if

SLIITYBRAR

Illlll'illlllllnwlmmluluwm

3 1293 01016 8213

This is to certify that the

dissertation entitled

IDENTIFYING THE DETERMINANTS OF A KAIZEN-SUGGESTION
SYSTEM AND ASSESSING ITS IMPACT ON PLANT-LEVEL
PRODUCTIVITY: A POOLED CROSS-SECTIONAL AND TIME
SERIES ANALYSIS

presented by
Wen-Jeng Lin

has been accepted towards fulfillment
of the requirements for

doctoral degree in Jghilosoph'y
(Industrial Relations
and Human Resources)

 

 

 

 

Date 5&4:-

MSU i: an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

 

LIBRARY
Michigan State
Unlverslty

 

 

 

PLACED! RETURN BOXtonmavomhdnckomtmmywrocord.

TO AVOID FINES Mum on or Moro dd. duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

IDENTIFYING THE DETERMINANTS OF A KAIZEN-SUGGESTION SYSTEM AND
ASSESSING ITS IMPACT ON PLANT-LEVEL PRODUCTIVITY: A POOLED
CROSS-SECTIONAL AND TIME SERIES ANALYSIS

BY

WEN-JENG LIN

A DISSERTATION

SUBMITTED TO
MICHIGAN STATE UNIVERSITY
IN PARTIAL FULFILLMENT OF THE REQUIREMENT
FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

SCHOOL OF LABOR AND INDUSTRIAL RELATIONS

APRIL 1995

ABSTRACT

IDENTIFYING THE DETERMINANTS OF A KAIZEN-SUGGESTION SYSTEM
AND ASSESSING ITS IMPACT ON PLANT-LEVEL PRODUCTIVITY: A
POOLED CROSS-SECTIONAL AND TIME-SERIES ANALYSIS
BY

Wen-Jeng Lin

Kaizen has been viewed as the key to Japanese competitive
success. Kaizen—suggestion systems have thus drawn heightened
research interest. Unfortunately, however, there is a paucity
of studies that have evaluated.kaizen-suggestion.systems. This
paper is one step toward increasing understanding of kaizen-
suggestion systems. It serves as an exploratory effort to
examine the determinants of suggestions made in the kaizen-
suggestion system and the impact of adopted suggestions on
organizational effectiveness.

In the determinant level, the empirical results provide
initial support that cumulative experience in intangible
suggestion making (i.e., those suggestions where dollar
savings cannot be estimated), management training and top
management participative style played crucial roles in
determining tangible suggestion making (i.e., those
suggestions where dollar savings can be estimated).

In the outcome level, the findings are that there is an
accumulative effect of suggestions successive incremental
improvements in productivity, labor efficiency and product
qualityu A. lagged, effect exists between suggestion

implementation and economic gains. Improvements in

productivity, labor efficiency' and. quality' are not only
dependent on the present volume of suggestions accepted but
dependent on the past volume of suggestions adopted. However,
there is a different pattern of delayed effect over time
between tangible and intangible suggestions.

Overall tangible suggestions have longer lag structure
than intangible suggestions. That is, economic gain responses
to tangible suggestions generally last longer than do response
to intangible suggestions. There are also different effects on
productivity, labor efficiency and quality between the two
types of suggestions. Tangible suggestions have a greater
effect on productivity gains and high labor efficiency but
have a smaller effect on product quality improvement. In
contrast, intangible suggestions have a larger effect on
product quality improvement but have a smaller effect on both
productivity and labor efficiency improvements.

This study has highlighted the continuous and incremental
improvement with what Imai (1986) termed "Japanese competitive
success." The effects of tangible and intangible suggestions
occur gradually and as a continuous incremental process.
Although the improvements are subtle in the short term, when
sustained over time, the long term improvements are

considerable.

ACKNOWLEDGMENTS

I wish to express deep appreciation to Dr. Joel Cutcher-
Gershenfeld, my mentor and committee chairperson, for his
continuous and patient support. He introduced me to the area
of Japanese management practices and kaizen—suggestion system.
His guidance and.scholarly example during the three-year Japan
Project--the Seeds of Change--will have a profound impact on
my future research and career.

I am also grateful for the guidance received from Dr.
Karen Roberts, Dr. Ellen Kossek, Dr. Kevin Ford and Dr.
Georgia Chao. I also extend my thanks to Dr. Ed. Montemayor
for his help and guidance in using statistical software.

I anlgreatly indebted to Nippondenso, U.S. Inc. and those
who work in the company, Michael Gagnon, Sue Flees, Doug
Butler, Roy Roemen and Mark Herzing, for their assistance in
providing data and practical guidance for my research.

I also would like to acknowledge all members of the
Japanese research team—-the Seeds of Change--at the School of
Labor and Industrial Relations, Michigan State University.
This dissertation emerged out of parallel field research of
this research team. Thus this dissertation is an extension of
this team's aims and spirit. Special thanks are accorded to
Betty Barrett, Takashi Inaba and Arthur Wheaton for their
personal advice and encouragement.

Finally, to my parents, wife Ming—Lin, daughters Yen-Ju

iv

v
and Yi-Jeng I owe the deepest appreciation and gratitude.

Their sacrifices have made my education possible.

TABLE OF CONTENTS

CHAPTER ONE: INTRODUCTION AND OVERVIEW

The Basic Concepts of Kaizen, Suggestion
and Productivity

Importance of the Study
Aims of the Study
Contributions

Outline of this Dissertation

CHAPTER TWO: METHODOLOGY

The Research Site and Background of
the Kaizen-Suggestion System

Data Source and Collection
Method of Data Analysis

Operationalization of Variables

CHAPTER THREE: IDENTIFYING THE DETERMINANTS OF

SUGGESTION’MAKING

Introduction
Relevant Literature Review

The Conceptual Model the Determinants of
Effective Kaizen-Suggestion System

The Operational Model and Hypotheses
Research Models

Empirical Results and Implications

vi

13

21

23

24

26

26

29

32

39

44

44

47

59

6O

72

73

vii

CHAPTER FOUR: ASSESSING THE IMPACT OF THE KAIZEN-
SUGGESTION SYSTEM ON PLANT-LEVEL
PRODUCTIVITY

Introduction
Literature Review

The Conceptual Model of Effect of the
Kaizen Suggestion System on Productivity

Hypotheses
Research Models

Empirical Results and Implications

CHAPTER FIVE: CONCLUSIONS AND SUGGESTIONS

Appendix A: The Selection Process of the Optimal
Pattern of Lagged Intangible Suggestion
Variables with Tangible Suggestion As
Dependent Variable

Appendix B: The Selection Process of the Optimal
Pattern of Lagged Training Variables with
Tangible Suggestion As Dependent Variable

Appendix C: The Selection Process of the Optimal
Pattern of Lagged Training Variables with
Intangible Suggestion As Dependent Variable

Appendix D: Hypotheses of Control Variables for
Productivity and Labor Efficiency Models
(Chapter Four)

Appendix E: Hypotheses of Control Variables for
Quality Model (Chapter Four)

Appendix F: Relationships Between Suggestions,

Control Variables, and Organizational
Effectiveness and Efficiency Measures

LIST OF REFERENCES

93

93

94

114

126

139

141

173

190

191

192

193

194

195

196

Table 1-1

Table 1-2

Table 1-3

Table 1-4

Figure 3-1

Figure 3-2

Table 3-1

Table 3-2

Table 3-3

Table 4-1

Table 4-2

Table 4-3

Table 4-4

Figure 4-1

Table 4-5

LIST OF TABLES AND FIGURES
The Difference Between Japanese and
American Suggestion System 6

Core Assumptions of Kaizen-Oriented
and Traditional Suggestion Systems 7

The Difference Among Kaizen,
Organizational Development, and
Reengineering Approaches 10

Annual Percent Changes in Manufacturing
Productivity, 1960-1992 14

A Conceptual Model of the Determinants
of Effective Kaizen-Suggestion System 59

An Operational Model of the Determinants
of Effective Kaizen-Suggestion System 62

Means, Standard Deviations, and
Correlation Matrix 75

The Determinants of Tangible

Suggestion Making 80
The Determinants of Intangible

Suggestion Making 84
The Impact of the Suggestion Programs

on Productivity 96
The Impact of QCs on Productivity 101

The Effects of Financial Involvement
Programs on Productivity 106

The Effects of Worker Participation on
Productivity 109

A Conceptual Model of Effects of the
Kaizen-Suggestion System on
Productivity 118

Relationships Between Tangible

Suggestions, Intangible Suggestions,

and Organizational Effectiveness and
Efficiency Measures 143

viii

Table

Table

Table

Table

Table

Table

Table

4-8

4-9

4-10

ix

Results of Regression Analyses of the
Effect of Suggestions and All Other
Variables on Productivity

Results of Regression Analyses of the
Effect of Suggestions and All Other
Variables on Labor Hours (A Three-
Month-Lag Model for Suggestion
Variables)

Results of Regression Analyses of the
Effect of Suggestions on Labor Hours
(Lagged Three Months)

Results of Regression Analyses of the
Effect of Suggestions on Labor Hours
(Lagged Four Months)

Results of Regression Analyses of the
Effect of Suggestions and All Other
Variables on Labor Hours (A Four-
Month-Lag Model for Suggestion
Variables)

Results of Regression Analyses of the
Effect of Suggestions and All Other
Variables on Product Quality

Results and Policy Implications for
Tangible and Intangible Suggestions

145

150

152

153

156

160

166

CHAPTER ONE

INTRODUCTION AND OVERVIEW

Kaizen has been viewed as the key to Japanese competitive
success (Imai 1986; Yasuda 1991; Japan Human Relations
Association 1992; Japanese Human Relations Association 1988).
As Imai put it:

Kaizen strategy is the single most important concept in

Japanese management--the key to Japanese competitive

success, ......... ,a strategy to cope with the challenges

of the 1980s, 1990s, and the beyond ........ ,Japanese

Companies have successfully designed, manufactured, and

marketed competitive products using kaizen strategy

(1986, Pp XXIX-XXXI).

The kaizen-suggestion systems also meet employees' expectations
of involvement in organizational decision making (Lawler,199l)
because the kaizen-suggestion can serve as a form of
communication (Miner 1969; Kossen 1983; French 1984; Klotz
1988) or a form of employee involvement (French 1984; Mattes
1992) . In recent years, the topic of Japanese management
practices in general and kaizen-suggestion system in
particular have received much attention. Unfortunately,
however, there is a paucity of studies that have evaluated
either the determinants or the outcomes of kaizen-suggestion
systems. Research that has been.done has either theoretical or
methodological weaknesses that may limit understanding of the

nature of the kaizen-suggestion system. This dissertation

attempts to fill these gaps and make such a contribution.

2

The concept of a kaizen—suggestion system is not only
important in the implicationszof continuous improvement itself
but also has links in the literature on productivity, quality
and employee involvement. It is an important phenomena, in
part, because of the way it integrates across all three areas.
Imai(1986) has suggested that improved productivity and
quality are two major outcomes of kaizen activity. Whenever
and wherever improvements are made in companies, these
improvements are eventually going to lead to improvements in
such areas as quality and productivity. Further, a kaizen-
suggestion system can serve as a form of communication or a
form of employee involvement. The kaizen-suggestion plans give
employees opportunities to participate in company matters and
decision making.

This chapter provides an introduction.and.overviewrof the
entire dissertation. The introduction and overview include:
(1) the basic concepts of kaizen, suggestion and.productivity,
(2) why it is of importance to examine the kaizen-suggestion
system, (3) aims of the study, (4) the potential limitations
of this study, and (5) an outline of the subsequent chapters

in the dissertation.

The Basic Concepts of the Kaizen, Suggestion
and Productivity

Kaizen

The Japanese term "kaizen" originally came from the

3
Chinese. The word "kai" means change or correct, whereas "zen"
means good.or satisfactory; Everything changed fromtthe status
quo to a better situation or correction of mistakes or errors
means "kaizen."

The Japanese term "kaizen" means continuous improvement
in day-to-day life. Imai gives us a clearer picture of the
meaning of kaizen in industrial organizations. Imai (1986,
p.25) suggests that there is one major difference between
kaizen and innovation. Innovation usually calls for a sudden
change and this may require a considerable increase in
investment. In contrast, kaizen occurs gradually and as a
continuous incremental process. Everyone is involved in the
process of change since the change process of improvement
itself originates through discussions. Therefore, kaizen is an
umbrella idea covering most "uniquely Japanese" practices such
as suggestion systems, TQC (total quality control), QC
circles, TPM (total productive maintenance), just-in-time
delivery, zero defects, and productivity improvement. Lillrank
and Kano (1989) also provide a precise definition of kaizen
that is characterized by 1) improvement that combines both
innovation and maintenance, 2) improvement that takes place in
small steps, 3) improvement that involves everyone, and 4)
improvement that emphasizes the production process (p.28).

One of the most important features of kaizen is process-
oriented thinking rather than result-oriented thinking. It

does not mean that the results are not important. The logic

4

here is that once the process is improved, a better result
always accompanies it. By emphasizing concern with the work
process as much as with the result, a kaizen approach can help
management and workers be good problem-solvers when things go
wrong and better problem-solvers at other times. Employees
under the kaizen concept are concerned with how to get things
done, whether it can be done better, and whether other ways
have been tried. Once a task is completed, the questioning
keeps on going. The focus shifts from what was performed to
how it was performed, how it can be done more accurately and
more efficiently next time.

A good example of process-oriented management is that in
Japanese companies in general and at the research site in this
study (Nippondenso, U.S.) in particular, you can find charts
and graphs posted all over the place, measuring production,
product.defects, suggestions adopted” accident rate, training,
skill improvement, and so on. Every team and department has
its own charts and graphs, which are updated regularly. Team
members usually are responsible for creating these charts and

graphs.

Employee Suggestion System

As mentioned by Imai, suggestion systems are the
centerpieces of kaizen practices. The suggestion system is an

integral part of individual-oriented kaizen. It is a vehicle

5
for carrying out the maxim that one should work smarter, not
harder (Imai 1986, pp 110-111).

In Japan, suggestion systems are part of ongoing daily
company improvement efforts. The major goals of suggestion
systems are how to solve work problems and develop efficient
and effective work methods. Some examples are as follows:

Making the job easier and safer

Removing drudgery and nuisance from the job

Making the job more productive

Improving product quality

Eliminating waste in overproduction, transportation,

inventory, etc.

Under kaizen systems, employees typically make suggestions on
hOW’tO solve work.problems, develop efficient work methods, or
improve*working'conditions. Each.suggestion.is taken.seriously
and is followed up on.

Japanese managers practice kaizen by encouraging constant
incremental process improvements. They solicit and reward all
suggestions, home runs as well as singles. U.S. companies, on
the contrary, primarily look for the home run suggestion, the
one-time dramatic event. 1n: addition, the following data
provided by the Japan Human Relations Association also can
illustrate the difference between Japanese and American

suggestion systems:

6

Table 1-1: The Difference Between Japanese and American
Suggestion systems

Japan United States
1,936,738 # of potential participants 9,194,476
47,926,020 # of suggestions 1,246,749

24.70 # of suggestions per 0.14
person
$3.26 prize money per suggestion $416
$231,770 economic benefit produced $19,995
per each 100 potential
participants

Adopted from Rehfeld E. J. (1994). Alchemy of a Leader:
Combining Western and Japanese Management Skills to Transform
Your Company. New York: John Wiley & Sons. Inc.

A.multitude of minor problem-solving suggestions seems to mean
much more to employees than a handful of big cost-saving
ideas. Besides the different aims of the two systems, Table 1-
2 also shows other major differences in core assumptions

between kaizen-oriented and traditional suggestion systems.

 

Ehmh-unonm

Emmenmcoq

coNflHom mEfiB

 

oHQHmmmm nos: moofluocsm
Monuo >9 Umuhommsm ..
mummmamz Emumoum cofiummmmsm

mo mmfiuno now 0» HmummusH

mHODMEAUHOOU Emumonm
smNflmx >3 Umuhommdm in A.oum
.mnmmcfimcm .mooMHB coaaflxm
.Houmuwafiomm .mhmcmmq EmmB
.muoxnoz cofluosoonmv mofludo

now m.oco>H0>m ou HohmmusH

mmoHSOmmm cosmofloma

 

smuflsflq

mumHmEOO

coma uom>mamm Cu mmoood

 

Uflmcfihuxm SHHHmEHHm

UflmcflHuGH >aflnmefinm

mcum3mm

 

 

 

 

 

3mm has: ucoEm>Ho>cH mo mamom

mandcfl>flUnH mmsouw Hm>mq
mnoaumasnn¢_fiouuhm nnoaumfisnu4_aounhm

noauuommnm Honoauavnufi nofiumommam wounofiuouQONAaM Dana?

 

 

 

_fiouu>m n0aumommnm Honofluavsua was counoduounoufism

no mfiOauQfiflnmd ouoo “ana wanna

.mmcmazoca mo mcflmmwcumm wnu zmdoucB unoEo>0HmEH mooscflunou mcflUGMDmHmUnD
"cmuflmx can ucmEm>Ho>sH omonmEm .AwmmHV.Hm um caowcmnmumwuuonouso Eoum commend

 

=Go cod: HmummunH msoHDMHomO can“ newumummusH

 

mucm>o cmflmmcmm mucmsm>oumefi HmucmsonooH meooudo cmuowmxm

 

=mcsu neon mcwuufl£= =monnHm mowuuwnz Efid\mdoom

 

 

30am ummm GOADMDGoEoHQEH mo commm

 

 

 

 

Kaizen-Suggestion System

By integrating the two concepts of kaizen and suggestion
mentioned above, the kaizen-suggestion system can be defined
as a systemwide process designed to achieve continuous
incremental improvement in organizations by involving every
employee voluntary initiation of a steady stream of suggested
changes in daily operations, procedures and policies. This
definition helps to distinguish kaizen from other approaches
to organizational change and improvement, such as
organizational development (OD) and reengineering. Kaizen
stands in contrast to traditional OD approaches to change.
Organization development is a broad phenomenon involving great
diversity of planned interventions, including job redesign,
team building, survey feedback, employee involvement and so
on. OD is intended to "change the organization in a particular
direction, toward improved problem solving, responsiveness,
quality of work life, and effectiveness" (Cummings and.Worley
1993, p.3). Kaizen also stands in contrast to newer
reengineering approaches. Reengineering is tflua fundamental
rethinking and radical redesign of business processes to
achieve dramatic improvements in critical, contemporary
measures of performance, such as cost, quality, service and
speed (Hammer and Champy, 1993). Reengineering approaches
emphasize total rejection of the present system. In contrast,
the kaizen system begins with the present situation and

transforms it over time through small incremental improvement .

10
Table 1-3 presents the different features among kaizen, OD and
reengineering approaches. It may help us to better understand

the nature of different approaches to organization change.
Table 1-3: The Difference Among Kaizen, Organization Development and
Reengineering Approaches

Kaizen Organization Reengineering
Development

Focus *Processes *Processes/ *Results
Results
*People *Structure *Strategy
Aims/Impact Long-Term Medium-Term Short-Term
Payoff Payoff Payoff
Implementation Extended Moderate Quick
Time
Inhibits Short-Term Quick Payback Long-Term
Cost Savings Adaptability
Examples *Suggestion *Job Enrichment *Downsizing
Programs 1”Labor-Management *Restructuring
*QCs Cooperation

Organizational Productivity
In the last fifty years, the concept of productivity has
gradually come to be a key concept in business. Spurred by the

international concern, productivity has become a dominant

11
theme in assessing overall economic, organizational and
individual performance.
Productivity means different things to different people.
Different perceptions of productivity from an economist’s, an
industrial/organizational psychologist’s and a manager's

perspective will be discussed.

Economist's Perspective

Productivity is a function of all of the various inputs
to the production function which can be characterized as an
input/output relationship1 (Siegel 1980; Kendrick 1977;
Greenberg 1973; Mark 1971; International Labor Office 1969).
This efficiency-oriented definition derives from the basic
production function, which states that Q = F(K, L), where Q =
volume of output, K = capital inputs, L = labor inputs. There
are different ways of measuring productivity. The way in‘which

productivity’ is measured determines the meaning that it

 

1Siegel (1980)--Productivity is a family of ratios of (a)
quantity of output to (b) quantity of related resource input.
Kendrigk(l977)--Productivity is the relationship between
output and its associated inputs when the output and inputs
are expressed in real (physical volume) terms.

Greenberg (1973)--Measure of relationship between quantity of
resources used and quantity of output.

Mark (1971)--Productivity is loosely interpreted to be the
efficiency with which output is produced by the resources
utilized. A measure of productivity is generally defined as a
ration relating output (goods and services) to one or more of
the inputs (labor, capital, energy, etc.) which were
associated with that output.

International Labor Office (1969)--"The ratio between onxut

and input."

12
carries. There are two major methods by which productivity is
measured: (1) labor' productivity, (2) total factor
productivity. Basically, labor productivity is the ratio of
the quantity of output produced to the labor inputs (Stein
1971; Fourastie 1957). Total factor productivity is the ratio
of the quantity of output produced to a weighted combination
of the quantities of labor, capital and other resources that

produced it (Kendrick 1961; Solow 1957).

Industrial/Organizational Psychologist's Perspective

The literature fromtthelbehavioral sciences uses the term
productivity quite frequently. The I/O psychologists study
human behavior in a variety of forms of organizations and
social settings. For many researchers in this field, such
performance measures as personnel turnover, absenteeism,
accident rates, and grievances are considered productivity
criteria as much.as measures of production rate or quantity of
items pmoduced (Katzell, Bienstock, and Faerstein, 1977).
Usage of the term productivity'in.the more conventional sense,
such as labor productivity, physical output per worker, and
holding quantity constant (Work in America 1973; Sutermeister
1976), has grown parallel to the more flexible interpretation

of the meaning.

Manager's Perspective

For managers, productivity is important as a means of

13
organizational control. Management theorists argue that the
major function of management is the control of productivity.
Umstot, Bell.and Mitchel (1976) suggested that "concern for
productivity is still the dominant focus of managers" (p.379).
Beginning with Drucker (1954), productivity has been proposed
as an important element in the development of human resources.
This thinking continues today with dual emphasis on
productivity and the quality of work life (Hackman and Suttle,

1977) .

Importance of the Study

Why Kaizen-Suggestion Systems Meet Today’s Need
The Productivity and Quality Problems

Although several factors lead to economic decline, there
is general agreement that the productivity level, growth rate,
and quality of goods and services can all have a major
influence on the national economy in general and industrial
organizations in particular. The United States has the lowest
productivity growth rate among six major industrialized
countries in the past three decades. Table 1-4 combined from
twotdata sources illustrates a long-term trend of productivity
growth rate for six selected countries during the recent
period 1960 to 1992. In spite of its recent improvement (1991-
1992), the lag in productivity growth is putting the United

States behind five other industrialized countries.

14

Table 1-4: Annual Percent Changes in Manufacturing
Productivity, 1960-1992.

(Output per hour)

1960-1990 2 9 6.9 2.9 4.9 4.0 3 7
1960-1973 3 3 10 2 4.5 6.4 5.6 4 2
1973-1990 2.5 4 4 1.7 3 7 2 8 3.3
1973-1979 1 4 5.0 2.1 4.6 4 2 1.2
1979-1990 3.1 4.1 1.5 3.2 2.1 4.4
1906-1969 ' 3.1 4.1 1.5 3.2 2.1 4.4
1969-1990 2 5 3.6 1.3 1.1 4.5 0.9
1990-1991 1.9 4.3 0.6 -.1 3.0 3.9
1991-1992 4 3 -5.0 4 2 2.9 0 5 4.9

Source: (1) Dee! A.. and task. C. (1991). lanutacturing productivity
and labor costs in 14 economies. lonthly Labor Review. December. pp 24-37.

(2) fleet A.. lash. C.. and Sparks. C. (1993). International
comparisons of manufacturing unit labor costs. December. pp 47-58.

Quality is fast becoming one of the competitive issues of
the 1980s and 19908. Increased customer sensitivity gives it
new visibility. Pressures for improvement have become intense.

Juran (1981) presented his perception of the relative
quality of goods and services produced by the West and those
produced by Japan. Prior to World War II, American product
quality was regarded as the best, while Japanese product
quality was widely regarded as among the worst. During the
19808 this situation was reversed. Japan assumed a leading
position in global quality. A 1981 survey supported this
observation when it reported that nearly 50 percent of U.S.
customers felt the quality of American products had dropped

during the previous five years (Binstock, 1981). Another

15
survey which was conducted in the U.S. and Japan showed
similar results. Thirty percent of U.S. customers felt that
Japanese consumer products were of better quality than
American products, whereas only 21 percent of customers
responded that American consumer products were of better
quality than their counterpart’s. On the contrary, 54 percent
of Japanese customers felt that Japanese consumer products
were of better quality than American consumer products, while
only 16 percent of Japanese customers felt that American
consumer products were of better quality than Japanese
(Harrington 1991, p.3).

Slower productivity growth coupled with deteriorating
product quality has placed the United States at serious

competitive disadvantage in global markets.

The Expectations of Employee Involvement in
Decision Making

Lawler (1991) argued that employees have developed
expectations of participation in organizational decision
making due to increasing levels of education and democratic
philosophy in society. He argues that employees are
increasingly interested in advancing their legal and societal

rights to have a say in decisions in the organizations.

The Kaizen-Suggestion System: The Best Answer

As mentioned earlier, kaizen has been viewed as the key

to’Japanese competitive success (Imai 1986; Yasuda 1991; Japan

16

Human Relations Association 1992; Japanese Human Relations
Association 1988). Imai also.argues that improved productivity
and quality are two major outcomes of kaizen activity.
Whenever and wherever improvements are made in business, these
improvements are eventually going to lead to improvements in
such areas as quality and productivity. Quality and
productivity are usually improved through the elimination of
wastez, cost reduction, the introduction of new and more
competitive products, and by using and improving the latest
technology. Therefore, the kaizen-suggestion system can meet
businesses' needs for productivity and quality improvements.

Kaizen-suggestion systems also meet employees'
expectations of involvement in organizational decision making
because the kaizen-suggestion can be served as a form of
communication (Miner 1969; Kossen 1983; French 1984; Klotz
1988) or a form of employee involvement (French 1984; Mattes
1992). It provides a channel for employees to contribute their
ideas. Further, it provides an opportunity for employees to
show dissatisfaction with existing practices and procedures.
Reuter (1977) indicates the suggestion systems frequently
serve as a management safety valve that allows employees to

substitute a constructive solution to a problem that otherwise

 

2For'example, Nippondenso,U.S. and Toyota have identified
the seven types of waste: 1) waste from overproduction, 2)
waste of waiting time, 3) transportation waste, 4) processing
waste, 5) waste of motion, 6) waste from product defects, and
7) waste from inventory. The goal of the kaizen-suggestion
system is to eliminate these wastes which are generated in the
production processes.

17
might remain an unspoken complaint or irritation. Kaizen-
suggestion systems are also designed to provide a channel for
employees to communicate with. management and. provide an
opportunity for employees to share job-related technological
and administrative information for suggestion making.
Employees who make suggestions are given the responsibility to
carry them out and are encouraged, either individually or as
a team, to figure out how best to do so. Kaizen-suggestion
plans give employees opportunities to participate in company

matters and decision making.

Theoretical Issues

In recent years, the topic of Japanese management
practices in general and kaizen-suggestion programs in
particular have received much attention. Unfortunately,
however, there is a paucity of studies that have evaluated
either determinant .level or outcome level of the kaizen-
suggestion system. What research has been done has either
theoretical or methodological weaknesses that may limit
understanding of the nature of kaizen-suggestion system. Some
papers were presented as case studies (Smith 1968; Marmaduke
1946; Greenlaw 1980; Gunch 1991; Matthes 1992; Reuter 1976,
1977). Some studies broadly introduced the benefits of
suggestion programs in big organizations (Graf 1982; Wilce
1971). Because of the lack of theoretical or statistical

analyses, some important questions about the determinants and

18

effects of the kaizen-suggestion system remain unanswered.

Measurement/Methodoloqical Issues

Kaizen is culturally based and not easily understood by
non-Japanese. The problem is that we always have difficulty
understanding how to measure the results of kaizen and what
the best indicators of kaizen are. In.day-to-day working life,
managers speak of the kaizen of organizational performance;
workers speak of the kaizen of working conditions; industrial
engineers speak of the kaizen of operational efficiency. What
are the ultimate indexes that serve as a measure for checking
the results in an organization? What kind of index can
illustrate for us how things have been improved?

Another issue is measurement problems in suggestion
making. Typically, employee suggestion contributions are
measured by the volume of suggestions. This reflects the
quantity of suggestions rather than the quality of
suggestions. For example, the sum of ten "tiny" suggestions
may have fewer economic benefits than one "valuable"
suggestion. None of the existing studies has differentiated
the volume of suggestions from the quality of suggestions. In
addition, the differentiation. of 'tangible and intangible
suggestions is also important for research. A tangible
suggestion is a suggestion for which economic benefits or
dollar savings can be estimated. That is, cost reduction

results from suggestions that lead to tangible financial

19
benefits such as savings in labor, material, machine hours,
downtime, waiting time and so on. An intangible suggestion is
defined as a suggestion where economic benefits or dollar
savings cannot be estimated.

Those suggestions involving issues of safety, quality or
housekeeping' are all typical intangible suggestions.
Traditionallyy U.S. companies consider" only' tangible
suggestions. Intangible suggestions are a matter of no
important. Actually, however, a multitude of minor suggestions
may mean much more to the employees than a handful of big
ideas. Employees' high level needs can be satisfied and their
problem-solving skills can be improved through creation and
implementation of numerous intangible suggestions. In
addition, a tangible suggestion that may result in huge cost
savings for an organization could be generated by employee who
has had.proficient and rich experiences in.making’ a number of
intangible suggestions. That is because cumulative
proficiency, rich experiences, skills, and knowledge in making
intangible suggestions are useful for the employee to‘generate
tangible suggestions that are usually' more difficult to
generat and implement than intangible suggestions. None of the
existing studies has differentiated tangible suggestions from
intangible suggestions. Yet they shouLd not be ignored in
kaizen-suggestion research.

Productivity is typically defined as a ratio of outputs

to inputs (Siegel 1980; Kendrick 1977; Greenberg 1973; Mark

20

1971; International Labor Office 1969). Outputs are goods and
services produced by an organization. Inputs include labor,
capital, materials, and energy. The most common productivity
indicators are really measures of labor productivity. It can
be regarded as a "hard" measure of organizational performance.
However, few of the existing studies either in employee
suggestion research or in employee involvement research have
adopted a "hard" index of productivity. On the contrary, the
use of "soft" measures of productivity in terms of job
satisfaction or employee morale in these research areas is
prevalent. Much of the research concerned with cost reduction
or cost savings claimed. for' employee suggestion implies
productivity improvements. These measurement problems in
kaizen, suggestion and productivity should be examined and
resolved in a comprehensive study.

Methodology is another issue that is a matter of concern
in the existing studies. Cross-sectional analysis is the only
method used in employee suggestion research. It raises the
following questions. If I assume that a tangible suggestion
is associated with employees’ experiences and contributions in
making intangible suggestions, time is a critical variable.
Again, if we assume productivity not only depends on the
present volume of suggestions accepted, but depends on past

volume of suggestion accepted, time is a crucial factor too3.

 

3More detailed hypotheses will be discussed in chapter
three and chapter four.

21
These assumptions cannot be examined and tested using cross-
sectional analysis. The general relationship between tangible
suggestions and intangible suggestions as well as the
relationship between suggestions and productivity can only be
understood via longitudinal analysis. This nethodological

weakness will be improved in this study.

Aims of the Study

The aim of the study is twofold:

(1) To identify the specific factors that influence the number
of adopted suggestions.

(2) To assess the specific ways that suggestions made in the
kaizen—suggestion system impact on productivity and other
measures of organizational effectiveness.

