwise specified. Zero or aggregate time period reported for cross-sectional studies. Longitudinal studies report comparison time periods for adoption period equal to zero.
Table 2.2: Other relevant features of practice performance studies.
20
28
CHAPTER 3
RESEARCH PROPOSITIONS 3.0 Research Propositions - Objectives This chapter introduces the specific research questions addressed in this study. The questions are structured in the form of five general propositions at two levels. The first level consists of the general relationship between three ideas: lean manufacturing management practices, operations financial performance, and business financial performance. The second level consists of the relationships between specific lean practices and specific measures of financial performance. Figure 3.1 illustrates the structure of the research propositions in the form of a diagram. The concept of lean manufacturing management is captured in two ways—either as a categorical strategic archetype or as individual practices. The first case contends that lean practices tend to act in mutually supportive sets of interrelated practices rather than as standalone mechanisms or practices in simple linear combinations. The lean practices included at the outset of the study are listed in Table 2.4 and consist of five internally oriented practices and three externally oriented lean practices. The remainder of the chapter develops five general propositions about the relationships between lean manufacturing management practices and financial performance as follows:
1. Lean archetype versus operations financial performance (P1), 2. Individual lean practices versus operations financial performance (P2), 3. Lean archetype versus business financial performance (P3), 4. Individual lean practices versus business financial performance (P4), and 5. Operations financial performance versus business financial performance assuming
that lean archetypes have specific relationships between operations financial performance measures that can be used to identify lean firms (P5).
Theoretical and empirical support for each proposition to justify its inclusion in this study is provided.
29
Research Proposition Diagram
Business Performance
ROESales GrowthStock Return
Lean
ArchetypeIndividual Practices
Operations Performance
Asset ProductivityEmployee
ProductivityGross Margin
RatioCash-to-Cash Cycle TimeTotal Cycle
Time
P1P2
P4
P3
P5
Figure 3.1: Diagram of research propositions.
30
CHAPTER 4
METHODS 4.0 Methods - Objectives The research design consists of surveying a select set of manufacturing firms, for which long-term archival accounting data is available, on their use of lean manufacturing management practices. To establish the relationship between lean manufacturing management practices and financial performance, it is necessary to create a research design that effectively captures both variables. The design used in this study combines an analysis of archival financial performance data with an empirical survey of practice data from a sample frame of manufacturing companies. The chapter is divided into four sections. The first section describes the creation of the sample frame. The second describes the formation of a set of financial performance variables. The third describes how the lean practice variables were created and how the companies were categorized into lean and non-lean firms. The final section describes the statistical methods used to compare lean and non-lean companies and analyze the effect of individual practices on financial performance. Figure 4.1 provides a high level block diagram of the research design and Table 4.1 provides a detailed overview of the research methods applied in this study.
Sample Frame Firms(N = 316)
Survey Respondent Firms(n = 42)
Lean Practice Implementation Analysis
Lean/Non-leanClassification
Financial Performance:Operations and Business
Practice/Performance-------------------------
Classification/PerformanceAnalysis
Results
Figure 4.1: Block diagram of research design.
31
Research Methods Overview
Section Title Highlights 4.0.0 Methods - Objectives Overview of chapter structure. 4.1.0 Sample Frame - Objectives • Overview of frame selection criteria
• Descriptive data for: o Total assets o Number of employees o Total inventory o Net sales
4.1.1 Availability of Archival Financial Data Data sources include:
• Compustat • CRSP
4.1.2 Business Scope Reported participation in only one 4-digit SIC 4.1.3 Size Restriction 50-9,000 employees 4.1.4 Industry Group Manufactures (SIC):
• 28: Chemicals • 35: Industrial, Computer Equipment • 36: Electronic and Components • 38: Measuring, Control Instruments; Medical
4.1.5 Years of Data Available • Reported for 11 years
• Performance based on last 5 years (1998-2002) 4.2.0 Financial Performance Variables - Objectives Criteria for selection:
• Measure sustained performance • Publicly available • Valid and reliable • Recognizable to operations managers • Comparable
4.2.1 Business Financial Performance Variables Variables selected for business financial performance:
• ROE • Sales Growth • Stock Return
4.2.2 Operations Financial Performance Variables Variables selected for operations financial performance:
• Asset productivity • Employee productivity • Gross margin ratio • Cycle time (several measures)
4.2.3 Financial Performance Median z-score Calculation • Comparison portfolio selection criteria:
o Industry match (2-digit SIC) o Size match (+/- 50% of total asset size)
32
• Median z-score o Formulation o Tests for industry and size effect control
effectiveness 4.3.0 Lean Manufacturing Practice Variables - Objectives Constructs included:
Internally oriented lean practices • Just in time production methods (JIT) • SPC tools to monitor quality (SPC) • Employee involvement (EMP) • Group technology to enhance the flow of products (GT) • Total productive maintenance (TPM)
Externally oriented lean practices • Communication with suppliers (SCM) • Customer involvement (CUS)
4.3.1 Survey Instrument • 36 items covering 8 practice areas
• 5-point Likert scales 4.3.2 Data Collection • Target high-level operations managers
• Web-based survey • Contacts by mail and phone • “Performance ranking” provided as incentive
4.3.3 Survey Sample Validation Frame and sample comparison for:
• Representativeness • Response bias
4.3.4 Practice Construct Formulation • Factor analysis
• Construct validity and reliability • Factor score measurement
4.3.5 Lean Archetype Formation • Cluster analysis to form lean and non-lean clusters
• Validation through: o Rationalizing practice levels within clusters o Discriminant analysis o Predictive validity with respect to operations
financial performance measures 4.4.0 Relationship Analysis – Objective: Describe how the relationships between lean practice variables
and financial performance were analyzed. 4.4.1 Lean Archetypes and Operations Financial Performance • Wilcoxon rank-sum test comparing lean and non-lean
archetypes with respect to performance. • Logistic regression of lean and non-lean archetype
classification on a set of performance variables 4.4.2 Individual Lean Practices and Operations Financial Performance • Logistic regressions for each performance variables as
above or below industry median performance on all practice measures
4.4.3 Lean Archetypes and Business Financial Performance • Wilcoxon rank-sum test comparing lean and non-lean
33
archetypes with respect to performance. • Logistic regression of lean and non-lean archetype
classification on a subset of performance variables 4.4.4 Individual Lean Practices and Business Financial Performance • Two sample t-test comparing above and below industry
median performers on their level of practice implementation • Logistic regressions for each performance variables as
above or below industry median performance on all practice measures
4.4.5 Lean Operations and Business Financial Performance • Logit models developed in the survey sample are used to
predict lean and non-lean classification in non-respondents in the sample frame based on operations financial performance measures.
• Business financial performance of classified companies is compared using Wilcoxon rank-sum tests.
4.5 Summary The study’s research method is summarized.
Table 4.1: Research methods overview and highlights.
4.2 Financial Performance Variables 4.2.0 Objectives This study examines the effect of lean manufacturing practices on two levels of financial performance: operations and business. Variables at the operations financial level tend to be influenced directly by the operation’s function. Operational variables used in this study include asset productivity, employee productivity, gross margin ratio, and cycle time. Business financial variables are not directly attributable to any specific function. Return on equity, sales growth, and stock return are the business financial variables used. Table 4.3 provides a listing of the study’s financial performance measures. One criterion for selecting financial variables maintains that the variables must measure sustained performance and not be unduly influenced by abnormal changes in a particular year’s performance. To address this criterion the five-year median performance for each financial variable is used. 3 The use of a five-year performance window (1998-2002) and its median, assures less bias by outlier data than the five-year mean performance. This increased confidence that the performance being explained is sustainable. Several general objectives guide the selection of financial variables. The variables must be readily available. Public sources of information, such as those supplied by Compustat and CRSP databases were used exclusively (Compustat, 2003; CRSP, 2003). The use of
3 The value of using the median as the primary indicator of central tendency in financial data analysis is illustrated by the following story: “A local bar is populated with a dozen democrats and a dozen republicans. Bill Gates, Chairman of Microsoft, walks in the door and the democrats think that in general they are no better off (median net worth has changed only slightly). The republicans think that on average they are all multi-millionaires (mean net worth has changed dramatically).” – Attribution unknown.
34
public data sources minimized the problem of missing data that is usually encountered when one requests sensitive information from companies. Since the data is public, the financial elements of the study can be replicated and verified. The Compustat and CRSP databases, although not perfect, have been thoroughly researched and tested as tools for conducting financial research on companies. A drawback of using an existing financial database is that the availability of component data constrains variable formulation. Another standard in the selection of financial variables holds that they be common and understandable to operations managers. This research does not aspire to extend the art of financial analysis; rather, value to practitioners is a key consideration. This study also could serve as a baseline for future research looking to improve financial measures. That financial variables will be more or less effective in predicting or reflecting the effects of certain lean manufacturing management practices is to be expected. Neither does this study propose an optimum set of measures. An effort was made to be comprehensive yet parsimonious in the selection of financial variables. The archival studies in the literature review chapter cover over twenty-eight distinct financial measures. This study focuses on eight measures at the operational level, including a set of five distinct, yet related, cycle time measures, and three business level measures for a total of eleven financial measures. A final objective in creating the set of financial variables is comparability. Comparability between and among companies and industries is a critical requirement. This issue is addressed at two levels. First, ratios of absolute performance were created by applying a scaling factor. For example, equity book value, number of employees, and cost of goods sold are used as scaling factors for income before extraordinary expenses (ROE), operating income (employee productivity), and total inventory (inventory cycle time), respectively. Secondly, individual company performance is compared to that of a matched portfolio of similar companies. Matching is based on a minimum of a match on 2-digit SIC code and on total asset size in 2001 to with +/- 50 %. Several researchers assert that industry and size are two of the most influential determinants of firm financial performance (Barber et al., 1996; Dess, Ireland, & Hitt, 1990; Mauri & Michaels, 1998; Moch, 1976). A “median z-score” is created for each company financial measure providing a measure of company performance with respect to the median value of the comparison portfolio4.
4 The term “median z-score” is used to indicate a measure that is calculated based on a value’s difference from the median and the sum of squared defenses from the median. Normal z-score calculations use the mean. See Sec. 4.3.3 for details.
35
Return on Equity ROESales Growth SGStock Return SR
Return on Cash Adjusted Assets ROCAEmployee Productivity EPGross Margin Ratio GMRCash to Cash Cycle time CTCTotal Cycle time CTTInventory Cycle time CTIPayables Cycle time CTPReceivables Cycle time CTR
Business
Operational
Table 4.3: Financial performance measures.
4.2.1 Business Financial Performance Variables Three overall measures of business financial performance are utilized in this study: return on equity (ROE), sales growth (SG), and stock return (SR). To ensure sustained performance, five years of data from 1998 to 2002 was utilized. Two different measures of central tendency are applied to the business financial performance measures (the median is used for ROE and SG and the mean for SR). Applicable sections explain the reasons for selecting different measures. Compustat and CRSP databases supplied the raw financial data.5 4.2.1.1 Return on Equity (ROE) Return on equity is a well-accepted measure of overall business performance. ROE is given a high level of credibility by value investors such as Warren Buffet because of its comprehensive nature and companies’ inability to manipulate its value (Vick, 2001). ROE measures a firm’s ability to generate profits in comparison to the historical book value of its assets and is computed as income before extraordinary items divided by the book value of common equity. The average equity for the year is assumed the most relevant with respect to evaluating current year returns. Common equity is essentially assets minus liabilities. In equation form:
( )2
606018
1 tt DDD
EquityCommonofValueBookItemsaryExtraordinBeforeIncome
+=
−
Variables beginning with “D” stand for data items available in Compustat. A complete listing of data items and their descriptions is included in Table 4.4.
5 Individual company years with missing or zero values in the financial ratios numerator or denominator were excluded from the analysis in all cases.
36
D1 = Cash and Short-Term Investments (MM$) D3 = Inventories, Total (MM$) D6 = Assets Total (MM$) D12 = Sales, Net (MM$) D13 = Operating Income Before Depreciation (MM$) D18 = Income Before Extraordinary Items (MM$) D29 = Employees (M) D41 = Cost of Goods Sold (MM$) D60 = Common Equity, Total (MM$) D70 = Accounts Payable (MM$) D151 = Receivables, Trade (MM$)
Table 4.4: Reference list of Compustat data descriptions (Compustat, 2003)6
4.2.1.2 Sales Growth (SG) Sales growth measures a firm’s performance with respect to increasing its revenues over the previous year’s. Because a key criterion in valuing stock is a company’s ability to generate ever-increasing returns for its investors, growth is an imperative for most traded businesses. By looking at sales growth in relation to a matched set of comparison companies (e.g. “median z-score”), the sales growth measure provides a sense of how well the company is taking market share from its competitors. If its growth rate is greater than the median value for the comparison companies then a firm is ostensibly gaining market share. Sales growth is computed as the change in sales from the previous year divided by sales in the previous year. In equation form:
1
1
1 121212
−
−
−
−=
t
tt
t DDD
SalesSalesinChange
4.2.1.3 Stock Return (SR) Stock return is a market-based measure indicating the premium that the market places on the potential earnings stream for a company. The average annual return for a given company is calculated as the arithmetic sum of the monthly returns adjusted for dividends and splits for the five-year period (60 months) divided by five. The CRSP database provides the adjusted monthly return value. RET is the “Holding Period Total Return.” “Returns are holding period returns from month-end to month-end, not compounded from daily returns, and ordinary dividends are reinvested at month-end” (CRSP, 2003). Since the return is reset and calculated monthly, indexing to a specific time point is not necessary and the arithmetic mean is an appropriate alternative to the geometric mean that is often used in stock return calculations.
6 M=1000; MM=1,000,000.
37
Since SR is a market-based measure, it is assumed that trend and variability, or risk, is included. In equation form:
55
60
1
60
1∑∑== = mm
RETReturnStockMonthly
A complete explanation of the algorithm CRSP uses for calculating the monthly return values (RET) is provided on the database website)7. With respect to the stock return data, two methodological issues are noteworthy. The first is that SR stands as the only financial measure in this study to use the mean, rather than the median, as its measure of central tendency. Since 60 stock return data points were available for each firm, as opposed to only five annual data points, it may be assumed that a stock return average value is less susceptible to outliers. In addition, researchers have found that stock return data distributions generally do not suffer as much from heavy tails (outliers) as other financial measures (Barber et al., 1996). The second issue is that SR uses calendar year data as opposed to the other financial measures that use fiscal year data. Over 50% of the companies in Compustat have December close dates, making the distinction less relevant. It was deemed more relevant to match stock performance timing across companies and business cycles than to match the fiscal accounting calendar year8.