The first objective of this study is to examine the
extent to ‘which situational factors are determinants of
adopted.tangible and.intangible suggestions. This will be done
by analyzing the effects of lagged suggestion variables as
well as organizational variables. First, what factors
determine tangible suggestion contributions. Second, does past
intangible suggestion. making' affect current tangible
suggestion making? If yes, what is the lagged effect between
adopted tangible suggestions and adopted intangible
suggestions. The case of a Japanese-owned company,

Nippondenso,U.S., will be used to seek insight into these

22
questions.

The second objective of this study is to assess the
impact of suggestions made in the kaizen-suggestion system on
productivity, labor hour inputs, and product quality. While
studying the impact of the kaizen-suggestion system, I will
take into account other control variables that might also
influence organizational performance, such as technology,
training, product defects, industrial accidents, absenteeism,
organizational structure, etc. Three separate models will be
examined.to test the joint effects of suggestion variables and
other control variables.onjproductivity, labor hour inputs and
product quality.

Nippondneso,U.S. (NDUS) has been chosen as the case study
site. Two features make it ideal for an examination of
systematic situational determinants of employee suggestion
contributions as well as their effects on productivity and
other measures of organizational effectiveness. First, this
dissertation builds on an Michigan State University (MSU)
project studying the cross-cultural diffusion of U.S. and
Japanese work. practices. Three years ago, I started to
participate in this project supervised by an international
group of scholars based at MSU’s School of Labor and
Industrial Relations. We began a study of shopfloor work

practices in Japanese-affiliated factories in North America‘.

 

‘This project is supervised by Dr. Joel Cutcher-
Gershenfeld (Michigan State University) and Dr. Micho Nitta
(University of Tokyo).

23
The initial stage of this project was completed in 1992 but
research processes (e.g., study group meetings, the project-
book-writing sessions) continued. Thus this dissertation
emerged out of parallel field research, and I take advantage
of this three-year ongoing learning experience in Japanese
shopfloor ‘work: practices. However, because this research
project studied Japanese work practices, in general, this
study can help us further understand. one Japanese work
practice, the kaizen-suggestion system, in particular. Second,
NDUS was awarded the best employee suggestion program among
medium to large size manufacturing companies by the National
.Association. of Suggestion. System. in 1992. Therefore the
kaizen-suggestion system at NDUS can be viewed as a benchmark
employee suggestion system in the U.S.. A study of this
benchmark employee suggestion system is helpful for us to

better understand the nature of other suggestion systems.

Contributions
This paper serves as an exploratory effort to address
these research questions by examining the determinants of
suggestion made in the kaizen-suggestion.system.and.the impact
of adopted suggestions on productivity, labor efficiency and
product quality. It is exploratory in the sense that (1) no
other study in which a world-class benchmark kaizen-suggestion

system has been assessed in this way, and (2) it is based on

24
a rich body of plant-level data seldom available to
researchers to test the strengths of determinants and.outcomes
of the kaizen-suggestion system. Because of the lack of an
adequate theoretical construct for relating these concepts,
this paper can only serve for proposition generation rather

than formal theoretical testing.

Outline of This Dissertation

Chapter One generally describes the importance of the
study, the rationale underlying a number of research aims and
the potential contributions and limitations of this
dissertation.

Chapter Two discusses the methodology of study which
include a description of the research site as well as the
background of the kaizen-suggestion system, data source and
collection, method of data analysis, and Operationalization of
variables.

Chapter Three identifies the determinants of adopted
suggestions including a relevant literature review,
hypotheses, and research model. The results of the data
analysis for each hypothesis of those factors that influence
the number of adopted suggestions are contained in the last
section of this chapter. Finally, policy implications will
follow.

Chapter Four assesses the impact of the kaizen-suggestion

25

system on productivity and other measures of organizational
effectiveness. This includes a relevant literature review,
theoretical model, hypotheses, and.researttimodel. The results
of the data analysis for each of the hypotheses on the
outcomes of a kaizen-suggestion system are contained in the
last section.of this chapter. Inferential statistics of pooled
cross-sectional and time-series analysis will be used to
assess whether a hypothesis is statistically supported.

Chapter Five contains a summary of results and a
discussion of implications for theory and practice. Finally

directions for future research follow.

CHAPTER TWO

METHODOLOGY

This chapter describes the research.site, data source and
collection, Operationalization of variables, research.models,

and method of data analysis.

The Research Site and Background of the
Kaizen-Suggestion System

The site for the research reported here is a non-
unionized, medium-size company which is 100% Japanese owned.
Open in 1986, the facility occupies approximately 850,000 sq.
ft. During the period of the study, the work force included
approximately 1250 employees. The plant produces heat transfer
products for several major U.S. automobile manufacturers with
1992 sales of approximately $450 million. These include air-
conditioning systems (evaporators, cooling units, and
condensers), heater systems (heater cores, blowers, and air
ducts), electric fan and shroud.assemblies, and.radiators. The
plant produces several product lines with two technological
levels: (1) automated mass production lines for high-speed
assembly, and (2) flexible and automated production line
capable of finishing a wide range of products in random order
and production lines easily adaptable to model change.

The company had suggestion programs with some other

26

27

kaizen activities such as QCs and Total Productive Maintenance
since 1989. There are three general purposes of the kaizen-
suggestion system: 1) employees participation, 2) employees’
development and education, and 3) company’ 5 economic benefits.
As a business philosophy, the kaizen-suggestion system focuses
on improving the process and how employees do things. It
provides an organized systematic medium to recognize employees
for their problem-solving skills, while demanding their
involvement and commitment. Kaizen-suggestion systems focus on
employee involvement in.implementation of an idea, rather than
the idea itself.

The economic benefits are generally gained from

eliminations of seven types of waste5 by involving employees

 

5 (1) Waste from overproduction: This waste is created.by
producing goods over and above the amount required by the
market. The by-products of overproduction include extra
inventory, extra space, extra handling, extra interest
charges, extra overhead, etc..

(2) Inventory waste: As discussed above in connection with
waste of overproduction, excess inventory increase the cost of
a product.

(3) Waste of waiting time: For example, instead of occupying
machines totoverproduce, operator should remain ideal when the
required amount of work is finished. If supervisors cannot
better assess the capacity and control the situation well,
waste of waiting is created.

(4) Transportation waste: For example, transportation waste is
created when incoming material stored in the warehouse before
it is brought to the line rather than delivering directly the
material down the line.

(5) Processing waste: This waste is created from the ongoing
work processes or even the processing method itself, which.may
be a source of problems, resulting in unnecessary waste.

(6) Waste of motion: For example, walking is one kind of
wasteful movement, especially when one person is responsible
for operating several machines. Machines should be placed so
that the operator’s walking time is minimized.

(7) Waste from product defects: For example, when defects

28
and management. Waste at NDUS is operationally defined as
anything other than.the minimum.amount of equipment, material,
parts, space,‘ and labor inputs, which is absolutely
essentially’ to add 'value to the jproduct. By' diligently
practicing problem solving with as many people as possible,
much.of the current waste will be reduced. While each.person’s
idea.will be used to help the improvement of plant operations,
the results can. be obtained. by' implementing improvement
activities in the most integrative fashion so that each set of
improvements can be tied with the others. Consequently, the
more the elimination.of‘waste, the lower the production costs,
the higher the added value to the product. To administrate and
coordinate a kaizen-suggestion system, the System
Administrator is responsible for chairing the Steering
Committee, comprised of both production and non-production
employees, while playing a pivotal role in the administration
of the suggestion system. The primary function of the
committee is to review suggestion evaluations, approve awards
after calculations, and decide employee appeals associated
with suggestion evaluations. Employee suggestions are
classified as "tangible" or "intangible" based on whether
dollar savings.can.be estimated..And.these suggestions will be
evaluated according to the following general criteria: (1)

originality and creativity, (2) application and the effort of

 

occur at one station, operators at subsequent stations waste
time waiting, thereby adding cost to the product and adding to
production lead time.

29
the employee to implement it, (3) potential benefits, and (4)
employees’ skill level relative to the difficulty level of the
idea he/she suggests. Recognition for employees is based on
earning assign.points that are calculated.for each suggestion.
A gift certificate is awarded based on points earned and the
certificate can be redeemable at Service Merchandise stores.
There is an additional milestone points award process which
generates prizes such as a car or a trip to Japan as a result

of accumulated participation over many years.

Data Source and Collection

Data source and collection in this dissertation were
heavily based on previous plant tours, individual interviews,
shop-floor observations and preliminary empirical study (Lin,
1993) at NDUS.

Quantitative as 'well as qualitative techniques were
employed in this study. The results of the qualitative method
are used to enrich the quantitative analysis.

The data used in this study were department-level data6
from the Production Section for the months January 1991 to
August 1994. I include in my sample all departments in the
Production Section with complete data. The departments with

incomplete data. were excluded” As a result, I found 19

 

‘Individual suggestions are aggregated.intc»a department-
suggestion data to connect to departmental productivity and
other variables.

30

departments that conformed to the requirement. Consequently,
the available data represent 19 out of 29 observations, for 44
equivalent time periods. The departments are cross-sectional
and the "monthly" observations for a given cross-section are
arrayed in a time series. Totally, 836 observations were
included for pooled cross-sectional and time-series data
analysis in this study.

Additional qualitative field data collection is employed
to enrich the quantitative analysis. The qualitative methods
include focus group interviews and examination of appropriate
documents and records. Individual interviews key officials on
historical, and administrative and strategic issues are also
employed.

The development of interview questions was heavily based
on previous plant tours, individual interviews, shop-floor
observations and a preliminary empirical study (Lin, 1993) at
NDUS. These intensive research activities were parts of the
MSU project on cross-cultural diffusion of the U.S. and
Japanese work practices.

Overall, four general questions were asked for focus
group interviews. Each is represented below, with a
description of the specific information that was being

solicited through questions.

1. Please trace the full process history of a suggestion:

* idea generation;

31
* pre-suggestion-stage implementation;
* ongoing operation.
This question is designed to understand the full process of
the kaizen-suggestion system and the reasons why time lag
exists between suggestions adopted and productivity
improvements. It is also designed to elicit information about

the role of team leaders in the process of suggestion.making.

2. To what extent and in what ways does the kaizen-suggestion
system affect your job performance:
* Affective level: job satisfaction and work motivation;
* Cognitive level: job skills and knowledge.
The question is designed to help understand the employees
evaluation of the kaizen-suggestion system. Information also
is gathered to test the conceptual model of the impact of the
kaizen-suggestion system on individual performance which.will

be discussed in chapter four.

3. To what extent and in what ways does a kaizen-suggestion
system impact on organizational effectiveness:

* productivity;

* working hours

* quality.
This question is designed to obtain information about the
changes in productivity, quality and safety after the

intervention of the kaizen-suggestion plan.

32
4. Please trace the idea generation of a tangible suggestion
in comparison to an intangible suggestion.
This question is designed to help identify the
determinants of tangible and. intangible suggestions. The
information also clarifies the relationship between tangible

and intangible suggestions.

Method of Data Analysis

The technique of pooled cross-section and time-series
data analysis in general and classical pooling with a cross-
sectionally correlated and time-wise autoregressive estimation
in specific is employed in this study. There are several
alternatives for estimating the pooled data: (1) OLS
regression.estimation, (2) the covariance model, (3) the error
component model, and (4) classical pooling with a cross-
sectionally heteroskedasticity and time-wise autoregressive
model (Dielman, 1989). However, these alternatives are
inappropriate for the nature of data in this study. Therefore,
classical pooling with a cross-sectionally correlated and
time-wise autoregressive estimation is a preferred estimation
model in this study, and this will be explained later in this
section.

Firstly, I want to examine and discuss the appropriation
of alternative models here. The simplest method to estimate

the pooled data is to perform OLS regression on the entire

33
data set. Its underlying assumption is that there are no
complexities in the error structure. It seems to be suspect.
The common model for pooled.cross-section and time-series

data can be formed as:

Y1, = the level of the productivity measure in department

I i at time t.

i = 1, 2, 3, ........... ,19 (the number of department)

t = 1, 2, 3, ........... ,44 (the number of time points)

Xi,t = the corresponding measure of tangible and

intangible suggestion in department i at time t
a1 = the unknown parameter which measures the impact of
the tangible and intangible suggestion
and ei,t = the disturbance term which measure the impact of
all variables not in the equation.

With everything defined in equation 2.1, the issue is
simply one of estimating the sets of regression coefficients
(a,). The trouble is the increased structural complexity of
the disturbance (s) when pooling is attempted. When one
attempts to pool cross-section and time-series observations,
one must be aware of the fact that violations of the classical
least square assumptions are likely. Cross-sectional data are

often characterized by heteroscadasticity, while time-series

34
data may provide serially correlated disturbance terms. Thus,
it is possible or very probable that the pooled disturbance
term contains three types of perturbations: (1) cross-section
disturbances, (2) time-series-related disturbance, and (3) a
combination of both (Pindyck and Rubinfeld, 1976).

In the current examples this can be understood by noting
that the relationship between the disturbances of the
departments at some specific point in time may be different
from the relationship between the disturbance of one specific
department at two different times. More specifically, measures
of departmental productivity, labor efficiency and product
quality are relatively stable over time, but vary greatly from
department to department. Therefore, the disturbance for
department i at time t is highly correlated with the
disturbance for the same department at an earlier time point.
It is also expected that this correlation.will decrease as the
time points get further apart. Similarly, when the departments
are tightly linked to other departmentsfi it implies that the

disturbance for department i at time t‘Will be correlated.with

 

7NDUS can be featured as lean production system which is
defined in terms of a combined system including: customer-
driven priorities, just-in-time delivery between customers and
suppliers, little internal inventory between stations and
suppliers, broad team responsibilities for monitoring work
methods, processes, motion that eliminate waste and a
commitment to continuous improvement. See Womack, Jones, and
Roos (1990). There are high levels of team interdependence and
high.labor/management support.forwcontinuous improvement under
lean production system. See Cutcher-Gershenfeld et al. (1994) .
It suggests that it is very likely that departments are highly
interdependent at NDUS, U.S..

35

the disturbance for department j at the same time. Further,
since each department is subject to similar external effects
(e.g. , changing in company's production system, human resource
policy, etc.), it is likely that contemporaneous correlation
exists”. What this implies from a practical standpoint is that
ordinal least square of coefficients (a,) will be unbiased,
but inefficient (Johnston, 1972). Therefore, OLS estimation is
not considered desirable.

Secondly, the covariance model recognizes that pooling
may lead to variable cross-section and time-series intercepts,
and adds dummy variables to characterize each cross-sectional
unit and time period. However, it uses a substantial number of
degree of freedom (Judge, Griffith, Hill, Lutkepohl, and Lee
1985; Sayrs 1989; Dielman 1989).

Thirdly, classical pooling with a cross-sectionally
heteroskedatic and time-wise autoregressive model is based on
two assumptions: 1) the error variance differs between cross
sections, and 2) time-series residuals are autocorrelated
(Kmenta 1971). This model is not considered desirable because
this model lacks the assumption of contemporaneous correlation
that very likely exists when departments are highly
interdependent.

Finally, the error component model assumes nonhomogeneous

intercepts but assumes there are independent, identically

 

8This correlation between disturbances of different
cross-sections at the same point in time is called
contemporaneous correlation. See Kmenta (1971, pp 512-514).

36
distributed random variables rather than fixed (Judge et al.
1985) . Judge et al. argued that this independence allows
constant autocorrelation of disturbances from different time
periods. It also implies the contemporaneous correlation
between the disturbances of two-section units is the same for
every pair of cross-section units and that the correlation
between the disturbances of a given cross-section unit is
constant over time and same for every cross-section unit.
Despite the fact that model assumes a fairly sophisticated
error structure, it still cannot account for error structure
complexities in this study. It is because some cross-section
or time-series relevant variables may complicate the error
structure. For example, an adopted suggestion in department i
may improve productivity in department 1 itself as well as
other departments j, k, ....,z if this suggestion idea can
eliminate cross-departmental or even plant-wide waste. Another
example, an adopted suggestion in department i at time t may
prolong influence on productivity at time t+1, t+2, ...... ,t+n.
Additionally, as discussed earlier, measures of departmental
productivity is relatively stable over time, so the
disturbance for department i at time t is highly correlated
with the disturbance for the same department at an earlier
time point and the correlation declines as the disturbances
become further apart in time and that it can be different from
department to department. To account for such complexities,

the error component model is not considered desirable. Thus it

37
is preferable to use a procedure that can account for
heteroscedasticity, autocorrelation, and contemporaneous
correlation among the disturbances. Consequently, a cross-
sectionally correlated and time-series autoregressive
estimation is a preferred model in this study.

Basically, these difficulties can be overcome by a double
transformation of the original data (i.e. Generalized Least
Square) to deal with the aforementioned problems of
heterscedasticity, serial correlation, anui contemporaneous
correlation in the cross-sections (Kmenta 1971, pp 512-514).

Specifically, we assume, for instance, the model to be:

Yi,t = a0 + a1 x1i,t + a2 X21,t + a3 X31,t + a, X4i,c + 81,: (2'2)

Where
Yi,t = the measure of departmental productivity,
labor efficiency and product quality,
an,...xnfl_= the indicator variables for the tangible
suggestions, intangible suggestions, technology,

training, respectively.

and a1....a,:= the productivity improvement associated
with the tangible suggestion, intangible suggestion,

technology, and training.

38

1, 2, 3' and 4 are estimated in a double transformation process

as:

(351; (#‘B‘IXW '1 (x"‘9'1y‘)

The variables X' and Y' are transformed to account for an error
term that exhibits the following distribution:

Emit) =01, (heteroskedasti vi ty)

E(€,-,8jt) =01]. (mutualcorrelation)

51591313 t,1+p.it (autoregression)

and
E<ei,t-1pjc) =0
EWichc) =¢1j
E<Mitpjs) =0 ( t’ts)
i, j = 1, 2, ......... N

39

In this study, in summary, a cross-sectionally correlated
time-wise autoregressive model is employed.

SHAZAM software program (White, 197?) is employed in this
dissertation. The SHAZAM performed a cross-sectionally
correlated and time-wise autoregressive model under Kmenta’s
(1971, pp 512-514) model specifications and conditions. A
Generalized Least Square procedure and a cross-sectionally
correlated and time-wise autoregressive model are used on the

model described in Kmenta.

Operationalization of Variables

Productivity

When productivity is examined longitudinally, identifying
changes directly attributable to the kaizen-suggestion system
is difficult. Productivity can.be improved.either due to*waste
elimination in rework, waiting, motion, inventory, processing,
transportation, overproduction, or due to more effective use
of the existing means of production, which is the result of
improvement suggestions. The waste elimination in the
production system can be measured by changes in specific
inputs of the production system such as labor hour inputs.
More effective use of the existing means of production can be
measured by changes in specific outputs of the production

system such as output quantity. In order to include both input

40
and output elements, the definition of productivity is
therefore confined to the ratio between output and input.
Labor productivity data that are measured.by the ratio between
total quantity outputs and total working hour inputs were
employed in the study. Departmental productivity data were
calculated from combining productivity results for different

teams within the departments. A brief explanation follows:

: QI+QZ+ ............ +Qn
5P TH1 + THZ + ............ + THn

 

Pb

where

PDEP = Departmental productivity

Q1 ..... Qn = Quantity produced from team 1 to team N within
a department

TH,...JHL,= Total hours worked from team 1 to team N

within a department

Working Hours
The number of regular shift hours worked by full-time

employees.

Absenteeism
The absent ratio derived from unpaid absent (unplanned

and unscheduled absent).

41
Quality of product
A quality index derived from a count of the number of

customer claims for product defects.

Safety

A safety index derived from OSHA log for the number of
recorded nurse’s visit for occupational injuries and

illnesses.

Tangible Suggestion

A tangible suggestion is defined as a suggestion that
economic benefits or dollar savings can be estimated.
That is, cost reduction result from suggestions that can
lead to tangible financial benefits such as savings in
labor, material, machine hours, downtime, and so on. In
other words, tangible suggestions can result in
quantitative or tangible effects. Increased production
volume, shortening of product manufacturing time or
reduced defect costs are all good examples. Tangible
suggestions could be both "evolutionary" and
"revolutionary" (Mirvis, 1988, 1990) in terms of their
impact on economic gains. A tangible suggestions also
could be both "technical" and "administrative" (Evan and
Black, 1967; Daft and Becker, 1978) in terms of the
nature and category of the adopted suggestions. Tangible

suggestion is m4044easured as its total volume of

42
suggestion ideas for the statistical analysis in this
study. This variable reflects the quantity of adopted

suggestions in a department.

Intangible Suggestion

An intangible suggestion is a suggestion that economic
benefits or dollar savings cannot be estimated. Such
suggestions always involve intangible items such. as
safety, quality, housekeeping improvement, and so on. In
other' words, intangible suggestions could. result in
qualitative or intangible effects. Improved quality of
work, improved product quality, improved information
feedback, improved managerial skills, better
communication or increased.quality-consciousness are all
good examples. Intangible suggestions could be both
"evolutionary" and "revolutionary" (Mirvis, 1988, 1990)
in terms of their impact on economic gains. Intangible
suggestions also could Ix: both "technical" and
"administrative" (Evan and.Blact, 1967; Daft and Becker,
1978) in terms of the nature and category of the
suggestions. The variable of total volume of intangible
suggestions will be employed in the regression analysis.
This variable also reflects the quantity of adopted

suggestions in a department.

43
Traininq Hours
The number of credit hours employees earned in the
classroom training in a department. It is broken down by
three categories: technical training, self-actualization
training and. management training. Self-actualization
training is basically used in the employee orientation.
It also can be defined as "individual and organizational

development" training.

Technology

Dummy variable with 0 for assembly department and with

value 1 for machining department.

Overtime Hours
The number of overtime hours worked by employees in a

department.

Workforce size

The workforce of a department.

Top management leadership

Transition of top management on October 1992.
Dummy variable with value 0 for January 1991 to October

1992 and with value 1 for other months.

CHAPTER THREE

IDENTIFYING THE DETERMINANTS OF SUGGESTION MAKING

Introduction

This chapter discusses the relevant literature review. A
conceptual model of the determinants of suggestion making‘will
follow. Based on the conceptual model, hypotheses will be
proposed. Then regression analyses of the determinants of
tangible and intangible suggestions will be used to test the
hypotheses. Finally, policy implications will be discussed.

Because very few studies9 have evaluated the determinants
of effective suggestion.systems, the first section will review
the relevant literature which will be based on relevant
studies in employee involvement and participation in decision
making, though not to employee suggestion making directly.
Additionally, most independent variables used in the empirical
study of this chapter will not be discussed in the following

section of literature reviewu None of these variables has been

 

9Fewstudies have evaluated the determinants of effective
suggestion system. These studies assessed the determinants of
suggestion system by collecting data from either field (Burke,
Hoffman, Kazer, and Hall 1982; Pizam 1972; Ekval 1971) or
laboratory (Steinberg, 1981) . Two major influential forces were
studied: personal determinants (e.g. motivation, personality,
age, education, intelligence, satisfaction, creativity,
technical interests, occupation, length of service, etc.) and
situational determinants (e.g. money reward, supervisory
encouragement, freedom of communication, diversity of
workforce, coordination, etc.)

44

45

studied in the area of employee suggestion making. It is also
because none of the existing studies has applied pooled
departmental data in their statistical analyses. More
specifically, some individual factors (e.g. personality,
ability, extrarole behavior, and job complexity) that have
been examined in the few previous studies cannot be measured
and estimated in a time-series study. They may be only
appropriate in a cross-sectional study.

Because this dissertation will suffer from insufficient
literature and theoretical foundation, establishing an
exploratory'conceptual model of the determinants of suggestion
making may fill this big gap. Again, although both individual
and situational factors are included in the model, only
situational factors are employed in the empirical study.
First, as mentioned above, individual variables are difficult
to measure and estimate in a time series study. Second, this
is because departmental level of influence is more important
than individual level of influence at NDUS. In NDUS, there is
a strong organizational culture supporting the suggestion
program; therefore, lower variance exists at individual levels
and higher 'variance exists at departmental levels.
Consequently, situational variables at departmental levels
will be employed and analyzed in this chapter.

The empirical results will be reported in two separate
models: one for tangible and one for intangible suggestions.

It is important to distinguish between tangible and intangible

46

suggestions. First, suggestion making may reflect a learning
process. What workers have learned. in. making intangible
suggestions eventually will be transformed into the knowledge
needed for generating a new tangible suggestion. The
conceptual model, empirical results, and policy implications
proposed.in this chapter are based.on this assumption. Second,
the effects of tangible and intangible suggestions on
organizational effectiveness are different. Tangible
suggestions generally create quantitative effects whereas
intangible suggestions usually create qualitative effects.
Quantitative change which is derived from tangible suggestions
and qualitative change WhiChL is derived from intangible
suggestions have different policy implications to
organizations. The theoretical model, empirical results, and
policy implications proposed in the next chapter (chapter 4)
will be based on this assumption. For these reasons, I argue
that a distinction between tangible and intangible suggestions
is important for studying kaizen suggestion systems. Two
different models and regression analyses will be tested in
this chapter and the next chapter.

Policy implications will be discussed in the last
section. If the assumption that intangible suggestion making
has positive effects on tangible suggestion.making is true, it
will have profound policy implications for most traditional
U.S. companies. That is because these firms only pay attention

to tangible suggestions and give little attention to

47

intangible suggestions.

Literature Review

This literature review will consist of examining
selected major personal and situational variables and
interpreting their impact on employee involvement or worker
participation.behavitmn As discussed in the first chapter, an
employee suggestion plan can be viewed as a form of employee
involvement. Thus, the prior literature in this area will
provide the basis to better understand the employee suggestion
behavior.

The individual difference variables selected for
inclusion in this report are (1) ability, (2) personality; (3)
organizational extrarole behavior, and (4) demographic
variables (i.e. gender, tenure, and education). The
situational variables selected for inclusion in this review
are (1) job complexity, and (2) supervisory style.

Beginning with the personal variables, the discussion to
follow will examine the existing evidence concerning the
relation of each of the above variables to employee

involvement.

Personal Variables
Ability
Pizam (1974) suggested that some personal characteristics

such as creativity and ideas source should influence the

48

number of suggestions employees submit and the number of
suggestions adopted within a period of time. The higher a
person is on scores of creativity and idea source, the more
suggestions he/she submits. The study showed that structural
characteristics (i.e., supervisory' encouragement, the job
he/she does, or the work environment) were not good.predictors
of suggestion making.

In a typical EI study, the surveys of attitudes towards
participation have suggested.that level of skill and.knowledge
have some influence on the perception of employee involvement
(Holter, 1965; Smith, 1955; Pym, 1965). In particular, more
highly skilled employees are more likely to be interested in
both direct and indirect participation and are more prepared
to act in.a representative capacity.

In addition, some recent studies have shown the

importance of the role of subordinate competence between

employee participation in decision making (PDM) - performance
relationships . Locke and Schweiger (1979) postulate that
individual factors (e.g. group member knowledge) and

organization factors (e.g. task attributes) condition the
effectiveness of PDM including, presumably, its ability to
augment performance. Locke and Schweiger (1979) and Locke,
et. al. (1980) reasoned. that group> member' knowledge and
competence have a bearing upon the connection between PDM and
performance. They tend to view competence as a potential

moderator variable. Their position would be strengthened if

49
it could be shown that participation enhances the performance
of more competent employees but fails to accentuate the

performance of less competent workers.

Personality

It has long been assumed that the effects of a
participative versus non-participative managerial style would
depend on the kind of people being supervised. Vroom (1960)
hypothesized that the relationship between psychological
participation and both job satisfaction and job performance
varied with the strength of the need for independence and the
degree of authoritarianism. The results showed that the
feeling of participation in decision making generally had a
positive impact on attitudes (or satisfaction) and performance
(or effectiveness). A highly authoritarian personality was
virtually unaffected by the opportunity to participate; those
low on authoritarianism and with a high need for independence
reacted most positively. It should be noted that it would be
dangerous to generalize from this study; the sample consisted
of supervisors and.not blue-collar workers and the study dealt
with perceived.rather than.actual.participation“ Other similar
studies have been.examined.in testing of Vroom’s hypothesis or
a hypothesis similar to it and treated need for independence
and/or authoritarianism-like variables as moderators in the
relationship between participative leadership and job

performance and satisfactitmn Support for the relationship is

50

provided in studies by Mitchell, Smyser, and Weed (1975) and
Runyon (1973). No support for the relationship is found in
studies by Abdel-Halinland.Rowland (1976), Sadler (1970), Tosi
(1970), Searfoss and Monczka (1973), and Vroom and Mann
(1960). Finally, two other studies used the path-goal theory
of leadership effectiveness (House & lMitchell, 1974) to
examine personality-participation-effectiveness relationship .
Schuler’s study (1976) indicated that subordinates’ PDM was
satisfying to low authoritarian subordinates regardless of the
degree of task repetitiveness, but it was satisfying to high-
authoritarian subordinates only on tasks with low
repetitiveness. Furthermore, Abdel-Halim’s (1983) findings
suggested that high PDM was satisfying to low need-for-
independence subordinates regardless of the degree of task
repetitiveness; and high PDM was satisfying to high need-for-
independence subordinates only on low task repetitiveness.

Finally, Ekvall (1971) has indicated that successful
suggester seems to relate to their reaction of an active
personality and they need to adopt to change. The primary
motivation behind employee suggestion making goes beyond the
awards and promotion of the system. However, Whitwell (1965)
indicated that the suggestion system.can be viewed.as contract
offered by the firm to purchase employees’ ideas. The primary
motivation behind employee participation in the suggestion

system is receiving a reward at the conclusion of the process.

51
Organizational Extrarole Behavior

The emergence of research on helping behavior can be
directly traced to a number of theoretical sources:
Gouldner’s (1960) proposition regarding the prevalence of the
universalistic norm of reciprocity; and Leeds’ (1963)
suggestion regarding the prescription of the norm of giving.
These theories discuss social conditions for helping behavior,
or offer a cognitive and motivational basis for helping
behavior. Such questions as why people are often apathetic
and.do not help others, what conditions facilitate helping, or
what personal characteristics are associated with the tendency
to help have guided the study of helping behavior.

With a few exceptions, organizational scientists only
recently have begun to include ideas related to prosocial
behavior in studies of behavior in work organization (Brief
and Motowidlo, 1986). Katz and Kahn (1966) distinguished
between inrole and extrarole behaviors and suggested that
organizations depend on both kinds of employee actions. They
have indicated the many occasions in which organizational
functioning depends on extrarole behavior--behavior that
cannot be prescribed or required in advance for a given job.
These behaviors include any of those gestures that lubricate
the social machinery of the organization but that do not
directly adhere in the usual notion of task performance.
Organ (1988) and Staw & Boettger (1990) have broadened their

conceptualization of the performance contract to include

52
extrarole behavior. Organ and Konovsky (1989) have examined
the effects of individual affect and cognition on an
individual’s performance of citizen behaviors.

There are several similar concepts that have been
referred to as helping behaviors in different ways. For
example, Katz and Kahn (1966) used the term extrarole behavior
to refer to proactive behavior to achieve organizational
welfare. Brief and Motowidlo (1986) used the term prosocial
behavior to refer to positive social acts carried out to
produce and maintain the well-being and integrity of others.
Organ (1988) used citizenship behaviors to refer to those
employees who contribute helping behavior that cannot be
required for a given job. In the present study, I use Katz
and Kahn’s (1966) term, extrarole behavior, because
suggestions are something extra. Improvement suggestions from
employees are not normally expected to maintain a responsible
attitude toward the business, beyond the responsibilities
specifically assigned to them. They are something "extra" to
the company--something beyond the call of duty.

Extrarole behaviors are behaviors that are performed by
organizational members voluntarily and these contributions are
not inherent in formal role obligations. Examples of
extrarole behaviors include helping co-workers, supervisors,
subordinates with a job-related problem, helping to keep the
work area clean, tolerating temporary impositions without

complaint, talking favorably about the organization to

53
outsiders, protecting and conserving organizational resources,
and suggesting improvements in production or administrative
procedures.

There is no conceptual basis for thinking that employee
suggestion contributions would be related to organizational
extrarole behavior. Brief and Motowidlo (1986) suggested that
"attempting to suggest procedure, administrative or organiza-
tional improvements is another prosocial expression."
Employee suggestion behavior involves going beyond required
job assignment to perform some voluntary activity with the
intent of helping the organization and benefitting others.
Extrarole activities include such gestures as helping co-
workers, supervisors, subordinates with a job-related problem
(Puffer, 1987; Smith, et al., 1983) and showing thoughtful and
sympathetic attention to the need of other employees. For
most shop-floor employees, investigating an organizational
problem and suggesting changes for improving it is not a
formal job requirement. Instead, it is a voluntary gesture
that goes beyond their obligatory job assignments. For these
employees, suggestion contribution is clearly an

organizational extrarole behavior.

Demographic Variables
It is prevalent to stress the positive impact of
education on the potential for employee involvement. Zupanov

and Tannenbaum (1968) suggested that higher aspirations and

54

greater interest in participation tended.to have higher levels
of educational attainment. As employees’ training and
schooling increase, they will be better able and.more eager to
participate. Furthermore, participation itself appears to be
a very intense form of education (Pateman, 1973). Moreover,
participation perhaps creates demand on the part of the
workers for“ more general education and training courses
(Jenkins, 1973).

Some of the findings on scientific and R & D innovations
might be applicable to employee suggestion systems also. It
would be particularly interesting to examine studies relating
length of service to innovativeness. Katz (1982) and Smith
(1970) found a curvelinear relationship between the mean
tenure of members in project groups and ratings of their
group’s performance. Smith, however, found a positive linear
relationship between mean group tenure and group performance
as measured by patents and technical papers. These
researchers saw a lack of development of necessary role and
status relationships in groups with low mean tenure and
suggested that groups with high mean tenure may have isolated
themselves from important outside sources of information.
Pelz and Andrews (1976) also found this curvelinear
relationship but they found that projects with long-tenured
members could generate an intellectual competitiveness that

maintained a high performance ratio.

55

Situational Variables

Situational determinants are different from personal
determinants because they are the result of the circumstances
in which an individual finds himself/herself in relation to
his/her environment. Besides individual differences,
situational factors also play a very important role in how
people act. Social psychologists have employed a variety of
terms to describe the necessity' of using' both sets of
concepts. For example, Lewin’s (1951) general theoretical
formulation that behavior (B) is a function (F) of the person
(P) and the Situation (S), B = F(P,S) is a good example. Two
important situational determinants of employee suggestion
behavior in this study are: job complexity, and supervisory

style.

Job Complexity
The nature of the job is an important determinant of how
people act, and differences in tasks and task characteristics
have been shown to mediate differences in individual and
social behavior (Hackman, 1969a, 1969b). House and Mitchell
have stated clearly the difference between the nature of jobs
and PDM as follows:
When...task demands are ambiguous (or
non-repetitive) , participative leadership

will have a positive effect on the
satisfaction and motivation of the

subordinates, regardless of the
subordinate’s predisposition toward...
authoritarianism cnr need for

independence. when task demands are

56

clear (repetitive), subordinates who are

not authoritarian and who have high needs

for independence...will respond favorably

to leader participation and their

opposite personality types will respond

less favorably. (1974, p 93)
In other words, when the task is highly repetitive or routine
and subordinates are not allowed to make their own work
decisions, participation in decision making would have little
effect. Besides, Abdel-Halim (1983) investigated the effects
of task and personality characteristics on subordinate
responses to participatory decision making with a sample of
229 supervisory and non-supervisory employees in a large,
retail drug company suggesting that high need-for-independence
subordinates performed better and were more satisfied with
high participation only for non-routine tasks. Furthermore,
Hackman and Lawler (1971) indicated that perceived job
complexity is related positively to employee motivation, job
satisfaction, and other job performance. When jobs are high
on the four core dimensions (skill variety, autonomy, task
identify and feedback) , employees with moderately high desires
for higher order need satisfaction, tend to work harder, be
more satisfied, be absent from work infrequently, and.be rated
by supervisors as doing high quality work. Brief and Aldag
(1975) replicated Hackman and Lawler’s (1971) study and found
similar results. This study provided strong support for the
presence of positive correlations between the employee’s

perception. of his job characteristics and his effective

responses to that job. That is, an employee’s perceptions of

57
each core dimension was significantly (p< .05) related to his
level of internal work motivation, general job satisfaction,
and job involvement with the exception of the correlation
between task identify and internal work motivation. Several
studies also found that monotonous, repetitive jobs are
positively related to job dissatisfaction, absenteeism, and

turnover (Blanner, 1964; Guest, 1955; Walker, 1950).

Supervisory Leadership Style

Theoretically, supervisory style should influence
employee participation in decision making and the amount of
innovation in an organization. Steinberg (1981) suggested
that a. participative leadership style as opposed to an
authoritarian style causes more task-related idea generation,
and that the presence of a formal participative decision
making system would result in more task-related idea
generation than in those firms that do not have such.a system.

Some other PDM-innovation relationship research has been
conducted which collected data from non-industrial settings.
Fairweather, Sanders, and Tornatzky (1974) found that there
was a strong relationship between PDM and the degree of change
observed in federal psychiatric hospitals. They suggested
that "the degree to which involvement across disciplines,
across social status levels, and with more groups produced
greater change." Moreover, similar results were observed in

a study of welfare organizations. IHage and Aiken (1967) found

58

a positive relationship between PDM and the rate of program
change, and a negative relationship between more hierarchial
authority and program change. These studies implied that the
increased interaction provided by PDM created a greater degree
of perceived participation by employees, and that perceived
participation appeared to be associated with performing an
innovation.

Another classic studies in the PDM-innovation
relationship, though not aimed at innovation directly, also
supported.the idea that there was positive correlation.between
two variables. Maier (1953) found that a group’s resistance
to change could be sharply reduced by training the leader in
group decision procedures. This study suggested that PDM
could lessen resistance to innovation or change.

In sum, the impact of selected individual and situational
variables on employee involvement/suggestion making discussed

above can be conceptualized in Figure 3-1.

59

Figure 3-1: A Conceptual Model of the Determinants of
Effective Kaizen-Suggestion System

 

 

PERSONAL VARIABLES:

*Ability
*Personality
*Extrarole
Behavior
*Demographic
Variables

 

 

 

‘EMPLOYEE INVOLVEMENT/

SUGGESTION MAKING

 

 

SITUATIONAL VARIABLES:

*Job Complexity
*Supervisory Leader-
ship Style

 

 

 

 

60

The Operational Model and Hypotheses

An Operational Model of the Determinants of Effective Kaizen-
Suggestion System

As mentioned above, despite the fact that a number of
studies have evaluated the factors that influence employee
involvement behavior in.general, few studies have examined the
determinants of an (effective kaizen-suggestion system in
particular. Both situational and personal factors have been
examined in Employee Involvement research, seldom have studies
evaluated the situational constraints of suggestion system in
specific. Research that has been done almost always
concentrates on personal factors such as motivation,
personalityy intelligenceq age, education” occupation,
creativity (Burke et al. 1982; Pizam 1974; Ekvall 1971;
Steinberg 1981). None of the personal factors has been
supported as a powerful determinant of suggestion making.
Thus, situational variables should be carefully examined in a
kaizen-suggestion study. In this study, only situational
factors will be examined. This is based on two reasons. First,
as mentioned.earlierq in.NDUS, a strong organizational culture
supports the suggestion system and thus variance at the
departmental level is higher than individual level. Second, a
review of employee involvement and employee suggestion
literature suggests that the personal variables are not good
predictors of making suggestions in one hand, and the

situational determinants of suggestion generation have not

61
been examined on the other. Situational factors, thus, should
be investigated in the study of employee suggestion systems.

In figure 3-2, situational predictors of suggestion
making are broken down by group and environmental factors.
Group variables include a top management participative style
and group size, while environmental variables include
training, technology, absenteeisnl and. overtime. Most
independent variables used in this chapter have not been
discussed in the literature review section. It is partly
because none of these variables has been studied in the area
of employee suggestion making and partly because none of
existing studies has applied pooled departmental data in their
statistical analyses.

These group and environmental factors are expected to
relate to employee suggestion generation. The question is how
do they' work? Further; what, is the :relationship Ibetween
current suggestions and those adopted in the past? Do
accumulated experiences and skills in making intangible
suggestions in the past help stimulate or catalyze more new
improvement ideas which are then reflected in tangible
suggestion making? In this study, I will examine this
"snowball effect" of suggestion generation. An operational
model of the determinants of effective kaizen-suggestion model
is presented in Figure 3-2. The creation of hypotheses in the
following section will be based on the operational model

presented in Figure 3-2.

 

62

Figure 3-2: An Operational Model of the Determinants of
Effective Kaizen-Suggestion System

 

 

 

 

 

GROUP FACTORS INTANGIBLE SUGGESTIONI NVIRONMENTAL FACTORS
TGENERATED AT TIME t-l
*Top Management *Training
Leadership *Technology
*Group Size *Absenteeism
*Overtime
*Technology

 

 

 

 

L

1TANGIBLE SUGGESTIONP

 

 

 

GENERATED AT TIME t

63
Hypotheses

At NDUS, employees who make suggestions are given the
responsibility to implement them. They are encouraged to
consult with their team leaders, supervisors, or engineers and
figure out hOW'beSt to do so. Employees are allowed to test or
find alternative solutions. Allowing employees to solve
problems, under a kaizen system, fosters individual learning
(Florida and Jenkins, 1993). Hence, for a suggester,
suggesting is a learning process. Consequently, a kaizen-
suggestion system not only facilitates current suggestion
making but accumulates knowledge needed for generating new
ideas, new suggestions in the future.

In general, tangible suggestions are ideas related to
production or technological improvements, while intangible
suggestions are ideas related to nonproduction or
nontechnological improvements. Tangible suggestions tend to be
more difficult to generate than intangible suggestions. In
other words, generating a tangible suggestion requires more
knowledge, skills, and ability than generating an intangible
one. A tangible suggestion can hardly be generated without
proficient problem-solving skills, practiced job knowledge,
and enriched suggestion experiences. And these skills,
knowledge and experiences could be accumulated from previous
experiences in making intangible suggestions. An intangible
suggestion itself could be a very subtle organizational

improvement, but an intangible suggestion is very crucial and

64

fundamental for the accumulation of skills and knowledge
needed for making a tangible suggestion, or even making
another" intangible suggestion. SDI other *words, Iknowledge
accumulation in making intangible suggestion is the basis of
tangible suggestion making. Thus, the total volume of adopted
intangible suggestions can.be viewed.as the effect of learning
and knowledge accumulation. Consequently; a positive effect of
this knowledge accumulation on tangible suggestion making is
expected.

As mentioned earlier, tangible suggestions tend to be
more difficult to generate than intangible suggestions.
Generating a tangible suggestion requires more knowledge,
skills and ability than generating an intangible one.
Knowledge and skills accumulation in making intangible
suggestions helps group members build their competency within
the group, and in turn, increased group competency helps
stimulate 'more tangible suggestions ‘making. Therefore, a
positive effect of this group competency accumulation on
making tangible suggestions is also expected. On the basis of
these reasoning, the first hypothesis can, be stated as

following:

Hypothesis 1.
The greater the current and past volume of adopted

intangible suggestions, the greater the current volume of
adopted tangible suggestions.

Employees can make more valuable suggestions only when

65
they have the required skills and knowledge to understand how
improvements could be made and how problems can be solved.
Training is the best way to improve employees’ skills and
knowledge, as well as change their attitude toward.improvement
activities. When employees are trained to operate, maintain,
and repair equipment, they are capable of sharpening their
skills and knowledge of production technologies and
understanding the overall.production system1in.ways that might
lead to insights on how improvements could be made and how
problems could be solved. Viewed dynamically, training
increases skills and knowledge, which.increases the ability to
generate improvement ideas and solve problems. Further,
because tangible suggestion generally are ideas related to
production or technological improvements, technical training
may be more helpful for tangible suggestion making than
management and self-actualization training. On the contrary,
because intangible suggestions generally are ideas related to
nonproduction, nontechnological or administrative
improvements, management and self-actualization training may
be more important than technical training.Thus on the basis of
this reasoning, the expected relationship between training and

number of adopted suggestions can be stated as follows:

Hypothesis 2.

The greater the number of training credit hours, the
greater the number of the departmental tangible and
intangible suggestions.

66
Hypothesis 2.1
Technical training will be more important than management
self-actualization training for making tangible
suggestions.
Hypothesis 2.2
Management and self-actualization training will be more

important than technical training for making intangible
suggestions.

Two broadly different production technologies (machining
vs. assembly) are used in the production system at NDUS. The
nature of jobs seems to be linked to production technologies.
The jobs are more complex on machining lines than those jobs
on assembly lines. Thus we expect job complexity to be
positively related to suggestion generation because complex
tasks create more chances for jobholders to suggest change for
improvement in job design; innovative work methods, production
procedures,cnradministrative procedures. In.contrast, simple
tasks have relatively few interdependent components and
require fewer skills, reducing the likelihood that useful
alternative work methods could be determined even if the
employee’s ideas were solicited. In other words, when a job
is simple and the best way of performing a task is obvious,
employee involvement and suggestion making probably are not
necessary (Hatcher, et. al., 1989) . The expected relationship
between job complexity and suggestion contribution is stated

in Hypothesis 3.

67

Hypothesis 3.

Machining departments will have more tangible and

intangible suggestions than assembly departments.

Absenteeism may be negatively related to suggestion
generation. Dissatisfaction and low organizational commitment
probably are two major reasons for absenteeism. An employee
who is dissatisfied with his/her job or organization and has
a low degree of organizational commitment may have no
intention to participate in improvement activities in general
and suggestion making in particular. On the other hand, it is
impossible for an employee to make any suggestion when he/she
absent from work. Even when he/she comes back to work, he/she
may have little time to participate in suggestion making
activities because he/she has to pay more attention to catch
up his/her work. Thus a negative relationship may exist
between absenteeism and suggestion making.
Hypothesis 4.

The higher the absenteeism.ratio, the lower the volume of
tangible and intangible suggestions.

Suggestion making requires substantial time away from
production tasks to investigate problems, consult with
resources (i.e., team leaders, supervisors, engineers, etc.),
test solutions, write a suggestion form, and.so on. Suggesters
always participate in these suggestion making activities in
addition to their duties. Therefore suggestions cannot be made

if suggesters work overtime too often or too long. Lack of

68

free time seriously interferes with employees’ involvement in
suggestion making activities. Even if employees have time
after they come home from work, they are often so emotionally
and physically drained that they have little energy left to
think about improvement ideas. Based on this reasoning, a
negative relationship is expected between overtime and
suggestion.

Hypothesis 5.

The greater the number of overtime hours a department

has, the fewer tangible and intangible suggestions the

department makes.

The size of the workforce in a department may be
positively related to its suggestion contributions. Employees
in a bigger organization may have more opportunities to
interact and discuss with other employees than employees in a
smaller organization. Imai (1986) has argued that the change
process or improvement itself originates through discussions.
High degree of discussion or interaction with others (e.g.
teammates, coworkers in other teams, team leaders, etc.) is
helpful for' employees to stimulate improvement ideas or
brainstorm better solutions. Further, a bigger organization
tends to have a higher degree of diversity in people, work
processes, or equipment than a smaller organization. Employees
in an organization with a higher degree of diversity may have
broader views and tend to be more creative and open minded,
which is the foundation of an effective kaizen-suggestion

system. Thus, from this reasoning, I propose the next hypothesis:

69

Hypothesis 6.

Big departments have more tangible and intangible
suggestions than small departments.

The dummy variable for "year" is to examine the effect of
top leadership on suggestion generation. The observational
period was divided by two stages: period one: January 1991 to
October 1992; period two: November 1992 to June 1994. The cut
off point (i.e. October 1992) of the observational period was
set at the month of the transition of two top managers who are
in charge of the kaizen-suggestion system. The former and
present managers were described as representing two extreme
points of the authoritarian-democratic continuum
respectively”. Leadership has been identified as one of the
major factors that affect the performance of the kaizen-
suggestion system at NDUS. Thus, the leadership style of top
management should be examined in the model of the determinants
of suggestion making.

Theoretically, supervisory style should influence
employee participation in decision making. A participative
style of leadership can create a more open communication
system and increase interaction among employees and their

supervisors, co-workers and subordinates for the purpose of

 

10This is based on individual interviews with key
officials who are in charge of the implementation of the
kaizen-suggestion program in HR department and with an IE
engineer who is partially responsible for the implementation
of kaizen activities in the plant.

7O
supervisors, co-workers and subordinates for the purpose of
discussing and resolving work-related procedures and issues.
The increased interaction and discussion of work-related
procedures and issues provided by a participative organization
increases the opportunity for expressing an idea. In other
words, the greater the amount of work-related interaction that
takes place,the greater the chance that innovative ideas will
arise from this interaction (Hoffman & Maier, 1961).
Furthermore, Likert (1967) and Lowin (1968) indicated that a
more open and less critical atmosphere will lessen the
inhibitions that employees may have about expressing an idea
regarding work-related procedures or processes. The
implications of research on leadership style and participative
climate are that the more participative environment an
organization has, the more likely the employees have work-
related interactions, the greater the chance that suggestion
ideas will arise from this interaction. Therefore, an
organization with a participative leadership style may have
more adopted suggestions than an organization with an
authoritarian leadership style. In other words, the volume of
adopted suggestions in period two may be more than in period
one. However, we should. note that the: different results
between two periods of time may also reflect a matured effect

of the kaizen-suggestion system“. The next hypotheses can be

 

11When employees have long been lived with the kaizen-
suggestion system, the accumulation of problem-solving skills,
knowledge, and experiences in generating ideas helps employees

71
described as follows:

Hypothesis 7.
A positive relationship exists between a top leadership

participative style and the number of adopted tangible
and intangible suggestions.

Summapy of Hypotheses

Hypothesis 1.
The greater the current and past volume of adopted

intangible suggestions, the greater the current volume of
adopted tangible suggestions.

Hypothesis 2.
The greater the number of training credit hours, the
greater the number of the departmental tangible and
intangible suggestions.

Hypothesis 2.1
Technical training will be more important than management
self-actualization training for making tangible
suggestions.

Hypothesis 2.3
Management and Self-actualization training will be more
important than technical training for making intangible
suggestions.

Hypothesis 3.

Machining departments will have more tangible and
intangible suggestions than that of assembly departments.

Hypothesis 4.

The higher the absenteeism ratio, the lower the volume of
tangible and intangible suggestions.

 

 

make more volume or higher quality of suggestions over time.

72
Hypothesis 5.
The greater the number of overtime hours a department
has, the fewer tangible and intangible suggestions the
department makes.

Hypothesis 6.

Big departments have more tangible and intangible
suggestions than small departments.

Hypothesis 7.
A positive relationship exists between a top leadership

participative style and the number of adopted tangible
and intangible suggestions.

Research Models

Multiple regression analysis is employed to estimate the
model of the determinants of suggestion making. The analysis
of the determinants of suggestion making is fundamentally
dynamic it can only be understood via longitudinal analysis
and thus is best suited to time-series analysis. Since the
focus is not on differences among departments, however, making
monthly aggregates for any one department is inadequate. Thus,
data for departments are pooled for the period tested to
provide the aggregate data.base for estimating the regression.
On the other hand, because pooled cross-sectional and time-
series data increase degrees of freedom and thus reduces the
variance of the estimators of the regression, the possibility

of significant results will increase.

73

Model of determinants of Adopted Tangible Suggestions

Tang“, = a0 + a1 Intangi, H, + a2 Mgttrani, H,
+ a3 SelftranL bk + a1 Techtrant pk
+ a5 Tech“ + a,3 Absenti,t
+ a7 Overtimei,t + 31 SizeL,
+ a9 Leadershipi,t + e1,t

Model of determinants of Adopted Intangible Suggestions

Intangm = a0 + a1 Mgttrani, H, + a2 Selftranil M,
+ a3 TechtranL bk + a, Tecth
+ a5 Absenti,t + a6 OvertimeLt
+ a, Sizer: + a8 Leadershipi,t
+ eh,

Where

Tanthfl,= Volume of adopted tangible suggestion in department 1 for
month t-k

IntangL Pk = volume of adopted intangible in department 1 for month
t-k

MgttranL bk = Management training hours earned in department i for
month t-k

Selftrani, N, = Self-actualization training hours earned in department
i for month t-k

TechtranL bk = Technical training hours earned in department i for
month t-k

Tecth = Degree of technology in department i at month t
Absenteeismi,t = Absent ratios in department i at month t
Overtimei,t = Overtime hours in department i at month t
SizeLt = Size of workforce in department 1 at month t

Leadership,‘t = Dummy variable for the transition of top management
in department 1 at month t

74
Empirical Results and Implications

Pooled cross-sectional and time-series regressions were
estimated with data from the 19 departments for the
observational periods January 1991 to August 1994. The total
volume of tangible suggestions was used, as a dependent
variable; the current (t) and past volume (t-1/,i.t-2) of
intangible suggestions addressing a learning effect of problem
solving and participation, and the various training hours,
overtime, absent, workforce size, technology and leadership
as control variables. The major aim of this model is to

examine the knowledge accumulation effect of intangible

suggestions on tangible suggestion making.

Relationships Between Tangible Suggestions, Intangible
Suggestions, and Other Control Variables

Table 3-1 shows means, standard deviations and
correlations for twelve variables across the 19 departments
from January 1991 to August 1994. The data reveal a strong
connection between tangible suggestion and various intangible
suggestions, providing support for Hypothesis 1---that is,
that the current and past volumes of adopted intangible
suggestions are associated with.the current volumes of adopted
tangible suggestions. More management training is related to
more the volume of adopted tangible suggestions. A.more highly
participative leadership style of top management is related to
a greater volume of adopted tangible suggestions. There is

also evidence that more management training is strongly

mu.

66. 6a.- 66. 66.- 66. 66. 66.- 6a.- 6H.- 6H.- 66. 66. «6. ass-666666 666666666: son .«a
HH.- «6.- 6a.- 56.- 56.- 66.- 66. 66. 66. 66. 66. 65. smoaoenuoa .««
«6. 6H. 66. «6. as. an. 6«. n«. #6. «6.« 66.H o«uuu ago-n: .6«
65. n«. «H. 6«. «6. «6. «6. H6.- 56.«56 66.666 os«uuo>o .6
66. 56. a«. 56. 56. 56. «6.- 6H.6« «6.6« ou«6 cououuuox .6
66. «H. 66. 66. 66. «6.- 66.5H «6.6« ma«6«uua Hau«aaooa .5
66. 66.- «6.- A6.- 66.- «5.5 «5.6 maanaaua no«6un««6666¢-6H66 .6
66. 56. 6«. 66. «6.6« 66.«a ma«a«aua 666366666: .6
56. 66. 6«. 66.5H «5.«H .«-u. ao«6-666:6 o«aamnuuuu .6
56. 6«. 66.5” «5.«« .«-6. no«uaomusm canamaauna .n
6«. 66.5« «5.«a .6. noes-666:6 oaa«mauuan .«
s«.H 66. .6. aoaauomuam o«n«6aaa .«
i -..:-:... ..... Mid- . m - w-----m::-m:-:m:-:.au----s...
nuaunnluul nnnnnnn unlit-illintuit.-lanaiin.-alafllildnfidflldikldnxrllnuInIIInullinnnlullilclulnlouu

Gnu I 2. Manual nodusasuuou use .euowusuben masons-.5 .essel ”H-n sands

76
associated with more adopted intangible suggestions.
Correlation analysis reveals associations, but it does
not provide tests of causality. Further hypothesis tests will

be examined in the next section.

The Selection Processes of Various Lagged Variables

There were some problems that had to be solved before I
conducted a regression analysis with a lagged hypothesis of
knowledge accumulation. There is 1K3 theoretical basis for
making a clear-cut decision as to just what the maximum direct
lag of intangible suggestions should be. Griliches (1967) has
indicated "Do not expect the data to give a clear-cut answer
about the exact form of the lag. The world is not that
benevolent. One should try to get more implications from
theory about the current form of the lag and impose it on the
data." Nevertheless, in this study I have no theoretical
basis on which to select the most appropriate pattern of
lagged relationships. Nor am I aware of any suggestion study
that has successfully determined an operational solution to
the problem based on theoretical reasoning. in: solve this
problem, I conducted direct distributed lag estimates and
selected the most "benevolent" results. I estimated the
parameters of truncated versions of from one through six-
period lag (i.e. six months or two quarters).

The overtime-pattern of effects for intangible suggestion

variables can be examined in Appendix A. The intangible

77

suggestion effect on tangible suggestions in Appendix A peaks
at current period and declines through t-6. Obviously the
regression.coefficients begin.to be insignificant and small or
wrong-signed when Intang t-3 is included in the equations;
equation (2) in Appendix A may therefore be considered the
most accurate equation. which is still significant.
Consequently, the current and a lag of one and two months of
intangible suggestions were included. in the full models
presented in Table 3-2.

Another problem for model selection is how to select the
most appropriate pattern of lagged variables of the various
training hours. There is also no theoretical basis for
supporting a lag relationship between training and tangible
suggestion making or how long a lag will be. Sue Flees,
specialist of HRM and coordinator of suggestion program at
NDUS, suggested that it seemed to take time for transfer of
training from basic knowledge to applied knowledge. Since the
task situations are so different, learning in training
settings has little relationship to suggestion making at work
settings. Therefore, it can.be assumed that the learning that
occurred in the training settings takes time to transfer and
aid performance in making suggestions. Appendix B shows the
-selection processes of the optimal pattern of lagged training
‘variables. The optimal results were obtained with a lag of
cone month. 'Fhe signs of regression coefficients are "correct"

:in.all three training variables. The regression coefficients

78
are insignificant and small or' wrong-signed with. either
current period or a lag of two months. The results suggest
that the more training a department took in the last month,

the more adopted tangible suggestions in the current month.

Empirical Results from the Model of Determinants of Adopted
Tangible Suggestions

After the selection of the optimal pattern of lagged
variables of intangible suggestions and training hours, the
full model of determinants of adopted tangible suggestions was
conducted. The results are represented in Table 3-2.

The current and past volume of adopted intangible
suggestions are found to have positive effects on tangible
suggestion making. The current and past volume of adopted
intangible suggestions might be expected to work in the
equations of the determinants of adopted tangible suggestion
for one of two reasons: either because they represent past
individual learning and knowledge accumulation that affect
current tangible suggestion making, or because intangible
suggestion making improves employees’ motivation or
satisfaction that increase current tangible suggestion making.
Intangible suggestions have greatest effect on tangible
suggestions at current period (P < .005), is small for a lag
of one month, goes up again for a lag of two months (p <
.005). The regression.coefficients of 0.015, 0.003, and 0.007
in equation 1 in table 3-2 exhibit a reverse bell-shaped

distribution around a central lag of one month. Similar

79
results are obtained from regression coefficients in equations
2, 3, and 4 in table 3—2. This may be interpreted by saying
that one part of tangible suggestions increases simultaneously
with intangible suggestions and a second part of tangible
suggestion gains which show a distributed lag behind
intangible suggestion making at current period.

More extensive employees and.management participation in
training programs lead to more tangible suggestion making. In
equation 1, 2, and 3, table 3-2, the association between
Mgttran and tangible suggestion is statistically significant
at the 0.05 level. However, the other associations between
the training indices (Self-actualization and Technical) and
tangible suggestion in equation 1, 2, and 3, table 3-2, are
not statistically significant. It is partially opposite to
the hypothesis that technical training may be more important
than management training to help employees make tangible
suggestions. It is interesting that management training is
designed for leaders at all levels (i.e., team leaders,
coordinators, departmental supervisors, and.so‘on)lbut not for
workers. 1n: may be interesting to observe that when team
leaders or supervisors learn managerial and interpersonal

skills in the training programs12 and return to working

 

12Michael Gagnon, manager of Organizational Development
at NDUS, supported this idea that some management courses such
as 'problem solving, creativity skills, risk taking, and
communication especially help team leaders or supervisors
guide their members and solicited their new ideas to make more
suggestions.

80

Table 3-2: The Determinants of Tangible Suggestion Making

Explanatory Tangible Suggestion
Variables (l) (2) (3) (4)
Intercept b 099 .108' 122 .131"'
t 875 1.888 1.151 3.883
Intang, b 015'" . 015'" . 015'" . 016'"
B 213 .217 .215 .227
t 5.663 5.765 5.742 6.087
Intangp1 b .003 003 .003 .003
B .044 .043 .045 .048
t 1.171 1.133 1.209 1.275
IntangH b .007'" .007"' 007'" .007'"
B .097 .096 097 .095
t 2.593 2.573 2.595 2.536
Mgttranh, b .003' .003' 003’
B .043 .044 .043
t 1.756 1.836 1.766
Selftranh, b .004 .004 .004
8 .024 .027 .024
t .918 1.092 .939
T.Chtr‘n‘_‘ b e0002 e"'o‘ e0001
B 003 .006 .001
t .111 .273 .060
Overtime b -.3E-04 -.Sl-04
3 .e01‘ -6022
t -.459 -1.l76
Absent b .011 .005
B .014 .006
t .527 .240
Size b -.3E-04 -.0005
B -.0006 -.001
t “s02: '65‘5
Tech b -.009 -.103
a '6035 'e038
t -16085 -1e183
Leadership b .167"' .154"'
B .069 .064
t 2.575 2.435
R’ .134 .131 .134 .120
1 11.601 15.583 14.148 37.695

r-test for all estimates significant at the .01 level.

0 - p < .05 level
8* - p < .01 level
0" - p < .005 level

b is the regression coefficient.
B is the standardized Coefficient.

t is the t value.

81
settings, they may be enthusiastic about encouraging and
guiding their team members to make more suggestions.

More democratic leadership style, as indicated by a dummy
variable in Leadership, may be associated with more tangible
suggestions. In equation 1 and 3, table 3-2, the association
between Leadership and tangible suggestions is statistically
significant at the .005 and .01 level respectively. It
suggests that the volume of adopted tangible suggestions in
period two (i.e. January 1991 to September 1992) is more than
that in period one (i.e. October 1992 to August 1994). A much
more democratic and participative organizational climate has
been observed and described in period two than in period one.
However, note that there is a highly participative base-line
at NDUS, so variance is around an already high level. It
supports the hypothesis that the more participative
environment an organization has, the more likely it is for
employees to have work-related interactions, and the greater
the chance that suggestion ideas will arise from this
interaction; We also should note that the increased pattern
could. reflect a 'matured effect of the kaizen-suggestion
program.

The equations reported in ‘table 3-2 provide little
evidence that more overtime hours hinder tangible suggestion
making. A higher absenteeism ratio led to more tangible
suggestions, not fewer as predicted by a "demoralized"

hypothesis. The larger workforce size lead to fewer

82
suggestions, not more as predicted” Finally, machining
departments have fewer tangible suggestions than assembly
departments, again a result opposite to the prediction.

With regard to the impact of individual learning, as
measured by adopted intangible suggestions, the regression
reported in equation 4 in table 3-2, shows a statistically
significant effect for current period suggestions (P < .005)
and for a lag of two months (P < .01) on tangible suggestion
making. The R2 for equation 4 is .120. The R2 rises from
.120 (equation 4) to .134, .131 and .134 (equations 1, 2, and
3 respectively) when all or part of the control variables
appear in the regressions, indicating a small effect for
control variables on. making tangible suggestions. IMore
control variables included in equations 1, 2, and 3 increased
explanatory power very little (the R2 rises from .120 to
.134, .131 and .134 respectively), whereas the F values
dropped.dramatically from 37.695 to 11.601, 15.583, and 14.148
respectively. It suggests that a set of intangible suggestion
variables has a strong cumulative and statistically

significant effect on tangible suggestion making.

Empirical Results from the Model of Determinants of Adppted
Intangible Suggestions

Table 3-313 reports the results of regression estimates

 

” Various training variables with a lag of one month are based
on the results of selection process of the most appropriate
pattern of lagged training variables. The regression.estimates
are presented in Appendix C.