7 “Holding Period Return Variable Name = RET A return is the change in the total value of an investment in a common stock over some period of time per dollar of initial investment. RET(I) is the return for a sale on day I. It is based on a purchase on the most recent time previous to I when the security had a valid price. Usually, this time is I - 1. Returns are calculated as follows: For time t (a holding period), let:
t’ = time of last available price < t r(t) = return on purchase at t’, sale at t p(t) = last sale price or closing bid/ask average at time t d(t) = cash adjustment for t f(t) = price adjustment factor for t p(t’) = last sale price or closing bid/ask average at time of last available price < t.
then r(t) = [(p(t)f(t)+d(t))/p(t')]-1 t’ is usually one period before t, but t’ can be up to ten periods before t if there are no valid prices2 in the interval. A series of special return codes specify the reason a return is missing.” (CRSP, 2003) 8 This highlights the fact that the fiscal year data used for the other measures does not match. A tradeoff is necessary between data objectivity and time matching. Estimating non-fiscal year-end data raises issues of
38
4.2.2 Operations Financial Performance Variables Compustat accounting data is used to assess operations financial performance with respect to four areas: Asset Productivity, Employee Productivity, Gross Margin Ratio, and Cycle Time. 4.2.2.1 Asset Productivity – Return on Cash-Adjusted Assets (ROCA) Return on cash-adjusted assets measures the return on book value of assets adjusted for cash balances and short term investments such as marketable securities. According to one study, “The most important adjustment to total assets can be the deduction of cash and marketable securities from the book value of total assets” (Barber et al., 1996). Some cash is necessary for operations, but much of the variability in total assets results from financing activities such as issuing of securities (Barber & Lyon, 1996). By removing cash, this study focuses on the productivity of physical assets such as equipment and machinery. The assets deemed most relevant are those held during the year in which the income is generated. Therefore, the average asset value is used as the scaling factor. In equation form:
2)]16()16[(
1311 tttt DDDDD
AssetsofValueBookAdjustedCashAverageiation ore DeprecIncome BefOperating
−+−=
−−
4.2.2.2 Employee Productivity (EP) Operating income per employee measures productivity on a per person basis. Since operating income is revenue less operating expense it measures the value-added per employee. Value enhancement and waste elimination are primary objectives of lean manufacturing practices (Monden, 1983; Womack et al., 1991). Value addresses both the revenue and the cost aspect of the operation. Most of the archival studies cited in the literature review that address employee productivity use sales revenue as the numerator in a ratio to the number of employees. Sales figures are more likely to be publicly available than operating income. This results in a focus only on the revenue enhancing contributions of employees (Boyd et al., 2002; Chang & Lee, 1995; Huson & Nanda, 1995). Only Easton and Jarrell (1998) incorporated employee productivity in terms of operating income in researching the effects of TQM on financial performance. In equation form, employee productivity in the current study is expressed as:
2)2929(
131 tt DDD
EmployeesofNumberTotalAverageiation ore DeprecIncome BefOperating
+=
−
data availability and assumptions. A strong argument can be made that by using 5-year median and mean values the impact of any time shift is reduced.
39
4.2.2.3 Gross Margin Ratio (GMR) The current study defines GMR as the ratio of the cost of goods sold to the selling price. Lean manufacturing practices should act to reduce the cost of goods sold by eliminating waste and improve sales revenues by improving delivery and quality. Both the cost of goods sold and sales are income statement items taken from the same corresponding year. In equation form:
1241
DD
Salesods SoldCost of Go
=
Note that because the ratio is calculated as the cost of goods sold divided by sales, the measure has the typical characteristic of being a maximum of one9 (i.e. CGS=Sales), decreasing with reductions in cost or increases in sales. Lower values are better than higher ones for this measure. 4.2.2.4 Cycle Time Time, in addition to quality and cost, is a focus of lean manufacturing. In order to understand the role that time plays in the practice-performance relationship, this study included several measures of cycle time as performance variables. Operations cycle time is defined as the total time to complete a transaction from the time it enters a system to the time it exits the system. Cycle time includes all waiting, processing, and transfer time. This study is concerned with two aggregate measures of cycle time and their component measures. The two aggregate measures are total cycle time (CTT) and cash-to-cash cycle time (CTC). The component parts that make up both of these aggregate measures are accounts payable cycle time (CTP), inventory cycle time (CTI) and accounts receivable cycle time (CTR). The relationship between these cycle times is depicted graphically in Figure 4.2. All of these measures and their formulations are explained in the following sections. Included cycle time measures use Little’s Law10 (Anupindi, Chopra, Deshmukh, Mieghem, & Zemel, 1999), to convert accounting information on the quantity of transactions in a system to cycle time by dividing by the transaction throughput rate. For example, inventory cycle time is calculated as average annual inventory dollars divided by the cost of goods sold in a year. Cycle time is used rather than its inverse, transaction
9 It is possible that a company could have a GMR of greater than one (i.e. selling at a loss) for a period, especially if large amounts of overhead are allocated. The use of a 5-year median value and the elimination of start-up phenomena by making sure the companies have at least 11 years of reported data minimized the occurrence of this event. However, it is surprising to note that 35 of the 316 sample frame companies (11.1%) had median GMR’s greater than one for the 5 year period studied. The actual maximum value for all the sample frame companies was 9.07. Most of the companies with GMR’s over one were in the chemical and allied products industry (33/35). 10 Little’s Law: CT = WIP/R; where CT = cycle time in units of time, WIP = work in process in units, and R = throughput rate in units/time.
40
turns11, because of its additive properties. Cycle times can be summed or subtracted to create meaningful aggregate values whereas turns cannot.
Cash-to-Cash Cycle Time
Inventory (CTI)Accts
Receivable (CTR)
Accts Payable (CTP)
CTC= CTI + CTR - CTP
Total Cycle TimeCTS = CTI + CTR + CTP
Figure 4.2: Relationship between cycle time measures used in this study.
4.2.2.4.1 Cash-to-Cash Cycle Time (CTC) CTC measures the average time between a company’s payment to vendors and money collection from customers. Proponents of lean manufacturing hold that a significant benefit is the improvement in cash availability made possible by inventory and scrap rate reductions (Hall, 1983; Heard, 1995). To calculate CTC, accounts receivable cycle time (CTR) and inventory cycle time (CTI) are added together and accounts payable cycle time (CTP) is subtracted. This can be expressed in equation form as:
CTC = CTI + CTR - CTP 4.2.2.4.2 Total Cycle Time (CTT) CTT uses the same components as cash-to-cash cycle time but it sums all three elements, expressed in equation form as:
CTT = CTI + CTR + CTP
CTT differentiates from CTC in that it assumes as its objective the reduction of the aggregate transaction processing time rather than the time between vendor payout and receipt of customer funds. 11 Example: Inventory turns = Cost of Goods Sold / Average Inventory.
41
4.2.2.4.3 Inventory Cycle Time (CTI) CTI accounts for the average time goods spend in the manufacturing process from raw materials to finished goods. CTI is calculated as the average of the current and the previous year-end inventories divided by the cost of goods sold. It is the reciprocal of inventory turns per year, expressed in equation form as:
t
tt
D
DD
ods SoldCost of GoInventoryTotalAverage
412
)33( 1 +=
−
4.2.2.4.4 Accounts Receivable Cycle Time (CTR) CTR is the average time it takes the company to be paid. CTR is the average time between product shipment and payment receipt from the customer and is calculated as average of the current and previous year-end accounts receivables values divided by sales for that year. Use of trade receivables limits the value to receipts generated from sale of goods and services in the ordinary course of business (Compustat, 2003). Thus, accounts receivable cycle time can be thought of as the reciprocal of trade receivable turns expressed in equation form as:
t
tt
D
DD
SalesceivablesReTradeAverage
122
)151151( 1 +=
−
4.2.2.4.5 Accounts Payable Cycle Time (CTP) CTP is the average time it takes the company to pay its vendors, or the average time between the receipt of raw materials and vendor payment. CTR is calculated as average of the current and previous year-end accounts payable values divided by cost of goods sold for that year and can be thought of as the reciprocal of payables turns expressed in equation form as:
t
tt
D
DD
ods SoldCost of GoPayableAccountsAverage
412
)7070( 1 +=
−
Note that CTP is the one financial performance measure in this study whose direction of improvement is open to debate. A longer, or higher, value for CTP is interpreted as a longer time to pay vendors. This is a positive attribute when applied in the cash-to-cash cycle time calculation where CTP is subtracted to reduce CTC. It is a negative attribute when calculating total cycle time where CTP is summed to CTT. The implication of the CTP’s dual characteristic is applied later in the analysis to distinguish between the effects of pursuing cash-to-cash cycle time reduction or total cycle time reduction.
42
4.2.3 Financial Performance Median z-score Calculation The median value for each financial measure described above, except stock return, which utilizes the mean, was calculated for each sample frame company for the five-year period 1998 to 2002. Although the measures are scaled ratios, controlling for industry and firm size further enhances the comparability of the performance between companies. Industry and size are two obvious factors affecting financial performance. In order to control for these factors, while creating measures that can be compared across all 316 sample frame companies, a “median z-score” (zp), was calculated for each company financial performance variable. zp essentially is a statistical normalization using the portfolio median value rather than the mean. To calculate zp a comparison portfolio of companies is first identified for each company based on a 2-digit SIC code and size match to within +/- 50% of its total asset value (Barber et al., 1996). Easton and Jarrell (1998) use a similar technique comparing firm performance against a portfolio of three companies matched for industry and size “to the extent possible.” Other archival studies use a single company match. Balakrishnan et al. (1996) and Kinney & Wempe (2002) use a 3-digit SIC match and select a Compustat-listed firm closest in net sales. A minimum of five companies was included in each portfolio12 to reduce the effect of individual portfolio companies on the comparison value. A z-score was then calculated for each measure based on the difference between each sample frame company’s five-year median measure of performance and the median measure of its distinct set of portfolio companies. Expressed in equation form for a given financial performance measure and company:
( )
11
2
−
−
−=
∑=
p
n
kk
P
n
PmdP
PmdPzp
Where: P = five-year median performance for a given measure and company, Pmd = Median performance of the five-year median performance for the
portfolio of comparison companies (i.e. median of medians), Pk = five-year median performance value for member k of the comparison
portfolio for the company. k = Set of comparison portfolio companies. np > 1 Number of values comprising the set of comparison portfolio values
for a particular measure13.
12 A portfolio may contain firms not included in the 316 company sample frame. Potential portfolio candidate firms were expanded to include firms that did not meet the 50 to 10,000 employee criteria, but did meet the SIC code and asset matching requirements. 385 firms with 30 to 18,000 employees were allowed in the comparison portfolios, but not included in the sample frame to increase the likelihood of obtaining five companies with the request industry and asset size match. 13 The number of performance values for a particular measure in a particular portfolio may be less than the minimum of five companies in each portfolio due to missing values.
43
Median z-scores provide a ready means for distinguishing between high and low performers. A company with a z-score above zero (positive) is performing above the industry median for that particular performance measure and companies below zero (negative) are performing below median. Whether a high or low score is desirable depends on the particular measure. For example, a positive or higher z-score on ROE is good and a negative, or lesser z-score, is bad. Conversely, lower, or negative, z-scores are generally good for cycle time measures (e.g. takes less time to complete) and gross margin ratio (e.g. CGS<Sales).
Med
ian
z-sc
ore
SRSGROE
7.5
5.0
2.5
0.0
-2.5
-5.0
Distributions for Business Financial Performance Variables in Sample Frame
Note: Two outliers for ROE not shown at -15.6 and 40.9
Figure 4.3: Box plots for business performance measure median z-scores.
44
Med
ian
z-sc
ore
CTTCTCGMREPROCA
10
5
0
-5
-10
Distributions for Operations Financial Performance Variables in Sample Frame
Note: Outlier for CE not shown at 12.2
Figure 4.4: Box plots for operations performance measure median z-scores.14
4.3 Lean Manufacturing Practice Variables 4.3.0 Objectives This section describes the process of formulating the manufacturing practice variables used in this study. Overall, 48 companies from the 316-firm sample frame provided valid responses to a web-based survey instrument designed to capture the extent of practice implementation in the eight lean practice areas identified in the literature review. A final set of seven lean practices (Table 4.6) was identified through factor analysis for classifying lean and non-lean firms. The design of the survey instrument, the data collection process, the survey sample validation, the derivation of factor scores for each of the lean practices, and the formation of lean and non-lean archetype classifications are described in the following sections.
Internally oriented lean practices Just in time production methods (JIT) SPC tools to monitor quality (SPC) Employee involvement (EMP) Group technology to enhance the flow of products (GT)
14 Box plots for CTI, CTP, and CTR are not show due to space limitations, but demonstrate similar characteristics to the two aggregate cycle time measures shown.
45
Total productive maintenance (TPM) Externally oriented lean practices
Communication with suppliers (SCM) Customer involvement (CUS)
Table 4.6: Lean practices included in this study.
4.3.1 Survey Instrument The survey instrument consists of 36 items covering the eight practice areas identified in the literature review. The questions are set up on a five-point Likert scale (Hair, Anderson, Tatham, & Black, 1998) to measure the extent of implementation of the practices described by each of the items. The scale ranged from none (1) to complete implementation (5). Intermediate scale values were referenced to little (2), some (3), and extensive (4). The majority of the survey items were taken from those shown to have good construct validity and reliability in Shah’s dissertation (Shah, 2002). Items were added or modified based on a review by experts in the field including academics, lean consultants, and practicing manufacturing managers. To the degree that existing questions and constructs applied, the collection of practice data was largely confirmatory. A prime consideration in the design of the survey instrument was keeping it short and focused in order to obtain an adequate response rate. In addition to questions for the seven practice areas supplied by Shah (2002), five questions on total productive maintenance (TPM) were taken from Cua et al. (2001) and Koufteros et al. (1998). The questions selected for use in this survey focused on the autonomous maintenance aspect of TPM rather than the technology and proprietary equipment aspects also included by Cua et al. (2001). The autonomous maintenance aspect of TPM is the core element of the program that differentiates it from technology and automation practices. Appendix B provides a spreadsheet form of the survey showing those items used to develop the final factor scores for each practice construct. The actual survey was web-based. As an advantage of using a web-based instrument, automatic prompts asked respondents to go back and complete any questions that they skipped; thereby this study avoided problems with missing practice data. In fact, no practice data was missing for any of the 48 respondents. 4.3.2 Data Collection Once the 316-company sample frame was identified, the problem remained as to how to get an adequate response rate from knowledgeable, high-level operations managers within the target companies. The first approach was to locate company officers and contact information in the Standard&Poor's (2003) register of corporations, directors, and executives. Phone and mailing address information for corporate officers with titles such as vice president of operations, manufacturing, or supply chain were recorded. Titles of quality manager and chief executive officer were recorded if company employees with
46
operations related titles were not listed. A sample cross check against company web site information lead to a 100 percent check of the name, title, address and phone for each company against company web site information. In the event of conflicting information, the web site data were assumed to be more valid. A maximum of four formal attempts per company were made to solicit a survey response. The first two were by letter. An initial solicitation letter was sent to 378 firms including those comprising the final sample frame in August of 2003. The letter generally followed the recommended format for mailed surveys according to Dillman (2000). Example letters for the first and second mailing are included as Appendix C and D, respectively. Each was written on The Ohio State University, Fisher College of Business letterhead and included an endorsement letter from the Center for Excellence in Manufacturing Management at the college. The main thrust of the solicitation campaign was to encourage the respondents to visit the website and complete the questionnaire. After the first mailing, a sample of six non-responding companies was called by the researcher to gauge the effectiveness of the initial mailing. Some changes were made, including the inclusion of a company specific statement such as, “Company X Performance Analysis Results Enclosed” under the return address and a separate ranking and percentile sheet was redesigned in an effort to differentiate the solicitation letters from other mail. A unique aspect of the solicitation approach was to provide potential respondents with business and operations performance-ranking data in the solicitation letters. Each company was provided with their overall ranking and percentile performance on business and operations measures as a way to incite curiosity and response. As a “reward” for completing a response, respondents were allowed access to a webpage showing the ranking and percentile performance ratings for all the sample frame companies. Appendix E provides a partial listing from this webpage. To calculate each company’s rank and percentile performance, the individual z-scores for each measure were first ranked from one to 378. Business and operational ranking were based on the sum of the ranks of the individual measures in that particular category. The percentile performance simply indicates the company’s position in the ranking, expressed as a percentage. A percentile rating of 1% means that the company is in the top 1% of performers for that particular metric. The rankings were not used in this study except to improve the response rate and provide respondents with “up front value.” Note that the summing technique used to arrive at the business and operations performance ranking assumes equal weighting of included measures. This weighting scheme was intentionally simple and meant to induce response. A more proper holistic ranking scheme would account for different weightings with respect to varying emphasis warranted in different business and operations strategies. The response to the two-letter mailing was disappointing. Only seventeen firms (4.6%) completed the surveys and nine more declined participation (2.4%). On September 10, 2003, five interviewers from the The Ohio State University’s Center for Survey Research were provided a phone script and trained to call all non-responding, non-declining
47
companies. When an appropriate person could not be contacted directly, a voice message was left. This first phone call campaign resulted in 17 additional responses (34 total). On September 16, 2003, a second and final telephone campaign focused on reminder calls to individuals who seemed likely to respond based on the first phone call and companies in under-filled industry and size categories. This final campaign, in addition to personal calls and e-mail follow-ups by the researcher, resulted in an additional 16 responses (50 total) by mid October 2003. The final adjusted tally for the survey data collection is 48 valid responses from a validated sample frame of 316 for a response rate of 15.2%. Two of the responses were eliminated because, with the addition of the 2002 fiscal year data, they no longer met the sample frame criteria. One company was no longer listed in Compustat and one company dropped well below the fifty-employee minimum. The 316 sample frame was corrected for companies that failed to meet the sample frame criteria in 2002 fiscal data, had mail returned with no forwarding address, or did not have a validated phone number. The 316-sample frame can be broken down into 48 responders, 25 declining companies, and 243 non-responding companies. Table 4.7 shows a breakdown of respondents, non-respondents, and decliners in the sample frame by industry code.
Variable Response N 28 35 36 38 Mean Median Min MaxAssets - Total (MM$) None 243 57 40 68 78 272.8 71.9 5.20 2335.4
Yes 48 14 12 11 11 363.5 67.8 5.80 2865.8Decline 25 10 4 6 5 350.0 142.0 4.00 2105.0Overall 316 81 56 85 94 292.7 72.1 4.20 2865.8
Employees - Number (M) None 0.9 0.3 0.05 7.7Yes 1.0 0.3 0.05 7.5
Decline 0.9 0.4 0.06 8.2Overall 0.9 0.3 0.05 8.2
Inventory - Total (MM$) None 32.5 9.2 0.00 468.9Yes 34.6 10.1 0.00 329.1
Decline 31.6 5.0 0.00 253.5Overall 32.7 9.3 0.00 468.9
Sales - Net (MM$) None 195.2 54.8 0.00 1921.0Yes 231.3 65.3 6.70 2206.0
Decline 173.8 64.9 3.30 1215.4Overall 199.0 56.2 0.00 2206.0
SIC
Table 4.7: Descriptive statistics by survey response category.
The sample frame was designed to increase the responses’ validity for a single high-level respondent by selecting companies with relatively limited product lines and capping the number of employees at 9,000. Table 4.9 provides a listing of self-reported titles for all of the respondents. The title reporting with the highest frequency was vice president or director of operations (15), followed by vice president or director of manufacturing (eight). Exactly half (24) of the respondents reported having the terms “operations” or “manufacturing” in their title. Six respondents were company presidents or chief executive officers. Only three respondents reported titles perceived as being lower than
48
the vice president or director level (i.e. plant manager, production manager, and senior staff scientist). Only one respondent did not supply a title. Overall, the titles seem high enough in terms of typical manufacturing firm hierarchies, to assume that the survey respondents have a reasonably broad knowledge of the firm’s operations and practices.
Frequency Self Reported Respondent Title 15 Sr. VP, VP or Director Operations 8 VP or Director of Manufacturing 7 President, CEO, and/or COO 2 Executive Vice President 2 Vice President 1 Director of Finance 1 General Manager 1 Global Manufacturing Services Manager 1 Managing Director, Global Materials & Supply Chain 1 Plant Manager 1 Production Manager 1 Sr. Staff Scientist 1 Sr. Vice President 1 Therapeutics Operations Manager 1 VP Corporate Quality 1 VP Engineering 1 VP Manufacturing and Quality 1 VP Materials Management 1 VP Quality 1 VP Technology 1 Not reported
50 Total
Table 4.9: Self reported respondent titles sorted by frequency.
4.3.4 Practice Construct Formulation This section describes the process of forming the seven practice variables constructs that captured the extent of lean practice implementation. The five internally and two externally oriented lean practices are listed in Table 4.6. A factor analysis was performed on the data items contained in the 48 valid responses to the survey. The analysis sought to exploit relationships between data items previously explored in literature (Cua et al., 2001; Koufteros et al., 1998; Shah, 2002). The overall objective was to establish the extent to which each company surveyed had implemented each distinct practice. 4.3.4.1 Factor Analysis The data were factor-analyzed using SPSS 11.5 for Windows (SPSS, 2002). A seven-factor solution was found to be relatively robust and corresponded closely to expectations derived from previous research. The solution was then confirmed through a maximum likelihood analysis with Oblimin rotation (SPSS, 2002). Each factor is composed of
49
three to five items with factor loadings above 0.38. All Eigen values were above 1.0. Cumulative variance explained for all seven constructs was 78.2 percent. Table 4.10 shows the items, factor loading, and factor score weights for their respective constructs. The table is sorted in order of highest factor loading within each practice construct. Cross-loading was minimal with only two group technology questions on the similarity of routing and processing requirements (GT1 and GT2) loading at similar levels on statistical process control. These items were retained to maintain content validity. An eight-factor solution was attempted, but the loading was not specific to the expected supplier JIT factor. That construct was dropped and items deemed relevant to JIT (i.e. SJT1: “Our key suppliers manage inventory in our facilities”) and supplier communication (i.e. SJT3: “Suppliers are directly involved in the new product development process”) were included in those constructs. Appendix F includes complete tables of the factor loadings and factor score coefficients. Overall, the factor analysis is considered quite successful from the standpoint that seven of the constructs loaded as expected even though the sample size was relatively small.
50
ItemFactor
Loading
Factor Score
Weight
Just-in-time (JIT) Production Methods [α=.85]JIT2 0.94 1.015 Production at stations is "pulled" by the current demand of the next stationJIT1 0.85 0.020 We use a "pull" production systemJIT3 0.78 0.009 We use Kanbans, squares, or containers as signals for production controlJIT4 0.48 -0.002 Production is "pulled" by the shipment of finished goods
SJT1 0.46 0.009 Our key suppliers manage inventory in our facilitiesStatistical Process Control (SPC) [α=.85]
SPC2 0.68 0.368 Statistical techniques are used to reduce process varianceSPC1 0.63 0.240 Equipment and processes on the shop floor are currently under SPCSPC3 0.62 0.139 Charts showing defect rates are used as tools on the shop-floorSPC4 0.52 0.081 We use fishbone-type diagrams to identify causes of quality problems
Employee Involvement in Problem Solving (EMP) [α=.87]EMP3 0.88 0.393 Shop-floor employees lead product/process improvement effortsEMP1 0.75 0.361 Shop-floor employees are key to problem solving teamsEMP2 0.71 0.256 Shop-floor employees drive suggestion programs
Group Technology (GT) [α=.88]GT4 0.91 0.617 Families of products determine our factory layoutGT3 0.78 0.267 Equipment is grouped to produce a continuous flow of families of productsGT2 0.46 0.106 Products are classified into groups with similar routing requirementsGT1 0.38 0.063 Products are classified into groups with similar processing requirements
Total Productive Maintenance (TPM) [α=.93]TPM5 0.93 0.998 We do preventive maintenanceTPM4 0.77 0.046TPM2 0.55 0.024
Supplier Communication (SCM) [α=.72]SCM3 0.81 0.367 We require major suppliers to contribute to cost and quality improvement effortsSCM2 0.80 0.472 We give our suppliers feedback on quality and delivery performanceSJT3 0.50 0.129 Suppliers are directly involved in the new product development process
Customer Involvement (CUS) [α=.70]CUS2 0.76 0.372 Customers frequently share current & future demand information with marketingCUS1 0.72 0.363 Our customers are actively involved in current and future product offeringsCUS3 0.49 0.175 Our customers give us feedback on quality and delivery performance
We emphasize good maintenance as a strategy for achieving quality and schedule compliance
Records of routine maintenance are kept
Internally oriented lean practices [Cronbach's Alpha]
Externally oriented lean practices
Table 4.10: Factor analysis items, loading, factor score weights and reliability.
4.3.5 Lean Archetype Formation Lean practices have been described as a strategic configuration or holistic set of mutually supportive practices (Shah, 2002; Shah et al., 2003; Ward et al., 1996). Therefore, one approach to testing the hypothesis that lean manufacturing practices result in higher levels of operations and business financial performance is to identify companies that have a configuration of lean practices and compare their performance to those that do not. The majority of past survey-empirical studies used regression models to test the relationship between a set of lean practices and performance (Table 2.2). Callen et al. (2000) used a minimum practice-level heuristic to classify companies into lean and non-lean practitioners. Shah (2002) used a multifaceted approach including clustering similar to that used in this study. This study also takes a multifaceted approach to the problem and uses cluster analysis as a repeatable, systematic, and non-arbitrary method to identify lean
51
and non-lean companies. It also employs discriminant analysis and logistic regression models to validate the cluster analysis. The discriminant analysis identifies which practices contribute most to “leanness.” The logistic regression identifies which practices within lean are most relevant to financial performance. 4.3.5.1 Cluster Analysis A hierarchical cluster analysis was performed on the seven lean practice variables using the squared Euclidian distance between variables and Ward’s method of optimizing the minimum variance within clusters (SPSS, 2002). Ward’s method takes as an objective function the minimization of the within-groups sum of squares error because “[t]his method tends to find (or create) clusters of relatively equal sizes and shapes as hyperspheres.” (Aldenderfer & Blashfield, 1984, p43). Ward’s technique also tends to combine clusters with a small number of observations (Hair et al., 1998) which is an important feature since the sample includes only 48 observations. A hierarchical approach and squared Euclidian distance were selected because of their straightforward interpretation and dominance in the frequency of applied use (Aldenderfer et al., 1984). Shah (2002) used the same method in discussing lean configurations, albeit with a much larger sample size. Sample size becomes important in determining the number of clusters that can be extracted. Cluster analysis is a non-inferential statistical method in which variables are grouped based on logical heuristics. There are no tests to prove statistically whether a particular clustering is valid. Therefore, the most objective approach to confirm clustering is replication with a second independent sample. Since, the relatively small sample size (48) precludes this option; this study validates the clustering in three other ways. First, the clusters are examined for consistency with expectations from theory with respect to practice levels within groups. Lean clusters should have significantly higher mean practice levels than non-lean clusters. Expectedly, lean practices act as a highly interrelated set and lean cluster members will not be high in only one or two practices. General sense-making of the data also contributed to the evaluation of the validity of the cluster analysis. Second, a discriminant analysis was performed to evaluate the significance of the practice variables as a set in determining cluster membership wherein the relative contribution of each lean practice is examined. Lastly, the clusters were compared with respect to predictive validity. Differences between clusters on the operations financial performance variables were compared. The assumption was made that lean and non-lean companies are likely to be different on one or all of the operations performance variables, with cycle time being the most likely. A logistic regression analysis was performed to see if a set of operations financial performance variables could predict cluster membership. Logistic regression was used as opposed to discriminant analysis because of its less stringent assumptions with respect to predictor variables and its similarity to regression in interpretation. This study also
52
employs logistic regression in other areas of the research, and a degree of parsimony in method use is desirable. 4.3.5.1.1 Cluster Results Two and three-cluster solutions were derived from a hierarchical cluster analysis of the factor scores for the seven lean practices. Both solutions were examined for their predictive validity and consistency with expected practice levels within groups. Table 4.11 shows the mean factor scores for the two and three cluster solutions. The two-cluster solution distinctly forms one lean (A) and one non-lean (B) cluster. Cluster A includes all positive mean lean practice scores and Cluster B includes all negative mean practice scores. This is sensible from the standpoint that lean practices tend to act as an integrated set. The two clusters differ significantly on only four of the seven practice factors. The two-cluster solution is not significantly different on TPM or either of the external lean practice factors (i.e. SCM, CUS). In addition, when clusters A and B were compared with respect to differences in the set of operations financial performance measures, only receivables cycle time (CTR) was significantly different (p = .016). Cash-to-cash cycle time (CTC) and inventory cycle time (CTI) were moderately15 significantly different at p= .069 and p = .065 respectively16. All three cycle time measure differences were in the expected direction with the lean firms (Cluster A) having lower cycle times.
15 The term “moderately significant” is used to designate a statistical test result that has a p-value less than 0.1 but greater than the 0.05 cutoff value typically used for Type I Error. The term “significant” alone is used when the p-value is less than 0.05. 16 Differences compared using a non-parametric Wilcoxon rank-sum test SPSS. 2002. SPSS for Windows, 11.5 ed.: SPSS, Inc.
53
t-test (A & B only)
A B C SignificanceTwo Cluster Solution
n = 24 24
Internally oriented lean practicesJust in time production methods (JIT) 0.80 -0.80 ***SPC tools to monitor quality (SPC) 0.30 -0.30 **Employee involvement (EMP) 0.53 -0.53 ***Group technology (GT) 0.37 -0.37 ***Total productive maintenance (TPM) 0.12 -0.12
Externally oriented lean practicesCommunication with suppliers (SCM) 0.05 -0.05Customer involvement (CUS) 0.18 -0.18Three Cluster Solution
n = 24 18 6
Internally oriented lean practicesJust in time production methods (JIT) 0.80 -0.64 -1.27 ***SPC tools to monitor quality (SPC) 0.30 -0.39 -0.04 **Employee involvement (EMP) 0.53 -0.43 -0.83 ***Group technology (GT) 0.37 -0.88 1.16 ***Total productive maintenance (TPM) 0.12 -0.50 1.01 **
Externally oriented lean practicesCommunication with suppliers (SCM) 0.05 -0.11 0.12Customer involvement (CUS) 0.18 0.02 -0.78
Note: Significant difference for 2-tailed test*** p<.01, ** p<.05, * p<0.1
Mean Factor Scores
Cluster Designation
Mean Factor Scores
Table 4.11: Mean values for two and three cluster analysis solutions.
4.3.5.1.2 Cluster Validation The three-cluster solution is more appealing from both a conceptual and predictive standpoint. When a third cluster was extracted, a small group of six firms with a unique set of practice characteristics was identified as part of the non-lean (B) cluster (Table 4.11). These firms are low in all lean practices except group technology, total productive maintenance, and supplier communication, where they scored high relative to the other two clusters. These firms arguably pursue some type of automation/technology-focused strategy. Although it is of interest as a strategic alternative to a balanced lean approach, the fact that only six firms are present in that cluster make it impractical to carry them forward in the current study. With the C Cluster removed, TPM becomes significantly different between Clusters A and B. Henceforth, A and B will be designated as L (lean) and N (non-lean) clusters. The fact that the external practice factors supplier communication (SCM) and customer involvement (CUS) were not significantly different between the two large clusters can be rationalized in several ways. The external lean factors could be less central to a core implementation plan with practicing firms. Lean manufacturing typically displays a
54
pattern of starting on the manufacturing floor and working its way both down and up the supply chain. It could also be that the external practices are more difficult to implement than internal ones since each depends on the cooperation of an external partner not under the direct control of the firm. Lastly, the sample size may not be large enough to pick up the difference for those practices. A discriminant analysis supports the L and N cluster solution. The analysis was run to test the significance of the practice variables as a set as well as the contributions of the individual practice variables in determining firm classification. The discriminant function is significant (p= .000) and the canonical correlation was .855 (SPSS, 2002). Box’s M test was not significant (p<0.1), thus indicating that the assumption of equal population covariance matrices holds. Table 4.12 shows lean practice variables ordered by absolute size of correlation within the function structure (SPSS, 2002). JIT, GT, and EMP seem to be integral determinants of lean classification, while the two external practices (i.e. SCM and CUS) make little difference at all. The predictive capability of the Lean (L) and Non-lean (N) clusters provides a compelling argument for their selection as the primary classification categories used in this study. Table 4.13 shows the medians and significance for Wilcoxon rank-sum tests for all the operations financial performance variables. Note that for the three cluster solution (L/N), all of the cycle time measures (except payables) are significant (p<0.05) or moderately significant (p<0.1) and in the expected direction for the lean cluster. It is likely that cycle time would be the key measure affected by lean practice implementation. The fact that the gross margin ratio is moderately significant and in the unexpected direction, (i.e. lean firms have lower gross margins, thus poorer performance) was surprising at first. However, the result is probably an artifact of competitive pricing or “margin squeeze.” One of the primary reasons firms implement lean practices is as a strategic response to competitive price pressure. Lean implementation and tight margins are expected to be correlated. Non-lean firms have larger margins and do not feel the imperative for implementing lean. The argument is not that lean causes worse margins, but that if firms in an intensively competitive pricing environment did not implement lean, they might not have a feasible margin. The fact that asset and employee productivity are not significantly different for lean and non-lean firms is more problematic. Proponents of lean argue that lean plants are more productive. Partial explanations may lie in the long-term nature of assets or admonitions against using lean as a downsizing or layoff program. In any case, the lean and non-lean clusters by way of a three-cluster solution have significantly more predictive power than the two-cluster result. Therefore, the lean and non-lean classifications used in the rest of this study are those from the three-cluster solution. Overall, the logistic regression analysis provides support for the predictive validity of the L and N cluster solution and indicates that cycle time performance, with respect to competitors, is the primary determinant of cluster membership.