83

of determinants of adopted intangible suggestions. In the
equations reported in table 3-3 the same patterns hold as in
table 3-2 except workforce size and leadershiph More training
credit hours generally caused more intangible suggestions.
However, only management training was statistically
significant at .005 level. It partially supports the
hypothesis that management and self-actualization training may
be more important than technical training for employees making
intangible suggestions. ‘There was no evidence to support that
Overtime and Workforce Size affected intangible suggestion
making. Absenteeism and Technology did not affect intangible
suggestion making in the hypothesized manner. The results
were opposite to the predictions. In equation 1 and 4 in
table 3-3, more participative climate led to fewer intangible
suggestions, not more as I predicted before, and the
coefficients were statistically significant at .05 level. The
hypothesis is not supported. Roy Roemen, superintendent of
production at NDUS, suggested that it may be because the
evaluation committee members (included himself) have set more
rigid standards for adopting intangible suggestions in the
past two years. It suggests intangible suggestions decrease
over time that seems to have nothing to do with top management
leadership style.

R2 for equations 1 through 4 in table 3-3 are .025, .021,
.019, and .024 respectively, indicating that only about 2

percent of the variations in intangible suggestion making

84

Table 3-3: The Determinants of Intangible Suggestion Making

--------------- ..............................................................................

Explanatory Intangible Suggestion
Variables (l) (2) (3) (4)
Intercept b 6.215"' 4.999"' 5.704"' 5.783”'
t 4.163 5.691 4.052 5.047
Mgttranh, b .068"' .065"' .064"' .066"'
B .061 .059 .059 .061
t 3.364 3.321 3.401 3.338
Selftranv, b .049 .049 .052 .046
B .022 .022 .023 .020
t 1.263 1.279 1.399 1.200
Techtranp, b .005 .005 .001 .006
B .005 .005 .001 .006
t .314 .276 .087 .339
Overtime b -.001 -.0004
B -.019 -.015
t -.512 -.555
Absent b .288 .338 .315
B .026 .031 .029
t 1.286 1.522 1.414
Sise b .021 .012 .005
B .028 .017 .007
t .696 .508 .211
Tech b -.632 -l.217
B -.016 -.031
t -.477 -.842
Leadership b -1.865' -1.938'
B - 053 -.055
t -1.697 -1.773
2’ .025 .021 .019 .024
r 2.695 3.496 3.279 3.375

P-test for all estimates significant at the .01 level.
9 - p < .05 level
0* - p < .01 level

0.. - p < .005 level

b is the regression coefficient.

8 is the standardised Coefficient.

t is the t value.

85
could be accounted for by these situational variables.
However, the F values for these 4 equations are all
statistically significant at .01 level. This should not
surprise us because the sample size of this study is 836. It
can be concluded that the linear relationship between these 8
independent variables and intangible suggestions is not zero
in the population, with a less than 1 percent chance of doing

so erroneously.

Summapy of Tests of Hypothesis

Knowledge Accumulation Hypothesis:

Positive relationships were found between tangible
suggestion making in month t and intangible suggestion making
in month t-k. The current period of tangible suggestion
making is based on current and past learning and knowledge
accumulation in making intangible suggestions. Thus, this

hypothesis is strongly supported.

Training Hypothesis:

Only for management training programs were statistically
significant found between training and both tangible and
intangible suggestion making. This hypothesis is partially

supported.

Overtime Hypothesis:

A negative relationship was found in both cases of

86
tangible and intangible suggestion making. The more overtime
hours a department has, the fewer tangible and intangible
suggestions the department makes. However, these findings are
not statistically significant, so we must conclude that there

is little evidence to support this hypothesis.

Absent Hypothesis:

For' both tangible and intangible cases, a. positive
relationship was found between suggestion making and the
absenteeism ratio. The results were opposite to what I

expected. This hypothesis is rejected.

Workforce Size Hypothesis:

A negative relationship was found between tangible
suggestion making and workforce size. 'The result was opposite
to the prediction. This hypothesis is rejected.

For intangible suggestion making, a positive relationship
between suggestion. making' and. workforce size was found.

Nevertheless, the result was nonsignificantly supported.

Technology Hypothesis:

For the technology dummy variable, negative relationships
were found between suggestion making and technology in both
tangible and intangible suggestion cases. Machining
departments do not have more adopted suggestions than assembly

departments as I expected. This hypothesis is rejected.

87
Leadership Hypothesis:

This hypothesis is strongly supported for the case of
tangible suggestion making and rejected for the case of
intangible suggestion making.

For the leadership dummy 'variable, a positive
relationship was found between tangible suggestion making and
leadership style; whereas a negative relationship was found
between intangible suggestion making and leadership style.

In sum, the analysis of plant-level data from NDUS
indicates that where there was more extensive individual
learning and knowledge accumulation through intangible
suggestion making, more management training, and a more
democratic organizational climate, tangible and intangible

suggestions are significantly increased.

Policy Implications

The significant results from the impact of intangible
suggestions on tangible suggestions may be striking in light
of the fact that most of the companies in the US only pay
attention to tangible suggestions, giving little attention to
intangible suggestions. The empirical results suggest that
employees may hardly make any tangible suggestion without
accumulating experience, skills, and knowledge by making
intangible suggestions. Allowing employees to make intangible
suggestions may, therefore, foster more tangible suggestions.

For a suggester, suggesting is a learning process. Making an

88
intangible suggestion currently helps ani employee to
accumulate knowledge needed for generating new tangible
suggestions in the future. Expecting employees to make more
complex or valuable suggestions without experience in making
simple or small suggestion is unrealistic. Therefore,
companies that implement or are interested in an employee
suggestion system should pay more attention to intangible
suggestions. .After' all, intangible suggestions are the
knowledge basis for tangible suggestions. Even if companies
do not treat intangible suggestions as more important than
tangible suggestions, at least they should be treated as
equally important as tangible suggestions. Furthermore, in
order to foster more tangible suggestions from knowledge
accumulated in making intangible suggestions, a learning
climate should. be created” .At INDUS, management should
facilitate employees’ problem solving and learning from
experience. Members from different teams or departments are
encouraged to discuss and resolve problems with each other.
More intensive interactions among or within teams may solicit
or generate more new ideas about how to solve problems. Past
adopted intangible suggestions also can. be developed. as
educational materials for the training sessions for QCs’
members or team members. Employees may catch cues or get some
ideas that lead to make more tangible or intangible
suggestions. On the other hand, employees may also receive a

message that company is not only concerned for machine or

89

organizational effectiveness but also concern people and non-
financial things. Regression analyses provide evidence of
statistically significant association between both tangible
and intangible suggestions, and. management training, but
nonsignificant for self-actualization and technical training.

At NDUS, employees who make suggestions are given the
responsibility to implement and operate them. However, before
they operate them, they are encouraged and in most cases they
have to consult with their team leaders or supervisors about
what the ideas are and figure out how best to do them. Team
leaders and supervisors play a very important role in the
kaizen-suggestion system to guide and help their members to
make and operate their suggestion ideas. In some cases, team
leaders even gave their members a rough idea and let them
investigate problems and finally made a formal suggestion.
Thus, it may be not so surprising to us that management
training (only offered to management) is more important than
self-actualization and technical training.

There are many management courses offered by NDUS“.
Among these courses, interpersonal skills training might be
the most important for team leaders to help their members make

an improvement suggestion. A person who demonstrates

 

1“Some selected management courses include: problem solving,
creativity skills, risk taking, leadership skills,
interpersonal communication, managing conflict, performance
improvement, maintenance improvement, productive listening,
and so on. All these courses may be helpful for team leaders
and supervisors to handle the kaizen-suggestion more
successfully.

90

effective interpersonal behavior has acquired the ability to
sustain trust, openness, emotional support, and.the expression
of strong feelings which are pivotal in the process of making
suggestions. Team members may lose interest in making
suggestion, if their leaders do not give them either physical
or emotional support. 'Without team leaders’ trust and
openness, team members will hesitate to talk to their leaders
even if they have a valuable idea. It suggests that
management support in the process of suggestion making is very
important to a successful kaizen-suggestion system. It
implies that companies should link management training to the
Kaizen-suggestion system at the strategic level. It also
implies that management training programs should be designed
to support kaizen processes and activities at the practical
level. Well-designed management training programs not only
train team leaders to help their members get involved in
kaizen activities, but educate them to reduce their resistance
to change.

Work teams and a bottom-up approach are frequently
mentioned as being features of Japanese management. Japanese
top managers rarely act as strategic analysts and sometimes
are characterized as invisible leadership (Kagono et al.
1985). The significant results from the impact of top
leadership style on tangible suggestions may challenge this
traditional argument about Japanese business leadership. The

regression analysis shows that democratic and visible

91

leadership from top management is an essential factor for the
effective kaizen-suggestion system. It was supported in the
regression analysis of tangible suggestion making, but was not
supported for the case of intangible suggestion making. The
data suggest participative leadership is essential for
successful kaizen processes. It implies that a more
democratic climate should be developed to promote and support
a kaizen-suggestion system.

The jointly significant effects of management training
and top management leadership on the kaizen-suggestion
activities implies that management support is the key for the

successful kaizen-suggestion system.

Summapy and Discussion

Most independent variables used in the empirical study of
this chapter were not discussed.in the literature reviewu This
is in part because none of these variables has been studied
and partly because none:of existing studies has applied.pooled
data.in.their statistical analyses. This dissertation.suffered
from insufficient literature and theory.

The results of this chapter provide initial support for
the proposition that the current (t) and past (t-1, and t-2)
adopted intangible suggestions are strongly associated with
the current adopted tangible suggestions. It implies that
suggestion making is a learning process. What worker have

learn in making intangible suggestions eventually will be

92
transformed to the knowledge needed for generating a new
tangible suggestion. An intangible suggestion itself may be a
subtle organizational improvement but it is very important and
fundamental for the accumulation of knowledge needed for
making tangible suggestions.

However, the initial evidence provided in this chapter
should be interpreted carefully. For example, the empirical
results from this chapter suggest that a significant finding
in the same month can be interpreted as a reflection of the
fact that groups high on intangible suggestions will also be
high on tangible suggestions. The additional significance of
the lag in the same regression suggested.a second.phenomena in
which.intangible suggestions predict tangible suggestions. IUI
interesting question is raised here. Will the same results
hold when tangible and intangible suggestions are reversed in
a regression equation. If yes, it implies that a reciprocal
relationship between tangible and intangible suggestions may
exist. Thus, it needs a further examination.

Finally, we should note that since the effect in table 3-
2 is not significant for all lagged months, these findings are
suggestive but not definitive. The R? in table 3-2 and 3-3
regressions are small, so we should interpret the results
carefully, Also, because the data.on.topimanagement leadership
style is based on a relatively limited measure, it must be

treated with more caution.

CHAPTER FOUR

ASSESSING THE IMPACT OF THE KAIZEN-SUGGESTION SYSTEM ON
ORGANIZATIONAL EFFECTIVENESS

Introduction

In this chapter, three issues will be addressed. First,
the relevant literature will be reviewed. Second, a conceptual
model of the effect of kaizen-suggestion system on
productivity will be established to clarify the influential
process of such systems on productivity. Finally, empirical
results will be discussed and the policy implications will be
elaborated.

The first part of the chapter will contain comprehensive
literature review and it will still tell us little about
kaizen program performance effects. A major difficulty for
literature review about kaizen suggestion systems is the lack
of a core literature, either theoretical or empirical. It
pushed me to be more comprehensive than usual in relevant
areas such. as employee involvement or QCs. Therefore a
comprehensive literature review in traditional suggestion
programs, QCs and employee involvement will be presented in
the first part of this chapter. The concept of a kaizen-
suggestion system is not only important in an implication of

continuous improvement itself but also has links in the

93

94

literature to productivity, quality and employee involvement.
It is an important phenomena, in part, because of the way it
integrates across all three areas. For example, the QC
literature adds the notion of group-generated suggestions, but
it fails to link the idea of system change. The employee
involvement literature provides the notion of communication,
decision making and problem solving, but it fails to link to
suggestion behavior itself.

The second part of this chapter will examine how the
kaizen-suggestion system. can. be expected to affect
productivity. A conceptual model will be established to aid
the analysis of the effects of suggestion making.

Because of the lack of core literature and theory, and.as
a result of the analysis, I will be presenting an inductive
model built on field observations, statistical analysis and

literature. Finally, policy implications will follow.

Literature Review
Introduction
As mentioned above, few studies have examined the impact
of the suggestion system on organizational effectiveness in
general and on productivity in specific. Due to the lack of
core literature in this area, it is not possible to make a
complete revieW'of these studies. In addition to a very narrow

employee suggestion literature, some relevant studies on

95
worker participation or employee involvement will be reviewed
in this chapter, though not linked to employee suggestion
making' directlyx This literature review'*will consist of
examiningWQuality'Control Circles, financial involvement (e.g.
Scanlon Plans), and other forms of worker participation and
interpreting their impact on productivity. In some cases,
actually, QCs and financial involvement use some form of
suggestion program; thus, QCs, financial involvement and the

kaizen-suggestion system are similar in their nature.

Research on Employee Suggestion Program

Few studies have evaluated the organizational impact of
employee suggestion systems. Most of the suggestion research
is case study reports of success in the practitioner
literature. What research has been done suggests that the
companies 'which. implement employee suggestion. system
experience reduced costs, greater employee satisfaction, and
better working conditions (see table 4-1). These studies have
either theoretical or methodological weakness that may limit
understanding of the full influential process of the
organizational impact of the .kaizen-suggestion system.
Generally, lack of theoretical, methodological and statistical
analysis hardly qualifies them as studies; however, they can
provide us with useful and fruitful information. The employee
suggestion literature contains innumerable stirring

testimonials and success stories.

96

Table 4-1: The Impact of the Suggestion Programs on Productivity

Study

Seimer
(1959)

Denz (1946)
Rand Inc.

Loesges
(1946)

Reuter
(1977)

Paulson
(1971)

NASS (1975)

Wilce (1971)

French
(1984)

French
(1984)

Sample

127 indi-
viduals in
two steel
fabricating
companies

Remington

Western
Electric
Company, Inc.

228 companies

General

Electric

General Dynamics

Standard Tele-
phones & Cables
Ltd.

Westinghouse

Hughes Aircraft
Co.

Productivity
Criteria

“NH

mU'lnhU-JNH

Profits

Costs

Safety

Job satisfaction
Product quality
Job security

.Costs

Costs

Job satisfaction
Relationship
between supervisors
and subordinates
Skill development

Costs

Costs

Costs

Costs

Costs

Results

mU'lubh-DNH

Increased
Reduced
Improved
Mixed
Improved
Increased

A feeling of
solidarity with
the company

Cost savings of
$200 to $400
per year

WNH

Reduced
Mixed
Mixed

Improved

Cost savings
of 25 million
in 1970

Cost savings
of $3,339,153
in 1974

Cost savings
of $25,000
in one year

Cost savings
of $1,446,505

Cost savings
of $24 million
in the first

97

Table 4-1: (Continued)

Study Sample Productivity Results
Criteria

nine months of

1962
Gunsch United Electric 1. Costs 1. 60% cut in
(1991) Controls Co. inventory
2. Efficiency 2. 90% reduction

in the time
needed to complete
a project

3. A consistent on-
time delivery rate
of 95%, up from
65% from 1987 to
1990

These cases are not concerned specifically with productivity
(defined as a ratio of outputs to inputs) but the kinds of
cost reduction claimed for employee suggestions certainly
imply productivity improvements.

Most cases concern.particular employee suggestions that
"hit a home run"--that is, proposing one suggestion that saved
a huge sum of money (Paulson 1971; NASS 1975; Wilce 1971;
French 1984), while some cases examine some outcomes other
than cost savings, such as employees’ attitude, job security,
safety, super-subordinate relations, and skill development
(Semimer 1959; Denz 1946; Reuter 1977).

There is one problem with the successful case study
reports. Stories of huge savings generated. by a single
suggestion may imply the overall employee suggestion effort is

highly successful and the organization is a high-involvement

98
organization, when in reality there are very few suggestions

generated.

Research on Quality Circles

Essentially, a quality circle is "a small group of
employees from a common work area who get together regularly
to identify and.generate solutions for problems they encounter
in their work situation" (Ledford, Lawler, and Mohrman 1986,
p.256). Membership of the circle typically comprises between
four and a dozen.people who meet under the guidance of a group
leader. They are aided by a facilitator, who trains members,
provides a source of information and encouragement, and act as
a liaison between the circle and remainder of the
organization. Generally, a senior manager is responsible for
ensuring that the circles are able to acquire sufficient
resources to allow the completion of tasks and sufficient
authority put ideas into practices. In addition, the circles
are initiated and driven by a steering committee which acts
across the establishment or organization as a whole (Thompson
1982; Ingle 1982; Crocker and Charney 1984). There are two
distinct sets of objectives behind the implementation of
quality circles. First, they are introduced in order to
increase productivity and reduce costs, improve product
quality and service, basically aims which relate to enhanced

organizational effectiveness. Secondly, they save as a further

99
motive improvement in employee satisfaction and commitment.
However, this literature review will focus on the former
objective rather than latter objective because it is
consistent with the topic, although the two objectives are
closely related and likely to be included in any rationale for
the introduction of quality circles.

Each study’s sample size, productivity criteria, and
major findings is presented in Table 4-2. Although many of
these studies utilized multiple outcome measures, productivity
related measurements were reported only.

Closer inspection of Table 4-2 reveals two types of
studies: those which report data with and without statistical
analysis. Interestingly, all the studies that did not use
statistical analyses to support their conclusions reported
positive results (Murray 1981; Nelson 1980; Industrial Week
1979; Juran 1978; Yager 1979; Arbose 1980; Donovan and van
Horn 1980; Tortorich 1981). In contrast, only two of the six
studies with statistical analysis report positive results
(Marks et al. 1985; Jenkins and Shimada 1984), with one study
reporting negative result (Srinivason, 1983) and five studies
reported. nonsignificant results (Harper' and. Jordon 1982;
Norris and Cox 1987; Mohrman and Novelli 1985; Guantilake
1984; Wolfe 1985). The first cluster of evaluation reports
consists of the anecdotal appraisals and cost savings data
offered by program sponsors as evidence of program

accomplishments. Such reports frequently provide estimates of

lOO

anticipated savings rather than actual cost reductions. These
reports rarely mention how costs and benefits are estimated.
When the estimating procedure is explicit, the figures are
usually based on the estimated value of QC suggestions prior
to implementation. This is an important issue, because many
suggestions are actually never implemented or are implemented
only after a long period of time (Mohrman and Novelli 1985;
Wayne, Griffin, and Bateman 1986). These reports may Hake
overbroad assumptions regarding the productive utilization of
work time stemming from labor saving efficiencies. Thus, the
findings of such reports must be viewed with some measure of
skepticism (Steel and Shane, 1986).

The second cluster of studies provided relatively
rigorous theoretical analysis and scientific statistical
measurement, but these studies rarely provided systematic
examinations of the circle’s effectiveness. One reason that it
may'be difficult towmeasure productivity, even, when.using SQC
methods, is that this approach does not directly measure
productivity. When analyzing productivity increases, most
studies do not report exactly which method they are utilizing
to evaluate productivity measures. Moheman and.Novelli (1985)
found some improvement in productivity, but could not
determine whether or not it was due to the quality circle
program. In addition, Steel and Shane (1986) indicated that
"the majority of studies constituting the quality circle

evaluation literature are, at best, seriously flawed and, at

Table 4-2: The Impact of

Study

Murray
(1981)

Nelson

(1980)

Nelson
(1980)

Nelson
(1980)

Main (1980)

Industry
Week (1979)

Juran (1978)

Yager (1979)

Arbose
(1980)

Sample

300 QCS in
Honeywell

General
Electric Co.

Morton
Chemical Co.

Purchasing
department of

Westinghouse Co.

One department
in Cincinnati
Milacron

Hughes Aircraft
Co.

Television
division of
Motorola Co.

15 QCs in
Lockheed Co.

4 companies

101

QCs on Productivity

Productivity

Criteria

1. Cost Savings

1. Control costs

1. Control costs

1. Control costs

1. Production
rejection rate

1. Cost reduction

1. Product defects

1. Costs savings

1. Return on investment

Results

1. Cost Savings of
$500,000

1. Cost savings of
$15,000 per year

1. Cost savings of
$300,000 a year

1. Cost savings of
$636,000 a year

1. The rejection
rate on an item
was reduced from
50% to zero

1. Annual savings of
$45,000 from the
reduction of
defects and
another $48,000
from the redesign
of sample boards
for assembly work

1. Defect rate was
reduced from 1.8
to 0.04 per tele-
vision set

1. Cost savings of
$2, 844 , 000 in the
first two years
of operation

1. The ROI is esti-
mated at from
five to ten to
one

Table 4-2:

Study

Marks et al.

(1986)

Norris and
Cox (1987)

Jenkins and
Shimada
(1983)

Mohrman and
Novelli
(1985)

Donovan and
Van Horn
(1980)

Tortovich
et a1.
(1981)

(Continued)

Sample

46 circle
members and

46 non-members
in a manufactur-
ing

112 circle
members and

121 nonmembers
in an electronic

450 production
personnel were
divided by 11
QC groups

156 food ware-
house personnel

120 assembly
line workers
(10 QC groups)

94 assembly
line workers
(11 QC groups)

80 assembly
line workers
(3 QC groups)

872 individuals
(463 QC groups)

102

productivity
9mm

1. Productivity

1. Job performance
-dependability
-quantity/qua1ity
of work
-cooperativeness
-safety/health

1. Production
quantity
Quality
Re-work costs

WM

1. Productivity

changes in produc-

tivity

1. Unit assembly
costs

1. Unit assembly
costs

1. Unit assembly
costs

1. Productivity
2. Quality

1.

Results

Overall partici-
pant ’ 3 performance
in productivity,
efficiency, and
working hours are
better than non
participants
(significant
difference
favoring QC
groups)

. No significant QC

effects

Significant
increase in three
of four producti-
vity criteria for
QC groups

No reliable

46% reduction in
in costs over 2
years

36% reduction in
in costs

Significant
difference

favoring QC
groups

Significant
difference

favoring QC
groups

Table 4-2: (Continued)

Study

Guantilake
(1984)

Wolfe (1985)

Srinivason
(1983)

Sample

2 QC groups and
1 non-QC groups
in 2 hospitals

3 QC groups and
3 non-QC groups
a county govern-
ment

Computer firm
(sample size
was not reported)

103

 

Productivity
Criteria

1. Productivity
2. Quality

1. Productivity
2. Quality

1. Productivity

Results

Nonsignificant
improvement in
productivity and
quality by QC
groups

Nonsignificant
improvement in
productivity and
quality by QC
groups

Significant
difference not
not favoring QC
groups

104
worst, potentially misleading. If the level of scientific
rigor found in other field research domains such as job
redesign, survey feedback, and.goal setting may be employed as
a yardstick, then the quality circle literature exhibits

generally inferior quality" (pp 450-451).

Research on Financial Involvement Programs

One of the most significant growth areas of employee
involvement in recent years has been in the field of financial
participation (Poole 1986, 1989; Smith 1986; Lawler 1986;
O’Dell 1981). There are basically three types of Plans: (1)
employee share ownership (e.g., ESOPs), (2) Profit sharing,
and (3) gainsharing plans (e.g. Scanlon Plans, Improshare,
etc.). Poole (1989, pp 70-72) has suggested five sets of
reasons for the introduction of financial involvement; there
are 1) moral commitment by employers, 2) staff retention, 3)
employee involvement, 4) improved industrial relations
performance, and 5) protection against takeover. The most
important factor of financial involvement.probably'is category
(3), which is broad and seems to incorporate a number of
different sets or reasons. Numerous studies supported.the idea
that financial involvement programs that may account for their
success through employee participation (Frost, Wakeley, and
Ruh 1974; Lawler 1986; O’Dell 1981; White 1979; Graham-Moore
and Ross 1983; Ross 1983; Hatcher and Ross 1991; Miller and

Schuster 1987; Klein and Hall 1988; Long 1978, 1980).

105

Financial involvement programs are as much an approach to
participative management as they are a pay plan. Typically,
gainsharing programs, for instance, use some form of
suggestion program as their way of implementing participative
management (Frost, Wakely and Ruh, 1974; White 1979). In the
Scanlon Plan, for example, written suggestions are solicited
and committees are established to process them. Most
literature on the Scanlon Plan cites not only participation
and. communication, but 'willingness, cooperation, and
acceptance of change that occur because of the process of the
suggestion system (Scanlon 1984; Lesieur and Pucket 1968;
Schultz 1958; Ross 1969; Ross and Jones 1972; Northrup and
Young 1968). In addition to suggestion systems, most
gainsharing plans include a committee that is created to
manage the plan and communicate the results.

Studies in the effects of financial involvement programs
on productivity are presented.in.Table 4-3. All of the studies
used statistical analyses to support their conclusions and
found a positive impact on productivity. Although many of
these studies used productivity data for outcome measures,
different measurements for'jproductivity' were found. Some
studies used input measures such as hours worked (Shatter
1984; Doherty 1989). Some studies used output measures
including those of quantity and quality of production and of
cost effectiveness (Schuster 1984; Hitcher and Ross 1991;

Doherty 1989). And some other studies used

2106

Table 4-3: Effects of Financial Involvement Programs on Productivity

Study

Schuster
(1984)‘

Kaufman
(1992)b

FitzRoy and
Kraft
(1987)C

Cable and
Wilson
(1990)c

Wadhwani
and Wall
(1990)C

Kruse
(1934)d

Conte,
Tannenbaum
McCulloch

Hatcher
and Ross
(1991)'

Sample

890 union
production/
repair workers
in two divisions
of an aircraft
repair facility

112 companies

65 firms in the
West Germany
metalworking
industry

61 firms in the
West Germany
metalworking
industry

219 manufacturing
companies in Great
Britain

2 companies

98 employee-
owned firms

An automobile
supplier

Productivity
Criteria

Quantity of units
produced and hours
worked

Productivity

Productivity

Productivity

Productivity

Productivity

Productivity

Quality

Results

Statistically
significant
increases in the
time series analyses
for productivity
data

1.The median produc-
tivity increased by
8% in the first year
2. The cumulative
p:rc>d11c:t iiri t)!
increased by 17.5%
in the third year

Profit sharing has
strong effects on
productivity

Overall.productivity
differentials of 20-
30% in favor of

profit - sharing firms

Significant differ-
ence favoring
profit - sharing firms

Slightly difference
favoring the firm
practicing ESOPs

Managers felt
employee ownership
has a positive (1981)Cl
effect on profit and
productivity

Significant.increase
in product quality

Table 4-3: (Continued)

Study Sample
Doherty An aerospace
et al. firm

(1989)e

A non-profit

A manufacturing

2107

Productivity
Criteria

1. Productivity
2. Quality
3. Cost savings

1. Productivity
2. Cost saving

Cost savings

Cost savings

Results

1.Productivity
improved by 35.3%
over the base period

quality

2.Improved by 44.1%
over the base period
(nonsignificant)

3.52.0 million total
gross savings

1.Productivity
improved by 11% over
the base period

2.$3.0 million in
savings

\

$9.25 million in the
potential net value
of the cost savings

$6.6 million in the
potential net value
of the cost savings

A bank
a Scanlon Plans
b. Improshare
c: Profit sharing
d~ ESOPs
e General gainsharing plan

108

"real" productivity measures in terms cf 21 ratio relating
output to inputs (FitzRoy and Kraft 1987; Cable and Wilson
1990; Wadhwani and wall 1990). Quantitative or "hard"
productivity data were used. as an outcome of financial
involvement programs with the exception of Conte’s et al.
(1981) study. In their study, an indirect assessment of the
productivity performance of ESOPs was conducted. They asked
managers to evaluate their productivity performance since the
intervention of ESOPs.

Basically, most studies claimed that financial
involvement programs produce a more democratic environment as
well as superior‘ channels for’ information-processing' and
conflict resolution. Better conflict resolution reduces labor
turnover and hence increase workers’ tenure with the firm.
Longer average tenure then.translate into higher productivity.
Profit sharing, gainsharing and ESOPs also provide better
incentives and possibilities for workers to acquire human

capital.

Research on Other Forms of Worker Participation

The introduction of new institutional arrangements for
promoting collaborative problem solving between management and
workers has been one of the more widely recognized
transformation since the early 19708 (Kochan, Katz, and

McKersie 1986; Lawler 1991; Cooke 1990). They argue that

1139

Table 4-4: The Effects of Worker Participation on Productivity

Study

Rosenberg
and
Rosenstein
(1980)‘

Bragg and
Adrews
(1973)b

Goodman
(l979)°‘

Nurick
(1985)d

Taylor,
Friedman

and Couture

(1987).

Trist,
Susman and

Brown (1977)‘

Buller and

Bell (1986)'

Katz, Kochan
and Gobeille

(1983)d

Kochan, Katz

and Mower
(1984)”

Sample

262 individuals
in a manufacturing
company

32 hospital
laundry workers

a coal mining
effect

245 employees
in utility
engineering
section

100 telephone
employees

24 coal miners

53 coal miners

5 GM plants
with highest
QWL ratings
for the period
1977-1979

110 observations
from 5 auto plants

Productivity
Criteria

Productivity

Productivity

Productivity

Productivity

Productivity
Quality of service

Productivity

Productivity

Product quality

Productivity
Product services

quality

Results

Positively corrected

Increased

Slightly positively

effect

No effect

1. Improved
2. Higher

Higher

Increased

1. Increased by 1.5%
2. Decreased by 2.4%

1 . Increased by 73 . 3%
2 . Increased by 74 . 6%

Table 4-4: (Continued)

Study Sample

Katz, Kochan 66 responds from
Weber 25 manufacturing
(1985)d plants

French, Ross 800 manufacturing
Kirbby, workers

Nelson and

Smith (1958)‘

22 individuals in
a manufacturing
plant

Schuster
(1983)e

Voos (1987)b 343 Wisconsin firms
with bargaining
units of 3 east 50
employees

Cutcher- 37 Monthly

Gershenfeld observations

2110

Productivity

Criteria

1. Product quality
2. Direct-labor
efficiency

Productivity

Productivity

Productivity
Product quality
Unit labor cost

(AND-4

1. Productivity variance
2. Net return to direct

Results

1. r = .26“

2 r = .17“
Increased
Significant higher
1. Positive effect
2. Positive effect
3. Positive effect

1. Work areas with
transformational'
labor-management
relations have
higher produc-
tivity than work
areas with

"traditional"
relations

1. Firms with joint
labor-management
programs have
greater improve-
ment in product
quality than
firms with no
participation
programs

(1991) across 25 labor hours
work areas in
Xerox Co.
Cooke 1. 194 unionized 1. Perceived extent of
(1992) plants surveyed change in product
in 1986 quality
2. 70 unionized
and 61 non-
unionized
firms surveyed
in 1988
a: Group-based participation
b: Worker participation programs
c: Job rotation
d: QWL programs
e: Labor-management committees
f: Employee participation in production change
i

: The term "transformational" refers to the labor-management relations in a‘work

areas that can be characterized by increased cooperation and improved dispute

resolution (Cutcher-Gershenfeld,

1991).

111
changes in external economic conditions (e.g. the heightened
threat from foreign competition) has provided the need for
employee involvement.programs. The wider utilization.of‘worker
participation systems is also triggered by the desire of
managers to try the innovation to improve company profits and
stagnating productivity.

In. this review, worker" participation includes joint
problem-solving, work group discussion, labor-management
committee, job design (i.e. job enrichment, job enlargement,
job rotation, and job switching). Each study’s sample size,
intervention, productivity criteria and results is presented
in table 4-4.

The sample size ranged from 32 to 800. Most of sample
subjects were engaged in manufacturing jobs. The most often
used economic performance measure in this review was
productivity which was defined as employee output per hour.
All of the studies found a positive effect of worker
participation on productivity with exception of Katz et al.
(1983). Two studies (Katz et al. 1983; and Katz et al. (1985)
found mixed effects on direct-labor efficiency which was
defined as the ration of actual hours of labor input to
standardized hours. Katz et al. (1985) found that direct-labor
efficiency had decreased. Although direct-labor efficiency had
decreased in the five GM plants with high.QWL activity, it was
higher than in the five plants with the low QWL activity. This

suggests that the employees involved.in QWL programs were more

112

efficient than those employees not involved in QWL programs.