55
Function 1 JIT .757 GT .576 EMP .383 SPC .239 TPM .203 SCM .056 CUS .055
Table 4.12: Structure matrix for Lean and Non-lean Cluster discriminant function.
Wilcoxon Rank-Sum Test
Lean (L) Non-lean (N) Significancen = 24 18
P-valueAsset Productivity (ROCA) 0.37 0.31
Employee Productivity (EP) 0.20 0.34Gross Margin Ratio (GMR) -0.01 -0.30 *
Cash-to-Cash Cycle Time (CTC) -0.16 0.49 **Total Cycle Time (CTT) -0.41 0.25 *
Inventory Cycle Time (CTI) -0.27 0.65 **Payables Cycle Time (CTP) -0.16 -0.44
Receivables Cycle Time (CTR) -0.28 0.45 **Note: Significant difference for 2-tailed test
** p<.05, * p<0.1
Cluster Designation
Median Z-scores
Table 4.13: Median z-scores for operations financial performance variables.
56
4.5a Binary Logistic Regression: Cluster versus ROCA, EP, GMR, CTC Variable Value Count Cluster N 17 (Event) L 24 Total 41 * NOTE * 1/42 cases contained missing values Logistic Regression Table Odds 95% CI Predictor Coef SE Coef Z P Ratio Lower Upper Constant -0.591 0.435 -1.36 0.18 ROCA 0.428 0.368 1.16 0.25 1.53 0.75 3.16 EP -0.476 0.435 -1.09 0.27 0.62 0.26 1.46 GMR -0.426 0.497 -0.86 0.39 0.65 0.25 1.73 CTC 0.902 0.468 1.93 0.05 2.46 0.98 6.17 Test that all slopes are zero: G = 9.058, DF = 4, P-Value = 0.060
4.5b Binary Logistic Regression: Cluster versus ROCA, EP, GMR, CTT Variable Value Count Cluster N 17 (Event) L 24 Total 41 * NOTE * 1/42 cases contained missing values Logistic Regression Table Odds 95% CI Predictor Coef SE Coef Z P Ratio Lower Upper Constant -0.391 0.405 -0.97 0.33 ROCA 0.673 0.440 1.53 0.13 1.96 0.83 4.65 EP -0.621 0.436 -1.42 0.16 0.54 0.23 1.26 GMR -0.115 0.519 -0.22 0.83 0.89 0.32 2.47 CTT 1.175 0.658 1.78 0.07 3.24 0.89 11.77 Test that all slopes are zero: G = 8.473, DF = 4, P-Value = 0.076
Figure 4.5: Binary logit models predicting cluster membership including cash-to-
cash cycle time (a) or total cycle time (b). 4.5 Summary In order to test relationships between lean practices and business financial performance, this study depended largely on the formation of a valid variable set. The method combines statistical and heuristic techniques found in finance and operations management empirical studies. A sample frame was designed to increase confidence that single respondent surveys accurately reflected the lean practices implemented in a sample firm. Care was taken to ensure comparability of financial performance, across industries and firm size, by using a matched portfolio of comparison firms and a portfolio median-based median z-score comparison measure. The use of five-year median values for performance measures increased confidence that they measured sustained performance. Lean practice scores were based on questions and constructs previously identified in
57
operations management literature and confirmed by high loadings and reliabilities in the sample data set. The lean constructs covered the lean practice spectrum with a more comprehensive question set than previous studies combining empirical and archival data. Lean practice factor scores were used to identify lean and non-lean archetypes through cluster analysis. A discriminant analysis confirmed the clusters to be highly associated with JIT, group technology, and employee involvement and less so with external practices like supplier communication and customer involvement. The lean clusters, through non-parametric 2-sample tests and logistic regression, were shown to be predictive of superior cycle time and inferior gross margin performance. The resulting variables were then analyzed to shed light on the relationship between lean manufacturing management practices and financial performance. Lean and non-lean clusters were compared on operations and business financial measures using non-parametric Wilcoxon rank-sum tests. The level of individual practice usage was compared for above- and below-industry median performers on those same measures. Logit predictive models were developed and applied to a sample of survey non-respondents to explore the relationship between lean operations financial performance and business financial performance.
58
CHAPTER 5
RESULTS 5.0 Results – Objectives and Overview Three variable sets were developed in the previous chapter: lean manufacturing practice variables, operations financial performance variables, and business financial performance variables. This chapter investigates the relationships between each of these variable sets. Figure 5.1 represents pictorially the relationships investigated. To enhance efficient communication the relationships are labeled L-O for the lean practice to operations financial, L-B for the relationship between lean practices and business financial performance, and O-B for the relationship between operations and business financial performance variables. This chapter addresses the analysis of the relationships L-O, L-B, and O-B, in that order. Figure 5.1 also shows an important distinction in the sources of the data for the three variable sets. A key feature of this study is its use of different sources for practice and performance data. Perceptual surveys of high-level executives provide it with practice data, and performance data derives from archival public data sources. The key results of this study can be summarized as follows:
Lean versus Operations Financial Performance (L-O): Lean companies differ from non-lean companies with respect to operations financial performance. The only performance measures on which lean firms are significantly better than non-lean firms are time-related. Lean firms’ overall cycle times are shorter (total and cash-to-cash) and they have shorter inventory and receivables cycle times. Lean companies do not have significantly different asset or employee productivity. The only operations financial measure for which lean firms are significantly worse than non-lean ones is gross margin ratio. Lean firms operate with narrower gross margins than non-lean firms do. Lean versus Business Performance (L-B): Lean firms, as a classification described by a set of integrated practices, did not demonstrate significantly better business-level financial performance than non-lean firms for the sample companies. Business financial performance was based on separate measures of ROE, sales growth, and stock return. However, companies that had higher than competitive median ROE had significantly higher implementation levels of employee involvement than below-median firms. Employee involvement is a core practice in lean and may even be a prerequisite for successful implementation of other lean practices (Flynn et al., 1995). Alternatively, the externally-oriented lean practice of supplier communication was significantly related to below competitive median ROE.
59
Operations versus Business Financial Performance (O-B): Based on the relationship observed between lean practice and operations financial performance in the survey sample, lean companies were postulated to have a characteristic relationship between operations and business financial variables that can be used to predict firms likely to have implemented lean practices. This characteristic relationship was modeled using logistic regression to create a “lean financial signature.” When applied to the 244 survey non-respondents in the sample frame, this model predicts that lean companies have significantly better ROE. This is the case even though the lean firms do not have significantly better asset or employee productivity and have narrower gross margins. The only competitive strength contributing to these lean companies’ ROE performance is faster cycle time.
Practice and Performance VariablesArchival Data:• Compustat/CRSP• N = 316• Base measures - 5 year median
performance (1998-2002).• Measures are normalized
differences from the median performance of matched comparison company portfolios
LeanPractices
BusinessFinancial ($)Performance
OperationsFinancial ($)Performance
“L-O”
“O-B”
“L-B”
Perceptual Data:• Survey of executives• n = 48/316• Base measures are factor
scores for 7 lean constructs• Lean and non-lean clusters
Figure 5.1: Relationships between variables investigated.
5.1 Lean versus Operations Financial Performance (L-O) 5.1.2 Analysis and Results To establish the relationship between lean manufacturing practices and operations performance, it was necessary to validate the formation of lean and non-lean archetypes (clusters). In the Methods Chapter, an analysis proposed that the seven lean practices acted as a gestalt, or balanced set of practices, in determining a company’s lean or non-leanness. If that were indeed the case, the expectation is that operations performance
60
would be related strongly to lean categorical classification and not necessarily to individual practice levels or even the simple linear combinations of those practices. 5.1.2.1 Lean Archetype versus Operations Financial Performance Two statistical analyses were performed in order to test the supposition that leanness--as captured in archetypal classification--is related to operations performance. The first analysis examined the relationship between lean classification and eight individual measures of operations financial performance. Table 4.13 in the Methods Chapter presents the analysis results. Figures 5.2 and 5.3 plot operations financial performance by cluster membership. The median z-scores may be interpreted as measures of distance above or below the competitive median performance for that particular metric. Wilcoxon rank-sum tests of median z-scores for lean and non-lean firms are not significantly different for asset or employee productivity. The difference was in the expected direction for asset productivity, with the median z-score for lean being higher (0.37 versus 0.31). Employee productivity lay in the unexpected direction with the median z-scores for non-lean firms being higher. The amount of emphasis to place on these directional observations is limited since statistical tests did not find a significant difference. The effect of sample size on the likelihood of not finding a significant difference in performance between lean and non-lean companies when one truly exists (Type II Error) was analyzed by resampling the original with replacement to obtain sample sizes of 100, 200, and 300. The results are reported in Table 5.1. The difference in asset productivity is only detectable at a z-score difference 3 times that found in the original sample (0.18 vs 0.06) and sample size of 100. Employee productivity is not found to be significantly different at any difference or sample size using this method. Gross margin ratio was found to be moderately significant but also in the unexpected direction. Lean companies tend to have narrower gross margins than non-lean companies do. As discussed in the Methods Chapter, this outcome might have been foreseen. A potential reason for selecting lean as a strategic improvement program is profit margin pressure. Where lean companies significantly differentiate themselves is with respect to cycle time performance. They have significantly lower aggregate cycle time as measured by cash-to-cash cycle time and component-level cycle time as measured by inventory and receivables performance. Total cycle time, as another aggregate measure, is also moderately significantly better for lean companies. Payables cycle time is not significant although the difference is in the unexpected direction in that non-lean companies pay their vendors faster than lean companies.
61
Data4.83.62.41.20.0-1.2-2.4
Cluster
ROCA
EP
GMR
L
N
L
N
L
N
ClusterLN
Dotplot of ROCA, EP, GMR vs Cluster
Figure 5.2: Plots of lean (L) and non-lean (N) clusters’ operations financial
performance (i.e. ROCA, EP, GMR).
Data3.22.41.60.80.0-0.8-1.6
Cluster
CTC
CTT
CTP
CTI
CTR
L
N
L
N
L
N
L
N
LN
ClusterLN
Dotplot of CTC, CTT, CTP, CTI, CTR vs Cluster
Figure 5.3: Plots of lean (L) and non-lean (N) clusters’ cycle time operations financial performance.
62
Difference P-value Difference P-value Difference P-value Difference P-value
ROE 0.22 0.790 0.24 0.202 0.22 0.181 0.26 0.007SG 0.11 0.453 0.13 0.050 0.14 0.000 0.03 0.106SR 0.24 0.905 0.23 0.484 0.30 0.563 -0.07 0.645
ROCA 0.06 0.809 0.18 0.042 0.01 0.234 0.18 0.016EP -0.14 0.929 -0.10 0.894 -0.15 0.968 -0.10 0.110GMR 0.30 0.077 0.16 0.115 0.24 0.001 0.35 0.000CTC -0.65 0.040 -0.60 0.000 -0.59 0.000 -0.59 0.000CTT -0.65 0.066 -0.44 0.014 -0.42 0.001 -0.64 0.000Notes: 1. Nominal sample size. Adjusting for missing data: SR = 23/15, CTC = 24/17, CTT = 24/17.
2. Values based on single, simple random selection from original n=24/18 sample values with replacement to sample sizes of n=100, 200, and 300.3. Difference based on median Z-scores lean minus non-lean.4. P-values based on 2-tail Wilcoxon rank sum test. Values </=0.10 are in bold italics.
Business Financial Performance
Operations Financial Performance
Sample Firms (Lean/Non-lean) Single Resample with Replacement (Lean versus Non-lean)
n=24/18 (1) n=100 n=200 n=300
Table 5.1: Lean versus non-lean firm median z-score differences for the original
sample and resampled data with replacement. 5.1.2.2 Individual Lean Practice versus Operations Financial Performance An alternative to explaining differences in operations performance based on cluster membership is an examination of the relationship between individual lean practices and operations financial performance. To accomplish this, eight individual logistic regressions were run to test whether the main effects of the seven lean practices linearly combine to predict differences in the eight measures of operations performance. Factor scores for all seven lean practices were entered into the model simultaneously. Operations financial performance was categorized as above or below competitive median performance based on positive or negative median z-scores for a particular performance measure. All 48 respondent companies were included in the analysis since, in this case, lean cluster membership is not considered. Of the eight logit models run for each of the operations financial performance variables, none were significant at the p<0.1 level except for payables cycle time (p=.000). Any one of eight significant models is easy to discount using the familywise error argument presented above. A one-in-three probability exists that one of the models would be significant just by random chance. It is interesting to note that the only two coefficients that were significant in this particular model were supplier communication (p=.004) and group technology (p=.019). Supplier communication was positively correlated with accounts payable cycle time. This could be cynically interpreted as “talk is cheap,” or the observation that companies communicating a lot with their suppliers do not pay quickly. Group technology is correlated in the expected direction (i.e. higher use of group technology is associated with lower payables cycle time), but a logical justification for the relationship is obscure. All models were rerun with the two externally oriented
63
practice (i.e. SCM and CUS) variables removed to see if the reduced partitioning of the variance would yield significant models, but the results were similar to the previous ones. None of the regressions were significant17 at p<0.1. 5.1.2.3 Lean versus Operations Performance Summary Lean archetypes were found to be significantly different from non-lean archetypes with respect to cycle time and gross margin measures of operations financial performance. Lean firms have lower aggregate cycle time and narrower gross margins. Alternately, the individual practices that make-up lean were not found to be linearly related to operations performance. This gives credence to the notion of lean acting as a balanced and integrated set of practices. 5.2 Lean versus Business Financial Performance (L-B) 5.2.2 Analysis and Results The analysis of lean manufacturing management practice versus business financial performance progresses in a manner similar to that between lean and operations performance. The differences between lean and non-lean archetypes were compared and then individual lean practices examined for their relationship to each of the three business financial performance measures. 5.2.2.1 Lean Archetype versus Business Financial Performance The supposition that business performance is related to lean cluster membership was tested in two ways. Business performances on three measures (i.e. ROE, sales growth, and stock return) were compared for lean and non-lean clusters with individual Wilcoxon rank-sum tests. Cluster differences were also tested using logistic regression with the three business measures variables as predictors of cluster membership. Neither approach revealed a significant relationship between cluster membership and any of the business performance variables. Table 5.2 displays the results of the Wilcoxon rank-sum tests. While none of the differences is significant, the lean cluster has higher median z-score values on all three measures. The effect of sample size on the likelihood of making a Type II Error was analyzed by examining differences between resampled data for each of the business financial performance measures (Table 5.1). ROE was only significantly different at the n=300 level and sales growth at the n=100 to 200 level. Stock return never demonstrated a significant difference at these sample levels. A series of plots shown in Figure 5.4 show the relative positions of the individual data points. The approach that used logistic regression yielded no significant results. A model entering all three measures of business performance as main effects in predicting cluster membership was significant only at the p=.34 level.
17 A similar result was obtained using standard regression models predicting operations financial performance as a continuous variable.
64
Wilcoxon Rank-Sum
TestLean (L) Non-lean (N) Significance
P-valueReturn on Equity (ROE) 0.38 0.16 0.78
n 24 18Sales Growth (SG) 0.06 -0.06 0.45
n 24 18Stock Return (SR) 0.42 0.18 0.89
n 23 15Notes: 1. Significant difference for 2-tailed test
2. Stock return n smaller due to missing data.
Cluster Designation
Median Z-scores
Table 5.2: Lean cluster difference test results for business financial performance.
Data2.71.80.90.0-0.9-1.8-2.7
Cluster
ROE
SG
SR
L
N
L
N
L
N
ClusterLN
Dotplot of ROE, SG, SR vs Cluster
Figure 5.4: Plots of lean (L) and non-lean (N) clusters versus business financial
performance measures. 5.2.2.2 Individual Lean Practice versus Business Financial Performance An alternative to explaining differences in business financial performance by way of lean/non-lean archetypes is to examine the relationship between individual lean practices and business performance. This relationship was tested in two ways. The first ran individual two-sample t-tests comparing above and below competitive median business financial performers with respect to lean practice implementation. The second examined
65
three logistic regression models using the seven practice factor scores as covariates and above and below-competitive median business financial performance as the dependent variable. The results of the 2-sample comparisons are shown in Table 5.3 which indicated that high ROE performers (above competitive median) tend to have only moderately significant lower levels of supplier communication and significantly higher levels of employee involvement than low ROE performers. Other practice variables did not differ between high and low performers. The pattern was repeated with respect to above- and below-competitive median sales growth performers with the exception that the difference was more significant in supplier communication practice implementation. High stock return performers demonstrated a negative association with both forms of external lean practices. Supplier communication was moderately significantly different and customer communication was significantly different. As previously discussed, the use of individual comparison tests must be interpreted with consideration of familywise error. For 21 comparisons, it is likely that at least one of the comparisons would be significant at the p<0.05 level purely by chance. The familywise chance of making a Type I Error is 66 percent18. The logistic regression approach compensates for a portion of the familywise error by simultaneously testing all of the lean practice variables. Of the three models (i.e. ROE, SG, and SR), only return on equity was significant (p=.049). Neither sales growth nor stock return was significant at the p<0.1 level. Figure 5.5 shows the model for ROE. Supplier communication and employee involvement were the only significant practices at p=.035 and p=.092, respectively. The odds ratio for each significant coefficient in the model is a marginal likelihood. The confidence interval provided in Figure 5.5 is with respect to the odds ratio. In this case, the likelihood of a firm having above-competitive median performance decreased by 3.2 times (reciprocal of .031) for every one unit increase in SCM factor score with all other factors being equal. Supplier communication seemed to have a negative relationship with ROE. The likelihood of above-median ROE performance increased by 2.3 times for every one-unit increase in employee involvement. EMP had a positive relationship with ROE. Plots in Figure 5.6 graphically depict the breakout of above and below-competitive median ROE groups with practice implementation score as the x-axis. The logit model offers support for the findings in the two sample t-tests, namely that supplier communication and employee involvement are negatively and positively associated with ROE, respectively. 5.2.3 Lean versus Business Performance Summary While lean archetypes per say were not found to be associated with above or below-competitive median business financial performance, one of the three primary differentiating practices of lean and non-lean firms was found to be associated with
18 659.)05.1(1)1(1 21 =−−=−−= c
FWαα
66
above-median business financial performance. Employee involvement was found to be associated with higher ROE performance.
ROE SG SRSupplier communication SCM (0.44) * (0.55) ** (0.36) *
Just in Time JIT 0.36 0.11 (0.07)Statistic Process control SPC 0.30 0.10 (0.06)Employee Involvement EMP 0.59 ** 0.57 ** 0.19
Group Technology GT (0.15) (0.02) (0.18)Total Productive Maintenance TPM (0.25) 0.20 (0.14)
Customer Involvement CUS 0.30 (0.04) (0.51) **Note: Significant difference one tailed t-test: **p<.05, *p<.10
Practice Factor Score Differences(Above minus Below Competitive Median
Performers)Practices
Table 5.3: Comparisons between above and below competitive median business
financial performers with respect to lean practice level. Binary Logistic Regression: NROE versus SCM, JIT, SPC, EMP, GT, TPM, CUS
Variable Value CountNROE Hi 32 (Event)
Lo 16Total 48
Logistic Regression TableOdds
Predictor Coef SE Coef Z P Ratio Lower UpperConstant 0.996 0.400 2.49 0.01SCM -1.164 0.551 -2.11 0.04 0.31 0.11 0.92JIT -0.180 0.456 -0.39 0.69 0.84 0.34 2.04SPC 0.825 0.548 1.51 0.13 2.28 0.78 6.68EMP 0.851 0.504 1.69 0.09 2.34 0.87 6.29GT -0.026 0.403 -0.06 0.95 0.97 0.44 2.15TPM -0.830 0.507 -1.64 0.10 0.44 0.16 1.18CUS 0.589 0.474 1.24 0.21 1.80 0.71 4.57
Test that all slopes are zero: G = 14.112, DF = 7, P-Value = 0.049
95% CI
Figure 5.5: Binary logit model predicting above (Hi) or below (Lo) competitive
median ROE performance.
67
Data2.11.40.70.0-0.7-1.4-2.1
ROEq
SCM
EMP
Hi
Lo
Hi
Lo
ROEqHiLo
Dotplot of SCM, EMP vs ROEq
Figure 5.6: Plots of above (Hi) and below (Lo) competitive median ROE groups for
lean practice scores on SCM and EMP (Data). 5.3 Lean Operations versus Business Financial Performance (O-B) 5.3.1 Overview Lean classification was not significantly related to business financial performance in the survey sample companies. Employee involvement, as a core element in lean implementation programs, was found to be positively related to better ROE performance. Lean practices as an integrated set did not have the expected positive association with any of the three measures of business performance. This section takes another approach to testing the question of whether lean archetypes tend to have better business financial performance than non-lean archetypes. As noted earlier, lean classification does seem to be significantly related to operations financial performance, most appreciably with respect to better cycle time performance. This observation is derived from the two logit models regressing lean cluster membership against a parsimonious set of operations financial performance variables. The two models were moderately significant at p=.060 and p=.076 respectively. The models are presented as:
Model 1: CTCGMREPROCANPNPLn 902.426.475.428.591.)(1(
)( +−−+−=⎥⎦⎤
⎢⎣⎡
−
Model 2: CTTGMREPROCANPNPLn 175.1115.621.673.391.)(1(
)( +−−+−=⎥⎦⎤
⎢⎣⎡
−
68
Besides differences in coefficients, the two models also differ in their aggregate cycle time terms. Model 1 uses cash-to-cash cycle time (CTC) and Model 2 uses total cycle time (CTT). Cycle time has the only significant coefficient in either model at p=.05 and p=.07, respectively (ref. Figures 4.2a and b for full model details). The logit models are interpreted as the log of the odds that a firm is non-lean. These models can be applied within the sample frame to predict which companies are likely to be lean or non-lean. Effectively, the models can be thought of as lean “financial signatures” (lean signatures) for firm operations performance. Theoretically, the models can be used to predict lean cluster membership in an independent sample of companies. 5.3.2 Analysis and Results The survey non-respondents in the sample frame make up a readily testable independent sample. Survey respondents and non-respondents were shown to not be significantly different with respect to industry composition, size, and common measures of financial performance. Several precautions were taken in preparing the data and applying the lean signature in the sample frame19. The survey non-respondent sample frame data set was trimmed to keep the independent variable data within the range of the data used to develop the models. Table 5.4 shows the survey sample data range, the non-respondents before trimming, and the resulting data ranges and sample size. The survey-responding firms used to generate the lean signature models were excluded from the new sample, thereby ensuring independence. A 75% probability level was used to delineate lean or non-lean archetype membership. The 75% value was arbitrarily selected as a trade-off between developing a lean versus non-lean contrast and retaining an adequate sample size. The 75% cutoff was tested20 for sensitivity from 65 to 80% without an appreciable difference in the outcome for the total cycle time model (Table 5.6). The cash-to-cash cycle time model proved much less robust, showing only moderately significant results for lean having lower sales growth at 75%.
19 These precautions are also described in the methods section 4.5.3. 20 Increments of +/-5% above and below 75% were tested.
69
Survey Sample (n=42)ROCA EP CE CTC CTT
Max 5.368 4.784 5.163 3.092 2.949Min -2.189 -2.123 -2.979 -1.539 -1.393
Survey Non-respondents (n=274)Max 2.936 4.288 12.243 5.568 6.971Min -8.773 -7.555 -4.504 -2.336 -2.929
Trimmed Survey Non-respondents (n=244)Max 2.936 4.288 3.913 2.925 2.913Min -1.923 -2.112 -2.440 -1.506 -1.383
Note: Includes 5 of 6 TPM/GT cluster companies who responded to the survey, but were not categorized in lean nor non-lean clusters.
Table 5.4: Survey sample data ranges versus non-respondent data pre- and post-trimming.
Wilcoxon Rank-Sum Test
Lean Non-leanBusiness
Return on Equity ROE 0.14 -0.34 0.01**Sales Growth SG -0.01 0.01 0.85Stock Return SR 0.27 0.21 0.57
Operations Lean Signature
Asset Productivity ROCA 0.01 0.06 0.62Employee Productivity EP 0.18 0.12 0.85
Gross Margin Ratio GMR 0.15 -0.39 0.00***Total Cycle Time CTT -0.73 1.65 0.00***
Component Cycle TimePayables CTP -0.24 1.05 0.00***Inventory CTI -0.62 1.49 0.00***
Receivables CTR -0.25 0.29 0.00***Notes: 1. Significant difference for 2-tailed test.
2. ***p<.01, **p<.05
Median Z-scoresPredicted Archetype P-value
SignificanceFinancial Performance Variables
Table 5.7: Predicted differences in financial performance variables using Model 2 (CTT).
5.4 Summary The relationships between lean manufacturing management practices and operations financial and business financial performance was examined from several perspectives. Lean practices were shown to be significantly related to operations financial performance
70
(L-O) as an integrated set and as individual practices. Lean practices were not significantly related to business financial performance (L-B) as a set when tested for the companies in the survey sample. However, two analyses did imply a substantive positive relationship between lean practice and business financial performance in the form of ROE. Employee involvement, as a core lean practice, is significantly and positively related to ROE performance. In addition, when a logit model developed using the surveyed sample was used to differentiate lean and non-lean firms in a separate sample, the lean firms demonstrated a significantly better ROE performance. Sales growth and stock performance did not seem significantly related to either leanness as a strategic archetype or as individual practices.
71
CHAPTER 6
DISCUSSION 6.0 Discussion – Objectives and Overview This chapter provides an in-depth discussion of key results as they pertain to the relationship between lean manufacturing management practices and financial performance. Key results will be covered in the logical order developed in previous chapters. Covered first is the composition of practices making up lean strategic archetypes. Next, the relationship between lean, as a strategic archetype and as individual practices, and operations financial performance are discussed, followed by lean’s relationship to business level financial performance. Comments are made concerning managerial implications and a comparison of the results to those of other studies. The final section of this chapter discusses threats to the validity of the results. 6.1 Lean Archetype Practice Composition Understanding how lean practices come together to form a unified strategic approach for a firm is a crucial building block for research into the connections between leanness and financial performance. A key finding of this study is that lean practices act as integrated, synergistic set in affecting operations financial performance. Lean practices are not linearly additive in their affect on performance. This is demonstrated by the finding of a significant positive relationship between lean archetype classification and cycle time performance and the lack of significance in a regression model sought to identify linear combinations of lean practices. Investigation of the relative composition of practice implementation levels is a central factor in understanding how lean affects financial performance. JIT practice emerges as the primary discriminant factor distinguishing lean and non-lean firms followed by group technology and employee involvement (Table 4.12). This result varies somewhat from Shah (2002), who found employee involvement followed by supplier communication and JIT to have the highest discriminant loadings. All loadings above 0.3 are considered substantive (Hair et al., 1998). Both studies found JIT practices, such as the use of Kanbans for production control, to be one of the top three discriminators of leanness and both have employee involvement in common. Just-in-time, group technology, and employee involvement are found to be the top determinants of lean categorization in the current study. This adds confidence that the categorization correctly distinguishes lean and non-lean firms. Shah (2002) and the results of this study differ in determining the role of external practices like supplier communication and customer involvement in distinguishing lean and non-lean firms. No differences between lean and non-lean firms, with respect to their
72
implementation of external practices, were illustrated in this study. This can be viewed as a positive attribute of lean since supplier communication is the only significant practice found to be negatively correlated to ROE performance when regressed against all individual lean practices (ref. Figure 5.5). A potential interpretation of the joint results of external practices being non-significant in discriminating for leanness and the presence of a significant negative correlation to business financial performance maintains that external lean practices are difficult to implement in both extent and effectiveness. It is possible that the lack of a meaningful difference for lean firms in the extent of external practice implementation could indicate that those particular practices are difficult to execute. Even though a firm may attempt to implement a lean practice, it may not be competently performed. Supplier communication and customer involvement requires a communication protocol and infrastructure crossing organizational boundaries. That supplier communication is negatively correlated to performance indicates that those firms that do implement it are not obtaining the desired positive effect. An alternative explanation for the external practice findings may be that the primacy of implementation is indicative of the sequence in which the practices are implemented. This argument follows the vein of capability building put forward by Ferdows and De Meyer (1990) and Noble (1995). Manufacturing organizations initially tend to implement lean manufacturing practices in their core operations and then to integrate vertically—either back to suppliers or forward to customers. As the last ones implemented, external practices are not as far along the learning curve and, therefore, capability is less well developed. The finding also could be attributed to a reversal in proposed causation. Companies that have “problem” suppliers and customers tend to implement more practices to attempt to mitigate their negative effects. This would explain why firms with higher levels of supplier communication tend to have lower ROE. The result may also hint at the power of the supplier in the firm-supplier relationship. The relative buyer-supplier power relationship was not studied in this research. 6.2 Lean versus Operations Financial Performance (L-O) Lean archetypes were found to perform better than non-lean companies with respect to measures of cycle time. However, surprisingly, they were not found to perform better with respect to measures of asset or employee productivity. This finding is in contrast to MacDuffie (1995) who found a strong positive correlation between measures of both employee involvement and its interaction with JIT systems with labor productivity as measured by the number of hours required to build an automobile. This study finds that the primarily mode of lean’s effect on operations performance is a reduction in overall transaction time. Some evidence suggests that lean firms operate with lower profit margins than non-lean firms. These results contrast to tests of the relationship between individual practice
73
measures and performance, which found no significant relationship. This adds credence to the proposition that lean acts as a unified set of interrelated, synergistic practices in affecting operations financial performance. Lean archetypes were shown to have significantly better cycle time performance than non-lean firms. This corresponds well with both past empirical studies and theory regarding lean’s focus on waste reduction, especially in the form of excess inventory. Cycle time is the total time it takes for a transaction, or unit of production, to be processed from start to finish. The fact that both cash-to-cash and total cycle time seem to be equally effective in distinguishing between lean and non-lean firms indicates that payables cycle time, or the time between receipt and payment for raw material, is not a significant factor in measuring lean performance (Figures 4.2a and b). Examination of Table 4.13 supports the observation that shorter inventory and receivables cycle times are the distinguishing measures of lean operations financial performance. Lean firms operate with less total inventory than their competitors operate and are paid in less time by their customers. There are some indications that lean companies operate with a lower gross margin (GMR) than non-lean companies (Table 4.13)21. In other words, lean companies operate with less difference between selling price and cost of goods sold. This finding conflicts with Proposition 1 that lean companies have better financial performance on all five operations measures. A possible explanation for this observation is that lean practice implementation is a competitive response to price pressure. Low margins are a cause of lean implementation rather than a result; therefore, there tends to be a positive correlation between low gross margins and leanness. Huson and Nanda (1995) observed that although the direct effect of JIT is to lower cost, the margins following JIT implementation tend to decrease more than in non-JIT firms. Both the Huson and Nanda (1995) results and those of this study stand in contrast to Callen et al. (2000) who found a found a positive relationship between JIT-TQM implementation and contribution margin per sales dollar. The net implication is that managers should be wary of gross profit margin improvement as a goal or measure of effective lean implementation. A comparison of above- and below-competitive median GMR performers within lean firms did not indicate a significant difference in implementation for any of the seven individual lean practices. Lean practices should act to improve or maintain profit margins through waste elimination and cost reduction. A larger sample size of lean companies (greater than 200 based on Table 5.1) may demonstrate definitively whether or not levels of lean implementation are associated with improved margins.
21 Gross margin ratio (GMR) is the measure used in this study and appears in all tables and formula. Since it is basically calculated as cost of goods sold divide by sales, lower values are “better” (larger difference between CGS and sales). A higher GMR value corresponds to a “lower gross margin.”