Discussion

Table 4-1, 4-2, 4-3 and 4-4 give summaries of 9 employee
suggestion studies, 17 QC studies, 9 financial participation
studies, and 13 other forms of worker participation studies
respectively. Forty-one of the forty—eight studies found a
positive effect for involvement programs on productivity.
These results are consistent with a finding of recent review
of participation research. A meta-analysis by Miller and Monge
(1986) produced.a weighted.mean correlation of r = .15 for the
25 participation-performance correlations included in their
analysis. Three different models were tested in this meta-
analytical study. A cognitive model in which participation was
predicted to have a stronger influence on productivity and
satisfaction for decisions about which employees had
knowledge. An affective model, where it was proposed that
participation would lead to the attainment of higher order
needs (i.e., self-expression, respect, independence) which
would lead to an increase in satisfaction. The last model
discussed was a contingency model, in which theorists
predicted that participation would affect satisfaction and
productivity in different ways across individuals.

When applying a cognitive model, it is predicted that
when employees participate in the decision making process,

they' will attain knowledge that will lead to increased

113

productivity. These results should be even stronger when the
employees participate in.decisions in which they have specific
knowledge. If the cognitive model of worker participation
applies to employee suggestion behavior, it implies that the
effectiveness behind an employee suggestion program stems from
the fact that suggestions are made by employees who know the
problems and.areas of concern for their department. It assumes
that the best person to make a decision affecting the job, is
the individual working in that job. By including employees in
decisions that affect them.and.their jobs, employee suggestion
programs should make the employees even more aware of what is
taking place in their areas.

Further, some other meta-analytical studies (Wangner and
Gooding 1987a, 1987b; Gauzzo, Jackson, and Katzell 1985) also
suggested that employee involvement modestly influences job
performance. However, three narrative literature reviews
(Cotton et al. 1988; Locke and Schweiger 1979; Schweiger and
Leana 1986) found that the relationship between participation
and performance is unclear.

Finally, few studies have really established.the complete
process of the organizational impact of participation. That
is, the influential processes between participation
interventions and productivity are unclear. The question is
how participation affects productivity. If participation
really is positively associated with productivity, is it

because participation directly causes productivity

114
improvements, or is it because participation indirectly causes
productivity improvements via improved job satisfaction or
improved job skills and knowledge. Failure to examine these
intervening variables will seriously limit understanding of
the full influential process between participation and
productivity. Thus, I will attempt to establish a full
theoretical and conceptual model of the effect. of the kaizen-

suggestion system on productivity in the next section.

The Conceptual Model of Effects of the Kaizen-Suggestion
System on Productivity

No study, either theoretical or empirical, has examined
how the kaizen-suggestion system can be expected to affect
productivity. It was discovered that a major difficulty was
the lack of conceptual models available to aid the analysis of
the effects of suggestion making. In this study, I adopt
Sutermister’s(l969) concept that.productivity'i8 a:function.of
technological improvement and human contributions. In other
words, productivity "is not determined solely by how hard and
how well people work. The technical factors play a role,
sometimes an overwhelmingly important one, sometimes a minor
one"(p.5).

The employee suggestion program is a form of employee
involvement or worker participation. Classical studies have
argued that employee involvement can achieve higher job

satisfaction and thereby achieve higher organizational

115

performance. Further, employees improve their job skills,
knowledge and social skills via problem solving and this, in
turn, will improve productivity. Finally, employee suggestion
ideas can lead to labor-savings or capital-savings through
eliminating' the waste in. jprocessing, waiting time,
overproduction, inventory, motion, transportation, defects
etc. and this, in turn will improve productivity too.

Therefore, I argue that a kaizen-suggestion program may
affect productivity by altering (I) work efforts or job
satisfaction, (II) the productive skills of the labor force,
and (III) the organizational efficiency. These three
influential paths of kaizen-suggestion on productivity will be
examined in the model. The process model of the effect of a
kaizen-suggestion program on productivity is represented in
figure 4—1. (I), (II), and (III) mark in figure 4-1
representing three different influential paths of kaizen-
suggestion on productivity improvement. (I) and (II) represent
motivation and ability. First two paths of the figure show
that performance level (i.e., productivity) is a function of
one’s motivation and total job capability. When both
motivation and ability are high, maximum performance can be
achieved. Third path of the figure shows that productivity is
a function.of output and input. When output is increased while
maintaining input, or input is decreased while maintaining
output, productivity improvement can be achieved. By adopting

Sutermeister’s (1969) concepts, path (I), (II) in the figure

116
4-1 can be categorized as human influence, while (III) can be

categorized as technological influence“.

Motivapion Influences
Suggestion Contribution and Psychological Impact

Suggestion contribution to the company should influence
employees’ perceptions of shared common goals, a feeling of
solidarity with the organization, and support of the
organization or loyalty. Being given an opportunity to
participate in decision making (i.e., suggestion making) may
create several kinds of important perceived similarities with
management. There is likely to be, first, a greater sense of
approximate similarity status with management. There is no
longer a wide gap between two kinds of people--those who give
orders and those who take them. To the extent that decision
making is shared, all are on a level where they can contribute
ideas, have them heard, and perhaps make an impact. A second
kind of perceived similarity likely to arise from joint
decision making is a similarity of values and goals. Through
the process of mutual influence which comes with sustained

interaction and through the process of actual agreement on

 

15From an economist’s perspective, the equipment, tools,
knowledge, and skill that go into the transformation process
to convert inputs to outputs are referred to technology
(Chinloy, 1981) . From a I/O psychologist’s perspective, skills
and knowledge in that transformation process are referred to
an individual’s ability or job-capability (French 1958; Wagner
and Hollenbeck 1992). In this study, I adopt I/O
psychologist’s definition.

117
decisions, the perception of shared common goals, and a
feeling of solidarity or loyalty is likely to emerge (Long,

1978, 1980; Patchen, 1970).

Psychological Impact, Job Satisfaction, and
Organizational Performance

Job satisfaction may be influenced indirectly by
suggestion contribution through psychological impact
processes. The perceived importance of the suggestion plan
and perceived influence on decision making is expected to
influence job satisfaction. If employees feel that a'
suggestion program is very important for their financial
benefits, job security, or job involvement, they may be more
satisfied with their job. Moreover, if employees feel that
they have formal or informal influence on decision making via
a suggestion system, they also may be more satisfied with
their job.

On the other hand, employee suggestion programs, through
job satisfaction, lead to improved organizational performance
(both industrial relations and productivity) . Job satisfaction
traditionally was the major independent variable for job or
organizational performance. A causal relationship between
satisfaction and. performance was assumed, that is, high
performance leading to higher job satisfaction. Locke (1970)
has suggested that performance is primarily' a cause of

satisfaction and only indirectly a result of satisfaction.

118

Figure 4-1. A Conceptual Model of Effects of the Kaizen-

Suggestion system.on Productivity

 

[Suggestion Generatgd]

 

[Resource Support}

 

 

 

[Suggestion lementedl

Human Influence

Technological Influence

 

 

 

 

 

 

I II
J
PSYCHOLOGICAL INPACT: ONGOING PROBLEM SOLVING
TRAINING

 

 

*Perceived Importance
of the Plan
*Perceived Influence
of Decision Making
*Perception of Shared

 

 

 

Common Goals , 3:

*A Feeling of Soli- PROBLEM-SOLVING SKILLS
darity with the AND KNOWLEDGE LEARNING:
Organization

*Support of the Orga- *Increased Problem-
nization Solving Skills

*Increased Job
Knowledge

 

 

 

 

I;
[gob Satisfaction]

 

 

.______:1. - *
INDUSTRIAL PERFORMANCE: [INPROVED JOB CAPABILITIES!

 

 

*Absenteeism Rate
*Turnover Rate
*Industrial Accident
Rate

*Orievance Rate
'Discipline Rate

 

 

 

 

 

 

 

III

 

 

SUGGESTIONS IN TECH-
NOLOGICAL CHANGES

*Production.Processes,
Facilities, and
Equipment

*Quality of Inputs

*Control System

:H&$£££222L_____

 

 

F‘

 

[nomwm>nwnnmmr

 

 

 

—O(pnoouc'rmrr mam

119

Recently, however, a cyclical model has been examined.in.which
satisfaction and performance are cause of each other; and the
closer linkage runs from performance to satisfaction.
French,Israel,and As (1960) have suggested that job
satisfaction is au1 intervening ‘variable ibetween. employee
participation in decision making and organizational
performance. Similarly, job satisfaction can be regarded as an
intervening variable between suggestion contribution and
organizational effectiveness. Therefore, the theory of the
impact of the organizational performance of suggestion program
may include two stages: suggestion contribution leading to
high job satisfaction, and then high job satisfaction leading
to high organizational performance. Traditional industrial
and organizational researchers assumed that high job
satisfaction lead to reducing absenteeism, turnover, and
industrial accident rates. Moreover, it was typically assumed
by those researchers that job satisfaction was positively
associated with "job performance" such as productivity,
product quality.

In a sense, workers make daily decisions concerning
whether or not they will appear for work. We would assume
these decisions to be predictable from information about the
anticipated consequences of the alternative. If the
consequences expected from not working are more attractive
than those expected from working, the worker would be

predicted to be absent. On the other hand, if the reverse is

120
true, the worker would be predicted to report for work.
Therefore, job satisfaction would be negatively related to
absences. Some studies support this hypothesis (Herman, 1973;
Vroom, 1962, 1964). In the model, it would.be assumed that if
a suggestion.progranlis attractive enough.for worker to report
for their work, absenteeism will relatively go down.

Hill and.Trist (1953) have suggested that accidents, like
absenteeism and turnover, reflect the strength of motivation
on the part of the individual to withdraw from a work
situation. In support of this view they found that accident
rates are positively associated with other forms of absences
and most strongly associated with the least sanctioned forms
of absence. If this interpretation.is correct, we should.also
expect to find a negative relationship between job
satisfaction and industrial accidents. Thus, dissatisfied
workers should be more likely to have accidents in order to
remove themselves from their unpleasant work situations
(Vroom, 1964, p. 180). Stagner, Flebbe, and.Wood (1952) found
a correlation of -.42 between.the mean job satisfaction scores
and variables for 12 shops in a railroad. We can also assume
that, in the model, suggestion contribution positively affects
jobisatisfaction.and thereby reduce industrial accident rates.

Moreover, job satisfaction would 'result in higher
performance such as productivity because workers will
demonstrate their gratitude for rewards received from

management by increasing their output or that a satisfied

121
worker is more likely to accept managerial goals of higher
production.(Vroom, 1964, p. 182). Some studies have suggested
that increases in satisfaction might result in higher
performance (Vroom, 1960; Locke, 1970; Slocum, 1970). In the
model, we would assume that suggestion contribution cause
higher job satisfaction and thereby increase productivity

improvement.

Industrial Relations and Productivity Improvement

Industrial relations performance may influence economic
performance (i.e.productivity improvement). Katz, Kochan, and
Gobeille (1983) first introduce.both industrial performance
and economic performance into empirical research of worker
participation programs. They use six variables of industrial
relations performance to predict economic performance. These
variables include grievance rate, absenteeism rate, discipline
rate, contract demands, negotiation time, and attitudinal
climate. The evidence has shown that performance of
industrial relations significantly influences the economic
performance of 18 plants adopting QWL Programs within a
division of General Motors. Kochan and Katz (1988, p. 360)
also suggest that "industrial relations performance affects
economic performance." They indicate that the plants with
relatively good industrial relations performance also have
relatively higher productivity and quality. Therefore,

industrial relations performance might be a good predictor of

122

firm overall productivity performance.

Job Capability Influence

Suggestion Making, Training and Problem-Solving Skill
Learning

Employee problem-solving programs (groups vs.
individuals) or kaizen activities generally include a training
component that can improve employees’ job capabilities (Japan
Human. Relations .Association 1992; Pike and. Barnes 1994;
Atkinson 1990; Denton 1991). Kaizen training seeks to provide
the ability to be involved, effectively, in participative
problem solving and to support involvement. Usually, the
objectives of this kind of training are to:

1. Give employees the opportunity to learn by actual
experience the problems of management;

2. Solve a specific problem or situation that impedes the
effectiveness of the organization;

3. Make fuller use of the know-how and resources of kaizen
members;

4. Encourage learning by doing and risk-taking. (Saint 1974,
p.143)

After learning specific skills and knowledge needed for
suggestion generation, workers then apply these skills in the
solution of a problem. Workers build their compentency by
solving' problems. Employee job capabilities can also Ibe

improved through this learning process of problem solving.

123
Because of the opportunity to integrate their learning with
action, the method of learning is highly effective. To
implement continuous improvement suggestions, employees must
consult with supervisors and seek their advice, and such
communication can be highly instructive. Actually, "this is
probably the most effective on—the-job training a person can
get" (Japanese Human Relations Association 1992, p.77).
Therefore, a kaizen-suggestion system can lead to the

improvement of job skills and knowledge.

Training, Skill Learning and Productivity improvement

From a management perspective, any improvement in
productivity' is reflected. in cost reduction (Lawler and
Ledford 1982, p.301; Gregerman 1984, p.130). Generally, cost
reduction in the manufacturing sector can be brought about
either by increasing machine uptime or by reducing manning.
Reducing machine downtime and manning requires that shopfloor
workers be skilled (Ishida 1993). For example, if the standard
job cycle at any particular station is particularly long or
heavy, thus creating a bottleneck and slowing the process, the
individuals may seek to automate portions of that station’s
workload. The result is increased.output per hour and this, in
turn, improves productivity. Without required job skills and
knowledge, however, such.improvements cannot.be'made. Further,
if work processes are interrupted due tx: a machine

malfunction, thus increasing machine downtime, workers have to

124
diagnose the trouble and try to repair it as best they can.
Thus one result of increased worker skills is the reduction of
machine downtime and. thereby" improve jproductivity, Again
without required skills and knowledge, workers may have no
idea how to deal with the machine problems and even cannot do
simple trouble shooting and problem solving. Therefore, the
effect of skills is very strong on input measures of
productivity, The cost savings of reducing downtime and
manning lie on the input side of the productivity equation.
That is, the efficiency of transforming inputs into outputs
increases as these particular types of labor-related input
costs decline. In sum, improved job skills and knowledge are
by-products of the kaizen-suggestion system. Once workers’ job
skills and knowledge are improved, they can improve
organizational productivity by doing machine maintenance,
trouble shooting, problem solving, changing parts or even
repairing equipment. Consequently, the kaizen-suggestion
system leads to organizational productivity improvement via

improved employee job capabilities.

Technological Influence
Kaizen-Suggestion, Technological Change and Productivity
Japan Human Relations Association (1988) stresses that a
suggestion system is to achieve:
1) Improvements in work methods;

2) Improvements in tools, machinery, and equipment;

126
For example, if one suggestion can minimize transportation,
energy, or inventory costs or reduce waste in processing”,
it can reduce input levels and thereby increase productivity.
(3) Improvement suggestions in control systems:
For example, if a suggestion can increase automaticity, it can
reduce work operational time as well as increase output
levels, and this, in turn, improving productivity.
4) Improvement suggestions in changing product design:
For example, if one suggestion can improve the quality of
services to users (including convenience, flexibility,
durability, reliability, and safety), it increases the added
value of a product and thereby improves productivity in the
long run. Once suggestions lead to technological changes in
production processes, facilities, equipment, quality of
inputs, control system and product design, productivity can be
improved by decreasing input level, increasing output level,

or changing both input and output levels.

Hypotheses

Prgdugtivity and Labor Inpu; Models

Employee suggestions can improve productivity through

both human influences and technological influences as

 

16The kaizen-suggestion system in Cannon Co. is to eliminate
following wastes: 1) waste caused by work-in-process, 2) waste
caused by defects, 3) waste in equipment, 4) waste in expense,
5) waste in indirect labor, 6) waste in planning, 7) waste in
human resources, 8) waste in operations, and 9) waste in
startup (Dyer 1987, pp 17-18).

125
3) Improvements in organization and safety;
4) Improvements in transportation;
5) Improvements in cost-cutting;
6) Improvements in energy conservation;
7) Improvements in clerical work; and
8) Improvements in sales operations.
Clearly, most goals of the suggestion system are to change
production technology and thereby improve productivity. This
is especially correct for the first two objectives of the
suggestion system as listed above. Gold (1975, 1979) and.Gold,
Peirce, Rosegger (1970) have argued the technological changes
which are most likely to affect productivity are largely
encompassed by the following categories: 1) changes in the
nature of production facilities, and equipment, 2) changes in
the quality of inputs, 3) improvements in control system, and
4) changes in production design. Clearly, the above goals of
the suggestion system can be broken down by these five
categories.
(1) Improvement suggestions in changing the nature of
production facilities and equipment:
For example, if one suggestion improves work method. or
process, it can reduce work operational time and thereby
increase efficiency. Again, if one suggestion extends the
machine’s life, the machine becomes more profitable and return
on investment is greater.

(2) Improvement suggestions on changing the quality'of inputs:

127

discussed in the last section. Suggestion contribution to the
company may influence employees’ perceptions of shared common
goals, a feeling of solidarity with the company, and support
of the organization. Improved employee attitude and morale,
thus, may lead to improvements in productivity. In addition,
employees learn problem-solving skills and knowledge in the
process of suggestion making. Increased job skills and
knowledge, thus, lead to productivity improvements.

Improvement suggestions may improve productivity' by
changing technology. There are several ways that employee
suggestions can improve productivity. First, suggestion ideas
improve work methods and processes, which may contribute to
productivity improvement. Second, suggestions improve
equipment, tools, and machinery, which may also lead to
improvements in productivity. Third, better operating
conditions, production quality, and safety environment may
also contribute to productivity improvement. Fourth,
improvement suggestions that reduce the input requirements and
ease the problems of the company may lead to productivity too.
Finally, improvement ideas for reducing transportation, energy
costs and other wastes in processing may also result in
significant productivity improvement. We might expect,
therefore, a positive relationship between suggestions and

productivity.

Hypothesis 1.

The greater the number of the adopted tangible or

128

intangible suggestions, the greater the productivity

gains.

Generally, tangible suggestions may result in
technological changes and tangible financial benefits such as
labor savings, material savings, downtime reduction, output
increases, etc. Most of them result in technological and
production improvements that directly lead to productivity
improvements. On the other hand, intangible suggestions
generally contribute to production quality, safety,
housekeeping, convenience improvements, or improvements in the
quality of working conditions. Most of them result in non-
technological and.non-production improvements that indirectly
contribute to productivity improvements. The impact of
intangible suggestions on productivity might be smaller than
that of tangible suggestions. Therefore, the expected
relationship between tangible and intangible suggestion

effects on productivity can be described in hypothesis 2:

Hypothesis 2.

Tangible suggestions have greater effect on pnihcthdty
than intangible suggestions.

In measuring the impact of suggestions on productivity,
time is a critical variable. As mentioned above, total volume
of adopted suggestions in the current period of time is
expected to positively relate to productivity during the same

period of time. The interesting question raised here is

129

whether productivity not only depends on present volume of
suggestion accepted, but also depends on past volume of
suggestions accepted? There are several reasons why there
might be a lag effect in the suggestion-productivity
relationship (i.e. a lapse of time between a change in an
explanatory variable (suggestion) and a change in the
dependent variable (productivity)). First, therermight be time
lags in the process of transformation.of suggestion ideas into
operations and of operations into productivity gains. It also
takes time when old types of machinery, tools, equipment, or
regulations are replaced by new technology or organization
rules. Secondly, it takes time to train employees to learn.new
work methods and operate machinery, tools, or equipment. It
also takes time for employees to live with new organizational
rules and regulations. Finally, since behavior is often based
on habit, employees who are used to an old way may resist the
new way. Hence, we might expect that there are time lag
effects of suggestions on productivity.

Although there might be a delayed or lag effect between
suggestion activity and productivity performance, there may be
a different pattern of effects over time between tangible -and
intangible suggestions. In general, most tangible suggestion
ideas result in technological and production improvement,
while intangible suggestion ideas result in nontechnological
and nonproduction improvement. Technological and production

improvement generally involves the reallocation of resources

130

such as renewal of equipment, labor and capital readjustment,
so technological and production.improvement which.results from
tangible suggestion generation may take time to respond to
productivity gains. On the other hand, nontechnological and
nonproduction which result from intangible suggestion
generation.may also take time to respond.to productivity gains
but it may be much shorter than technological and production
improvement. Therefore we might expect that tangible
suggestions have a longer lag structure than intangible
suggestions. In other words, productivity response to tangible
suggestions is slower than to intangible suggestions. Thus on
the basis of this reasoning, the hypothesis 3 can be stated as
follows:
Hypothesis 3.

A time lag effect exists between suggestions adopted

and productivity gains.
Hypothesis 3.1.

Current and past adopted suggestions are relevant in

determining productivity improvements.
Hypothesis 3.2.

The lag for tangible suggestion effect is longer than
for intangible suggestion effect.

Based on the discussion and reasoning of the relationship
between suggestions and productivity, further examination in
the relationship Ibetween. suggestion. and "labor input8"--

measured by total working hours--will be presented in this

131
section.

Generally, suggestions in improving work methods,
processes, or equipment may reduce operational time. However,
operational time may be increased shortly after the
implementation of suggestions simply because workers lack
experiences in how to transform suggestion ideas to
operations. For example, when old types of machines, tools or
equipment, are replaced by new machinery, workers could make
mistakes in operating the new machines, which should slow the
work processes and thereby increase operational time.
Operational time may be lsot in "trial and error" in the
initial stage of suggestion implementation. After this stage,
operational time may decrease in response to improvement
suggestions. From this reasoning, the next hypothesis can be
stated as follows:

Hypothesis 4.1

There will be a lagged effect with a curvelinear

relationship between working hour inputs and

suggestions.
Hypothesis 4.2.

Tangible suggestions have a greater effect on working
hours than intangible suggestions.

In addition. to adopted. suggestions, there 'are many
control variables with effects on product ivity and labor
inputs that can. be anticipated. These control variables

include quality, safety, absenteeism, training, technology and

132
organizational size. The expected relationships between these
control variables and productivity or labor inputs are simply
explained as follows.

If an organization produces products with defects, it
hinders organizational productivity growth. A deterioration in
the quality'of the products or services can.disrupt schedules,
delay' deliveries, increase rework, increase scrap, waste
manpower and materials and machine time, and.increase warranty
cost. Reworking products, inspecting parts, and the product
lost due to scrap all lower productivity. Thus, a negative
relationship is expected between the product defects and
productivity.

An unsafe environment may inhibit individual and
organizational. performance. lMore specifically, the safety
issue confronting organizations is cost related (Bittel,
1968). Thus, the payoff in productivity could be substantial.
Above tangible and intangible costs reside in the input side
of the productivity equation. That is, efficiency of
transforming inputs into outputs declines as these particular
types of labor-related input costs rise. Therefore, a positive
relationship is expected between industrial accidents and
labor inputs, and thus a negative relationship is expected
between industrial accidents and productivity.

Absenteeism is generally regarded as costly to an
organization (Katz, Kochan, and Keefe 1987; Katz, Kochan, and

Weber 1985; Katz, Kochan, and Gobeille 1983). The costs of

133

paying absent employees and their replacements, of filling
vacancies made by those who left, and of performing at lower
efficiency by the substitutes, reside on the input side of the
productivity equation. If outputs remain unchanged, increased
labor-related input costs decline productivity. Thus, a
negative relationship exists between absenteeism and
productivity.

Training has long been used for improving productivity
(Katzall and Guzzo 1983; Kopelman 1986). One objective of
training is to change employees’ skills, behaviors, and
attitudes in a way that will enhance job performance, either
immediately or in the long run. In their meta-analysis
comparing the effects of various productivity programs, Guzzo,
Jette, and Katzell (1985) found that training was the most
powerful means of increasing productivity. The effect of
training was strongest on output measures of productivity.
Improved productivity is achieved by transforming better
employees skills, behaviors, or attitudes into higher level of
outputs. Thus, a positive relationship between training and
productivity is expected.

"Productivity is not determined solely by how hard and
how well people work. The technical factors play a role,
sometimes an overwhelmingly importance, sometimes a minor one "
(Sutermeister 1969, p.5) . Thus, the degree of technology would
have an effect on productivity. In this study, the definition

of technology is the machinery and equipment employees have to

134

work with. Generally, this can be dichotomized to "machining"
and "assembly." Technology in terms of machining and assembly
affects productivity in two different ways. First, the higher
the degree of technology, the more efficiency in transforming
inputs into outputs. In other words, productivity in machining
areas may be higher than productivity in assembly areas.
Secondly, employees who work on machining lines always have
more skills and knowledge than employees who work on assembly
lines. Skills and knowledge result in the ability that is one
of the determinants of productivity improvement. This
reasoning also supports the argument that productivity in
machining areas may be higher than productivity in assembly
areas.

In addition to the dominant technology employed by the
firm, organizational structure also affects both aggregate and
individual level productivity. The term "organizational
structure" is extremely broad, its definition depends on the
particular school of thought that one is currently reading. In
this study, organizational size will be employed. Anecdotal
evidence supports the ideas that organizational size is
negatively related to its effectiveness. Dalton, Todor,
Spendolini, Fielding and Poter (1980) have indicated that
subunit size is negatively related to organizational
performance. The larger the size of the subunit, the lower the
level of performance (in 5 out of 6 cases). On the contrary,

Cummins and King (1973) found that size was positively related

135
to performance on structured, routine tasks, but negatively
related to performance on unstructured, ambiguous tasks.
The various hypotheses about expected relationships among

these control variables will be presented in Appendix D.

Quality Model
Japanese .Human. Relations .Association (1989) has
identified two broad sets of objectives of kaizen activity:

1. Tangible results: improve efficiency of operations, reduce
prime costs, eliminate poor quality.

2. Intangible results: improve safety, quality, environment,
and service.

In addition to productivity improvement, quality improvement
is also a pivotal goal of the kaizen-suggestion system. There
are two different ways that quality can be improved via a
kaizen-suggestion program. First, quality improvements
directly result from suggestions that lead to any idea that
reduces or eliminates defects. Secondly, quality improvements
can be accomplished indirectly via suggestions that lead to
any idea that improves technology or production system such as
work processes, equipment, etc. In other words, once
technology or production systems are improved, better quality
of products can be produced. On the basis of this reasoning,
I propose hypothesis 5:

Hypothesis 5.

A negative relationship exists between the volume of
adopted tangible and intangible suggestions and the

136

number of customer claims for product defects.

As mentioned above, because most quality improvement may
be derived from intangible suggestions other than tangible
suggestions, intangible suggestions are expected to have
greater effects on quality than tangible suggestions.

The next hypothesis is:

Hypothesis 6.

Intangible suggestions have greater effect on quality

than tangible suggestions.

In addition to adopted suggestions, there are also many
control variables with effects on quality that can be
anticipated.iri‘this model. These control variables include
training, absenteeism, safety and over time. The expected
relationships between these control variables and quality can
be explained as following.

The relationship between training and quality improvement
is pretty straightforward. Training is one of the sources of_
labor quality change. The objective of training is to change
workers’ skills, knowledge, behavior and attitude. Employees’
capabilities and motivation should be improved via effective
training programs. High quality workers produce high quality
products. Training, thus, may positively influence product
quality.

Absenteeism may be negatively related to quality. Poor

quality of products or service might be produced because of

137
generally higher scrap or spoilage of substitutes. Quality of
products or services may also suffer because of the absentee’s
poor work motivation. Thus, a negative relationship may exist
between absenteeism and quality.

Unsafe environment should hinder organizational
effectiveness in terms of quality performance. A good product
is unlikely to be produced without a safe environment. Unsafe
equipment, tools or machines will very possibly'produce flawed
products. Further, product quality' may' suffer" because a
replacement for the recuperating worker is always less skilled
in the substitutive job. Thus a positive relationship is
expected between the number of accidents and the number of
customer claims for product defects.

Overtime may be negatively related to quality in two
different ways. First, flawed products can be reworked with
additional labor inputs and these additional labor inputs
could be reflected in more overtime hours. Thus, more overtime
hours may imply more flawed products in an organization.
Secondly, if workers work excessive overtime, they'may'perform
the job incorrectly, thereby producing defective products
simply because of fatigue. Consequently, overtime may be
related to, or lead to poor quality products or services. The
greater the number of overtime hours, the larger the number of
customer claims for product defects.

The various of hypotheses about expected relationships

among these control variables will be presented in.Appendix.E.

138
Summary of Hypotheses

Hypothesis 1.

The greater the number of adopted tangible or intangible
suggestions, the greater the productivity gains.

Hypothesis 2.

Tangible suggestions have greater effect on pnxmcohdty
than intangible suggestions.

Hypothesi 3.

A time lag effect exists between suggestions adopted
and productivity gains.

Hypothesis 3.1.
Current and past adopted suggestions are relevant in
determining productivity improvements.

Hypothesis 3.2.

The lag for tangible suggestion effect is longer than
for intangible suggestion effect.

Hypothesis 4.1.
There will be a lagged effect with a curvelinear
relationship between working hour inputs and
suggestions.

Hypothesis 4.2.

Tangible suggestions have greater* effect on. working
hours than intangible suggestions.

Hypothesis 5.
A negative relationship exists between the volume

adopted tangible and intangible suggestions and the
number of customer claims for product defects.

139

Hypothesis 6.

Intangible suggestions have greater effect (n1 quality
than tangible suggestions.

Research Models

Productivity Model

Where

1,11 = a0 + a1 TANGi,t-k + 8.2 INTANGi't_k
+ a3 MGTTRANLt + a4 SELFTRANL,
+ as TECHTRANiIt + a6 QUALITYLt
+ a7 SAFETYiIt + a8 ABSENTi't
+ a9 TECHL, + a1, 51213,,t + ei,t
Pr- = Productivity level in department i at month t.
Tangi'w, = The volume of adopted tangible suggestions
in department 1 for month t-k.
Intang”,k = The volume of adopted intangible
suggestions in department 1 for month t-k.
Mgttrani't = Management training credit hours in
department i for month t.
Selftrani,t = Self-actualization training credit hours
in department 1 for month t.
Techtrani,t = Technical training credit hours in

department 1 for month t.

140

Qualitth = The customer claims for product defects in
department 1 at month t.

Safety,“t = The occupation injuries and illnesses in
department 1 at month t.

Absenti,t = The absent hours in department i at month t.

Techi,t = The degree of technology used in department
i at month t.

Sizei,t = The size of work force in department i at
month t.

Labor Inputs Model

Hl,t = a0 + a1 TANGi,t-k + a2 INTANGi't_k
+ a3 SAFETYi't + a, QUALITY,”

+ a5 ABSENTi,t + 91,:

Where

Hi,t = Labor hour inputs in department 1 at month t.

Tangi'b, = The volume of adopted tangible suggestions
in department i for month t-k.

Intang”,k = The volume of adopted intangible
suggestions in department 1 for month t-k.

SafetyL, = The occupational injuries and illnesses in
department i at month t.

Qualityi,t = The customer claims for product defects in
department 1 at month t.

Absenti,t = The absent hours in department i at month
t.

141

MW

Qi't = a0 -+ an TANGL,* + a.2 INTANGi,k
+ a31MGTTRANi,t + a4 SELFTRANL,
+ a5 TECHTRANLt + a6 SAFETYLC
+ a3 OVERTIMEi,t + a5 ABSENT”

Where

Qi't = The customer claims for product defects in
department i at month t.

Tangi'bk = The volume of adopted tangible suggestions
in department i for month t-k.

Intangilbk = The volume of adopted intangible
suggestions in department 1 for month t-k.

Mgttrani,t = Management training credit hours in
department 1 for month t.

Selftrani't = Self-actualization training credit hours
in department i for month t.

TechtranL, = Technical training credit hours in
department i for month t.