74
Neither asset productivity (ROCA) nor employee productivity (EP) differs between lean and non-lean archetypes in this study. ROCA results contrast somewhat with Fullerton et al., (2003) who found a positive relationship between what they called JIT manufacturing (i.e. focused factory, group technology, reduced setup times, TPM, multifunctional employees, and uniform workload) and ROA and ROS. They found no relationship between measures of financial performance and JIT unique effects (i.e. Kanban and JIT purchasing) although they did control for inventory financial benefits, which are a primary contribution from JIT implementation. The ROCA measure in this study includes inventory, excludes liquid assets, and compares the firm median five-year performance with that of a comparison portfolio of companies. The lack of a significant result for ROCA could result from the fact that many plant and equipment assets are long term. Even as lean practice implementation reduces the requirements for floor space and increases the useful productivity of equipment, those assets remain on the accounting books. If sales growth is flat, the only way ROCA can be affected is through inventory reduction and reductions in operating expense. If that impact is not significant in relation to physical assets, the measure does not move. In any case, ROCA does not seem to be overtly responsive to lean practice implementation. That employee productivity is not significantly different between lean and non-lean was somewhat expected. The literature review studies that included a measure of return per employee were split 50/50 on the finding of a significant relationship (Table 2.6). MacDuffie (1995) found a significantly positive relationship between lean practice implementation and a more direct measure of productivity (labor hours). Experience in implementing lean programs attests to the likelihood of a labor requirement reduction. A typical consultant recommendation relative to implementing lean manufacturing is to anticipate productivity improvements with either layoff up-front or plans for redeployment of excess workers. One possible explanation for the lack of a significant finding is that the survey data lags behind the performance data. The performance data represents a time span of 1998 to 2002 while the survey was taken in the fall of 2003. A premise of this research acknowledges that lean implementation takes time. An oft repeated nostrum of lean production recalls that it took Toyota 40 to 50 years to achieve their current level of lean practice implementation, and the process continues today (Drickhamer, 2004; Womack et al., 1991). In the latest generation of lean implementations, the expectation is that substantive results should be evident in two to five years. In fact, substantial improvements in inventory reduction are often achieved in the first six to 12 months. If this is indeed the case, the significance of cycle time improvement and non-significance of productivity improvements could be indicative of a lag in performance effect, with cycle time improvements leading the improvement in productivity. An assumption made in interpretation on all practice-performance relationship analysis results is that lean practices take time to implement and institutionalize and that performance effects are cumulative. The company performance data are from the years
75
1998 to 2002 and the survey data represents lean practice implementation levels in the summer and fall of 2003. The survey data collection lags the financial performance data by at least one year. However, the assumption that lean practice take multiple years to implement and that the results are cumulative means that indications of changes to performance should be evident in the earlier time period, especially for firms with high levels of practice implementation. A conscience decision was made in this research to focus on currently observed levels of implementation rather than more suspect “time since initiation.” Under the assumption of time to implement and cumulative effects, finding evidence of a significant effect of lean practice on financial performance may represent a conservative bias. The alternative assumption is that firms with better financial performance tend to be better able to implement lean practices. The research methods employed in this study does not allow a definitive resolution as to which assumption is correct. However, the use of the median for five years of sustained performance reduces the negative implications of the reversal in causality (performance practice). An interpretation of the results in the reversed direction is that firms with consistently higher than competitive median financial performance implement lean and sustain good performance. In either direction of causation, lean is at least not detrimental to financial performance. Another possible explanation lies in the relationship between lean implementation and capacity. It is possible that productivity increases do not show up as readily in firms where output is restricted by the market demand rather than plant capacity. In the case of a market restriction, lean implementation translates into an increase in relative capacity rather than output. Since a major lean production principle asserts that building more than is needed is a waste and warns against utilization as a goal, this finding is consistent with lean thinking. These results bear significant implications for managers attempting to measure or justify lean implementation programs. It is unlikely that lean practice implementation will show immediate results in asset or employee productivity. Alternatively, one may postulate that total cycle time is responsive to lean practice implementation. Total cycle time should be the financial accounting-based measure of choice in either monitoring or justifying lean programs. 6.3 Lean versus Business Financial Performance (L-B) Lean archetypes were not found to have significantly different performance than non-lean firms when tested with respect to business-level financial performance measures in the sample database (i.e. ROE, sales growth, and stock return). However, individual lean practices were found to be significantly related. Employee involvement demonstrated a positive relationship to above-competitive-median ROE, while the practice of supplier communication demonstrated a negative relationship (Figure 5.5). This indicates a weak relationship between business financial performance and lean strategy in general but a strong relationship with specific aspects of lean.
76
Compared to previous studies, the results of this study are interesting in two respects. The first is in not finding a significant relationship between lean manufacturing management practices and stock return. All of the cited JIT and TQM studies that examined stock performance found a positive correlation (Easton et al., 1998; Hendricks et al., 1996; Howton et al., 2000). The Hendricks and Singhal (1996) and Howton et al. (2000) studies looked at short-term effects on stock price preceding and following JIT and TQM announcements (-200 to -11 days versus announcement) and can be attributed to the perceived stock market valuation of those practices. The Easton and Jarrell (1998) TQM study is more in line with the current study by examining the performance over a similar five year period. Both positive stock performance results of the Easton and Jarrell (1998) and positive ROE results of this study could be based on the underlying mechanism of employee involvement fundamental to both successful TQM implementation and lean (Flynn et al., 1995; MacDuffie, 1995). The second difference occurs in the finding of a significant relationship between lean practices and ROE. The other study to look at ROE was Boyd et al. (2002). Unlike this study, they did not find a significant relationship between JIT and ROE; however, the sample size was small and did not control for industry and size. It also is possible that studies of “only JIT” overlook the importance of employee teams and problem-solving to achieve sustained financial results. It is significant that the human resource aspects of lean seem to be of most significance in moving financial measures. This emphasizes that it is the “soft,” cultural side of lean that really makes continuous improvement possible in that lean becomes an employee commitment rather than an abstract philosophy. The importance of employee involvement as a crucial infrastructural element in successful lean implementation is supported in studies by Flynn et al. (1995), MacDuffie, 1995), and Sakakibara et al. (1997). It is noted that both this study and Shah (2002) found the presence of employee involvement to be a differentiator between lean and non-lean firms. The importance of having both technology practices (such as JIT and group technology) and human element practices (such as employee involvement) is supported by other empirical studies in technology implementation where human resource management and technology together lead to further advantage (Boyer, Leong, Ward, & Krajewski, 1997). The lack of a significant result for sales growth in this study was disappointing. The quality and delivery improvements attributed to lean implementation by previous studies (Table 2.5) would lead one to expect an increase in sales in comparison to non-lean firms. This did not occur. A possible explanation could be in the tight profit margin characteristic identified for lean firms in this study. Without a clear costs advantage, lean firms must compete on speed and quality alone. Price reductions are never an attractive competitive option when costs are high. It is possible that lean firms lose out by competing with firms that offer lower prices, poorer quality and delivery and customers simply choose price as the primary selection criteria. The requirement for only profitable growth may be limiting criteria for lean firms with tight margins. It may be that lean firms with higher levels of effective implementation do have higher sales growth rates, but the sample size in this study is too small to detect it.
77
There are three general managerial implications of the relationships found between lean manufacturing management practices and high-level business financial performance:
1. Expectations of direct short-term impact on sales growth and stock performance should be conservative. The evidence does not support a strong direct connection.
2. Over the long term, ROE should be considered a primary measure of lean practice implementation effectiveness at the business financial performance level.
3. Investing in employee involvement is a likely avenue for increasing firm value and laying a foundation for effective lean implementation.
6.4 Lean Operations versus Business Financial Performance (O-B) The previous section discussed the lack of significance between lean archetype classification in general and business level performance and a significant relationship between employee involvement, a key lean practice, and ROE. A logistic regression model was developed to identify lean archetypes by their “financial signature.” Using this model and survey non-respondents in the sample frame, 41 lean and 19 non-lean firms were identified with a high level of certainty. A comparison of lean and non-lean firms identified by this means indicated that lean firms have significantly better ROE performance and no difference with respect to sales growth and stock performance. The result that lean firms have better ROE is even more interesting when considered in conjunction with the results on the other performance measures. Lean firms have better ROE performance than non-lean ones even though they tend to have lower gross margins and are not significantly different with respect to asset or employee productivity (ref. Table 5.7). In fact, these lean companies’ only advantage is their faster cycle times; turning inventory, receivables, and payables faster than non-lean companies. Since they operate with lower margins their ROE performance likely results primarily from a higher volume of business in the same amount of time. These results reinforce the management implications offered in previous sections, namely that ROE is a strong business level indicator of lean performance and that cycle time is a key measure at the operations level. Alternately, lean is not a sole recipe for increasing sales growth or stock price. At the operations level, companies that emphasize asset and employee productivity measures are likely to be disappointed in the performance of their lean implementation over the short-term. 6.6 Summary The primary determinants of leanness are the implementation of JIT, group technology, and employee involvement practices. The finding of employee involvement as a key determinant agrees with other studies and, in conjunction with its strong individual practice correlation to above-competitive-median ROE performance, makes it a strong candidate for a central role in any lean implementation program. A primary benefit, and measure, of lean implementation at the operations level is an improvement in total cycle time. This supports previous studies and theory on the way lean acts to reduce inefficiency or waste in processes, especially in the form of excess
78
inventory. The implication for management is that cycle time, rather than asset or employee productivity, should be the primary metric and expectation for lean implementations. Although lean is thought to have a significant role in cost reduction, this study does not find that to be the case. Leanness seems to be associated with lower gross margins. It is not clear whether lower margins are a result or a cause of lean practice implementation. It is likely that companies implement lean practices as a strategic reaction to competitive price pressure and that without the cost-reducing effects of lean, overall business-level financial performance measures, such as ROE, would be flatter or negative. The results of this study did find moderate support for the proposition that lean companies have better ROE performance. Although not evident from the survey sample, a process identifying lean firms based on a “lean financial signature” in the non-respondent sample frame did show a positive relationship. Lean firms can be successful solely by competing on time. Lean firms have better cycle time performance but lower gross margins and not significantly different asset and employee productivity than non-lean firms. Firms implementing lean improvement programs should expect improvements in cycle time and ROE performance.
79
CHAPTER 7
CONCLUSION 7.1 Research Objective Lean, the manufacturers’ paradigm of choice, is the heir apparent to mass-production practice. Not since the turn of the century when Henry Ford supplanted craft manufacturing with his high volume, low cost, automated approach has there been such a growing consensus as to what fosters a manufacturer’s competitiveness. Although measuring the popularity of lean was not a goal of the current study, the proportion of lean and non-lean firms in the sample was not found to be significantly different. Twenty-four of 42 sample firms from four different industries were identified as being lean. Lean is no longer a rare approach. An understanding of what it means to be a lean manufacturer continues to develop. Originally, lean manufacturing was merely perceived as casual imitation of Japanese manufacturing practices, such as quality circles and Kanbans, which could be applied individually to achieve desired effects. Lean is now construed more properly as a holistic set of interrelated, synergistic practices that work in concert to achieve significant improvements in cost, time, and quality performance. Currently, lean marks the culmination of our search toward understanding efficient, competent manufacturing. Over the last three decades, practitioners and researchers have been piecing together the relationships between of total quality management, just-in-time, and total productive maintenance, once thought to be individual paradigms. Although it is unlikely that the realm of lean practice has been fully defined or its potential fully realized, we now know enough about the key practices within lean to know what to focus on when examining a company’s current or proposed practice implementation. Even as understanding, use, and advocacy for lean practices have taken hold, some reservation remains. The direct link between lean practice implementation and financial performance has not been consistently demonstrated by ostensibly “good” research studies. The approach of traditional operations management empirical research is limited by their restriction to plant-level data and a limited access to proprietary archival financial data. Traditional financial researchers possessed the skill to access and analyze business-level archival data but lacked an in-depth understanding of the paradigms operating in lean manufacturing. This lack of understanding, in combination with a lack of direct access to production floor practices, bars them from deriving definitive conclusions about the relationships between practices and financial performance. To more definitively assess the relationship between lean practice and financial performance this study rectifies these shortcomings and combines comprehensive surveys of manufacturing management practices with archival financial data. This study finds lean to be a distinct, differentiatable collection of practice areas--namely just-in-time,
80
group technology, employee involvement, statistical process control, and total productive maintenance. The implementation of these practices as an integrated set differentiates lean companies from their competitors. Although the nature of social scientific research precludes definitive answers, this study concludes that lean companies do perform better than non-lean ones on time-based measures of operations financial performance and at least one measure of overall business financial performance. 7.2 Key Findings The relationship between lean manufacturing management practices, operations financial performance, and business financial performance can be viewed as a hierarchical relationship. Business financial performance resides at the top of that hierarchy. The three measures of business performance used in this study gauged the impact of lean practice implementation--sales growth, stock return, and return on equity–and resulted in a significant positive relationship between lean practice and ROE but none to sales growth or stock return. It is possible that ROE stands out because it is most purely a function of internal operations management practices. Both sales growth and stock return are significantly influenced by external forces, as sales growth is a function of customer perceptions of value and competitive positioning and stock return is largely determined by investor perceptions of future value. Interestingly, all three measures provide a different perspective on value. ROE takes an internal, operational view of value, whereas sales growth takes a customer view and stock return an investor view. It may be that this study’s results demonstrate customer’s and investor’s lack of maturity in fully appreciating lean manufacturing’s benefits. Customers are likely more short-term cost-conscious than aware of the benefits lean offers in the form of simultaneously-improved quality, delivery, flexibility, and total cost of ownership. This speculation finds support in that this study did not find that supplier communication and customer involvement significantly contributed to the determination of lean classification. Even ostensibly lean companies have difficulty valuing and managing their connections to customers and suppliers. Similarly, investors may also require more education and experience in the benefits of lean to fully appreciate it. It is not obvious why firms with above-competitive median ROE performance do not also have above-competitive median stock return. ROE is generally accepted to be positively correlated to market-to-book, when adjusted for the cost of equity capital (Fruhan, 1979; Woo, 1984; Varaiya, Kerin, & Weeks, 1987). ROE was not adjusted for the cost of equity capital in this study. An alternate explanation suggests that investors do not understand the lean mechanisms underlying the higher ROE performance and so do not foresee the performance continuing or else view early lean implementation as just another short-lived “program of the month.” Value investors such as Warren Buffet likely would take another view advocating a combination of solid ROE performance with good underlying management practices as a worthy investment. Another finding of this study perceives the important role that cycle time plays as a goal and measure of lean performance. Of the operations financial measures used in this study, only the cycle time measures demonstrated consistently improved performance
81
with the implementation of lean practices. The strength of the cycle time finding maybe considered partial affirmation of the “theory of swift and even flow” (Schmenner & Swink, 1998). This theory is consistent with lean philosophy in holding that the reduction of throughput time and variability are the primary determinants of improved productivity. Although variability was not a measure incorporated in this study, the importance of median total transaction time as a primary focus and result of lean practice implementation is supported. Asset and employee productivity were not significantly affected by lean implementation and gross margins in lean firms were actually narrower in firms not implementing lean practices. Cycle time is effective because it is a measure of a fundamental result of focusing on lean. Lean firms focus on reducing waste in inventory and transaction time, which results in lower total cycle time. At the outset of this study, two measures of overall cycle time performance were identified. One emphasized the role of lean in freeing up cash for the firm (cash-to-cash cycle time). The other looked at the role of lean as primarily a mechanism to reduce total transaction time (total cycle time). Although both measures were significant in various tests, ultimately, total cycle time proved the more robust predictor of upper level business performance. This seems to indicate that lean firms are best advised to think of lean as a paradigm for reducing all forms of transaction time and not as merely a mechanism for freeing up cash. The practices that most clearly distinguished lean and non-lean firms were just-in-time, group technology, and employee involvement, all of which are internal practices under the sole control of operations management. JIT and GT represent well-defined, easily verifiable practices. One way to view JIT and GT are as “hard” lean practices since physical evidence of their implementation can be readily verified through inspection of the production floor. One would expect to find Kanbans in use and the demand pull originating from the customer traceable back through the production line to the raw material. JIT provides the underlying paradigm and mechanisms for production control. Group technology is evidenced by routing sheets and the grouping of equipment into product-family focused cells. The paradigm drives the facilities layout and routing of product through the factory. Both JIT and GT leave hard, physical evidence of their implementation and use. Evidence of employee involvement may be less readily apparent based on physical evidence. Employee involvement represents the soft, less tangible, human resource dimension of lean; the fact that shop floor employees “are the key to problem-solving” and “lead product/process improvement efforts” may not be readily apparent upon inspection. A more in-depth knowledge of the operation’s workings or discussions with employees would be required to identify the implementation of employee involvement. Regardless of the difficulties in identifying the presence or absence of employee involvement, the salient observation is that it is not solely the hard practices that promote the effectiveness of lean implementations.