Safetyi,t = The occupational injuries and illnesses in
department i for month t.

OvertimeL bk = The overtime hours in department i for
month t-k.

Absenti'bk = The absent hours in department i for month

t-k.

142
Empirical Results and Implications

The data for this pooled cross-sectional and time-series
study consist of 19 cross-sectional departments and 44
time-periods thus 836 observations are estimated” This
chapter examines pooled departments’ productivity, labor hour
inputs, and product quality with current and lagged volume of
tangible and intangible suggestions, and other control
variables such as safety, workforce size, overtime,
absenteeism, technology, etc. My previous study of kaizen-
productivity connection which consisted of 44 cross-sectional
work teams and 17 time periods had indicated that there was a
strongest response of productivity to tangible suggestions
with an average lag of four months, while there was a
strongest response of productivity to intangible suggestion
with an average lag of two months. In this dissertation I
estimated.the parameters.by using direct distributed.lag model
with.three-period lags (i.e. lagged one, two and.three months)
for both tangible and intangible suggestions in all three
models. This will allow full consideration of lags of up to
one quarter, but will still leave open the possible effects of

longer lags.

Relationships Between Tangible suggestions, Intangible
Suggestions, and Organizational Effectiveness and Efficiency

Table 4-5 presents means, standard deviations, and
correlations between tangible suggestions (t, t-l, t-2, t-3,

t-4), intangible suggestions (t, t-1, t-2, t-3, t-4) and

mod

ha. no. vo. we. so. «0.. NO. no. no.. «0.: no.) No.- «v. ma. >06dd90 .nH
an. ac. ho. ho. ho. no. «0.: co. co. fio.u Hc.u n.00wv a.noom IHSON Manda .NH

66. 66. 66. 66. «6. 66. 66. 66. 66. 56. 66.66 66.«66 666>6uosvoum .66
.6-6. «6666-6656

56. 66. 66. 66. 6«. 66. 66. 66. 66. 66.56 65.«6 o6n6uanuaH.66
.6-6. 6666666696
56. 66. 66. 66. 6«. 66. 66. 66. 66.56 65.«6 6666666666 .6
26-6. ao6uuomman
56. 66. 66. 66. 6«. 66. 66. 66.56 65.«6 o6n6mauun6 .6
.6.». a6666666=6
56. 66. 66. 66. 6«. 66. 66.56 65.«6 6666666666 .5
any sedueemmsm
66. 66. 66. 66. 6«. 66.56 «5.«6 66666666a6 .6
.6.». 6666-66696
6«. 6«. 66. 66. 6«.6 66. 66666668 .6
.6.». 6666-66696
6«. 6«. 66. 6«.6 66. 66666668 .6
.«-6. ao6u-oumsu
66. 6«. 6«.6 66. 66666668 .6
16-63 no6uuoumna
6«. 6«.6 66. 66666668 .«
Au. nouueemmnm
6«.6 66. 66666665 .6
«6 66 66 6 6 5 6 6 6 6 « 6 .o.6 666:
uo6nu6ua>

.wno I z. esudesel hoasaoauum use
eeeus>6uueuum Heao6usn6semuo one .edoaueeMMSm eanwmnsunu .eaouuesmmSm eaadunsa asetusm eadneaowususu 6m-v sands

144

organizational effectiveness and efficiency (i.e.

productivity, quality and labor hours). The data shows the
strongest intercorrelations between suggestion variables
(tangible and intangible) and productivity. Correlations range
from .17 to .19 for tangible suggestions, and range from .12
to .15 for intangible suggestions. The correlations provide
strong support for Hypotheses 1 and 2---that is, (1) the more
current and past volume of adopted tangible and intangible
suggestions are associated with productivity, and (2) tangible
suggestions have greater effect on productivity than that of
intangible suggestions. The correlations between tangible,
intangible suggestions, and productivity also reveal a pattern
of lagged effect, providing support for Hypothesis 3. The
correlations for tangible suggestion effects are smaller in
the current period (t), peak at t-1 and t-2, decline
thereafter. Tangible suggestion exhibit lag functions which
peak at t-l and t-2 other than the current period (t). The
same lagged patterns hold for intangible suggestion effects.
The finding strongly supports Hypothesis 1 to 3. Further

hypothesis tests will be reported in the next section.

Empirical Results from the Productivity Model

The level of departmental productivity' was used as
dependent variables; the current (t) and past (t-1, t-2, and
t-3) volume of tangible and intangible suggestions addressing

a joint effect of kaizen processes and worker participation;

Table 4-6: Results of Regression Analyses of the Effect of Suggestions and

Productivity
Explanatory Productivity
Variables (l) (2) (3)
Intercept b 95.501"' 95.457"' 97.748"'
t 155.690 159.390 156.930
Tang, b . 325'" . 392'"
B .012 .014
t 3.532 4.459
TangM b .543'" .568'”
B .020 .021
t 5.091 5.625
TangM b .378'" .421'"
B .014 .015
t 3.546 4.154
Tang,., b . 383'" . 380'"
B .014 .015
t 4.178 4.312
Intang, b -.004 -.002
B “.002 “.001
t -.606 -.284
Intang,_1 b .002 -.003
B .001 -.002
t .306 -.462
Intangh, b .001 .016'
B .007 .008
t 1.553 2.307
Intang“, b -.010 -.005
B “600‘ -6002
t -1.358 -.648
Mgttran b .020"' .020"' .013'
8 .010 .001 .006
t 3.022 2.848 1.900
Selftran b .030' .031‘ .029'
B .007 .007 .007
t 1.976 2.117 2.147
Techtran b -.007 -.006 -.0003
a “600‘ -0003 -6000:
t -.955 -.846 -.052
Safety b .081 .064 .091
B .003 .002 .003
t .947 .771 1.168
Quality b -.643"' -.665"' -.795"'
8 -.008 -.008 -.010
t -2.741 -2.912 -3.624
Size b .275"' .275"' .252'"
B .199 .199 .182
t 19.699 20.286 17.803
Absent b .001" .001"' .0005
B .008 .008 .004
t 2.545 2.709 1.526
Tech b .005 .005' .003
B .005 .005 .003
t 1.565 1.711 1.065
R‘ 419 .421 .339
P 36.839 49.808 35.153

145

All Other Variables on

P-test for all estimates significant at the .01 level.
9 e p < .05 level
*0 - p < .01 level

0'0 - p < .005 level

b is the regression coefficient.

8 is the standardized coefficients.

t is the t value.

146

and the training, safety, quality, size, absent, and
technology as control variables . Table 4 - 6 reports these
regressions.

The over-time pattern of joint effects for both tangible
and intangible suggestion variables is presented in equation
1 in table 4-6; and simple effects for tangible and intangible
suggestion 'variables are presented in equation 2 and 3
respectivelyu The tangible suggestion effect in equation 2 is
smaller in the current period, peaks at t-1, declines
thereafter. Tangible suggestions exhibit lag functions which
peak at t-1 other than the current period. The regression
shows statistically significant effect for current period and
for three lagged variables on productivity improvement at .005
level. It suggests that the current level of productivity is
a function of the current and previous volume of the tangible
suggestions. The regression analysis suggests that each
additional tangible suggestion increases the productivity by
.392, .568, .421, and .380 for current period and a lag of one
month, two months, and three months respectively. Note that
the regression coefficient for a lag of three months in
equation 2, table 4-6, still show positive signs and
significance, so that the duration of lag effects of tangible
suggestion may be longer than three months.

The intangible suggestion variables in equation 3,
table 4-6, show negative signs at current period and a lag of

one month, peaks at t-2, damp out at t-3 with a negative sign

147

again. The association between productivity and intangible
suggestion with a lag of two months is statistically
significant at .05 level. Regression analyses show that
intangible suggestions have negative effects on productivity
at current period and t-l, and have significant positive
effects on productivity at t—2. It suggests that intangible
suggestions seem to have no effect or even a minuscule
negative effect on productivity until a lag of two months.
Even though regression analyses provide evidence of
statistical significance of the lagged intangible suggestions
(t-2), the estimated size of the effects is weak (standardized
coefficient is .008).

The equations 1, 2, and 3 in table 4-6 show that the more
current and past volume of adopted tangible and intangible
suggestions are associated with productivity. It is
interesting to note that when both tangible and intangible
suggestions are jointly employed in the equation 1 of table 4-
6, the coefficient for intangible suggestions with a lag of
two months is no longer statistically significant in
comparison with the equation 3, suggesting that the effect of
intangible suggestions on productivity is mediated through
their' effects (n1 tangible suggestions. It. implies that
intangible suggestions may be through tangible suggestions
that lead to improved productivity. In other words, the
processes of the impact of suggestions on productivity may

include two stages: more adopted intangible suggestions

148

stimulating more tangible suggestion making, and in turn more
adopted tangible suggestions leading to productivity gains.
Therefore, tangible suggestion can be regarded as an
intervening variable between intangible suggestion and
productivity gain. It partly supports the results of the
connection between intangible and tangible suggestions found
in chapter 3.

Interestingly enough, there are different pattern effects
between tangible and intangible suggestions. Tangible
suggestions have strong and significant effects on
productivity after a lag of three months, and its duration
seems to be longer than that. In contrast, intangible
suggestions only have greater effects on productivity in.a lag
of two months and that this effect weakens at t-3. The
empirical results, thus, suggest that tangible suggestions
seem to have longer duration of lag effect, while intangible
suggestions tend to have shorter duration of lag effect. It
implies that tangible suggestions paid off slower but longer
than intangible suggestions.

There are also different magnitude effects between
tangible and intangible suggestions. Both R2 and F value for
equation 2 (only tangible suggestion variables are included)
are greater than those in equation 3 (only intangible
suggestion variables are included). In addition, in equation
1, accumulated standardized coefficient of tangible

suggestions (.060) is also greater than that of intangible

149
suggestions (.016). Therefore, tangible suggestions tend to
have greater effects on productivity than that of intangible
suggestions. It implies that tangible suggestions paid off
more than intangible suggestions.

More extensive management and self-actualization training
lead to higher productivity. In equation 1 through 3, the
association between management training and productivity is
statistically significant at either .005 or .05 level; and the
association between self-actualization training and
productivity is consistently statistically significant at the
.05 level. Interestingly enough, there is no relationship
between technical training and productivity. Even the
regression results consistently show negative signs.

In equation 1, 2, and 3, quality measured by product
defects is found to have a negative effect on productivity.
It suggests that the higher the number of product defects, the
lower the level of productivity. All coefficients in the
three equations are statistically significant at .005 level.

Technology is found to have a positive effect on
productivitym It suggests that the machining departments tend
to have higher productivity performance than the assembly
departments. However, the evidence from regression analyses
is weak.

Technical training, safety, workforce size and
absenteeism did not affect productivity in the hypothesized

manner. The results are contrary to the hypotheses. More

1150

Table 4-7:
Variables on Labor Hours

Results of regression analyses of the effect of suggestions and All Other
(A Three-Month-Lag Model for Suggestion Variables)

--------o---n---------------—--—--———---------—-—---------—------—------—---------—

Explanatory Labor Hours
Variables (1) (2) (3)
Intercept b 6381.100”' 6250.600"' 6347.800"'
t 70.755 90.851 78.575
Tang, b .940 1.532
B .0002 .0004
t .133 .226
Tang51 b .028 -2.844
B .7E-05 -.0007
t .003 -.349
'ra-ng..2 b 17.245' 14.307'
B .004 .004
t 1.998 1.745
Tangp. b 13.302' 11.039
B .003 .003
t 1.879 1.609
Intang, b 3.394"' 3.481"'
B .013 .013
t 4.338 4.656
IntangM b -1.199 -1.155
B -.004 -.004
t -1.393 -1.404
IntangH b -.802 -.648
B -.003 -.002
t -.885 -.743
Intangb, b -3.390"' -3.156"'
B -.012 -.012
t -3.522 -3.721
Safety b 72.275"' 66.775"' 75.525"'
B .019 .018 .020
t 7.499 6.920 8.015
Quality b 115.730"' 131.000“' 117.690"'
B .010 .012 .011
t 4.888 5.545 5.086
Absent b -l7.929' -19.505" -16.003'
B -.006 -.007 -.005
t -2.251 -2.472 -2.141
R2 .128 .098 .131
F 10.960 12.868 17.841
F-test for all estimates significant at the .01 level.
* a p < .05 level
** = p < .01 level
*‘* = p < .005 level

b is the regression coefficient.
B is the standardized Coefficient.
t is the t value.

151

technical training leads to lower levels of productivity, not
higher as hypothesized. Higher number of accidents leads to
higher levels of productivity, not lower as predicted” ZBigger
workforce size leads to higher level of productivity, not
lower as predicted by an "efficient resource" theory, and the
coefficients were statistically significant at the .005 level.
Finally, a higher absenteeism ratio leads to higher level of
productivity, not lower as expected, and the coefficients were
significant at the .005 or .01 level.

In summary, the empirical results reported in table 4-6
show that more adopted improvement suggestions, more
‘management and. self-actualization. training, fewer jproduct
defects, and relatively high technology are associated with

productivity improvement.

Empirical Results from the Model of Labor Hour Inputs

Table 4-7 reports the results of the regression estimates
of the model of labor inputs with lagged suggestions of three
months. When I examined the regression results about the lag
pattern of tangible and intangible suggestions, I found that
there might be a problem in selecting the optimal lag.
Equation.1 and.3 in table 4-7 show that intangible suggestions
have a positive effect on productivity at current period and
negative effects thereafter. It suggests that intangible
suggestion making may be expensive in the short run but it

pays off in the long run. Considering the pattern of

152

Table 4-8: Results of RegressionHAnalyses of the Effect of Suggestions on Labor Hours (Legged

Three Months)

~~¢~~pn~~ad---wano~-purn-an..-—nun-.66..--66-.-enema—enn—nut—ame-——‘-~----me-am—~¢-—----—-—-—~--—-—~——-

Intercept

Tang,

Tangb,

Tang“,

Tangz-a

Intang,

IntangH

Intang,.2

IntangH

(7010' "(110' "(110‘ "(HO

an‘ "010'

”(110‘

15

12

-2

-2

-2

-3

-4

.600'"
.212

.245
.001
.750

.749
.0002
.105

.723'
.004
.213

.080'
.003
.159

.872'"
.011
.725

.043'
.008
.308

.595.H
.010
.843
.537'"
.013
.275
.065

.191

1.

.300'"
.678

.695
.002
.549

.527
.002
.210

.505
.002
.047

.409

.002
.238

.009

829

6420.
38.

-1.
.007
.311

-2

-2.
.009
.676

14.

000
179

.953'"
.011
.054

904'

283'"

.449."
.013
.385

.065

F-test for all estimates significant at
t a p < .05 level

it
if.

p < .01 level
p < .005 level

b is the regression coefficient.
8 is the standardized Coefficient.
t is the t value.

.01 level.

1553

Table 4-9: Results of Regression.Ana1yses of the effect of Suggestions on Labor Hours (Legged
Four Months)

Explanatory Labor Hours
Variables (1) (2) (3)
Intercept b 6249.400"' 6198.300"' 6344.600'"
t 31.126 41.437 34.071
Tang, b -8.099 -9.861
B .002 .002
t -l 177 -1.579
TangM b -8.072 -15.975'
B -.002 -.004
t -.926 -2.041
TangM b 1.864 -6.140
B .0005 -.002
t .208 -.769
Tanga. b -12.948 -18.704"
B -.003 -.005
t -l.510 -2.427
TangH b -35.811"° -37.886"'
B -.009 -.010
t -5.289 -6.154
Intang, b 3.340 3.093"'
B .012 .012
t 4.401 4.242
IntangH b -1.273 -1.S93'
B -.005 -.006
t -1.316 -1.909
IntangH b -1.989' -1.851'
B -.007 -.007
t -1.954 -2.096
Intang,.3 b -2.49S"' -2.687"'
B -.009 -.010
t -2.470 -3.070
Intangp. b 1.131 1.376'
B .005 .005
t 1.479 1.763
R2 .086 .050 .068
F 7.727 8.657 12.031

F-test for all estimates significant at the .01 level.
* = p < .05 level
** = p < .01 level

*** = p < .005 level

b is the regression coefficient.

B is the standardized Coefficient.

t is the t value.

154

intangible suggestions, it is reasonably assumed that tangible
suggestions may have a similar pattern as intangible
suggestions. As I examined equation 1 and 2 in table 4-7, the
lagged pattern of tangible suggestions seems to only tell us
half the story, That is, tangible suggestions cost more labor
inputs but they do not pay off. I suspected that adopted
tangible suggestions may have paid off longer than a lag of
three months. In addition, my previous work for pooled team
data suggested that tangible suggestions tend to have longer
effects on productivity than that of intangible suggestions.
Considering these two reasons, I conducted two tests to
examine whether adopted tangible suggestions pay off longer
than a lag of three months.

Regression results with three months lag and four months
lag are reported in table 4-8 and.4-9 respectively, Examining
equation 2 in table 4-9, although there is no similar lagged
pattern as intangible suggestions as I expected, the
regression result is much better than equation 2 in table
4-8. Coefficients for the four-month-lag model all become
"correct" with.negative signs, and the coefficient for t-1, t-
3, and t-4 are statistically significant at .05, .01, and .005
level respectively. R? rises from .009 (equation 2 in table
4-8) to .050 (equation 2 in table 4-9). It suggests that a
four-month-lag model has much stronger explanatory power than
a three-month-lag'model, Regression analysis in equation.1 in

table 4-9 shows the similar results that R2 (.086) for four-

155

month-lag model is higher than R2 (.065) for three-month—lag
model in table 4-8. The coefficient of tangible suggestions
with a lag of four months in table 4-9 is statistically
significant at .005 level too. It suggests that tangible
suggestions pay off at t-4. Therefore, if this lagged
variable (t-4) is not included in the regression estimation,
it may obscure the effect of tangible suggestions. In
addition, interestingly enough, comparing equation 3 in table
4-8 and 4-9, there was no change for R2 and even worse for F
value when.I include one more lagged intangible suggestion (t-
4). It suggests that intangible suggestions with a lag of
four months seem to have no effect on labor inputs.

The empirical results from table 4-9 suggest that
tangible suggestions may have no pmsitive effect on labor
inputs until a lag of four months. Based on this test, I will
re-estimate regressions with four months lag for both tangible
and intangible suggestions. New regression estimates are
presented in table 4-10.

Table 4-10 presents initial exploratory findings with
total labor hours as the dependent variable. The major goal of
these regression analyses is to examine 'to what extent
improvement suggestions (both tangible and intangible) reduce
hours worked. However, "total labor hours" is not a typical
dependent variable, so the analyses were exploratory.

Equation 1 in table 4-10 shows that the current and past

volume of adopted tangible suggestions have negative, but

1£56

Table 4-10: Results of Regression of Analyses of the Effect of Suggestions and All
Other Variables on Labor Hours (A Pour-Month-Lag Model for Suggestion Variables)

Explanatory Labor Hours
Variables (1) (2) (3)
Intercept b 6325.100”' 6262.300"' 6281.700"'
t 65.833 89.132 74.003
Tang, b -1.410 -3.660
B - 0004 -.001
t - 183 -.495
Tangb- b -4.978 -11.316
B -.001 -.001
t -.527 -1.264
Tangp, b 6.341 4.448
B .002 .001
t .653 .485
Tang,_3 b -8.412 -9.191
B -.002 -.002
t -.894 -1.022
TangM b -37.oas"' -34.621“'
B -.010 -.009
t -4.833 -4.655
Intang, b 3.523'“ 3.502'"
B .013 .013
t 4.337 4.684
Intangb- b -.879 -.818
B -.003 -.003
t -1.003 -.999
Intangpz b - 161 .052
B -.001 .0002
t -.177 .060
Intang,.3 b -2.084' -2.215"
B -.008 -.008
t -2.230 —2.526
IntangH b 3.004"‘ 2.761"'
B .011 .010
t 3.432 3.315
Safety b 76.387"‘ 71.466"' 76.767“‘
B .020 .019 .020
t 7.781 7.265 8.159
Quality b 90.279"' 113.980"' 115.550"'
8 .008 010 .010
t 3.791 4 805 4.948
Absent b -23.836"' -20.635" -17.020'
B -.008 -.007 -.006
t -2.974 -2.573 -2.295
R2 152 117 .144
F 11.367 13.633 17 430
F-test for all estimates significant at the .01 level.
* = p < .05 level ** = p < .01 level *** a p < .005 level

b is the regression coefficient.
B is the standardized Coefficient.

t is the t value.

157

delayed effects on the current labor inputs. However, the
association between tangible suggestions at t, t-l, t-2, t-3
and labor inputs is weak. Even the coefficient of tangible
suggestions at t-2 is positive. Regression analysis shows
that coefficient for a lag of four months is statistically
significant at .005 level. It indicates that each additional
tangible suggestion reduces labor inputs by approximately 37
hours after‘ a lag' of four' months. It reflects that 37
operational hours gained four months later once the suggested
changes are well established.

Intangible suggestions in equation 1, table 4-10, show a
different pattern from tangible suggestions. The
implementation of intangible suggestions may increase labor
inputs at current period t but it may reduce labor inputs at
t-l, t-2, and t-3. It suggests that intangible suggestion
making may be expensive in the short run but it pays off in
the long run” However, we should note that coefficients of t-
1, t—2, and t-3 are small, even though the coefficient with a
lag of three months (t-3) is statistically significant at .05
level.

There is a different over-time pattern between tangible
and intangible suggestions. The strongest response of labor
efficiency to tangible suggestions occurs at t-4, whereas the
strongest response of labor efficiency to intangible
suggestions occurs at t-3. It suggests that tangible

suggestions tend to have longer duration of lag effect than

158

intangible suggestions. In other words, tangible suggestions
may pay off later than intangible suggestions. There is also
a different magnitude effect between tangible and intangible

suggestions. The coefficient of tangible suggestions at t-4

in equation 1 is 37.085, while the coefficient of intangible
suggestions at t—3 is only 2.048. It suggests that each
additional tangible suggestion may reduce labor inputs by
about 37 hours after' a lag «of four 'months, while each
additional intangible suggestion only reduces 2 hour inputs
after a lag of three months. Tangible suggestions tend tijay
off much more than intangible suggestions.

In sum, the results from table 4-10 indicate that labor
efficiency is a function in part of current and past volumes
of adopted tangible suggestions. Tangible suggestions have a
bigger effect on labor efficiency than intangible suggestions
but they pay off a little bit slower than intangible
suggestions.

In equation 1, 2, and 3 in table 4-10, a higher
industrial accident rate, as indicated by a higher number in
Safety, leads to more labor hours. The association between
the number of accidents and labor hours is all statistically
significant at .005 level in three equations. The magnitudes
of the effects of accidents are also sizeable. For example,
the coefficient in equation 1, table 4-10, implies that if
industrial accidents was increased by one case, labor inputs

would increase by about 76 hours.

159

More product defects, as indicated by Quality, also
caused more labor hours. In equation 1, 2, and 3 in table 4-
10, the coefficients between Quality and labor hours are all
statistically significant at .005 level. For example, the
coefficient in equation 1, table 4-10, implies that each
additional product defect causes 90 hours more labor inputs.

The association between absenteeism and labor hours is
opposite to the prediction in all three cases. Higher
absentee rates are associated with fewer labor hours, not more
as expected, and coefficients are significant at .05, 1, and
5 percent level in equation 1, 2, and 3 respectively.

Regression coefficients in equation 1, 2, and 3 of table
4-10 show that more current and previous volume of adopted
tangible and intangible suggestions and fewer' number' of
industrial accidents and product defects are associated with

high labor efficiency.

Empirical Results from Quality Model

The number of product defects, as indicated by Quality,
used as dependent variable; the current (t) and.previous (t-1,
t-2, and t—3) volume of adopted tangible and intangible
suggestions addressing kaizen (or improvement) effect; and
training, safety, overtime, and absent as control variables.
The empirical results are presented in table 4—11.

The lagged.pattern.of joint effects for both tangible and

intangible suggestions reports in equation 1, while the over-

160

Table 4-11: Results of Regression Analyses of the Effect of Suggestions and All Other
Variables on Product Quality

Explanatory Quality
Variables (1) (2) (3)
Intercept b 101”' 103"' 099'"

t 6 328 7 355 6 199
Tang, b - 006' -.006"

B - 016 - 017

t -l 682 -2 381
TangM b 001 - 001

B 004 - 003

t 379 - 389
Tangb, b -.003 -.002

B - 008 -.007

t - 804 -.833
TangH b -.008' -.006'

B -.022 -.017

t -2.262 -2.306
Intang, b -.0007' -.0007'

B -.029 -.O29

t -2.079 -2.291
Intang,,1 b -.001"' -.001"'

B -.O45 -.046

t -2.919 -3.278
Intang,_2 b .001" .0009”'

B 041 .041

t 2.395 2.612
Intang,_J b .0002 .6E-04

B 008 .003

t .478 .156
Mgttran b .0003 .0002 .0003

B .012 .008 .007

t 1.006 .716 .631
Selftran b -.0007 -.0008 -.0006

B -.012 -.014 -.012

t -1.123 -1.391 -l.124
Techtran b -.0005' -.0005' -.0006'

B -.020 - 022 -.023

t -1.829 -2.236 -2.198
Safety b .039"° .035"' .038"'

B .112 .103 .110

t 7.608 8.355 7.943
Overtime b .8E-05 .SE-OS .4E-05

B .011 .007 .006

t .627 .443 .354
Absent b .004 .003 .004

B .014 .011 .017

t 1.061 .913 1.326
R2 119 .119 111
F 7 909 11.224 10 335

F-test for all estimates significant at the .01 level.
* = p < .05 level
** = p < .01 level
*** a p < .005 level
b is the regression coefficient.
B is the standardized Coefficient.
t is the t value.

161
time patterns of simple effect for tangible and intangible
suggestions are presented in equation 2 and 3 respectively.

The regression results in equation 2 of table 4-11,
tangible suggestions are found to have a negative effect on
product defects. Greater current and past volumes of adopted
tangible suggestions lead to fewer product defects. Tangible
suggestions exhibit lag functions which multipeak in the
current period and a lag of three months. It suggests that
the current quality improvement is partly the result of the
current improvement tangible suggestions and partly the result
of the previous (especially at t-3) improvement tangible
suggestions.

The regression in equation 3 in table 4-11, intangible
suggestions are also found.to have negative effects on quality
at t and t-l. The coefficients of intangible suggestions at
t and t-1 are statistically significant at .05 and .005 level
respectively. Initial suggestions at the initial month and a
lag of one month jointly negatively influence product defects,
which suggests that the greater number of intangible
suggestions at current and last month are associated with
fewer' product defects at current period. The regression
coefficients become wrong-signed for a lag of two and three
months. Therefore, the maximum lag for the effect of
intangible suggestions on quality may be only one month.
However, it may be difficult to interpret the intangible

suggestions at t-2 in which its coefficient is positive and

162
statistically significant at .005 level.

As mentioned above, the tangible suggestion effects in
equation 2, table 4-11 keep negative and significant till a
lag of three months, whereas the intangible suggestion effects
in equation.3 of table 4-11 keep negative and significant only
to a lag of one month. The results, thus, suggest that
intangible suggestions seem to pay off faster than intangible
suggestions. In the equation 1 reported in table 4-11 the
same pattern holds except tangible suggestions at t-l.

Intangible suggestions not only paid.off quicker but tend
to have greater magnitude effects and seem to be statistically
better than tangible suggestions. In equation 1, the sum of
standardized coefficients for intangible suggestions at t and
t-l is .074, while the sum of standardized coefficients for
tangible suggestions at t, t-2, and t-3 is only .046. Again,
comparing equation 2 and 3, the sum of standardized
coefficients for tangible suggestions (taking t and t-1 for
their negative signs) and.intangible suggestions (taking t, t-
1, t-2, and t-3 for their negative signs) is .075 and .044
respectively. The same results hold. Therefore, intangible
suggestions have greater effect on quality and pay off faster
than tangible suggestions.

More extensive technical training leads to fewer product
defects, as indicated by Quality. In equation 1 through 3,
the association between technical training and quality

improvement is all statistically significant at .05 level. On

163
the other hand, management and self-actualization training
have no effect on quality improvement.

A higher number of industrial accidents, as indicated by
Safety, is associated with more product defects. The
coefficients of Safety in equation 1 through 3 in table 4-11
are all statistically significant at .005 level. Product
quality tends to suffer from an unsafe environment.

More extensive use of overtime leads to more product
defects. However, the regression results in equation 1
through 3 provide little evidence to support this argument.
Further, a higher absent rate causes more product defects.
The evidence from regression analysis in table 4-11 is also
weak.

In sum, regression coefficients in equation 1, 2, and 3
of table 4-11 indicate that more current and previous adopted
intangible and tangible suggestions, more extensive technical
training and fewer number of industrial accidents are
associated with quality improvement. The pattern and effect
of intangible suggestions on quality improvement are different
from tangible suggestions. Intangible suggestions have
greater effect on.quality'improvement and.pay off quicker than

that of tangible suggestions.

Summapy of the Findings

Both tangible and intangible suggestions are good

predictors of productivity gains, high labor efficiency and

164

product quality improvement. lflmaempirical results from table
4-6, 4-10, and 4-11 show that the greater current and.previous
volumes of-adopted.tangible and.intangible suggestions lead to
higher productivity, lower labor inputs and fewer product
defects. However, there are different lagged patterns for
tangible and intangible suggestions. Consistently, tangible
suggestions have a longer duration of lag effect than
intangible suggestions in all three models. It implies that
intangible suggestions tend to be faster and easier to
transfer from. kaizen (improvement) knowledge to economic
benefits than of tangible suggestions. Further, there is also
a different effect between tangible and intangible
suggestions. Tangible suggestions have stronger effect on
productivity improvement and higher labor efficiency, whereas
intangible suggestions have stronger effect on quality
improvement.

More extensive use of training programs is associated
with high productivity and low product defects. However,
different training programs have different effects on
productivity gains and quality improvement. More extensive
use of management training or self-actualization training is
associated with productivity improvement and high labor
efficiency but not for quality improvement. In contrary, more
extensive use of technical training is associated with quality
improvement but not for productivity improvement and high

labor efficiency.

165

A higher number of industrial accidents, as indicated by
Safety, is associated with low labor efficiency and product
quality, but there is no relationship between the number of
industrial accidents and productivitqn .Additionally, a higher
number of product defects, as indicated. by Quality, is
associated with lower levels of productivity and low labor
efficiency. The regression results from table 4-6, 4-10, and
4-11 show that workforce size, absenteeism, overtime, and
technology seem to be poor predictors of these three dependent

variables of organizational economic performance.

Policy Implications

Increasing competitive pressures in the 19808 led many
companies to identify productivity and quality improvements
as major competitive objectives. Thus, continuous
productivity and quality improvements became important means
to obtaining a competitive advantage that could transfer to
high profitability. In this chapter, the empirical results
show that there is a cumulative effect of suggestions on
successive incremental improvements in productivity and
quality. Improvements in manufacturing process or better
process flow management which are derived from adopted
tangible and intangible suggestions increase the possibility
of incremental improvements in overall productivity and
quality.

My analysis suggests that it is important to distinguish

166
between.tangible and.intangible suggestions, because there are
distinct policy' implications between these two types of
suggestions. Table 4-12 depicts between tangible and
intangible suggestions along several dimensions. First,
tangible suggestions result in measurable financial benefits
such as labor savings, material savings, energy savings,
downtime reduction, output increase, etc. On the other hand,
most intangible improvements result in unmeasurable production
quality improvement, safety improvement, housekeeping
improvement or improvement of quality of working conditions.
Thus, tangible suggestions create quantitative or "tangible"

effects, while intangible suggestions lead to qualitative or

Table 4-12: The Results and Policy Implications for
Tangible and Intangible Suggestions

 

 

Tangible Intangible
Suggestion Suggestion
Nature of the - Quantitative - Qualitative
Outcomes - Results-oriented - Process-oriented
Pattern of the - Long-term gains - Short-term gains
Results
Magnitude of - Medium- to large-scale - Small- tx: medium-
Effects continuous improvement scale continuous
improvement
Driven by - Efficiency-demand - Quality-demand
change change
Aims/Impact - Cost-reduction - Quality-
change enhancement

change

167
"intangible" effects. To extend this line of thinking,
tangible suggestions can be characterized as result-oriented
whereas intangible suggestions can be characterized as
process-oriented.

Based on the regression results, productivity, labor
efficiency and quality improvement not only depend on present
volume of suggestions accepted but depend on past volume of
suggestions adopted. The analysis suggests that a lagged
effect exists between suggestion implementation and economic
gains. However, there are different patterns of delayed
effects over time between tangible and intangible suggestions.
Overall, tangible suggestions have:a longer lag structure than
intangible suggestions. In other words, productivity and
quality responses to tangible suggestions are slower than to
intangible suggestions. Thus, tangible suggestions may lead
to long-term (up to four months) gains for organizations,
whereas intangible suggestions may cause short-term (up toione
or two months) gains for organizations. It implies that
organizations can encourage employees to generate and then
adopt more tangible suggestions as a long-term strategy for
enhancing organizational effectiveness, while more extensive
adoption of intangible suggestions can.be used.as a short-term
strategy for improving organizational effectiveness.

The empirical results also indicate that tangible
suggestions generally have greater effect on economic benefits

than intangible suggestions. However, there are different

168
effects on a variety of dependent variables between tangible
and intangible suggestions. Tangible suggestions have a
larger effect on productivity gains and high labor efficiency
and.a smaller effect on quality improvement. On the contrary,
intangible suggestions have a greater effect on quality
improvement and a smaller effect on productivity and labor
efficiency improvement. It implies that if organizations
attempt to gain competitive advantage by pursuing a strategy
of "quality enhancement," intangible suggestions should not be
ignored. More extensive adoption of intangible suggestions
may be an intelligent tactic for these companies. Firms with
a "quality-enhancement" strategy should establish a more open
and.process-oriented suggestion system to encourage employees
to make more intangible suggestions. A traditional suggestion
system. that only" emphasizes tangible suggestions may' be
incompatible with a "quality-enhancement" strategy. On the
other hand, if firms attempt to gain competitive advantage by
pursuing' a strategy' of "cost reduction," more extensive
adoption of tangible suggestions may be an effective

tactic“. Based on the above reasoning, a simple conclusion

 

17We should. note that it does not mean that intangible
suggestions have no effect on productivity improvement. As
mentioned earlier in table 3-1 of chapter 3 and in table 4-1
in this chapter, present and.past volume of adopted intangible
suggestion are highly related to current volume of adopted
tangible suggestion; and the effect of intangible suggestions
on productivity may be mediated through their effects on
tangible suggestions. Thus, intangible suggestions may be not
as important as tangible suggestions for a firm pursuing a
strategy’ of cost reduction, but they' cannot be totally
ignored.

169

of "suggestion—change" connection can be drawn here. The need
for tangible suggestions may be driven by organizational
"efficiency-demand change" and those tangible suggestions thus
will result in "cost-reduction change." On the other hand,
the need for intangible suggestions may be the result of an
organizational "quality-demand change" and the results from
those intangible suggestions thus will lead to "quality-
enhancement change."

In the kaizen-suggestion system, employees' suggestions
are built on. past knowledge and. practice that lead to
organizational continuous improvement. IRegression results
suggest that the accumulation of tangible and intangible
suggestion making cause improved productivity and quality.
Individual suggestion making is a learning process and a
process of knowledge accumulation. Learning and knowledge
accumulation by individual workers contributes to
organizational learning. Cutcher-Gershenfeld et al. (1994)
have argued. that kaizen. activities involve employees to
collect and interpret data in suggestion programs, which is a
planned learning process. Florida and Jenkins (1993) also
have indicated that suggestion systems in Japanese companies
allow employees to utilize their creativity on solving
problems that foster individual learning. The kaizen-
suggestion.program at NDUS has provided.maximum opportunities
for individual and organizational learning and in turn

achieved better organizational performance. It implies that

170
firms can obtain continuous productivity gains and quality
improvement by creating a learning organization for improved
technology, manufacturing process, better managerial process
or better product quality.

As reviewed earlier in this chapter, the use of a kaizen-
suggestion. progranl is grounded in the theory’ of worker
participation. The literature on suggestion programs, QCs and
financial involvement strongly' argues for) their' positive
effects on organizational effectiveness. This chapter
supports these arguments and it can also be viewed as making
a valuable contribution to those literatures. Allowing
employees to be involved in day-to-day kaizen activities
enriches their knowledge and skills of the manufacturing
process and production technologies that might lead to

improved organizational effectiveness.

Summing Up

In the first part of the chapter a comprehensive
literature review was presented, even so it still told us
little about kaizen program performance effects. A major
difficulty for a literature review about kaizen suggestion
systems is that there is no core literature, either
theoretical or empirical. It pushed me to be more
comprehensive than usual in relevant areas such as employee
involvement or QCs. Therefore a comprehensive literature

review in traditional suggestion programs, QCs and employee

171
involvement was presented in the first part of this chapter.
The concept of a kaizen-suggestion system is not only
important as an implication of continuous improvement itself
but also has links to the literature on productivity, quality
and employee involvement. It is an important phenomenon, in
part, because of the way it integrates across all three areas.

Increasing competitive pressures in the 19808 led many
companies to identify productivity and quality improvements
as major' competitive objectives. 'Thus, continuous
productivity and quality improvements became important means
to obtaining a competitive advantage that could transfer to
high profitability. In this chapter, the empirical results
show that there is a cumulative effect of suggestions on
successive incremental improvements in productivity and
quality. Improvements in the manufacturing process or better
management system.which are derived from adopted tangible and
intangible suggestions increase the possibility of incremental
improvements in overall productivity and quality.

My analysis suggests that it is important to distinguish
between tangible and. intangible suggestions, because the
effects of tangible and intangible suggestions on
organizational effectiveness are different. ‘Tangible
suggestions generally create quantitative effects whereas
intangible suggestions usually create qualitative effects.
Quantitative change, which is derived from tangible

suggestions, and qualitative change, which is derived from

172
intangible suggestions, have different policy implications to
organizations. If organizations attempt to gain competitive
advantage by pursuing a strategy of "quality enhancement,"
intangible suggestions should not be ignored. More extensive
adoption of intangible suggestions may be an intelligent
tactic for these companies. On the other hand, if firms
attempt to gain competitive advantage by pursuing a strategy
of "cost reduction," more extensive adoption of tangible

suggestions may be an effective tactic.

CHAPTER FIVE

CONCLUSIONS AND SUGGESTIONS FOR RESEARCH

Banker (1993) has argued that "what is important for
sustained competitive advantage is productivity gain that is
sustained from period to period, and not just a transitory
improvement in performance in.a particular period that cannot
be replicated.in subsequent periods." (P. 27). This study has
highlighted the continuous improvement with what Banker termed
"sustained competitive advantage. " The Japanese term "kaizen"
means continuous improvement that occurs gradually and as a
continuous incremental process. This implies small and
incremental change. Although the improvements are subtle in
the short run, sustained over time, they are considerable.

This thesis has examined the determinants of an effective
kaizen-suggestion system and its impact on productivity and
quality. Because no other studies have evaluated the
organizational determinants and impact of kaizen-suggestion
systems, this paper serves as an exploratory effort to examine
the causes and effects of a kaizen-oriented suggestion system.

In testing the hypothesized determinants and effects of
the kaizen-suggestion system, using Nippondenso, U.S. as the
case study, empirical examinations were presented in chapter

3 and 4 respectively. In this final chapter of the paper,

173

174
three issues will be addressed. First, the major findings
will be summarized and the implications of the research will
be elaborated. Second, the limitations of the study will be
examined. Third, suggestions for future research will be

made.

Major Findings and Implications

Kaizen.has been viewed.as the key to Japanese competitive
success (Imai 1986; Yasuda 1991; Japan Human Relations
Association 1992; Japan Human Relations Association 1988).
Imai (1986) also-argues that improved productivity'and.quality
are two major outcomes of kaizen activity. Whenever and
wherever improvements are made in business, these improvements
are eventually going to lead to improvements in such areas as
quality and productivity. This study attempts to answer a
number of questions about these arguments. If kaizen is so
important to business competition, what factors determine
these kaizen activities? If kaizen is a learning process
(Florida and Jenkins, 1993; Cutcher-Gershenfeld et al. 1994),
are current tangible suggestions associated with past
accumulative intangible suggestions? If yes, what is the
lagged effect between adopted tangible suggestions and
intangible suggestions? What is the effect of management on
suggestion making? Furthermore, if kaizen is important to
business success, how does it work? Does a greater volume of

suggestions lead to improved productivity and quality? If

175

kaizen suggestions can improve organizational productivity and
quality, are they achieved immediately or at some later time?
If there is a lag relationship between suggestion
implementation and productivity improvement, what will the lag
structure look like? Is there a different pattern of the
effects of tangible and intangible suggestions on productivity
or quality?

To answer these questions, ijroposed.and tested.a cross-
sectional and.time-series model of the determinants and.impact
of the kaizen-suggestion program. It is the very first
attempt to employ this statistical technique and time lag
model in.employee suggestion research” 'The major findings and
their implications are summarized as follows.

The results of this thesis provide initial support for
the proposition that the current (t) and past (t-l, t-2).
adopted intangible suggestions are strongly related to the
current adopted tangible suggestions. It implies that
suggestion making is a learning process. What workers have
learned in making intangible suggestions eventually will be
transformed to the knowledge needed for generating a new
tangible suggestion. An intangible suggestion itself may be
a very subtle organizational improvement, but it implies that
it is very crucial and fundamental for the accumulation of
knowledge needed for making tangible suggestions. Using a
baseball analogy, a tiny intangible suggestion can be viewed

as a single whereas a big tangible suggestion can be seen as

176

a home run. However, accumulation of numerous tiny singles
may be more valuable than a home run. A person can hit a home
run that may be accumulated from previous experience in
hitting numerous singles. To encourage constant incremental
process improvements, those firms with a kaizen-oriented
suggestion system solicit and reward all suggestions, home
runs as well as singles. From the perspective of knowledge
accumulation, a company with a traditional suggestion system
may learn something from a firm with a kaizen-oriented
suggestion. system. Further, a. "healthier“ culture for
suggestion making should be created to accumulate knowledge
and encourage continuous improvement, At Nippondenso, U.S.
Inc., for example, no matter what suggestions (tangible or
intangible) employees make, the company pays serious attention
to them equally. Employees are encouraged to make as many
intangible suggestions as they can. Thus, employees do not
feel embarrassed if they only make a small tiny suggestion.
If an organization does realize the importance of intangible
suggestions, continuous organizational improvement becomes
feasible and possible.

Regression analyses provide evidence of statistically
significant associations between.both tangible and intangible
suggestions, and management training. The more management
training team leaders or departmental supervisors attend, the
more adopted tangible and intangible suggestions in these

departments. When team leaders or departmental supervisors

177

learn.more interpersonal or leadership skills, a more open and
supportive culture can be created. Leaders’ personal skills
and.an open culture are important for creative idea generation
and the idea’s Operationalization. This illustrates the fact
that management support in the process of suggestion.making is
pivotal for an effective suggestion system. To solicit more
kaizen ideas, management training should be linked to the
kaizen-suggestion system at the strategic level. Management
training courses should be designed to support kaizen
activities and processes at the practical level. Leaders
should be trained to become resources for their members in a
suggestion system.

However, we should note that since the effect of
intangible suggestions on tangible suggestions in the Chapter
3 is not for all lagged months, these findings are suggestive
but not definitive. The R? in the Chapter 3 regressions are
also small, so we should interpret the results carefully.
Again, the data on top management leadership style is based on
a relatively limited measure, so it must be treated with more
caution.

The empirical results suggest that the effect of
intangible suggestions on productivity may be mediated through
their effect on tangible suggestions. It implies that the
processes of the impact of suggestions on productivity may
include two stages: (1) more adopted intangible suggestions

stimulating more tangible suggestion making; (2) and in turn,

178
more adopted tangible suggestions leading to productivity
improvement. Tangible suggestion therefore can be viewed as
an intervening variable between intangible suggestion and
productivity’gain. This finding was based.on simple regression
analyses rather than more complex mediation tests, so the
finding may be suggestive but not definitive.

The results of this study also provide initial support of
the proposition that tangible suggestions tend to have strong
effect on productivity till a lag of three months, and its
duration.may be longer than that. Intangible suggestions only
have greater effect on productivity for a lag of two months
and that this effect obviously weakens at t-3. Thus, tangible
suggestions have a longer duration of lag effect than
intangible suggestions. It implies that tangible suggestions
pay off slower and longer than intangible suggestions.
.Additionally, regression. analyses show' that tangible
suggestions tend to have greater effects on productivity than
intangible suggestions. It implies that tangible suggestions
pay off more than intangible suggestions. In sum, tangible
suggestions pay off slower, but more than intangible
suggestions.

A four-month-lag model of labor inputs is statistically
much better than a three-month-lag model. The empirical
results from the four-month-lag model show that the strongest
response of labor efficiency to tangible suggestions occurs at

a lag of four months, whereas the strongest response of labor

179

efficiency to intangible suggestions occurs at a lag of three
months. It implies that tangible suggestions pay off later
than intangible suggestions. Although tangible suggestions
pay off slower than intangible suggestions, they tend to pay
off much more than intangible suggestions. The regression
results imply that each additional tangible suggestion may
reduce labor inputs by approximately 37 hours after a lag of
four months, while each additional intangible suggestion only
reduces about 2 hour inputs after a lag of three months.

The empirical results provide evidence that a greater
current and past volume of adopted tangible and intangible
suggestions lead.to fewer product defects. lHowever, there are
different patterns and effects between tangible and intangible
suggestions. The strongest response in quality improvement to
intangible suggestions and to tangible suggestions occurs at
a lag of two months and four months respectively. It implies
that intangible suggestions pay off quicker than tangible
suggestions. Further, the regression results show that
intangible suggestions not only pay off faster but have
greater effect (x1 quality improvement than tangible
suggestions. Intangible suggestions have greater estimated
effect size and are statistically better than tangible
suggestions. It implies that intangible suggestions also pay
off more than tangible suggestions.

By integrating three models, the empirical results show

that there is accumulative effect of suggestions on successive

180
incremental improvements in productivity, labor efficiency and
quality. Improvements in technology and manufacturing
process, which are derived from adopted tangible and
intangible suggestions, lead to incremental improvements in
overall productivity and quality. Improvements in
productivity and quality not only depend on present volume of

suggestions adopted but depend on previous volume of

suggestions adopted. A lagged effect exists between
suggestion implementation and economic gains. However, a

different pattern of delayed effects over time between
tangible and intangible suggestions is found. Overall,
tangible suggestions have a longer lag structure than
intangible suggestions. It implies that economic benefits
(i.e. improved productivity, labor efficiency and quality)
respond to tangible suggestions later than to intangible
suggestions. 'Thus, more extensive adoption. of tangible
suggestions can serve as a long-term strategy for enhancing
organizational effectiveness, while more extensive adoption of
intangible suggestions can be used as a short-term strategy
for enhancing organizational effectiveness. .Additionally,
there are different effects on these three dependent variables
in relation to tangible and intangible suggestions. Tangible
suggestions have a bigger effect on productivity gains and
high labor efficiency but have a smaller effect on quality
improvement. By contrast, intangible suggestions have a

larger effect on.quality improvement but have a smaller effect

181

on both productivity and labor efficiency improvements. It
implies that if firms attempt to«gain competitive advantage by
pursuing a "quality-enhancement" strategy, more extensive
adoption of intangible suggestions is an intelligent tactic.
Thus, a more open and process-oriented suggestion system
should be established to link to a "quality-enhancement"
strategy. A more open, process—oriented or kaizen-oriented
suggestion system not only emphasizes tangible suggestions but
also rewards intangible suggestions. A traditional suggestion
system that only' emphasizes tangible suggestions may' be
incompatible with a quality-enhancement strategy. On the
other hand, if organizations attempt to gain competitive
advantage by pursuing a strategy of "cost reduction," more
extensive adoption of tangible suggestions is an effective
tactic. Considering the above reasoning, we realize the
suggestion-change connection. The need for tangible
suggestions may be driven.by organizational efficiency-demand
change and those tangible suggestions thus may result in cost-
reduction change. On the other hand, the need for intangible
suggestions may be due to organizational quality-demand change
and the results of those intangible suggestions thus may lead
to quality-enhancement change.

To support a kaizen-oriented suggestion system, a
learning organization should be created to reinforce
individual learning and foster knowledge accumulation.

Employees’ suggestions are built on past knowledge and

182

practice that lead to continuous organizational improvement.
The exposure of employees to a learning culture makes
knowledge accumulation and individual learning easier. The
greater the individual learning and knowledge accumulation,
the greater number of kaizen suggestions an organization will
have. A learning organization fertilizes a suggestion program
with a more open, creative and supportive climate. A more
open, creative and supportive culture is required for an
effective suggestion program.

The empirical results also provide initial evidence that
more extensive use of training leads to improved productivity,
labor efficiency and product quality. However, there are
different effects among these three types of training
programs. More extensive use of management and self-
actualization training is associated with high productivity
and labor efficiency, while more extensive use of technical
training is associated with high.product quality. It implies
that if an organization attempts to gain advantage by pursuing
a "cost-reduction" strategy, a ‘management or self-
actualization training program may be more effective than a
technical training program to attain its aim. On the other
hand, if an organization. attempts to gain advantage by
pursuing a "quality-enhancement" strategy, a technical
training program may be more effective than management or

self-actualization training for its goal achievement.

183
Research Limitations

This thesis suffers from. some limitations. A. major
limitation.of this dissertation is lack of previous studies in
"kaizen" or "suggestion systems". Because very few studies
have evaluated either the determinants or the outcomes of an
effective suggestion system, this dissertation suffers from
insufficient theoretical foundation.

Second, its focus on a single site case restricts its
generalizability. The suggestion program at NDUS is a kaizen-
oriented system. The company practices kaizen by encouraging
constant incremental process improvement. Thus, NDUS solicits
and rewards all suggestions, tangible and intangible. On the
contrary, companies with traditional suggestion systems
routinely look for the home run suggestions, the one-time
dramatic events. The nature and process of a kaizen-oriented
suggestion system are quite different from a traditional
suggestion system. It may be difficult to apply the results
obtained from a kaizen-oriented suggestion system to a
traditional one. A single-site and uni-system design may
restrict its generalizability.

Third, because:of the economic and longitudinal nature of
this study, only organizational and economic variables have
been considered. Individual behavioral variables were
excluded. I have argued that suggestions may affect
productivity by altering (1) work efforts or job satisfaction,

(2) the productive skills of the labor force, and (3) the

184
organizational efficiency. Without behavioral variables in
the empirical models, high unexplained variance is expected.
Consequently, since this study considers only organizational
and.economic variables, and other individual factors have been
excluded, the findings may be limited in their

generalizability.

Suggestions for Future Research

This study' suggests future research in five areas.
First, suggestion variables used in this thesis were measured
by the volume of adopted tangible and intangible suggestions.
They reflect quantity of suggestions rather than quality of
suggestions. It‘would.be interesting to examine the effect of
qualitysuggestions - reflected in suggestion points or cost
savings.18

Second, some technology-related. variables should be
considered in the future research. Roy Roemen, superintendent

in.production.at.NDUS, has been interviewed and indicated that

 

18In NDUS, a suggestion.point index is total points integrated
from points awarded from both tangible and intangible
suggestions. Some criteria will be applied in the evaluation
of suggestion such as savings, originality, creativity and so
on. A cost savings index is total cost savings derived from
the sum of total tangible suggestions. Unfortunately, because
existing data for points of tangible and.intangible suggestion
were integrated in one, a simple effect of tangible or
intangible suggestions cannot be estimated. It makes the
comparison between the quality of tangible and intangible
suggestions impossible. Therefore, the effect of quality of
suggestions was not examined in this study. A. separate
estimation of points (quality) of tangible and intangible
suggestions thus is suggested.

185

many suggestions (both.tangible and.intangible) have been made
and accepted in his department shortly after the company
adopted new machines or manufacturing processes. It was
simply because employees tried to solve new problems and made
their jobs easier. This suggests an interaction. between
suggestions and technology--what the Japanese call "people
giving wisdom to the machines". It implies that a reciprocal
relationship between technology and suggestions exists. It may
be interesting to examine: whether advanced technology leads
to more improvement suggestions, or more suggestions lead to
improved technology, or they interact reciprocally. This will
be an interesting topic for further study.

Third, some individual-related variables also should be
considered in future research. Sue Flees who is HRM specialist
and is in charge of the kaizen-suggestion system also has
suggested that some individual variables might be expected to
be associated with the performance of suggestion making at
NDUS. These variables may include personality, motivation,
work experiences and superior-subordinate relationship. This
suggests the possibility of determinants of individual
difference on suggestion making. However, we should note that
individual factors (e.g. personality, motivation, etc.) cannot
be measured and estimated in a time-series study. It may be
only appropriate in a cross-sectional study.

If more technology-related and individual variables were

available, it might be possible to do a broader study of these

186
indicators that would have greater generalizability.

The results from chapter three provide initial support
for the proposition that the current (t) and.past (t-1, and t-
2) adopted intangible suggestions are strongly associated with
the current adopted.tangible suggestions. However, the initial
evidence provided in this chapter should be interpreted
carefully. For example, the empirical results from chapter
three suggest that the significant finding in the same month
can be interpreted as a reflection of the fact that groups
high on intangible suggestions will also be high on tangible
suggestions. The additional significance of the lag in the
same regression suggested a second phenomena that intangible
suggestions predict tangible suggestions. An interesting
question is raised here. Will the same results hold when
tangible and intangible suggestions are reversed in a
regression equation? That is, using intangible suggestions
predicts tangible suggestions. If yes, it implies that a
reciprocal relationship between tangible and intangible
suggestions may exist. Therefore, further examination of the
tangible-intangible connection is needed.

In Chapter 3, I only utilized lagged intangible
suggestions to predict tangible suggestions. We can't
distinguish whether there is a two-way relationship, a one-way
causal relationship or a common antecedent driving the
emergence of both. In other words, we cannot realize how past

adopted tangible suggestions influence current tangible

187

suggestion making. We also cannot tell how lagged tangible
suggestions interact with lagged intangible suggestions to
influence current tangible suggestion.making, Therefore, a new
model should be conducted to test this more complicated set of
associations between tangible and intangible suggestions. We
may have lagged tangible suggestions, lagged intangible
suggestions and an interaction (i.e., lagged tangible
suggestions X lagged intangible suggestions) all utilized to
predict tangible suggestions in the new model.

Finally, a multi-site and bi-system can be used in
studies conducted on the determinants and effects of
suggestion programs. For example, a traditional suggestion
system can be used as a control group. Its determinants and
effects then can be compared with that of a kaizen-oriented
suggestion system. Mixed studies incorporating two
perspectives would yield more comprehensive organizational

policy recommendations.

In Closing

This dissertation opened with Imai’s argument that "
kaizen has been viewed as the key to Japanese competitive
success." In this thesis I have tried to explain if and how
it works. Clearly, kaizen suggestions through people lead to
improved productivity, labor efficiency and quality without

any major or extra capital investments. It challenges

188
conventional economic thinking.

This dissertation.has raised.fundamental questions in the
first chapter about the relationship between kaizen and
innovation. A8 I argued earlier, kaizen is not innovation.
kaizen does not replace innovation. Rather, kaizen and
innovation should be "complementary." Imai (1986) has
indicated "kaizen improves the status quo by bringing added
value to it. It is bound to yield positive results if results
are continued toward a clearly defined goal. ------ As soon as
kaizen’s marginal value starts declining, one should turn to
the challenge of innovation. Top management’s job is to
maintain a balance between kaizen and innovation, and it
should.never forget to look for innovative opportunities." (pp
228-229).

This dissertation has linked the concept of a kaizen
suggestion system to the literature on productivity, quality
and employee involvement. It is an important phenomena, in
part, because of the way it integrates across all three areas.
I also have proposed the conceptual or process models of the
determinants and the outcomes of a kaizen suggestion system.
They can serve as theoretical foundation for further study.

By examining fundamental assumptions about the lagged
effect of kaizen suggestions in both determinant and outcome
levels, the phenomena of knowledge accumulation and
organizational learning in the research site were found. A

kaizen-suggestion system helps accumulate knowledge and.makes

189
organizational learning easier.

A. kaizen—suggestion. systenl solicits and. rewards all
suggestions, home runs as well as singles. Its strategy
strives to give undivided attention to both tangible
(quantitative) and intangible (qualitative) effects, to both
process and result, to both long-term.and short-ternlgains, to

both cost-reduction and quality-enhancement change.

APPENDICES

1530

Appendix A: The Selection Process of the Optimal Pattern of Lagged
Intangible Suggestion Variables with Tangible Suggestion
as Dependent Variable.

Explanatory Tangible Suggestion
Variables (1) (2) (3) (4) (5) (6)
Intercept b .157'“ .131'" .129'" .121'" .108'" .105'"
t 5.505 3.883 3.744 3.521 3.174 3.061
Intang, b .018"' .016“' .015”' .015"' .015“' .014"'
B .257 .227 .223 .216 .211 .211
t 7.326 6.087 5.892 5.654 5.493 5.502
Intang,_l b .005' .003 .002 .002 .002 .002
B .078 .048 .039 .033 .025 .023
t 2.189 1.275 1.005 .839 .622 .567
Intang,_2 b .007"' .006' .005' .005' .004
B .095 .087 .075 .679 .065
t 2.536 2.242 1.873 1.691 1.609
IntangH b .002 .001 -.0001 -.0003
B .028 .016 -.002 -.004
t .739 .408 -.060 -.107
IntangH b .003 .002 .002
B .050 .031 .026
t 1.316 .797 .645
Intang,_5 b .005' .005'
B .078 .072
t 2.039 1.838
Intang“, b .002
B .022
t .581

p < .05 level

p < .01 level

p < .005 level

b is the regression coefficient.
B is the standardized Coefficient.
t is the t value.

fit
it.

1£91

Appendix B: The Selection Process of the Optimal Pattern of Legged Training Variables with
Tangible Suggestion as Dependent Variables.

Explanatory Tangible Suggestion
Variables (1) (2) (3)
Intercept b .112"' .081' .124'"
t 2.602 1.947 2.894
Intangt b .015"' .015"' .016"'
B .221 .219 .227
t 5.897 5.849 6.063
Intang.1 b .003 .003 .003
B .045 .219 .227
t 1.195 1.176 1.209
Intangbz b .006" .007" .007"
B .092 .095 .096
t 2.453 2.545 2.539
Mgttran, b .003‘
B .042
t 1.747
Selftrant b -.003
B -.019
t -.786
Techtran, b .0003
B .005
t .213
Mgttran,_1 b .003‘
B .042
t 1.749
Selftran,_1 b .004
B .026
t 1.047
TechtranH b .001
B .007
t .288
Mgttran,_2 b -.7E-04
B -.0009
t -.041
Selftran,.2 b 002
B 011
t 445
TechtranH b 0004
B 005
t 215

* = p < .05 level
** = p < .01 level
*** = p < .005 level

b is the regression coefficient.
B is the standardized coefficient.
t is the t value.

192

Appendix C: The Selection Process of the Optimal Pattern of Lagged Training Variables with
Intangible Suggestion as Dependent Variable.

Explanatory Intangible Suggestion
Variables (1) (2) (3)
Intercept b 5.921"' 4.937"’ 6.361"'
t 8.339 7.032 8.953
Mgttran, b .003
B .003
t .172
Selftran, b -.0004
B -.0002
t -.011
Techtran, b .002
B .002
t .133
MgttranM b 063"'
B .057
t 3.369
Selftran,.1 b .051
B .023
t 1.383
TechtranH b .022
B .002
t .136
MgttranH b -.021
B - 019
t -1.077
Selftranb2 b -.041
B -.018
t —1.108
TechtranH b .011
B 011
t .689

* = p < .05 181181
** a p < .01 level
*** = p < .005 level

b is the regression coefficient.
8 is the standardized Coefficient.
t is the t value.

193

Appendix D: Hypotheses of Control Variables for Productivity
and Labor Efficiency Models (Chapter Four)

Hypothesis 1.1.

The more the number of product defects, the lower the
level of productivity.

Hypothesis 1.2.

The more the number of product defects, the more the
operational time.

Hypothesis 2.1.

The more the number of accidents, the lower the level 6
productivity.

Hypothesis 2.2.

The more the number of accidents, the higher the level
of labor inputs.

Hypothesis 3.1.

The higher the ratio of absent, the lower the level of
productivity.

Hypothesis 3.2.

The higher the ratio of absent, the higher the level of
labor inputs.

Hypothesis 4.
A positive relationship exists between management,
self-actualization and technical training and
productivity.

Hypothesis 5.

The level of productivity in machining departments will
be higher than that of assembly departments.

Hypothesis 6.

The larger the size of the department, the lower the
level of productivity.

194

Appendix E: Hypotheses of Control Variables for QualityiModel
(Chapter Four)

Hypothesis 1.1.

A negative relationship exists between the number of
training hours and the number of customer claims for
product defects.

Hypothesis 1.2.
Technical training have greater effect on product
quality improvement than that of management and self-
actualization training.

Hypothesis 2.

The higher the ratio of absent, the more the number
of customer claims for product defects.

Hypothesis 3.

A positive relationship exists between the number of
the accidents and the number of customer claims for
product defects.

Hypothesis 4.
The more the number of overtime hours, the more the
number of customer claims for product defects.

HH.: 00.: 0H.: «0.- 00.: m0.: mH.: H«.: mH.: H0. 00. 00. m0. no. no. 00.
«0. 0H. 00. «0. HH. HH. «0. 00. «H. 0«. 0H. 0H. 0«. ««- n0.:
ah. nH. nH. vH. Hm. 00. 00. 0«. n0. n0. «0. «0. «0. «0.:
m0. «0. H«. mm. 0H. «0. 0m. 00. 00. 00. >0. «0. «0.:
0v. vH. m0. «0.: 0H. H0.: 00. 00. M0. 00. 00. «0.:
00. 00. n0.: 0H. 00. m0.- v0.: 00.: «0.: H0.: H0.
0H. 00. ««. 0H. «H. HH. m0. «0. 0H. H0.:
0«. mm. 0H. MH. vH. «H. mH. mH. 00.
FH. m0. v0. 00. «0. «0.: «0. «0.
«n. «0. «0. «0. >0. «0. H0.:
mH. vH. 0H. 0H. «H. 0H.
hm. «v. #0. av. v«.
«n. «0. vv. mH.
hm. av. «H.
on. mH.
vH.
:-0« mH 0H 0H 0H mH VH nH «H HH 0H m m h o m
.............................. 6.23:;

00.
«0.:
00.
H0.:
m0.
«0.
n0.
v0.
n0.:
00.
0H.
0H.
0«.
mH.
nH.
mH.

0«.