82
In fact, it is the soft practice implementation of employee involvement that seems to make the most difference with respect to business financial performance. Employee involvement was the only lean practice to demonstrate a positive relationship to ROE as a measure of business-level financial performance. The positive relationship between employee involvement and ROE was clearly demonstrated even as the relationship between lean archetypes and ROE proved less strong. This is an interesting result because employee involvement is a soft practice, being the least visible and least formulaic of the lean practices. Human resource managers should be gratified that this study verifies that the human asset makes the biggest difference in lean company financial performance. 7.3 Managerial Implications 7.3.0 Overview and Objectives The direct, specific managerial implications of the study results have already been covered in the Discussion Chapter (6) in relation to specific findings. In this section, two more general implications are discussed. The first arises from the implications of the results with respect to a lean implementation model. The second develops the results in absolute terms to gain a better appreciation for the potential impact of lean on firm financial performance. 7.3.1 Lean Implementation In an application sense, this study’s findings form the basis for a rudimentary implementation model for lean manufacturing firms. First, organize facilities into product-focused flows in line with the tenants of group technology. Next, implement a product flow control system based on Kanbans and pull production. Product should move through the factory based on demand initiated by the customer and signaled back through the production line. At each stage of implementation, involve the employees who will be responsible for running the line and solving problems that arise in the future. To aid employees in running the line and solving problems help them to develop their skills in statistical process control and total productive maintenance. As a first order measure of the financial effectiveness of this implementation, monitor total cycle time. The expected outcome of a successful implementation should be an improvement in business-level ROE performance. Figure 7.1 depicts this model in graphical form. It should be noted that this lean implementation model needs additional verification and testing. The small sample size and cross-sectional nature of this study cannot explicitly attest to the correctness of this model. No information in the study permits an analysis of how the firms in this study came to be lean over time; likely, many unique paths were followed. The model in Figure 7.1 describes but one logical and parsimonious approach that explains the results of the study.
83
OperationsLevel
Monitor:Total Cycle
Time
BusinessLevel
Result:Improved
ROEGroup
TechnologyJIT
Production
Employee Involvement
SPC
TPM
Lean Implementation Plan
Feedback
Figure 7.1: Proposed lean implementation model based on study results.
7.3.2 Absolute Magnitude of Lean Effects Managers should be encouraged to implement lean because it is generally associated with higher levels of ROE. The difference in median ROE median z-scores for companies predicted to be lean and non-lean was 0.57 (Table 5.5). Since z-scores are somewhat difficult to interpret, a translation is warranted. If the 0.57 z-score difference is applied to the predicted non-lean companies in the non-respondent sample from which the score was derived, it represents a median absolute difference of 0.143. The calculation of the median absolute difference is accomplished through a simple manipulation of the median z-score given in the Methods Chapter22. In other words, an extremely non-lean company (75% probability) could expect to improve its five-year median ROE performance by as many as 14.3 points if it implemented lean manufacturing management practices! Alternatively, an extremely lean company would have a five-year median ROE 20.8 points lower had it not implemented lean. In an absolute sense, this result is extraordinary. Companies stand to benefit substantially from implementing lean manufacturing management practices. However, caution should be used in interpreting this result. Although a logical and conservative prediction method was followed in deriving this result from the non-respondent sample frame, a significant ROE difference was not found in the original sample. The non-respondent sample frame result may be an artifact of the prediction process and not solely lean practice. For example, although the use of a 75% probability
22 ( )
PPmdn
PmdPzzDifferenceAbsoluteMedian
p
n
kk
differenceP
p
−
⎥⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢⎢
⎣
⎡
+−
−+=
∑=
1)( 1
2
84
criterion increases confidence that the firms so identified are truly lean or non-lean, it likely also ensures that the firms are atypical or extreme in their practice. Thus, results should be interpreted with caution. A performance difference substantiated in the original sample frame was that of total cycle time performance (CTT). Lean companies had a median median z-score 0.66 lower than non-lean companies. This translates, for non-lean companies that implement lean, into a potential 75-day reduction in five-year median total cycle time and represents an increase of 89 days for companies had they not implemented lean manufacturing. The median five-year median total cycle times were 275 and 188 days for non-lean and lean groups, respectively, and represent a substantial difference as a percentage. Lean companies realize a 32 percent reduction in total transaction time that can be applied directly to competitive advantage in the form of lower operating costs, lower inventory costs, faster delivery, and shorter lead times. The shorter cycle times of lean firms indirectly translate into better quality and increased flexibility. This study finds a substantial cycle time advantage for companies implementing lean manufacturing management practices. 7.4 Future Research This study raises other potential research questions and opens the door for further study of the practice-performance question. One obvious avenue for future research is to look at the performance of the same companies over time. The current survey represents a snapshot interval of company practice, while the dependent variable in this study was the sustained performance of these companies for the years 1998 to 2002. Since the Compustat and CRISP data for each of theses companies are updated every year, it will be possible to compare performance trends over time to a known time point and level of lean implementation. Interesting questions may then be asked on the sustainability of lean performance results and the effect of practice mix on sustainability. Lags in performance may also be examined. Possibly, the reason an ROE effect was found while a stock return effect was lacking was that stock performance lags ROE performance. The design of this research allows for at least one future analysis without the necessity of conducting another survey. A relationship that was not examined in this research is the effect of lean practice on financial performance variability. Value investors typically look at three aspects of firm performance. Two of these, sustained ROE performance and good management practice, were examined in the current study. The third aspect is variability in financial measures such as ROE (Vick, 2001). Financial performance variability should be reduced by lean practice implementation. Lean strives to “make value flow without interruption” (Womack et al., 1996) and reduce all sources of variation in the manufacturing process (Schonberger, 1996). An expanded study examining not only the effects of lean on the median 5-year performance but also the variation over that time period would no doubt prove enlightening.
85
This study also expands the opportunity for researchers to examine additional industries. The median z-score method employed here effectively controls for industry and size effects. Based on this result, it would be possible to expand the sample frame to include all manufacturing SIC codes, which would allow a broader generalization of the findings and would also increase the power of the findings by increasing the sample size. There is also no obvious reason why the same research approach would not apply to service industries. Services are currently the largest sector of the US economy, representing over 80% of jobs and growing (Jacobs, 2001, Table 2-1, pp. 161-164). Researchers such as Bowen and Youngdahl (1998) make a strong case for the application of lean practices to services. The research design would include identifying single SIC code service providers in the Compustat database and adapting the survey questions to service operations. The goal of such research aim to find if lean has a similar effect on the financial performance of service companies as the effect it has on manufacturing firms. Finally, this research uncovered a potential third strategic archetype. Six companies were identified in the original sample that did not seem to fit in either the lean or non-lean categories. These firms possess very high levels of group technology and total productive maintenance practices while they simultaneously implement low levels of other lean practices. There are representatives from all the industry codes included in this study, thereby reducing the likelihood that this archetype is industry specific or limited to only process intensive manufacturers. It would be interesting to obtain a larger sample size of these companies to see if their strategy is more or less effective than lean in achieving sustained financial results.
86
BIBLIOGRAPHY Aldenderfer, M. S. & Blashfield, R. K. 1984. Cluster analysis. Beverly Hills: Sage
Publications. Anupindi, R., Chopra, Deshmukh, Mieghem, & Zemel. 1999. Managing business process
flows (1st ed). Upper Saddle River, NJ: Prentice Hall. Balakrishnan, R., Linsmeier, T. J., & Venkatachalam, M. 1996. Financial benefits from
JIT adoption: Effects of customer concentration and cost structure. Accounting Review, 71(2): 183.
Barber, B. M. & Lyon, J. D. 1996. Detecting abnormal operating performance: The empirical power and specification of test statistics. Journal of Financial Economics, 41(Issue 3): 359.
Bettis, R. A. 1996. SMJ 1995 best paper prize to Danny Miller. Strategic Management Journal, 17(7): 503.
Biggart, T. B. & Gargeya, V. B. 2002. Impact of JIT on inventory to sales ratios. Industrial Management & Data Systems, 102(4): 197-202.
Billesbach, T. J. & Hayen, R. 1994a. Long-term impact of just-in-time on inventory performance measures. Production & Inventory Management Journal, 35(1): 62.
Billesbach, T. J. & Hayen, R. 1994b. Long-term impact of just-in-time on inventory performance measures. Production & Inventory Management Journal, 35: 62.
Blumenthal, R. G. 1998. 'Tis the Gift to Be Simple: Why the 80-year-old DuPont model still has fans. CFO Magazine.
Bowen, D. E. & Youngdahl, W. E. 1998. "Lean" service: in defense of a production-line approach. International Journal of Service Industry Management, 9(3): 207-225.
Boyd, D. T.; Modified duPont Model; http://www.clt.astate.edu/dboyd/modified_dupont_model.htm; Jan 2004.
Boyd, D. T., Kronk, L., & Skinner, R. 2002. The effects of just-in-time systems on financial accounting metrics. Industrial Management & Data Systems, 102(3): 153-164.
Boyer, K. K., Leong, G. K., Ward, P. T., & Krajewski, L. J. 1997. Unlocking the potential of advanced manufacturing technologies. Journal of Operations Management, 15(4): 331-347.
Bozarth, C. & McDermott, C. 1998. Configurations in manufacturing strategy: A review and directions for future research. Journal of Operations Management, 16(4): 427-439.
Bremner, B. & Dawson, C. 2003. Can anything stop Toyota? BusinessWeek, Vol. 3858. Business Source Premier; Sept 30, 2003.
Callen, J. L., Fader, C., & Krinsky, I. 2000. Just-in-time: A cross-sectional plant analysis. International Journal of Production Economics, 63(3): 277-301.
Chang, D. & Lee, S. M. 1995. Impact of JIT on organizational performance of U.S. firms. International Journal of Production Research, 33(11): 3053.
Claycomb, C., Germain, R., & Droge, C. 1999. Total system JIT outcomes: Inventory, organization and financial effects. International Journal of Physical Distribution & Logistics Management, 29(9/10): 612.
87
Cohen, J. & Cohen, P. 1983. Applied multiple regression/correlation analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: L. Erlbaum Associates.
Compustat; http://wrds.wharton.upenn.edu/ds/comp/index.shtml; August 2003. Center for Research in Security Prices (CRSP);
http://wrds.wharton.upenn.edu/ds/crsp/index.shtml; August 2003. Cua, K. O. 2000. A theory of integrated manufacturing practices: Relating total quality
management, just-in-time and total productive maintenance. unpublished dissertation, University of Minnesota.
Cua, K. O., McKone, K. E., & Schroeder, R. G. 2001. Relationships between implementation of TQM, JIT, and TPM and manufacturing performance. Journal of Operations Management, 19(6): 675.
Davy, J. A., White, R. E., Merritt, N. J., & Gritzmacher, K. 1992. A derivation of the underlying constructs of just-in-time management systems. Academy of Management Journal, 35: 653.
Dess, G. G., Ireland, R. D., & Hitt, M. A. 1990. Industry effects and strategic management research. Journal of Management, 16(Issue 1): 7.
Dillman, D. A. 2000. Mail and internet surveys : The tailored design method / Don A. Dillman (2nd ed). New York: J. Wiley.
Drickhamer, D. 2004. Lean manufacturing: The 3rd generation. Industry Week/IW, 253: 25-30: Penton Publishing.
Easton, G. S. & Jarrell, S. L. 1998. The effects of total quality management on corporate performance. Journal of Business, 71(Issue 2): 253.
Ferdows, K. & De Meyer, A. 1990. Lasting Improvements in manufacturing performance: In search of a new theory. Journal of Operations Management, 9(2): 168-184.
Field, A. P. 2000. Discovering statistics using SPSS for windows: Advanced techniques for the beginner. London; Thousand Oaks: Sage Publications.
Flynn, B. B., Schroeder, R. G., & Sakakibara, S. 1994. A framework for quality management research and an associated measurement instrument. Journal of Operations Management, 11(4): 339-366.
Flynn, B. B., Sakakibara, S., & Schroeder, R. G. 1995. Relation between JIT & TQM: Practices and performance. Academy of Management Journal, 38(5): 1325-1360.
Fullerton, R. R. & McWatters, C. S. 2001. The production performance benefits from JIT implementation. Journal of Operations Management, 19(1): 81.
Fullerton, R. R., McWatters, C. S., & Fawson, C. 2003. An examination of the relationships between JIT and financial performance. Journal of Operations Management, 21(4): 383.
Fruhan, W. E. 1979. Financial strategy : studies in the creation, transfer, and destruction of shareholder value. Homewood, Ill.: R. D. Irwin.
Graham, B. & Dodd, D. L. 1934. Security analysis. New York: Whittlesey House, McGraw-Hill book company, inc.
Graham, B. 1949. The intelligent investor, a book of practical counsel. New York: Harper.
Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. 1998. Multivariate data analysis (5th ed). Upper Saddle River, NJ: Prentice Hall.
88
Hall, R. W. 1983. Zero inventories. Homewood, IL: Business One Irwin. Heard, E. 1995. Cycle time - A cash machine for companies! Unpublished. Hendricks, K. B. & Singhal, V. R. 1996. Quality awards and the market value of the firm:
An empirical. Management Science, 42(Issue 3): 415. Heskett, J. L., Jones, T. O., Loveman, G. W., Sasser, W. E. J., & Schlesinger, L. A. 1994.
Putting the service-profit chain to work. Harvard Business Review, 72(2): 164-174.
Howton, S. D., Higgins, E. J., & Biggart, T. B. 2000. The information content of just-in-time inventory system adoption announcement. Journal of Business Finance & Accounting, 27(5/6): 711.
Huson, M. & Nanda, D. 1995. The impact of just-in-time manufacturing on firm performance in the US. Journal of Operations Management, 12(3/4): 297.
Inman, R. A. & Mehra, S. 1993. Financial justification of JIT implementation. International Journal of Operations & Production Management, 13(4).
Jacobs, E. E. 2001. Handbook of U.S. labor statistics (5th ed): Bernan Press. Kaplan, R. S. 1983. Measuring manufacturing performance: A new challenge for
managerial accounting research. The Accounting Review, 58(4): 686-705. Keppel, G. 1991. Design and analysis : A researcher's handbook (3rd ed). Englewood
Cliffs, NJ: Prentice Hall. Kinney, M. R. & Wempe, W. F. 2002. Further evidence on the extent and origins of JIT's
profitability effects. Accounting Review, 77(1): 203. Kirk, R. E. 1995. Experimental design: Procedures for the behavioral sciences (3rd ed).
Pacific Grove, CA: Brooks/Cole. Koufteros, X. A., Vonderembse, M. A., & Doll, W. J. 1998. Developing measures of
time-based manufacturing. Journal of Operations Management, 16(1): 21-41. LexisNexis.
Lexis/Nexis, Reed Elsevier Inc, 2004. MacDuffie, J. P. 1995. Human resource bundles and manufacturing performance:
Organizational logic and flexible. Industrial & Labor Relations Review, 48(2): 197.
Mauri, A. J. & Michaels, M. P. 1998. Firm and industry effects within strategic management: An empirical examination. Strategic Management Journal, 19(Issue 3): 211.
McDougall, D. C. 1988. Learning the ropes: How to tell when you've found an effective performance measurement system for manufacturing. Operations Management Review, (Fall 1987 & Winter 1988): 38-48.
Meyer, A. D., Tsui, A. S., & Hinings, C. R. 1993. Configurational approaches to organizational analysis. Academy of Management Journal, 36(6): 1175-1195.
Miller, D. 1986. Configurations of strategy and structure: Towards a synthesis. Strategic Management Journal, 7: 233.
Miller, D. 1996. Configurations revisited. Strategic Management Journal, 17(7): 505-512. Miller, J. G. & Roth, A. V. 1994. A taxonomy of manufacturing strategies. Management
Science, 40(3): 285-304. Minitab Release 14.1. 2003. Minitab Inc.
89
Moch, M. K. 1976. Structure and Organizational Resource Allocation. Administrative Science Quarterly, 21(Issue 4): 661.
Monden, Y. 1983. Toyota production system: Practical approach to production management. Norcross, GA: Industrial Engineering and Management Press Institute of Industrial Engineers.
Noble, M. A. 1995. Manufacturing strategy: Testing the cumulative model in multiple country context. Decision Sciences, 26(5): 693-721.
Nunnally, J. C. 1978. Psychometric theory (2d ed). New York: McGraw-Hill. Ono, T. 1988. Toyota production system: Beyond large-scale production. Cambridge,
MA: Productivity Press. Peterson, P. P. 1989. Event studies: A review of issues and methodology. Quarterly
Journal of Business & Economics, 28: 36: College of Business Administration/Nebraska.
Sakakibara, S., Flynn, B. B., Schroeder, R. G., & Morris, W. T. 1997. The impact of just-in-time manufacturing and its infrastructure on manufacturing performance. Management Science, 43(9): 1246-1257.
Schmenner, R. W. & Swink, M. L. 1998. On theory in operations management., Journal of Operations Management, Vol. 17: 97-113.
Schonberger, R. 1982. Japanese manufacturing techniques: Nine hidden lessons in simplicity. New York: Free Press.
Schonberger, R. 1986. World class manufacturing: The lessons of simplicity applied. New York, London: Free Press; Collier Macmillan.
Schonberger, R. 1996. World class manufacturing: The next decade: Building power, strength, and value. New York: Free Press.
Siegel, S. & Castellan, N. J. 1988. Nonparametric statistics for the behavioral sciences (2nd ed). New York: McGraw-Hill.
Shah, R. 2002. A configurational view of lean manufacturing and its theoretical implications. Unpublished Dissertation, The Ohio State University, Columbus, OH.
Shah, R. & Ward, P. T. 2003. Lean manufacturing: Context, practice bundles, and performance. Journal of Operations Management, 21(2): 129-150.
SPSS. 2002. SPSS for Windows, 11.5 ed: SPSS, Inc. Standard & Poor's. 2003. Standard & Poor's register of corporations, directors and
executives. New York,: Standard & Poor's Corp.
Upton, D. 1998. Just-in-time and performance measurement systems. International Journal of Operations & Production Management, 18(11/12): 1101.
Vick, T. P. 2001. How to pick stocks like Warren Buffett: Profiting from the bargain hunting strategies of the world's greatest value investor. New York: McGraw-Hill.
Varaiya, N., Kerin, R. A., & Weeks, D. 1987. The Relationship between Growth, Profitability, and Firm Value. Strategic Management Journal, 8(5): 487-497.
Ward, P. T., Bickford, D. J., & Leong, G. K. 1996. Configurations of manufacturing strategy, business strategy, environment, and structure. Journal of Management, 22(4): 597-626.
Wiens, E.; Cobb-Douglas Production Function; http://www.egwald.com/statistics/restrictedcobbdouglas.php3; Oct 11, 2003.
90
Wharton Research Data Services; http://wrds.wharton.upenn.edu/home/index.shtml; Oct 2003.
Womack, J. P., Jones, D. T., Roos, D., & Massachusetts Institute of Technology. 1990. The machine that changed the world: Based on the Massachusetts Institute of Technology 5-million dollar 5-year study on the future of the automobile. New York: Rawson Associates.
Womack, J. P., Jones, D. T., & Roos, D. 1991. The machine that changed the world: The story of lean production (How Japan's secret weapon in the global auto wars will revolutionize western industry) (1st HarperPerennial ed). New York, NY: HarperPerennial.
Womack, J. P. & Jones, D. T. 1996. Lean thinking: Banish waste and create wealth in your corporation. New York: Simon & Schuster.
Woo, C. Y. 1984. An Empirical Test of Value-Based Planning Models and Implications. Management Science, 30(9): 1031-1050.
91
APPENDIX B
Complete implementationExtensive implementation
Some implementationLittle implementation
No ImplementationSupplier Communication
SJT3 Suppliers are directly involved in the new product development process 1 2 3 4 5SCM2 We give our suppliers feedback on quality and delivery performance 1 2 3 4 5SCM3 We require major suppliers to contribute to cost and quality improvement efforts 1 2 3 4 5
Complete implementationExtensive implementation
Some implementationLittle implementation
No ImplementationJust-in-time (JIT) Production Methods
SJT1 Our key suppliers manage inventory in our facilities 1 2 3 4 5JIT1 We use a "pull" production system 1 2 3 4 5JIT2 Production at stations is "pulled" by the current demand of the next station 1 2 3 4 5JIT3 We use Kanbans, squares, or containers as signals for production control 1 2 3 4 5JIT4 Production is "pulled" by the shipment of finished goods 1 2 3 4 5
Statistical Process Control (SPC)SPC1 Equipment and processes on the shop floor are currently under SPC 1 2 3 4 5SPC2 Statistical techniques are used to reduce process variance 1 2 3 4 5SPC3 Charts showing defect rates are used as tools on the shop-floor 1 2 3 4 5SPC4 We use fishbone-type diagrams to identify causes of quality problems 1 2 3 4 5
Employee Involvement in Problem SolvingEMP1 Shop-floor employees are key to problem solving teams 1 2 3 4 5EMP2 Shop-floor employees drive suggestion programs 1 2 3 4 5EMP3 Shop-floor employees lead product/process improvement efforts 1 2 3 4 5
Group TechnologyGT1 Products are classified into groups with similar processing requirements 1 2 3 4 5GT2 Products are classified into groups with similar routing requirements 1 2 3 4 5GT3 Equipment is grouped to produce a continuous flow of families of products 1 2 3 4 5GT4 Families of products determine our factory layout 1 2 3 4 5
Total Productive Maintenance (TPM)TPM2
1 2 3 4 5TPM4 1 2 3 4 5TPM5 We do preventive maintenance 1 2 3 4 5
Complete implementationExtensive implementation
Some implementationLittle implementation
No ImplementationCustomer Involvement
CUS1 Our customers are actively involved in current and future product offerings 1 2 3 4 5CUS2 Customers frequently share current & future demand information with marketing 1 2 3 4 5CUS3 Our customers give us feedback on quality and delivery performance 1 2 3 4 5
Survey of Manufacturing Practices
SECTION III: The following questions are associated with your product line's operating practices with respect to key customers. Please indicate the extent of implementation of each of the following practices in your
operations for this product line.
SECTION I: The following questions are associated with your product line's operating practices with respect to key suppliers. Please indicate the extent of implementation of each of the following practices in your
operations for this product line.
SECTION II: The following questions are associated with your product line's operating practices. Please indicate the extent of implementation of each of the following practices in your operations for this product line.
Records of routine maintenance are kept
We emphasize good maintenance as a strategy for achieving quality and schedule compliance
92
APPENDIX C EXAMPLE INITIAL SOLICITATION LETTER
[An endorsement letter from the Center for Excellence in Manufacturing Management at Fisher College will also
be attached.]
Highlighted fields are customized using the existing research database.
August 3, 2003
Ms. Ellen Bewell – Director of Manufacturing XYZ Company 105 Corporation Place Mapleton, Iowa 47853
Dear Ms. Bewell:
I am writing to ask your help in a study of manufacturing firms in North America. This study is part of an effort to understand the relationship between manufacturing practices and business level financial performance. As Director of Manufacturing, you have the overall knowledge of your company’s operating practices that will allow you to make a valuable contribution to this study. XYZ Company was one of approximately 400 manufacturing firms selected for this study. The criteria for selection was the availability of public financial accounting data on your firm in Standard and Poor’s Compustat database and the fact that the database reports that you primarily operate in one product area, namely the dairy products industry. In fact, I have already completed the first phase of the research, which includes a comparative analysis of your business and operations financial performance against the other manufactures in the study. In order to control for differences in industry and size, your company’s performance on each dimension was compared to a select portfolio of companies of similar industry and asset size. My analysis shows that your aggregate business performance ranks you number 54 (1 being the best) and your operations performance ranks you number 106. Your relative performance, expressed as a percentile (1 meaning the top 1% of performers), on the individual dimensions included to derive the ranking is as follows:
Business: Ranked 54
Return on equity: 94% Sales Growth: 90% Stock Return*: 50%
Operations: Ranked 106 Return on cash adjusted assets: 94% Employee productivity: 72%
93
Cost efficiency: 11% Cash-to-cash cycle time: 95%
* 50% used if stock return data was not available. This study seeks to understand how your manufacturing and operation practices contribute to your performance. Results from this survey will be used to analyze which practices are most valuable to a manufacturing firm and in what competitive situations. The survey utilizes simple 5-point scale questions to assess the extent to which a variety of what are currently believed to be “best practices” in operations are used in your operations. This study is intended solely for the purpose of furthering knowledge in the field of manufacturing. By better understanding which practices are contributing most to the bottom line, companies can make better decisions of where to invest their limited time and other resources. Your answers are confidential and will be released only as summaries in which no individual’s answers can be identified. When you visit the survey web site http://www.research-center.info/ your name will be deleted form the mailing list. This survey is voluntary. You can help by taking a few minutes to share your knowledge of your manufacturing operations. If for some reason you prefer not to respond please let me know by visiting the web site and checking “decline” on the introduction page. As an additional incentive for visiting the web site and completing the survey, the web site has a listing of all the 600+ companies in the study and their respective performance ratings. If you have any questions or comments about this study, I would be happy to talk with you. My number is 614 292-4136 or you can e-mail me at [email protected] . Thank you very much for helping with this important study. Respectfully yours,
Eric O. Olsen PhD Student Dr. Peter Ward – Principal Investigator (614) 292-5294 [email protected] Fisher college of Business, 600 Fisher Hall, 2100 Neil Ave, Columbus, OH 43210
P.S. If by some chance, I have made a mistake and you are not the most appropriate person at XYZ Company to respond to this survey, please forward it to the proper person.
94
APPENDIX D EXAMPLE FOLLOW-UP LETTER
[An endorsement letter from the Center for Excellence in Manufacturing Management at Fisher College will also
be attached.]
Highlighted fields are customized using the existing research database.
August 25, 2003 Jane Bewell - VP of Graduate Students Formal Company Name 2003 Olsen Drive Columbus, OH 43210 Dear Ms. Bewell: A couple weeks ago, I wrote asking for your help in a study of manufacturing practices and financial performance. If you have responded already, you have my thanks. If, however, you have not yet taken the time to complete the survey, please take a moment to do so now or pass this along to the most appropriate person in your organization. To complete the survey, go to www.research-center.info and enter pass code A26. Your answers will be kept strictly confidential and will be released only in aggregate form. This survey is voluntary. You can help by taking a few minutes to share your knowledge of Informal Company Name’s manufacturing operations. If you prefer not to respond, please let me know by checking “decline” on the introduction page. The first phase of this research involved a comparative analysis of the financial performance of your business and operations against 377 other manufactures in the study. Informal Company Name’s performance was compared to companies of similar industry and asset size, and your relative performance is reported on the following page. Data from this survey will be analyzed to determine which practices are most valuable to a manufacturing firm and under which competitive circumstances. By better understanding which practices are contributing most to the bottom line, companies can make better decisions of where to invest their limited time and other resources. As an incentive for completing the survey, the web site lists all 378 companies in the study and their respective performance ratings. If you have any questions or comments about this study, I would be happy to talk with you. My number is 614 292-4136 or you can e-mail me at [email protected] . Thank you very much for helping with this important study.
95
Respectfully yours, Eric O. Olsen Ph.D. Student Dr. Peter Ward – Principal Investigator (614) 292-5294 [email protected] Fisher college of Business, 600 Fisher Hall, 2100 Neil Ave, Columbus, OH 43210
[Next page]
Informal Company Name’s PERFORMANCE ANALYSIS
Business: Ranked 54 of 378 Performance Percentile (1 = Top 1% of performers)
Return on equity: 94 Sales Growth: 90 Stock Return*: 50
* 50 used if stock return data was not available. Operations: Ranked 106 of 378
Performance Percentile (1 = Top 1% of performers)
Return on cash adjusted assets: 94 Employee productivity: 72 Cost efficiency: 11 Cash-to-cash cycle time: 95
Note: Data for all the companies include in this analysis may be obtained by completing
the survey of manufacturing practices at the following web address:
www.research-center.info
Informal Company Name’s pass code is A26
More information on this research project may be obtained by contacting:
Eric Olsen [email protected] The Ohio State University
http://fisher.osu.edu/~olsen_54/ Fisher Hall, rm 251A office: 614 292-4136 2100 Neil Avenue fax: 614 292-1272 Columbus, OH 43210
96
APPENDIX E SAMPLE OF RANKINGS AND PERCENTILE PERFORMANCE RATING VIEWABLE UPON COMPLETION OF THE WEB-BASED SURVEY BY
RESPONDENTS
Performance Percentiles and Overall Performance Rankings for Companies Solicited for Participation in this Study
Percentile: 1 = Top 1% of Performers Rank: 1 = Best Percentile: 1 = Top 1% of Performers Rank: 1 = Best
Company Name
Return on Equity
PercentileSale Growth Percentile
Stock Return Percentile (4)
Business Performance
Rank
Return on Cash Adjusted
Assets Percentile
Employee Productivity Percentile
Cost Efficiency Percentile
Cash to Cash Cycle Time Percentile
Operations Performance
Rank
AAON INC 4 8 50 26 3 25 92 4 60ABAXIS INC 57 14 50 123 47 49 92 13 196ABIOMED INC 83 80 50 316 96 92 14 57 282ACRES GAMING INC 88 44 50 268 55 55 17 58 160ACTEL CORP 14 59 50 124 10 1 1 71 21ADAPTIVE BROADBAND CORP. 87 100 99 377 96 94 56 92 369ADVANCED PHOTONIX INC 63 56 50 242 31 39 24 60 105ADVANCED TISSUE SCI 85 43 12 181 30 25 45 59 114AEROSONIC CORP 49 38 50 167 42 45 42 90 217AEROVOX INC 81 76 81 340 79 78 78 18 277AETRIUM INC 58 85 97 346 60 63 36 83 257ALANCO TECHNOLOGIES INC 61 40 50 206 54 62 22 42 162ALARIS MEDICAL INC 99 72 50 323 74 65 28 72 251ALKERMES INC 84 31 31 211 80 75 84 50 330ALLIANCE PHARMACEUTICAL CP 88 95 50 338 72 89 64 21 267ALLIED HEALTHCARE PRODS INC 79 91 50 321 57 56 82 61 276ALTERA CORP 12 22 50 45 10 2 25 70 39AMARIN CORP PLC 73 39 50 233 12 2 46 17 18AMERICAN MEDICAL TECHNOL INC 38 89 17 195 45 36 16 89 167AMERICAN SCIENCE ENGINEERING 50 33 50 152 45 41 72 40 197AMERICAN TECH CERAMICS CORP 25 37 50 103 28 45 39 86 191AMERICAN VANGUARD CORP 11 49 50 99 20 11 52 69 100 Included at the bottom of the table:
Notes: 1. These relative performance ratings are to be used for academic research only. They are to be distributed only with the permission of the author (Eric Olsen).
2. Relative ratings are applied across industries and firm sizes. To be most meaningful results should be compared between firms of similar size competing in similar industries.
3. A more detailed explanation of how relative performance ratings are calculated and may be interpreted is available to survey participants in the form of annotated results. Other interested parties may contact the principle researcher, Eric Olsen.
4. Stock return percentile performance was set to 50% if stock return data were not available.