666 n z

66. 66. 66. 66. 65. 5666666668 .6«

66.- 66. 66. 66.6 66.6 66666 66.-ha .6«

66. 66. 66.- 56.«56 66.666 os6uuo>o .66

«6.- 66.- 66.- 66.6« 66.6« 6666 66666666: .66
manHsua

66. 66.- «6.- 66.56 «6.66 666686668 .56

ms6sHeuH d06u

66.- 66.- 66.- «5.5 «5.6 -666666664-6666 .66
manHsuH

66. 66.- 66. 66.66 66.«6 666866666: .66

56. 66. 66. 6«.6 66. 56.666 .66

«6.- 66.- «6.- «6. 66. 6666696 .66

66. 66.- 66.- 6.6666 6.6666 .6566 666666: .«6

66. 66. 56. 66.66 66.«66 >66>6uuneoua .66
66-6. 66666-6696

66. 66. 66. 66.56 65.«6 6666666666.66
26-6. 6666666696

66. 66. 66. 66.56 65.«6 6666666666 .6
6«.». 6666666656

6«. 66. 66. 66.56 65.«6 6666666666 .6
.6.». 6666666696

66. 6«. 66. 66.56 65.«6 o6n6666666 .5
A». nOHuaemmsm

66. 66. 6«. 66.56 «5.«6 6666666666 .6
.6-60 6666666696

6«. 66. 66. 6«.6 66. o6n66868 .6
.6.». 66666-6666

6«. 6«. 66. 6«.6 66. 66666668 .6
16.». 6666666696

66. 6«. 6«.6 66. 66666668 .6
.6.». 6666666666

6«. 6«.6 66. 66666668 .«
.u. QOHueemmsm

6«.6 66. 66666668 .6

6 « 6 .n.6 use:

.6800 savanna.
eeuseeex «UGOHOHuum one eeede>6uoeuuu HsnOHDeeHasmuo one .eeHnsHMs> Houunou .eso6ueemmsm sestuem eoHnenoHusHem ”h RHoseoo<

REFERENCES

LIST OF REFERENCES

Arbose, J. R. (1980). Quality Control Circles: The West
adopts a Japanese concept. International Management,
December, 31-39.

Abdel-Halim, A.A. (1983). Effects of task and personality
characteristics on subordinate response tijarticipative
decision making, Academy of Management Journal, 26(3),
477-484.

Abdel-Halim, A.A., and Rowland, K. M. (1976) . Some personality
determinants of the effects of participation: A further
investigation, Personnel Psychology, 29, 41-55.

Atkinson, P. E. (1990). Creating culture change: The key to
successful total guality management. San Diego:
Pfeiffer Company.

Banker, R. D. (1993). Accounting for productivity gains. In

Ijiri, Y. (ed.) Creative and innovative approaches to
the science of management. Connecticut: Quorum Books.

Binstock, S. L. (1981). Americans express dissatisfaction

with quality of U.S. goods. Quality Progress, January,
p.13.

Bittel, L. R. (1968). What evepy supervisor should know. New
York: McGraw-Hill.

Burke, M. J., Hoffman, R. G., Hazer, J. T., and Hall, T. L.
(1982). Predicting suggestion-making intentions and

behaviors. Managerial Psychology. 3(2), 1-15.

Buller, P. F., and Bell, C. H. Jr. (1986). Effects of team
building and goal setting on productivity: A field

experiment. Academy of Management Journal, 29(2), 305-
328.

Bragg, J. E., and Andrews, I. R. (1973). Participative
decision making: An experimental study in a hospital.

Journal of Applied Behavioral Science, 9(6), 727-735.

Blanner, R. (1964). Alienation and freedom. Chicago:
University of Chicago Press.

Brief, A.P., and Aldag, R. J. (1975). Employee reactions to
job characteristics: A constructive replication,

Journal of Applied Psychology, 60, 182-186.

Brief, A.P., and Motowidlo, S. J. (1986). Prosocial

196

197

organizational behavior, Academy of Management Review,
11, 710-725.

Cummings, T. G., and Worley, C. G. (1993). Organization
development and change, New York: West Publishing Co.

Cable, J., and Wilson, N. (1990). Profit-sharing and
productivity: Some further evidence, The Economic
Journal, 100, 550-555.

Cooke, W. N.(1990). Labor-management cooperation. Michigan:
W.E. Upjohn for Employment Research.

Chinloy, P. (1981). Labor productivity. Cambridge, Mass: Abt
Book.

Cutcher-Gershenfeld, J.(1991). The impact on economic
performance of a transformation in workplace relations.
Industrial and Labor Relations Review, 44(2), 241-260.

Cutcher-Gershenfeld, J., Nitta, M., Barrett, 8., Belhedi,N.,
Coutchie, C., Inaba, T., Ishino, 1., Lee, S., Lin W-J,
Monthersell, S., Mulder, J., Rabine, S., Ramanand, S.,
Strolle, M., and Wheaton, A., (1994). Employee

involvement and kaizen: understanding contingous

improvement through the harnessing of knowledge,
Unpublished manuscript, School of Labor and Industrial

Relations, Michigan State University.

Crocker, O. L., and Charney, C. (1984). Quality Circles: A

guide to participation and productivity. New York:
AMACOM.

Cummins, R. C., and King, D. C. (1973). The interaction of
group size and task structure in an industrial
organization. Personnel Psychology, 26, 87-94.

Conte, M., Tannebaum, A. S., and McCulloch, D. (1981).
Employee ownership. Michigan: The Institute for Social
Research, The University of Michigan.

Campbell, J. P., Campbell, R, J. et al. (1988). Productivity

in organizations: New perspectives from industrial and
organizational psychology. San Francisco: Jossey-Bass

Publishers.

Doherty, E. M., Nord, W. R., and McAdam, J. L (1989),
Gainsharing and organization development: A productive

synergy. The Journal of Applied Behavioral Science,
25(3), 209-229.

Denz, F. A. (1946). Why a suggestion plan? in American

198

Management Association (Ed.), Getting and using
employees’ ideas. New York: AMA, 3-6.

Dyer, C. (1987). Cannon production system: Creative
involvement of the total workforce. Cambridge, Mass:
Productivity Inc..

Daft, R., and Becker, S. (1978). The innovative

organization: Innovation adoption in school
organizations. New York: Elsevier.

Denton, D. K. (1991). Horizontal Management: Beyond total
customer satisfaction. New York: Lexington Books.

Donovan, M., and Van Horn, B. (1980). Quality Circle program
evaluation. Transactions of the Second Annual IAQC
Conference, 5, 463-470.

Daltn, D. R., Todor, W. D., Spendolini, M. J., Fielding, G.
J., and Poter, L. W. (1980). Organizational structure
and performance: A critical review. Academy of Management
Review, 5, 49-64.

Drucker, P. R. (1954). The Practice of Management, New York:
Harper & Row.

Dielman, T. E. (1989). Pooled cross-sectional and time-series
data analysis. New York: Marcel Dekker, Inc.

Ekvall, G. (1976). Creativity of the place of work. Stockholm:
The Swedish Council for Personnel Administration.

Even, W., and Black, G. (1967). Innovation in business

organization: Some factors associated with success or
failure, Journal of Business, 40, 519-530.

French, J. R. P., Israel, J., and As, D. A. (1960). An
experiment in a Norwegian factory: Interpersonal
dimensions in decision Making. Human Relations, 13, 3-19.

French, J. R. P., Ross, I. C., Kirby, S., Nelson, J. R., and
Smyth, P. (1958). Employee participation in a program of
industrial change, Personnel, 35(3), 16-29.

Frost, C., Wakeley, L., and Ruh, R. (1974). The Scanlon Plan

for organizapion development: Identity, participapiop,
and egpality. East Lansing: Michigan State University.

Fairweather, G.W., Sanders, D. H., and Tornatzky, L. Z.

(1974). Creating change in mental health organizations.
New York: Pergamon Press, Inc.

199

FitzRoy, F. R., and Kraft, K. (1987). Cooperation,
productivity, and profit sharing, Quarterly Journal of
Economics, 23-35.

Fourastie, J. (1957). Productivity. prices. and wages. Paris:
Organization for Economic Cooperation and Development.

Florida, R., and Jenkins, D. (1993). Advanced production
systems and learning in the Japanese transplants: A
Research Prospectus. Paper Presented in the Conference
of Workshop on Apprenticeship Education in Japan.
University of Pittsburgh. April 22-25, 1993.

French, W. (1984). The Personnel management process - Human
resources administration. Atlanta: Houghton.Mifflin Co.,

527.

French, E. G. (1958). The interaction of achievement
motivation and ability in problem solving success.

Journal of Abnormal and Social Psychology, 7, 306-309.

Gouldner, A.W. (1960). The norm of reciprocity: A.preliminary
statement, American Sociological Review, 25, 161-178.

Griliches, Z. (1967). Distributed lags: A survey.
Econometrica, 16-49.

Goodman, P. S. (1979). Assessing organizational change: The

rushton gpality of work of experiment. New York: John
Wiley & Sons.

Gunatilake, S. (1984). An.exploratory'study'of quality circles
and team building in two hospital settings, Dissertation
Abstracts International, 45, 1723B.

Graf, L.A. (1982). Suggestion program failure: Causes and
remedies, Personnel Journal, 450-454.

Guest, R.H. (1955). Men and.machines: An assembly-line worker
looks at his Job. Perspnnel, 31, 3-10.

Gunsch, D. (1991). Employees team up with HR, Personnel
Journal, October, 68-71.

Greenlaw, P. S. (1980) . Suggestion systems: An old approach to

a new problem. The Personnel Administrator, January, 49-
54.

Gregerman, I. B. (1984). Productivity improvement: A.gpide for
small business. New York: Van Nostrand Reinhold Company.

Gold, B. (1975). Technological change: Economics, management

200

and environment, Oxford: Pergamon.

Gold, B. (1979). Productivity, technology, and capital:
Economic analysis. managerial strategies. and.government

policies. Lexington: D.C. Health and Company.

Gold, B., Pierce, W., and Rosegger, G. (1970). Diffusion of
major technological innovations in U.S. iron and steel
manufacturing. Journal of Industrial Economics. July.

Greenberg, L. (1973). A practical guide to productivity
measurement. Washington, D.C.: The Bureau of National

Affairs.

Guzzo, R. A., Jette, R. D., and Katzell, R. Z. (1985). The
effects of psychologically based intervention programs on
worker productivity: A meta-analysis, Personnel
Psychology, 38, 275-291.

Graham-Moore, B., and Ross, T. L. (1983). Productivity
gainsharing. Englewood Cliff, NJ: Prentice-Hall.

Hatcher, L., Ross, T. L., and Collins, D. (1989). Prosocial
behavior, job Complexity, and suggestion contribution

under gainsharing plans. The Jourpal of Applied
Behavioral Science, 25, 231-248.

Herman, J. B. (1973). Are situational contingencies limiting
job attitude-job performance relationships?

Organizational Behavior and Human Performance, 10, 208-
224.

Hill, J. M., and Trist, E. L. (1953). A Consideration of
industrial accidents as a means of withdrawal from the
work situation. Human Relations, 6, 357-380.

Hackman, J.R. (1969a). Nature of the task as a determine of
job behavior. Personnel Psychology, 22, 435-444.

Hackman, J.R. (1969b). Toward understanding the role of tasks
in behavioral research. Acta Psychologica, 31, 97-128.

Hackman, J. R., and Lawler, E. E. (1971). Employee reactions

to job characteristics, Journal of Applied Psychology,
55, 259-286.

Hackman, J. R., and Suttle, J. L. (1977). Improving life at
work. Santa Monica, CA: Goodyear Publishing Co. Inc.

Hatcher, L., and Ross, T. L. (1991). From individual
incentives to an organization-wide gainsharing plan:
Effects on teamwork and product quality. Journal of

201

organizational Behavior, 12, 169-183.

Hatcher, L., Ross, T. L., and Collins, D. (1989). Prosocial
behavior, job complexity, and suggestion contribution

under gainsharing plans. The Journal of Applied
Behavioral Science, 25, 231-248.

Hammer, M., and ChamPY- J. (1993). Reenginerring the

corporation: A manifesto for business revolution, New
York: HarperCollins Publishers, Inc..

Hage, J. and Aiken, M. (1967). Social change in complex
organizations. New York: Random House.

Harrington, H. T. (1991). Business process improvement. New
York: McGraw-Hill, Inc.

Holter, H. (1965). Attitude toward employee participation in
company decision making processes. Human Relations, 18,
297-321.

House, R.J., and Mitchell, T. R.. (1974). Path-Goal theory of
leadership, Journal of Contemporapy Business, 3, 81-98.

Ingle, S. (1982). Quality Circles master guide: Increasing
productivity‘with.people power. New York: Prentice-Hill.

International Labor Office (1969). Measuring labor
productivity. Geneva: ILO.

Industry Week (1979). A.quality concept catches on worldwide.
Industpy Week, 16 April, 125.

Ishida, M. (1993). Managerial Control and skill formation in
Japanese auto companies. Paper presented in Asia Regional

Conference in IR.

Imai, M. (1986). Kaizen: the key to Japan’s competitive
success. Toronto: McGraw-Hill Publishing Company.

Japan Human Relations Association (1988). The idea book:

Improvement through total emplgyee involvemeng.
Cambridge, Mass: Productivity, Inc..

Japan Human Relations Association (1992). Kaizen Teian 1:

Developing systems for continuous improvement through
employee suggestions. Cambridge, Mass: Productivity

Press.

Judge, G. G., Griffiths, W. e., Hill, R. C., Lutkepohl, H, and

Lee, T. (1985). The theogy and practice of econometrics.
New york: John Wiley and Sons.

202

Jenkins, D. (1973). Job power: Blue and white collar
democracy. Garden City, New York: Doubleday.

Juran, J; M. (1978). Japanese and Western.Quality--A.contrast,

Quality Progress, 11, 12.

Juran, J. M. (1981a). Product quality--a prescription for the
West. Part 1: Training and improvement programs.

Management Review, June, 9-14.

Juran, J. M. (1981b). Product quality-—a Prescription for the
West. Part II: Upper-management leadership and employee
relations. Management Review, July, 57-61.

Jenkins, T. M., and Shimada, J. Y. (1983). Effects of Quality

Control Circles on worker performance: A field
experiment. Paper presented.at 43rd Annual Meeting of the

Academy of Management, New York.

Johnston, J. (1972). Ecgnometric Methods. New York: McGraw-
Hill.

Kmenta, J. (1971). Elements of Eggnomstrics. New York: The
Macmillan Company.

Katz, D., and Kahn, R. L. (1966). The social psychology of
organizations. New York: Wiley.

Katz, H., Kochan, T. A., and Gobeille, K. R. (1983).
Industrial relations performance, economic performance,
and QWL programs: An interplant analysis. Indpstrial and
Labor Relations Review, 37(1). 3-17.

Katz, H. C., Kochan, T. A., and Keefe, J. H. (1987),
Industrial relations and productivity in the U.S.

automobile industry. Brppkings Paper on Egonomig
Activity, 3, 685-715.

Katz, H. C., Kochan, T. A., and Weber, M. R. (1985)..Assessing
the effects of industrial relations systems and efforts
to improve the quality of working life on organizational

effectiveness, Academy of Management Journal, 28(3) , 509-
526.

Kochan, T. A., Katz, H. C., and Mower, N. R. (1984). Worker

participation and American Unions: Threat or opportunity?
Michigan: W.E. Upjohn Institute for Employment Research.

Kochan, T. A., Katz, H. C., and McKersie, R. B.(1986). The
transformation of American industrial relations. New
York: Basic Books, Inc.. -

203

Katz, R. (1982). The effect of group longevity on project
communication and performance. Administrative Science

Quarterly, 27, 81-104.

Klotz, V. (1988). Staff suggestion schemes, International
Labor Review, 27(3), 335-353.

Kossen, S. (1983). The human side of organizations. New York:
Harper & Row Publishers, p. 79.

Katzell, R. A., and Guzzo, R. A. (1983). Psychological
approaches to productivity improvement. Americap

Psychologists, 38, 468-472.

Katzell, R. A., Bienstock, P., and Faerstein, P. H. (1977). A

guide to worker productivity experiments in the Unissd
States, 1971-1975, New York: New York University Press.

Kopleman, R. E. (1986). Managing productivity in
organisations. New York: McGraw-Hill.

Kendrick, J. W. (1961). Productivity prends in the U.S..

Princeton, N. J.: Princeton University Press.

Kendrick, J. W. (1977) . Underspanding productivity. Baltimore,
MD: John Hopkins Press.

Kendrick, J. W. (1984). Improving company productivity:

Handbook with case studies in collaboration with Amer; csn
Productivity Center, The John Hopkins University Press,

Baltimore.

Kagono, T., Nonaka, I., Sakakibara, K., and Okumura, A.

(1985). Strapegic versus evolutionapy management; A Us-

Japan comparison of sprategy and structure, Amsterdam:
North-Holland.

Klein, K. J., and Hall, R. J. (1987). Correlates of employee
satisfaction with stock ownership: who likes an ESOP

most? Journal of Applied Psychology, 73(4), 630—638.

Kaufman, R. T. (1992). The effects of Improshare on
productivity, Industrial and Labor Relations Review,
45(2), 311-322.

Kruse, D. (1984). Employee ownership and employee attitudes:
Two case studies. Pennsylvania: Norwood.

Lin, W-J. (1993). The impact of a kaizen-suggestion system on

productivity: A pooled cross-sectional and time-series
analysis. Unpublished manuscript, School of Labor and

 

204
Industrial Relations, Michigan State University.

Louviere, V. (1973). Where ideas aren’t kept in camera,
Nation’s Business, June, 15.

Long, R. J. (1978). The effect of employee ownership on
organizational identification, employee job attitudes,
and organizational performance: A tentative framework and
empirical findings. Human Relations, 31(1), 29-48.

Long, R. J. (1980). Job attitudes and organizational
performance under employee ownership. Academy of

Management Review, 23(4), 726-737.

Locke, E. A. (1970). Job satisfaction and job performance: A

theoretical analysis. Organizapional Behavior and Human

Performance, 3, 484-500.

Leeds, R. (1963). Altruism and the norm of giving. Merrill-
Palmer angterly, 9, 229-240.

Lewin, K. (1951). Frontiers in gropp dypamics in field phsggy

in social science. Ed, by Cartwright D. New York: Harper,
188-231.

Lillrank, P., and Kano, N. (1989). Continuous improvemenp:
Quality cpntrol circlss in Japaness industrial. Ann

Arbor, University of Michigan: Center for Japanese
Studies.

Loesges, R. E. (1946). Building a suggestion program in
American.Management Association (Ed.), Getping and gsipg
employees’ ideas. New York: AMA, 15-19.

Lesieur, F. G., and Puckett, E. S. (1969). The Scanlon Plan
has proved itself, Harvard Business Review, September-
October, 109-118.

Lawler, E. E., and.Ledford, G. E. (1982). Productivity and the

quality'of work life, National Productivity Review, 1(1),
295-307.

Lawler, E. (1991). High.involvement.msnagement. San.Francisco:
Jossey-Bass.

Locke, E.A., and Feren,D. B., McCaleb, V. M., Shaw, K, N., and
Denny, A. T. (1980). The relative effectiveness of four
methods of motivating employee performance. In K. Duncan,
M. Gruneberg and.D. Wallis (Ed.), Changes in Working Life
(p. 363-388), Chichester, England: John Wiley.

Locke, E.A., and Schweiger, D. M. (1979). Participation in

205

decision making: One more look. In B.M. Staw (Ed.),

Research in Organization Behavior, (Vol. 1, p. 265-339),
Greenwich, Connecticut: JAI Press.

Maier, N.R.F. (1953). Principles of human relations
applications to management, New York: Wiley.

Matthes, K. (1992). What’s the big idea? - Empower employees
through suggestion. H.R. Focus, Oct., p. 17.

Miner, J.B. (1969). Personnel psychology. London: The
Macmillan Co., p. 271.

Mitchell, T.R., Smyser, C. M., and Weed, S. E. (1975). Locus
of control: Supervision and work satisfaction, Academy of

Management Journal, 18, 623-631.

Marmaduke, H. C. (1946). The Illinois central suggestion
system. In American Management Association (Ed. ) , Getting
and Using Employees’ Ideas. New York: American Management
Association, 23-25.

Miller, K. I., and Monge, P. R. (1986). Participation,
satisfaction, and productivity: A meta-analysis review.

Academy of Management Journsl, 29(4), 727-753.

Mohrman, S. A., and Novelli, L. Jr. (1985). Beyond
testimonials: Learning from a quality circles program,

Journal of Occupatignal Bshavior, 6, 93-110.

Morton, C. P. (1971). ’Drop Dead’ say some employees, but
suggestion boxes thrive, Napional Observer, June 18.

Miller, C. Z., and Schuler, M. H. (1987). Gain sharing plans:
A comparative analysis. Organizational Dypamics, 16(1),
44-67.

Main, J. (1980). The battle for quality begins, Forpune, 29
December, 28-33.

Mirvis, P. (1988). Organizational development, Part I: An
evolutionapy perspective. In Woodman, R., and Pasmore,

W. (eds.) Research in organizational change and
development (Vol.2). Greenwich, CT: JAI Press.

Mirvis, P. (1990). Organizational development, Part II: An
revolutignapy perspective. In Woodman, R., and Pasmore,
W. (eds.) Research in organizational change and
development (Vol.4). Greenwich, CT: JAI Press.

Mark, J. A. (1977). Concepts and measures of productivity. In
Bureau of Labor Statistics, The meaning and measurement

206

of productivity. Washington, D.C.: U.S. Department of
Labor.

Murray, T. J. (1981). The rise of the productivity manager,
Dun’s Review, 117, 64-69.

Marks, M. L., Mirvis, P. H., Hackett, E. T., and Grady, J. F.
(1986). Employee participation in a Quality Circle
program: Impact on quality of work life, productivity,

and absenteeism, Journal of Applied Psycholggy. 71(1),
61-69.

Norris, D. R., and Cox, J. F. (1987). Quality Circles
programs: Volunteering for participation. Journal of

Occupational Behavior, 8, 209-217.

Northrup, H. P., and Young, H. A. (1968). The cause of
industrial peace revisited. Ipdustrial and Labor

Relapions Review, October, 31-47.

Nelson, J. (1980). Quality Circles become contagious,
Industrial Week, 30, 34-40.

Nippondenso Manufacturing U.S.A. Inc. (1990). NDUS Spggestign
System Handbook. Battle Creek, Michigan.

National Association of Suggestion Systems (1975). Suggestion
Newsletter, 5.

Nurick, A. J. (1985). Participation in organization shange:
The TVA esperiment. New York: Praeger Publishers.

O’Dell, C. (1981). Gsinshsring; involvement, incentive, snd
productivity; New York: American Management Association.

Organ, D. (1988). Qrgsnizatipnal citizenship pehaviop: Ths
good sgldier sypdrome. Lexington, MA: Lexington Books.

Organ, D.W., and Konovsky, M. (1989). Cognitive versus
affective determinants of organizational citizenship

behavior, gournal of Applied Psychology, 74, 157-164.

Pateman, C. (1973). Participation and democratic theopy.
Cambridge: Cambridge University Press.

Pelz, D.C., and Andrews, F. M. (1976). Scientists in
organizations (Rev. Ed.), Ann Arbor, Michigan:
Institute for Social Research.

Pizam, A. (1974). Some correlates of innovation within
industrial suggestion system, Personnel Psychology, 27,
63-76.

207

Puffer, S.M. (1987). Prosocial.behavior, noncompliant.behavior
and work performance among commission sales people.

Journal of Applied Psychology, 72, 612—621.

Pike, J., and Barnes, R. (1994). TQM in action: A practical

approach to continuous performance improvement . New York:
Chapman & Hall. .

Pym, D. (1964). Is there a future for wage incentive systems?
British Journal of Industrial Relations, 2, 379-397.

Patchen, M. (1970). Participation, Achievement, and
Involvement on the job. New Jersey: Prentice-Hall.

Poole, M. (1986). Toward a new industrial democracy: Workers’
participation in indgsppy. Routedge Kegan Paul, London.

Poole, M. (1989). The prigins pf sconomic democracy: Profit
sharing and employes shareholding schemss. Routledge,

London.

Pindyck, R. S., and.Rubinfeld, D. L. (1976). Econgmsprig modsl
and econgmic foresssps. New York: McGraw-Hill.

Reuter, V; G. (1976). A.new*look.at suggestion system, gggppsl
of Syspems Mansgement, January, 6-15.

Reuter, V. G. (1977). Suggestion system--utilization,

evaluation, and implementation. California Mspagsmenp
Review, Spring, 78-79.

Rosenberg, R. D., and.Rosentein, E. (1980). Participation and

productivity: An empirical study, Indusprial and Labs;
Relations Rsvisw, 33(3).

Ross, T. L. (1969). The ascogntant’s role in partisipapive
decision.making.‘Unpublished Ph.D dissertation, Michigan

State University.

Ross, T. L., and Jones , G. M. (1972). An approach to
increased productivity: The Scanlon Plan, Financial
Executives, 40, 23-29.

Ross, T. L. (1983). Why Pg fails in some firms, In Graham-

Moore, B., and Ross, T. L. (Eds.), Prodgctivity
gainsharing. Englewood Cliff, NJ: Prentice-Hall.

Runyon, K. (1973). Some interactions between personality
variables and management styles, Journal of Applied
Psychology, 57, 288-295.

Sadler, J. (1970). Leader style, confidence in management and

208

job satisfaction, Journal of Applied Behavioral Sgience,
6, 3-19.

Schuler, R.S. (1976). Participation with supervisor and sub-
ordinate authoritarianism: A Path-Goal theory reconcile-

ation, Administrative Science Quarterly, 21, 320-325.

Smith, G. (1986). Profit sharing and employee share ownership
in Britain, Employment Gazette, September, 380-385.

Smith, A.Z. (1968). Harnessing the power of ideas, Personnsl
Journal, 770-775, 779.

Smith, C.A., Organ, D. W., and Near, J. P. (1983).
Organizational citizenship behavior: Its nature and

antecedents, gourpsl pf Applied Psychology, 68, 653-663.

Smith, C.G. (1970). Age of R & D groups: A reconsideration,
Human Relations, 23, 81-96.

Smith, P.C. (1955). The prediction of individual differences
in susceptibility to industrial monotony, Journsl of
Applied Psychology, 39, 322-329.

Smith, A. E. (1968). Harnessing the power of ideas. Psrspnnsl
Journsl, November, 770-775.

Staw, B.M. and Boettger, R. D. (1990). Task revision: A
neglected form of work performance. Acadsmy of Management
Journal, 33, 534-559.

Slocum, J. W. (1970). Performance and satisfaction: An

analysis. Industrial_gelations- 9. 431—436.
Sayrs, L. W. (1989). Ppolsg pime sspies analysis. Newbury

Park, California: SAGE Publication Inc.

Stagner, R., Rlebbe, D. R., and Wood, E. F. (1952). Working on
the railroad: A study of job satisfaction. Personnel
Psychology, 5, 293-306.

Srinivason, C. (1983). Influence of quality circles on
productivity, group behavior, and interpersonal behavior:
An exploratory micro-organizational development

perspective. Dissertation Abstragts Internatignal, 43,
3055A.

Steel, R. P., and Shane, G. S. (1986). Evaluation research on
quality circles: Technical and analytical implications,
Human Relations, 39(5), 449-468.

Sutermister, R. A. (1969). People and Productivity. New York:

209
Mcgraw-Hill.

Siegel, I. H. (1980). Company productivity: Measurement for
improvement, Michigan: The W. E. Upjohn Institute for

Employment Research.

Stein, H. (1971). The meaning of productivity. In Bureau of

Labor Statistics, The meaning and measuremenp of
productivity. Washington, D.C.: U.S. Department of

Labor.

Saint, A. (1974). Lesrning at work: Hgman resourses snd
organizational development . Chicago: Nelson-hall company.

Schuster, M. (1984). The Scanlon Plan: A longitudinal

analysis, Journsl of Applied Bshavioral Sciense, 20(1),

23-28.

Schuster, M. (1983). The impact of union-management
cooperation on productivity and employment. Indusprial
and Labo; Relations Rsview, 33(3), 355-367.

Solow, R. M. (1957). Technical change and the aggregate

production function. RevieW'of Eggnpmigs spd Spspispiss,
39(3), 312- 320.

Scanlon, J. N. (1948). Profit sharing under collective

bargaining: Three case studies, Indpstpisl and Lspp;
Rslspipps Rsyisw, October, 58.

Shultz, G. P. (1958). Worker participation on production
problem. In Lesieur, F. G. (Ed.), Tps Ssanlop Plsp: A

fronpier in labo; mansgemsnt cpopsgspigp, Cambridge:
Technology Press of MIT.

Steingerg, R. A. (1981). Partisipspive lesdspship as s

determinant f i a u es '0 ner ti n- An

szperimenpal simplatign. Michigan State University,
Unpublished Ph.D Dissertation.

Thompson, P. C. (1982) . ua i Ci le : How m ke h m w rk
in Amsrisa. New York: AMACOM.

Tosi, H.A. (1970). Reexamination of personality as a
determinant of the effects of participation, Pegsonnel

Psychology, 13, 125-139.

Tortorich, R., Thompson, P., Orfan, C., Layfield, D., Dreyfus,
C., and.Kelly, M. (1981). Measuringworganizational impact

of quality circles, The Quality gircles Journal, 4(4),
24-34.

210

Taylor, T. D., Friedman, D. J., and Couture, D. (1987).
Operating without supervisors: A experiment,

Organizational Dypamics, 15, 26-38.

Trist, E. L., Susman, G. I., and Brown, G. R. (1977). An
experiment in autonomous working in an American
underground coal mine, Human Relations, 30(3), 201-236.

Umstot, D. D., Bell, C. H., and.Mitchel, T. R. (1976). Effects
of job enrichment and task goals and satisfaction and
productivity: Implications for job design. Jouppal of

Applied Psycholggy, 61(4), 379-394.

Voos, P. B. (1987). Managerial perceptions of the economic
impact of labor relations programs. Industrial and Labo;
Relations Rsview, 40(2), 195-208.

Vroom, V.H., and Mann, F. (1960). Leader authoritarianism and
employee attitudes, Personnel Psychology, 13, 125-139.

Vroom, V. H. (1962). Ego-involvement, job satisfaction.and job
performance. Psrsopnel Psychplogy, 15, 159-177.

Vroom, V. H. (1964). Work spd Mgtivation. New York: John
Wiley & Sons.

Vroom, V. H. (1960). Soms Eersgpality Determinapts pf she
Ef c s f ' ' ' . Englewood Cliffs, New Jersey:

Prentice-Hall.

Wilce, H. (1971). Getting value from suggestion schemes,

Inpepnspional Mspsgspspp, 26(12), 31-37.

Wolfe, P. J. (1985). Quality Circle intervention: Structure,
process, results. is ' n ract e a ional,
45, 2184A.

White, J. (1979). The Scanlon Plan: Causes and correlates of
success. Acsdsmy pf Msnsgemsnt Jogrnsl, 22, 292-312.

Womack, J., Jones, D. T., Ross, D. (1990). The machine that
ghanged ths wgrld . New York: Macmillan Publishing
Company.

Wadhwani, S., And Wall, M. (1990), The effects of profit-
sharing on employment, wages, stock returns and
productivity: evidence from UK micro-data, The Economic
Journal, 100, 1-17. ‘

Walker, C.R. (1950). The problem of the repetitiveness job.
Harvard Business Review, 28, 54-58.

211

Work in America (1973). Report of a special task force to the

Secretary of Health, Education and Welfare, Cambridge,
Mass: MIT press.

Wayne, S. J., Griffin, R. W., and Bateman, T. S. (1986).
Improving the effectiveness of quality circles, Personnel
Administrator, May, 79-88.

Whitewell, W. C. (1965). Suggestion systems are profitable.
Connecticut: Prentice-Hill Inc.

Wagner, J. A. III, and Gooding, R. Z. (1987a). Effects of
societal trends on participation research. Administrapive

Science Quarterly, 32, 241-262.

Wagner, J. A. III, and Gooding, R. Z. (1987b). Shared
influence and organizational behavior: A meta-analysis of
situational variables expected to moderate participation-

outcome relationships, Academy of Managemenp dpurpsl, 30,
524-541.

Wagner, J. A. III, and Hollecbeck J. R. (1992). Management of
organizapional behavior. Englewood Cliffs, NJ: Prentice
Hall, Inc. Chapter 4.

Yasuda, Y. (1991). 40 years, 20 million ideas--§hs Toypps
Suggespion System, Cambridge, Mass: Productivity Press,

Inc..

Yager, E. (1979). The Quality Circle explosion, Tpsipipg_spd
Develppmsnp dourpal, 35, 98-105.

Zupanov, J., and Tannenbaum, A. S. (1968). The distribution
of control in some Yugoslav industrial organizations as
perceived by members. In Tannenbaum A. S. (Ed.) QQDLLQI
in Organizations. 91-109. New York: McGraw-Hill.

"I1111111111E5